The Journal of Solid Waste Technology and Management

THE JOURNAL OF
SOLID WASTE TECHNOLOGY

AND MANAGEMENT
Formerly The Journal of Resource Management and Technology (Volumes 12-22)
Formerly NCRR Bulletin (Volumes 1-11)
May 2016
Volume 42
Number 2
THE JOURNAL OF SOLID WASTE TECHNOLOGY AND MANAGEMENT

ISSN: 1088-1697
FOUNDER:
Iraj Zandi
University of Pennsylvania
U.S.A.
ASSOCIATE
EDITORS
Haluk Akgun
Department of Geological Engineering
Middle East Technical University
Ankara 06531, Turkey
Email: hakgun@metu.edu.tr
Ofira Ayalon
Samuel Neaman Institute, Technion, Haifa;
Natural Resources & Environmental Research
Center, Faculty of Management
University of Haifa
Haifa, 3498838 , Israel
Email: aofira@gmail.com
Cristina Braga
Universidade Federal do Paran
Setor de Tecnologia
Departamento de Engenharia Qumica
Centro Politcnico Cx.P. 19011
Curitiba - PR - 81531-990, Brasil
EDITOR:
Ronald L. Mersky
Widener University
U.S.A.
SENIOR ASSOCIATE EDITOR:

Wen K. Shieh
University of Pennsylvania
U.S.A.
Paul Phillips
School of Environmental Science
University College Northampton
Boughton Green Road
Northampton, NN2 7AL, U.K.
Richard Marsh
Cardiff University
U.K.
Email: paul.phillips@northampton.ac.uk
Adam Read
Knowledge Leader
Waste & Resources Management
AEA
The Gemini Building, Harwell IBC
Didcot Oxon OX11 0QR, U.K.
Email: Adam.Read@uk.aeat.com
David Smith
The Regional Municipality of Niagara (retired)
Canada
Email: dgsmith1917@hotmail.com
EDITORIAL BOARD
Email: crisbraga@avalon.sul.com.br
Shoou-Yuh Chang
Department of Civil Engineering
North Carolina A&T State University
Greensboro, NC 27411, U.S.A.
Email: chang@ncat.edu
Mervat El-Hoz
University of Balamand
P.O.Box 100, Tripoli, Lebanon
Email: mervat.elhoz@balamand.edu.lb
Noah I. Galil
Department Civil Engineering
TechnionIsrael Institute of Technology
Haifa 32000, Israel
Email: galilno@tx.technion.ac.il
Ilona Srvri Horvth

Swedish Centre for Resource Recovery
University of Bors
SE-501 90 Bors, Sweden
Email: ilona.horvath@hb.se
Chukwu Onu
Southern University
Southern Branch Post Office
Baton Rouge, LA 70813, U.S.A.
Email: onu@engr.subr.edu
Magdy Abdelrahman
North Dakota State University
U.S.A.
Amimul Ahsan
Universiti Putra Malaysia (UPM)
Malaysia
Steve Bloomer
University of Teesside
U.K.
Sarvesh Chandra
Indian Institute of Technology Kanpur
India
Franco Medici
University of Rome La Sapienza
Italy
Yusuf Mehta
Rowan University
U.S.A.
Ilan Nissim
Israel Ministry of Energy and Water Resources
Israel
Terry Tudor
University College Northampton
U.K.
N.C. Vasuki
Delaware Solid Waste Authority (retired)
U.S.A.
Ming-Yen Wey
National Chung Hsing University
Republic of China
Keith P Williams
Cardiff University
U.K.
Anita Zvodsk
Barry University
U.S.A.
Jess Everett
Rowan University
U.S.A
Patrick Hettiaratchi
University of Calgary
Canada
Isam Janajreh
Masdar Institute
Abu Dhabi
Gennaro J. Maffia
Manhattan College
U.S.A.
ABSTRACTED/INDEXED IN: CAS database (Chemical Abstracts), Engineering Abstracts (Compendex), Environmental Abstracts, Environmental
Periodicals
Pollution
Abstracts,
All-Russian Institute
of Scientific
and Technical
(VINITI, REFERATIVNYI
ZHURNAL
), rests upon the
The
JournalBibliography,
of Solid Waste
Technology
and Management,
is published
by Widener
UniversityInformation
School of Engineering.
The responsibility
for contents
SCOPUS,
scholar,
EBSCO, SCImago,
GreenFILE
authors
andGoogle
not upon
the University.
This journal
is available by subscription and may be purchased at the rates posted at ingentaconnect.com/content/jswt/jswt. Editorial and subscription address is: Department of Civil Engineering, Widener University, One University Place, Chester, PA 19013-5792, U.S.A.; Telephone (610)
499-4042; Fax (610) 499-4461. Email: solid.waste@widener.edu. Web site: solid-waste.org. Copyright 2016 by Widener University. Printed in U.S.A.
THE JOURNAL OF
AND MANAGEMENT
May 2016
Volume 42
71
Number 2
HYDROGEOCHEMISTRY AND STABLE ISOTOPES OF A SOLID WASTE DISPOSAL AREA

FROM GUALEGUAYCH,ENTRE ROS, ARGENTINA
Romina Sanci, Hctor Osvaldo Panarello, Fernanda Cravero

87
WASTE-TO-ENERGY CHARACTERIZATION AND MODELING: A CASE STUDY AT THE US

NAVAL ACADEMY
P.A. Caton, J. Schmidt, C. Adsit, E. Bermudez, C. Chase, T. Kerner

98
COMPOSTING OF RUMEN CONTENT WASTE USING ANAEROBIC-ANOXIC-OXIC (A2/O)

SYSTEM
Rhenny Ratnawati, Yulinah Trihadiningrum, Sri Rachmania Juliastuti

108
OPPORTUNITY OF BIOCHEMICAL PROCESS FOR PHOSPHOGYPSUM UTILIZATION
Ye Chernish, L. Plyatsuk
116
EVALUATION AND QUANTIFICATION OF HOUSEHOLD WASTES SENT FOR RECYCLING

THROUGH A SELECTIVE WASTE COLLECTION SYSTEM
Vanessa A. Mantovani, Sandro D. Mancini, Thalita R. Barros, Jos L. Ferraz

128
IMPERIAL SMELTING FURNACE SLAG AS FINE AGGREGATE IN CEMENT CONCRETE

MIXES
Dr. S.B Patil, Dr. A.K. Vyas, Dr. A.B. Gupta, Rashmi S. Patil
THE JOURNAL OF
AND MANAGEMENT
May 2016
Volume 42
137
Number 2
USE OF RECYCLED AGGREGATE CONCRETE IN PAVEMENT CONSTRUCTION
Roz-Ud-Din Nassar, Parviz Soroushian

145
A MULTI-CRITERIA EVALUATION OF THE METHODS FOR RECYCLING SCRAP TIRES
Shoou-Yuh Chang, Frank Gronwald

157
ASSESSMENT ON NON-LINEAR MODELS FOR DEMONSTRATING HEAVY METALS

RELEASE BEHAVIOR FROM SOLID WASTE BLOCK
M. Aminul Haque, M. A. Hoque
HYDROGEOCHEMISTRY AND STABLE ISOTOPES OF A SOLID

WASTE DISPOSAL AREA FROM GUALEGUAYCH,
ENTRE ROS, ARGENTINA
Romina Sanci1*, Hctor Osvaldo Panarello*, Fernanda Cravero+
*Instituto de Geocronologa y Geologa Isotpica
Universidad Buenos Aires - Consejo Nacional de Investigaciones Cientficas y Tcnicas, Pabelln INGEIS,
Ciudad Universitaria, C1428EHA, Ciudad Autnoma de Buenos Aires, Argentina
+CETMIC (Centro de Tecnologa de Recursos Minerales y Cermica)
Camino Centenario y 506, Gonnet, La Plata, B1897ZCA, Argentina
Tel: +54 1147833021/3022/3023; Fax: +54 1147833024
rominasanci@hotmail.com; romina@ingeis.uba.ar
ABSTRACT
Major and trace elements, pH, electrical conductivity and organic matter were determined in
groundwater and subsoil samples collected from an urban solid waste final disposal site
(Gualeguaych city, Argentina) in order to quantify leachate impact on the quality of these resources. Geochemical modeling results showed mixing process of pristine water and leachate
in different proportions which promote, in turn, other processes in the free aquifer such as cation exchange, outgassing of carbon dioxide , precipitation and dissolution of minerals. Anomalous electrical conductivity values of subsoil samples confirmed the leachate impact too. Soilwater interaction allowed the development of natural attenuation processes in water: low dissolved organic carbon (DOC), iron and aluminum concentrations were associated with high
amounts in subsoil. High clay content (smectites) and high cation exchange capacity, organic
matter and metals analysis suggesting that the exchange and or sorption process were effective. 13C-DOC constant values ratified the DOC sorption in sediments rather than oxidation.
Other natural processes such evaporation and marine ingressions were detected through
geochemical models. 18O y 2H values suggested not only a meteoric origin for the groundwater but also evaporation and mix process of waters. 13C-DIC demonstrated that CO 2 outgassing could have happened during the mixing.
Keywords: leachate; contamination; environmental isotopes; geochemical processes
INTRODUCTION
The urban solid waste (USW) final disposal sites are characterized for presenting different environmental problems
related with pollutants that result from the physical, chemical
and biological processes that act upon the waste. Leachate

formation is one of the problems associated with these sites,
since water percolation through the deposited waste extracts
both organic and inorganic materials that dissolve and infiltrate into the porous media and could deteriorate soil and
_________________________________________
Corresponding author
HYDROGEOCHEMISTRY AND STABLE ISOTOPES OF A SOLID WASTE DISPOSAL AREA FROM GUALEGUAYCH
71
groundwater quality of the area and surrounding areas due to

migration. This leachate composition will depend on initial
humidity and waste composition, the amount of water that
infiltrates by rain or that could enter the landfill as groundwater, on cover material permeability, the reactions associated
with the degradation of the materials which form the cover
and age of USW (Bocanegra et al., 2001). Pollutants from
these leachates usually correspond to three great groups:
dissolved organic matter, inorganic elements (Ca2+, Mg2+,
Na+, K+, NH 4 +, Fe, Mn, Al, Cl-, SO 4 2- and HCO 3 -) and heavy
metals (Cd, Cr, Cu, Pb, Ni and Zn) (Christensen et al., 2001;
Kjeldsen et al., 2002). Once these compounds are incorporated into the subsoil, in places where there are no environmental practices as to how to collect and treat them, different
mobilization pathways are presented: compounds are either
trapped in the soil because they are either dissolved in the
aqueous phase or occupying exchange places, adsorbed on
inorganic constituents (clays or Fe, Mn and Al oxides), associated with organic matter and/or precipitated as pure or
mixed solids, or they can move to surface or ground waters
(Garca and Dorronsoro, 2005). The polluting nature of this
antropic source results in the emission of pollutants even after
its closure and for a time span that can involve years. Therefore, in the areas where this kind of activities takes place,
without control of contaminants, it is imperative to evaluate
the processes that influence the generation, transport and
destination of the pollutants. In order to explain and predict
changes in groundwater composition near and away from
contaminant source, the hydrogeochemical data sets contain
the information on governing geochemical processes (cation
exchange, proton buffering, mineral precipitation, degassing,
redox as degradation of DOC inside the plume). Modeling is
essential to deduce and quantify these processes in order to
demonstrate and understand these mechanisms. Also, it is not
only necessary to have hydrochemistry and mineralogical
data from a sector of free aquifer of Cura Stream Basin but
also information about the unsaturated zone affected by the
disposal of waste. It is known that most processes that cause
attenuation or remove pollutants from the subsoil are much
more important in the soil profile as a result of its huge content of organic matter and clays and its great bacterial population. For example, heavy metals (Hg, Pb, Zn, Ag, etc.) are not
always able to reach groundwater because processes like
adsorption on clay surfaces or on organic matter or the formation of insoluble complexes may fix or delay their movement (Ogundilan and Afolabi, 2008). The same effect is
achieved by the cation exchange capacity of the soil that
plays an important role in the retention of metal cations generated through natural or anthropic processes. Therefore, and
in order to be able to detect possible natural attenuation of
contamination , it is important to know the cation exchange
capacity of the soil in the study area (emphasizing the detection of organic matter and clays) and evaluate the precipitation and/or buffering processes (e.g. van Breukelen et al.,
2003; Berj et al., 2005). Also, the use of environmental isotopes in water (hydrogen, oxygen and carbon) in conjunction
with major ions contributes to detect contamination by leachate (e.g. North et al., 2004; Castaeda et al., 2012).
For this research, we first checked if there were mixing
72
processes between leachate and groundwater that could affect

the quality of this resource, quantifying its impact. In order to
do this, a geochemical model was elaborated with data from
major elements, temperature, pH, electrical conductivity,
dissolved organic carbon and some metals from the unconfined aquifer by means of the NetPath program (Plummer et
al., 1991; Plummer et al., 1994) which enabled to identify the
reactions that help to explain the physico-chemical characteristics observed to explain along the groundwater flow. These
same reactions were validated in turn, by field observations
and mineralogical data of subsoil samples and the use of
isotopes in groundwater. Besides, in order to evaluate the
impact of leachate in the unsaturated zone, subsoil samples
were taken from up to 2 m deep both in pristine and USW
affected areas taking into account that electrical conductivity
is an indicator of soil pollution. Thus, this parameter was
quantified as well as some of the elements that may contribute to generate it e.g. some cations -Ca, Mg, Na, K- and
heavy metals, together with cation exchange capacity, organic matter, carbonates and texture.
Area Descriptions
The study area is the Municipal USW final disposal site of
Gualeguaych, Entre Ros, Argentina. The zones filled with
USW cover approximately 34 ha and do not have subsoil gas
extraction systems or leachate collection systems. This site is
confined to the low area of El Cura Stream Basin, towards the
south of the cited city, in the vicinity of the confluence of El
Cura Stream and the Gualeguaych River (Figure 1). Two
geomorphological units are widely distributed in the basin:
the loessic plain (LP) and the alluvial plain (AP) formed
chiefly by two reddish-brown loessic levels (called Pampean
sediments) and silty to silty-clayey materials, respectively.
These two levels are recognized within LP as Hernandarias
Formation (Middle Pleistocene) and Tezanos Pinto Formation
(Upper Pleistocene), and they have large lithological and
hydrogeological similarities. Materials conforming these
formations are silts and silty clays of aeolic origin (loess)
with variable proportions of fine to very fine sand and different degrees of carbonate cementation and disseminated calcium carbonate concretions of approximately 2 cm long
(Iriondo, 1980). Regarding the mineralogy, both levels contain in their sand fraction mainly plagioclase, quartz, volcanic
glass, microcline and lithic fragments; in their pelitic fraction,
quartz, phyllosilicates, plagioclase, amorphous minerals and
in a smaller proportion, potassium feldspar; in their clay
fraction, in a major proportion, smectites, with micas and
illites (Boujon and Sanci, 2014). Between both loessic levels,
discontinuous deposits of greenish silt with a high content of
marine fossils were observed (Boujon et al., 2011; Boujon
and Sanci, 2014), in coincidence with the levels described by
Guida and Gonzlez (1984) referred as the Belgranense
marine ingression of the Late Pleistocene. The Hernandarias
Formation presents numerous interbedded fluvial sediments
in its interior, which confer it a marked heterogeneity, and
gypsum accumulations towards its base and calcareous concretions or discontinued levels of tosca towards its top
(Iriondo, 1996).
JOURNAL OF SOLID WASTE TECHNOLOGY AND MANAGEMENT
VOLUME 42, NO. 2
MAY 2016
FIGURE 1
The study area (urban solid wastes final disposal site) showing the sampling locations, the predominant direction of groundwater
regional (SW to NE) and local flow (S to N), and geomorphological units (loessic plain and alluvial plain). Right upper quadrant
indicates the geographic location and confluence of El Cura Stream and Gualeguaych River.
In El Cura stream basin, the Pampeano or Pampean
sediments are located below the soil cover on sites
topographically high and intermediate (between 50 m and 10
m contour lines) (Boujon and Sanci, 2014). It is a really
continuous and its thickness ranges between 10 m and 45 m.
Its hydrogeological importance is given because it behaves as
free aquifer and contains the water table. The regional flow
network obtained from water levels measured in the free
aquifer within the basin showed that groundwater moves
predominantly from SW to NE, in coincidence with the regional topographic slope of El Cura stream basin and the
increase of salinity (from about 90 mg/L to 859 mg/L ) in the
same direction (Boujon y Sanci, 2014). Water from the free
aquifer is for human and animal use and irrigation purposes.
Locally, the USW final disposal site has 10 piezometers
placed up and down stream (M9w, P10, P6, P7, P5, P11, P9,
P12, P22, TZ1) whose depths never exceeds 4 meters. The
local conceptual model for this area shows that groundwater,
collected by piezometers P (Figure 1), corresponds to shallower flow lines (S-N oriented) and less influenced by the
regional flow system (Boujon et al., 2011; Boujon, 2014).
The predominant soils within the LP geomorphological
unit were the Vertic Argiudolls, subgroup of soils classified
according to the Soil Taxonomy (Soil Survey Staff, 2014).
They are characterized by a powerful mollic epipedon and a
thick well-structured argillic horizon underneath, with these
typical horizons: A1 A2 BA Bt1 Bt2 BCk and Ck (Boujon
and Sanci, 2014). In general, A denomination relates to a
mineral horizon obscured by organic matter, Bt o Bk to

mineral horizon formed by accumulation of clays or calcium
carbonate respectively, and C to mineral horizon little
affected by pedogenic processes. On the other hand, in the
AP unit, soils are poorly developed; there is a wide predominance of acquic moisture regime soils, simple profiles and
evidences of repeated and continuous soil rejuvenation. Typic
Endoaquols are the predominant ones (A-AC-C horizons).
Methods
A physico-chemical analysis of the unconfined aquifer
was performed from 10 samples from piezometers located
inside the Municipal USW disposal ground (Figure 1). The
following parameters were determined: temperature, pH,
electrical conductivity (E.C.) carbonates (CO 3 2-), bicarbonates (HCO 3 -), sulfates (SO 42-), chlorides (Cl-), calcium
(Ca2+), magnesium (Mg2+), sodium (Na+), potassium (K+) and
dissolved organic carbon (DOC). pH, E.C. and temperature
were measured in situ and samples were taken and kept refrigerated until their analysis in the laboratory. CO 3 2- and
HCO 3 - determinations were carried out 24 hours after receiving the samples by titration with sulfuric acid (H 2 SO 4 ). Magnesium (Mg2+), calcium (Ca2+), sodium (Na+) and potassium
(K+) were determined by atomic absorption spectrometry
(Buck Scientific 200A), sulfate (SO 4 2-) by inductively coupled plasma atomic emission spectrometry (BAIRD-ICP
2070) and chloride (Cl-) by titration with silver nitrate
73
(AgNO 3 ). Dissolved organic carbon (DOC) was determined

by the SM (Standard Methods) 5310B analytical technique.
Samples for dissolved methane analysis were taken and analyzed following Lomond and Tong (2011) procedure . For
analysis of heavy metals in water, samples were filtered
through Whatman G / F of 0.45m pore size that were stored
in 100 ml sterilized flasks and acidified and sterilized
acidified with nitric acid to pH less than 2. The quantification
of metals was performed by atomic emission spectrometry
inductively coupled plasma (ICP).
Isotopic ratios of hydrogen (2H), oxygen (18O) and
carbon (13C) were measured in well pristine water (M9W),
the leachate (LX1), P samples and PP sample. During 4
months prior to the survey, monthly integrated samples were
collected in an ad hoc container placed at the landfill surveillance building in order to verify isotopic values (18O and 2H)
of study area (sample PP). Carbon inorganic dissolved (DIC)
from water was precipitated as BaCO 3 adding BaCl2 to 100
mL of sample previously adjusted to pH 11 (Valencio et al.,
2003). The precipitates underwent in vacuo heat treatment at
350C to remove organic matter interferences and then they
were directly introduced into the extraction system. Carbon
isotopic ratios 13C/12C were measured on CO 2 extracted from
carbonates with phosphoric acid 100% according to
McCreas (1950) method modified by Panarello et al. (1982).
The cryogenically purified CO 2 was contrasted with an internal reference (Carrara marble). Data was informed as isotopic
deviations ( ) relative to the international reference standard V-PDB, with an analytical uncertainty of 0.1. Isotopic analyses (2H and 18O) were made by laser spectroscopy
in a Los Gatos Research (LGR) Liquid Water Isotope Analyzer (OA-ICOS: Off-Axis Integrated Cavity Output Spectroscopy) at the Stable Isotopes Laboratory at the INGEIS.
The reference standard used is the Vienna Standard Mean
Ocean Water (V-SMOW). Isotopic results are expressed as ,
defined as = 1000(RS - RP)/ RP , where is the isotopic
deviation in ; S: sample; P: International reference; R:
isotopic ratio (2H/1H, 18O/16O, 13C/12C).
DOC samples were prepared based on the technique described in Gandhi et al. (2004). A sample volume containing
about 50 umol C was transferred to 50mL beakers (that were
previously decontaminated with 50% nitric acid) and placed
in an oven at ca.50C, till dryness. 1mL of ultra-pure water
was then added to the sample together with 85% phosphoric
acid (ca. 100-300 uL) and agitated to remove DIC. More
ultra-pure water was added to reach 2mL. 150uL of the concentrated sample were then transferred to tin capsules for
liquids in several aliquots. The samples were dried in an oven
(ca.50C) after each transfer. The closed capsules were then
measured in a Carlo Erba Elemental Analyzer 1180, coupled
to a continuous flow Thermo Scientific Delta V mass spectrometer through a Thermo ConFlo IV interface.
The hydrogeochemical data were subjected to
hydrogeochemical modeling. Inverse mass-balance models
were performed using NETPATH (Plummer et al., 1991). In
order to determine if the quality of groundwater has been
affected by leachate, ion Cl- was used in calculations as it is a
conservative ion. Piezometer M9w (pristine) and sample LX1
were taken as initial water and P6, P11, P9,P 12 and P22 as
74
final water due to their geographical location (downstream)

from USW and LX1. Since the plot does not have a leachate
collection system from which pure leachate could be extracted, a most possible representative sample, named LX1 was
taken from a small depression next to the waste heap from
where that effluent (very dark in color) could be captured
before its infiltration. In TZ1, due to its high chloride content
(higher than LX1), a typical marine ingression water was
added to the mixing (Plummer et al., 1994). The chemical
and sedimentary characteristics of the study area, with the
presence of discontinuous levels of marine ingressions, could
support this assumption. The program required for its operation that there is at least one phase for each of the incorporated ion (constraints). Therefore the minerals for modeling
were chosen from background data and identifications performed by X-ray diffraction.
The analysis of the soil samples from each site included
pH (soil-water ratio 1:2.5; Peech 1965), electrical conductivity of the saturated paste extract (soil-water ratio 1:2.5) (Bower & Wilcox 1965), total carbon (Richter & Wistinghausen
1981), cation exchange capacity (CEC), exchangeable Ca,
Mg, Na and K (Chapman 1965) and calcium carbonate
equivalent. The texture of the samples from each site was
determined using the gravimetric method. Subsoil samples
were dried at 40 C, filtered by 2 mm, homogenized and
stored in plastic bags before analysis. The total content of
metals (Fe, Mn, Al, Zn, Cu, Cr) in higher EC levels (25-50
cm) was also determined. XRD was performed using a Rigaku D-Max III, with Cu-Ka radiation and a graphite monochromator operated at 35 kV and 15 mA. XRD patterns were
recorded on whole rock from 2 to 60 2 and on the fraction
<2 m from 2 to 30 2 to check for impurities. Once carbonates were eliminated, the fraction < 2 m was obtained by
centrifuging a suspension at 750 rpm for 3.3 min. In order to
check for the presence of expandable clay minerals, samples
were intercalated with ethylene glycol.
Results
Physico-chemical and isotopic data (13C-DIC, 13CDOC, 18O and 2H) from the samples taken in September,
2012, are shown on Table 1. As it can be seen in Figure 2, the
influence of USW on the free aquifer resulted in a rise in Cl-,
SO 4 2-, Na+, HCO 3 -, Ca2+, Mg2+ and K+ compared to the values from the reference water for this study, M9w. Some exceptions to the rule are the Ca2+, Mg2+, Na+ and HCO 3 - concentrations from P10, the Mg2+ and Na+ from P7 and the
SO 4 2- from P9. Even if the differences are small, it is worth
pointing out that Cl- (conservative element and leachate indicator) concentrations increase considerably in all cases in
waters from the piezometers located downstream of the USW
disposal area. At TZ1, the chloride ratio was much higher
than at the other sampling sites (a 34:1 ratio), thus, the influence of another process other than mixing was considered.
The existence of leachate is not only confirmed by the presence of chlorides but also by an increase in the electrical
conductivity in water, an increase in DOC and a decrease in
dissolved oxygen when compared with values from M9w. pH
from M9w and LX1 are slightly alkaline whereas those from
VOLUME 42, NO. 2
MAY 2016
TABLE 1
Physico-chemical and isotopic data (13C-DIC, 13C-DOC, 18O and 2H) from water samples taken in September 2012.
M9w: pristine water; LX1: leachate. ND: not determinated
CDOC
CDIC
37.8
-24
-12.6
-1.8
31.2
-25.9
-11.8
-3.1
-10
16.7
-23.9
-6.7
-5.2
-25
6.5
-25
1.3
-5.7
-27
13
Sample
pH
CE S/cm
T C
O 2 mg/L
HCO 3 mg/L
P10
7.08
810
17.6
3.5
265
P7
6.96
920
17.6
3.2
477.3
P6
7.01
3340
17.5
2.8
1794
P11
7.39
1840
18.6
4.3
1004
Cl mg/L
2-
2+
2+
Mg mg/L
Na mg/L
K mg/L
NO 3
mg/L
CH 4
mg/L
SO4 mg/L
Ca mg/L
71
66.8
20.7
5.2
135.8
10.5
2.8
35.5
45.7
55.7
12.7
112.6
16.3
0.9
177.5
62
50.7
19.4
795.4
3.1
1.1
35.5
61.5
47
11.3
382.6
2.1
2.1
DOC
mg/L
13
18
P9
6.96
1730
17.4
2.8
1034.4
89
16.9
87
25.6
322
3.5
3.2
8.2
-24.7
6.1
-6.0
-29
P12
6.91
2020
17.4
2.9
1074
159.2
31.4
89.5
39.9
266.2
4.3
1.6
6.3
-25.7
2.5
-5.3
-26
P22
7.01
1480
16.3
4.3
730.5
88.8
63.6
90.8
28.1
205.4
5.8
4.1
3.1
-24.8
-5.6
-4.9
-24
M9w
7.4
757
18.5
6.2
417
13
25
51
13
146
2.4
3.28
nd
1.5
nd
-12.6
-4.8
-27
LX1
8.17
3207
22
0.2
1394
195
37.2
106
33.7
262.5
170
10.5
243
-25.4
9.1
-2.4
-4
TZ1
7.96
2710
18.1
3.2
710.3
443.8
157
53.2
60.5
500.8
0.8
3.9
4.4
-24.2
-12.5
nd
nd
FIGURE 2
Concentrations of major elements compared to the values from de reference water (M9w).
P samples are neutral. On the other hand, data represented on

a Piper diagram (Figure 3) show that the water samples, with
the exception of TZ1, fall into the HCO 3 - Na+ water type,
pointing out that the prevailing process has presumably been
cation exchange. TZ1 is sodium chloride-bicarbonate water
that would indicate the mixing of waters from the first type
with another source with a higher chloride content, compatible with a marine ingression. Metallic trace elements are
found in low concentrations (Table 2) and below the levels
allowed for human consumption according to the Argentine
Food Law, except for Mn that exceeds 0.10 mg/L in P7 P6 P9
P12 and P22. In all cases, the highest content of dissolved
metals was found in LX1, where Fe, Mn, Al, Zn, Cu and Cr
were present, with the exception of As, which presented the
highest concentration in piezometers P. Cd, Cr and Pb were
not found in those piezometers and Fe was only found in P10,

P7 and P6, whereas Al was found in P10 and P7 and Cu, in
P10, P11 and TZ1. Arsenic anomaly in samples P (higher
values than LX1) probably corresponds to background values
in the Chaco-Pampa Plain (Nicolli et al., 2012).
For the geochemical modeling, the following phases were
considered: Calcite (CaCO 3 ), due to its abundance in the
aquifer either in nodules, continuous levels or cementation;
(H-exchange) to represent the exchange of Ca2+ by Na+ and
H+ through clays (mainly montmorillonite-smectite), due to
the nature sodium bicarbonate water and neutral pHs; CO 2
phase as exchange with the atmosphere since it is an open
system; K-montmorillonite as capable mineral phase of incorporating K+ excess of the mixing; and Gypsum as source
of sulfate, characteristic of arid environments such as loess.
75
FIGURE 3
Piper diagram showing the chemical composition of
groundwater from wells.
The models selected to explain groundwater composition
(Table 3) show that the main operating processes are the
mixing with leachate, water evaporation, outgassing of CO 2 ,
cation exchange, gypsum salts dissolution or precipitation
and K + adsorption in the interlayer (exchangeable cation) of
soil montmorillonites already formed. The mixing percent-
ages found in the models reflect the influence of leachate on

water in all the samples, with percentages that vary according
to the analyzed piezometer. P22 and P12 are the most affected samples (35-66% of LX1). Since most piezometers chemical composition can be interpreted as mixtures between
M9W and LX1, taking into account that M9W (pH = 7.4),
and LX1 (pH =8.3) the mixture would result a water with
increased pH. Since the pH of such solutions are neutral and
lower than M9W, an additional source of H+ must be considered. The increment of H+ (and Na+) in the solutions can thus
be explained by positive transport of these ions that are interchanged by Ca2+ in active clays. These exchange processes
that take place in the free aquifer are held by the presence of
exchanging clays such as smectite, which are characteristic of
loess and alluvial plain. Also, Figures 3 shows the sodium
character of the waters. In the case of Mg2+/ Na+ exchange,
the transfer process (with positive and negative values) is
more variable. The geochemical modeling for TZ1 (Table 3)
gave similar processes with 2.4% leachate and 1.3% saline
water typical of a marine ingression.
The mixing percentages thrown by geochemical modeling
would indicate the presence of DOC in final water (1.5mg/L
for M9w and 243 mg/L for LX1). However, in Figure 4a, it
can be observed that DOC measured along a regional flow
line (P10 P11 P12 P22 TZ1) is not constant and decreases
downstream. Therefore, the possibility of an aerobic
oxidation of organic matter due to O 2 availability in water
(replacing the CO 2 phase with CH 2 O and O 2 and the constraint redox) was considered without obtaining valid results.
On the other hand, when results from 13C-DOC and 13CDIC (presented in Table 1) are spatially linked along the
regional flow line (Figure 4.A), little variations can be observed for the first parameter (between -25.7 and -24)
and an enrichment in 13C-DIC from P10 onwards (from 12.6 to 2.5). This same trend can be observed if the
results from a local flow line (P7 P6 P12 P22, Figure 4.B) are
analyzed, with a decrease in DOC from P7 onwards, almost
TABLE 2
Concentrations of trace metallic elements in piezometers and leachate expressed in mg/L, and levels (*) allowed for human
consumption to the Argentine Food Low (ANMAT, 2007).
76
Sample
Fe
Mn
Al
As
Cd
Zn
Cu
Cr
Pb
P10
0.221
0.022
0.125
0.019
0.031
0.018
P7
0.138
0.818
0.014
0.013
0.011
P6
0.058
0.719
0.03
0.009
P11
0.05
0.004
P9
0.274
0.021
0.007
P12
0.376
0.02
0.007
P22
0.121
0.018
0.022
TZ1
0.003
0.039
0.025
0.003
LX1
2,297
2,176
10.48
0.022
0.177
0.045
0.057
Lmit*
0.30
0.10
----
0.01
0.005
0.005
0.005
VOLUME 42, NO. 2
MAY 2016
TABLE 3
Results of geochemical processes in groundwater from wells P6, P11, P9, P12, P22, using NETPATH. Processes refer to mixing between M9w and LX1;
evaporation (positive sign); dissolution (positive sign) or outgassing (negative sign) of CO 2 in an open system; Ca2+/Na+-H+ exchange (positive sign indicates
the Ca2+ adsorption and the release of Na+ and H+ with molar ratio 1:1.6 Na+ and 0.4H+; Mg2+/Na+ exchange (positive sign indicates the Mg2+ adsorption and
the release of Na+ with molar ratio of 1:2); dissolution (positive sign) or precipitation (negative sign) of gypsum; precipitation of K-montmorillonite (negative
sign).* Only Exchange (Ca2+ adsorption and the release of Na+).
Evaporation
Factor
Mixing process (%)

Final water
Initial
M9w
water
Initial water
LX1
Dissolution or
precipitation
of calcite
Dissolution or
outgassing of
CO 2
Ca2+/Na+ and
H+ Exchange
Mg2+/Na+
Exchange
Dissolution or
precipitation
of gypsum
Precipitation
of K-mont
(mmol/L)
(mmol/L)
(mmol/L)
(mmol/L)
(mmol/L)
(mmol/L)
P6
79
21
3,461
-5,662
1,107
0.483
-0.1
-2,859
P11
97
1,920
0.812
0.771*
0.317
0.069
-0.495
P9
80
20
1,789
-4,395
0.148
0.118
-0.187
-2,656
P12
34
66
1,223
-15,498
0.254
-0.258
-0.078
-8,327
P22
65
35
1,136
-9,129
0.045
-0.178
0.277
-4,445
TABLE 4
Results of geochemical processes in groundwater from wells TZ1 using NETPATH. Processes refer to mixing
between M9w, LX1 and typical marine ingression (Plummer et al., 1991).
Processes
Evaporation
factor
Mixing process (%)
Final
water
Initial
water
M9w
TZ1
96
Initial water LX1
Initial
water
Ingresion
2.7
1.3
Dissolution
of calcite
Dissolution
of CO 2
(mmol/L)
1,523
Ca2+/Na+
Exchange
Mg2+/Na+
Exchange
Dissolution
of gypsum
k-mont
(mmol/L)
(mmol/L)
(mmol/L)
(mmol/L)
0.980
-0.333
0.415
-0.941
FIGURE 4
DOC (mg/L), 13C-DOC, HCO 3 - (mg/L) and 13C-DIC measured along a regional (A) and local (B) flow line.
constant 13C-DOC values between 25.9 and -23.9 and

enrichment in 13C-DIC (from -11.8 to 1.3).
The width of the unsaturated zone at some points of the
municipal site was minimal, with water depths that ranged

from 0.90 m to 0.20 m at P10, P7, P6, P9, P11, P12 and P22.
Thus, subsoil samples obtained from those piezometers (up to
77
2 m) would correspond to the unsaturated zone as well as part

of the saturated zone. Table 5 shows the results of sand, silt
and clay percentages, pH, CO 3 %, OM %, EC, CEC y exchangeable cations, taking as a reference the values from
SP10 for the loessic plain and from AP for the alluvial plain.
Regarding soil texture, SP10 has a greater content of sand
than AP (21- 47% vs. 1-9%), and the latter shows a clay
content (36-61% vs. 16-41%). Silt percentages are variable
and this fraction is present in both soundings (between 5 and
53%). pH ranges from 6.9 to 10.3 with prevailing values from
8 to 10, which indicate alkaline conditions in the subsoil.
CO 3 2- and CEC values from both units are similar to the
reference values, but anomalies are found in %OM and EC.
In the case of %OM, anomalies can be observed in almost all
the levels of SP12 and SP22. In the case of EC, anomalies are
found at SP9, SP12 and SP22, mainly at levels 25-50 cm. At
these 25-50 cm levels with high EC, Al, Cr, Cu, Fe, Mn and
Zn concentrations were measured (Table 6). When these
values are compared with the values from alluvial plain pristine sites (AP), metallic anomalies are observed, SP9 shows
higher concentrations of Al, Cu, Fe, Mn and Zn, SP12 of Cu,

Fe, Mn and Zn and SP22 shows higher concentrations of Al,
Cr, Cu, Fe and Zn. SP7 (located upstream of USW and in the
loessic plain) shows values even lower than AP (in the alluvial plain). Although CEC varies from level to level since it
depends on the content and type of clay minerals and organic
compounds, it could be said that the analyzed data, as a
whole, show that CEC values are naturally greater in AP
(between 31-56 cmolc/kg) than in SP10 (between 14-31
cmolc/kg) and that, in these samples, CEC values are mainly
correlated with clay content in a 66% (Figure 5). When comparing the sum of exchangeable bases (S = Ca2+ +Mg2+ + K+
+ Na+) with CEC, it can be observed that S is greater at some
levels, mainly at SP12 and SP22, that would indicate that
apart from exchangeable cations, soluble salts are also present. When this relation is analyzed in terms of % of base
saturation (V=(Ca+Mg+K+Na/CIC)x100)), V is over 50% in
many levels, which indicates a saturation of the adsorbent
compound regarding these cations.
The X-ray diffraction mineral analysis (Figure 6) of the
TABLE 5
Results of sand, silt and clay percentages, pH, CO 3 %, OM %, EC, CEC y exchangeable cations, taking as a reference the values from SP10
for the loessic plain and from AP for the alluvial plain.
Sand/Silt/Clay
(%)
pH
CO 3 (%)
OM (%)
EC (dS/m)
CEC (cmolc/kg)
Ca2+ (cmolc/kg)
Mg2+ (cmolc/kg)
Na+ (cmolc/kg)
K+ (cmolc/kg)
78
Deep level (cm)

0-25
25-50
75-100
125-150
175-200
0-25
25-50
75-100
125-150
175-200
0-25
25-50
75-100
125-150
175-200
0-25
25-50
75-100
125-150
175-200
0-25
25-50
75-100
125-150
175-200
0-25
25-50
75-100
125-150
175-200
0-25
25-50
75-100
125-150
175-200
0-25
25-50
75-100
125-150
175-200
0-25
25-50
75-100
125-150
175-200
0-25
25-50
75-100
125-150
175-200
SP10
47.8/5.5/16.8
42.8/30.5/26.8
30.3/27.9/41.8
21.6/51.6/26.8
40.6/32.7/26.8
6.9
8.5
9.5
9.7
9.5
1.4
1.4
7.6
1.5
2.4
3.63
0.99
0.33
0.29
0.21
0.27
0.55
2.19
2.03
1.04
14.8
20.3
29.4
22.1
31.2
6.4
6.3
6.6
2.9
4.8
1.7
2.4
4.2
2.2
3.4
2.1
2.4
17.8
17.6
14.5
1.4
3.6
0.9
0.8
0.4
SP7
52.2/34.4/13.4
51.1/30.5/18.4
45.8/42.1/12.2
39.8/31.9/28.4
53.3/23.3/23.4
10.3
10.1
9.3
9.3
9.5
2.8
3.1
3.3
2.7
2.2
0.20
0.19
nd
nd
0.17
1.56
1.16
0.67
0.81
0.65
9.6
12.1
11.6
18.7
15.9
2.6
2.8
6.2
5
4.2
0.4
0.8
4.8
5.3
2.7
15.7
16.1
13.3
15.4
9.9
0.4
0.3
0.5
0.7
0.4
PA
9.9/48.4/41.8
9.4/53.8/36.8
4.2/34/61.8
1/37.2/61.8
SP9
10.7/52.6/36.8
7.2/36.1/56.8
6.4/31.9/61.8
35.3/27.9/36.8
SP12
22.4/60.9/16.8
19.7/38.6/41.8
12.1/26.1/61.8
49.8/23.4/26.8
7.8
8.1
8.2
7.8
8.6
8.2
SD
9.2
8.4
8.5
SD
9.1
2.3
4.7
8.6
5
2
5.9
1.7
1.9
1.3
4.9
14.1
1.7
1.41
1.06
0.88
0.35
1.93
0.84
0.98
0.35
1.45
1.92
1.32
0.26
0.21
0.19
0.66
1.76
1.26
4.86
nd
0.65
2.77
5.88
nd
0.66
31.8
26
43.4
56.5
35.1
41
43
30.3
18.3
32
49.4
18.7
23.2
29.6
36.4
25.8
13.2
20.3
7.8
7.3
6.4
14.6
20.8
8.8
2.5
2.8
8
8.7
4.9
7.8
7.1
3.9
3.8
6.7
8.3
3.3
0.5
0.6
5.4
11.2
10.1
19
20.2
7.2
9.6
19
23.3
4.8
0.5
0.4
0.7
0.8
0.4
0.8
0.7
0.4
0.7
0.7
0.9
0.4
SP22
4.2/47.5/48.4
4.1/41.2/54.7
6.3/47.8/45.9
1.2/55.4/43.4
1/50.6/48.4
7.3
7.5
8.2
8
7.9
1.9
3.5
2.3
1.9
2
4.67
2.39
1.08
0.90
0.89
0.62
4.29
1.07
0.34
0.23
27.8
29.3
29.1
27.2
26.8
17.1
13.2
13.2
15.7
16.5
6.7
7.7
7.3
7.3
6.4
7.2
17.4
16.2
8.9
3.4
0.7
0.5
0.6
0.7
0.8
TZ1
52.8/32.9/14.3
45.3/28/26.8
37.6/33.2/29.3
27.1/38.7/34.3
37.6/30.6/31.8
8.5
9.1
8.7
9.8
9.7
2.1
1.8
1
5.8
3.5
1.71
0.34
0.67
nd
0.33
0.12
0.86
0.64
1.24
1.2
13.9
16.9
19.3
25.2
23
6.8
12.5
13.2
15.2
16
1.6
2.8
2.4
3.1
3.4
0.8
11.1
6.5
23.6
17
0.3
0.4
0.5
1.2
1
VOLUME 42, NO. 2
MAY 2016
TABLE 6
Concentrations of trace metallic elements in sediments (level 25-50 cm).
Sample
Al
Cr
Cu
Fe
Mn
Zn
(mg/kg)
(mg/kg)
(mg/kg)
(mg/kg)
(mg/kg)
(mg/kg)
PA
4272
2.6
6.4
2601
241
9.2
SP9
6631
12.8
4070
511
19.1
SP12
4321
8.4
2813
474
10.5
SP22
9722
5.9
19.2
7104
52
18.8
SP7
2548
1.9
5.2
1647
313
5.3
FIGURE 5
CEC values correlated with clay content from subsoil samples.
loessic plain (LP), represented by P10, indicates the presence
of quartz, potassium feldspars-plagioclase and clays with
variable proportions of tridymite-cristobalite in all levels and
of calcite from 50cm onwards. In the alluvial plain sediments
(AP), the same minerals were identified in all the levels with
the exception of cristobalite, which was only found in the
first 25 cm. In both cores, clay content increased towards the
deeper levels. The fraction smaller than 63 microns from
levels 25-50 cm at AP and V10 coincides with whole sample
analysis for this level, regarding its mineralogy. Clay fraction
analysis for levels 25-50 at AP and VZ2 indicate the presence, in decreasing order, of smectites, illites and kaolinites,
with a prevalence of the first one (Figure 7). When
discontinuous levels SP9, SP12 and SP22 were compared
with PA, it was found the same mineralogy, as in TZ1.
DISCUSSION
Figure 8a (2H vs. 18O cross plot) show the isotope composition of a mill network producing for the water table and
surface waters sampled during March, 2010 (Boujon, 2014)

together with the local meteoric water line (LML) from the
station INGEIS Ciudad Universitaria, Buenos Aires
(INGEIS-CU), the closest to the study area (180 km southwards), whose parameters do not differ significantly from
that of Santa Fe (260 km NW) (Dapea and Panarello, 20042007). Therefore, as there is no LML from Gualeguaych, the
LML from Ciudad Universitaria can be used without generating important errors. In addition the piezometer series was
also plotted. Free aquifer isotope data, as indicated by the
mill network, fall in a cloud around the average weighted
value of INGEIS-CU station i.e. 18O ~ -5 and 2H~ -26.
On the other hand water from the Arroyo El Cura and tributaries and the Gualeguaych River exhibit more depleted 2H
and 18O values. Piezometers values distribute along the
MWL showing rapid response to the rain. As established by
Gat (2010) this picture is related to a Rayleigh fractionation
process produced by the so called amount effect in which
the isotopic composition of a precipitation event become
more depleted as the rain out proceed. The mills respond to
whole precipitation whereas piezometers infiltrate only during a period of time and become saturated leaving to the
79
FIGURE 6
Identification of minerales by x-ray diffraction (XDR) of SP10 y AP (whole sample and fraction smaller than 63 m).
Clay, Qtz (quartz), Fd and Ca (potassium feldspars-plagioclase).
runoff the remaining, more negative fraction of rain. This
runoff reaches the creeks and rivers which discharge these
depleted water that can be mixed with the base flow (free
80
aquifer).
Figure 8b present a detail the isotopic composition of the
piezometers, together with a sample of the leachate (LX1)
VOLUME 42, NO. 2
MAY 2016
FIGURE 7
Identification of minerales by x-ray diffraction (XDR) of SP22 y AP (clay fraction).
Sm (smectite), I (illite), K (kaolinite), Qtz (quartz).
FIGURE 8
2H vs. 18O cross plot showing the isotope composition of a mill network producing for the water table and surface waters
sampled during March, 2010 (Boujon, 2014) ) (a) together with the local meteoric water line (LML)
and piezometer series (September 2012) (b).
and M9W corresponding to the average value of pristine
water. PP represents the water input as precipitation. Isotopic
composition of these rains is compatible with the prevalence
of recycled rains (Gat, 2010), characterized by a high deuterium excess, and that statistically occur during May-June
months in that region (Dapea, 2007). In Figure 8b, it can be
seen that the isotopic composition of the waters - with the
exception of P10 - can be explained as a ternary mixing with
variable proportions from samples M9W, LX1 and PP. In
turn, P9, P12, P11, P22 and P6 line up along a line with a
slope of 4.64, which can be interpreted as the result of an
evaporation line with a greater gradient (in accordance with
the meteorological conditions of the area) with varying contributions of water isotopically similar to that from M9w. It is
worth mentioning that, given its position on the LML, P7
could be the result of the above-mentioned mixing, but this is

not supported by the hydrochemical characteristics or by the
geographical location of the piezometers with respect to the
USW location. Isotopic differences between P7-P10 (upstream from the USW) and the other P piezometers (P9, P12,
P11, P22 and P6 from downstream) could be explained by the
incidence of the materials that were removed to cover the
SIW (heterogeneous and non-compact), fact that affects the
last group of piezometers and not P7-P10. This situation
probably generates a differential vertical infiltration of rain in
the ground and thus the seepage of volume elements with
different isotopic composition due, as mentioned, to the
amount effect.
Leachate chemical composition significantly varies according to the nature of waste, its age and the practices asso-
81
ciated to the USW burial. In general, all species concentrations are much higher in a young leachate than in an old one.
In the former case, dissolved salts and heavy metals contents
are much higher than the permissible limits for drinking water (Ogundilan and Afolabi, 2008). In this case, even though
LX1 falls within the ranges proposed for this type of effluents
(Table 7), the major and trace elements concentrations measured exceed the permissible limits for drinking water of the
Argentine Food Law only in Cl-, higher than 350 mg/L at
TZ1, total hardness (as CaCO 3 ), higher than 400 mg/L at
LX1 and Mn, higher than 0.10 mg/L at P7, P6, P9, P12 and
P22. This situation can occur either because LX1 type water
leached as a result of the action of groundwater coming from
precipitations (LX1 does not differ significantly from the P
waters, although it is more concentrated- Figure 3 and it falls
on the meteoric line-Figure 5), or it is linked to the age of
USW burial and its relation with the state of degradation.
There is a strong relationship between the state of refuse
decomposition and its associated leachate characteristics
(Farquhar and Rovers, 1973). Generally, it is accepted that
landfills undergo at least four phases of decomposition, (1) an
initial aerobic phase, (2) an anaerobic acid phase, (3) an initial methanogen phase, and (4) a stable methanogen phase
and an additional aerobic or humic phase of decomposition
has been proposed (Bozkurt et al., 2000). Geochemical study
of gases made on site (concentrations of CO 2 , CH 4 , N 2 and
O 2 as well as carbon isotope composition of the CO 2 )
showed that the process underway is an aerobic phase of
biodegradation (Sanci et al., 2012). Lack of methane and
other evidence allowed to conclude that the time since the
end of operations in the MSWFDS (14 years) makes it possible to assume that we are in the presence of a posthumous
stage of biodegradation or postmethanogen phase. Sampling
2012 is also not found CH 4 dissolved in water.
Several authors (Christensen et al. 2001; Kjeldsen et al.,
2002; Bjerg et al., 2005) refer to an acid phase and a
methanogenic phase when talking about leachate, distinguishing how the concentration of several parameters vary in each
phase. Although it is true that LX1 could be a postmethanogenic leachate, since the presence of CH 4 in water could not
be corroborated (Table 1), it was observed that it affects the
quality of the pristine water far beyond the evaporation indicated by geochemical models. DOC, EC and concentrations
at P and TZ1 piezometers are higher than at M9w, especially

in Cl-, HCO 3 -, SO 4 2- and Na+ (Figure 2). These parameters
are all indicative of leachate. TZ1 is also influenced by the
marine ingression, whose field evidences were also seen in a
sounding near TZ1 and within the municipal site (Boujon and
Sanci, 2014). Moreover, there are EC and OM anomalies in
subsoil samples from P12 and P22 (with higher concentrations than the reference values), as well as in the water from
them, which showed the highest mixing ratios (35% and
65%, Table 3).
pH values are variable if we consider the mixing ratio
indicated by the geochemical models (Table 3) and the pH
value of the initial waters (M9w with 7.4 and LX1 with 8.17).
The exception is P11 that only has a 3% of LX1 and the same
pH as M9w. P12, P9 and P22 show more acid pHs, even
more than pristine water, which lead to think that besides the
exchange of Ca2+ for Na+ (sodium bicarbonate waters) there
was an exchange with H+ according to the following reaction:
Na+H+X + Ca2+
CaX + Na+ + H+
When there are acid inputs, pH decreases but, since

bicarbonates are present, this excess is released as CO 2 and
CO 2 + H 2 O).
pH remains near neutrality (HCO 3 - + H+
The process is, in short, a neutralization that justifies the pH
obtained. In the case of our study, ground materials act
mostly as aquifer sediments since groundwater depths are
shallow. Therefore, ion exchange and cation retention can
occur either from clays or from organic matter in the subsoil
(both present in the site). The results of geochemical models
show a positive transfer of Na+ and H+ to the water with Ca2+
adsorption and, in Figure 2, Ca2+ does not seem to increase in
a significant way when compared to M9w. Since organic
matter also exchanges and subsoil data indicate that there is
base saturation (Table 5), it is possible to think that the acidity is provided by the organic acids present in the subsoil.
OM anomalies detected at P12 and P22 are possibly due to
the presence of leachate acids. Organic acids provide H+ to
the soil solution at a variable pH (Conti, 2007).
On the other hand, 13C-DIC in wells P6 P9 P11 P12 P22
is enriched with respect to 13C-DIC background (-12.6 ).
High 13-DIC values are generally observed for landfill
TABLE 7
Typical range of concentrations of major and trace elements concentrations proposed by
Kjeldsen et al. (2002) and Christensen et al. (2001).
82
Sample
pH
CE
Cl-
SO42-
Ca2+
Mg2+
Na+
K+
HCO3-
LX1
8.1
2940
195
37.2
106
33.7
262.5
170
1394.12
Typical Range
4.5-9
2500-35000
150-4500
Aug-50
Oct-00
30-15000
70-7700
50-3700
610-7320
Sample
Cd
Cu
Cr
Pb
Zn
Fe
Mn
As
LX1
0.045
0.057
0.177
2,297
2,176
0.022
Typical Range
0.0001-0.4
0.005-10
0.02-1.5
0.001-5
0.03-1000
Mar-00
0.03-1400
0.01-1
VOLUME 42, NO. 2
MAY 2016
leachate and are thought to be related to production of

isotopically positive CO 2 by methanogenesis (Hackley et al.,
1996), and a lesser extent by outgassing of CO 2 (Baedecker
and Back, 1979; Hornibrook et al., 2000). Since no CH 4 was
found in water (consistent with the presence of dissolved O 2 ),
the isotopic enrichment would be due to CO 2 degassing (as
indicated by geochemical models, Table 3), which must have
taken place in the final waters to compensate the carbon excess produced by mixing and evaporation. Moreover, 13C
values of final waters are also influenced by the M9w and
LX1 mixing ratio. M9w and LX1 water types have their own
isotopic signatures: M9w 13C, -12.6 and LX1 13C,
+9.1. It is worth mentioning that if we calculate 13C values of final waters from the mixing percentage and the 13C
values of each initial water, estimated 13C values of final
waters show less enrichment, that is why it is supposed that
there is fractionation through degassing. Evaporation does not
generate isotopic fractionation in 13C (Mook, 2000). P11 only
has 3% of LX1 (it could be said that it has no leachate mixing) and it is the only that shows CO 2 dissolution, thus a 13C
value of +1.3 is probably the result of intake of heavy
subsoil CO 2 generated during the USW degradation (Sanci et
al., 2012).
DOC decreases from 16 to 3.1 mg/l in downstream direction, while no dilution appears to take place. On the contrary,
the isotopes 18O and 2H contribute to validate the evaporation
process indicated by the models. The deuterium excess value
of the samples (located over the local meteoric water line,
Figure 8) may indicate exchange between H from water with
that from the dissolved CH 4 (Clark and Fritz, 1997), but we
found no CH 4 that would justify this. It is believed that the
evaporation could be due to a process occurring prior to infiltration and/or evaporation in the aquifer consistent with the
rise of the up-going flow at a discharge area, as in this case.
13C-DOC showed a tendency to stay constant with distance
from the landfill and an inverse correlation with DOC concentration was not found. Degradation of certain organic
compounds favors the 12C with respect to the 13C isotopes,
which results in enrichment for 13C in the residual fraction
(Rayleigh process, van Breukelen et al., 2003). Therefore, the
no observed correlation between increasing 13C-DOC and
decreasing DOC concentrations. It is more likely to have
occurred DOC sorption in sediments rather than oxidation.
Evidence could be that the DOC in P12 and P22 decreases
and in those wells (with LX1 higher percentage) increases
organic matter (Figure 4). Besides the DOC in the final water
should be more concentrated when taking into account the
mixing ratios between M9W (1.5 mg / l) and LX1 (243 mg /
l), and that does not happen.
According to Kawahigashi et al. (2006) sorption of dissolved organic matter (DOM) on mineral phases is an important geochemical process for carbon preservation and
element cycling in soils and aquatic systems. Sorption is
more rapid than microbial degradation (Qualls & Haines,
1992; Kalbitz et al., 2003) and, therefore, is more efficient in
removing DOM from water percolating through soil. Also,
Kjeldsen et al. (2002) say that the acid phase, the leachate
may show low pH values and high concentrations of many
compounds, in particular high concentrations of easily de-
gradable organic compounds as volatile fatty acids. But in the

later stable methanogen phase, the pH increases and the degradability of the organic carbon is lowered dramatically. The
acids that accumulated in the acid phase are converted to
gases and cellulose and hemicellulose decomposition also
begins. The pH increases as acids are consumed. In the stable
methanogenic phase, the methane production rate will reach
its maximum, and decrease thereafter as the pool of soluble
substrate (carboxylic acids) decreases. In this phase, the rate
of CH 4 production is dependent on the rate of cellulose and
hemicellulose hydrolysis. The pH continues to increase to
steady state pool concentrations that are on the order of a few
mg/L. Some COD is present in the leachate, but it is mostly
recalcitrant compounds such as humic and fulvic acids
(Barlaz and Ham, 1993; Christensen et al., 1994). In this way
compounds of high molecular weight and rich in acidic functional groups and aromatic moieties sorb most strongly
(Kawahigashi et al. 2006).
Total soil contents of Fe, Mn and Al at P9, P12 and P22
increase with respect to AP (Table 6) and in those piezometers, Fe and Al are not detected in the water and water Mn
content decreases. It cannot be ascertained that those soil
anomalies are only the result of the presence of oxides, since
they were not detected by X-ray diffraction, but there is a
possibility, since metals tend to fix in the soil at basic pHs
and thus its leaching rate is low (Stumm and Morgan, 1996).
Therefore in P9 P12 and P22 (where water and sediment was
measured) could be said that the precipitation of trace
element (as oxides) would control the concentration of Fe and
Al in water and that both contribute to the adsorption of
organic matter (the sorption depends much on the contents of
Al and Fe oxides and carbon in soils, Kahle et al, 2004).
Metals that are added due to leachate input can be captured
by the formation of secondary minerals such as
montmorillonite, illite, kaolinite, trace metals precipitation
and adsorption (McLean and Bledsoe, 1992). These processes
are the ones that control the concentration of many metallic
elements in water.
CONCLUSIONS
The results obtained from physico-chemical and isotopic
data provided useful evidence of leachate migration in the
groundwater. The presence of leachate is indicated by anomalies in the content of the inorganic conservative element in
waters from well P (located downward the USW disposal
site) in relation to the content of those elements in M9w (non
contaminated water). High values in major elements in the
water are produced not only by antropic processes but also by
other natural ones such as evaporation in P6, P11, P12 y P22
and marine ingression in TZ1. When all these processes are
put together, the quality of water change mainly due to presence of those elements which are in concentrations unfit for
human consumption. What is clear is that the soil EC increases in wells affected by USW what would indicate the presence of soluble salts coming from the leachate.
Environmental isotopes (carbon, oxygen and hydrogen)
were a useful tool in validating the geochemical models. 18O
83
y 2H values suggest not only a meteoric origin for the

ground waters but also that samples P6 P11 P9 P12 y P22
have undergone a evaporation and mix process of waters of
the type PP, M9w y LX1. Moreover, C isotopes, particularly
those values enriched in 13C-DIC, demonstrated that CO 2
outgassing could have happened during the mixing. No evidences were found (such as CH 4 dissolved in water) that
could identified another process, such as metanogenesis, that
could have been responsible for the 13C enrichment.
Soil-water interaction allow the existence in the area of
natural attenuation processes in the water such as organic
matter sorption in the subsoil, as shown by the lowering in
the concentration of DOC in water, the rising of OM in the
subsoil (mainly in those samples affected by mix of waters
such as P12 y P22) and the constant values of 13-DOC in
water. Furthermore, precipitation and/or adsorption of trace
elements like Fe and Al, control the concentration of these
elements in water, since they are not present in wells P11 P12
P22 P9 and TZ1 but are found in higher amounts in the subsoil. Those wells are located in the alluvial plain (higher clay,
CEC and OM content), suggesting that the exchange and/or
sorption processes were effective. Smectite was identified as
the main clay mineral. This mineral not only exchange Ca2+
and/or Mg2+ for Na+, but also could neutralize water, exchanging metals and H+ and incorporated excess K+ from
leachate.
ACKNOWLEDGEMENT
This study was supported by Project PICT 2010 2749 of
National Agency for Scientific and Technological Promotion
from Argentina. We would like to special acknowledge the
Stable Isotopes and Agroambiente Laboratories of INGEIS
for collaborating with us in this study and Mr. Llambias and
Mr Giordanengo for their participation in the field work.
REFERENCES
ANMAT (2007). Cdigo Alimentario Argentino. Ley 18284.
Captulo XII: Agua Potable. Bebidas hdricas, agua y agua
gasificada.
http://www.anmat.gov.ar/alimentos/normativas_alimentos
_caa.asp
Auge, M., C. Snchez, & M. Santi, (2005). Hidrogeologa
de la regin arrocera de Entre Ros. IV Congreso Hidrogeolgico Argentino, Actas CGO_01: pp. 1-9, Ro Cuarto, Argentina.
Baedecker, M. & W. Back, (1979). Hydrogeological process
and chemical reactions at a landfill. Groundwater, Volume 17, pp. 417-429.
Barlaz, M.A. & R.K. Ham (1993) Leachate and gas generation. In D.E. Daniel (Ed.) Geotechnical Practice for
Waste Disposal (p. 113). London: Chapman and Hall.
Bjerg, P.L., H.J. Albrechtsen, P. Kjeldsen, & T.H. Christensen, (2005). The groundwater of waste disposal facilities. In B.S. Lollar (Ed.), Tratise on Geochemistry (Volume 9), Environmental Geochemistry (pp. 579-612). New
84
York: Springer.
Bocanegra, E., H. Massone, D. Martinez, E. Civit, & M.
Farenga, (2001). Groundwater contamination: risk management and assessment for landfills in Mar del Plata, Argentina. Environmental Geology, Volume 40(6), pp. 732741.
Boujon, P. & R. Sanci, (2014). Evaluacin de la vulnerabilidad del acufero libre en la cuenca del Arroyo El Cura,
Gualeguaych, Entre Ros. Revista de la Asociacin Geolgica Argentina, Volume 71(2), pp. 274-289.
Boujon, P. (2014). Estudio de la contaminacin de aguas
subterrneas por mtodos geofsicos e hidrogeolgicos.
Tesis doctoral, Facultad de Ciencias Exactas y Naturales,
Universidad de Buenos Aires
Boujon, P., R. Sanci, F. Stckli, C. y Dapea, (2006). Caracterizacin hidroqumica e isotpica de las aguas subterrneas de la ciudad de Gualeguaych, Entre Ros, Argentina. VIII Congreso Latinoamericano de Hidrogeologa
(ALSHUD), Memorias Anais: 32, Asuncin, Paraguay.
Boujon, P., C. Pomposiello, & R. Sanci, (2011). Estudios
geofsicos en sitio de disposicin final de residuos slidos
urbanos ciudad de Gualeguaych, Entre Ros, Argentina:
evidencia de contaminacin. VII Congreso de Hidrogeologa y V Seminario Hispano Latinoamericano sobre temas actuales de la Hidrologa Subterrnea. Hidrogeologa
Regional y Exploracin Hidrogeolgica, pp. 205-212. Salta, Argentina.
Bower, C.A. & L.V. Wilcox, (1965). Soluble salts. In C.A,
Black (Ed.). Methods of soil analysis (pp. 933-940). Madison, WI, USA: American Society of Agronomy.
Bozkurt, S., L. Moreno, & I. Neretnieks, (2000). Long-term
processes in waste deposits. The Science of the Total Environment, Volume 250, pp. 101-121.
Castaeda, S.S., R.J. Sucgang, R.V. Almoneda, N.D.S. Mendoza, & C.P.C. David, (2012). Environmental isotopes
and major ions for tracing leachate contamination from a
municipal landfill in Metro Manila, Philippines. Journal
of Environmental Radioactivity, Volume 110, pp. 30-37.
Chapman, H.D. 1965. Cation exchange capacity. In: C.A.
Black (Ed.). Methods of soil analysis (pp. 891-900). Madison, WI, USA: American Society of Agronomy.
Christensen, T.H., P. Kjeldsen, H.J. Albrechtsen, G. Heron,
& P.H. Nielsen, (1994). Attenuation of landfill leachate
pollutants in aquifers. Critical Reviews in Environmental
Science & Technology, Volume 24, pp. 119-202.
Christensen, T.H., P. Kjeldsen, P.L. Bjerg, D.L. Jensen, J.B.
Christensen, A. Baun, H.J. Albrechtsen, & G. Heron,
(2001). Biogeochemistry of landfill leachate plumes.
Applied Geochemistry, Volume 16, pp. 659-718.
Clark, I.D. & M. Fritz, (1997). Environmental Isotopes in
Hydrogeology. New York, Lewis Publishers.
Conti, M. 2007. Materia orgnica, En M. Conti (Ed) Principios de edafologa, Buenos Aires: Facultad de Agronomia, UBA, pp. 67-85.
Dapea, 2007. Istopos estables livianos. Su aplicacin en
hidrologa e hidrogeologa. Tesis doctoral, Facultad de
Ciencias Exactas y Naturales, Universidad de Buenos Aires.
Dapea, C. & H.O. Panarello, (2004). Composicin isotpi-
VOLUME 42, NO. 2
MAY 2016
ca de la lluvia de Buenos Aires. Su importancia para el estudio de los sistemas hidrolgicos pampeanos. Revista
Latino-Americana de Hidrogeologa, Volume 4, pp. 1725.
Dapea, C. & H.O. Panarello, (2007). Estacin Santa Fe. Red
Nacional de Colectores. Argentina (resumen), V Congreso
Argentino de Hidrogeologa, Paran, Entre Ros, IAH
Universidad Nacional de Entre Ros, pp. 187-198.
Farquar, G.J. & F.A. Rovers, (1973) Gas production during
refuse decomposition.Water Air and Soil Pollution, Volume 2, pp. 483495.
Gandhi, H., T.N. Wiegner, P.H. Ostrom, L.A. Kaplan, & N.E.
Ostrom, (2004). Isotopic (13C) analysis of dissolved organic carbon in stream water using an elemental analyzer
coupled to a stable isotope ratio mass. Rapid Communications in Mass Spectrometry, Volume 18, pp. 903906.
Garca I. C. y Dorronsoro (2005). Contaminacin por metales
pesados. Tecnologa de Suelos. Universidad de Granada.
Departamento de Edafologa y Qumica Agrcola.
http://edafologia.ugr.es. Gat, J.R. (2010). Isotope Hydrology. A Study of the Water Cycle. Centre for Environmental Technology. London: Imperial College.
Guida, N. M. y Gonzlez, 1984. Evidencias paleoesturicas
en el sudoeste de Entre Ros, su evolucin con niveles marinos relativamente elevados del Pleistoceno Superior y
Holoceno. 9 Congreso Geolgico Argentino, Actas 3: pp.
577-594, Bariloche.
Hackley, K.C., C.L. Liu, & D.D. Coleman, (1996). Environmental isotope characteristics of landfill leachates and
gases. Ground Water, Volume 34(5), pp. 827836.
Hornibrook, E.R.C, F.J. Longstaffe, & W.S. Fyfe, (2000).
Evolution of stable carbon isotope compositions for methane and carbon dioxide in freshwater wetlands and other
anaerobic environments. Geochimica et Cosmochimica
Acta, Volume 64 (6), pp. 10131027.
Iriondo, M. (1980). El Cuaternario de Entre Ros. Revista
de la Asociacin de Ciencias Naturales del Litoral, Volume 11, pp. 125-141.
Iriondo, M. (1996). Estratigrafa del Cuaternario de la cuenca
del Ro Uruguay. XIII Congreso Geolgico Argentino y
III Congreso de Exploracin de Hidrocarburos. Actas 4:
pp. 15-25, Buenos Aires.
Kahle M., M. Kleber R. Jahn (2004). Retention of dissolved
organic matter by phyllosilicate and soil clay fractions in
relation to mineral properties. Organic Geochemistry,
Volume 35, pp. 269276.
Kalbitz, K., J. Schmerwitz, & D. Schwesig, (2003) Biodegradation of soil-derived dissolved organic matter as related to its properties. Geoderma, Volume 113, pp. 273
291.
Kawahigashi, M., K. Kaiser, A. Rodionov, & Guggenberger,
G. (2006). Sorption of dissolved organic matter by mineral soils of the Siberian forest tundra. Global Change
Biology, Volume 12, pp. 18681877.
Kjeldsen, P., A. Morton, A. Barlaz, P. Rooker, A. Baun, A.
Ledin, & T.H. Christensen, (2002). Present and LongTerm Composition of MSW Landfill Leachate: A Review. Critical Reviews in Environmental Science &
Technology, Volume 32(4), pp. 297336
Lomond, J.S. & A.Z. Tong, (2011). Rapid Analysis of Dissolved Methane, Ethylene, Acetylene and Ethane using
Partition Coefficients and Headspace-Gas Chromatography. Journal of Chromatographic Science, Volume 49,
pp. 469-475.
MacCrea J.M. (1950). On the isotopic chemistry of carbonates and paleotemperature scale. Journal of Chemical
Physics, Volume 18(6), pp. 849857
McLean, J.E. & B.E. Bledsoe, (1992). Behavior of metals in
soils. Technology Innovation Office, Office of Solid
Waste and Emergency Response, 1992 - Hazardous waste
sites - EPA/540/S-92/018, 25 pages.
Mook, W.G. (2000). Natural abundance of the stable isotopes
of C, O and H. In W.G. Mook, (Ed.) Environmental isotopes in the hydrological cycle. Principles and applications (pp. 89-120). Paris: UNESCO.
Nicolli, H.B., J. Bundschuh, M.C. Blanco, O.F. Tujhneider,
H.O. Panarello, C. Dapea, & J.E. Rusanski, (2012). Arsenic and associated trace-elements in groundwater from
the Chaco-Pampean plain, Argentina, Results from 100
years of research. Science of the total Environment, Volume 429, pp. 36- 56.
North, J.C., R.D. Frew & B.M. Peake, (2004). The use of
carbon and nitrogen isotope ratios to identify landfill
leachate contamination: Green Island Landfill, Dunedin,
New Zealand. Environment International, Volume 30(5),
pp. 631637.
Ogundiran, O.O. & T.A. Afolabi, (2008). Assessment of the
physicochemical parameters and heavy metals toxicity of
leachates from municipal solid waste open dumpsite. International Journal of Environmental Science & Technology, Volume 5 (2), pp. 243-250.
Panarello, H.O., C.M. Garcia, S.A. Valencia, & E. Linares,
(1982). Determinacin de la composicin isotpica del
carbono en carbonatos, su utilizacin en Hidrogeologa y
Geologa. Revista de la Asociacin Geolgica Argentina,
Volume 35(4), pp. 460-466.
Plummer, L.N., E.C. Prestemon, and D.L. Parkhurst, (1991).
An interactive code (NETPATH) for modeling NET geochemical reactions along a flow PATH: U.S. Geological
Survey Water-Resources Investigations Report 91-4078,
227 p.
Plummer, L.N., E.C. Prestemon, and D.L. Parkhurst, (1994).
An interactive code NETPATH) for modeling NET geochemical reactions along a flow PATH--version 2.0: U.S.
Geological Survey Water-Resources Investigations Report
94-4169, 130 p.
Qualls, R.G. & B.L. Haines, (1992) Biodegradability of
dissolved organic matter in forest through fall, soil solution, and stream water. Soil Science Society of America
Journal, Volume 56, pp. 578586.
Richter, M. & E. Wistinghausen, (1981). Unterscheidbarkeit
von humusfraktione in boden bei unterscheidlicher
bewirtschaftung. Z. Pflanz. Bodenkunde, Volume 144,
pp. 395406.
Sanci, R., H.O. Panarello, & H. Ostera, (2012). CO 2 emissions from a municipal site for final disposal of solid
waste in Gualeguaych, Entre Rios Province, Argentina.
Environmental Earth Science, Volume 66, pp. 519528.
85
Soil Survey Staff (2014) Keys to Soil Taxonomy, 12th ed.

USDA-Natural Resources Conservation Service, Washington, DC.
Stumm, W. & J.J. Morgan, (1996). Chemical Equilibria and
Rates in Natural Waters. In John Wiley & Sons (Ed.)
Aquatic Chemistry (pp. 764-768). New York.
Valencio, S., H. Ostera, & H. Panarello, (2003). Monitoring
86
13
C in CO 2 soil gas from a landfill: first results from a

case study. IV South American Symposium on isotope
Geology, Short Papers I: pp. 128-130, Salvador, Brasil.
Van Breukelen, B.M., F.M. Roling, J. Groen, J. Griffioen, &
H.W. Van Verseveld, (2003) Biogechemestry and isotope geochemistry of a landfill leachate plume. Journal
of Contamination of Hydrology, Volume 65, pp. 245-268.
VOLUME 42, NO. 2
MAY 2016
WASTE-TO-ENERGY CHARACTERIZATION AND MODELING: A

CASE STUDY AT THE US NAVAL ACADEMY
P.A. Caton*, J. Schmidt, C. Adsit, E. Bermudez, C. Chase, T. Kerner
Department of Mechanical Engineering, US Naval Academy
590 Holloway Road
Annapolis, MD 21402
patcaton@usna.edu
ABSTRACT
This study focused on the potential for energy recovery of the low-food waste stream via gasification on a college campus. Waste was sampled using standard sampling protocols from several locations in both academic and residential buildings. Each sample was analyzed qualitatively (categorically) and quantitatively (proximate/ultimate analysis). Relative to average US
domestic municipal solid waste (MSW), these waste samples contained higher levels of paper
and plastic, comparable food waste, and lower levels of other components. Overall, the waste
had lower moisture content and higher energy content than average domestic US MSW. Fuel
lean combustion was simulated with excess air and suggested campus waste would produce
higher reactor temperatures than average, domestic MSW. The potential for gasification to a
producer gas was simulated using a chemical equilibrium approach. Cold gas efficiencies of
85% followed by electric power generation with typical efficiency would result in the potential
offset of 2.5% of institutional electricity use, but a combined heat and power approach could also offset a substantial amount of the thermal energy requirements of the institution. A simplified
three-component surrogate waste sample was proposed; analysis of two candidate mixtures
showed overall good representation of actual campus waste and can facilitate future experimental efforts.
Keywords: Waste characterization, gasification, waste-to-energy, college campus waste,
chemical equilibrium, combined-heat-and-power
INTRODUCTION
Waste-to-energy conversion provides a way to utilize
wastes and reduce disposal costs while simultaneously
providing a source of power to offset conventional fuel use.
Combustion-based mass burn facilities for waste-to-energy
conversion have existed for decades. Waste-to-energy conversion via gasification is an emerging technology that has
some interesting advantages and challenges. Pytlar reviewed
waste biomass gasification and pyrolysis facilities in North
America in 2010 and suggests these facilities have an advantage over traditional mass burn combustion facilities due
to the potential for reduced emissions and reduced odor. He
identified 22 facilities at 19 sites that focus specifically on
heat and/or power generation using biomass wastes with gasification. The majority of these installations, however, are
designed to handle a specific waste feedstock (e.g. chicken
litter) or a variety of similar feedstocks (e.g. rice husks and
straw) [1]. Chaivichitmalakul et al. identify biomass waste
gasification as an important technology for the developing
world, and identify several economic factors that are im-
_______________________________________
*Corresponding author
WASTE-TO-ENERGY CHARACTERIZATION AND MODELING: A CASE STUDY AT THE US NAVAL ACADEMY
87
portant in successful implementation in the context of development [2].

However, applying gasification to a waste-to-energy process is challenging. For example, Davis et al. summarize the
implementation problems from a recent effort to gasify organic wastes from a food distribution center at a US military
base. The system was intended to process over 3000 lbm per
day of waste to produce 70 kW of electricity and 88 kW of
thermal energy. Unexpected problems occurred with waste
pre-processing (forming briquettes required by the gasifier),
controlling sulfur content in the producer gas, and the unexpected additional personnel required to maintain operation.
However, they report important lessons learned that could
facilitate future successful implementation of similar technology [3].
Often, the desire to incorporate waste resources in a gasification system results in a first step of co-firing the waste
with a conventional energy feedstock. The Nuon Power facility in Buggenum, Netherlands, indicated successful results
with co-firing coal and 30% biomass in an integrated gasifier
combined cycle (IGCC) power plant [4]. Gaqa et al. report
synergistic effects of co-firing conventional fuels such as coal
with biomass residues (wood) due to the significantly higher
volatile content of the wood, which facilities the overall reaction [5]. Paskach and Reardon suggest the advantages of initial gasification of biomass prior to co-firing with coal in a
conventional combustor, as the gasification process affords
the ability to clean up the producer gas, removing components that could potentially harm the combustor, and therefore allow a greater range of initial feedstocks [6]. Most of
these efforts with biomass involve a relatively homogeneous
feedstock, unlike typical mixed municipal solid waste
(MSW).
Successful integration of wastes into a gasification wasteto-energy system, with or without co-firing, is likely to be
very dependent on understanding the waste supply, the technology used, and the needs of the industry or community being served. For example, sugar cane processing requires a
substantial amount of steam but produces a relatively well
characterized waste (bagasse fiber). Turn et al. present a case
study analysis of a biogas integrated combined cycle
(BIGCC) heat and power application to a typical sugar cane
plant in Hawaii. The case study results show that this plant
could produce a net 22 MW e of electrical power while supplying most, but not all, of the thermal steam demand for the
sugar processing [7]. Mochida et al. report experimental results of high temperature steam gasification of woody biomass residue for smaller scale distributed combined heat and
power generation in Japan, a need made all the more cogent
after the major earthquake experienced there in 2011. Based
on experimental data, they modeled both supplying syngas to
an absorption refrigeration unit and producing combined heat
and power with a diesel generator. Their analysis showed
overall economic feasibility with payback periods on the order 7 years, without accounting for reduced costs of waste
disposal [8]. In general, the gasification industry is working
to commercialize technology that will work with heterogeneous waste streams [9, 10, 11].
Colleges and universities are unique communities, often
88
like small independent towns, and often with smaller-scale

combined heat and power installations on-site. The waste
stream from colleges and universities may also be somewhat
unique relative to the overall average domestic MSW. Implementations of waste-to-energy using gasification on campuses have had notable successes and failures, but there is a
growing impetus to implement these technologies on campus
[12].
The objective of this study is to investigate the possibility
of using waste-to-energy via gasification at a college campus
(the US Naval Academy, Annapolis, MD) and suggest tools
to model the process. The campus is home to approximately
4400 students and 600 faculty. The campus is arranged with
most residential buildings segregated from the academic and
administrative buildings. All student meals are served in one
central food service cafeteria, allowing food service waste to
be easily segregated from the main trash dumpsters, which is
already done in practice. For this reason, this source of waste
was excluded from this present analysis. In this study, waste
was sampled and analyzed to determine any differences from
average domestic US MSW, and its potential performance
was simulated in a gasification regime. A surrogate waste
recipe was suggested and tested to determine its utility in
representing average campus MSW. Institutional data on total
waste production and energy consumption are used to gauge
possible energy offsets with a waste-to-energy strategy.
Methods
In order to experimentally characterize waste from the
institution, the authors considered how to obtain a representative sample of waste. Waste produced on campus is expected
to vary by location and by time of year. Location-based differences occur because different types of waste are produced
depending on proximity to certain functions such as food
service, the post office, residential verses academic buildings,
etc. Time-of-year variations can occur due to events such as
student move-in, move-out, high numbers of visitors during
graduation events, or students and faculty being gone for holidays. To minimize geographical variations, waste was collected from several strategic locations around campus, as
described below. To minimize time-of-year variations, the
authors collected waste during a typical week of the academic
year, not during the relatively short periods that would tend to
change the composition and amount of waste.
Waste was collected from areas around the main dormitory buildings and from various trash bins around one of the
main academic buildings. The former waste samples were
designed residential and the latter as academic/administrative, given the similarities between activities in
the academic and administrative spaces, and the co-locations
of many of these functions. The residential waste was collected from four different trash dumping locations around the
eight buildings that comprise the central dormitory. The locations were chosen based on knowledge of different activities
that occur in parts of the dormitory (one site near a snack
center, one site near a post office, etc.) such that a representative average of the dormitory could be established.
To capture waste over a wide area, waste was collected
VOLUME 42, NO. 2
MAY 2016
from large bags used by janitorial staff which enclose waste

from a multitude of smaller collection sites in the area. In this
way, several large bags represented the trash produced over a
wide area near that dumpster location. The total amount of
waste collected was approximately 54 kg (120 lbf). The sorting was conducted in a controlled environment indoors to
avoid environmental difficulties such as rain or wind. Each
sample of waste from each location was sorted into categories
and weighed on a laboratory scale to establish the weight
contribution of each category to the overall waste stream. The
sampling categories were chosen based on ASTM D5231,
though some of the categories were combined for convenience. Yard waste was not specifically recorded due to insignificant quantities. Materials that were too small or too difficult to separate (e.g. dust, coffee grounds, leaf fragments, and
small pieces from other categories) were classified as other
and represent a source of sampling error.
Two methods were utilized to determine the overall chemical analysis of the institutional waste stream and compare it
to data for average US domestic MSW. The first method relied upon the sampled composition of the waste stream (e.g.
amount of paper, plastics, textiles, etc.) as described above.
Published data from Ref. [13] on the typical chemical analyses of each component were then used to construct a composite waste analysis. For example, the total moisture in the
overall waste stream can be calculated by adding up the
amount of moisture that should exist in each component of
the waste stream. This approach relies on published analyses
for moisture, ash, elemental composition, and energy content
for a variety of components (see Ref. [13]).
The second method relied on direct chemical analysis of a
smaller sample size of the waste. In this approach, all collected waste from each location was combined and thoroughly
mixed into a cone. This cone was quartered to reduce the
sample size to a 5 US gallon (3.79 L) size suitable for testing.
The coning and quartering approach is described in Ref. [14],
and approximately followed the ASTM D5231 specification.
Multiple 5 gallon samples were constructed for each location
(academic/administrative and residential), excluding the metals and glass fraction, and were then analyzed for proximate,
ultimate, and energy content analysis by ALS Environmental
(Tuscon, AZ) using ASTM D7582, ASTM D5373, ASTM
D4239, and ASTM D5865 specifications with oxygen measured as a difference. While direct measurement is usually
desirable, the possible disadvantage of this approach is the
additional sampling error introduced by the coning and quartering approach required to reduce sample size.
Two surrogate waste samples were proposed and tested
based on categorical analysis of the actual campus MSW. The
surrogate waste samples contained three components: waste
shredded office paper, waste shredded plastic bottles, and a
standard commercial moist dog food, the latter included to
represent the organic fraction and to supply the required
moisture. The two surrogate mixtures were analyzed using
the same protocols as above by ALS Environmental and
compared with actual measured sampled of campus waste.
The approach to modeling the stoichiometry and energy
conversion of the waste fuel, using the proximate and ultimate analysis, is similar to the approach used by Vaezi et al.
and Edwards and Simon [15, 16]. The overall stoichiometry

of the modeled waste is determined according to the general
reaction:
CH a SbOc N d Sie Fe f Alg Cah ... + H 2O H 2O + ( O2 + 3.76 N 2 )
CO2 + H O H 2O + SO
SO2 + O O2 + N N 2 + ASH
CO
2
[1]
where are stoichiometric coefficients for reactants and

are the stoichiometric coefficients for the products. The mineral content in the waste fuel includes the elements Si, Al, Ti,
Fe, Ca, Mg, K, Na, and P. The ASH in the product stream is
modeled with the following mixture of ten components: SIO 2
(46%), Al2 O 3 (13%), TiO 2 (1%), Fe 2 O 3 (20%), CaO (11%),
MgO (1.5%), K 2 O (1%), Na 2 O (5%), SO 3 (1%), and P 2 O 5
(0.5%) based on Ref. [17].
Chemical reaction and conversion of the waste was modeled using a first-law energy balance for a flow reactor, given
by:
R
0 = m i hi (TR ) m i hi (TP ) Q
=
0 m FUEL hFUEL + m OX hOX + m MOD hMOD + m MG hMG , R m GAS hGAS m MG hMG , P m ASH hASH Q
[2]
where R denotes reactants, P denotes products, h is the massis the rate of heat transfer, and the subbased enthalpy,
scripts are defined as follows: FUEL (waste fuel), OX (oxidizer, usually air or O 2 ), MOD (moderator, often steam), GAS
(output gas flow from gasifier), MG (metals and glass), ASH
(ash). In the energy balance above, inputs to the reactor have
positive sign and outputs have negative sign. The relative
mass flow rates in the above balance can be related back to
the stoichiometry and proximate/ultimate analyses, and the
enthalpy of the fuel related to the higher heating value of the
waste fuel.
Combustion conditions were modeled using lean operating conditions and by fixing the output stoichiometry assuming major combustion products (CO 2 , H 2 O, SO 2 , O 2 , N 2 )
and ash. The energy balance above was then solved for the
resulting product temperature in the reactor. Gasification was
modeled using chemical equilibrium with constrained enthalpy and pressure. The sensible and latent enthalpies of ash,
metals, and glass were included, making the energy balance
implicit in temperature. Solution was obtained by iterating
using a bounded bisection numerical approach to ensure convergence. Chemical equilibrium was calculated using
MATLAB with Cantera, an open-source suite of functions
designed to facilitate thermochemical calculations. A GRI-3.0
species set was utilized for property data, modified to include
sulfur species important in waste gasification, with a total of
57 species including hydrocarbons C1 C3, and common
species from the elements H, O, N, Ar, and S [18].
The metals/glass and ash sensible and latent enthalpy
terms are estimated using characteristic values available in
the literature (e.g. Ref [13]):
89
relative frequency of garbage truck pickups for all waste receptacles. Each receptacle, and the building it served, was
classified as either residential or academic/administrative.
The relative frequency of trash pickup and the size of each
receptacle was then used to estimate the distribution of waste
between these two groups. The bottom row of Table 1 show
that the split was nearly uniform, with 51% of waste estimated to come from academic/administrative buildings and the
remaining 49% from residential buildings.
Once the institutional waste was sampled and categorized,
smaller composite samples were constructed and analyzed
per the ASTM specifications outlined in the Methods section
and in Table 2. Table 2 compares the academic/administrative, residential, and combined waste data from
a proximate, ultimate, and energy content analysis. The
Comp column gives calculated or reported values based on
adding up the contributions of each component using published component data from Ref. [13]. The Direct Meas.
column gives values based on direct chemical analysis of a
representative sample of the waste. The error bounds shown
in the table for the measured quantities are based on the
standard deviation of measurements of multiple samples
). While the measured residential samples showed a
(
relatively large range of values, in all cases the moisture from
institutional waste was lower than the EPA average MSW on
the order of 5% - 10% absolute (5 10 kg less water per 100
kg as-received waste) as shown in Figure 3. The lower moisture content could facilitate many energy conversion processes and is due to the higher fractions of dry components (paper
and plastics) in the institutional waste. While ash from institutional waste showed as much as a 1.5% increase absolute,
the elemental analyses (C, H, O, N, S) are comparable between institutional waste and average domestic MSW.
The melting point of ash was assumed to occur at 800C.
Results and Discussion

TABLE 1 compares the sampled categories of waste from the
academic areas and residential areas with the composition of
average domestic MSW reported by EPA in 2012 for the
United States [19]. These data are shown graphically in Figure 1 as differences from the average domestic MSW values;
positive bars in this figure indicate greater quantity of that
component than as reported by EPA for domestic MSW, and
negative bars indicate less. Institutional waste showed much
greater paper and plastic and less textiles, wood, and metals
than the domestic MSW. These differences are not surprising
when considering the activities of an academic institution,
which involve significant paper use. Student lifestyles often
involve late hours, snacking and drinking outside of meal
times, which could explain the greater fraction of plastic
waste and similar fraction of food waste, despite the main
food waste stream from the central cafeteria being excluded
from these samples.
Note that no rubber/leather or yard waste was recorded in
the samples. If present, the fraction of rubber/leather in the
waste samples would have been counted as textile waste, but
this fraction would be very small. Yard and leaf waste is seasonal, and over the course of this study, no significant yard or
leaf waste was measured.
To determine the relative contributions of academic/administrative buildings and residential buildings to the
overall waste flow, records were gathered to determine the
TABLE 1
Comparison of Academic and Residential Waste with Domestic US MSW
Component
EPA Reported MSW 2012
Academic/Admin
Residential
Combined
Paper
14.8%
38.0%
31.0%
34.5%
Plastics
17.6%
34.1%
32.2%
33.2%
Rubber/Leather
3.8%
0%
0%
0%
Textiles
7.4%
0%
1.0%
0.5%
Wood
8.2%
0%
0.1%
0.1%
Food/Garbage
21.1%
21.1%
22.1%
21.6%
Yard/Leaves
8.8%
0%
0%
0%
Metals
9.0%
1.6%
3.1%
2.4%
Glass
5.1%
3.6%
7.0%
5.2%
Other
4.4%
1.6%
3.5%
2.5%
51%
49%
Fraction
90
VOLUME 42, NO. 2
MAY 2016
FIGURE 1
Differences among institutional waste from academic/administrative building and residential buildings shown as an absolute
difference (%) from US domestic MSW as reported by EPA in 2012.
TABLE 2
Proximate, ultimate, and energy content analysis based on published values and direct measurement of waste samples.
EPA
Component
Academic/Admin
Residential
Combined
Comp
Comp
Direct Meas.
Comp
Direct Meas.
Comp
Direct Meas.
Moist. (AR wt%)
32.2%
20.8%
22.6% 2.2%
22.4%
26.9% 5.2%
21.6%
21.0% - 28.4%
Ash (AR wt%)
5.3%
6.8%
6.4% 0.1%
6.8%
4.9% 2.2%
6.8%
4.5% - 6.8%
C (MF wt%)
51.7%
51.1%
51.5% 0.6%
51.4%
52.9% 3.4%
51.2%
50.2% - 54.2%
H (MF wt%)
6.6%
6.5%
6.6% 0.1%
6.5%
6.4% 0.7%
6.5%
6.1% - 6.9%
O (MF wt%)
32.6%
33.3%
32.7% 0.3%
32.7%
33.6% 3.1%
33.0%
31.5% - 34.8%
N (MF wt%)
1.2%
0.4%
0.8% 0.1%
0.4%
0.4% 0.1%
0.4%
0.5% - 0.7%
S (MF wt%)
0.2%
0.1%
0.1% 0.03%
0.1%
0.1% 0.03%
0.1%
0.1%
HHV (AR)
16.0
19.7
17.9 1.1
19.5
16.9 1.9
19.6
15.9 18.9
HHV (MF)
23.6
24.9
23.1 0.8
25.2
23.2 2.5
25.0
21.5 24.8
HHV (MAF)
25.6
27.3
25.2 0.7
27.6
24.8 2.3
27.4
23.5 26.5
The energy content analyses (as given by the higher heating value) show consistent and expected trends, although the
measured samples give a consistently lower heating value
than those based on the measured components. The reason for
this is unclear, but the discrepancy is small (0.2 0.9 MJ/kg)
when accounting for the given error bound. The higher heating values reported here are based on removing the noncombustible metals and glass. Table 1 show that metals and
glass compose 14.1% of US domestic MSW and were measured at 7.6% of institutional waste. If metals and glass are
included, each HHV would be reduced by that fraction: The
higher heating value of US domestic MSW would reduce
from 16.0 to 13.7 MJ/kg and that of the component-based

institutional waste from 19.6 to 18.1 MJ/kg. Because the average waste reported by EPA cannot be measured for direct
comparison, the best comparison is with the values determined by component sampling, which indicate a significant
energy content advantage for institutional waste. This higher
energy content is due to higher plastic and lower moisture
content. The higher heating value on an as-received basis, but
excluding non-combustible metals and glass, is shown in
Figure 2.
Combustion was modeled in a constant pressure flow reactor at overall lean conditions characterized by the relative
91
FIGURE 2
Energy content difference (HHV AR) from domestic MSW.
FIGURE 4
Adiabatic flow reactor temperature with excess air ratio (relative air-fuel ratio) from 1 to 2, comparing US domestic MSW
as reported by EPA in 2012 with two estimates of institutional
waste based on component sampling and based on direct
sample measurement.
of plausible values for reactor temperature; note that in all

cases, however, the campus waste is predicted to produce a
higher reactor temperature than average domestic MSW. The
increase in temperature can be attributed to the lower moisture and higher energy content of campus waste, resulting
from greater paper and plastic composition.
Gasification of waste was modeled in a similar constant
pressure reactor at rich conditions using chemical equilibrium. The adiabatic gasifier temperature is shown in Figure 5
on a typical footprint diagram with the molecular oxygen-tocarbon ratio on the horizontal axis and the total water-to-
FIGURE 3
Moisture difference (absolute %) from domestic MSW.
air-fuel ratio, . A value of unity indicates stoichiometric
operation; values greater than unity indicate fuel-lean conditions, and a value of 2 implies twice the amount of air required for stoichiometric combustion. The first law energy
balance for this reactor, given in the Methods section, was
solved for the resulting reactor temperature at adiabatic conditions using three different waste composition values: US
domestic MSW as reported by EPA, campus waste as characterized by the relative amounts of each category, and campus
waste as directly measured. The results are shown in Figure
4. The two characterizations of campus waste provide a range
92
FIGURE 5
Gasification temperature.
VOLUME 42, NO. 2
MAY 2016
carbon ratio on the vertical axis (both molar ratios). The institutional waste based on component-sampling was used in the
model. This waste already contains moisture corresponding to
a water-to-carbon ratio of 0.34, indicated by the dashed horizontal line on the figure. Operation of the gasifier in a regime
requiring more moderation (greater water) would require the
addition of additional steam during gasification. The diagonal
line on the lower-left of the plot indicates the materialbalance limit; operation below this line has insufficient oxygen to oxidize all the carbon in the fuel. For greater explanation and other examples of the footprint diagrams, see Ref.
[17].
On the same type of diagram, Figure 6 shows the cold-gas
efficiency and calculated methane concentration in the output
gas based on equilibrium. As expected, the model shows that
methane is only produced in significant concentration for rich
reactions with high levels of steam moderator, conditions
allowing a low gasifier temperature. These conditions also
produce the highest cold-gas efficiency. The model indicates
that for molar oxygen-carbon ratios below 0.3, cold-gas efficiencies of 80-90% are possible.
Two surrogate waste samples were proposed in order to
provide a standard approach to experimental testing of campus waste. The largest components of campus waste are paper, plastic, and organic food waste, together composing
nearly 90% of the sampled waste. Therefore, a threecomponent surrogate was proposed. The criteria for choosing
the three components were that they be readily available at
low cost and of relatively standard composition. Paper was
included as shredded office paper, which is often available in
great quantities from office environments. Plastics were included as shredded plastic bottles (such as water bottles and
sports drink bottles) from recycling containers present all
over campus. Usually, these bottles are empty, containing
few non-plastic elements, and are easily shredded using a
commercial shredder. The organic food waste component was
represented by a commercial moist-style canned dog food.
FIGURE 6
Gasification cold gas efficiency and equilibrium methane output concentration.
The analysis of the dog food, available on the label and typical of different brands, indicated crude protein, fat, fiber, and
moisture content of >10.0%, >6.0%, <1.0%, and <75.0%,
respectively. By adjusting the relative fractions of these components, the authors hypothesized that a reasonable surrogate
mixture could be developed that would facilitate future experimental investigation.
Surrogate sample #1 was created by approximating the
composition of each surrogate component (using Ref. [13])
and matching to the experimentally determined campus
MSW analysis. The result was that this first surrogate blend
included approximately equal fractions of paper, plastic, and
organics (34.7%, 32.9%, and 32.4% respectively). Surrogate
sample #2 included the same relative fractions of paper and
plastic, but with reduced organics from approximately 30% to
20%, which was predicted to result in a waste surrogate moisture content near the low end of the actual measured range.
Each surrogate mixture and the results comparing analyses of
the mixtures to the actual waste are given below in Table 3.
In Table 3, the predictions for proximate, ultimate, and
energy content analysis for each of the samples, based on
adding up the contributions of each component, are given in
the Comp column. The actual measured values of a representative sample of the surrogate waste mixture are compared
with actual campus waste under the Direct Meas. columns.
The predicted results based on adding up contributions of
each component compared well with the direct analysis of the
surrogate mixtures expect for the ash quantity. As expected,
the surrogate sample #1 has moisture content at the top of the
range of the actual campus waste while surrogate sample #2
has moisture content near the bottom of this range. The two
surrogate samples also result in energy content values that
span the range of actual campus waste.
Figures 7 and 8 show the comparison among the two surrogate samples and the actual measured values of campus
waste graphically. These two figures are multi-axis radar
type plots which enable comparison of the two surrogates
with the actual measured values along multiple criteria (ash,
moisture, and energy content on one figure and C, H, O, N, S
elemental composition on the other). The two surrogate samples are indicated by the solid and dashed lines on the figures;
the actual measured values of campus waste (the targets)
are indicated by the bold portions of each axis. Ideally, the
surrogate samples would intersect each axis at some portion
of the bold line. Figure 7 shows that both surrogate samples
span the moisture and energy content of actual campus waste
well; however, both samples under-represent the amount of
ash measured in the actual waste. On the elemental analysis
shown in sample #1 does a better job of representing the elemental content, but both surrogates do a reasonable job representing the actual waste. Fine adjustment of the samples
could improve the representation further. The discrepancy
with ash could be resolved by the addition of a fourth, high
ash component. Alternatively, the effect of reduced ash from
6% to 2% may be an acceptable discrepancy given the experimental convenience of the three component surrogate, and a
low ash testing mixture may even be desirable to facilitate
experimental combustion or gasification studies.
93
TABLE 1
Comparison of the two surrogate samples to measurements of Proximate, ultimate,
and energy content analysis of actual campus waste.
Surrogate Sample 1
Surrogate Sample 2
Paper
34.7%
40.3%
Plastics
32.9%
38.3%
Organics
32.4%
21.4%
Actual Combined Waste
Comp
Direct Meas.
Comp
Direct Meas.
Comp
Direct Meas.
Moist. (AR wt%)
27.4%
29.2%
20.2%
22.3%
21.6%
21.0% - 28.4%
Ash (AR wt%)
6.7%
2.2%
7.0%
2.6%
6.8%
4.5% - 6.8%
C (MF wt%)
51.4%
52.7%
51.7%
56.5%
51.2%
50.2% - 54.2%
H (MF wt%)
6.5%
6.2%
6.7%
7.7%
6.5%
6.1% - 6.9%
O (MF wt%)
32.2%
37.2%
32.5%
31.9%
33.0%
31.5% - 34.8%
N (MF wt%)
0.5%
0.6%
0.4%
0.4%
0.4%
0.5% - 0.7%
S (MF wt%)
0.2%
0.2%
0.1%
0.1%
0.1%
0.1%
HHV (AR)
18.4
16.6
20.4
19.7
19.6
15.9 18.9
HHV (MF)
25.3
23.4
25.6
25.3
25.0
21.5 24.8
HHV (MAF)
27.8
24.2
28.0
26.2
27.4
23.5 26.5
Available institutional waste collection records were reviewed for 10 months between April 2012 and January 2013
to estimate the total quantity of waste produced. Food waste
from the central cafeteria was excluded from the accounting,

as indicated previously. The data were sufficient to capture
monthly variations in waste produced. For example, reduc-
FIGURE 7
Comparison of ash content, moisture content, and HHV
among the two surrogate samples (solid and dashed lines) and
the actual measured campus waste values (indicated as bold
lines on each axis).
FIGURE 8
Comparison of elemental content (C, H, O, N, S) among the
two surrogate samples (solid and dashed lines) and the actual
measured campus waste values
(indicated as bold lines on each axis).
94
VOLUME 42, NO. 2
MAY 2016
tions in waste may be observed during the summer months

and over winter holidays when fewer students, faculty, and
staff are present. A large increase in waste is observed during
May when many visitors arrive for graduation events over an
extended period. Table 3 indicates that the average annual
waste production was estimated at 1.88 x 106 kg/year (approximately 2100 tons). Average haul-away disposal costs
during this period were $52/ton ($57/metric ton), with total
annual cost of over $100,000.
Using an energy content value estimated previously as
18.1 MJ/kg (including all moisture, ash, metals, and glass),
the total gross energy content of this waste stream is 33.9 x
106 MJ/year. In a gasification regime with cold-gas efficiency
of 85%, the remaining net energy content of the produced gas
would be 28.8 x 106 MJ/year. Using this gas in an electrical
generator (e.g. diesel or gas turbine based electrical generation system) with overall gas-to-electricity conversion efficiency of 30% would result in 8.65 x 106 MJ/year of electrical energy (2.40 x 106 kW-hr/year).
To compare these data with facility energy use, institutional energy data were reviewed for the period of September
2013 September 2014. The available data indicate energy
use by the categories shown in Table 4. The steam is produced mainly by combustion of natural gas, and the chilled
water produced by electricity-driven refrigeration cycles.
Thus, electricity is the main energy carrier for the institution,
but natural gas usage for direct heating (0.0482 x 106 MJ/yr)
or for hot water production mainly for building heat (0.246 x
106 MJ/yr) is also a significant energy carrier and cost (8.5 x
106 $/yr). A waste-gasification-to-electricity system as modeled here would allow offset of approximately 2.5% of institutional electricity use with a corresponding savings of nearly
$350,000 annually. From a different perspective, this quantity
of electricity would be more than sufficient to supply power
for all faculty desktop computers (assuming usage of 3 kWhr/day for 600 faculty) and all student laptops (assuming usage of 0.50 kW-hr/day for 4400 students). However, if a
waste-to-heat strategy was used by direct combustion, the
amount of waste would be more than sufficient to offset the
natural gas used for creating hot water or direct heating at
substantial cost savings. A combined strategy, for example in
which electricity was produced and waste heat recovered to
offset heating costs could result in maximum savings and
offset of conventional fuel used. Note that a complete economic analysis was not part of this study, and a full financial
analysis would necessarily involve a separate effort and a full
accounting of upfront and ongoing costs.
CONCLUSIONS
To investigate the potential of waste-to-energy conversion
at the US Naval Academy, waste was sampled from multiple
locations in academic and residential buildings. Bulk food
waste from the central cafeteria resulting from food spoilage,
preparation, and post-meal-waste was not considered in this
study. This waste stream is already isolated from other waste
streams. Waste sampling showed subtle differences between
the academic and residential buildings, mostly in the greater
paper content from academic buildings. When comparing the
sample waste with that reported by the US EPA for domestic
MSW, institutional waste showed significantly higher levels
of paper and plastics, with significantly lower fractions of
textiles, wood, and metals. Despite the fact that food waste
from the central cafeteria was not considered in the sampling,
food waste was comparable as that reported by EPA for
MSW.
Proximate, ultimate, and energy content analysis were
TABLE 4
Estimated total institutional waste, energy content, gasification output,
and electrical generation capacity on a monthly and annual basis.
Estimated Waste
Facility Waste
0.156
x 106 kg/month
1.88
x 106 kg/year
33.9
x 106 MJ/year
Gross Energy Content

HHV (AR, with metals and glass)
Gross E waste
18.1 MJ/kg
2.82
x 106 MJ/month
Post Gasification with CGE = 85%

HHV (AR, with metals and glass)
Net E gas
15.4 MJ/kg
2.40
x 106 MJ/month
28.8
x 106 MJ/year
0.721
x 106 MJ/month
8.65
x 106 MJ/year
Electricity Generation with = 30%

Net W elec
95
TABLE 5
Estimated total institutional annual energy usage and cost for September 2013 September 2014.
Estimated Facility Energy Use
Cost
Electricity
x 10 kW-hr / year
x 10 MJ/yr
x 106 USD ($)
97.4
351
13.9
x 106 BTU/yr
Hot Water
233
0.246
7.83
Chilled Water
160
0.168
2.46
Natural Gas
45.6
0.0482
0.676
Propane
1.07
0.00113
0.118
Fuel Oil
23.1
0.0244
0.639
determined for institutional waste using two approaches. In

the first method, the analyses were based on published data
for the amount of each component measured in the waste
sample. This approach allowed direct comparison with the
average domestic MSW reported by EPA. In the second approach, waste samples were further sampled using standard
size reduction methods and directly chemically analyzed by
ASTM specifications. Chemical analyses were comparable
between the two approaches, although energy content by subsampling and direct measurement (the second approach)
yielded results slightly but consistently lower than those predicted by the first approach. Overall, the analyses showed
that institutional waste has significantly lower moisture content (5% - 10% absolute), slightly higher ash content (1%-2%
absolute), and likely higher energy content (1-3 MJ/kg, approximately 10%). The differences are due primarily to the
greater paper (a dry component) and plastic (highly energetic
component) in institutional waste.
Gasification was modeled using an adiabatic steady-flow
reactor that accounted for approximate composition of the
ash, the sensible enthalpy of metals, glass, and ash, and the
enthalpy of fusion of the ash. Chemical equilibrium analysis
with constant enthalpy and pressure constraints indicated that
with some additional steam input for moderation, cold gas
efficiencies of ~85% are possible with operating temperatures
of 900-1000 K.
Two simplified surrogate waste samples were created to
represent actual campus waste using three readily available
and low cost components: shredded office paper, shredded
drink bottles, and commercially available moist dog food.
One sample had approximately equal quantities of these
components, the other contained approximately a composition of 40%, 40%, and 20%, respectively, of the three components. Overall, the two samples captured the actual chemical analysis of campus waste well, giving moisture and energy content values that spanned the actual measured ranges
and well-representing the elemental analysis. Both proposed
surrogate mixtures underestimated ash, but this may have
96
only modest experimental impact and could be a benefit for

laboratory work.
The quantity of generated waste was estimated using institutional records at 1.88 million kg per year with haul-away
disposal costs of over $100,000. Based on estimated energy
content, this amount of waste is equivalent to 33.9 million MJ
per year (based on higher heating value). With a cold gas
efficiency of 85%, this waste could be converted to 28.8 million MJ per year of producer gas, and with a final gas-toelectricity conversion efficiency of 30%, would result in 8.65
million MJ per year (2.4 million kW-hr) of electricity generation. This amount of electricity would be sufficient to offset
approximately 2.5% of annual institutional electricity use.
With a combined heat and power approach, it is likely that a
majority of the thermal energy requirements could also be
met while still providing some of this electrical power offset.
The analysis here does not consider the reduction in disposal
costs associated with using waste in an on-campus waste-toenergy process.
Acronyms
EPA
MSW
CGE
IGCC
BIGCC
MW e
MG
MAF
ASH%
W%
C%
E%
MW i
MG%
a
b
Environmental Protection Agency

municipal solid waste
cold gas efficiency
integrated gasification combined cycle
biogas integrated gasification combined cycle
megawatt electricity
metals and glass
moisture and ash free
percentage ash (weight basis)
percentage moisture (weight basis)
percentage carbon (weight basis)
percentage of any element (C, H, O N, S)
(weight basis)
molecular weight
percentage metals and glass (weight basis)
empirical coefficient for H
empirical coefficient for S
VOLUME 42, NO. 2
MAY 2016
c
d
empirical coefficient for O

empirical coefficient for N
relative air-fuel ratio
monoatomic oxygen in ash
8.
stoichiometric coefficient-reactants
R
P
hi
stoichiometric coefficient-reactants
reactants
products
enthalpy of species i
9.
10.
ACKNOWLEDGEMENTS
The authors would like to acknowledge the helpful guidance and advice from Mr. Michael Partyka, Ms. Eileen Cox,
Ms. Louise Becnel, Mr. Kelly Delikat. This study was partially supported by the DARPA Service Academies Challenge
Award HR0011411994.
11.
12.
REFERENCES
1.
2.
3.
4.
5.
6.
7.
Pytlar, T.S. Status of Existing Biomass Gasification and

Pyrolysis Facilities in North America, Proceedings of the
18th Annual North American Waste-to-Energy Conference, MAWTEC18, Orlando, FL, USA, May 11-13,
2010.
Chaivichitmalakul, S., Simonton, J.L. A Comparative
Feasibility Study of Biomass Gasification Between Developed and Developing Countries: An Economic Perspective on Waste and Residue, 30th Annual Conference
of American Society for Engineering Management,
Springfield, MI, USA, October 14-17, 2009.
Davis, J., Gelman, R., Tomberlin, G., Bain, R. Hawaii
and Guam Energy Improvement Technology Demonstration Project, National Renewable Energy Laboratory
Technical Report NREL/TP-7A40-60868, March 2014.
Available online at http://www.nrel.gov/docs/fy14
osti/60868.pdf, accessed January 2015.
Wolters, C. and Kanaar, M. Update of Nuon Power
Buggenum Plant Performance and Fuel Flexibility,
presentation at 2005 International Conference on Clean
Coal Technologies for our Future, CCT2005, CCT conference, Sardinia, Italy, October, 2005.
Gaqa, S., Mamphweli, S., Katwire, D., Meyer, E. Synergistic Evaluation of the Biomass/Coal Blends for CoGasification Purposes, International Journal of Energy
and Environment, Volume 5, No. 2, pp. 251-256, 2014.
Paskach, T.J. and Reardon, J.P. Gasification: Eliminating
Risks Associated with Co-Firing Biomass, Proceedings
of the ASME 2010 Power Conference, Chicago, IL, July
13-15, 2010.
Turn, S.Q., Bain, R.L., and Kinoshita, C.M. Biomass
Gasification for Combined Heat and Power in the Cane
Sugar Industry, Int. Sugar Journal, Volume 104, No.
13.
14.
15.
16.
17.
18.
19.
1242, pp. 268-273, 2002.

Mochida, S., Abe, T., Yasuda, T., Gupta, A.K. Combined Heat and Power System with Advanced Gasification Technology for Biomass Wastes, in Cleaner Combustion and Sustainable World, Qi, H. and Zhao, B. Eds.
DOI 10.1007/978-3-642-30445-3_111, Springer-Verlag,
Berlin, 2013.
Beaty, P. Gasifiers that Work, presentation at 2012 Gasification Technologies Workshop, Kingsport, TN, USA,
April 25-26, 2012.
Ridjan, I., Mathiesen, B.V., and Connolly, D. A Review
of Biomass Gasification Technologies in Denmark and
Sweden, Department of Development and Planning,
Aalborg University, ISBN 978-87-91404-54-2, 2013.
Patel, J. and Salo, K. Carbona Biomass Gasification
Technology, presentation at the TAPPI International
Conference on Renewable Energy, Atlanta, GA, USA,
May 10-11, 2007.
Bain, R.L. USA Biomass Gasification Status, National
Renewable Energy Laboratory, IEA Bioenergy Task 33,
April 18, 2012. Available online at http://www.ieatask
33.org/app/webroot/files/file/2012/USA.pdf,
accessed
January 2015.
Niessen, W.R. Combustion and Incineration Processes:
Applications in Environmental Engineering, 4th Ed.,
CRC Press, 2010.
Gerlach, R.W., Dobb D.E., Raab G.A., Nocerino J.M.
Gy sampling theory in environmental studies. 1. Assessing soil splitting protocols, J. Chemometrics, Volume 16, pp. 321-328, 2002.
Vaezi, M., Passandideh-Fard, M., Moghiman, M.,
Charmchi, M. Modeling Biomass Gasification: A New
Approach to Utilize Renewable Sources of Energy, Proceedings of the 2008 ASME International Mechanical
Engineering Congress and Exhibition, IMECE2008, Boston, MA, USA, October 31-November 6, 2008.
Edwards, C.F. and Simon, A.J. ME370B Energy Systems
II: Modeling and Advanced Concepts, Thermo Calculations with Coal and Biomass, course notes for ME370B
course, Stanford University, Winter 2008.
Higman, C. and van der Burgt, M. Gasification, 2nd Ed.,
Elsevier, 2008.
Goodwin, D.G. An Open-Source, Extensible Software
Suite for CVD Process Simulation. Chemical Vapor
Deposition XVI and EUROCVD 14, ECS Proceedings
2003, 2003-08, M. Allendorf, F. Maury, and F.
Teyssandier, Eds., The Electrochemical Society 2003,
pp. 155-162.
Municipal Solid Waste Generation, Recycling, and Disposal in the United States: Facts and Figures for 2012,
U.S. Environmental Protection Agency, Solid Waste and
Emergency Response (5306P), EPA-530-F-14-001, February 2014. Available online at http://www.epa.gov/
solidwaste/nonhaz/municipal/msw99.htm, accessed January 2015.
97
COMPOSTING OF RUMEN CONTENT WASTE USING

ANAEROBIC-ANOXIC-OXIC (A2/O) SYSTEM
Rhenny Ratnawati*1,3, Yulinah Trihadiningrum1, Sri Rachmania Juliastuti2
1
Department of Environmental Engineering, Institut Teknologi Sepuluh Nopember, Surabaya, Indonesia

2
Department of Chemical Engineering, Institut Teknologi Sepuluh Nopember, Surabaya, Indonesia
3
Department of Environmental Engineering, Universitas PGRI Adi Buana, Surabaya, Indonesia
rhenny11@mhs.enviro.its.ac.id
ABSTRACT
Rumen content waste is generally dried and dumped into municipal landfill by the slaughterhouse (SH) in Indonesia. A composting process using anaerobic-anoxic-oxic (A2/O) method
was selected for treating the SH solid waste for its advantage in nitrogen removal. The objectives of this study were to investigate optimum each composting duration, and effect of dolomite
addition for alkalinity control. Six standing reactors of 60 L capacity were used in this
experinment. Fifty kg of rumen content was placed in each reactor. The A2/O configuration
periods of 25-10-15, 15-10-25, and 30-10-10 days were applied in this research. The
composting process was conducted for 50 days. Aerobic condition was maintained using aeration rate of 0.5 L (kg-1 dry weight solid waste) min-1. Anoxic condition was established by 60 minute aeration rate of 0.28 L (kg-1 dry weight solid waste) min-1. The composting process was
conducted for 50 days. This research showed that the optimum anaerobic-anoxic-oxic time period for composting the rumen content waste was 25-10-15 days. In this reactor, the moisture
content was decreased from 83% to 72%. Temperature range was 29-34C, pH values slighty
fluctuated from 7.57 to 8.70. Inorganic nitrogen concentration decreased from 0.29 to 0.07%,
and organic nitrogen concentration decreased from 2.04 to 1.19%. Dolomite addition did not
give (P value >, = 0.05). This research proved that A2O method is appropriate for treating solid
waste with high nutrient contents.
Keywords: Anaerobic-anoxic-oxic, composting, rumen content waste
INTRODUCTION
Rumen content is one of the slaughterhouse (SH) solid
waste materials. This waste is rich in organic matter, and
nutrient such as nitrogen and phosphorus [1]. It also contains
various microorganisms such as phycomycetes fungi, protozoa, and bacteria which are probably pathogenic [2]. In
developing countries such as Indonesia, the SH waste is
usually disposed of in landfill or burried. These disposal
methods cause not only problems related to odour, flies and
hygiene, but also introduce enteric pathogens and excess

nutrients into surface and ground water [3,4].
Composting using anaerobic or aerobic system has been
successfully used for disposal of dead carcasses [5-7] and
manure [8-10]. However, Roy et al. [1] reported that only
2.70% of organic matter and 3.81% of nitrogen could be
degraded during 90 days of anaerobic composting of the
rumen content waste. This condition was probably related to
the low of C/N ratio of rumen content waste due to the high
nitrogen content [11]. In addition, aerobic composting con-
____________________________________________________
98
VOLUME 42, NO. 2
MAY 2016
serves more nitrogen content than in anaerobic process [1],

whereas large amount of NH 3 emission during anaerobic
decomposition could reduce the quality of compost as an
organic fertilizer and also creates severe odor problems [12].
Therefore, an alternative composting system for rumen content waste treatment using anaerobic-anoxic-oxic (A2/O)
methods, was proposed in this work.
The A2/O composting method is adopted from a
biological nutrient removal process, which is widely used in
wastewater treatment system. This treatment involves three
distinct stages, namely: anaerobic, anoxic and oxic phase.
Organic matter is transformed to methane and ammonia
under anaerobic phase. Ammonia and the remaining part of
organic matter, together with sulfide compounds, are
completely oxidized by nitrification, during the oxic phase,
and denitrification during the anoxic phase processes [13].
Peng et al. [14] used A2/O method for treating wastewater
with high nutrient contents with nitrogen removal efficiency
of 90%. Kargi and Uygur [15] used anaerobic, anoxic, and
oxic phases in removing of C, N, and P. The removal
efficiencies of NH 3 -N, NH 4 -N, and PO 4 -P from synthetic
wastewater were 89, 94, dan 99% respectively.
Iqbal et al. [16] compared three conditions os MSW
composting processes, i.e. oxic, oxic-anaerobic, and
anaerobic for 91 days. The results showed that maturity of
the compost product was faster in oxic condition than in
other methods. The nutrient content of the compost product
in the aerobic reactor was higher than those in aerobicanaerobic and anaerobic reactors in consecutive order. The
objectives of this study were to investigate the optimum anaerobic-anoxic-oxic phase duration, and the effect of dolomite addition for alkalinity control during the composting
process.
SH, which was located in Surabaya City, Indonesia. The

samples were air-dried for 2 days and stored in the laboratory at room temperature before used. Rumen content waste
were: pH value of 8, organic carbon of 14.05%, N-total of
1.23%, and moisture content of 79.61%.
Experimental design
Six experimental conditions were tested in duplicate during 50 days using laboratory-scale reactors. The experimental design comprised three anaerobic-anoxic-oxic configuration periods (25-10-15, 15-10-25, and 30-10-10 days)
and two dolomite doses (without dolomite addition and with
2% dolomite addition). Detailed experiment condition is
given in Table 1. The aerobic phase was maintained by aeration rate of 0.5 L min-1 kg-1, whereas anoxic condition was
made by aeration rate of 0.28 L min-1 kg-1 for 60 minutes
day-1. This aeration rate was applied according to Bari [17],
Tchobanoglous et al. [18], Tanpanich et al. [19], Saithep et
al. [20], Bernal et al. [21], Guo et al. [22], using perforated
pipe of 1 inch diameter for equal air distribution in this
reactor. The experiments were operated continuously at
normal room temperature.
Standing reactors of 60 L volume capacity were prepared
for this study. Design of composting reactor is shown in
Figure 1. About 50 kg of rumen content waste was placed in
these reactors. Each reactor was equipped with a perforated
pipe to flow air into the reactor, and a ventilated pipe which
was connected with gases adsorbent unit at the top of reactor. The bottom of each reactor contained 20 cm thick gravel
as leachate drainage layer. Air flow through the composting
reactors was provided by electric air pump (Swan Air Compressor Model SV212, Germany). The air flow was measured with air flow meter to ensure that it was maintained at
the required level.
MATERIALS AND METHODS
Sample and data analysis

Rumen content waste
Samples from each reactor was collected at day 0, 7, 14,
22, 27, 34, 41, and 50 for temperature, pH, moisture content,
Fresh cattle rumen content was collected from Pegirians
TABLE 1
Experiment condition
Reactor
A2/O configuration periods (days)
Dolomite doses (%)
R1
25-10-15
R2
25-10-15
R3
15-10-25
R4
15-10-25
R5
30-10-10
R6
30-10-10
COMPOSTING OF RUMEN CONTENT WASTE USING ANAEROBIC-ANOXIC-OXIC (A2/O) SYSTEM
99
40
Anaerobic
Temperature (oC)
38
Anoxic
Oxic
36
34
32
30
28
26
24
0
10
20
30
40
50
Time (days)
R1 (25-10-15 days; without dolomite)
R2 (25-10-15 days; with dolomite)
(a)
40
Anaerobic
Temperature (oC)
38
FIGURE 1
Design of composting reactor
organic carbon, and inorganic and organic nitrogen analyses.
Temperatures were measured using glass thermometer. The
pH value was measured after mixing the rumen content
waste with distilled water of a 1:10 ratio using a CyberScan
510 pH meter (Vernon Hills, USA). Organic nitrogen was
analyzed using Kjedahl nitrogen digestion and distilling apparatus of Gerhardt KBL 8 S (Bonn, Germany). Uncertainty
analysis was done by measuring standard deviation of each
measurement.
34
32
30
28
26
24
0
10
20
30
40
50
Time (days)
(b)
40
Temperature (oC)
Effects of different anaerobic-anoxic-oxic configuration

periods and dolomite doses during the composting process
on the evaluated indicators were tested using ANOVA at a
95% confidence limit.
Oxic
36
Anaerobic
38
Statistical analysis
Anoxic
Anoxic
Oxic
36
34
32
30
28
26
24
0
RESULTS AND DISCUSSION
20
30
40
50
Time (days)
Changes in temperature
Temperature values during composting are shown in
Figure 2. Initial temperature of all treatments was 301C,
and gradually increased but still remained below 40C.
Figure 2a shows that initial temperature in A2/O phase
periods 25-10-15 days (R1 and R2 reactors) was 312.12C
and 30.50.71C, respectively. During the anaerobic phase,
reactor without dolomite (R1 reactor) has a value range from
312.12C to 31.51.41C. Temperature increased to
343.54C at the anoxic phase. Then in the oxic phase, the
temperature decreased until the end of the research. The
final temperature in R1 reactor was 301.06C. Temperature
100
10
(c)
FIGURE 2
Temperature value during composting process
was observed at R2 reactor (with dolomite addition) during
both anaerobic and anoxic phases, by an average of 1C
higher or lower than the initial temperature. At the end of
composting, the final temperature in R2 reactor was
290.71C. In both R1 and R2 reactor showed the stable
temperature at the end of composting.
VOLUME 42, NO. 2
MAY 2016
10
Anoxic
Oxic
pH
6
0
10
20
30
40
50
Time (days)
(a)
10
Anaerobic
Anoxic
Oxic
6
0
10
20
40
30
50
Time (days)
Changes in pH value
(b)
10
Anaerobic
Anoxic
Oxic
pH
The pH variations during the composting process are

shown in Figure 3. Initial pH value of all reactors was an
average of 7.80.3. A2/O phase periods 25-10-15 days were
R1 (reactor without dolomite) dan R2 (reactor with
dolomite), during anaerobic phase pH values were from
7.20.23 to 8.00.7. (Figure 3a). The variation trends for R1
and R2 reactor were approximately similar, i.e., 7.90.64
and 7.70.17, respectively, on the end of anaerobic phase in
every A2/O configuration treatment. Mineralization of organic matter and conversion of organic nitrogen to ammonia
occurred during the initial phases of composting, causing a
rise in the pH value in the later phases [23]. Consequently,
the alkalinity of all treatments was increased and reached the
average value of 8.30.3 (R1) and 7.90.18 (R2) in anoxic
phase. In the end of oxic phase the alkalinity was increased
to 8.70.21 at R1, whereas the alkalinity for R2 was
decreased to 7.70.75.
Figure 3b showed that initial pH value in A2/O phase
periods 15-10-25 days were 8.01.41 and 8.10.06,
respectively, in R3 (reactor without dolomite) and R4
(reactor with dolomite). In R3, during anaerobic phase, pH
value decreased until the end of anaerobic phase with the pH
value of 7.90.65. Then pH value increased during anoxic
phase to become 8.30.21. During oxic phase, pH value in
R3 increased until the end of composting with the final pH
value of 9.00.71. Whereas in R4, similar with R3, during
anaerobic phase pH value decreased until the end of oxic
phase with the pH value of 7.60.06. Then gradually
increased during oxic phase until the end of composting. The
Anaerobic
pH
Figure 2b showed that initial temperature in A2/O phase

periods of 15-10-25 days in R3 (without dolomite) and R4
(with dolomite) reactors were 311.41C and 31.52.12C,
respectively. During the anaerobic phase, temperature
increased, then reduced to 29.54.24C and 30.31.77C,
respectively at the end of anoxic phase. At the end of
composting, the final temperatures were 332.12C and
311.41C, respectively.
Temperature in R5 (reactor in A2/O phase periods of 3010-10 days, without dolomite) and R6 (reactor in A2/O phase
periods of 30-10-10 days, with dolomite) reactors increased
to 352.12C and 372.83C, respectively, during the anaerobic phase, and then reduced to 330.71C and
35.50.71C, respectively during the anoxic phase (Figure
2c). Then during the oxic phase until the end of composting,
temperatures gradually decreased to 28.50.35C and
32.80.35C, respectively.
The highest temperature in the reactors was 37C. This
temperature was not sufficient for pathogen elimination.
Tchobanoglous et al. [18] stated that the pathogens and parasite eggs could only be destroyed at 60-70C. In addition,
three different A2/O configuration periods and dolomite
application did not impact on temperature change (P value >,
= 0.05) during composting process. Standard deviation in
temperature measurement has a value range of 0.00-3.18.
This value occured because of differences in the conditions
of the second reactors.
6
0
10
20
30
40
50
Time (days)
(c)
FIGURE 3
pH value during composting process
final pH value in R4 was 8.20.27. Reactors with dolomite
addition have pH value more stable than without dolomite
addition. Stylianou at al. [24] stated that dolomite addition
can keep pH value.
Reactor in A2/O phase 30-10-10 days at R5 (reactor
without dolomite) and R6 (reactor with dolomite) have
initial pH value of 8.00.03 and 7.60.16, respectively
(Figure 3c). Then during anaerobic phase pH value in both
101
90
Anaerobic
Moisture content (%)
R5 and R6 increased until the end of anoxic phase and were

8.20.84 and 8.20.12, respectively. During oxic phase, pH
value increased to become 9.20.71 in R5, whereas
decreased to become 7.11.07 in R6.
The study also showed that the addition of three different
A2/O configuration periods and dolomite application had no
significant impact on pH value change (P value >, = 0.05).
Standard deviation in pH measurement has a value of 0.001.07.
Anoxic
Oxic
86
82
78
74
70
66
Changes in moisture content
10
20
30
40
50
Time (days)
102
(a)
90
Anaerobic
Anoxic
Oxic
86
82
78
74
70
66
0
10
20
30
40
50
Time (days)
(b)
90
Moisture content during the composting process was 7284% (Figure 4). The initial moisture content was 79-81%,
and the final compost product was 72-80%. Figure 4a shows
that initial moisture content in A2/O phase period 25-10-15
days were 80%0.30 in reactor without dolomite (R1), and
79%1.51 in R2 (reactor with dolomite). At anaerobic
phase, moisture content in both R1 and R2 gradually
increased until the end of anoxic phase. Moisture contents at
the end of anoxic phase were 83%1.75 and 82%0.34 in
R1 and R2, respectively. Then during oxic phase moisture
content sharply decreased until at the end of composting.
The final moisture contents in R1 and R2 were 72%2.11
and 72%0.75, respectively.
Subali and Ellinawati [24] reported that moisture content
in the composting process decreased using continuous aeration until the end of composting period. The addition of dolomite was capable to degrade the organic material and to
reduce the moisture content [25]. The moisture content
should have been between 50-60% during the composting
process. The optimum value for compost maturity was 55%
[21] and the optimum moisture contents were between 5070% with the average 60% [26]. Excess of moisture content
could reduce oxygen availability during the oxic phase [27].
This condition might extend the composting process.
Ratnawati and Trihadiningrum [11] also reported that the
disadvantages of composting process of rumen content waste
required long time period.
Initial moisture content in A2/O phase periods 15-10-25
days were 81%0.3 and 79%1.51 in R3 (reactor without
dolomite) and R4 (reactor with dolomite), respectively
(Figure 4b). Similar with previous reactor (R1 and R2),
moisture content slightly decreased in R3 and R4 with the
value of 82%0.4 and 80%1.61, respectively, during
anaerobic phase until at the end of anoxic phase. The lowest
value of moisture content at the oxic phase was 80%0.43
(R3), and 730.42 (R4).
Figure 4c shows that in R5, reactor with A2/O phase
periods 30-10-10 days and without dolomite addition have
initial moisture content of 80%2.6. Whereas in reactor with
dolomite addition (R6) have moisture content of 79%0.71.
During anaerobic phase, in both R5 and R6 moisture content
slightly increased until the end of anoxic phase. Moisture
contents at the end of anoxic phase were 83%1.2 and
82%1.58 in R5 and R6, respectively. Then moisture
content gradually decreased from oxic phase to the end of
composting with the final moisture contents of 74%1.4
(R5) and 783.2 (R6).
Anaerobic
Anoxic
Oxic
86
82
78
74
70
66
0
10
20
30
40
50
Time (days)
(c)
FIGURE 4
Moisture content value during composting process
In addition, the study also showed that the addition of
dolomite had significant impact on moisture content change
(P value <, = 0.05), whereas there were no significant
(P value >, = 0.05) differences between three different A2/O
configuration periods. Standard deviation in moisture
VOLUME 42, NO. 2
MAY 2016
Organic compound + nutrients compost +

new cells + CO 2 + CH 4 + H 2 O + NH 3 +
H 2 S + energy microorganisms
(1)
In anoxic condition, elimination of nitrogen occurs through

nitrification and denitrification processes. During
nitrification process, ammonium is converted to nitrite and
nitrate by Nitrosomonas (equations 2 and 3). Nitrate is
further converted to nitrite by Nitrobacter and finally to
nitrogen gas in denitrification process (equations 4 and 5), as
described by Salimin dan Rachmadetin [30].
Nitrification:
NH 4 + + 3/2 O 2 NO 2 - + H 2 O + 2 H+
(2)
NO 2 - + O 2 NO 3 -
(3)
Denitrification:
4.50
4.00
3.50
3.00
2.50
2.00
1.50
1.00
0.50
0.00
-0.50 0
-1.00
Anaerobic
Anoxic
10
20
Oxic
30
40
50
Time (days)
(a)
Organic Nitrogen (%)
Initial organic nitrogen concentrations in all reactors

were 2.04% (Figure 5). Organic nitrogen concentrations
tended to increase in the first 34 days of composting period,
then decreased at the end of composting period.
Organic nitrogen concentrations in A2/O phase periods of
25-10-15 days were in a range from 0.02%0.75 to
3.68%0.13 (Figure 5a). In R1 (reactor without dolomite),
organic nitrogen concentrations decreased from the beginning of day 22 of composting period (during anaerobic
phase) to 3.68%0.13. Then it was slightly decreased in day
27 to 3.39%0.21 (during anoxic phase). During oxic phase
until at the end of composting period, organic nitrogen concentrations decreased to 0.070.35%. The highest organic
nitrogen concentration removal efficiency was in R1
(41.47%).
There were similar patterns between R1 and R2. In R2
(reactor with dolomite), organic nitrogen concentrations
increased in the 14 days of composting period until 27 days
of composting period (anarobic and anoxic phase), then it
decreased at the end of composting period (0.02%0.75).
Organic nitrogen concentrations decreased because of transformation and transfer of organic nitrogen concentrations
during composting. The processes were biological transformation (ammonificition, nitrification, denitrification, and
immobilization), transfer to leachate, and transfer to
atmosphere and conversion to NH 3 , N 2 O, N 2 [28]. The organic nitrogen concentration removal efficiency in R2 was
19.15%.
Four stages of biodegradation processes, namely
hydrolysis, acidogenesis, acetogenesis, and methanogenesis,
occured in anaerobic phase. The organic material will be
biodegraded to produce CH 4 and CO 2 (equation 1). The
anaerobic process can be described as suggested by
Narkhede et al. [29] as follows:
4.50
4.00
3.50
3.00
2.50
2.00
1.50
1.00
0.50
0.00
-0.50 0
-1.00
Anaerobic
Anoxic
10
20
Oxic
30
40
50
Time (days)
(b)
Organic nitrogen dynamics
content measurement has a value of 0.00-3.20.
4.50
4.00
3.50
3.00
2.50
2.00
1.50
1.00
0.50
0.00
-0.50 0
-1.00
-1.50
Anaerobic
10
Anoxic
20
30
Oxic
40
50
Time (days)
(c)
FIGURE 5
Organic nitrogen concentration content
during composting process
Substrate + NO 3 - NO 2 - + CO 2 + H 2 O
(4)
Substrate + NO 2 - N 2 + CO 2 + H 2 O + OH-
(5)
103
The organic nitrogen concentrations in A2/O phase periods of 15-10-25 days in R3 and R4 were from 0.02%0.04
to 3.44%0.34 (Figure 5b). In R3 (reactor without dolomite), the organic nitrogen concentrations increased in the
beginning of composting until the end of anoxic phase from
3.21%0.65 to 3.41%0.07. Then it decreased towards the
end of composting process to 0.040.35%. The organic nitrogen removal efficiency in R3 was 40.35%.
Whereas in R4 (reactor with dolomite), organic nitrogen
concentrations increased from 2.04 to 3.46% during the first
7 day composting process (anaerobic phase). During anoxic
phase (day 22), the organic nitrogen concentrations increased to 3.44%0.34, and decreased during oxic phase
until the end of composting period. The final organic
nitrogen concentration in R4 was 0.06%0.27. The organic
nitrogen removal efficiency in R4 was 15.36%.
Organic nitrogen concentrations in A2/O phase periods of
30-10-10 days were from 0.01%0.16 to 3.74%0.13
(Figure 5c). In R5 (30-10-10 days; without dolomite), during
anaerobic phase, organic nitrogen concentrations increased
from 2.04%0.42 to 3.74%0.13. Then it decreased to
1.29%0.28 at day 34 (anoxic phase). Then it decreased
during oxic phase to 0.01%0.16 until the end of composting period. The organic nitrogen concentrations removal
efficiency in R5 was 36.77%.
Organic nitrogen concentration in R6 (30-10-10 days;
with dolomite) increased from 2.04%0.42 to 3.74%0.13
during anaerobic phase. It decreased to 1.29%0.28 in
anoxic phase, and to 0.03%1.07 in the end of composting
period. The organic nitrogen concentrations removal efficiency in R6 was 17.65%.
The research also showed that the addition of three different A2/O configuration periods and dolomite application
had no significant impact on organic nitrogen concentrations
change (P value >, = 0.05). Standard deviation in moisture
content measurement has a value of 0.01-1.07.
Inorganic Nitrogen (%)
(6)
104
Oxic
0.60
0.40
0.20
0.00
0
10
20
-0.40
30
40
50
Time (days)
(a)
1.20
Anaerobic
Anoxic
Oxic
1.00
0.80
0.60
0.40
0.20
0.00
0
10
-0.20
20
30
40
50
Time (days)
(b)
1.20
Anaerobic
Anoxic
Oxic
1.00
0.80
0.60
0.40
0.20
0.00
-0.40
The inorganic nitrogen concentrations during the composting process were shown in Figure 6. Initial inorganic
nitrogen concentrations during the composting process were
from 0.25%0.04 to 0.48%0.26. Organic nitrogen concentrations tended to decrease from the beginning to the end of
composting period in all reactors. Reactors without dolomite
addition tended to have higher value of inorganic nitrogen
concentrations than those of the reactors with dolomite addition. Inorganic nitrogen concentrations in reactors without
dolomite addition were in a range from 0.01%0.0 to
1.05%0.06, and in reactors with dolomite addition were in
a range from 0.02%0.05 to 0.48%0.26. In general, pH
value higher than 7.5 could inhibit nitrification process. The-
Anoxic
0.80
-0.20
Inorganic nitrogen dynamics
Anaerobic
1.00
-0.20
Organic compound + O 2 + nutrients compost +

new cells + CO 2 + H 2 O+ NH 3 + SO 4 2- + energy
1.20
Microorganisms will oxidize organic compounds and

produce CO 2 , H 2 O, NH 3 , and SO 4 2- under aerobic condition
(Theresa et al. [31]), as shown in equation 6.
10
20
30
40
50
Time (days)
(c)
FIGURE 6
Inorganic nitrogen concentration content
during composting process
se conditions supported NH 3 volatilization [10].
The inorganic nitrogen concentrations in A2/O phase
periods 25-10-15 days were from 0.02%0.27 to
VOLUME 42, NO. 2
MAY 2016
0.56%0.03 (Figure 6a). In R1 (reactor without dolomite),

during anaerobic phase, inorganic nitrogen concentration
fluctuated then decreased to 0.21%0.03 at the end of
anaerobic phase. Then sharply decreased during anoxic and
oxic phase till the end of composting period. The final
inorganic nitrogen concentration in R1 was 0.07%0.02.
The inorganic nitrogen concentrations removal efficiency
was 75.43 (R1)
Reactor with dolomite (R2) has similar trend in changes
of inorganic nitrogen concentration. From the beginning of
composting until the end of composting period, inorganic
nitrogen concentrations were sharply decreased to
0.02%0.27 (the final inorganic nitrogen concentrations).
The inorganic nitrogen concentrations removal efficiency
was 95.32% (R2). Ratnawati and Trihadiningrum [33] also
reported that inorganic nitrogen concentration gradually
decreased until the end of composting process.
Inorganic nitrogen concentrations in A2/O phase periods
of 15-10-25 days in R3 (without dolomite) and R4 (with
dolomite) were in a range from 0.02%0.05 to 1.05%0.06
(Figure 6b). Inorganic nitrogen concentrations in R3 and R4
tended to increase in the first 14 days of composting period
(during anaerobic phase), then sharply decreased during
anoxic phase till the end of composting period
0.04%%0.03. The inorganic nitrogen concentrations removal efficiency was 90.06%.
Whereas in R4, during anaerobic phase, inorganic nitrogen concentrations slightly increased to 0.35%0.02. Then
during anoxic phase, there were similar fluctuations of inorganic nitrogen concentrations and it gradually decreased
from oxic phase to the end of composting period (oxic
phase). The final inorganic nitrogen concentrations were
0.06%0.06. The inorganic nitrogen concentrations removal
efficiency was 76.05%.
Inorganic nitrogen concentrations in A2O phase periods
of 30-10-10 days were from 0.03%0.03 to 1.00%0.11
(Figure 6c). In R5 (without dolomite), during anaerobic
phase, inorganic nitrogen concentrations sharply increased to
0.79%0.12 (day 14), and 1.00%0.11 (day 27). Then
sharply decreased during anoxic phase to 0.09%0.08, and
gradually decreased during oxic phase until the end of
composting period. The final inorganic nitrogen concentrations were 0.01%0.06.
Reactor with dolomite (R6), inorganic nitrogen concentrations during anaerobic phase decreased to 0.06%0.05.
Then gradually decreased during anoxic phase until the end
of composting period. The final inorganic nitrogen concentrations were 0.03%0.03.
The research also showed that the addition of three different A2/O configuration periods and dolomite application
had no significant impact on inorganic nitrogen concentrations change (P value >, = 0.05). Standard deviation in
inorganic nitrogen concentrations measurement has a value
range of 0.00-0.32.
waste was 25-10-15 days. In this reactor, moisture content

was decreased from 83% to 72%. Temperature range was
29-34C, pH values slighty fluctuated from 7.57 to 8.70.
Inorganic nitrogen concentration decreased from 0.29 to
0.07%, and organic nitrogen concentration decreased from
2.04 to 1.19%. Dolomite addition showed that it had no significant impact on temperature, pH values, inorganic and
organic nitrogen concentrations in all treatments (P value >,
= 0.05). Therefore dolomite addition was not needed for SH
composting process. This research proved that A2 O method
is appropriate for treating solid waste with high nutrient
contents.
ACKNOWLEDGMENT
This research was supported by the Directorate of Higher
Education, the Ministry of Research, Technology, and Higher Education of the Republic of Indonesia, for the Research
Grant Award PUPT 2014, Contract No. 016452.41/
IT2.7/PN.01.00/2014. Thank you to Mr. Andy Mizwar and
Ms. Deqi Raditya for proof reading this manuscript and Ms.
Desty Ayu Permatasari for helping this research.
REFERENCES
[1]
[2]
[3]
[4]
[5]
[6]
CONCLUSIONS
This research showed that the optimum anaerobicanoxic-oxic time period for composting the rumen content
[7]
Roy B.C., Khan M.R.I., Rahman M.M., Salleh

M.A.M., Ahsan A., Amin M.R. (2013): Development
of a convenient method of rumen content composting.
Journal of Animal and Veterinary Advances, Volume
12, pp. 1439-1444.
Olaniyi C.O. (2010): Economic evaluation, growth
performance and nutrient utilization of rumen content
in the nutrition of african catfish (clarias gariepinus)
burchell, 1822. In: Shriver AL (Hrsg.), Proceedings of
the Fifteenth Biennial Conference of the International
Institute of Fisheries Economics & Trade. International
Institute of Fisheries Economics & Trade, USA, July
13-16, 2010, Montpellier, France.
Akinyeye A.J., Solanke E.O., Akinmolayan O.A.
(2012): Physico-chemical analysis of abattior effluent
contaminated soil in Ikpoba, Benin, Edo State.
Research Journal of Pharmaceutical, Biological and
Chemical Sciences, Volume 3, 1079-1084.
Magaji J.Y., Chup C.D. (2012): The effects of
abattoir waste on water quality in Gwagwalada-Abuja,
Nigeria. Ethiopian Journal of Environmental Studies
and Management, Volume 5, pp. 542-549.
McCaskey T.A., Little J.A., Krotz R.M., Lino S.P.,
Hannah T.C. (1996): Onfarm composting feasible for
disposal of swine carcasses, Highlights Agric. Res.
Alabama Agric. Exp. Stat., pp. 18-20.
Morrow W.E.M., OQuinn P., Barker J., Erickson G.,
Post K., Mc Caw M. (1995): Composting as a suitable
technique for managing swine mortalities. Swine
Health and Production, Volume 3, pp. 236-243.
Stanford K., Larney F.J., Olson A.F., Yanke L.J.,
105
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
106
McKenzie R.H. (2000): Composting as a Means of

Disposal of Sheep Mortalities. Compost Science &
Utilization, Volume 8, pp. 135-146.
Dolliver H., Gupta S., Noll S (2008): Antibiotic
Degradation during Manure Composting. Journal of
Environmental Quality, Volume 37, pp. 1245-1253.
Lee J.E., Rahman M.M., Ra C.S. (2009): Dose effects
of Mg and P0 4 sources on the composting of swine
manure. Journal of hazardous materials, Volume
169, pp. 801-807.
Tiquia S.M., Richard T.L., Honeyman M.S. (2002):
Carbon, nutrient, and mass loss during composting.
Nutrient Cycling in Agroecosystems, Volume 62, pp.
15-24.
Ratnawati R., Trihadiningrum Y. (2014): Slaughterhouse solid waste management in Indonesia. Journal
of Biological Researches, Volume 19, pp. 69-73.
Kithome M., Paul J.W., Bomke A.A. (1999):
Reducing Nitrogen Losses during Simulated
Composting of Poultry Manure using Adsorbents or
Chemical Amendments. J. Environ. Qual., Volume
28, pp. 194-201.
Baeza J.A., Gabriel D., Lafuente J. (2004): Effect of
internal recycle on the nitrogen removal efciency of
an anaerobic/anoxic/oxic (A2/O) wastewater treatment
plant (WWTP). Process Biochemistry, Volume 39,
pp. 16151624.
Peng Yong-zhen, Xiao-lian W., Bai-kun L. (2006):
Anoxic Biological Phosporus Uptake and the Effect
of Excessive Aeration on Biological Phosporus Removal in The A2O Process. Desalination, Volume
189, pp. 155-164.
Kargi F., Uygur A. (2003): Nutrient removal performance of a five-step sequencing batch reactor as a
function of wastewater composition. Process Biochemistry, Volume 38, pp. 1039-1045.
Iqbal M.K., Khan R.A., Nadeem A., Hussnain A.
(2012): Comparative Study of Different Techniques
of Composting and Their Stability Evaluation in Municipal Solid Waste. J. Chem. Soc. Pak, Volume 34,
pp. 273-282.
Bari Q.H. (1999): Effect of Different Modes of
Aeration on Composting of Solid Waste in A Closed
Sistem. Thesis for Doctor of Philosophy at the
University of Hong Kong.
Tchobanoglous G., Burton L., Franlin, Stense, David
H. (1993): Wastewater Engineering Treatment and
Reuse, Fourth Edition. McGraw Hill Book Co.,
Singapore.
Tanpanich S., R. Chindachia, S. Duriyaprapan (2009):
Rate of Composting and Quality of Compost Under
Different Passively Aerated Composting. KKU Sci. J.,
Volume 37 (Supplement), pp. 153-161.
Saithep N., S. Dheeranupatana, P. Sumrit, S. Jeerat, S.
Boonchalearmkit, J. Wongsanoon, C. Jatisatienr
(2009): Composting of Tobacco Plant Waste by Manual Turning and Forced Aeration System. Maejo Int.
J. Sci. Technol., Volume 3(2), pp. 248-260.
Bernal M.P., Alburquerque J.A., Moral R. (2009):
[22]
[23]
[24]
[25]
[26]
[27]
[28]
[29]
[30]
[31]
[32]
[33]
Composting of Animal Manures and Chemical

Criteria for Compost Maturity Assessmen, A Review,
Bioresource Technology, Volume 100, pp. 5444-5453.
Guo R., Li G, Jiang T., Schuchardt F., Chen, T., Zhao
Y., Shen Y. (2012): Effect of Aeration Rate, C/N
Ratio and Moisture Content on The Stability and
Maturity of Compost, Bioresource Technology,
Volume 112, pp. 171-178.
Bishop P.L., Godfrey C. (1983): Nitrogen
transformations during sludge composting. Biocycle,
Volume 24, pp. 3439.
Stylianou M.A., Inglezakis V.J., Moustakas K.G.,
Loizidou M.D. (2008): Improvement of the quality of
sewage sludge compost by adding natural
clipnotilolite. Desalination, Volume 224, pp. 240249.
Subali B., Ellianawati (2010): Effect of Composting
Against Time Ratio C/N and Total Moisture In Compost. The 24th HFI National Conference East Java and
Yogyakarta, Semarang.
Polprasert, C. (2007): Organic Waste Recycling Technology and Management: 3rd edition. London: IWA
Publishing.
Himanen M., Hanninen K. (2011): Composting of
Bio-Waste, Aerobic and Anaerobic Sludges-Effect of
Feedstock on The Process and Quality of Compost.
Bioresource Technology, Volume 102, pp. 2842-2852.
Schaub S.M., J.J. Leonard. (1997): Composting: An
Alternative Waste Management Option for Food Processing Industries. Trends in Food Science V Technology.
Guardia A., Mallard P., Teglia C., Marin A., Le Pape
C., Launay M., Benoist J.C., Petiot C. (2010):
Comparison of five organic wastes regarding their
behaviour during composting: Part 2, nitrogen
dynamic, Waste Management, Volume 30, pp. 415
425.
Narkhede S.D, Attarde S.B., and Ingle S. (2010):
Combined Aerobic Composting of Municipal Solid
Waste and Sewage Sludge. Global Journal of Environmental Research, Volume 4 (2), pp. 109-112.
Salimin Z., Rachmadetin J. (2011): Denitrification
Liquid Radioactive Waste Containing Nitric Acid with
Biooxidation Process. The 9th Waste Management
Technology National Conference, Cilegon: Pusat
Teknologi Limbah Radioaktif-BATAN, Departement
of Engineering, Universitas Sultan Ageng Tirtayasa.
Theresa M.I., Milind V., Evangelyn K., Alocilja C.
(2007): Aerobic In-Vessel Composting Versus
Bioreactor Landfilling Using Life Cycle Inventory
Models, Clean Techn Environ Policy, Volume 10, pp.
3952.
Ratnawati R., Trihadiningrum Y., Warmadewanthi,
Herumurti W. (2014): Nitrogen and Phosphate
Concentration Patterns in the Aerobic and A2O
Composting Processes of Slaughter House Solid
Waste. 2nd International Postgraduate Conference on
Biotechnology (IPCB2014), 26-28 August 2014 at
Soka University, Tokyo, Japan.
VOLUME 42, NO. 2
MAY 2016
107
OPPORTUNITY OF BIOCHEMICAL PROCESS FOR

PHOSPHOGYPSUM UTILIZATION
Ye Chernish, Ph.D.
L. Plyatsuk, DSc., Professor
Department of Applied Ecology, Faculty of Technical Systems and Energy Efficient Technologies
Sumy State University, Sumy, 40007, Ukraine
liza_chernish@mail.ru
ABSTRACT
This paper focuses on the study of the phosphogypsum (PG) bioconversion process to elementary sulfur with using the bio-desulphurization system. Objectives of this work regards: production microbially generation (MG) gas under bio-sulfidogenic conditions; the immobilization of
Thiobacillus sp. on different support mediums (lavsan fiber and granular activated carbon
(GAC)) and the effect of contact time and culture pH on H 2 S removal. Anaerobic microbiological degradation (AMD) of sewage sludge with phosphogypsum containing sulfur containing
substances produces hydrogen sulfide in microbially generation gas. The optimal loading dose
of sewage sludge (SS) and dose of phosphogypsum were determined. The system of sulfide
bioconversion consisted of the scrubber connected to bio-desulfurization unit. The enrichment
culture of T. intermedius and T. ferrooxidans were used in bio-desulphurization system. The optimum operating parameters of desulfurization unit were determined. The usage of lavsan fibre
provides maximum H 2 S removal (89.78% w/w) and maximum bacterial growth reached
(3.51010 CFU/g) that proves to be more effective than that usage of granular activated carbon.
Therefore, pH and contact time were found to be a critical factors in the operation of the H 2 S
bioconversion system of sulfide with using different support mediums. The modification of biodesulfurization system was carried out using acidophilic mode.
Keywords: Anaerobic microbiological degradation, phosphogypsum, desulphurization, elemental sulfur
INTRODUCTION
Environmental problems take the lead in this big list of
challenges that humanity face nowadays. Stabilization
heavy metals by converting them to more chemical stable
forms that are less likely to leach is the one of priority task
for ecological safety. Designing an environmental friendly
and sustainable products and processes, inventing new ways
to recycle and reuse waste is a major challenge for future
engineers. Thus, the basic factors of the deposition process
of heavy metals by biogenic hydrogen sulfide and sulfide
conversion to elementary sulfur and produce microbially
generated gas were investigated.
108
Millions of tons of phosphogypsum (PG) is stacked worldwide every year and is progressively considered as an asset
more than an environmental burden. Phosphogypsum (PG) is
a waste by-product from the processing of phosphate rock by
the wet acid method of fertiliser production, which currently accounts for over 90% of phosphoric acid production.
World PG production is variously estimated to be around
100-280 Mtons per year [1] and the main producers of phosphate rock and phosphate fertilisers are in the USA, China,
Russia, Africa and the Middle East. In Ukraine the most
common of all kinds of gypsum wastes is phosphogypsum.
Phosphogypsum is formed in an amount of about 100 tons
annually in Sumy region (Ukraine). Currently, over 14 mil-
VOLUME 42, NO. 2
MAY 2016
lion tons of phosphogypsum was accumulated [2]. Solid

waste generated in the process of sulfuric acid decomposition
of natural phosphate raw material and the solid phase (calcium sulfate) separation from phosphoric acid solutions.
Reclamation includes a cover of vegetation to improve
aesthetic qualities, to reduce erosion and improve the quality
of surface runoff water. There are several problems associated with establishing vegetative cover on phosphogypsum,
including the residual acidity, nutrient deficiencies and low
nutrient-holding capacity, and the tendency for caking and
crust formation. In addition, there are other potential environmental problems [3].
The potential risk in estuarine zones is a consequence of a
dam that is not totally watertight allowing some metal-rich
efuents from the leaching of phosphogypsum, along with
some partial collapses during storm events, to reach the rivers. [4-7].
In fact, mineral phosphates, phosphate fertilizers,
phosphogypsum and phosphoric acid plant effluents generally
have lower radioactivities than the international limit value
prescribed by the European Atomic Energy Community. In
any case, only those materials meeting such safety requirements would be used. Pilot studies have demonstrated the
very low degree of risk and it is likely that re-use of
phosphogypsum will become more common in the near future [8]. These regulations will be the constraints that will
control the design of any process for cleaning PG in order to
be ready for reuse. Finding a safe and environmentally friendly use for this waste byproduct is imperative.
PHOSHPOGYPSUM TREATMENT
The methods described in the literature to minimise the
negative effects of this waste are classified by treatment type,
i.e. physical, chemical, thermal, etc., and different suggested
applications for PG are detailed [9-12]. PG has been used in
the cement industry as a setting regulator in place of natural
gypsum [9] and in the gypsum industry to make gypsum plaster. PG also has been used as agricultural fertilisers or soil
stabilisation amendments [13, 14]. Different authors have
proposed different temperatures to produce anhydrite [15,
16].
A new direction in environmental biotechnology is an
attempt of the simultaneous biodegradation of mixed sewage
sludge with phosphogypsum for growth sulfate-reducing bacteria (SRB). It is effective to use gypsum wastes as a source
of sulfate in biosulfide treatment of sewage sludge. The system of anaerobic microbiological degradation (AMD) with
the heavy metals sedimentation by biogenic hydrogen sulfide
is the promising way of sewage sludge detoxification. Thus
the possibility of co-processing of sewage sludge and
phosphogypsum in the anaerobic condition with the heavy
metals sedimentation with biogenic hydrogen sulfide was
substantiated in the [2, 17]
In the AMD under biosulfide conditions microbially generated (MG) gas is formed that contain high concentration of
H 2 S. So, not all amount of hydrogen sulfide is used for HM
ions precipitation. Thus, it is necessary to recover residual
H 2 S. It can be used for removing heavy metals from the new

batch of stabile mixture of wastes [2] or to produce elementary sulfur. Many microorganisms have been used to remove
hydrogen sulphide, principally bacteria. Thiobacillus is most
prevalent species involved in H 2 S removal. Literature have
reported the use of different Thiobacillus species (thioparus,
thiobacilli, denitricans, thiooxidans, ferroxidans) and other
microorganisms (chlorobiaciae, xanthomonas) converting the
H 2 S into reduced sulphur compounds [18].
In biological sulfide oxidation end product will be produced based on the oxygen concentration. Under oxygen limiting conditions, that is at O 2 concentration below 0.1 mg/l,
sulfur is the end product of the sulfide oxidation, while sulfate is formed under circumstances of sulfide limitation [19].
The THIOPAQ [20] process has been developed to remove H 2 S from low pressure biogas streams. In this process,
a gas stream containing H 2 S contacts an aqueous soda solution containing thiobacillus bacteria in an absorber. System
pH ranges from 8.2 to 9. But this range of pH is inhibiting the
Thiobacillus growth in the biofilter. The sulfur oxidizing bacterium Thiobacillus sp. are the gram negative, aerobic, rod
shaped, motile cells, grow well at 20300C and a pH of
around 2 to 5 [21].
Desulfurization system cannot be using when concentration of H 2 S more than 2.5% of gas general volume [20].
THIOPAQ process can not be used for cleaning microbially
generated gas with sulfides concentration more than 40%
vol., which formed during AMD under bio-sulfidgenic condition.
The use of support medium for the immobilization of microorganisms is widely known to provide a surface for microbial growth and a shelter that protects the microorganisms
from inhibitory compounds. Granular activated carbon
(GAC) often using for immobilization of microorganisms. To
remove hydrogen sulphide, the packing material media were
used in conventional biofilter beds consisting mostly of peat
and compost, but a wide variety of other materials can been
used: ca-alginate, polypropylene Pall rings, porous lava,
wood, and polyurethane foam. Literature contains few accounts of the use of polyurethane foam to remove malodorous gases with biofilters [22, 23].

The anaerobic bioreactor is a continuous stirred tank reactor with an internal settling zone that has useful capacity of 5
dm3. Gas sampling was carried out in a special sampling bags
of inert plastic Teflon volume of 1 dm3. The system of sulfide bioconversion consisted of the scrubber (1 dm3) connected to bio-desulfurization unit. The bio-desulfurization unit
looked like a plexiglass column with biofilter which contained the support medium.
Study of the gaseous phase was carried out on a laboratory gas chromatograph SelmiChrom-1 (Ukraine). The thermal
conductivity detector, also known as a katharometer was
used. Argon was chosen as the mobile phase. The compounds
are separated by traveling through a coated column and separate based on size and intermolecular interactions. The sam109
ple is injected into the pre-PLOT (porous layer open tubular) column-1 with a porous layer of sorbent PoraPlot Q (styrene-divinylbenzene coating). The gas mixture passing
through the column-1 goes to one of the two katharometer
cells. The hydrogen, oxygen, nitrogen follow out of the column-1 as a single chromatographic strip, then methane, carbon monoxide (IV) and hydrogen sulfide are separated. The
ballast column-2 is filled with an inert carrier Chromaton NAW-DMCS. H 2 and N 2 were separated in the HP-Plot Mole
Sieve column-3. Registration and processing of chromatograms was performed by the software Multichrom version
1.52x.
The quality analyses of materials were performed on the
scanning electron microscope-microanalyzer REMMA-102
(Ukraine). The measurements were performed at five points
on each sample. Spectrometric data are processed with the
necessary calibration measurements, the decryption of x-ray
spectra, qualitative and quantitative electron probe analysis
were performed using software microanalysis system. The
sensitivity of the measurement was at 1%. Additionally surveyed elemental composition of waste by X-ray fluorescence
analysis, which made it possible to determine the concentration of elements in the ppm level. Photomicrographs were
obtained and processed using the digital output of image
SEO Scan ICX 285 AK-F IEE-1394 and morphometric
program SEO Image Lab 2.0 and on the scanning electron
microscope REMMA-102. pH was analyzed by pX-meter
pX-150 (ionometer) (Belarus).
Microbiological investigation of
Thiobacillus
Isolation of sulfide oxidizing bacteria was done from aerobic sewage sludge collected from municipal wastewater
treatment plant. The aerobic sludge samples were collected
and screened for the removal of big particles.
Then the sludge is kept in aerobic conditions by continuous aeration in order to prevent growth of any anaerobic bacteria for a period of 10 days at temperature of 35 oC.
According to [24] Thiobacillus sp. can growth in various
range of pH. When pH differs from optimal value, maintenance energy requirements increase that leads to decrease in
H 2 S removal and population growth. The enrichment culture
of T. intermedius and T. ferrooxidans were used in studied
biodesulphurization system utilizing a range of 2 pH 5.
The enrichment medium for Thiobacillus sp. cultivation
has the following composition: NH 4 Cl, 1.0 gm; K 2 HPO 4 ,
0.6 gm; CaCl 2 2H 2 O, 0.2 gm; FeCl 3 H 2 O, 0.02 gm [25]; we
tacked such sulfates: ZnSO 4 7H 2 O, 40 mg; CaSO 4 2H 2 O, 70
mg; MnSO 4 , 15 mg; Na 2 B 4 O 7 , 10 mg; distilled water, 1000
ml; pH, 5.
phosphogypsum improves soil structure due to the high content of cations Ca2 +.
Phosphogypsum consists mainly of calcium sulfate dehydrate (CaSO 4 2H 2 O) and contains impurities of not decomposed phosphate, phosphates and silicates. Figure 1 provides
the results of Energy Dispersive X-Ray Analysis (EDX) of
phosphogypsum.
The quantitative content of impurities depends on the
mineral composition of the feedstock, smooth flow of production and serviceability of equipment and technological
process, etc. In phosphogypsum samples, taken directly from
the production process flow sheet were found in the X-ray
fluorescence analysis of such metals (% of total mass): Fe
(0.010%), Ni (0.001%), Cu (0.003%).

Production MG gas under bio-sulfidogenic
conditions
Following doses of phosphogypsum (X 1 ) were taken: 10
g/dm3; 12 g/dm3; 14 g/dm3; 16 g/dm3. The required amount of
fresh SS was defined as the proportion of the SS mass. Initially, loading doses of SS (X 2 ) were chosen such as: 1; 2.5; 4;
5.5; 7.0%. pH in the bioreactor was maintained within 7.0
7.5 units. Retention time was 10 days.
Figure 2 shows the process of gas formation that was observed under bio-sulfidogenic conditions.
The chromatographic investigation showed that content of
biogenic hydrogen sulfide gas changed in the wide range,
depending on the rate of consumption of organic and mineral
substrates. The effect of SS loading dose (X 2 ) at the PG dose
(X 1 ) of 14 g/dm3 for the gassing process has been analyzed.
Thus the total volume of MG gas production was 10.63 dm3
under initial loading dose (X 2 ) of 0.035 dm3 (1.0%) (Figure
Phosphogypsum characteristics as
nutrients
Phosphogypsum is a source of calcium, phosphorus, sulfur and trace elements, so it can be used for chemical soil
improvement, for example, sandy soils. In this application of
110
FIGURE 1
X-ray microanalysis spectrum indicating the detected
elements in a sample of phosphogypsum: S 3 (39.22 % w);
C (38.73 % w); Si 2 (1.79 % w); P 2 5 (0.45 % w).
VOLUME 42, NO. 2
MAY 2016
FIGURE 2
Formation of MG gas depending on the loading dose of SS (X 2 ) at different PG doses (X 1 )
2). The lag phase of bacterial growth was 2 days. The content
of hydrogen sulfide in the gas phase was increased gradually.
The gas production began growth after increasing of SS loading dose to 0.088 dm3 (2.5%). Total volume of MG gas was
21.3 dm3 and hydrogen sulfide varied within 30.032.2 %.
Then the SS loading dose was increased to 0.14 dm3 (4%)
that provoked increase of MG gas production to 33.86 dm3
with 33.038.0 % hydrogen sulfide. The largest increase of
MG gas volume was get by increasing of the SS loading dose
to 0.19 dm3 (5.5%). The results showed (Figure 2) that the
total volume of MG gas was stabilized at 57.32 dm3, in average 1.63 m3 per 1 m3 SS. The MG gas consisted of 46.848.0
% hydrogen sulfide, 24.326.5 % methane and 18.220.4 %

carbon dioxide. Nitrogen and hydrogen were presented as
impurities. According to the results of laboratory tests loading dose (X 2 ) more than 5.5 % by SS weight was not provide
growth of total volume of MG gas. Thus after loading dose
(X 2 ) of 7.0% association microorganisms did not have time
to recycle supplied organic substrate, as evidenced by the
decrease in gas production (Figure 2).
When loading dose of SS at the constant level of 5.5%,
sulfidogenesis decreasing (hydrogen sulfide was not more
than 25%) was provoked mostly by reducing of PG dose to
10 g/dm3 (Table 1).
TABLE 1
Results of the chromatographic analysis of MG gas
PG dose, g/dm3
Compounds of the
gaseous phase
10
12
The systematic error, rel.%

14
16
The volume, %
hydrogen sulfide
24.11.10
38.71.86
47.40.6
47.01.12
4.0
carbon dioxide
37.51.18
22.01.18
19.31.12
19.51.10
3.0
methane
34.81.05
35.41.05
25.41.0
25.01.05
3.0
hydrogen
2.30.02
2.60.02
2.80.02
2.70.02
5.0
nitrogen
1.30.03
1.30.03
5.10.03
4.30.03
5.0
111
Thus depletion of sulfate ions sources became through the

natural process of using sulfate-reduction bacteria sulfur mineral supplements (phosphogypsum) and provoked decreasing
of hydrogen sulfide content in MG gas.
Immobilized biomass in biofilter

The biomass of Thiobacillus sp. was immobilized on support medium, which consisted of lavsan fiber (Figure 3).
Comparing to other fibers lavsan fiber is more resistant to
high temperatures, abrasion, oxidizers, acids, and other chemical agents. Lavsan fibers preparation as in the case of a polyamide fiber is a single production cycle, comprising the steps
of production of the monomer di (p-hydroxyethyl) terephthalate, a polycondensation and it is expressed by reaction:
OOC
CH2
n OH
CH2OH
CH2
OOC
CH3
OH
COO
CH2
COO
CH2
CH2
CH2
...
n-1
OH
CH2
...
OOC
COO
CH2OH
CH2
n-1 OH
CH2
(1)
Immobilization of microorganisms can be carried out on

granular activated carbon (GAC) (Figure 5). The total immobilized biomass at the end of batches was of 1.6109 CFU/g.
Biochemical reactions were aggressive (pH = 4.55.0) for
GAC and its structure was quickly destroyed during biofilter
operation (after 30 days of exploitation). It was washed out of
the reactor. This support medium was needed to be updated a
lot more often than with lavsan fiber (35 times more). Thus,
GAC destruction caused the thiobacillus biomass reduction.
The sulfate in the biofilter medium is a toxic compound.
The principal problem is that sulfate is a product of the hydrogen sulfide biological oxidation. Yang et al. (1994) [21]
found that sulphate concentrations of more than 24 mg inhibited the performance of a hydrogen sulfide removing compost
biofilter. Note, that immobilization of microorganism was
provided in GAC. In this work the elimination capacity decreased quickly when the concentration of sulfate was more
than 5 g/dm3, nevertheless the biofilter had a fast capacity of
recovery (replace media, decrease load). In all cases, the sulfate concentration was kept at 34 g/dm3.
After Thiobacillus sp. immobilization in lavsan fiber, sulphate concentrations was more than 24 mg and didnt inhibit
hydrogen sulfide removing. The inhibition process started
from 45 mg. General formation of sulfur is preferred because
it is insoluble and can be easily recovered from the water
stream. The formation of end product is not only dependent
on the sulfide concentration but also on the amount of oxygen
supply to the reactor. This is evident from the following general scheme of biological sulfide oxidation system:
In the Figure 4 we can see the bacteria of Thiobacillus sp.

immobilized on the surface of lavsan fiber.
The total immobilized biomass at the end of batches was
of 3.51010 CFU/g of lavsan fiber support medium. During
desulfurization system operating period this material didnt
require regeneration. The advantage of lavsan fiber serving as
the support medium is to adsorb a high initial concentration
of substrate and progressively release it for microbial degradation, hence acting as a buffer for the microorganisms. The
biofilter worked stable (120 days) without this support medium regeneration.
But oxygen concentration is firstly determined by aerobic

needs of Thiobacillus. Thus, we cannot decrease aeration
FIGURE 3
Photograph of lavsan fiber (general view)
FIGURE 4
Scanning electron micrographs of the lavsan fiber structure
with Thiobacillus biomass during immobilization process.
Magnification: 2 microns.
112
(2)
(3)
VOLUME 42, NO. 2
MAY 2016
FIGURE 5
Scanning electron micrographs of the structure GAC with
Thiobacillus biomass during immobilization process. Magnification: 5 microns.
volume to minimum because of metabolism bacteria inhibition.
Thus in our experiments the initial concentration of sulfide was water down after scrubber. Thus, process of formation sulfate in biofilter was prevented.
Effect of contact time and culture pH

Concentration of H 2 S gas in treated gas (45% vol) and
bacterial population at a regular interval (5, 10, 15 h) of contact time (Figure 6) consisted of 30 tests for each interval.
Concentration of H 2 S gas and bacterial populations were
measured at optimum contact time (10h) for various pH val-
FIGURE 6
H 2 S removal in treated gas depending on contact time
ues (4.0, 4.5, 5.0, 5.5, 6.0 and 6.5) of culture (Figure 7). This
experiment consisted of 30 tests for each pH value. The biomass was immobilized on support medium, which consisted
of lavsan fiber.
Removals of H 2 S from gas were 18.56, 87.53 and 70.31
% (w/w) at 5, 10 and 15 h of contact time with Thiobacillus
respectively. H 2 S bacterial desulphurization increases with
contact time increase (10 h) and then declines at 15 h (Figure
6). Because of bacterial populations increase (3.51010
CFU/g) with contact time increase up to 10 h and then declines (1.71010 CFU/g) at 15 h. Microbial growth retardation
is due to the accumulation of metabolic products (i.e., sulfuric
acid); and a lack of water and nutrient salts in the biofilter are
other possible causes of column breakthrough. Thus, maximum (87.53 % w/w) removal of H 2 S and maximum
Thiobacillus populations (3.41010 CFU/g) were observed at
10 h of contact time. Therefore, this contact time is optimum.
Bacterial populations slow down due to nutrient exhaustion and intoxication medium by H 2 S after 10 h. At early
period of hydrosulfide bio-desulphurization was coincided
with lag-phase of bacterial growth. Extent of lag-phase is
dependent on H 2 S concentration, initially containing high
values and leading to lower growth of Thiobacillus sp. and
H 2 S removal in bioreactor. H 2 S removal increased with initial culture pH increase up to 5.0 and then it declined (Figure
7).
Bacterial growth reached maximum (3.51010 CFU/g)
with maximum H 2 S removal (89.78% w/w) at initial culture
pH 5.0 at optimum contact time of 10 h. Variations (4.56.5)
in pH culture medium resulted in changes in activity of
Thiobacillus sp. and hence bacterial populations as well as
H 2 S removal.
CONCLUSION
The use of PG poses unsolvable and troublesome prob-
FIGURE 7
H 2 S removal in treated gas depending on pH
113
lems. Environmental concerns are related with the large

stockpiles of PG and their negative impact on surrounding
land, water and air. More than 85% of all PG generated each
year is dumped on land or in the sea. The possibility PG utilization in biotechnology was studied.
MG gas with high concentration of hydrogen sulfide was
produced during AMD of sewage sludge with
phosphogypsum under bio-sulfidogenic conditions. The optimal loading dose of sewage sludge (SS) and dose of
phosphogypsum were determined.
The gaseous phase was purified in the bio-desulfurization
system. Experimental study compared results of two kinds of
support medium such as lavsan fiber and GAC was performed. pH and contact time were found to be a critical factors in the operation of the H 2 S bioconversion system of sulfide with using different support mediums. After Thiobacillus
biomass immobilized on lavsan fiber bacterial growth
reached maximum (3.51010 CFU/g) with maximum H 2 S
removal (89.78% w/w) at initial culture pH 5.0 at optimum
contact time of 10 h. Immobilization of microorganisms on
GAC has shown that the total immobilized biomass at the end
of batches was of 1.6109 CFU/g. Biochemical reactions were
aggressive (pH = 4.55.0) for GAC and its structure was
quickly destroyed during biofilter operation (after 30 days of
exploitation). Thus the possibility of using lavsan fiber as
material of support medium was tested. The usage of lavsan
fibre proves to be more effective than that of GAC in terms of
bio-desulfurization system performance and immobilized
biomass viability.
The possible avenues for future research are development
of regression models of gas purification efficiency and the
formation of elemental sulfur during process of biodesulfurization.
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
REFERENCES
[13]
[1]
[2]
[3]
[4]
114
A.B. Parreira, A.R.K. Kobayashi Jr, O.B. Silvestre,

Influence of Portland cement type on unconfined compressive strength and linear expansion of cementstabilized phosphogypsum. Journal of Environmental
Engineering, Volume 129 (2003), pp. 956960.
L. Plyatsuk, E. Chernish, Intensification of the anaerobic microbiological degradation of sewage sludge and
gypsum waste under bio-sulfidogenic conditions.
Journal of Solid Waste Technology and Management,
Volume 40, No 1 (2014), pp. 10 23.
S.G. Richardson, Vegetation and overburden cover on
phosphogypsum: effects on Radon emission, runoff water quality, and plant uptake of Fluoride and Radium.
The Challenge of Integrating Diverse Perspectives in
Reclamation - Zamora, Connolly (1993), pp. 9098.
Rafael Prez-Lpez,, Jos Miguel Nieto, Israel LpezCoto, Juan Luis Aguado, Juan Pedro Bolvar, Mara
Santisteban,
Dynamics
of
contaminants
in
phosphogypsum of the fertilizer industry of Huelva
(SW Spain): From phosphate rock ore to the environment. Applied Geochemistry, Volume 25 (2010), pp.
705715.
[14]
[15]
[16]
[17]
J.H. Potgieter, S.S. Howell-Potgieter, A plant investigation into the use of treated phosphogypsum as a setretarder in OPC and an OPC/fly ash blend. Minerals
Engineering, Volume 14, Issue 7 (2001), pp. 791795.
A. Kazilinas, M. Baaus, Investigation of technogen
gypsum neutralization and dehydration processes in the
environment of raised water steam pressure. The 9th International Conference Modern Building Materials,
Structures and Techniques. Selected papers. Vol III.
Held on May 16-18, (2007), Vilnius, Lithuania, available at http://leidykla.vgtu.lt/conferences/MBM_2007/1
pdf/Kaziliunas.pdf
M. B. Rajkovi, K. Karljikovi-Raji, G. T.
Vladjisavlevi, I. S. iri (Yugoslavia) Investigation
of radionuclides in phosphogypsum. Measurement
Techniques March, Volume 42, Issue 3 (1999), pp.
299305.
David P. Borris, Patricia W. Boody, Phoshpodypsum.
Proceedings of the International Symposium on
Phosphogypsum. Utilization and/or Disposal of
Phosphogypsum. Potential Barriers to Utilization: Final
report, Lake Buena Vista, Florida (1980), p. 575.
D.A. Mays, J.J. Mortvedt, Crop response to soil applications of phosphogypsum. Journal of Environmental
Quality, Vol. 15, No. 1 (1986), pp. 7881.
I.S. Alcordo, J.E. Rechcigl, Phosphogypsum and other
by-product gypsums. Soil amendments and environmental quality. (1995), pp. 365425.
J. Murphy, J.P. Riley, A modified single solution
method for the determination of phosphate in natural
waters. Analytica Chimica Acta, Volume 27, (1962)
pp. 3136.
A.I. Akn, S. Yesim, Utilization of weathered
phosphogypsum as set retarder in Portland cement.
Cement and Concrete Research, Volume 34 (4) (2004),
pp. 677680.
N. Degirmenci, A. Okucu, A. Turabi, Application of
phosphogypsum in soil stabilization. Building and Environment, Volume 42 (9) (2007), pp. 33933398.
Papastefanou, C., Stoulos, S., Ioannidou, A.,
Manolopoulou, M., 2006. The application of
phosphogypsum in agriculture and the radiological impact. Journal of Environmental Radioactivity, Volume
89 (2), pp. 188198.
J.M. Hird, Managing disposal of phosphogypsum and
clays at Texasgulfs North Carolina phosphate complex.
Chapter 42 - Environmental Aspects in Phosphate Beneficiation. Beneficiation of Phosphate: Theory and Practice. H. El-Shall, B.M. Moudgil, R. Wiegel: Editors.
Published by the Society for Mining, Metallurgy, and
Exploration, Inc. Littleton, Colorado (1993), pp. 447
454.
S. Folek, B. Walawska, B. Wilczek, J. Miskiewicz,
Use of phosphogypsum in road construction. Polish
Journal of Chem. Tech., Volume 13, No 2 (2011), pp.
1822.
E.Y. Chernish, L.D. Plyatsuk, Patent of Ukraine for
invention 103087, Process for the treatment of organic
wastes with removal of heavy metals; applicant and re-
VOLUME 42, NO. 2
MAY 2016
[18]
[19]
[20]
[21]
cipient patent SSU. Published on 10.09.2013, bul.

17/2013.
M. Ramirez, J.M Gmez, and D. Cantero, Removal of
hydrogen sulphide by immobilized Thiobacillus
thioparus in a biofilter packed with polyurethane foam.
Bioresource Technology, Volume 100, Issue 21 (2009),
pp. 49894995.
P. Ravichandra, Gopal Mugeraya, A. Gangagni Rao, M.
Ramakrishna and Annapurna Jetty, Isolation of
Thiobacillus sp from aerobic sludge of distillery and
dairy effluent treatment plants and its sulfide oxidation
activity at different concentrations, Journal of Environmental Biology, Volume 28(4) (2007), pp. 819823.
THIOPAQ Bio-Desulfurization Process. Cameron.
Printed in USA, 07/10 TC9814-047 (2010), p. 2.
Yang, Y.A. and Allen, E.R. Biofiltration control of
hydrogen sulfide: Design and operational parameters,
[22]
[23]
[24]
[25]
J. Air Waste Manage. Assoc, no 44 (1994), pp. 863

868.
M. Ramirez, J.M Gmez, Germn Aroca and D.
Cantero, Removal of hydrogen sulphide by immobilized Thiobacillus thioparus in a biofilter packed with
polyurethane foam. Bioresource Technology. Volume
100, Issue 21 (2009), 49894995.
Carrier Byoung-Gi Park, Won Sik Shin1, and Jong Shik
Chung, Simultaneous Biofiltration of H 2 S, NH 3 and
Toluene using an Inorganic/Polymeric Composite. Environ. Eng. Res., Vol. 13, No. 1 (2008), pp. 1927.
W. Vishniac, M. Santer, The Thiobacilli. Bacteriological Review, 21 (1957), pp. 185213.
General Microbiology, 7th edition. Hans G. Schlegel
(Translated Margot Kogut). Cambridge University
Press (1993), pp. 233234.
115
EVALUATION AND QUANTIFICATION OF HOUSEHOLD WASTES

SENT FOR RECYCLING THROUGH A SELECTIVE WASTE
COLLECTION SYSTEM
Vanessa A. Mantovania, Sandro D. Mancinia *, Thalita R. Barrosa, Jos L. Ferrazb
a
UNESP Univ. Estadual Paulista, Campus Experimental de Sorocaba

b
Faculdade de Tecnologia de Sorocaba
Av. Trs de Maro, 511 - Alto da Boa Vista. CEP: 18087-180
Sorocaba, SP, Brazil
Tel.: + 55 15 3238 3409; Fax: + 55 15 3228 2842
mancini@sorocaba.unesp.br
ABSTRACT
Studies on solid waste production and composition can provide information to underpin the
implementation and/or expansion of selective waste collection systems. This paper presents
data about household solid wastes disposed of in the municipal sanitary landfill and collected
by the municipal selective waste collection system of Sorocaba, state of So Paulo, Brazil,
which was established in 2007. Discards of household solid wastes in the municipality were
found to average 612 g/person/day and the population that participates in the selective waste
collection program has the potential to ensure that 30% of its household solid wastes do not
end up in the landfill. However, this represents only 2.6% of total household solid wastes generated in the municipality. Excluding rejects (6.5 wt.% of the total selectively collected wastes),
a total daily amount of recyclable or reusable household wastes of 11 tons and 301 m3 is separated, with paper and plastics standing out as the most frequent items (about 53 wt.% and 79
vol.%). If this waste collection system and its current productivity were maintained and all possible household wastes generated in the city sent for recycling, todays infrastructure (space,
trucks and workers) would have to increase approximately 11-fold.
Keywords: Household wastes; selective waste collection; composition; waste generation per
capita; paper; plastics
INTRODUCTION
Knowledge about the wastes produced by a population,
such as composition and average quantity generated per person per day, can be useful for planning public policies aimed
at minimizing these wastes, including selective waste collection programs (Fehr, 2014; Mbeng et al., 2011; Meuser et al.,
2011; Papachristou et al., 2009; Qu et al., 2009; Thitame et
al., 2010; Zeng et al., 2005).

Selective waste collection programs are extremely helpful
for any society wishing to recycle its wastes (Pacheco et al.,
2012). However, in Brazil these systems have proved ineffective in preventing the final disposal, usually in landfills, of
more than 15% of total wastes generated (Fehr et al., 2009).
In 2010, Brazils federal government passed a law making
it obligatory for every state and municipality to establish
__________________________________________________
116
VOLUME 42, NO. 2
MAY 2016
quantifiable targets for reducing wastes discarded in landfills

and increasing the volumes sent for recycling through selective waste collection. These targets must be included in the
countrys state and municipal master plans for the coming 20
years, and were to be presented in 2014 (Brazil, 2010). The
establishment and particularly the achievement of these targets in Brazil is an urgent matter, since only 2.2% of Brazilian wastes were recycled and/or composted in 2008 (MMA,
2011). This number is very low compared to other countries
such as Malaysia (5%), Japan (40%), United States (33,6%)
and Europe as a whole (41%) (Eurostat, 2011; Moh & Manaf,
2014; USEPA, 2009).
This Brazilian law states the need for diagnosing the current situation of solid wastes in order to underpin actions and
programs aimed at achieving the established targets (Brazil,
2010). In this regard, waste composition (particularly by
weight) is an essential factor in understanding the problem.
Ten studies reported in the literature about the characterization of wastes were selected for a comparative analysis.
Each of these studies is based on a different methodology and
also has different objectives. In four studies, the samples
were described in volumetric terms (up to 2 m3 or garbage
bags collected directly from 47 to 300 households), and in six
others they were described in terms of weight (minimum of 2
kg and maximum of 400 kg per sampling). The number of
samples varied from 2 to 160 and the number of separated
items from 7 to 53. The duration of these studies was a minimum of two weeks and a maximum of one year, and the sites
from which the samples were obtained were usually landfills,
although in some cases they were obtained from garbage
collection trucks and, as mentioned earlier, directly from
households (Aguilar-Virgen et al., 2013; Bernache-Prez et
al., 2001; Bolaane & Ali, 2004; Eisted & Christensen, 2011;
Mancini et al., 2007; Meuser et al., 2011; Papachristou et al.,
2009; Qu et al., 2009; Thitame et al., 2010; Zeng et al.,
2005).
This paper reports on a study involving the characterization of household wastes in a Brazilian municipality where a
selective waste collection program was established in 2007.
To this end, per capita data is presented about the quantity of

wastes destined for recycling and for final disposal. The municipalitys recycling potential was quantified based on the
differences between disposal in areas covered and not covered by the program. With these results, it is possible to establish the real range, potential and problems of the system,
thereby facilitating the establishment of feasible targets for
the expansion of selective waste collection in the municipality, which will become obligatory in all Brazilian municipalities starting in 2014.

Sorocaba covers an area of 456 km2 and is located at the
geographic coordinates of 23 30 22 latitude South and 47
27 21 longitude West, in the state of So Paulo, Brazil. Its
population in 2011 was 593,775 (IBGE, 2010).
The municipality has a mixed garbage collection service,
called ordinary collection, which collects wastes usually three
times a week throughout its territory. At the time of this research, the citizens were instructed to bag their ordinary garbage and place it in containers (usually 240 L) distributed
along the citys sidewalks (Figure 1).
The selective waste collection service, however, does not
cover the entire municipality. It operates on a door-to-door
basis once a week and is carried out by workers organized in
cooperatives (scavengers association), who receive these
wastes directly from homeowners participating in the program. The recyclable and marketable wastes are sold to recyclers and the rest is considered rejects and is landfilled.
Figure 2 depicts a flow chart of the possible destinations
of the household solid wastes in Sorocaba.
To achieve the objectives of this research, the city was
divided according to the 6 regions of selective waste collection established by the local government (Simes et al. 2011).
Members of the research team then joined the workers on a
normal workday in a neighborhood where the cooperative
FIGURE 1
Example of containers distribution (blue) in a street in Sorocaba in 2011
117
FIGURE 2
Flow chart of the possible destinations of the household solid wastes in Sorocaba
considered that selective waste collection worked well.
On the visits to these six neighborhoods, 91 households
that participated in the selective waste collection program
were identified randomly. The wastes of the people collaborating with the selective collection program were collected
for subsequent sorting. The residents of the selected households were asked to point out the container in which they
deposited their ordinary wastes and indicate the other households that used it, in order to determine how many people
used each container. A total of 4 containers were identified in

each of the 6 neighborhoods covered by the selective waste
collection service. Before the ordinary garbage collector truck
arrived, the research team collected all the contents of the
containers for subsequent weighing.
Taking into account the areas covered by the cooperatives,
they were asked to draw up a list of the neighborhoods not
covered by selective waste collection service. One
neighborhood per region was then selected (Table 1), aiming
TABLE 1
Socioeconomic Class considered for each of the sampled neighborhoods.
Region
1
Socioeconomic Class
Selective Collection Attended Neighborhood
Medium
Selective Collection Non-Attended Neighborhood
Low
High
High
Medium
Low
Low

5
118
Medium
Low
Medium
Medium
Medium
VOLUME 42, NO. 2
MAY 2016
to cover the largest possible area and every possible socioeconomic class based on the So Paulo Social Vulnerability
Index (SEADE, 2010).
Like in the previous case, households were chosen randomly (total of 51) and the residents were interviewed. A
total of 2 containers were identified in each of the 6 neighborhoods not covered by the selective waste collection service. Again, before the ordinary garbage truck arrived, the
research team removed the contents of the containers for
weighing.
The samples were collected from the 12 neighborhoods in
the first half of July 2011 and the wastes were sorted no later
than 12 hours after collection. At least 100 kg of waste from
each of the neighborhoods (from the 6 containers plus selec-
tively collected wastes) were sampled. This amount was

similar to that sorted in previous studies (Barros et al., 2014;
Mancini et al., 2007) and followed the ASTM D5231
standard (2008).
Figure 3 shows the spatial distribution of the ordinary
garbage containers selected for the research, as well as the
division of the citys six large sectors of selective waste collection. To ensure that only wastes generated by households
would be measured, containers and homes close to commercial establishments, offices and industrial plants were avoided.
Table 1 describes the socioeconomic class considered for
each of the sampled neighborhoods.
As can be seen in Table 1, 8 out of 12 of the sampled
FIGURE 3
Spatial distribution of the sampled containers (6 per region) in the municipality of Sorocaba: () neighborhoods without
selective waste collection (2 containers); () neighborhoods with selective waste collection (4 containers)
119
neighborhoods were considered high or medium classes. This

sampling is considered adequate once, among Brazils 5,565
cities in 2010, Sorocaba ranks 47th on the Human Development Index, i.e., 0.798 (UNDP, 2013).
The selectively collected wastes were sorted and then
weighed in containers of known volume. A Micheletti MIC100 electronic digital scale (max 100 kg; 20 g graduation)
was used for quantities larger than 50 g, and a Plenna MEA
6100 portable electronic digital scale (max 5 kg; 1 g
graduation) for smaller quantities. The wastes were divided
into 41 different items in terms of weight and volume.
The following equations can be used to calculate some
important results about a selective waste collection system,
and were applied to this one.
a) The fraction potentially diverted from landfilling by people participating in the system, % diverted p , in weight (w)
and volume (v):
(w per capita selective collection *Wr)] x 100%

w per capita total
where: Wr = fraction of wastes collected selectively and sent

for landfilling (not marketable and not reusable) = fraction
rejected
f) Amount of household wastes discarded daily (D).
D= w per capita average *P
(1)
% diverted p (v) = (v per capita selective collection
n / v per capita total )*100
(2)
For this calculation, the total waste collected from the containers was divided by the number of people using each container, and by the number of days between the preceding
collection to the collection by the research team (usually two
days). This value was added to the quantity collected selectively divided by the number of people in the households
participating in the selective collection system and by the
number of days between the last selective waste collection
and the one carried out by the research team (usually 7 days).
b) Total amount of wastes collected by the selective collection system (wc):
w c = w cm *100%/W c
(7)
where: w per capita average = weight of wastes generated per person in the entire city;
P = population of the municipality.
g) Total household wastes generated per day in the city,
which would be sent to the sanitary landfill if the selective
waste collection system were not operating (w total ):
w total = w
% diverted p (w) = (w per capita selective collection

/ w per capita total )*100
(6)
landfilled
+ wc wr
(8)
where: w landfilled = amount of wastes landfilled per day;

wr = weight of rejects per day from selectively collected
waste.
h) Fraction of wastes diverted from landfilling by selective
collection:
R diverted = ( w c - w r) / w total (%)
(9)
i) Total amount of wastes that could be diverted from landfilling if the entire population participated in the selective
waste collection program (w diverted )
w diverted max = % diverted * w total
(10)
j) Multiplication factor to be applied in the hypothetical case
of the entire population participating in the selective waste
collection (F):
(3)
F = w diverted max / w c
(11)
where Wc = fraction of marketable waste;

wcm = amount of marketed waste.

c) Number of people participating in the selective waste
collection system (n):
n = w c / w per capita selective waste collection
(4)
d) Estimated volume collected daily (v c )

v c = n*v per capita selective waste collection
(5)
e) Fraction really diverted from landfilling by participants in

the system, %diverted, in weight:
% diverted =
[w per capita selective waste collection
120
Per capita data

Table 2 shows the daily generation of household wastes
per person in the neighborhoods covered and not covered by
the selective collection system in each region, and average
wastes generated by the city.
In Table 2, note that each person in the neighborhoods not
covered by selective collection generated 708 g/day of
household wastes and almost 6 liters of solid wastes (discarded in the landfill). In the neighborhoods covered by the
selective system (last four columns in Table 2), about 482 g
and 3.2 liters/person/day were discarded, but it is not known
VOLUME 42, NO. 2
MAY 2016
TABLE 2
Average daily household wastes generated per person in neighborhoods covered and not covered
by selective waste collection in each region, and average of the regions.
Neighborhoods Not Covered by Selective

collection
Ordinary collection
Ordinary collection
Neighborhoods Covered by Selective collection

Selective collection
REGION
Weight
(g/person.day)
Volume
(L/person.day)
Weight
(g/person.day)
Volume
(L/person.day)
Weight
(g/person.day)
Volume
(L/person.day)
409.6
3.4
308.1
2.9
176.8
4.8
1,525.7
13.0
607.4
4.2
398.4
6.3
702.9
5.8
476.0
2.4
262.2
6.7
471.6
3.6
504.5
3.2
93.6
3.1
571.4
4.4
580.8
3.9
92.0
2.5
566.9
4.4
414.4
2.9
320.3
13.3
Sorocaba
708.0
5.8
481.9
3.2
223.9
6.1
if these values represent only residents participating in the

selective collection system, since the same container could
contain wastes from households participating and not participating in the selective system. In the case of participants,
another 224 g and 6 liters of wastes per person were sent to
cooperatives.
Table 3 quantifies the wastes generated per person in each
of the regions and the average generated by the city. These
quantities were calculated based on the average per person in
the regions covered and not covered by the selective collection system. However, in the neighborhoods covered by
selective waste collection, it would be incorrect to add the
average values of the ordinary and selective collections
(shown in Table 2), since not everyone using the containers
participated in the selective collection system. Therefore, the

daily total wastes destined for ordinary collection and for
selective waste collection were added and then divided by the
total number of people using the containers.
In Table 3, note that, on average, each resident of
Sorocaba discarded about 612 g and 5 liters of household
wastes/day. The 612 g/person/day quantified in Sorocaba
exceed the amount of household wastes generated per person
in other cities for which there are similar data, such as Beijing
(China) 230 g/person/day (Qu et al., 2009), Gaborone
(Botswana) 330 g/person/day (Bolaane & Ali, 2004),
Sisimiut (Greenland) 365 g/person/day (Eisted & Christensen, 2011), Phnom Penh (Cambodia) 487 g/person/day
(Seng et al., 2011) and Guadalajara (Mexico) 508
TABLE 3
Generation of wastes per person in each of the regions and in the entire city, and respective margins of error.
Region
Weight (g)
Error a (g)
Volume (L)
Error a (L)
396.4
188.1
4.3
2.1
1061.6
349.5
8.5
3.8
585.5
156.8
4.4
1.4
513.8
150.4
3.9
1.2
601.9
407.5
4.5
1.7
510.6
174.5
5.2
1.2
Sorocaba
611.6
190.95
5.2
1.4
Error= (t90%.SD)/n, where t90%=2.015; SD=standard deviation found; n=6).
121
g/person/day (Bernache-Prez et al., 2001). However, studies

in Erbil (Iraq) and Ensenad (Mexico) found higher values:
654 g/person/day and 779 g/person/day, respectively (Aziz et
al., 2011; Aguilar-Virgen et al., 2013).
At this point, it is relevant to point out differences, in
terms of socioeconomic classes in the regions of the city.
Region 2 shows a predominance of high-income households
(Table 1) that produce more wastes destined for both selective and ordinary waste collection (Tables 2 and 3), which
denotes their higher purchasing power. The other regions had
middle and low socioeconomic classes (regions 1, 3, 4 and 5)
or only middle class (region 6), with amounts ranging from
122
400g to 600g daily, which is much lower than the amounts of

over 1kg/day found in region 2. Also note that low class
neighborhoods (regions 4 and 5) covered by the selective
waste collection system were the ones that contributed the
least to the system. This may also be attributed to the lower
level of environmental awareness among less educated people, which is more common among lower socioeconomic
classes.
Studies about selective waste collection
VOLUME 42, NO. 2
MAY 2016
Table 4 describes the composition of selectively collected

wastes (by weight and volume), divided into 41 items, and
the estimated daily amount collected, expressed in weight and
volume. It should be noted that some collected wastes are
rejected because they are non-recyclable or reusable, or
because there is no market for this type of product in the
region, so they are sent to the landfill. Based on this parameter, the following wastes were considered rejects: food leftovers, fabrics and shoes in poor condition, mixed and metallized packaging (mainly aluminized plastic films), hazardous
wastes, rubber, thermoset materials and eventual items. Some
fabrics and shoes in good condition are not commercialized,

but are not rejects because they usually end up being
distributed. Fabrics in Good Condition were considered
clothes that could still be used, and Shoes in good condition
were the ones that were not only in good condition but also
paired. These reusable wastes were not included in the flow
chart of Figure 2.
In Table 4, note that, out of 41 separate items, 15 had
higher margins of error than their average, because of the
wide variation in samples of selectively collected wastes.
This can be explained by the different consumption and
TABLE 4
Composition of Selectively Collected Wastes in Sorocaba, divided into 41 items,
with their respective margins of error, and estimated wastes collected daily, in weight and volume
Composition
Estimated daily
Errora(%)
(wt.%)
collection (kg)
Food leftovers
0.3
0.3
35.3
Fabrics in good condition
1.8
2.0
211.6
Fabrics in poor condition
0.2
0.4
23.5
Shoes in good condition
0.6
1.2
70.5
Shoes in poor condition
0.3
0.4
35.3
White paper
4.9
6.3
564.2
Magazine paper
4.2
2.9
493.7
Newspaper
3.9
3.5
458.4
Kraft paper
2.2
3.3
258.6
Paper in poor condition
0.3
0.4
35.3
Cardboard
18.9
4.5
2,209.8
Long-life packaging
4.9
1.5
575.9
Mixed packaging (paper and plastic)
0.2
0.2
23.5
Metalized packaging
0.7
0.2
82.3
Aluminum cans
2.4
4.0
282.1
Aluminum (others)
0.0b
0.1
0.0
Steel cans
3.2
2.2
376.1
Other Ferrous Metals
3.0
5.2
340.9
Copper wire
3.9
7.8
458.4
Glass (unbroken)
16.2
12.6
1,892.4
Glass (broken)
0.8
0.7
94.0
PET (colorless)
4.6
2.7
540.7
PET (colored)
1.1
0.3
129.3
PET oil*
0.4
0.2
47.0
PET (thermoformed)
1.1
0.7
129.3
HDPE(film)
2.4
2.0
282.1
HDPE (rigid)
2.8
1.4
329.1
PVC
2.3
3.3
270.3
LDPE (film)
1.0
0.5
117.5
LDPE (rigid)
0.1
0.1
11.8
PP(film)
0.6
0.5
70.5
PP(rigid)
0.9
0.4
105.8
Expanded polystyrene (EPS)
0.3
0.2
35.3
Polystyrene (rigid)
0.4
0.2
47.0
Other Plastics (rigid)
0.6
0.5
70.5
Other Plastics (film)
0.4
0.3
47.0
Thermosets
0.1
0.2
11.8
Rubber
0.1
0.2
11.8
Electric and electronic wastes
3.7
4.9
434.9
Hazardous wastes
1.3
1.3
152.8
Eventual items
3.3
3.1
387.9
100.0
11,754
a
Error= (t90%.SD)/n, where t90%=2.015; SD=standard deviation found; n=6).
b
indexes lower than 0,05%
Items
Composition
(vol.%)
0.1
1.0
0.0b
0.0b
0.0b
2.0
3.2
3.4
1.6
0.2
27.3
7.9
0.1
2.4
0.2
0.2
1.2
0.2
1.1
2.6
0.1
8.3
2.3
0.8
2.7
5.6
4.1
2.9
5.7
0.4
3.2
1.5
1.5
0.7
0.9
0.5
0.0b
0.0b
0.9
0.1
3.2
100.0
Errora(%)
0.1
1.5
0.1
0.0b
0.0b
1.0
3.4
2.7
1.6
0.2
6.6
1.2
0.1
0.4
0.2
0.4
1.3
0.3
2.2
2.0
0.2
4.3
1.8
0.5
1.5
1.7
2.2
3.0
2.4
0.7
1.4
1.1
1.3
0.6
0.6
0.6
0.0b
0.0b
1.6
0.2
6.2
Estimated daily
collection (m3)
0.3
3.2
0.0
0.0
0.0
6.4
10.2
10.9
5.1
0.6
87.0
25.3
0.3
7.7
0.6
0.6
3.8
0.6
3.5
8.3
0.3
26.6
7.4
2.6
8.6
17.9
13.1
9.3
18.2
1.3
10.2
4.8
4.8
2.2
2.9
1.6
0.0
0.0
2.9
0.3
10.2
320
123
disposal habits of users and also by the fact that some items,
among all the samples, came from only one region, e.g.,
shoes in good and poor condition, thermoset materials and
copper wire. Another reason for this outcome is the greatly
detailed division of the items, suggesting that separating the
wastes into fewer items could minimize this problem (see the
discussion of Table 5). Larger quantities of samples or larger
waste samplings could result in larger numbers of each item,
thereby reducing the margin of error.
As Table 4 indicates, 8.9 wt.% of the wastes are not
marketed and 6.5 wt.% and 5.9 vol.% of selectively collected
wastes, corresponding, respectively, to about 760 kg/day and
19 m3/day, are considered rejects. Studies on selective waste
collection in Brazil consider that reject indices below 7 wt.%
are low (Besen, 2006). Sorocaba is located in a region with a
high amount of industries and the recyclable market around
the city can be considered one of the most important of the
country. So, the number of items that are marketable in Sorocaba is certainly higher and the reject rate is probably lower
than in most parts of Brazil. According to a survey conducted
in 18 cities in every region in Brazil in 2012 (Sorocaba was
not included), the reject rate was higher than 17% (CEMPRE,
2013).
These rejects include hazardous wastes, since they may be
reactive, flammable, corrosive, toxic and/or pathogenic
(ABNT, 2004). Typical examples of such wastes found in
samplings are fluorescent lamps, which should be sent to
recycling plants licensed for this purpose. However, Sorocaba
(like the majority of Brazilian cities) has no specific
collection program for hazardous wastes.
In contrast to the 6.5% of rejects in terms of mass, 93.5%
of wastes, i.e., approximately 11 tons/day (and 301 m3/day),
can be considered recyclable or reusable, as indicated in Table 4.
In addition to the wastes that are not marketed, there are
some recyclable wastes that are difficult to sell seasonally or
always, such as PPfilm, Expanded polystyrene (EPS), Other

Plastics (rigid and film) and PET thermoformed (blisters,
boxes, etc.). Conversely, however, PP rigid , PS rigid and other
types of PET are relatively easy sell in the region. Like the
selectively collected waste in So Carlos (Brazil), in Sorocaba PET is also the most discarded and marketed type of plastic (in terms of weight and volume), usually in the form of
bottles, including cooking oil bottles (PEToil) (Matos &
Schalch, 2007).
The volume-related values in Table 4 are interesting for
an assessment of the physical space required for handling
wastes. Sorocabas cooperatives collect a total of 299 m3/day
of marketable wastes, which must be compacted and/or
stored uncompacted and sold as soon as possible. For example, cardboard and colorless PET, the bulkiest items (88
m3/day and 27 m3/day, respectively) are compacted on site
immediately after sorting.
Table 5 describes the composition of recyclable and reusable wastes of Sorocaba, divided into 7 items (instead of the
41 items listed in Table 4), and the estimated weight and
volume of wastes collected daily. As mentioned in our comments about the results listed in Table 4, grouping several
types of waste in a single group, such as plastics, allowed for
the reduction of errors (lower than the average).
In Table 5, note that paper (including cardboard) is the
item with highest incidence in recyclable waste, in terms of
mass. Recyclable paper corresponds to about 4 tons or 120
m3/day, independently of its classification, by the sorting
team, as being in bad (with excessive dirt and moisture) or
good condition. Paper is also the most common type of waste
in selective waste collection in Mashhad, Iran (51%)
(Farzadkia et al., 2012).
Another material that stands out is plastics, which appears
in second place in terms of mass and first place in terms of
volume, with approximately 2.2 tons and 131 m3 /day in the
form of films (bags) and rigid plastics (everything else). It
TABLE 5
Composition of Wastes Collected Selectively in Sorocaba, divided into 7 items,
with their respective margins of error, and estimated weight and volume of wastes collected daily.
Items
Composition
(wt.%)
Error *(%)
Estimated daily
collection (kg)
Composition
(vol.%)
Error *(%)
Estimated daily
collection (m3)
Paper (Including cardboard)
34.2
10.4
4,019.9
37.6
5.1
120.3
Plastic (Films and Rigid)
19.0
3.3
2,233.3
41.1
7.2
131.5
Glass
16.9
12.3
1,986.4
2.7
1.9
8.6
Metal
12.4
7.6
1,457.5
2.9
1.9
9.3
Long-life packaging
4.9
1.5
575.9
7.9
1.2
25.3
Electric and Electronic wastes
3.7
4.9
434.9
0.9
1.6
2.9
Shoes and Fabrics in good condition
2.4
2.3
282.1
1.0
1.5
3.2
Total
93.5
10990.0
94.1
301.1
Error= (t 90% .SD/n, where t 90% =2,015; SD=standard deviation found; n=6)
124
VOLUME 42, NO. 2
MAY 2016
can be estimated that 42.5 g/day per person of plastics is sent

to the selective waste collection system of Sorocaba, which is
more than the 23.1 g/day per person (only plastic packaging)
found by Chaerul et al. (2014) in Bandung City (Indonesia).
The combined amount of paper and plastics represents almost
80 vol.% and more than 50 wt.% of the wastes collected
selectively in Sorocaba. In the survey conducted in 18
Brazilian cities, paper represented 45.9 wt.% and plastics
15.6 wt.%, i.e., 61.5% of the total recyclable wastes in 2012
(CEMPRE, 2013).
Table 6 presents the results of Equations 1 to 11, which
are pertinent to the discussion presented herein.
As can be seen in Table 6, the current selective waste
collection system has the potential to divert almost 32 wt.%
(Eq. 1) and 66 vol.% (Eq. 2) of the waste generated per person/day away from landfilling. However, this volume should
be viewed with caution, since the two waste collection systems involve completely different types of packaging and
compaction. Thus, in a possible expansion of the selective
waste collection system, this percentage cannot be used for
sizing the remaining ordinary collection and much less for

calculating the volume occupied in landfills, which is determined not only by the composition but also by the degree of
decomposition of wastes, among other factors (Pulat &
Yukselen-Aksoy, 2013).
In Table 6, also note that when the rejects are not considered, the real diversion index is 30% (Eq. 6) in the case of
participants in the selective waste collection system, which
represent 9.2% of the population (Eq. 4). Considering the
entire population and the total waste produced (Eq. 8), it can
be concluded that the real diversion index is 2.6% (Eq. 9).
This result is in agreement with that reported by Fehr et al.
(2009), who claim that traditional selective waste collection
systems can divert, at the most, 15% of the garbage destined
for landfilling. However, the percentage of household wastes
sent for recycling in Sorocaba is higher than the overall Brazilian percentage of 1.4%. On the other hand, the absence of
composting in Sorocaba is in contrast to the existence of this
activity in the country, although Brazils national composting
index is low (0.8%) (MMA, 2011).
TABLE 6
Results of Equations 1 to 11 and explanations about how some parameters were obtained and/or determined.
Eq.
Parameter
Result
Observations
Observation / Reference
% diverted p (w)
32
w per capita selective collection = 223.9 g
Table 2, last line
w per capita total = 705.8 g*
% diverted p (v)
66
* 705.8 g = 481.9 g + 223.9 g
4th and 6th column
v per capita selective collection = 6.1 L
Table 2
v per capita total = 9.3 L*
5th and 7th column
* 9.3 L = 3.2 L + 6.1 L

3
wc
11,754 kg/day
(Table 4,4th column, last line)
Wc = 0.911
Table 4
w cm = 10,708 kg/day
Campos, 2012
52,497 inhabitants
w per capita selective collection = 223.9 g (Table 2, 6th

column)
vc
320 m3/day
v per capita selective collection = 6.1 L
(Table 4,7th column, last line)
(Table 2, 7th column)
% diverted
30
Wr = 0.065
Table 4
363.15 t/day
w per capita average = 611.6 g
Table 3
P = 593,775 inhabitants
IBGE, 2010
CETESB, 2012
w total
422,4 t/day
landfilled
= 411.4 t/day
w r = Wr* wc = 764 kg /day

9
% R diverted
2.6
10
w diverted max
127 t/day
11
11
Table 4
125
Table 6 also shows that the entire city discards 363

tons/day of household solid wastes (Eq. 7) and that, were the
entire population to participate in the selective waste collection program like the current participants do, it would be
possible to prevent about 127 tons/day of wastes from being
landfilled (Eq. 10), which would represent an 11-fold expansion of the program (Eq. 11). Based on the density of 36.6
kg/m3 (wc/vc), this expansion would represent about 3,470
m3/day of wastes, also 11 times higher than the current figure.
This estimate excludes the problems discussed earlier about
volumetric calculations (e.g., forms of compaction).
This means that if the same selective waste collection
system and its current productivity were maintained, the
current infrastructure of space, number of trucks and cooperatives members would also have to increase by approximately
11-fold for the system to service the entire city. This number
is probably overestimated, once optimization studies can be
carried out. For instance, if the whole city participates in the
selective collection, best-route software can be used to minimize the number of trucks and the fuel needs. Also, a financial analysis can be considered as a preliminary study for a
further strategic plan for this expansion.
CONCLUSIONS
It was found that, on average, each person in Sorocaba,
SP, Brazil, discards about 612 g/day of household solid
wastes. Excluding rejects (6.5% of the total), that are destined
incorrectly to the selective waste collection system, the
population that contributes to the selective waste collection
program has a potential to prevent 30% of their household
solid wastes from being landfilled. This population (9.2% of
the total) can divert from landfilling about 11 tons/day and
301 m3/day of recyclable or reusable wastes. This represents
only 2.6 wt.% of total household solid wastes generated in the
municipality. The most abundant item in the selectively collected waste is paper, followed by plastics (the former representing 34.2 wt.% and 37.6 vol.%, and the latter 19 wt.% and
41.1 vol.%).
All the data obtained in this survey can be used in planning the selective collection of increasingly large quantities
of wastes, contributing to compliance with the Brazilian law
that requires the establishment and achievement of targets. It
also can contribute toward more accurate calculations of
physical space, equipment, workers and transportation requirements, as well as better possibilities for selling or even
processing these wastes.
In this regard, if this selective waste collection system and
its current productivity were maintained, the current infrastructure of space, number of trucks and number of members
of the cooperatives would have to be increased by
approximately 11-fold in order to send for recycling all the
possible household wastes generated in the city.
Despite the results being specific for Sorocaba, they can
be seen as typical for Brazilian cities of similar population
and HDI, as comprehensive selective collection is not a reality throughout the country. Other cities in the world can produce similar percentage results when they present equivalent
126
situation in terms of population, HDI and have incipient selective collection systems. Therefore, this study can be useful
for all types of cities, as it represents a way of calculating,
estimating and obtaining important parameters regarding
selective collection.
ACKNOWLEDGEMENTS
The authors are indebted to Bruna L. Bento, Fernanda S.
Agostini, Guilherme Nagamatsu, Joo Victor R. Borgheresi,
Lucas F. C. Ribeiro, Nathlia S. S. Lima and Yuri Dadauto,
as well as the Sorocaba Municipal Administration and the
members of the Cooperatives Reviver, Catares, Ecoeso and
Coreso, for their assistance during this research. The authors
also thank Sanex Solues Ltda for its financial support.
REFERENCES
ABNT - Associao Brasileira de Normas Tcnicas (2004).
Resduos Slidos: Classificao, NBR 10.004 (Brazilian
Association of Techinical Stardards. Solid Wastes:
Classification). ABNT, Rio de Janeiro, Brazil.
Aguilar-Virgen, Q., P. Taboada-Gonzlez, S. Ojeda-Bentez
(2013) Seasonal analysis of the generation and composition of solid waste: potential use a case study. Environmental Monitoring and Assessment, Volume 185, pp.
4633445.
ASTM - American Society for Testing and Materials (2008).
ASTM D5231 92 - Standard Test Method for Determination of the Composition of Unprocessed Municipal Solid Waste. ASTM, Philadelphia, USA.
Aziz, S.Q., H.A., Aziz, M.J.K. Bashir, M.S. Yuosoff (2011)
Appraisal of domestic solid waste generation, components, and the feasibility of recycling in Erbil, Iraq.
Waste Management & Research, Volume 29, pp. 880
887.
Barros, T.R., S.D. Mancini, J.L. Ferraz (2014) Composition
and Quantification of the Anthropogenic and Natural
Fractions of Wastes Collected from the Stormwater
Drainage System for Discussions About the Waste Management and People Behavior. Environment, Development and Sustainability, Volume 16, pp. 415429.
Bernache-Prez, G., S. Snchez-Clon, A.M. Garmendia, A.
Dvilla-Villareal, M.E. Snchez-Salazar (2001) Solid
Waste Characterisation Study in the Guadalajara Metropolitan Zone, Mexico. Waste Management & Research,
Volume 19, pp. 413424.
Besen, G.R. (2006) Programas municipais de coleta seletiva
em parceria com organizaes de catadores na Regio
Metropolitana de So Paulo (Selective collection systems
in So Paulo Metropolitan Region). Masters Thesis.
Faculdade de Sade Pblica da USP, So Paulo, Brazil.
Bolaane, B., M. Ali, (2004) Sampling Household Waste at
Source: Lessons Learnt at Gaborone. Waste Management
& Research, Volume 22, pp. 142148.
Brazil (2010). Poltica Nacional de Resduos Slidos
VOLUME 42, NO. 2
MAY 2016
(National Policy for Solid Wastes). Presidncia da

Repblica (2010): http://www.planalto.gov.br/ccivil_03/_
ato2007-2010/2010/lei/l12305.htm (May 18, 2014).
Campos, C. (2012) Coleta seletiva deve atingir 100% da
cidade (Selective Collection must reach 100% of the city).
From Cooperativismo (2012): http://www.cooperativismo.
org.br/cooperativismo/noticias/noticia.asp?id=17254
(May 18, 2014).
CEMPRE - Compromisso Empresarial para a Reciclagem
(Business Commitment to Recycling). (2013) Pesquisa
Ciclosoft (Ciclosoft Research). From CEMPRE (2013):
http://www.cempre.org.br/ciclosoft_2012.php (May 18,
2014).
CETESB - Companhia Ambiental do Estado de So Paulo
(Environmental Company of So Paulo State) (2012).
Inventrio Estadual de Resduos Slidos Domiciliares
2012 (State Report of Household Solid Waste 2012).
From CETESB (2012): http://www.cetesb.sp.gov.br/solo
/publicacoes-e-relatorios/1-publicacoes-/-relatorios (May
18, 2014).
Chaerul, M., A.R. Fahruroji, T. Fujiwara, (2014) Recycling
of plastic packaging waste in Bandung City, Indonesia.
Journal of Material Cycles and Waste Management, Volume 16, pp. 509518.
Eisted, R., H. Christensen, (2011) Characterization of
household waste in Greenland. Waste Management, Volume 31, pp. 14611466.
Eurostat - Statiscal Office of the European Union. Municipal
Wastes Statistics. From: Eurostat (2011): http://epp. eurostat.ec.europa.eu/statistics_explained/index.php/Municipal
_waste_statistics (May 18, 2014).
Farzadkia, M., S. Jorfi, A. Hamideh, M. Ghasemi, (2012)
Evaluation of dry solid waste recycling from municipal
solid waste: case of Mashhad city, Iran. Waste Management & Research, Volume 30, pp. 106112.
Fehr, M. (2014) The management challenge for household
waste in emerging economies like Brazil: Realistic source
separation and activation of reverse logistics. Waste
Management & Research, Volume 32, pp. 3239.
Fehr, M., K.A. Sousa, M.Q. Santos, M.M.O. Domingues,
(2009) Using test communities to demonstrate household
waste separation procedures in Brazil. Environmental
Practice, Volume 11, pp. 2531.
IBGE - Instituto Brasileiro de Geografia e Estatstica
(Brazilian Institute of Geography and Statistics) (IBGE,
2010). Estimativas da Populao Residente nos
Municpios Brasileiros em 2011 (Population Estimates for
Brazilian Cities in 2011). From: IBGE (2010):
http://www.ibge.gov.br/home/estatistica/populacao/estima
tiva2011/POP2011_DOU.pdf (May 18, 2014).
Mancini, S.D., A.R. Nogueira, D.A. Kagohara, J.A.S.
Schwartzman, T. Mattos, (2007) Potential Recycling of
Urban Solid Waste Destined for Sanitary Landfills: the
Case of Indaiatuba, SP, Brazil. Waste Management &
Research, Volume 25, pp. 517523.
Matos, T.F.L., V. Schalch, (2007) Composio dos Resduos
Polimricos, Ps-Consumo, Gerados no Municpio de So
Carlos (Composition of Post-consumption Polymeric
Waste Generated in the Municipality of So Carlos).
Polimeros, Volume 17: pp. 346351.

Mbeng, L.O., T. Tudor, R. Fairweather, (2011) Survey of
Household Waste Generation and Composition to Drive
Strategy Development A Case of Three Residential Areas in Douala, Cameroon. The Journal of Solid Waste
Technology and Management, Volume 37, pp. 284296.
Meuser, H., K.S. Grewal, R. Anlauf, R.S. Malik, R.K.
Narwal, J. Saini, (2011) Physical Composition, Nutrients
and Contaminants of Typical Wastes Dumping Sites.
American Journal of Environmental Sciences, Volume 7,
pp. 2634.
MMA - Ministrio do Meio Ambiente (Ministry of
Environment) (2011). Verso Preliminar do Plano
Nacional de Resduos Slidos (First Version of the Solid
Wastes National Plan). From: MMA (2011):
http://www.mma.gov.br/estruturas/253/_publicacao/253_p
ublicacao02022012041757.pdf (May 18, 2014).
Moh, Y.C., L.A. Manaf, (2014) Overview of household
solid waste recycling policy status and chalenges in Malaysia. Resources, Conseervation and Recycling, Volume
82, pp. 5061.
Pacheco, E.B.A.V., L.M. Ronchetti, E. Masanet, (2012) An
overview of plastic recycling in Rio de Janeiro. Resources, Conservation and Recycling, Volume 60, pp.
140146.
Papachristou, E., H. Hadjianghelou, E. Darakas, K. Alivanis,
A. Belou, D. Ioannidou (2009) Perspectives for Integrated Municipal Solid Waste Management in Tessaloniki,
Greece. Waste Manage Volume 29, pp. 115862.
Pulat, H.F., Y. Yukselen-Aksoy (2013) Compaction behavior of synthetic and natural MSW samples in different
compositions. Waste Manage Res, Volume 31, pp. 1255
61.
Qu, X, Z. Li., X. Xie, Y. Sui, L. Yang, Y. Chen (2009) Survey of Composition and Generation Rate of Household
Wastes in Beijing, China. Waste Manage, Volume 29,
pp. 261824.
SEADE - Fundao Sistema Estadual de Anlise de Dados.
ndice de Vulnerabilidade Social (State System of Data
Analysis. Social Vulnerability Rate). So Paulo: SEADE,
2010. Avalilable from: http://www.iprsipvs.seade.gov.
br/view/index.php?prodCod=2 [Acessed 18.05.2014].
Seng, B., H. Kaneko, K. Hirayama, K. Katayama-Hirayama
(2011) Municipal solid waste management in Phnom
Penh, capital city of Cambodia. Waste Manage Res, Volume 29, pp. 491500.
Simes, G.V.B., J.L. Ferraz, S.D. Mancini, S.H. Bonilla,
W.A. Bizzo, Coleta Seletiva como Instrumento de
Polticas Pblicas: a Experincia do Municpio de
Sorocaba-SP (Selective waste collection as a public policy
tool: the Sorocaba/SP experience). 3rd International Workshop Advances in Cleaner Production Proceedings. So
Paulo: UNIP, 2011. Available from: http://www.advances
incleanerproduction.net [Accessed 18.05.2014].
Thitame, S.N., G.M. Pondhe, D.C. Meshram (2010) Characterisation and Composition of Municipal Solid Waste
(MSW) Generated in Sangamner City, District
Ahmednagar, Maharashtra, India. Environ Monit Assess,
Volume 170, pp. 15.
127
UNDP - United Nations Development Program. Ranking do

IDH dos municpios do Brasil (HDI rank of municipalities
in Brazil). Braslia: UNDP, 2013. Available from:
http://www.pnud.org.br/IDH/Atlas2013.aspx?indiceAccor
dion=1&li=li_Atlas2013 [Acessed 18.05.2014].
USEPA - United States Environmental Protection Agency.
Municipal Solid Waste Generation, Recycling, and Disposal in the United States: Facts and Figures for 2008.
128
Washington: USEPA, 2009. Available from: http://www.

epa.gov/osw/nonhaz/municipal/pubs/msw2008rpt.pdf
[Accessed 18.05.2014].
Zeng, Y., K.M. Trauth, R.L. Peyton, S.K. Banerji (2005)
Chacterization of Solid Waste Disposed at Columbia
Sanitary Landfill in Missouri. Waste Manage Res,
Volume 23, pp. 6271.
VOLUME 42, NO. 2
MAY 2016
IMPERIAL SMELTING FURNACE SLAG AS FINE AGGREGATE

IN CEMENT CONCRETE MIXES
Dr. S.B Patil* Dr. A.K. Vyas** Dr. A.B. Gupta** Rashmi S. Patil***
Professor, Department of Civil Engineering
*Dr. Babasaheb Ambedkar Technological University, Lonere 402 103 INDIA
**Malaviya National Institute of Technology, Jaipur 302 017 INDIA
*** M.Tech student PDA college of Engineering Gulbarga INDIA
ABSTRACT
Imperial smelting furnace (ISF) slag a waste byproduct from pyrometullurgical process of Zinc
production has been utilized as fine aggregate in Cement concrete mixes. Experiments on the
utilization of an industrial waste in Cement concrete mixes were carried out as particle size distribution of the slag is very close to that of fine aggregate used for Cement concrete mixes. Use
of an industrial waste in cement concrete mixes of grades M25 and M30 was made for replacing fine aggregate by ISF slag in the range of 0% to 100%. The Engineering properties of these
mixes were evaluated by Compressive Strength. The chemical composition of ISF slag shows
presence of heavy metals and the environmental acceptability of such mixes was determined
by Toxicity characteristics leaching procedure (TCLP) test recommended by USEPA. This paper presents the results of Strength and TCLP tests. Experimental investigations show that up
to 60% replacement of sand by Imperial smelting furnace (ISF) slag satisfies both strength and
environmental acceptability.
Keywords: Byproduct, Concrete, Slag, Zinc, Environmental
INTRODUCTION
Metal bearing wastes are hazardous if leaching of heavy
metals like lead, cadmium etc. is beyond permissible limit
prescribed by different regulatory agencies. The lack of technical and financial resources and the regulatory control for
the management of hazardous wastes in the past has led to the
unscientific disposal of such wastes in India, which has posed
serious risks to human, animal and plant life.
Three main waste minimization strategies collectively
called the 3Rs are Reduction, Reuse and Recycle the waste
products. For producing useful materials to human society
first step of the management strategy is to see that production
of waste must be minimum, second step is to utilize the waste
as raw material for some other product/process, third step is
128
to recycle the waste into usable forms as construction materials. Such waste products offer not only a partial solution to
environmental and ecological problems but may also yield
many useful construction materials [1, 2, 3, 4]. The aim of
recycling waste is not only to make the desired end product
less expensive, but also to provide a blend of tailored properties of the desired product, which is suitable for various specified purposes. Most common technologies of waste utilization include Waste to Waste Encapsulation and Stabilization;
Vegetation Based Remediation Technologies; Waste in the
form of Composts; Incineration; Use of waste in Construction
Sector; Technology of Managing Solid and Hazardous
Wastes in Cement Kilns; Manufacture of Blended Cements;
and Optimization Technique and Waste Immobilization
Techniques [5].
VOLUME 42, NO. 2
MAY 2016
The waste immobilization techniques prevent the free

movement of contaminants in waste. Three major groups of
immobilization techniques are the temporary containment,
solidification/stabilization, and vitrification techniques. The
selection of a containment system is based on the type of
waste materials, geohydrological, and geoenvironmental conditions of site containing waste. Solidification/stabilization
techniques (S/S) technique can be either chemical or
nonchemical processes. Chemical methods are very popular
and cost effective as compared to other techniques. Solidification/Stabilization (S/S) technique is the best-demonstrated
available technology (BDAT) for environmentally safe disposal of inorganic hazardous waste materials. The Jarosite
waste from Hindustan Zinc Limited (HZL) has been solidified in the past with different combinations of binder/waste
ratio in cement concrete mixes. Ordinary Portland Cement
(OPC), Sulphate Resistant Portland Cement (SRPC) and their
combinations were used in concrete cubes. It was observed
that 40% OPC and 60% SRPC with replacement of 60% of
sand by Jarosite waste exhibited desired design strength [6].
The utilization of hazardous and polluting industrial wastes
is a challenge for environmental protection. It is possible to
take advantage of their useful properties and to neutralize
their harmful characteristics thereby obtaining valuable materials. Acid jarosite processing waste is considered hazardous
because of its high acidity and concentration of zinc, cadmium and copper. Al-Surface Cleaning Waste (ASCW) was
used as a neutralizer of JW acidity. The new material resulted
from mixtures of JW, DFS and ASCW with or without small
additions of CaO or OPC, possessed significant strength and
water resistance making it suitable for use as construction
material [7]. Performance evaluation of solidified stabilized
product using aggregate based stabilization and subgrade
treatments was evaluated by John and John as a part of recycling, which can reduce energy consumption and conserve
natural resources with added benefits of reducing the price of
construction [8]. Experimental studies on utilization of ISF
slags in cement concrete mixes indicate that ISF slag aggregate can be utilized in cement concrete mixes by partial replacement of fine aggregate [9, 10,11].
Production of Zinc and byproduct ISF Slag

Zinc is produced at Chanderia unit of Hindustan Zinc Limited (HZL), India. The primary production of Zinc is covered
under the hazardous category Schedule-1, category 7.3 of
hazardous waste (management and handling) rules 1989,
MOEF, India as amended in May 2003 [12]. ISF slag from
Zinc industries is produced in huge quantity as a waste byproduct in pyrometallurgical process. It is a metal bearing
waste The ISF slag generated at Chanderia unit of Hindustan
Zinc Limited (HZL) is mostly stored in a separate dump yard
with minor commercial disposal. Though the slag is vitreous
in nature, the toxic constituents are in sufficient concentrations to be hazardous due to mobility of the heavy metal ions
by weathering action [13]. Heavy metals like lead and zinc
are the prominent contaminants present in this slag. Concentration of lead beyond 5 ppm, and zinc beyond 500 ppm is
considered hazardous as per guidelines of Environment Protection Agency.
Objective of the study

The main objective of this research study was to investigate
the potential use of ISF slag as fine aggregate in cement concrete mixes in place of fine aggregate. The fine aggregate
was replaced by ISF slag in select grades of mixes in a range
of 20%-100% by weight and tested for strength properties.
Environmental acceptability of all the mixes that satisfied
strength requirements was tested through leaching tests as
per USEPA recommendations. The information embodied in
this paper is directed towards developing design basis for the
use of ISF slag as aggregate in cement concrete mixes.
EXPERIMENTAL INVESTIGATIONS
The experimental protocols included concrete mixes prepared by replacement of sand by ISF slag to the tune of 20%,
40%, 60%, 80% and 100% by weight. The specimens were
tested for compressive strength and leaching of metals after
28 days of casting. Strength of specimens has been assessed
through compression test. In order to check leaching of metals TCLP test has been conducted on all specimens.
Materials and Methods

Prior to mix design physical properties of aggregates, cement and ISF slag used in the present study were determined
as per relevant Bureau of Indian Standards (BIS) (better specify the numbers and put them in ref list).
Waste. ISF slag produced by pyro process at Chanderiya
Lead Zinc Smelter (CLZS), Rajasthan, India is an inert, siliceous and vitreous material. It is tapped from Imperial Smelting Furnace in a molten glassy phase at a temperature of 1150
to 1200o C and is cooled and granulated. The annual production of ISF slag is 1, 00,000 MT. This slag is stored in a separate dump yard. It typically has a particle size distribution
between coarse and fine sand, and a higher specific gravity of
3.72. ISF slag is a complex combination of lead (0.5%), zinc
(6-10%), silica (18-22%), calcium oxide (15-19%), alumina
(7-9%), iron oxide (29-36%) and magnesium (2-3%) [14].
Binder. Binders are used to reduce leaching of contaminants
by physical confinement or by chemical fixation of contaminants like heavy metal ions. These binders convert the hazardous wastes into an environmentally acceptable form. In
cases where the waste is extremely soluble or no suitable
chemical binder is found, the waste may be contained by encapsulation in some hydrophobic medium such as bitumen or
poly ethylene. Cement has been used as binder in the present
study for Solidification/Stabilization process of the waste.
Cement. Cement is widely used as binder for Solidification/Stabilization technique for treatment of inorganic wastes.
IMPERIAL SMELTING FURNACE SLAG AS FINE AGGREGATE IN CEMENT CONCRETE MIXES
129
In the present study Ordinary Portland Cement (OPC) was

used to develop base line data for solidification/stabilization
of the slag. Its composition is much more consistent from
source to source. OPC was replaced by low strength Portland
Pozzolana Cement (PPC) in concrete mixes and possible
changes in engineering properties and environmental aspects
were observed. Properties of cement evaluated in the laboratory are shown in Table 1 and Table 2 for OPC and PPC respectively.
The results are well within the limits prescribed by Bureau
of Indian Standards (BIS).
Coarse Aggregate: Quartzite as coarse aggregate of 20 mm

nominal size was used in all the cement concrete mixes. Specific gravity and bulk density of coarse aggregate tested in the
laboratory were determined as 2.54 and 1529 kg/m3, respectively. Particle size distribution of coarse aggregate is shown
in Table 3.
The results of sieve analysis show that it is a well graded
aggregate.
Fine Aggregate: Fine aggregates used in the tests were
river sand and ISF slag. The results of sieve analysis of sand
are shown in Table 4.
Particle size distribution of fine aggregate shows zone III
grading as per Bureau of Indian Standards (BIS). Fineness
modulus of fine aggregate was determined as 2.54 in the laboratory.
Aggregate
Properties of coarse and fine aggregate used in cement
concrete mixes have been evaluated in the laboratory and
presented below.
TABLE 1
Properties of Ordinary Portland Cement
S. No.
Test
Test value
BIS
Limits
Ref. No.
1.
Initial setting time
72 min
>30 min
IS 4031 part 5
2.
Final setting time
245 min
< 600 min
IS 4031 part 5
3.
Compressive strength
44.5 N/mm2
> 43 N/mm2
IS 4031 part 6
TABLE 2
Properties of Portland Pozzolana Cement
S. No.
Test
Test value
BIS
Limits
Ref. No.
1.
Initial setting time
78 min
>30 min
IS 4031 part 5
2.
Final setting time
305 min
< 600 min
IS 4031 part 5
3.
34 N/mm2
> 33 N/mm2
IS 1489 part 1
TABLE 3
Sieve Analysis of Coarse Aggregate
130
S. No.
Sieve Size
Percentage passing
1.
25 mm
100
2.
20 mm
96.05
3.
12.5 mm
27.48
4.
10 mm
10.61
5.
6.3 mm
2.54
6.
4.75 mm
1.37
VOLUME 42, NO. 2
MAY 2016
TABLE 4
Sieve Analysis of Sand
S. No.
Sieve Size
Percentage passing
1.
4.75 mm
90.4
2.
2.36 mm
85.5
3.
1.18 mm
77.4
4.
600 micron
62.15
5.
300 micron
26.35
6.
150 micron
5.2
Solidification / Stabilization Technique: Solidification

/stabilization technique was adopted in order to utilize the
waste as fine aggregate in concrete mixes. The fine aggregate
(river sand) was replaced by ISF slag in varying proportion in
all the concrete mixes. Cement concrete mixes were designed
for M25 and M30 grade of concrete. Fine aggregate was replaced by ISF slag equal to the 20%, 40%, 60%, 80% and
100% by weight in all the concrete mixes. American Concrete Institute (ACI) method of mix design was followed in
the design of concrete mix in order to account for different
fineness modulii of ISF slag and sand used as fine aggregate.
Concrete mixes were designed with water cement ratio of
0.45 for M30 and 0.50 for M25 mixes. Mix proportions obtained for different concrete mixes are given in Table 5.
Specimens were cast with designated ratios without slag
(control) and with predetermined percentages of slag.
Cement Concrete Mix Design: American Concrete Institute
(ACI) method of mix design was adopted in view of maximum utilization of fine aggregate and with varying fineness
modulii of sand and slag as fine aggregate.
Compressive Strength Test: For compressive strength testing,
the cubes of dimension 150mm 150mm150mm. were prepared from the mixture containing different percentages of
ISF slag. The prepared specimens were cured for 7 and 28
days. The testing was performed on compression testing machine with the load being applied continuously without
shock. A sample fails when it loses its physical integrity by
falling apart. The total load at the failure of the test specimen
was recorded to the nearest 5 kN.
Leaching test: Water leachability of untreated waste is generally low and may mislead to conclusions that waste is not
hazardous. However, as leachability testing by the TCLP
(USEPA) [15] reveals that considerable leaching of heavy
metals occurs in an acetic-acid aqueous solution, the waste
has been classified as hazardous [16]. TCLP was adopted in
1986 by the USEPA as replacement for the EP toxicity (Toxicity Characteristics Leaching Procedure (TCLP) test methods), most of the state and central pollution control boards in
India have adopted the TCLP as a regulatory test for identification and characterization of hazardous wastes. TCLP is
used to determine if a waste is to be regulated as a hazardous
waste with the assumption that a dilution/attenuation factor of
100 can be applied to account for environmental factors that
would reduce contaminant concentration. The maximum contaminant level (MCL) for drinking water is multiplied by 100
to set the regulatory level in the TCLP that determines, if a
material is considered to be hazardous [17]. TCLP is also
widely used to evaluate the effectiveness of stabilization. In
this test method the stabilized material is crushed to a particle
size less than 9.5 mm and mixed with a weak acetic acid extraction liquid in a liquid to solid ratio 20:1 and agitated in
rotary extractor for a period of 18 hours at 302 rpm and
22C. The sample is filtered through 0.6 0.8 m glass fiber
filter and the filtrate is defined as the TCLP extract. This
TCLP extract is analyzed for wide variety of hazardous waste
constituents including volatile and semi volatile organics,
heavy metals etc [18,19]. In the present study two toxic heavy
metals namely lead and zinc were monitored in the extract.
Metal analysis was carried out using Atomic Absorption
spectrophotometer AAS 4139, ECIL at Hindustan Zinc Ltd.
laboratory at Chanderia, following methods described in
TABLE 5
Details of Mix Proportions
S. No.
Mix Grade
Proportions of Mix (Cement: Fine Aggregate: Coarse Aggregate)
Water/Cement ratio
1.
M25
1: 1.93: 2.68
0.5
2.
M30
1: 1.65: 2.41
0.45
131
APHA [20]. Investigations have been carried out to study the

behavior of cement concrete mixes with partial replacement
of fine aggregate by ISF slag in a range of 0% to 100% by
weight by ISF slag in cement concrete mixes.

Concrete mix designs were worked out without and with
predetermined slag content. Concrete specimens were tested
for compressive strength and leaching characteristics and the
results have been discussed. Test results have been compared
with the similar work done by other researchers in the past.
The results of the compressive strengths of M25 mixes with
OPC and PPC as binders are given in Table 6.
The results of compressive strengths of M25 mix with water cement ratio of 0.50 indicate that almost all the mixes with
PPC as binder satisfy the target strength of 31.6 N/mm2. It
was indicated that OPC (43 grade) was more efficient in imparting designed strength because OPC possesses higher
compressive strength (44.5 N/mm2) compared to PPC (34
N/mm2). However, the reduction of strength was significant

on increasing the ISF slag percentage in OPC based mixes.
The results of the compressive strengths of M30 mixes with
OPC and PPC as binders with water cement ratio 0.45 are
given in Table 7.
The results of compressive strengths of M30 mix with water cement ratio 0.45 indicate that for mixes with OPC as
binder all the samples up to 60% replacement of sand by slag
satisfied the target strength of 36.6 N/mm2. The target
strength of concrete mixes prepared with 80% and 100% slag
content could not be achieved. This might be due to the fact
that voids between the slag particles could not be filled by
finer sand particles. Moreover higher content of zinc and sulphates were observed in XRF report (Table 8) of cement concrete specimens, which play an important role in slow setting
and hardening of cement.
When OPC was replaced by PPC as binder almost all the
concrete mixes prepared could not exhibit target strength.
This may be due the fact that PPC has low strength (34 Mpa)
and was not able to bind the aggregate effectively. It may be
further noted that the reduction in strength in PPC based
TABLE 6
Compressive Strength of Concrete M25 Mix
S. No.
Compressive Strength (N/mm2)
Slag
(%)
Cement Used OPC
Cement Used PPC
7 days
28 days
7 days
28 days
1.
31.89
41.76
15.21
33.61
2.
20
26.02
39.39
14.54
28.97
3.
40
23.42
32.54
14.54
31.88
4.
60
24.38
32.88
17.51
34.50
5.
80
20.45
32.29
22.65
33.95
6.
100
24.08
31.62
20.57
31.06
TABLE 7
Compressive Strength of Concrete M30 Mix
S. No.
132
Compressive Strength (N/mm2)
Slag
(%)
Cement Used OPC
Cement Used PPC
7 days
28 days
7 days
28 days
1.
24.90
36.96
24.87
34.6
2.
20
31.70
41.92
19.51
32.94
3.
40
29.36
38.22
25.12
40.88
4.
60
28.18
37.22
22.38
31.55
5.
80
23.60
34.37
19.80
30.14
6.
100
22.50
33.92
19.55
31.33
VOLUME 42, NO. 2
MAY 2016
TABLE 8
XRF Spectroscopy Results of Cement Concrete Mixes
S. No.
Content
Concentration
M25 Control Mix
M25 with 60% Slag
SIO2
69.13%
54.88%
Al2O3
5.63%
6%
Fe2O3
2.34%
9.35%
MgO
0.37%
0.39%
CaO
12.7%
18.5%
Na2O
0.58%
0.35%
K2O
1.52%
1%
TiO2
0.2%
0.28%
P2O5
0.3%
0.07%
10
MnO
0.07%
0.53%
11
SO3
6948ppm
1784ppm
12
63ppm
86ppm
13
Cr
288ppm
234ppm
14
Ni
44ppm
3ppm
15
Co
8ppm
35ppm
16
Cu
84ppm
40ppm
17
Pb
2ppm
2294ppm
18
Zn
67ppm
25932ppm
samples was not significant compared to its control (no replacement), whereas with OPC samples the reduction was
large as the percentage replacement increased. Analysis of
leaching of metals for all the samples of M25 and M30 mixes
was conducted using TCLP test. The results are tabulated in
Tables 9 and 10.
TABLE 9
Leaching of Metals of Concrete M25 Mix
S. No.
Slag
Leaching of Metals after 28 Days in ppm
(%)
Cement Used OPC
Cement Used PPC
Lead
Zinc
Lead
Zinc
1.
BDL
BDL
BDL
BDL
2.
20
0.53
1.38
0.66
1.51
3.
40
0.72
1.58
0.69
2.52
4.
60
1.05
2.71
1.25
3.72
5.
80
1.82
3.72
3.08
6.90
6.
100
2.93
5.34
3.92
7.46
BDL: Below Detection Limit
133
TABLE 10
Leaching of Metals of Concrete M30 Mix
S. No.
Slag
(%)
Cement used OPC
Cement used PPC
Lead
Zinc
Lead
Zinc
1.
BDL
BDL
BDL
BDL
2.
20
0.48
1.25
0.58
1.18
3.
40
0.59
1.51
0.89
2.22
4.
60
0.85
1.98
1.59
3.34
5.
80
1.65
2.55
2.08
5.40
6.
100
2.07
3.58
2.50
7.09
BDL: Below Detection Limit
All specimens were tested for leaching after 28 days of

curing. Leaching of Lead and Zinc was found to be within the
limits prescribed (Either specify these limits in the table or in
a sentence before the current statement) by USEPA and BIS
for all test specimens (Refer Tables 9 and 10). It can be seen
that leaching of metals slightly higher in the mixes prepared
with PPC binders. It was observed that leaching of Zn and Pb
increased at a rapid rate irrespective of type of cement in
M25 concrete specimens with slag content of 60%, 80 % and
100% though it was still within the prescribed limits up to
100% replacement of sand by slag.
Leaching for M30 concrete specimens was observed at
relatively higher rate in specimens prepared with 80% slag
and more. This proves that leaching is restricted in relatively
rich mixes because of low water cement ratio that ensures
that adequate quantity of cement is available for coating the
aggregates. It was observed that leaching of Zn and Pb further increased appreciably in mixes with higher slag contents.
Specimens were also tested after 180 days of casting and it
was observed that leaching of Zn was drastically reduced
(Refer Table 11) as compared to that of Pb.

The study reveals that the ISF slag can be very well utilized
as fine aggregate in concrete mixes. It was also observed that
leaching of metals was drastically reduced with the age of
concrete mixes. It can be concluded that ordinary concrete
mixes M25 with OPC as binder for general construction work
with 60% fine aggregate replaced by ISF slag, have no adverse effect on the strength and such mixes are also acceptable from environmental considerations.
The compressive strength at 28 days lies between characteristics strength and target strength when lower strength PPC
(34 N/mm2) is used as binder in such mixes. Concrete mixes
with water cement ratio 0.6 and fine aggregate exceeding
35% of aggregate volume have shown leaching of lead exceeding the limit of 5 ppm as recommended by USEPA.
In field study on ISF slag a series of concrete mixes with
target strength 45-50 MPa for use in concrete pavement was
developed in BREs concrete lab [21] with replacement of
sand by ISF slag to the tune of 75% by volume. The aggregate used were crushed cheddar lime stone coarse aggregate
TABLE 11
Leaching of Metals of Concrete M30 Mix with OPC as Binder
S. No.
Slag
(%)
Lead
Zinc
1.
BDL
BDL
2.
20
0.37
0.16
3.
40
0.41
0.19
4.
60
0.53
0.53
5.
80
0.85
0.63
6.
100
1.24
0.86
Detection Limit
134
VOLUME 42, NO. 2
MAY 2016
and cheddar lime stone sand, together with marine dredged

fine Holme sand in the United Kingdom (UK). The compressive strength at 28 days of mix prepared with 50% slag
showed 39 MPa as compared to 44.5 MPa for control mix.
Compressive strength at 91 days was reported as 44.5 MPa as
compared to 48 MPa of control mix. Durability tests indicated that there was very little difference in the behaviour of
concrete containing ISF slag and the control mixes. The
leaching test (LS2-10) at 7 day age of concrete core (from
demonstration concrete roadway) and site compacted cubes
showed that levels of all metals were below the limits for
inert landfill.
The present study shows that complete ISF slag can be
used as fine aggregate in M25 and M30 concrete mixes with
ordinary Portland cement (OPC) as binder. In the present
study specimens were tested more critically following TCLP.
The specimens were crushed down to the size of 9.5 mm
sieve and tested in aqueous buffered solution in toxicity characteristics leaching procedure (TCLP) to check leaching. Test
results also showed that leaching of heavy metals in these
mixes was within safe limits specified by USEPA.
construction activities and hence help save the reserves of

natural resources.
REFERENCES
[1]
[2]
[3]
[4]
[5]
CONCLUSIONS
ISF slag in cement concrete mixes as a substitute to fine
aggregate was studied using Solidification/ Stabilization process for assessing its potential for large scale application in
construction industry. Both engineering properties and environmental acceptability criteria were studied in details.
Concrete mixes namely M25 and M30 have been studied
with reference to compressive strength and expected leaching
of heavy metals from slag aggregate and the major conclusions are listed below.
a. Concrete mixes M25 prepared with OPC and PPC as
binder also exhibited target strength of 31.6 MPa for
almost all specimens with 20 to 100% replacement of
fine aggregate by slag.
b. Mix M30 achieved target strength up to 60% slag as fine aggregate and characteristic strength was exhibited
by mixes with more than 60% slag to 100% slag content. Replacement of OPC by PPC could not achieve
target strength but characteristic strength was attained
for all slag contents.
c. Concrete mixes M25 and M30 were found to be totally
safe with respect to leaching of heavy metals.
Leaching of Lead and Zinc was found to be within the limits prescribed by USEPA and BIS for all test specimens. It
can be seen that leaching of metals slightly increased in the
mixes prepared with PPC binders. It was observed that leaching of Zn and Pb increased at a rapid rate irrespective of type
of cement in M25 concrete specimens with slag content of
60%, 80 % and 100% though it was still within the prescribed
limits up to 100% replacement of sand by slag. Thus on the
basis of the study it is recommended that this technique can
be safely used for M25 mixes with slag replacement up to
100% and M30 mixes using OPC with slag replacement up to
60%, which satisfy both Engineering and Environmental acceptability. This will encourage the use of waste materials in
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
Zaibo Li, Zhao Sanyin, Zhao Xuguang, He Tusheng

(2012), Leaching characteristics of steel slag components and their application in cementitious property prediction, Journal of Hazardous Materials, Volume 199
200, pp. 1-500.
Leila Seleme Mariano (2013), A Feasibility Study of
Recycling of Construction and Demolition Waste for
Structural Purposes issue3, August 2013.
S.F.U. Ahmed (2014), Properties of Concrete Containing Recycled Fine Aggregate and Fly Ash, Journal of
Solid Waste Technology and Management, Volume 40,
No. 1 February 2014.
Dr. Abdelkader T. Ahmed (2014), Environmental Properties of Waste and By-Product Materials Used in Constructions Volume 40, No. 2 May 2014.
N Jay. Meegoda, A.S. Ezeldin, Hsai-Yang Fang, and
Hilary I. Inyang, (2003), Waste Immobilisation Technologies, Practice periodical of Hazardous, Toxic, and
Radioactive waste management, January.
S.B. Patil, S. Monica, Dr. A.K.Vyas, Dr. A.B.Gupta
(2006), Application of Solidification Stabilzation
Techniquefor metal bearing solid and hazardous waste
from IndustriesA Case study National Conference on
Environmental Pollution and its control at Engineering
College Kota.
V.A. Mymrin, H.A. Ponte, P.R. Impinnisi (2005), Potential Application of acid jarositr waste as the main
component of Construction materials, Construction and
Building Materials, Volume 19, pp. 141-146.
John Robinette Christopher and John A. Epps (2010)
Energy emissions and price allocation with construction
Rehabilitation and Material alternatives for flexible
pavements Transport and Research Board.
C. Atzani, L. Massidda, U. Sanna (1996), Use of granulated slag from lead and zincprocessing in concrete technology, Cement and Concrete Research, Volume 26,
pp. 1381-1388.
S.B. Patil, A.K. Vyas, A.B.Gupta (2007) Suitability of
waste slags from Industries as Construction Material: a
case study of Imperial Smelting Furnace (ISF) slag.
Dr. S.B. Patil, Dr. A.K. Vyas, Dr. A.B. Gupta (2014),
Utilization of an Industrial Waste in Cement Concrete
Mixes, Journal of Solid Waste Technology and Management, Volume 40, No. 1 February 2014.
Hazardous waste (management and handling) rules
(1989), Ministry of Environment and Forest (MOEF),
India as amended in May 2003.
A. Agrawal, K.K. Sahu, B.D. Pandey, (2004), Solid
waste management in non-ferrous industries in India,
Resources conservation and recycling, Volume 42.
Hindustan zinc limited Letter.
135
[15]
[16]
[17]
[18]
136
Toxicity Characteristics Leaching Procedure (TCLP) test

methods as followed by United States Environmental
Protection Act (USEPA), vide No: S.W 846.
M Sild., M. Vondruska, M. Koutny, and V. Bednarik
(2004), Stabilization/solidification of noncombustible
industrial waste by asphalt emulsions. Pract. Period.
Hazard.Toxic Radioact. Waste Management.
B. Batchelor, 2006, Overview of waste stabilization
with cement. Waste Management, March 2006.
Ashok Dhariwal, (2006) Industrial waste utilization for
construction material purposes, National conference on
environmental pollution and its control, March, 2006 at
Kota, Rajasthan.
[19]
[20]
[21]
R. Barna, P. Moszkowicz, C. Gervais, (2004) Leaching

assessment of road materials containing primary lead and
zinc slags, Waste Management, Volume 24 (2004), pp.
945-955.
APHA Standard Method for the Examination of Water
and Waste Water, seventeenth Edition, American Public
Health Association, Washington , DC, USA.
A.M. Dunster, F. Marlinier, R.M. Harrex (2005)
Ferrosilcate slag from zinc production as aggregate
bound in bituminous mixes The Waste and Resource action Programme (WRAP) www.aggregain.org.uk
VOLUME 42, NO. 2
MAY 2016
USE OF RECYCLED AGGREGATE CONCRETE

IN PAVEMENT CONSTRUCTION
Roz-Ud-Din Nassar1*
Parviz Soroushian2
1
Faculty member of Department of Civil Engineering

American University of Sharjah, University City, Sharjah, UAE
2
Professor of Civil and Environmental Engineering at Michigan State University

nassarro@gmail.com, rnassar@aus.edu
Phone: +971 505713765
ABSTRACT
Field performance of recycled aggregate concrete (RAC) in pavement construction subjected to
heavy traffic loads under aggressive weathering conditions was investigated. Test results of laboratory-cured specimens and cores drilled from pavement sections after 270 days of concrete
age showed that the RAC mixes can show performance at par or even better than that of corresponding control (normal) concrete mixes. Later-age enhanced strength and durability attributes of RAC concrete suggest its suitability for use in concrete-based infrastructure such as
pavement construction. The effect of the quality of parent concrete from which recycled aggregate has been obtained was observed to be significant on the performance of the resulting RAC
mix.
Use of RAC in pavement-concrete is estimated to result in major energy and environmental
gains besides being a step forward in the direction of sustainable concrete construction. Large
scale use of recycled aggregate in concrete pavement construction can significantly economize
its construction.
INTRODUCTION
It has been estimated that more than 325 million tons of
construction and demolition (C&D) waste is generated each
years in the U.S. which is about 35 to 40 percent of the nations municipal solid waste stream [1-2]. According to the
U.S. Environmental Protection Agency (EPA), the U.S. output of C&D debris has increased by 25 percent from 1996 to
2003 with a recovery rate of only 48 percent [3]. Land filling
the C&D waste has not only environmental implications; it is
also an economically costly option with high tipping fee. The

adoption of alternative options to conventional demolition
and land filling of C&D waste, such as recycling as aggregate
in new concrete can result in; reduction of solid waste volume
having significant environmental benefits, besides savings of
costs of landfill tipping fee [1, 4].
Factors such as: the rapid depletion of main sources of
quality (virgin) aggregate in some regions; growing environmental concerns about dumping of C&D waste; and increasing scarcity of landfills, increasingly promotes the recycling
___________________________________
137
of concrete demolition waste in new concrete. Presently, out

of the over 2 billion tons of aggregate consumed each year in
the U.S, only 5 percent comes from recycled sources such as
demolished concrete [5-6]. A considerable proportion of recycled aggregate is used as roadfill which does not fulfill the
requirements of sustainable construction as virgin aggregate
continues to be used. It has been estimated that globally about
10 billion tons of sand and rock is consumed by concrete industry and it produces over 1 billion tons of C&D waste each
year [7]. Production and transport of virgin aggregates generate emissions representing 0.0046 million tons of carbon
equivalent for each ton of virgin aggregate, compared to only
0.0024 million tons of carbon equivalent per ton of recycled
aggregates [8]. Since, about 70% of concrete volume comprises of aggregate, the use of waste concrete as recycled
aggregate will be an important step in the direction of sustainable construction practice which can also divert huge
quantities of landfill-bound component of demolished concrete for production of an environmentally friendly concrete.
This situation encourages urgent steps towards increasing the
recycling rate of demolished concrete as aggregate in new
concrete construction.
The broad based use of recycled aggregate in concrete
requires having in-depth knowledge of its physical and mechanical characteristics that are significantly different from
virgin aggregate and affect the strength and durability attributes of the resulting RAC [9-13]. Among these; most important is the higher water absorption of recycled aggregate
(two to three times that of normal aggregate) and the increased shrinkage of the resulting RAC. These drawbacks are
largely resulted from the old mortar / cement paste clinging to
the surface of recycled aggregates. According to Hansen and
Narud [14], the volume percentage of the old mortar attached
to the surface of aggregate varies between 25% and 35%
when concrete with natural gravel is reduced to 16-32 mm
particle size, about 40% in the case of recycled aggregate
with 8-16 mm particle size, and near 60% in recycled aggregate with 4-8 mm particle size. Hasaba et al. [15] reported
that 35.5% of old mortar is attached to natural gravel in recycled aggregate with 5-25 mm particle size produced from
concrete with 24 MPa compressive strength. For the same
size of recycled aggregate, the attached mortar fraction increased to 36.7% and 38.4% when the recycled aggregate
was produced by crushing concretes with compressive
strengths of 41 MPa and 51 MPa, respectively. According to
Japanese studies [16], approximately 20% of cement paste is
attached to the recycled aggregate with 20-30 mm particle
size. Nixon [17] has concluded that the most significant difference between recycled aggregate and virgin aggregate is
the markedly higher water absorption of the recycled aggregate. Tavakoli and Soroushian [12-13] found that the water
absorption capacity of recycled aggregate reflects on the
amount of cement paste adhering to the surface of the aggregate particles. With increased quantity of the clinging paste,
the drying shrinkage of the resulting concrete increased accordingly. They further reported that recycled aggregate concrete shows higher drying shrinkage than normal concrete,
and the magnitude of the increase in drying shrinkage depends on the properties of the parent concrete from which the
138
recycled aggregate was obtained.

Various past researchers have suggested a variety of
treatments to overcome the inferiorities of recycled aggregate
and improve its performance for use in structural and pavement grade concretes [18-26]. Kou et al. [27], Salem et al.
[4], Ann et al. [28], Berndt [29] and Belagraa and Beddarr
[30] have investigated the effect of various mineral admixtures as partial replacement for cement on the properties of
recycled aggregate concrete. Nassar and Soroushian [26]
have reported significant improvement in strength and durability of RAC as a result of pozzolanic reaction of milled
waste glass with hydrates of cement when 20% of cement
was replaced with milled waste glass in RAC.
This research looks at the field performance of RAC in
pavement construction under heavy traffic loads and aggressive environmental conditions to assess its suitability for use
in concrete pavement construction. The RAC mixes used in
this study were proportioned through the use of class C fly
ash as partial replacement for cement using low water-cement
ratio. Field investigation of RAC was carried out through
construction of two experimental pavement sections. One
experimental section was constructed each on east and westbound Jackson road in Ann Arbor, Michigan in May through
June and their long-term performance under heavy traffic
loads and weathering conditions was compared with that of
respective control (produced with virgin aggregate) pavement-concrete. Both; control and RAC pavement concretes
were produced using class C fly ash as 25% of Portland cement replacement in concrete. Strength and durability tests
were carried out on the field concrete. Long-term tests on
cores drilled from pavement concretes were carried out at 270
days of concrete age to investigate the strength and durability
characteristics of in-situ concrete mixes.

Materials
Table 1 shows physical properties of coarse, fine and recycled aggregates used in the experimental program. As can
be seen recycled aggregate used in westbound pavementconcrete is of considerably better quality than that used in the
eastbound pavement-concrete. This difference in quality also
points at the superior quality of parent concrete of the recycled aggregate used in westbound RAC. Ordinary Portland
cement was used in both, control and RAC mixes. Table 2
presents the physical properties of class C fly ash and Portland cement, while the chemical composition of the two is
given in Table 3. Mixture designs of RAC and control concretes are shown in Table 4. RAC mixtures were produced
with 100% replacement of virgin aggregate with recycled
coarse aggregate and natural sand. Air entraining agent (with
brand name of Conair 260TM) and water reducing agent
(known by brand name Optiflo 500TM) were used in both
concrete mixtures. All batches of concrete were manufactured
in ready-mix concrete plant installed near the project site.
VOLUME 42, NO. 2
MAY 2016
TABLE 1
Physical properties of coarse and fine aggregates
Aggregate
Dry density
3
Bulk specific
Bulk specific
Absorption
Loss on
Type
kg/m
gravity
gravity (SSD)
(%)
abrasion (%)
C. aggregate
1721
2.52
2.60
2.22
22.4
R. aggregate (E.B.)
1416
2.30
2.43
4.14
33.5
R. aggregate (W.B.)
1438
2.41
2.52
3.12
26.6
2.88 (F.M.)
2.64
1.20
F. aggregate
C = coarse;
R = recycled;
F = fine;
E.B. = Eastbound
W.B. = Westbound;
F.M. = fineness modulus
TABLE 2
Physical properties of Class C fly ash and Portland cement
Parameter
% Passing # 325 mesh
Specific gravity
2
Specific surface area (Blain) (m /kg)
Fly ash
Cement
84.6
87.25
2.61
3.15
319
356
TABLE 3
Chemical composition of Class C fly ash and Portland cement
Chemical composition (%)
Fly ash
Cement
SiO 2
33.52
20.2
Al 2 O 3
17.35
4.7
CaO
29.11
61.9
Fe 2 O 3
5.14
3.0
SO 3
2.38
3.9
MgO
4.56
2.6
Na 2 O
0.68
0.19
K2O
0.97
0.82
Loss on ignition
0.67
0.79
Test Specimens
Representative concrete cylinders having 152 mm (6 in.)
diameter and 304 mm (12 in.) height were tested in compression at 7, 28 and 90 days of concrete age after moist curing in
lime-saturated water. Similarly, concrete beams were prepared from all concrete mixtures, and were tested for flexural
strength after 7, 28 and 90 days of moist curing. After 270
days of concrete age, cores were drilled from the constructed
pavement sections and tested for evaluation of the compres-
139
TABLE 4
Concrete mix designs
Control
RAC (E.B.)
RAC (W.B.)
1005
827
840
Sand (kg/m3)
739
739
739
w/cm ratio
0.42
0.42
0.42
Cement content (kg/m3)
234
234
234
Water content (kg/m3)
132
132
132
Fly ash (kg/m3)
78*
78*
78*
Air entraining admixture (ml/kg)
2.80
2.80
2.80
Water reducing admixture (ml/kg)
1.75
1.75
1.75
Coarse aggregate
3
(kg/m )
* = 25% replacement of cement with class-C fly ash
sive strength, water sorption, chloride permeability and abrasion resistance of field concretes. All tests were carried out
according to the provisions of relevant ASTM standards that
are presented in Table 5.
Experimental Pavement construction

Experimental, Jointed Plain Concrete Pavement (JPCP)
sections consisted of four-lane Concrete Boulevard and a bike
lane, one each on east and westbound Jackson Road. Each
pavement section was 250 feet long, 27 feet wide with 9
inches deep cross section. Control concrete-pavement sections consisted of same cross section. By comparison, the
placement, finishing and appearance of RAC pavementconcrete was no different from control pavement-concrete.
EXPERIMENTAL RESULTS AND

DISCUSSION
Test results of experiments on control and RAC materials
in fresh and hardened states have been summarized in Table
6.
Fresh concrete properties

As shown in Table 6 the east and westbound RAC mixtures exhibited higher air content than the corresponding control mixes at equal dosage of air entraining agent. This difference could be attributed to the entrapped air in the porous
coarse aggregate particles in RAC mixtures. The rougher
TABLE 5
ASTM standards followed in the experimental work
Test description
Specification
Slump
ASTM C 143
Density (fresh concrete)
ASTM C 138
Air content
ASTM C 231
ASTM C 39
Flexural strength
ASTM C 78
Sorption
ASTM C 1585
Chloride permeability
ASTM C 1202
140
VOLUME 42, NO. 2
MAY 2016
TABLE 6
Fresh and hardened concrete properties of control and RAC concrete mixes
Property
Eastbound Pavement-Concrete
Westbound Pavement-Concrete
Control
RAC
Control
RAC
52
63
55
58
Density (kg/m3)
2285
2208
2287
2227
Air Content (%)
7.5
8.0
7.0
7.5
Slump (mm)
Concrete Age (days)
28
90
28
90
28
90
28
90
Compressive
Strength (MPa)
28.35
34.39
39.56
18.21
26.56
32.45
26.34
32
35.49
26.78
35.09
37.45
Flexural Strength
(MPa)
5.54
7.12
8.16
2.83
4.35
6.15
4.38
5.17
6.62
4.18
5.39
7.47
Core Compressive
Strength at 270 days
(MPa)
32.65
29.75
32.69
33.75
Specimen Chloride
Permeability (Coulombs)
2995
3285
2980
3120
Core Chloride Permeability (Coulombs)
3335
3770
3320
3450
Abrasion Weight
Loss (gm)
0.35
0.38
0.34
0.34
texture of recycled aggregate is expected to result in lower

workability of RAC mixtures than control concrete mixtures.
However, the RAC mixtures used in east and westbound
pavements showed higher slump values than corresponding
control concrete mixtures. The higher air content seems to
have overcome the effects of rough surface texture of recycled aggregate in RAC mixtures, as entrained air is known to
be a workability aid. Fresh mix density of RAC mixtures was
lower than that of control mixtures due to the lower specific
gravity of recycled aggregates than virgin aggregate.
In the case of eastbound pavement concrete mixtures, the
control concrete showed higher strength than RAC mixture at
all ages. This trend was, however, reversed (refer Table 6) in
the case of westbound pavement concrete mixtures wherein
RAC showed higher strength than corresponding control concrete at all ages. The different trends could be explained
based on the fact that the strength of RAC depends upon the
quality of recycled aggregate and hence the strength and
quality of parent concrete from which the recycled aggregate
has been obtained. In this case the better quality of recycled
aggregate (pointing at superior quality of its parent concrete)
used in the westbound pavement-concrete resulted in production of RAC with improved compressive strength.
The fact that RAC concrete mix is able to gain higher
strength than the control, suggest that through the use of qual-
ity recycled aggregate, RAC mixtures could be as good as

concrete produced with virgin aggregate or could even show
better performance than the later.
Flexural strength
The flexural strength of the RAC mix was less than that of
the control mix at all ages for the eastbound pavementconcrete. As reflected in Table 6, test results showed opposite
trend for the westbound concrete mixes wherein RAC
showed significantly higher flexural strength than control mix
at the age of 90 days. At earlier ages the flexural strengths of
the two concrete materials was comparable. Significant difference in flexural strengths of the RAC mixes used in east
and westbound pavement could be explained based on the
difference in quality of the two types of recycled aggregates
as discussed in the case of compressive strength.
As in the case of compressive strength, RAC mix demonstrated the ability to gain equivalent or even better flexural
strength than that of control concrete mix, once again proving
the importance of the quality of recycled aggregate to improve the performance of RAC.
Compressive strength of concrete cores

Compressive strength of the insitu pavement-concrete was
measured by drilling cores from the control and RAC pavement sections. Cores having diameter of 102 mm (4 in.) with
141
heights varying from 178 to 203 mm (7 to 8 in.) were tested

in compression at 270 days of concrete age. The strengths of
cores exposed to field environment are found to be less than
those obtained using continuously moist-cured cylindrical
specimens produced using the same concrete. The higher
strength of molded specimens than that of specimens cored
from field concrete is due to better preparation and improved
curing of the former.
The strength for cores drilled from east and westbound
RAC pavements follows trends similar to those for molded
specimens. As shown in Table 6, at the age of 270 days, the
core compressive strength of eastbound RAC pavementconcrete is less than that of the control concrete. However at
same age, the compressive strength of westbound RAC mix is
higher than that of control mix. Test results for cored specimens also provide evidence for the better compressive
strength of westbound pavement-concrete brought about the
superior quality of the respective recycled aggregate which
originates from its parent concrete.
Chloride permeability
Improvement in resistance of concrete to chloride ion
permeation indicates towards enhancement in barrier qualities
of concrete against the ingress of aggressive solutions, which
critically influence its long-term durability. According to
ASTM C 1202, if the number of coulombs passed lies between 2000 and 4000, the chloride permeability of concrete is
considered low, and it is considered very low for the 100 to
1000 range. All concrete materials in this study provided low
chloride permeability levels (Table 6). The performance of
the RAC materials is quite comparable with that control concrete materials. The lower chloride permeation of westbound
RAC pavement-concrete can be associated with the better
quality of recycled aggregate used in this concrete. Test results of concrete cores drilled from the two pavements follow
the trend of the laboratory-cured specimens. In this case too,
the resistance to chloride permeability of the westbound RAC
is quite comparable with that of control concrete. These test
results suggest the improvement in the durability characteristics of RAC mix brought about by the improvement in quality
of recycled aggregate used in its production.
were sealed with epoxy to produce one-dimensional sorption

during the test. Figure 1 shows the, i (sorption per unit area
per unit density of water) vs. time1/2 plots of RAC and control
concrete mixtures used in the eastbound pavement. The corresponding sorption plots for the concrete mixtures used in
westbound pavement sections are shown in Figure 2. The
westbound RAC pavement- concrete showed comparable
performance when compared with that of corresponding control pavement-concrete. The comparable rate of sorption of
westbound RAC pavement- concrete to that of corresponding
control concrete may be attributed to the better quality of
recycled aggregate used in this concrete. The superior quality
of recycled aggregate used in westbound RAC mix proved to
be effective towards enhancing its durability and bringing it
at par with that of control concrete.
Abrasion resistance
Abrasion resistance of concrete is an important property
influencing the performance of concrete pavements and
floors subjected to abrasive action of traffic [32-35]. Field
FIGURE 1
Water sorption
Since concrete is mostly in an unsaturated state throughout its service life; sorption is therefore, the most important
mode of moisture transport in it. Water absorption is an important indicator of the durability characteristic of hardened
concrete as a direct relationship exists between water absorption and porosity of concrete. Improvement in water barrier
qualities of concrete can greatly enhance the long-term performance and service life of concrete in aggressive service
environments. In this study sorption tests were carried out
using disc specimens having 102 mm diameter and 51 mm
thickness that were cut out from concrete cores at the age of
270 days. Specimens were dried to constant mass after 25
days (average) of oven drying at 50o C following the procedure of ASTM C 1585 -04 [31]. Sides and top of specimens
142
FIGURE 2
VOLUME 42, NO. 2
MAY 2016
observation of the pavement-concrete after exposure to a

complete cycle of aggressive weather under traffic load did
not show any signs of scaling or other surface damage. Abrasion resistance test results of cores obtained from east and
westbound control and RAC pavement-concretes at the age of
270 days are also presented in Table 6. The abrasive weight
losses of east and westbound RAC pavement-concretes are
seen to be comparable to the corresponding control mixes.
The improved strength characteristics of the westbound RAC
pavement-concrete results in its improved abrasion resistance
which is seen to be at par with that of control concrete (noting
that a direct relationship exists between compressive strength
and abrasion resistance of concrete). Test results of the abrasion resistance; show the suitability of the use of RAC in
pavement construction.
ACKNOWLEDGEMENT
Part of this research project was funded jointly by FHWA
and Washtenaw County Road Commission, MI. Authors are
grateful to their financial support. Vital support of Roy
Townsend and Sheryl Soderholm Siddall of Washtenaw Road
Commission and Hugh Luedtke of Ajax Paving is gratefully
acknowledged as well.
REFERENCES
[1]
[2]
CONCLUSIONS
[3]
[4]
The use of good quality recycled aggregate can result in

production of RAC mixes with improved strength and
durability characteristics.
Quality of parent concrete from which the recycled aggregate has been obtained significantly affects the quality
of the resulting RAC mix in terms of strength and durability attributes.
Visual inspection of pavement sections constructed with
RAC mixes did not show any signs of deterioration such
as surface scaling after exposure to a complete cycle of
aggressive weather under heavy traffic load.
RAC mixes can be used in pavement construction provided that recycled aggregate used in it is of superior
quality yielding RAC mix with improved strength, sorption, chloride permeability, and abrasion resistance characteristics.
Use of recycled aggregate in concrete is an important
step forward in the direction of sustainable construction
practice which can result in important energy and environmental gains besides helping in construction of economical transportation infrastructure.
[5]
[6]
[7]
[8]
[9]
[10]
[11]
BIOGRAPHY
Dr. Roz-Ud-Din Nassar is a faculty member of Civil Engineering in the American University of Sharjah, UAE. He
received his BS from Balochistan Engineering University
Khuzdar, Pakistan; MS from National University of Sciences
and Technology (NUST), Pakistan; and PhD from Michigan
State University. His research emphasizes concrete durability
and environmentally-friendly and energy-efficient construction materials.
Dr. Parviz Soroushian, FACI, FASCE is a professor of
Civil and Environmental Engineering at Michigan State University. He received his BS from Tehran University; and PhD
from Cornell University. His research interests include materials science and engineering, and environmentally friendly
and energy-efficient construction materials and systems.
[12]
[13]
[14]
[15]
DOD, U.S., Selection of Methods for the Reduction,

Reuse, and Recycling of Demolition Waste in Unified
Facilities Criteria. 2002.
CMRA. Economical Good Sense. 2011 [cited 2012;
Available from: http://www.cdrecycling.org/.
EPA, EPA Estimates 170 million tons of yearly
Construction, Demolition Debris. 2009.
Salem, R.M., E.G. Burdette, and N.M. Jackson,
Resistance to freezing and thawing of recycled
aggregate concrete. ACI Materials Journal, 2003.
100(Compendex): pp. 216-221.
FS18199, Recycled AggregatesProfitable Resource
Conservation, USGS, Editor. 2000.
Mehta, P.K., Greening of concrete industry for
sustainable development. Concrete International,
2002: pp. 23-28.
Mehta, P.K., Global Concrete Industry Sustainability.
Concrete International, 2009. Volume 31(2): pp. 45-48.
EPA530-R-03-017, Background document for life-cycle
greenhouse emission factors for clay brick reuse and
concrete recycling, EPA, Editor. 2003.
Etxeberria, M., et al., Influence of amount of recycled
coarse aggregates and production process on properties
of recycled aggregate concrete. Cement and Concrete
Research, 2007. 37(Compendex): pp. 735-742.
Cuttell, G.D., et al., Performance of rigid pavements
containing
recycled
concrete
aggregates.
Transportation Research Record, 1996(Compendex):
pp. 89-98.
de Juan, M.S. and P.A. Gutierrez, Study on the
influence of attached mortar content on the properties of
recycled concrete aggregate. Construction and
Building Materials, 2009. 23(Compendex): pp. 872877.
Tavakoli, M. and P. Soroushian, Drying shrinkage
behavior of recycled aggregate concrete. Concrete
International, 1996. 18(Compendex): pp. 58-61.
Tavakoli, M. and P. Soroushian, Strengths of recycled
aggregate concrete made using field-demolished
concrete as aggregate. ACI Materials Journal, 1996.
93(Compendex): pp. 182-190.
Hansen, T.C. and H. Narud, Strength of recycled
concrete made from crushed concrete coarse
aggregate.
Concrete
International,
1983.
Hasaba, S., M. Kawamura, and K. Toriik, et al. Drying
143
[16]
[17]
[18]
[19]
[20]
[21]
[22]
[23]
[24]
[25]
144
shrinkage and durability of concrete made of recycled

concrete aggregates. Japan Concrete Institute, 1981.
Volume 3: pp. 55-60.
B.C.S.J, Study on recycled aggregate and recycled
aggregate concrete. Concrete Journal, Japan, 1978.
Volume 16(7): pp. 18-31.
Nixon, P., Recycled concrete as an aggregate for
concretea review. Materials and Structures, 1978.
Volume 11(5): pp. 371-378.
Shayan, A. and A. Xu, Performance and Properties of
Structural Concrete made with Recycled Concrete
Aggregate.
ACI
Materials
Journal,
2003.
100(Compendex): pp. 371-380.
Tam, V.W.Y., X.F. Gao, and C.M. Tam,
Microstructural analysis of recycled aggregate
concrete produced from two-stage mixing approach.
Cement and Concrete Research, 2005. 35(Compendex):
pp. 1195-1203.
Yang, K.-H., H.-S. Chung, and A.F. Ashour, Influence
of type and replacement level of recycled aggregates on
concrete properties. ACI Materials Journal, 2008.
105(Compendex): pp. 289-296.
Kou, S.C., C.S. Poon, and C. Dixon, Influence of fly
ash as cement replacement on the properties of recycled
aggregate concrete. Journal of Materials in Civil
Engineering, 2007. 19(Compendex): pp. 709-717.
Henry, M., et al., Balancing durability and
environmental impact in concrete combining low-grade
recycled aggregates and mineral admixtures.
Resources, Conservation and Recycling, 2011. Volume
55(11): pp. 1060-1069.
Abbas, A., et al., Durability of recycled aggregate
concrete designed with equivalent mortar volume
method. Cement and Concrete Composites, 2009.
Fathifazl, G., et al., New mixture proportioning
method for concrete made with coarse recycled concrete
aggregate. Journal of Materials in Civil Engineering,
2009. 21(Compendex): pp. 601-611.
Gokce, A., et al., Freezing and thawing resistance of
air-entrained concrete incorporating recycled coarse
aggregate: The role of air content in demolished
concrete. Cement and Concrete Research, 2004.
[26] Nassar, R.-U.-D. and P. Soroushian, Strength and

durability of recycled aggregate concrete containing
milled glass as partial replacement for cement.
Construction and Building Materials, 2012. Volume
29(0): pp. 368-377.
[27] Kou, S.-c., C.-s. Poon, and F. Agrela, Comparisons of
natural and recycled aggregate concretes prepared with
the addition of different mineral admixtures. Cement
and Concrete Composites, 2011. Volume 33(8): pp.
788-795.
[28] Ann, K.Y., et al., Durability of recycled aggregate
concrete using pozzolanic materials. Waste
Management, 2008. 28(Compendex): pp. 993-999.
[29] Berndt, M.L., Properties of sustainable concrete
containing fly ash, slag and recycled concrete
aggregate. Construction and Building Materials, 2009.
23(Compendex): pp. 2606-2613.
[30] Belagraa, L. and M. Beddar. Admixtures' effect on
mechanical strength of a concrete made of recycled
aggregate. in 2005 International Congress - Global
Construction: Ultimate Concrete Opportunities, July 5,
2005 - July 7, 2005. 2005. Dundee, Scotland, United
kingdom: Thomas Telford Services Ltd.
[31] ASTM C1585 - 04 Standard Test Method for
Measurement of Rate of Absorption of Water by
Hydraulic-Cement Concretes. 2006, American Society
for Testing and Materials.
[32] Fwa, T.F. and P. Paramasivam. Surface-deterioration
resistance of concrete pavement materials. in First
International Symposium on Surface Characteristics,
June 8, 1988 - June 9, 1988. 1990. State College, PA,
USA: Publ by ASTM.
[33] Kumar, B., G.K. Tike, and P.K. Nanda, Evaluation of
properties of high-volume fly-ash concrete for
pavements. Journal of Materials in Civil Engineering,
2007. 19(Compendex): pp. 906-911.
[34] Li, H., M.-h. Zhang, and J.-p. Ou, Abrasion resistance
of concrete containing nano-particles for pavement.
Wear, 2006. 260(Compendex): pp. 1262-1266.
[35] Nanni, A., Abrasion resistance of roller compacted
concrete.
ACI
Materials
Journal,
1989.
VOLUME 42, NO. 2
MAY 2016
A MULTI-CRITERIA EVALUATION OF THE METHODS FOR

RECYCLING SCRAP TIRES
Shoou-Yuh Chang1, Frank Gronwald2
1
DOE Samuel Massie Chair Professor, Dept. of Civil Engineering

North Carolina A&T State Univ., Greensboro, NC 27411
2
Research Assistant, Dept. of Civil Engineering

chang@ncat.edu
fsgronwa@aggies.ncat.edu
ABSTRACT
Worn out tires create numerous problems in the field of solid waste management. The tires are
very bulky and take up a lot of premium space in the landfills. Insects and rodents like to use
them as homes where they can spread diseases and if the tires catch fire they will emit toxic
gasses. There are several methods of disposing of these tires, but no universal method which
is considered the best for all situations. This study looks at each method of tire disposal on the
basis of cost/benefit analysis, environmental impact, and feasibility. It will be beneficial to all if
the best methods were known, and some methods were eliminated to reduce confusion. Four
different methods of Multi-Criteria Decision Analysis were performed to determine the optimal
solution through a weighted scoring system which accounts for criteria in cost/benefit analysis,
environmental impact, and energy production. The study found that retreading tires for reuse
was the best option. However, since that is not always a practical or long-term solution the
study found that either using the scrap tires to create crumb rubber or for the purpose of civil
engineering applications would be the optimal solutions.
Keywords: Tire recycling, multi-criteria decision analysis, tire derived fuel, pyrolysis
INTRODUCTION
Every year a large number of tires are discarded. The
highest portion of these is generated from automobiles. During a majority of the history of the automobile, the scrap tire
problem did not exist. The problem began in the early 1960s
and has only escalated significantly over the last 25 years. It
was reported that 266 million tires were scrapped in the US in
1996 (Everett and Douglah, 1998). About 20% of these tires
have been either recycled or burned as fuel, with the remaining 80% either stockpiled or transported to landfills. As of
2012 , between 7 and 8 billion tires were stockpiled in the
United States (EPA, 2015). Since such a large proportion of
waste tires are simply buried in landfill sites or just taking up
space, this creates a big issue as to what to do with them.
There have been numerous proposals over the years for using
these waste tires in environmentally beneficial applications to

consume as many waste tires as possible. Even today, this
dilemma still presents a considerable challenge. Most used
tires are still either sent to the landfills or tire graveyards,
where hundreds to thousands of tires create an eyesore and a
public health hazard. Discarded tires become breeding
grounds for mosquitoes and rodents, which carry a lot of diseases, including West Nile virus, encephalitis, dengue fever,
and hanta virus (Jang et al. 1998). Tire piles also attract children, who can injure themselves playing in them. Scrap tires
brought to landfills can also damage the landfill linings that
have been installed to help keep surface and groundwater free
from landfill contaminants.
Another problem with discarded tires is the risk of fire.
Tire pile fires can smolder for weeks to months, releasing
extremely toxic pollutants into the air, creating serious respir-
145
atory and other health problems for people in the vicinity and
many miles away (Amari et al. 2000). Runoff water from a
tire fire can also produce toxins, which could then contaminate the groundwater. Once a large tire pile ignites, it can be
extremely difficult and almost impossible, to extinguish. In
many cases, fires from large tire piles will be left to burn for
long periods of up to several months with the fumes being
visible for many miles. This is due to the intense heat and
smoke generation which give fire fighters an almost impossible task of putting out these fires. One of the main problems
is that the air and soil pollution will become much worse if
attempts are made to extinguish the fire using foam or water
(Amari et al. 2000). Because of these reasons, scrap tire fires
are usually permitted to burn out in a controlled manner until
the entire pile is exhausted. Governmental agencies were
forced to develop environmental policies to promote and assist the tire recycling industry in order to clean up the accumulation of stockpiles of scrap tires, including those in illegal
dump sites.
Tires are composed of vulcanized rubber and various reinforcing materials. The most commonly used rubber matrix is
the co-polymer styrene-butadiene (SBR) or a blend of natural
rubber and SBR (Jang et al. 1998). In addition to the rubber
compound, tires contain reinforcing fillers and fibers. The
reinforcing fillers are made of carbon black, and are used to
strengthen the rubber and aid abrasion resistance. The reinforcing fibers are almost always composed of textile or steel
fibers, usually in the form of a cord which are used to provide
reinforcing strength in tires. Extenders, such as, petroleum oil
are also added to control viscosity, reduce internal friction
during processing, and improve low temperature flexibility in
the vulcanized product (Jang et al. 1998).
In the early 1900s, the average recycled content of all
rubber products was over 50%. Many would say that the rubber reclaiming industry is as old as the industrial use of rubber itself. The proponent fueling the rubber reclaiming industry was the comparatively scarce supply of rubber at that
time. In 1910, one pound of natural rubber cost nearly as
much as one pound of silver and it made economic sense to
reuse as much of the rubber as possible (Reschner, 2008).
However by 1960, the recycling content in the traditional
rubber manufacturing industry dropped to around 20%. This
was due to the effects of cheap oil imports, the availability of
synthetic rubber, and the development of steel belted radial
tires. These factors have led to a steady decline of rubber
recycling. The newer steel belted radial tire designs made it
increasingly difficult to grind and slice those tires. Because of
this, the old methods for regrinding and reusing waste tires
have been made obsolete. As of 1995, the traditional tire and
rubber re-claimers could produce only about 2% recycled
material. New methods for recycling scrap tires have been
developed and are considered to be better options. These
methods will be discussed further in the next section.
Current Methods of Disposal of Used Tire

Resale. This option occurs because an individual will usually
purchase four tires at the same time when only one or two
tires need replacing. The other tires can be resold by someone
146
who specializes in used tires. The level of use of reused tires

has declined drastically. One of the major contributing factors
has been the decline in the use of snow tires. Years ago, a
large portion of car owners maintained a pair of snow tires,
which needed to be mounted on the car for the winter and
removed in the spring. These snow tires were often purchased
as retreads. The development of all-season tires virtually
eliminated the use of conventional snow tires.
Another important impact on the decline of reused tires is
the invention of radial tires. The radial tires have threads
which have more durability and enhanced service requirements such as low rolling resistance. The decline of the rubber reclaiming industry was the major cause of today's scrap
tire problem. If all 278,000 tons of reclaim used in 1960 had
been produced from scrap tires, over 50 million scrap tires
would have been required (Reschner, 2008).
Tire retreading. This is a practice done mostly for extending
the life of truck tires as their cost keeps increasing. Some of
these tires can be retreaded two to three times over the life of
the tire. Retreading a tire will not eliminate it from contributing to the scrap tire problem because the retreaded tire will
eventually wear out, probably in less time than a new tire.
However, retreading delays the appearance of those tires on
the scrap heap and thus reduces the number of scrap tires that
would have appeared if they had not been retreaded by lack
of new tire demand. The choice to resell or retread should be
made either by the end user or during collection as it is not
practical to sort them out at the tire recycling facility.
Landfilling. This involves the collection of waste tires by a
solid waste hauler who is paid to remove discarded tires from
dealer's property. Most landfills accept whole scrap tires only
at a hefty tipping fee because tires are awkward to handle and
difficult to compact. In many cases, scrap tires will work their
way to the top of a closed landfill over time, causing costly
damages to the landfill cover. In landfills, tires tend to float
or rise to the surface, piercing the landfill cover. Shredding or
splitting of tires is becoming increasingly common as part of
the disposal process. Shredding can reduce tire volume by up
to 75%. This can reduce transportation costs by 30 to 60 percent simply because fewer trips are required, and maximum
hauling weights may be achieved more easily. Shredder companies charge anywhere from $19 to $75 per ton to shred
discarded tires in the U.S (Reschner, 2008).
Civil Engineering Applications. Tire derived products are
mostly composed of 1 tire chips which can be used to replace conventional construction material, such as road fill,
gravel, crushed rock, or sand. The benefits of using tire chips
instead of conventional construction materials are a reduced
density, improved drainage properties, and better thermal
insulation (Scrap Tire News, 2011).
One interesting application is using scrap tires for the creation of underwater reefs and breakwaters. A project in Ft.
Lauderdale, Florida, U.S. has used over 3 million tires and
annually adds one million tires per year to reefs. Enthusiasm
for this use has waned, and discarded tire reefs are now only
built in minimal numbers. Currently, an estimated 120 000
VOLUME 42, NO. 2
MAY 2016
150 000 tires are used annually in the construction of reefs.

Costs are about $3.50 per tire. This compares to an average
$45.25 per ton to landfill tires in the northeastern United
States (Reschner, 2008).
Discarded tires for breakwaters which are filled with foam
are able to displace 200 pounds of water and can be used to
float a number of devices such as marinas and docks (Everett
and Douglah, 1998). Tire floats cost approximately 6 to 8
cents per pound, which is cheaper than the current popular
choice, foam filled plastic which sells for between 10 and 14
cents per lb (Reschner, 2008). Breakwaters and flotation devices are able to use approximately 30,000 to 50,000 tires per
year in the U.S.
Another popular application is to use the scrap tires as
playground equipment. Building tire playgrounds and other
small-scale local and backyard recreational uses will consume up to 7500 truck tires per year in the U.S. Other countries such as Japan and Korea will use a minimal amount of
discarded tires as playground equipment; the numbers consumed are unknown. Demand for tire playground products is
declining as the economy improves and schools and parks
select wooden playground equipment.
Erosion control is also a feasible application for scrap
tires. Discarded tires used with other stabilization materials to
reinforce an unstable highway shoulder or protect a channel
slope have been shown to produce stability. Discarded tires
were banded together and partially or completely buried on
unstable slopes. Construction costs were reduced from 50 to
75% compared to the lowest cost alternatives such as rock,
gabion, or concrete protection (Reschner, 2008).
Tire Derived Aggregate (TDA) uses shredded tire rubber
as a replacement for naturally mined materials such as gravel.
It is considered safe, reliable, and cost-effective. One of its
more useful benefits is that when used under transportation
structures such as highways and rail tracks, TDA acts to reduce noise and vibration that can affect nearby buildings and
communities. There have been tests to see whether the tires,
which contain petroleum products and metals, will leach toxics into groundwater (Chang et al. 2009). Although small
amounts of iron and manganese have been detected, they
were found to be at levels which were below health concerns.
Other tests also were performed using tiny shrimp that were
placed in the runoff water and they suffered no noticeable
health consequences (Chang, 2008).
Increasing the civil engineering market depends on three
factors: cost competitiveness of tire shreds when compared to
traditional construction materials, increased acceptance by
regulatory agencies, and increased recognition by scrap tire
processors of the emerging opportunities available for civil
engineering applications.
In 2011, Scrap Tire News found that the ground rubber
market contained two classes of particle sizes: ground rubber (10 mesh and smaller) and coarse rubber (4 mesh and
larger, with a maximum size of one-half inch). Over the past
several years there has been significant growth in the ordering of coarse sized particles. This particle range is used in
mulch, playground surfacing, running track material, soil
amendments and some bound rubber products.
Rubber Modified Asphalt (RMA). Ground rubber can be

mixed with asphalt to favorably modify the properties of the
asphalt for highway construction. Ground scrap tire rubber
can either be used as part of the asphalt rubber binder, seal
coat, cap seal spray or joint and crack sealant; or as an aggregate substitution. Currently, there is an increasing interest in
the benefits of rubber-modified asphalt, in most states currently using it, and in the other states considering its implementation. To a large extent, any large-scale increase in the
use of RMA is dependent upon the willingness of a state department of transportation to accept national test results and
begin its own state and local level programs (Dierks, 2011).
Currently there are five states which are already using RMA:
California, Arizona, Texas, Florida and South Carolina. The
use of RMA in an open graded friction course on highways
has successfully reduced road noise in Arizona and California. RMA uses between 500 and 2,000 scrap tires for each
lane mile of pavement (RMA, 2003). Other benefits include
reduction of maintenance costs, less cracking and rutting, and
anti-aging due to the anti-oxidants obtained from the tire
manufacture process. Pavements made from rubberized HMA
exhibit greater skid resistance than those made from conventional hot mix asphalt (RMA, 2003).
Asphalt rubber cant be placed during weather extremes.
Generally, the concrete pavement surface must be between 85
and 145 degrees F for the overlay to adhere and cure properly
(Scrap Tire News, 2011). In the winter months the asphalt
rubber acted as a tough surface so there was a reduction of
chain wear and the snow plows didnt chip away at the surface (Dierks, 2011). The drawbacks of RMA are its workability, recyclability, and the fumes emitted during the paving
process.
Spliting or punching of tires. Splitting involves the removal
of steel beads and then the desired shaped of the tire is
achieved using a stamp or punch. Products produced from
split tires include floor mats, belts, gasket, shoe soles, dock
bumpers, seals, muffler hangers, shims, washers, insulators,
and fishing and farming equipment (Jang et al. 1998). These
account for a small, basically static, portion of the scrap tire
marketplace. Machines are required to cut tires into strips
which are then fastened together to make the products listed
above. These machines cost about $15,000 - $18,000 each
and are very labor intensive (RMA, 2003). This operation
will only become profitable in an area of abundant and lowcost labor.
Tire Derived Fuel (TDF) Tire derived fuel was the first
market for scrap tires. From 1979 until 1992, TDF was the
essentially the only market for scrap tires. Beginning in 1992,
whole scrap tires began to be used as feedstock for ground
rubber and processed tires were applied to civil engineering
applications. The EPA considers the use of tire-derived fuels
to be a preferable alternative to the use of fossil fuels (US
EPA, 2011). Use of tire derived fuel reduces the amount of
fossil fuels that would otherwise be consumed. TDF is less
expensive than fossil fuels. One of the benefits of using TDF
is that it saves the petroleum reserves while eliminating tons
of scrap tires from landfills and scrap tire piles.
147
The burning of whole discarded tires also reduces the

need for expensive shredding operations. However, the burning of whole discarded tires requires a sophisticated hightemperature combustion facility to keep emissions within
environmental limits and use of equipment capable of handling whole discarded tires and feeding them into the combustion chamber (Lucchesi and Maschio, 1983). It has been
recognized that the BTU value of scrap tires via TDF exceeds
that of coal, while the sulfur content is in the same order of
magnitude or even lower. Cement kilns are by far the largest
users of TDF. Some cement companies have the capacity to
incinerate whole tires, thus being able to omit the comparatively expensive size reduction process. However, when account is taken of the capital investment needed for machines
for chipping the tires and the energy costs in operating such
machines, the cost benefits obtained by replacing coal or other fossil fuel with tire chips are often not substantial
(Lucchesi and Maschio, 1983).
Scrap tires (whole and shredded) have been used as a fuel
source by industries in many countries around the world including the U.S. This fuel is commonly referred to as tire
derived fuel. The tire-derived fuel market consumed 2484.4
thousand tons of scrap tires in 2003, up from 2,144.6 thousand tons in 2001 (Giray and Sonmez, 2004). The increase
was caused mainly by three factors: increased demand for
alternative fuels due to elevated energy prices, continued improvement in the quality and consistency of TDF and more
reliable delivery of a consistent TDF product. One scrap tire
can be shredded into boiler fuel and sold for a profit of about
$.50 (US EPA, 2011).
One of the more significant factors which favor the use of
large-scale TDF operators is the construction of permanent
TDF feeding systems (Giray and Sonmez, 2004). These
showcase that these major TDF end users have made a commitment to the continued use of TDF. This is very positive
for the long term stability of the TDF market and is an indicator that the use of TDF will grow past its current levels.
Pyrolysis. This is a method where tires are subjected to heat
without oxygen and the byproducts are gas, oil and carbon
char. Some of the obstacles are the high capital investment
required and steep operating costs (Raman et al. 1981). One
company, Carbon Green, has developed a method for making
pyrolysis more cost effective through the use of nanotechnol-
148
ogy to upgrade the carbon char into a cheaper and more lasting replacement for carbon black, a material used in the manufacture of car and truck tires (Baker et al. 2009).
The oil produced by the process is similar to oil conventionally produced from crude. Discarded tire pyrolysis is not
profitable at present because of lower crude oil prices. The
steel wires of steel-belted tires and char by-products are problem pyrolysis residues. Pyrolysis and its profitability was
thought to be linked to the price of oil, but even in the past
couple years with the price of oil rising above $150 a barrel,
the technology was not economically supported (Giray and
Sonmez, 2004). The other components as in carbon black and
charcoal may prove to be the indicators of financial stability.
The methane gas produced can also be used to help power the
facility. The carbon char, after extensive refining, can be sold
to a limited market as filler in some materials and as a coloring agent for some plastics. The liquid portion from the pyrolysis is a hydrocarbon material which is similar to home
heating oil or diesel fuel. The fuel has limited use due to the
contaminants from the processing of synthetic rubber. These
contaminants make the oil generally unfit to use except as a
waste fuel or as a feedstock for further refining depending on
the facility (Raman et al. 1981). Figure 1 shows the current
usage of scrap tires (Baker et al. 2009).
The next principle question is how to go about obtaining
the tires. There are a few alternatives that will be discussed in
further detail.
FIGURE 1: Current Levels of US Scrap Tire Use
VOLUME 42, NO. 2
MAY 2016
the same equipment as those used to have the tire be processed into ground rubber. Table 2 displays typical costs for
handling the tires. This shows the breakdown of the costs
associated with the steps of tire collection. There are approximately 100 tires in a ton. One of the biggest fluctuating factors is the cost of fuel.
Now that the numerous methods to dispose of used tires
have been explored, the objective of this study is determining
which method will give the greatest benefit at the minimum
amount of expense. The optimum outcome will provide for
an energy efficient solution that is cost effective. There are
many methods of evaluating alternatives based off of different criteria that can differentiate most economical and practical method.
Scrap Tire Acquisition

The two main options are either to have the customers
drop off or haul in their tires or run a collection route. (RMA,
2005) found the maximum outward distance for a profitable
tire collection route to be 150 miles. This is due to the largest
cost of transporting the tires. An industry rule of thumb for
trucking costs is $1 per mile which would be incurred regardless of the number tires loaded on a truck. This cost must be
applied for the entire round trip. This amount was calculated
based on an expected collection fee of $.75 per tire collected
and a $.50 per tire processing profit. Table 1 shows a comparison between the costs of sending scrap tires to the landfill as
opposed to recycling them.
Table 1 may be viewed as somewhat misleading due to
the fact that once the tires are placed in the landfill their costs
are complete while the recycled tires still need additional
costs for processing. As an industry standard, it costs an average of $0.05 every time a tire is handled (TDF, 2003) and
civil engineering applications are usually easier to start and
operate when compared to ground rubber operations. An average startup cost is $2 of capital expense for each tire to be
processed. One of the benefits of this industry is that the
equipment used for handling and shredding tires into TDF
and civil engineering application material is almost exactly
Methods of Multi-Criteria Decision Analysis

Multi-Criteria Decision Analysis (MCDA) is the process
of structuring and solving problems in the development phase
that involve multiple criteria. Usually, there is not one optimal solution that will be the best at meeting all of the different criteria. Thus it is necessary to find a way of differentiating among the solutions. The main goal of all the MCDA
methods is to evaluate and choose among the alternatives
TABLE 1
A Comparison of the Cost of Landfilling and Recycling
Disposal in Landfill
Hauling for Recycling
Capital and Installation Costs:
$0
$0
Municipal Landfill Costs:
$25,000
$0
Hauling Fee:
$3,750
$5,000
Total Cost
$28,750
$5,000
Total Recovered Income
$0
$3,125
Net Annual Cost/Benefit
($28,750)
($1,875)
TABLE 2
Cost of Handling Scrap Tires
Process
Cost Per Tire
Cost Per Ton
Labor to Load/Unload
$0.10
$10
Transportation
$0.25 to $0.35
$25 to $35
Whole Tire Disposal
$0.10 to $0.30
$ 10 to $30
Profit for Collector
$0.25 to $0.55
$25 to $55
TOTAL COLLECTION COST
$1.00
$100
149
based on these criteria using a systematic analysis. The difficulty of this originates from the realization that one or more
of criterion may be in conflict. There may no longer be a
unique optimal solution to the MCDA problem that can be
obtained without incorporating preference information.
In this situation, the optimal solution needs to be replaced by
a set of non-dominated solutions. A non-dominated solution
has the property that it is not possible to choose another solution without sacrificing in at least one criterion. Thus, it is
most advantageous to choose a solution from the nondominated set due to the probability of doing better in some
or all of the criteria, and not being the worst in any of them.
Unfortunately in most cases, the amount of non-dominated
solutions is too large to accurately make an efficient choice.
Thus, there must be a tradeoff for obtaining a favorable result
in certain criteria by having undesirable outcome in others.
MCDA methods can generally be classified into three
main categories (Belton and Stewart, 2002):
1. Value measurement models, which are represented by
numerical scores. They are created in order to represent
the degree to which one decision option may be preferred
to another. These scores are developed individually for
each criteria, and are then synthesized in order to effect
aggregation into higher level preference models;
2. Goal or reference level models, which use desirability
levels for each criterion. This process then works to uncover options which are closest to achieving these desirable outcomes;
3. Outranking models, that apply alternative courses of action for creating pairwise, terms for each criteria in order to identify the extent to which a preference for one
over the other can be asserted. In aggregating such preference information across all relevant criteria, the model
seeks to establish the strength of evidence favoring selection of one alternative over another.
Some of the drawbacks of these MCDA methods are the
reliance of the decision maker to be rational, have enough
knowledge of the situation and needs, and for the decision
maker to be consistent in all judgments made during the process. This paper attempts to compare several of the MCDA
methods in terms of addressing the inherent complexity of
recycling scrap tires.
Analytic Hierarchy Process (AHP)

Analytic Hierarchy Process (AHP) is a method that was
developed using the principles of decomposition, pairwise
comparison, and priority vector generation with synthesis.
The AHP method was created by Saaty in 1977 through the
derivation of his theory of prioritized hierarchies. AHP
operates using the development of various functions to rank,
select, evaluate and benchmark the available criteria (Wasil
and Golden, 2003). The scale ranges from 1 to 9 using 1 for
valued less than', to 5 for àbout equal' and to 9 for àbsolutely more important (Vaidya and Kumar, 2006).
According to Liberatore and Nydick (2008), the main
uniqueness of AHP is its inherent capability to weight a large
number of different factors, of different natures, including
150
both qualitative and quantitative data, in order to make a decision based on a numerical basis. The main reason AHP is so
effective is its use of pairwise comparison. The pairwise
comparisons between individual alternatives were found to be
more reliable in calculating the actual weights than estimating
them indirectly. This is due to the fact that it is generally easier to evaluate the relative weights of the criteria between two
alternatives than to do so instantaneously with respect to all
of the others (Salmeron and Herrero, 2005). This method is
represented by using the D-Sight software.
The formation of AHP involves four steps, as follows:
(Douligeris and Pereira, 1992)
Step 1: Setting up the hierarchy.
Step 2: Collecting input data by pair-wise comparisons of
decision elements. Every attribute on each level is compared
with their adjacent attributes in respect of their importance to
the criteria.
Step 3: Use the èigenvalue' method to estimate the relative
weight for the criteria.
Step 4: Aggregate the relative weights of criteria to arrive at
a set of scores for the decision alternatives. The scores reflecting the weights given to each attribute are adjusted and
then summed to yield a final score for each option.
One of the essential benefits of AHP is its flexibility of
integration with other techniques. Badri (2001) researched
combining AHP with the Goal Programming method. Goal
Programming (GP) is a mathematical variation of linear programming which is used to handling the dilemma of having
multiple goals which conflict. These two techniques were
first combined in the evaluation of quality control systems.
Kwong and Bai (2002) next combined AHP with Quality
Function Deployment (QFD) and applied the combined AHPQFD approach to aid new product development. Unfortunately, it was found in many instances that pairwise comparison
ratios of the general became subject to individual judgments
subjecting the AHP to imprecision.
Fuzzy Analytic Hierarchy Process (FAHP)

Because of that drawback, there was sometimes an inability to obtain the relative weights necessary to meet the requirements of the criteria. This led to the introduction of
fuzzy numbers into the evaluation of pairwise comparisons
(Leung and Cao, 2000). Chang et al. (2009) found that the
uncertainty in the preference judgments would give rise to
uncertainty in the ranking of alternatives as well as causing
difficulty in determining the consistency of preferences. Despite the convenience of AHP in handling both quantitative
and qualitative criteria of multi-criteria decision making
problems based on decision makers' judgments, the fuzziness
and vagueness existing in many decision making problems
may contribute to the imprecise judgments of decision makers in conventional AHP approaches. Given the subjective
and qualitative nature of some service evaluation criteria,
decision makers found it extremely difficult to express the
strength of their preferences and to provide exact pair-wise
comparison judgments. Thus, conventional AHP does not
reflect natural human thinking. In order to avoid these risks
VOLUME 42, NO. 2
MAY 2016
on performance, fuzzy AHP (FAHP), a fuzzy extension of

AHP, was developed to address hierarchical fuzzy problems.
Fu et al. (2006) studied the application of FAHP to assign
relative weights between factors which affected the entry to
an electronic marketplace in Taiwanese industry. A threelayer hierarchical structure was proposed, and data was gathered through questionnaires distributed among 20 executives
and other people to obtain reliable information about strategic
factors necessary for a corporate decision about going into a
new emerging market. Jagananthan et al. (2007) employed
fuzzy AHP to facilitate the selection and evaluation of new
manufacturing technologies and concluded that fuzzy preference programming was advantageous in solving this problem.
FAHP was able to overcome issues such as contradiction,
inconsistency on fuzzy judgment matrices, and group decision-making. The ability of FAHP to address the uncertainty
in the judgments made by the users is a primary benefit compared to traditional AHP. Since environmental issues can be
very complex with lots of uncertainty involved, this method
may prove more suitable to evaluate the alternatives.
However, it must be kept in mind that AHP and FAHP are
not both going to be applicable. This depends solely on how
certain the users are in applying the information and evaluation. If the user is fairly confident, the classical AHP method
may be applied. When the information is not certain, FAHP
will provide the more efficient option. Since in most instances there will be at least a few uncertainties in the decision
problem, the fuzzy approach is most often applied. This
method was performed through use of the Make It Rational
software.
Multi-attribute Value Theory (MAVT)

The primary purpose for MAVT is modeling a representation of the decision maker's preferential system into a value
function V(a) (Von Winterfeldt and Edwards, 1986):
V (a ) = F (V1 (a1 ).......F (Vn (a n ))
(1)
where alternative a represents the vector of the evaluation

criteria. The objective for the function is to calculate the value of a that will maximizes the overall value of V(a). This
value function can best determine the decision maker's preference when several of criteria are either uncertain, or conflicting. This may be further extended to incorporate probabilities and adapt to additional expectations (Belton and
Steward, 2002). The objective of decision makers for this
process is to identify the alternative a that maximizes the
overall value of V(a). The most widely used form of this form
of function F is an additive model:
V (a ) = w1V1 (a1 ) + w2V2 (a 2 ).....wnVn (a n )
(2)
where the total of all the weights will add up to a sum of 1.

Due to the fact that low numerical scores on some criteria
can be compensated for by high scores on other criteria,
MAVT is considered to be a compensatory method. The Analytical software package uses this method.
Potentially all Pairwise Rankings of all

Possible Alternatives (PAPRIKA)
The PAPRIKA method is composed of software that asks
the decision maker a series of questions involving two hypothetical alternatives involving just two of the criteria. These
questions force the decision maker to have a tradeoff between
the criteria. These questions are asked using all of the
tradeoffs between all of the pairs of hypothetical alternatives
until there is enough information to accurately rank the alternatives being considered. The number of questions is minimized through the application of the transitive property
where:
if alternative A is ranked above alternative B, and B
is ranked above alternative C, then, logically, A is
ranked above C (Hansen & Ombler, 2008).
This means that each time a question is answered, PAPRIKA will eliminate all other potential questions that would
be considered corollaries of those already answered. The total
number of questions depends only on the amount of criteria
and how many choices. From the answers, weights of the
criteria are obtained from mathematical methods such as linear programming. 1000 Minds software is shown to use this
method.
Methodology
In evaluating the various tire recycling and reuse methods,
one of the primary factors will be the cost. Individuals and
private companies most likely will not be willing to pay high
fees to dispose of their tires. Each method will have its cost
calculated for each step in its process. Any revenue that is
generated will also be considered in the cost analysis.
When obtaining scrap tires the first operation that typically needs to be performed is to reduce the volume by shredding the tire. A tire recycling plant will cost on average 5
million dollars to build and will be able to handle a daily load
of 2 to 3 million tons of scrap tires (Scrap Tire News, 2011).
Complex machinery is needed to grind radial tires into steel
particles and shreds of crumb rubber. The tires are typically
shredded into chips, mostly 2 in size. Through shredding,
the volume of the scrap tires can be reduced to about , thus
reducing storage and shipping costs. It also makes the tire
chips easier to handle with standard equipment like front end
loaders or bobcats. The smaller the tire chip the more expensive and less time intensive the process will be as shown in
Table 3.
The RMA reported the amount of profit per process in
their 2003 report. It is listed in Table 4. The steel itself can be
sold for $.40 per tire and can be added to each process. The
main factor that has since impacted these processes has been
the increasing price of fuel. Because of this some of the processes which are able to replace the use of petroleum such as
TDF, RMA, and Pyrolysis have become significantly more
profitable and popular. However, the price of oil is volatile
and a market bottom will wipe out many in those sectors.
Thus it is best to use the average figures and store the profits
in case of downturn. Ferrer (1996) developed an equation for
151
TABLE 3
Tire Chip Size with Cost and Process Rate
Size
Description
Application
(in)
Cost Per Ton
Process Rate
($)
(Tons/Hour)
Clean cut. Ply & bead steel

remains
Cement Kilns, Civil Eng.

Applications
10
10 to 12
2 minus
Minimal wire, cut beads

removed by magnets
Industrial, utility, pulp paper

mill boilers
25
1 nominal
Same as 2" minus with extra

shredded pass
Power utility boilers
10 to 30
4 to 5
1/2 minus
Truly wire free, requires add.

shredding equipment
Feedstock in crumb rubber,

playground, fields
25 to 55
2 to 3
TABLE 4
Energy Values Input
Process
Energy Required (kW*hr/kg)
Manufacture a Tire
32
Produce a Rubber Compound
25
Incinerate a Tire
-9
Shredding Scrap Tires
1.2
calculating the profit that could be made for each method:

P = TF + R - PC - TR D
(3)
where P represents the profit, TF is the tipping fee, R is the

revenue generated from the final product, PC is the process
cost, TR is the cost of transporting the tires, and D is the cost
to dispose of the left over waste or ash.
The next factor to consider is the environmental impact. It
is necessary to look at potential pollutants and the damage
they may cause as well as take into account any risks to peoples health. For this research air pollutants will be exclusively explored because they exist in greater concentrations than
trace water pollutants and are easier to measure.
A sometimes hidden benefit is the amount of energy produced by each process. There are some of these methods that
require more energy for the process that will be given off.
Exothermic processes may be lead to a trade-off between
lower profitability and higher energy production. The cost of
the energy generated could help mitigate the cost of extraction although this will fluctuate with energy prices. Table 4
shows the energy required for the primary steps of tire recycling and will be one of the main contributors in the energy
generation calculations. One of the downsides to incinerating
scrap tires is that it takes 3 to 4 times the energy to produce
tire rubber than the amount of energy recovered by pyrolysis
or TDF. Ideally, the use of scrap tires for its originally in-
152
tended purpose or some other closely related use makes by

far more sense than incineration, both environmentally and
economically.
A practical consideration is the accessibility and availability of the solution. A big portion of this will come from the
distance from the tire source to the destination and what restrictions will be placed on each process. Many solutions are
best suited for a regional area. Different states and municipalities have their own laws so what looks feasible in one area
may not work as well in others.
Each solution will be evaluated on how it produces on
each factor and a ranking system will be developed. The solutions that perform the best in all the categories will be given
better scores and a cutoff score would eliminate the possible
solutions which were weak in one or more factors.
Only MCDA methods that considered the existence of a
hierarchy of criteria with multiple objectives and alternatives
were considered for this research. Thus, many other methods
that dealt with multi-objective planning particularly those that
used single objective optimization with constraints as other
goals, or use penalty functions, were omitted.

In order to evaluate the processes of scrap recycling there
are two tasks that need to be performed. First numerical val-
VOLUME 42, NO. 2
MAY 2016
ues for each of the criteria (profitability, energy generation,

and air pollutants emitted) must be calculated on all the processes. These values will then be used by the MCDA methods to evaluate their effectiveness and determine which process will be the best option.
CRITERIA CALCULATION
The profitability of each method is calculated using the
costs determined from the data from Table 1 and Table 2 as
well as the incorporation of equation (3). Some factors that
were not considered due to their wide variation include the
cost of the land, energy in both consumption and production,
and the initial capital investment. The profitability of each
option was calculated and is shown in Table 5.
Many of the processes require shredding the tires which

will have a cost that will reduce the total profitability. Retreading the tires and selling them on the open market involves minimal initial cost and is thus the most profitable but
may not be practical for many damaged tires. Meanwhile,
sending tires to the landfill requires a tipping fee and generates no additional income. The processes that require incineration are also unprofitable due to the costs of shredding the
tires with the energy costs of burning is greater than the value
of energy that would be generated in return.
The next criteria that needed to be calculated were net
energy generation. This calculation was heavily influenced by
the data in Table 4 as well as some values obtained from research sources. At this point only tire derived fuels and pyrolysis generate energy and many of the others are either essentially neutral or are large energy consumers. Tire stock-
TABLE 5
Profitability for Each Process
Process
Profit per Tire
Retread and Resell Tires
$8.00
Ground Rubber
$2.00
Civil Engineering
$1.56
Tire Derived Fuel (TDF)
$0.54
Tire Stockpile
$0.00
Rubber Modified Asphalt (RMA)
-$0.48
Pyrolysis
-$0.74
Landfill
-$6.00
TABLE 6
Net Energy Produced
Process
Energy Consumed
Energy Produced
Net Energy Production
(kW)
(kW)
(kW)
Tire Derived Fuel
86.2
560
473.8
Pyrolysis
129.2
245.6
116.4
Tire stockpiles
Civil Engineering Applications
1.2
-1.2
Landfilling
1.2
-1.2
Crumb Rubber
1.2
-1.2
Rubber Modified Asphalt
86.2
-86.2
Reuse and Retread Tires
174.9
-174.9
153
piles contain a great store of potential energy but there is no

constructive method of harnessing it.
The final criteria were the environmental output for each
process. The release of toxic substances in the air for three of
the most common air pollutants were determined and then
added together to develop a relative index of pollutants. This
was necessary to simplify the process and prevent thousands
of tradeoff determinations within these criteria. The pollutant
information is given in Table 7.
Once these calculations were completed there was now a
basis for each of the MCDA methods to be run. In this study
profitability was given the primary influence as that appears
to be the most motivating influence of the tire recycling market. Many companies will seek other endeavors if they start
to lose money. The second higher consideration is to keep the
amount of air particles emitted to be as low as possible. All of
the MCDA methods were run and the weights that were determined by their questions are given in Table 8.
The weights are then used by the DCMA methods to obtain a score which is used to evaluate which process should
be applied. The scores obtained from each process are displayed in Table 9.
From Table 9 it appears that the resold and retreaded tires
would be the most viable option. Retreading and reusing tires

as long as they are safe is the least costly option and generates no air pollutants. This also helps to reduce the demand
for new tires. However, the benefits from not replacing are
temporary. Eventually the tire will sustain enough wear or
damage that repair is no longer a feasible option. Additionally
only a small percentage of the tires are in good enough condition and it is not practical to sort them out at the recycling
facility. Crumb Rubber then becomes the next best option
according to three of the DCMA methods. Civil Engineering
Applications also scores very highly and would be considered
an acceptable alternative. The only less than ideal application
is using the scrap tires for coral reefs which is a very expensive option. There is the potential for the tires to leak out their
toxic chemicals and kill the wild life. Another point is that
very few people live near the coastal areas which would mean
spending an additional cost for the transport of the tires to the
location. However, there is not much demand for this so very
few tires are applied in this fashion. Landfilling was shown to
be a bad option by all of the DCMA methods. It is clear that
TDF is the best option for energy production, reusing tires is
the best option for profitability, and landfilling is not a good
fit for any of the criteria.
TABLE 7
Emitted Air Pollutants for Each Process
SO x
NO x
CO
Total
Pollutants
(ppm)
(ppm)
(ppm)
(ppm)
Crumb Rubber
Landfilling
76.7
124.4
259.5
460.6
Tire stockpile Fires
275
112
116
503
Pyrolysis
269.6
156
40
465.6
Tire Derived Fuel
630
222
30
882
Process
TABLE 8
Calculated Criteria Weights for Each MCDA Method
Criteria
154
MCDA Method Weights
Analytical
Make It Rational
D-Sight
1000 Minds
Profit Per Tire
.573
.677
.501
.634
Net Energy Generated
.146
.182
.217
.151
Air Pollutants Emitted
.281
.141
.282
.225
VOLUME 42, NO. 2
MAY 2016
TABLE 9
Summary of DMCA Software Results
Process
Analytical
Make It Rational
D-Sight Score
1000 Minds Score
Mean Score
Reuse and Retread Tire
77.01
67.68
72.16
89.5
76.59
Crumb Rubber
57.23
14.21
74.36
66.7
54.25
Civil Engineering
Applications
66.93
17.93
67.25
64.9
53.13
Pyrolysis
25.18
32.54
33.88
56.1
46.93
Tire Stockpiles
46.79
48.52
40.93
47.4
45.91
Tire Derived Fuel

(TDF)
26.66
34.98
50.18
43.9
41.43
35.50
31.74
29.54
35.1
32.97
Landfilling
24.51
19.62
31.70
18.6
23.61
CONCLUSION
Unwanted tires cause a great many problems in the field
of solid waste management. They are very bulky and take up
a lot of premium space in the landfills. Other animals like to
use them as homes where they can spread diseases. There are
several methods of disposing of these tires, but no universal
method that is considered the best. The biggest factor is the
local market. This study looked at each method on the basis
of cost/benefit analysis, environmental impact, and feasibility. It will be beneficial to all if the best processes such as tire
retreading, crumb rubber, and civil engineering applications
were applied regularly and the inefficient processes such as
landfilling were eliminated entirely.
ACKNOWLEDGEMENTS
This work was sponsored by the Department of Energy
Samuel Massie Chair of Excellence Program under Grant No.
DE-NA0000718. The views and conclusions contained herein
are those of the writers and should not be interpreted as necessarily representing the official policies or endorsements,
either expressed or implied, of the funding agency. A previous version of this research was presented at the 24th International Conference on Solid Waste Technology and Management at Philadelphia, PA on the dates of March 30 to April 2,
2014.
REFERENCES
Amari, T., N.J. Themelis, and I.K. Wernik. (2000). Resource
Recovery from Used Tires. Resources Policy. Volume
25, pp. 179-188.
Baker, T.E., T.M. Allen, D.V. Jenkins, and D.T. Mooney.

(2003). Evaluation of the Use of Scrap Tires in Transportation Related Applications in the State of Washington.
Washington State Department of Transportation. Report
to the Legislature SHB 2308.
Badri, M.A. (2001). A combined AHP-GP model for quality
control systems. Journal of Production Economics. Volume 72. pp. 27-40.
Belton, S. and T.S. Stewart. (2002). Multiple Criteria Decision Analysis: An Integrated
Approach. Kluwer Academic Publishers. Massachusetts.
Birkholz, D.A., K.L. Belton, and T.L. Guidotti., (2003).
Toxicological Evaluation for the Hazard Assessment of
Tire Crumb for Use in Public Playgrounds. Journal of
Air & Waste Management Association. Volume 53, pp.
903-907.
Chang, N. B. (2008). Economic and policy instrument analyses in support of the scrap tire recycling program in Taiwan.
Journal
of
Environmental
Management
Volume 86, pp. 435-450.
Chang, C.W., Wu, C.R., and Lin, H.L. (2009). Applying
fuzzy hierarchy multiple attributes to construct an expert
decision making process. Expert Systems with Applications Volume 36, pp. 7363-7368.
Dierks, W. (2011). Incorporating the Use of Recycled Rubber. Accessed at http://www.trb.org/Main/Blurbs/Uses_
of_Recycled_Rubber_Tires_in_Highways_154432.aspx
on April 23, 2011.
Douligeris, C. and Pereira, I. (1992). An analytic hierarchy
process approach to the analysis of quality in telecommunication system. IEEE Journal. Volume 48. pp. 16841688.
Everett J.W., and S. Douglah, ( 1998). Scrap tire disposal
(II): case study and recommendations. Journal of Solid
Waste Technology and Management. Volume 25, pp. 5160.
155
Fu, H.P., Ho, Y.C., Chen, R.C., Chang, T.H., and Chein, P.H.
(2006). Factors affecting the adoption of electronic marketplaces. Journal of Operation and Production Management. Volume 26. pp. 1301-1324.
Giray, E.S. and O. Sonmez,(2004). Supercritical extraction
of scrap tire with different solvents and effect of tire oil on
the supercritical extraction of coal. Fuel Processing
Technology, Volume 85, pp. 251-265.
Jagananthan, S., Erinjeri, J.J., and Ker, J.I. (2007). Fuzzy
analytic hierarchy process based group decision support
system to select and evaluate new manufacturing technologies. International Journal of Advanced Manufacturing
Technology. Volume 32. pp. 1253-1262.
Jang, J.W., T.S. Yoo, J.H. Oh, and I. Iwasaki, (1998). Discarded tire recycling practices in the United States, Japan
and Korea. Resources, Conservation and Recycling.
Volume 22. pp. 1-14.
Kwong, C. and Bai, H. (2002). A fuzzy AHP approach to
the determination of importance weights of customer requirements in quality function deployment. Journal of
Intelligent Manufacturing. Volume 13. pp. 367-377.
Leung, L. and Cao, D. (2000). On consistency and ranking
of alternatives in fuzzy AHP. Journal of Operational Research. Volume 124. pp. 102-113.
Liberatore, M.J. and Nydick, R.L. (2008). The analytic hierarchy process in medical and health care decision making:
A literature review. European Journal of Operational Research. Volume 189. pp. 194-207.
Lucchesi, A. and G. Maschio, (1983). Semi active carbon
and aromatics produced by pyrolysis of scrap tires. Cons
Recycl. Volume 6. pp. 85-90.
Raman, K.P., W.P. Walawender and L.T. Fan, (1981)
Gasfication of waste tires in a fluid bed reactor. Cons
156
Recycl . Volume 4. pp. 79-88.

Reschner, K. (2008). Scrap Tire Recycling: A Summary of
Prevalent Disposal and Recycling Methods. Accessed at
http://entire-engineering.de/Scrap_Tire_Recycling.pdf on
April 10, 2011.
Rubber Manufacturers Association (RMA). (2003). Considerations for Starting a Scrap Tire Company: A Blueprint
for Planning a Business Strategy. RMA Special Report
GEN-060.
Salmeron, J.L. and Herrero, I. (2005). An AHP based methodology to rank critical success factors of executive information systems. Computer Standards and Interfaces
Scrap Tire News. (2011). Crumb Rubber Information. Accessed at http://www.scraptirenews.com/crumb.php on
March 25, 2011.
Saaty, T.L. (1980). The analytic hierarchy process. Mc-Graw
Hill, New York.
US EPA. (2011). Tire Derived Fuels. Accessed at
http://www.epa.gov/osw/conserve/materials/tires/tdf.htm
on March 14, 2011.
US EPA. (2015). State Scrap Tire Programs: A Quick Reference Guide. Accessed at http://epa.gov/osw/conserve/ materials/tires/basic.htm on June 24, 2015.
Vaidya, O.S. and Kumar, S. (2006). Analytic hierarchy process: An overview of applications. Journal of Operational Research. Volume 169. pp. 1-29.
von Winterfeldt D and W. Edwards. (1986). Decision analysis and behavioral research. Cambridge University Press.
Cambridge.
Wasil, E. and Golden, B. (2003). Celebrating 25 years of
AHP based decision making. Computers and Operation
Research. Volume 30. pp. 1419-1430.
VOLUME 42, NO. 2
MAY 2016
A MULTI-CRITERIA EVALUATION OF THE METHODS FOR

RECYCLING SCRAP TIRES
Shoou-Yuh Chang1, Frank Gronwald2
1
DOE Samuel Massie Chair Professor, Dept. of Civil Engineering

2
Research Assistant, Dept. of Civil Engineering

chang@ncat.edu
fsgronwa@aggies.ncat.edu
ABSTRACT
Worn out tires create numerous problems in the field of solid waste management. The tires are
very bulky and take up a lot of premium space in the landfills. Insects and rodents like to use
them as homes where they can spread diseases and if the tires catch fire they will emit toxic
gasses. There are several methods of disposing of these tires, but no universal method which
is considered the best for all situations. This study looks at each method of tire disposal on the
basis of cost/benefit analysis, environmental impact, and feasibility. It will be beneficial to all if
the best methods were known, and some methods were eliminated to reduce confusion. Four
different methods of Multi-Criteria Decision Analysis were performed to determine the optimal
solution through a weighted scoring system which accounts for criteria in cost/benefit analysis,
environmental impact, and energy production. The study found that retreading tires for reuse
was the best option. However, since that is not always a practical or long-term solution the
study found that either using the scrap tires to create crumb rubber or for the purpose of civil
engineering applications would be the optimal solutions.
Keywords: Tire recycling, multi-criteria decision analysis, tire derived fuel, pyrolysis
INTRODUCTION
Every year a large number of tires are discarded. The
highest portion of these is generated from automobiles. During a majority of the history of the automobile, the scrap tire
problem did not exist. The problem began in the early 1960s
and has only escalated significantly over the last 25 years. It
was reported that 266 million tires were scrapped in the US in
1996 (Everett and Douglah, 1998). About 20% of these tires
have been either recycled or burned as fuel, with the remaining 80% either stockpiled or transported to landfills. As of
2012 , between 7 and 8 billion tires were stockpiled in the
United States (EPA, 2015). Since such a large proportion of
waste tires are simply buried in landfill sites or just taking up
space, this creates a big issue as to what to do with them.
There have been numerous proposals over the years for using
these waste tires in environmentally beneficial applications to

consume as many waste tires as possible. Even today, this
dilemma still presents a considerable challenge. Most used
tires are still either sent to the landfills or tire graveyards,
where hundreds to thousands of tires create an eyesore and a
public health hazard. Discarded tires become breeding
grounds for mosquitoes and rodents, which carry a lot of diseases, including West Nile virus, encephalitis, dengue fever,
and hanta virus (Jang et al. 1998). Tire piles also attract children, who can injure themselves playing in them. Scrap tires
brought to landfills can also damage the landfill linings that
have been installed to help keep surface and groundwater free
from landfill contaminants.
Another problem with discarded tires is the risk of fire.
Tire pile fires can smolder for weeks to months, releasing
extremely toxic pollutants into the air, creating serious respir-
145
atory and other health problems for people in the vicinity and
many miles away (Amari et al. 2000). Runoff water from a
tire fire can also produce toxins, which could then contaminate the groundwater. Once a large tire pile ignites, it can be
extremely difficult and almost impossible, to extinguish. In
many cases, fires from large tire piles will be left to burn for
long periods of up to several months with the fumes being
visible for many miles. This is due to the intense heat and
smoke generation which give fire fighters an almost impossible task of putting out these fires. One of the main problems
is that the air and soil pollution will become much worse if
attempts are made to extinguish the fire using foam or water
(Amari et al. 2000). Because of these reasons, scrap tire fires
are usually permitted to burn out in a controlled manner until
the entire pile is exhausted. Governmental agencies were
forced to develop environmental policies to promote and assist the tire recycling industry in order to clean up the accumulation of stockpiles of scrap tires, including those in illegal
dump sites.
Tires are composed of vulcanized rubber and various reinforcing materials. The most commonly used rubber matrix is
the co-polymer styrene-butadiene (SBR) or a blend of natural
rubber and SBR (Jang et al. 1998). In addition to the rubber
compound, tires contain reinforcing fillers and fibers. The
reinforcing fillers are made of carbon black, and are used to
strengthen the rubber and aid abrasion resistance. The reinforcing fibers are almost always composed of textile or steel
fibers, usually in the form of a cord which are used to provide
reinforcing strength in tires. Extenders, such as, petroleum oil
are also added to control viscosity, reduce internal friction
during processing, and improve low temperature flexibility in
the vulcanized product (Jang et al. 1998).
In the early 1900s, the average recycled content of all
rubber products was over 50%. Many would say that the rubber reclaiming industry is as old as the industrial use of rubber itself. The proponent fueling the rubber reclaiming industry was the comparatively scarce supply of rubber at that
time. In 1910, one pound of natural rubber cost nearly as
much as one pound of silver and it made economic sense to
reuse as much of the rubber as possible (Reschner, 2008).
However by 1960, the recycling content in the traditional
rubber manufacturing industry dropped to around 20%. This
was due to the effects of cheap oil imports, the availability of
synthetic rubber, and the development of steel belted radial
tires. These factors have led to a steady decline of rubber
recycling. The newer steel belted radial tire designs made it
increasingly difficult to grind and slice those tires. Because of
this, the old methods for regrinding and reusing waste tires
have been made obsolete. As of 1995, the traditional tire and
rubber re-claimers could produce only about 2% recycled
material. New methods for recycling scrap tires have been
developed and are considered to be better options. These
methods will be discussed further in the next section.
Current Methods of Disposal of Used Tire

Resale. This option occurs because an individual will usually
purchase four tires at the same time when only one or two
tires need replacing. The other tires can be resold by someone
146
who specializes in used tires. The level of use of reused tires

has declined drastically. One of the major contributing factors
has been the decline in the use of snow tires. Years ago, a
large portion of car owners maintained a pair of snow tires,
which needed to be mounted on the car for the winter and
removed in the spring. These snow tires were often purchased
as retreads. The development of all-season tires virtually
eliminated the use of conventional snow tires.
Another important impact on the decline of reused tires is
the invention of radial tires. The radial tires have threads
which have more durability and enhanced service requirements such as low rolling resistance. The decline of the rubber reclaiming industry was the major cause of today's scrap
tire problem. If all 278,000 tons of reclaim used in 1960 had
been produced from scrap tires, over 50 million scrap tires
would have been required (Reschner, 2008).
Tire retreading. This is a practice done mostly for extending
the life of truck tires as their cost keeps increasing. Some of
these tires can be retreaded two to three times over the life of
the tire. Retreading a tire will not eliminate it from contributing to the scrap tire problem because the retreaded tire will
eventually wear out, probably in less time than a new tire.
However, retreading delays the appearance of those tires on
the scrap heap and thus reduces the number of scrap tires that
would have appeared if they had not been retreaded by lack
of new tire demand. The choice to resell or retread should be
made either by the end user or during collection as it is not
practical to sort them out at the tire recycling facility.
Landfilling. This involves the collection of waste tires by a
solid waste hauler who is paid to remove discarded tires from
dealer's property. Most landfills accept whole scrap tires only
at a hefty tipping fee because tires are awkward to handle and
difficult to compact. In many cases, scrap tires will work their
way to the top of a closed landfill over time, causing costly
damages to the landfill cover. In landfills, tires tend to float
or rise to the surface, piercing the landfill cover. Shredding or
splitting of tires is becoming increasingly common as part of
the disposal process. Shredding can reduce tire volume by up
to 75%. This can reduce transportation costs by 30 to 60 percent simply because fewer trips are required, and maximum
hauling weights may be achieved more easily. Shredder companies charge anywhere from $19 to $75 per ton to shred
discarded tires in the U.S (Reschner, 2008).
Civil Engineering Applications. Tire derived products are
mostly composed of 1 tire chips which can be used to replace conventional construction material, such as road fill,
gravel, crushed rock, or sand. The benefits of using tire chips
instead of conventional construction materials are a reduced
density, improved drainage properties, and better thermal
insulation (Scrap Tire News, 2011).
One interesting application is using scrap tires for the creation of underwater reefs and breakwaters. A project in Ft.
Lauderdale, Florida, U.S. has used over 3 million tires and
annually adds one million tires per year to reefs. Enthusiasm
for this use has waned, and discarded tire reefs are now only
built in minimal numbers. Currently, an estimated 120 000
VOLUME 42, NO. 2
MAY 2016
150 000 tires are used annually in the construction of reefs.

Costs are about $3.50 per tire. This compares to an average
$45.25 per ton to landfill tires in the northeastern United
States (Reschner, 2008).
Discarded tires for breakwaters which are filled with foam
are able to displace 200 pounds of water and can be used to
float a number of devices such as marinas and docks (Everett
and Douglah, 1998). Tire floats cost approximately 6 to 8
cents per pound, which is cheaper than the current popular
choice, foam filled plastic which sells for between 10 and 14
cents per lb (Reschner, 2008). Breakwaters and flotation devices are able to use approximately 30,000 to 50,000 tires per
year in the U.S.
Another popular application is to use the scrap tires as
playground equipment. Building tire playgrounds and other
small-scale local and backyard recreational uses will consume up to 7500 truck tires per year in the U.S. Other countries such as Japan and Korea will use a minimal amount of
discarded tires as playground equipment; the numbers consumed are unknown. Demand for tire playground products is
declining as the economy improves and schools and parks
select wooden playground equipment.
Erosion control is also a feasible application for scrap
tires. Discarded tires used with other stabilization materials to
reinforce an unstable highway shoulder or protect a channel
slope have been shown to produce stability. Discarded tires
were banded together and partially or completely buried on
unstable slopes. Construction costs were reduced from 50 to
75% compared to the lowest cost alternatives such as rock,
gabion, or concrete protection (Reschner, 2008).
Tire Derived Aggregate (TDA) uses shredded tire rubber
as a replacement for naturally mined materials such as gravel.
It is considered safe, reliable, and cost-effective. One of its
more useful benefits is that when used under transportation
structures such as highways and rail tracks, TDA acts to reduce noise and vibration that can affect nearby buildings and
communities. There have been tests to see whether the tires,
which contain petroleum products and metals, will leach toxics into groundwater (Chang et al. 2009). Although small
amounts of iron and manganese have been detected, they
were found to be at levels which were below health concerns.
Other tests also were performed using tiny shrimp that were
placed in the runoff water and they suffered no noticeable
health consequences (Chang, 2008).
Increasing the civil engineering market depends on three
factors: cost competitiveness of tire shreds when compared to
traditional construction materials, increased acceptance by
regulatory agencies, and increased recognition by scrap tire
processors of the emerging opportunities available for civil
engineering applications.
In 2011, Scrap Tire News found that the ground rubber
market contained two classes of particle sizes: ground rubber (10 mesh and smaller) and coarse rubber (4 mesh and
larger, with a maximum size of one-half inch). Over the past
several years there has been significant growth in the ordering of coarse sized particles. This particle range is used in
mulch, playground surfacing, running track material, soil
amendments and some bound rubber products.
Rubber Modified Asphalt (RMA). Ground rubber can be

mixed with asphalt to favorably modify the properties of the
asphalt for highway construction. Ground scrap tire rubber
can either be used as part of the asphalt rubber binder, seal
coat, cap seal spray or joint and crack sealant; or as an aggregate substitution. Currently, there is an increasing interest in
the benefits of rubber-modified asphalt, in most states currently using it, and in the other states considering its implementation. To a large extent, any large-scale increase in the
use of RMA is dependent upon the willingness of a state department of transportation to accept national test results and
begin its own state and local level programs (Dierks, 2011).
Currently there are five states which are already using RMA:
California, Arizona, Texas, Florida and South Carolina. The
use of RMA in an open graded friction course on highways
has successfully reduced road noise in Arizona and California. RMA uses between 500 and 2,000 scrap tires for each
lane mile of pavement (RMA, 2003). Other benefits include
reduction of maintenance costs, less cracking and rutting, and
anti-aging due to the anti-oxidants obtained from the tire
manufacture process. Pavements made from rubberized HMA
exhibit greater skid resistance than those made from conventional hot mix asphalt (RMA, 2003).
Asphalt rubber cant be placed during weather extremes.
Generally, the concrete pavement surface must be between 85
and 145 degrees F for the overlay to adhere and cure properly
(Scrap Tire News, 2011). In the winter months the asphalt
rubber acted as a tough surface so there was a reduction of
chain wear and the snow plows didnt chip away at the surface (Dierks, 2011). The drawbacks of RMA are its workability, recyclability, and the fumes emitted during the paving
process.
Spliting or punching of tires. Splitting involves the removal
of steel beads and then the desired shaped of the tire is
achieved using a stamp or punch. Products produced from
split tires include floor mats, belts, gasket, shoe soles, dock
bumpers, seals, muffler hangers, shims, washers, insulators,
and fishing and farming equipment (Jang et al. 1998). These
account for a small, basically static, portion of the scrap tire
marketplace. Machines are required to cut tires into strips
which are then fastened together to make the products listed
above. These machines cost about $15,000 - $18,000 each
and are very labor intensive (RMA, 2003). This operation
will only become profitable in an area of abundant and lowcost labor.
Tire Derived Fuel (TDF) Tire derived fuel was the first
market for scrap tires. From 1979 until 1992, TDF was the
essentially the only market for scrap tires. Beginning in 1992,
whole scrap tires began to be used as feedstock for ground
rubber and processed tires were applied to civil engineering
applications. The EPA considers the use of tire-derived fuels
to be a preferable alternative to the use of fossil fuels (US
EPA, 2011). Use of tire derived fuel reduces the amount of
fossil fuels that would otherwise be consumed. TDF is less
expensive than fossil fuels. One of the benefits of using TDF
is that it saves the petroleum reserves while eliminating tons
of scrap tires from landfills and scrap tire piles.
147
The burning of whole discarded tires also reduces the

need for expensive shredding operations. However, the burning of whole discarded tires requires a sophisticated hightemperature combustion facility to keep emissions within
environmental limits and use of equipment capable of handling whole discarded tires and feeding them into the combustion chamber (Lucchesi and Maschio, 1983). It has been
recognized that the BTU value of scrap tires via TDF exceeds
that of coal, while the sulfur content is in the same order of
magnitude or even lower. Cement kilns are by far the largest
users of TDF. Some cement companies have the capacity to
incinerate whole tires, thus being able to omit the comparatively expensive size reduction process. However, when account is taken of the capital investment needed for machines
for chipping the tires and the energy costs in operating such
machines, the cost benefits obtained by replacing coal or other fossil fuel with tire chips are often not substantial
(Lucchesi and Maschio, 1983).
Scrap tires (whole and shredded) have been used as a fuel
source by industries in many countries around the world including the U.S. This fuel is commonly referred to as tire
derived fuel. The tire-derived fuel market consumed 2484.4
thousand tons of scrap tires in 2003, up from 2,144.6 thousand tons in 2001 (Giray and Sonmez, 2004). The increase
was caused mainly by three factors: increased demand for
alternative fuels due to elevated energy prices, continued improvement in the quality and consistency of TDF and more
reliable delivery of a consistent TDF product. One scrap tire
can be shredded into boiler fuel and sold for a profit of about
$.50 (US EPA, 2011).
One of the more significant factors which favor the use of
large-scale TDF operators is the construction of permanent
TDF feeding systems (Giray and Sonmez, 2004). These
showcase that these major TDF end users have made a commitment to the continued use of TDF. This is very positive
for the long term stability of the TDF market and is an indica-
148
tor that the use of TDF will grow past its current levels.
Pyrolysis. This is a method where tires are subjected to heat
without oxygen and the byproducts are gas, oil and carbon
char. Some of the obstacles are the high capital investment
required and steep operating costs (Raman et al. 1981). One
company, Carbon Green, has developed a method for making
pyrolysis more cost effective through the use of nanotechnology to upgrade the carbon char into a cheaper and more lasting replacement for carbon black, a material used in the manufacture of car and truck tires (Baker et al. 2009).
The oil produced by the process is similar to oil conventionally produced from crude. Discarded tire pyrolysis is not
profitable at present because of lower crude oil prices. The
steel wires of steel-belted tires and char by-products are problem pyrolysis residues. Pyrolysis and its profitability was
thought to be linked to the price of oil, but even in the past
couple years with the price of oil rising above $150 a barrel,
the technology was not economically supported (Giray and
Sonmez, 2004). The other components as in carbon black and
charcoal may prove to be the indicators of financial stability.
The methane gas produced can also be used to help power the
facility. The carbon char, after extensive refining, can be sold
to a limited market as filler in some materials and as a coloring agent for some plastics. The liquid portion from the pyrolysis is a hydrocarbon material which is similar to home
heating oil or diesel fuel. The fuel has limited use due to the
contaminants from the processing of synthetic rubber. These
contaminants make the oil generally unfit to use except as a
waste fuel or as a feedstock for further refining depending on
the facility (Raman et al. 1981). Figure 1 shows the current
usage of scrap tires (Baker et al. 2009).
The next principle question is how to go about obtaining
the tires. There are a few alternatives that will be discussed in
further detail.
VOLUME 42, NO. 2
MAY 2016
and civil engineering application material is almost exactly

the same equipment as those used to have the tire be processed into ground rubber. Table 2 displays typical costs for
handling the tires. This shows the breakdown of the costs
associated with the steps of tire collection. There are approximately 100 tires in a ton. One of the biggest fluctuating factors is the cost of fuel.
Now that the numerous methods to dispose of used tires
have been explored, the objective of this study is determining
which method will give the greatest benefit at the minimum
amount of expense. The optimum outcome will provide for
an energy efficient solution that is cost effective. There are
many methods of evaluating alternatives based off of different criteria that can differentiate most economical an
Scrap Tire Acquisition

The two main options are either to have the customers
drop off or haul in their tires or run a collection route. (RMA,
2005) found the maximum outward distance for a profitable
tire collection route to be 150 miles. This is due to the largest
cost of transporting the tires. An industry rule of thumb for
trucking costs is $1 per mile which would be incurred regardless of the number tires loaded on a truck. This cost must be
applied for the entire round trip. This amount was calculated
based on an expected collection fee of $.75 per tire collected
and a $.50 per tire processing profit. Table 1 shows a comparison between the costs of sending scrap tires to the landfill as
opposed to recycling them.
Table 1 may be viewed as somewhat misleading due to
the fact that once the tires are placed in the landfill their costs
are complete while the recycled tires still need additional
costs for processing. As an industry standard, it costs an average of $0.05 every time a tire is handled (TDF, 2003) and
civil engineering applications are usually easier to start and
operate when compared to ground rubber operations. An average startup cost is $2 of capital expense for each tire to be
processed. One of the benefits of this industry is that the
equipment used for handling and shredding tires into TDF
Methods of Multi-Criteria Decision Analysis

Multi-Criteria Decision Analysis (MCDA) is the process
of structuring and solving problems in the development phase
that involve multiple criteria. Usually, there is not one optimal solution that will be the best at meeting all of the different criteria. Thus it is necessary to find a way of differentiating among the solutions. The main goal of all the MCDA
methods is to evaluate and choose among the alternatives
TABLE 1
A Comparison of the Cost of Landfilling and Recycling
Disposal in Landfill
Hauling for Recycling
Capital and Installation Costs:
$0
$0
Municipal Landfill Costs:
$25,000
$0
Hauling Fee:
$3,750
$5,000
Total Cost
$28,750
$5,000
Total Recovered Income
$0
$3,125
Net Annual Cost/Benefit
($28,750)
($1,875)
TABLE 2
Cost of Handling Scrap Tires
Process
Cost Per Tire
Cost Per Ton
Labor to Load/Unload
$0.10
$10
Transportation
$0.25 to $0.35
$25 to $35
Whole Tire Disposal
$0.10 to $0.30
$ 10 to $30
Profit for Collector
$0.25 to $0.55
$25 to $55
TOTAL COLLECTION COST
FIGURE
$1.00 1
Current Levels of US Scrap Tire Use
$100
149
based on these criteria using a systematic analysis. The difficulty of this originates from the realization that one or more
of criterion may be in conflict. There may no longer be a
unique optimal solution to the MCDA problem that can be
obtained without incorporating preference information.
In this situation, the optimal solution needs to be replaced by
a set of non-dominated solutions. A non-dominated solution
has the property that it is not possible to choose another solution without sacrificing in at least one criterion. Thus, it is
most advantageous to choose a solution from the nondominated set due to the probability of doing better in some
or all of the criteria, and not being the worst in any of them.
Unfortunately in most cases, the amount of non-dominated
solutions is too large to accurately make an efficient choice.
Thus, there must be a tradeoff for obtaining a favorable result
in certain criteria by having undesirable outcome in others.
MCDA methods can generally be classified into three
main categories (Belton and Stewart, 2002):
1. Value measurement models, which are represented by
numerical scores. They are created in order to represent
the degree to which one decision option may be preferred
to another. These scores are developed individually for
each criteria, and are then synthesized in order to effect
aggregation into higher level preference models;
2. Goal or reference level models, which use desirability
levels for each criterion. This process then works to uncover options which are closest to achieving these desirable outcomes;
3. Outranking models, that apply alternative courses of action for creating pairwise, terms for each criteria in order to identify the extent to which a preference for one
over the other can be asserted. In aggregating such preference information across all relevant criteria, the model
seeks to establish the strength of evidence favoring selection of one alternative over another.
Some of the drawbacks of these MCDA methods are the
reliance of the decision maker to be rational, have enough
knowledge of the situation and needs, and for the decision
maker to be consistent in all judgments made during the process. This paper attempts to compare several of the MCDA
methods in terms of addressing the inherent complexity of
recycling scrap tires.
Analytic Hierarchy Process (AHP)

Analytic Hierarchy Process (AHP) is a method that was
developed using the principles of decomposition, pairwise
comparison, and priority vector generation with synthesis.
The AHP method was created by Saaty in 1977 through the
derivation of his theory of prioritized hierarchies. AHP
operates using the development of various functions to rank,
select, evaluate and benchmark the available criteria (Wasil
and Golden, 2003). The scale ranges from 1 to 9 using 1 for
valued less than', to 5 for àbout equal' and to 9 for àbsolutely more important (Vaidya and Kumar, 2006).
According to Liberatore and Nydick (2008), the main
uniqueness of AHP is its inherent capability to weight a large
number of different factors, of different natures, including
150
both qualitative and quantitative data, in order to make a decision based on a numerical basis. The main reason AHP is so
effective is its use of pairwise comparison. The pairwise
comparisons between individual alternatives were found to be
more reliable in calculating the actual weights than estimating
them indirectly. This is due to the fact that it is generally easier to evaluate the relative weights of the criteria between two
alternatives than to do so instantaneously with respect to all
of the others (Salmeron and Herrero, 2005). This method is
represented by using the D-Sight software.
The formation of AHP involves four steps, as follows:
(Douligeris and Pereira, 1992)
Step 1: Setting up the hierarchy.
Step 2: Collecting input data by pair-wise comparisons of
decision elements. Every attribute on each level is compared
with their adjacent attributes in respect of their importance to
the criteria.
Step 3: Use the èigenvalue' method to estimate the relative
weight for the criteria.
Step 4: Aggregate the relative weights of criteria to arrive at
a set of scores for the decision alternatives. The scores reflecting the weights given to each attribute are adjusted and
then summed to yield a final score for each option.
One of the essential benefits of AHP is its flexibility of
integration with other techniques. Badri (2001) researched
combining AHP with the Goal Programming method. Goal
Programming (GP) is a mathematical variation of linear programming which is used to handling the dilemma of having
multiple goals which conflict. These two techniques were
first combined in the evaluation of quality control systems.
Kwong and Bai (2002) next combined AHP with Quality
Function Deployment (QFD) and applied the combined AHPQFD approach to aid new product development. Unfortunately, it was found in many instances that pairwise comparison
ratios of the general became subject to individual judgments
subjecting the AHP to imprecision.
Fuzzy Analytic Hierarchy Process (FAHP)

Because of that drawback, there was sometimes an inability to obtain the relative weights necessary to meet the requirements of the criteria. This led to the introduction of
fuzzy numbers into the evaluation of pairwise comparisons
(Leung and Cao, 2000). Chang et al. (2009) found that the
uncertainty in the preference judgments would give rise to
uncertainty in the ranking of alternatives as well as causing
difficulty in determining the consistency of preferences. Despite the convenience of AHP in handling both quantitative
and qualitative criteria of multi-criteria decision making
problems based on decision makers' judgments, the fuzziness
and vagueness existing in many decision making problems
may contribute to the imprecise judgments of decision makers in conventional AHP approaches. Given the subjective
and qualitative nature of some service evaluation criteria,
decision makers found it extremely difficult to express the
strength of their preferences and to provide exact pair-wise
comparison judgments. Thus, conventional AHP does not
reflect natural human thinking. In order to avoid these risks
VOLUME 42, NO. 2
MAY 2016
on performance, fuzzy AHP (FAHP), a fuzzy extension of

AHP, was developed to address hierarchical fuzzy problems.
Fu et al. (2006) studied the application of FAHP to assign
relative weights between factors which affected the entry to
an electronic marketplace in Taiwanese industry. A threelayer hierarchical structure was proposed, and data was gathered through questionnaires distributed among 20 executives
and other people to obtain reliable information about strategic
factors necessary for a corporate decision about going into a
new emerging market. Jagananthan et al. (2007) employed
fuzzy AHP to facilitate the selection and evaluation of new
manufacturing technologies and concluded that fuzzy preference programming was advantageous in solving this problem.
FAHP was able to overcome issues such as contradiction,
inconsistency on fuzzy judgment matrices, and group decision-making. The ability of FAHP to address the uncertainty
in the judgments made by the users is a primary benefit compared to traditional AHP. Since environmental issues can be
very complex with lots of uncertainty involved, this method
may prove more suitable to evaluate the alternatives.
However, it must be kept in mind that AHP and FAHP are
not both going to be applicable. This depends solely on how
certain the users are in applying the information and evaluation. If the user is fairly confident, the classical AHP method
may be applied. When the information is not certain, FAHP
will provide the more efficient option. Since in most instances there will be at least a few uncertainties in the decision
problem, the fuzzy approach is most often applied. This
method was performed through use of the Make It Rational
software.
Multi-attribute Value Theory (MAVT)

The primary purpose for MAVT is modeling a representation of the decision maker's preferential system into a value
function V(a) (Von Winterfeldt and Edwards, 1986):
V (a ) = F (V1 (a1 ).......F (Vn (a n ))
(1)
where alternative a represents the vector of the evaluation

criteria. The objective for the function is to calculate the value of a that will maximizes the overall value of V(a). This
value function can best determine the decision maker's preference when several of criteria are either uncertain, or conflicting. This may be further extended to incorporate probabilities and adapt to additional expectations (Belton and
Steward, 2002). The objective of decision makers for this
process is to identify the alternative a that maximizes the
overall value of V(a). The most widely used form of this form
of function F is an additive model:
V (a ) = w1V1 (a1 ) + w2V2 (a 2 ).....wnVn (a n )
(2)
where the total of all the weights will add up to a sum of 1.

Due to the fact that low numerical scores on some criteria
can be compensated for by high scores on other criteria,
MAVT is considered to be a compensatory method. The Analytical software package uses this method.
Potentially all Pairwise Rankings of all

Possible Alternatives (PAPRIKA)
The PAPRIKA method is composed of software that asks
the decision maker a series of questions involving two hypothetical alternatives involving just two of the criteria. These
questions force the decision maker to have a tradeoff between
the criteria. These questions are asked using all of the
tradeoffs between all of the pairs of hypothetical alternatives
until there is enough information to accurately rank the alternatives being considered. The number of questions is minimized through the application of the transitive property
where:
if alternative A is ranked above alternative B, and B
is ranked above alternative C, then, logically, A is
ranked above C (Hansen & Ombler, 2008).
This means that each time a question is answered, PAPRIKA will eliminate all other potential questions that would
be considered corollaries of those already answered. The total
number of questions depends only on the amount of criteria
and how many choices. From the answers, weights of the
criteria are obtained from mathematical methods such as linear programming. 1000 Minds software is shown to use this
method.
Methodology
In evaluating the various tire recycling and reuse methods,
one of the primary factors will be the cost. Individuals and
private companies most likely will not be willing to pay high
fees to dispose of their tires. Each method will have its cost
calculated for each step in its process. Any revenue that is
generated will also be considered in the cost analysis.
When obtaining scrap tires the first operation that typically needs to be performed is to reduce the volume by shredding the tire. A tire recycling plant will cost on average 5
million dollars to build and will be able to handle a daily load
of 2 to 3 million tons of scrap tires (Scrap Tire News, 2011).
Complex machinery is needed to grind radial tires into steel
particles and shreds of crumb rubber. The tires are typically
shredded into chips, mostly 2 in size. Through shredding,
the volume of the scrap tires can be reduced to about , thus
reducing storage and shipping costs. It also makes the tire
chips easier to handle with standard equipment like front end
loaders or bobcats. The smaller the tire chip the more expensive and less time intensive the process will be as shown in
Table 3.
The RMA reported the amount of profit per process in
their 2003 report. It is listed in Table 4. The steel itself can be
sold for $.40 per tire and can be added to each process. The
main factor that has since impacted these processes has been
the increasing price of fuel. Because of this some of the processes which are able to replace the use of petroleum such as
TDF, RMA, and Pyrolysis have become significantly more
profitable and popular. However, the price of oil is volatile
and a market bottom will wipe out many in those sectors.
Thus it is best to use the average figures and store the profits
in case of downturn. Ferrer (1996) developed an equation for
151
TABLE 3
Tire Chip Size with Cost and Process Rate
Size
Description
Application
(in)
Cost Per Ton
Process Rate
($)
(Tons/Hour)
Clean cut. Ply & bead steel

remains
Cement Kilns, Civil Eng.

Applications
10
10 to 12
2 minus
Minimal wire, cut beads

removed by magnets
Industrial, utility, pulp paper

mill boilers
25
1 nominal
Same as 2" minus with extra

shredded pass
Power utility boilers
10 to 30
4 to 5
1/2 minus
Truly wire free, requires add.

shredding equipment
Feedstock in crumb rubber,

playground, fields
25 to 55
2 to 3
TABLE 4
Energy Values Input
Process
Energy Required (kW*hr/kg)
Manufacture a Tire
32
Produce a Rubber Compound
25
Incinerate a Tire
-9
Shredding Scrap Tires
1.2
calculating the profit that could be made for each method:

P = TF + R - PC - TR D
(3)
where P represents the profit, TF is the tipping fee, R is the

revenue generated from the final product, PC is the process
cost, TR is the cost of transporting the tires, and D is the cost
to dispose of the left over waste or ash.
The next factor to consider is the environmental impact. It
is necessary to look at potential pollutants and the damage
they may cause as well as take into account any risks to peoples health. For this research air pollutants will be exclusively explored because they exist in greater concentrations than
trace water pollutants and are easier to measure.
A sometimes hidden benefit is the amount of energy produced by each process. There are some of these methods that
require more energy for the process that will be given off.
Exothermic processes may be lead to a trade-off between
lower profitability and higher energy production. The cost of
the energy generated could help mitigate the cost of extraction although this will fluctuate with energy prices. Table 4
shows the energy required for the primary steps of tire recycling and will be one of the main contributors in the energy
generation calculations. One of the downsides to incinerating
scrap tires is that it takes 3 to 4 times the energy to produce
tire rubber than the amount of energy recovered by pyrolysis
or TDF. Ideally, the use of scrap tires for its originally in-
152
tended purpose or some other closely related use makes by

far more sense than incineration, both environmentally and
economically.
A practical consideration is the accessibility and availability of the solution. A big portion of this will come from the
distance from the tire source to the destination and what restrictions will be placed on each process. Many solutions are
best suited for a regional area. Different states and municipalities have their own laws so what looks feasible in one area
may not work as well in others.
Each solution will be evaluated on how it produces on
each factor and a ranking system will be developed. The solutions that perform the best in all the categories will be given
better scores and a cutoff score would eliminate the possible
solutions which were weak in one or more factors.
Only MCDA methods that considered the existence of a
hierarchy of criteria with multiple objectives and alternatives
were considered for this research. Thus, many other methods
that dealt with multi-objective planning particularly those that
used single objective optimization with constraints as other
goals, or use penalty functions, were omitted.

In order to evaluate the processes of scrap recycling there
are two tasks that need to be performed. First numerical val-
VOLUME 42, NO. 2
MAY 2016
ues for each of the criteria (profitability, energy generation,

and air pollutants emitted) must be calculated on all the processes. These values will then be used by the MCDA methods to evaluate their effectiveness and determine which process will be the best option.
CRITERIA CALCULATION
The profitability of each method is calculated using the
costs determined from the data from Table 1 and Table 2 as
well as the incorporation of equation (3). Some factors that
were not considered due to their wide variation include the
cost of the land, energy in both consumption and production,
and the initial capital investment. The profitability of each
option was calculated and is shown in Table 5.
Many of the processes require shredding the tires which

will have a cost that will reduce the total profitability. Retreading the tires and selling them on the open market involves minimal initial cost and is thus the most profitable but
may not be practical for many damaged tires. Meanwhile,
sending tires to the landfill requires a tipping fee and generates no additional income. The processes that require incineration are also unprofitable due to the costs of shredding the
tires with the energy costs of burning is greater than the value
of energy that would be generated in return.
The next criteria that needed to be calculated were net
energy generation. This calculation was heavily influenced by
the data in Table 4 as well as some values obtained from research sources. At this point only tire derived fuels and pyrolysis generate energy and many of the others are either essentially neutral or are large energy consumers. Tire stock-
TABLE 5
Profitability for Each Process
Process
Profit per Tire
Retread and Resell Tires
$8.00
Ground Rubber
$2.00
Civil Engineering
$1.56
Tire Derived Fuel (TDF)
$0.54
Tire Stockpile
$0.00
-$0.48
Pyrolysis
-$0.74
Landfill
-$6.00
TABLE 6
Net Energy Produced
Process
Energy Consumed
Energy Produced
Net Energy Production
(kW)
(kW)
(kW)
Tire Derived Fuel
86.2
560
473.8
Pyrolysis
129.2
245.6
116.4
Tire stockpiles
1.2
-1.2
Landfilling
1.2
-1.2
Crumb Rubber
1.2
-1.2
86.2
-86.2
Reuse and Retread Tires
174.9
-174.9
153
piles contain a great store of potential energy but there is no

constructive method of harnessing it.
The final criteria were the environmental output for each
process. The release of toxic substances in the air for three of
the most common air pollutants were determined and then
added together to develop a relative index of pollutants. This
was necessary to simplify the process and prevent thousands
of tradeoff determinations within these criteria. The pollutant
information is given in Table 7.
Once these calculations were completed there was now a
basis for each of the MCDA methods to be run. In this study
profitability was given the primary influence as that appears
to be the most motivating influence of the tire recycling market. Many companies will seek other endeavors if they start
to lose money. The second higher consideration is to keep the
amount of air particles emitted to be as low as possible. All of
the MCDA methods were run and the weights that were determined by their questions are given in Table 8.
The weights are then used by the DCMA methods to obtain a score which is used to evaluate which process should
be applied. The scores obtained from each process are displayed in Table 9.
From Table 9 it appears that the resold and retreaded tires
would be the most viable option. Retreading and reusing tires

as long as they are safe is the least costly option and generates no air pollutants. This also helps to reduce the demand
for new tires. However, the benefits from not replacing are
temporary. Eventually the tire will sustain enough wear or
damage that repair is no longer a feasible option. Additionally
only a small percentage of the tires are in good enough condition and it is not practical to sort them out at the recycling
facility. Crumb Rubber then becomes the next best option
according to three of the DCMA methods. Civil Engineering
Applications also scores very highly and would be considered
an acceptable alternative. The only less than ideal application
is using the scrap tires for coral reefs which is a very expensive option. There is the potential for the tires to leak out their
toxic chemicals and kill the wild life. Another point is that
very few people live near the coastal areas which would mean
spending an additional cost for the transport of the tires to the
location. However, there is not much demand for this so very
few tires are applied in this fashion. Landfilling was shown to
be a bad option by all of the DCMA methods. It is clear that
TDF is the best option for energy production, reusing tires is
the best option for profitability, and landfilling is not a good
fit for any of the criteria.
TABLE 7
Emitted Air Pollutants for Each Process
SO x
NO x
CO
Total
Pollutants
(ppm)
(ppm)
(ppm)
(ppm)
Crumb Rubber
Landfilling
76.7
124.4
259.5
460.6
Tire stockpile Fires
275
112
116
503
Pyrolysis
269.6
156
40
465.6
Tire Derived Fuel
630
222
30
882
Process
TABLE 8
Calculated Criteria Weights for Each MCDA Method
Criteria
154
MCDA Method Weights
Analytical
Make It Rational
D-Sight
1000 Minds
Profit Per Tire
.573
.677
.501
.634
Net Energy Generated
.146
.182
.217
.151
Air Pollutants Emitted
.281
.141
.282
.225
VOLUME 42, NO. 2
MAY 2016
TABLE 9
Summary of DMCA Software Results
Process
Analytical
Make It Rational
D-Sight Score
1000 Minds Score
Mean Score
Reuse and Retread Tire
77.01
67.68
72.16
89.5
76.59
Crumb Rubber
57.23
14.21
74.36
66.7
54.25
Civil Engineering
Applications
66.93
17.93
67.25
64.9
53.13
Pyrolysis
25.18
32.54
33.88
56.1
46.93
Tire Stockpiles
46.79
48.52
40.93
47.4
45.91
Tire Derived Fuel

(TDF)
26.66
34.98
50.18
43.9
41.43
35.50
31.74
29.54
35.1
32.97
Landfilling
24.51
19.62
31.70
18.6
23.61
CONCLUSION
Unwanted tires cause a great many problems in the field
of solid waste management. They are very bulky and take up
a lot of premium space in the landfills. Other animals like to
use them as homes where they can spread diseases. There are
several methods of disposing of these tires, but no universal
method that is considered the best. The biggest factor is the
local market. This study looked at each method on the basis
of cost/benefit analysis, environmental impact, and feasibility. It will be beneficial to all if the best processes such as tire
retreading, crumb rubber, and civil engineering applications
were applied regularly and the inefficient processes such as
landfilling were eliminated entirely.
ACKNOWLEDGEMENTS
This work was sponsored by the Department of Energy
Samuel Massie Chair of Excellence Program under Grant No.
DE-NA0000718. The views and conclusions contained herein
are those of the writers and should not be interpreted as necessarily representing the official policies or endorsements,
either expressed or implied, of the funding agency. A previous version of this research was presented at the 24th International Conference on Solid Waste Technology and Management at Philadelphia, PA on the dates of March 30 to April 2,
2014.
REFERENCES
Amari, T., N.J. Themelis, and I.K. Wernik. (2000). Resource
Recovery from Used Tires. Resources Policy. Volume
25, pp. 179-188.
Baker, T.E., T.M. Allen, D.V. Jenkins, and D.T. Mooney.

(2003). Evaluation of the Use of Scrap Tires in Transportation Related Applications in the State of Washington.
Washington State Department of Transportation. Report
to the Legislature SHB 2308.
Badri, M.A. (2001). A combined AHP-GP model for quality
control systems. Journal of Production Economics. Volume 72. pp. 27-40.
Belton, S. and T.S. Stewart. (2002). Multiple Criteria Decision Analysis: An Integrated
Approach. Kluwer Academic Publishers. Massachusetts.
Birkholz, D.A., K.L. Belton, and T.L. Guidotti., (2003).
Toxicological Evaluation for the Hazard Assessment of
Tire Crumb for Use in Public Playgrounds. Journal of
Air & Waste Management Association. Volume 53, pp.
903-907.
Chang, N. B. (2008). Economic and policy instrument analyses in support of the scrap tire recycling program in Taiwan.
Journal
of
Environmental
Management
Volume 86, pp. 435-450.
Chang, C.W., Wu, C.R., and Lin, H.L. (2009). Applying
fuzzy hierarchy multiple attributes to construct an expert
decision making process. Expert Systems with Applications Volume 36, pp. 7363-7368.
Dierks, W. (2011). Incorporating the Use of Recycled Rubber. Accessed at http://www.trb.org/Main/Blurbs/Uses_
of_Recycled_Rubber_Tires_in_Highways_154432.aspx
on April 23, 2011.
Douligeris, C. and Pereira, I. (1992). An analytic hierarchy
process approach to the analysis of quality in telecommunication system. IEEE Journal. Volume 48. pp. 16841688.
Everett J.W., and S. Douglah, ( 1998). Scrap tire disposal
(II): case study and recommendations. Journal of Solid
Waste Technology and Management. Volume 25, pp. 5160.
155
Fu, H.P., Ho, Y.C., Chen, R.C., Chang, T.H., and Chein, P.H.
(2006). Factors affecting the adoption of electronic marketplaces. Journal of Operation and Production Management. Volume 26. pp. 1301-1324.
Giray, E.S. and O. Sonmez,(2004). Supercritical extraction
of scrap tire with different solvents and effect of tire oil on
the supercritical extraction of coal. Fuel Processing
Technology, Volume 85, pp. 251-265.
Jagananthan, S., Erinjeri, J.J., and Ker, J.I. (2007). Fuzzy
analytic hierarchy process based group decision support
system to select and evaluate new manufacturing technologies. International Journal of Advanced Manufacturing
Technology. Volume 32. pp. 1253-1262.
Jang, J.W., T.S. Yoo, J.H. Oh, and I. Iwasaki, (1998). Discarded tire recycling practices in the United States, Japan
and Korea. Resources, Conservation and Recycling.
Kwong, C. and Bai, H. (2002). A fuzzy AHP approach to
the determination of importance weights of customer requirements in quality function deployment. Journal of
Intelligent Manufacturing. Volume 13. pp. 367-377.
Leung, L. and Cao, D. (2000). On consistency and ranking
of alternatives in fuzzy AHP. Journal of Operational Research. Volume 124. pp. 102-113.
Liberatore, M.J. and Nydick, R.L. (2008). The analytic hierarchy process in medical and health care decision making:
A literature review. European Journal of Operational Research. Volume 189. pp. 194-207.
Lucchesi, A. and G. Maschio, (1983). Semi active carbon
and aromatics produced by pyrolysis of scrap tires. Cons
Recycl. Volume 6. pp. 85-90.
Raman, K.P., W.P. Walawender and L.T. Fan, (1981)
Gasfication of waste tires in a fluid bed reactor. Cons
156
Recycl . Volume 4. pp. 79-88.

Reschner, K. (2008). Scrap Tire Recycling: A Summary of
Prevalent Disposal and Recycling Methods. Accessed at
http://entire-engineering.de/Scrap_Tire_Recycling.pdf on
April 10, 2011.
Rubber Manufacturers Association (RMA). (2003). Considerations for Starting a Scrap Tire Company: A Blueprint
for Planning a Business Strategy. RMA Special Report
GEN-060.
Salmeron, J.L. and Herrero, I. (2005). An AHP based methodology to rank critical success factors of executive information systems. Computer Standards and Interfaces
Scrap Tire News. (2011). Crumb Rubber Information. Accessed at http://www.scraptirenews.com/crumb.php on
March 25, 2011.
Saaty, T.L. (1980). The analytic hierarchy process. Mc-Graw
Hill, New York.
US EPA. (2011). Tire Derived Fuels. Accessed at
http://www.epa.gov/osw/conserve/materials/tires/tdf.htm
on March 14, 2011.
US EPA. (2015). State Scrap Tire Programs: A Quick Reference Guide. Accessed at http://epa.gov/osw/conserve/ materials/tires/basic.htm on June 24, 2015.
Vaidya, O.S. and Kumar, S. (2006). Analytic hierarchy process: An overview of applications. Journal of Operational Research. Volume 169. pp. 1-29.
von Winterfeldt D and W. Edwards. (1986). Decision analysis and behavioral research. Cambridge University Press.
Cambridge.
Wasil, E. and Golden, B. (2003). Celebrating 25 years of
AHP based decision making. Computers and Operation
Research. Volume 30. pp. 1419-1430.
VOLUME 42, NO. 2
MAY 2016
ASSESSMENT ON NON-LINEAR MODELS FOR DEMONSTRATING

HEAVY METALS RELEASE BEHAVIOR FROM SOLID WASTE BLOCK
M. Aminul Haque1, 2*, M. A. Hoque1
1
Department of Civil and Environmental Engineering

Shahjalal University of Science and Technology, Sylhet, Bangladesh
2
Department of Civil Engineering, Leading University, Sylhet, Bangladesh

mahaque@lus.ac.bd
ABSTRACT
To reduce the solid waste load and minimize the contaminant like heavy metals migration to
the surrounding environment at landfill sites, solid waste was incorporated in paving mortar
block by following the solidification/stabilization treatment technique. Two well established nonlinear methods such as diffusion equation derived for a plane source model and empirical
method employing a polynomial equation were used for the better understanding of the heavy
metals like Fe, Cu and Ni migration phenomena. Experimental data representing the releasing
heavy metal from paving block were used for non-linear models for calibration. The accuracy
of the models was statistically evaluated followed by models parameter estimation. The study
showed that polynomial equation is better than diffusion equation for explaining experimental
observation. Moreover, in this current study, polynomial equation was extended further three
different longer terms namely Model 1, Model 2 and Model 3 respectively using the least
squares procedure for examining the best fit profile with the observations. All the three
extended models were justified against statistical point of view. Calibration results shows that
the polynomial equation with 2.5 degree (Model 2) explains better leaching behavior of Fe and
Cu, whereas the Model 3 having third degree polynomial equation is found to be perfect for
representing Ni release pattern from the solid waste block. Apart from the calibration
approach, the accuracy of Model 2 and Model 3 were validated with respective experimental
observation followed by the model parameters estimation with 95% confidence interval.
Keywords: Solid waste block, Cumulative fraction leached, Model calibration, Parameter
estimation
INTRODUCTION
In conventional solid waste management system such as
landfilling, contaminants like heavy metals spread into the
soil, surface water and ground water table [1,2] that could be
deleterious [3,4] for the surrounding environment at landfill
sites. In order to reduce the scale of pollution of waste at
landfill sites, S/S techniques are used to prevent or minimize
the release of hazardous compounds from contaminated

waste into the environment by producing a solid mixture [1]
and sustainable end product [5] that are non-hazardous or less
hazardous than the original waste which formation prevent
pollutants from migrating into the environment [6,7] and safe
disposal to land [8] or used as building elements [9,10]. The
performance of S/S is generally assessed with leaching and
extraction tests [11,12] to measure the potentiality of
contaminants like heavy metals release to the environment
___________________________________
* Corresponding author
ASSESSMENT ON NON-LINEAR MODELS FOR DEMONSTRATING HEAVY METALS RELEASE BEHAVIOR
157
from stabilized waste [13].

Several researchers have used different non- linear models
for explaining the migration characteristics and longer-term
leachability prediction of heavy metals from solidified waste
matrix using leaching tests. Ficks second law based diffusion
model have been used to evaluate the Cesium (137Cs) and
60
Co release from radioactive materials [14], Sodium (Na+),
Potassium (K+), Chlorine (Cl-) etc. leaching from steel
processing industry waste [15], arsenic (As) leaching rate
from arsenic containing solid wastes [16] and Cesium (137Cs)
leaching from radioactive wastes [17]. Some previous
researchers investigated heavy metals migration pattern from
solidified waste formations during curing period employing
polynomial equation based model such as 60Co leaching rate
from radioactive waste-concrete composites [18], 137Cs and
60
Co releasing rate from radioactive materials [14], Copper
aluminum oxychloride (CAOX) leaching mechanism from
industrial waste-cementbentonite clay matrix [19] and 137Cs
transport phenomena from radioactive waste [17]. Plecas and
Dimovic (2009) [14] observed 137Cs and 60Co migration rate
from radioactive waste composites using the one dimensional
differential model. Rate equation for coupled diffusion and
simultaneous first-order reaction based model was also
adopted by some researchers [17,18,19] for observing the
heavy metal leaching mechanism from different waste
materials.
The quantities of heavy metals migrated from solid waste
matrix through modelling is need to be investigated for better
projection of the heavy metal release phenomena. In addition,
the models need to be evaluated for the identification of the
best fit with experimental observation as the models are

sensitive to the property of metals. The accuracy of the
models were confirmed through statistically evaluation
followed by validation and checking the 95% confidence
interval of the estimated model parameters.
MATERIALS AND EXPERIMENTAL DATA

COLLECTING METHODS
Preparation of solidified waste mortar
specimen
In mixing compositions, landfill solid waste was mixed
with stabilizing binder like OPC as the replacement of fine
aggregate having mix proportion 1:2 with maintaining some
characteristics for each waste mortar specimen which are
represented in Table 1. Cubical block samples were cured
with solid/ liquid ratio of 4 in terms of volume [15] under the
average humidity of 93% 2 % at 25010 C in the laboratory
room. PH of curing water was taken 7.0 constantly by using
distil water.
Leaching test of heavy metals

ANS16.1 leaching test was performed for 1, 2, 3, 4, 5, 6, 7,
14, 21, and 28 curing days period [19]. Fe and Cu were
tested by using flame emission atomic absorption
spectrophotometer (Spectra AA Varian) and Ni was also
TABLE 1
Physical characteristics of solidified waste mortar specimen
SL No.
158
Parameters
Values
01
Dimension
5cmx5cmx5cm
02
Mixing proportion of ingredients
03
Unit weight of specimen
04
Unit weight of cement
690.39 kg/m3
05
Unit weight of sand
976.66 kg/m3
06
Unit weight of waste
414.01 kg/m3
07
Unit weight of water
334.84 kg/m3
08
Weight per specimen
260 1 gm
09
Replacement of sand by waste per specimen
10
Quantity of cement per specimen
86.67 gm
11
Quantity of sand per specimen
121.33 gm
12
Quantity of waste per specimen
52 gm
13
Water to cement ratio per specimen
0.485
1:2
2080 5 kg/m3
30% of the sand volume
VOLUME 42, NO. 2
MAY 2016
measured using by Hach DR/4000 Spectrophotometer (Using

method: 8037). Fe, Cu and Ni leaching test results were
recorded as cumulative fraction leached (CFL) in terms of
cumulative released to the curing leachant in relation to the
total availability of metal mass in the samples [15].
by
Leaching models and theoretical aspects
f = A 0 + A 1 t1/2 + A 2 t + A 3 t3/2 + A 4 t2
was fitted to the leaching data.
In this study, two well established methods (Method I and

II) based on diffusion model and polynomial equation based
model used for demonstrating the leaching phenomena from
solid waste matrix to justify against the experimental
observations showing heavy metal migration behavior from
SWMB. The theoretical aspects of both nonlinear models are
presented below:
Method I: Diffusion equation based on a plane source model
The study revealed that migration of metals in curing
leachant from the S/S matrix is diffusion controlled. To
identify the metal leaching mechanisms, some researchers
[9,14,15,17,18,19,20,21,22,23,24,25,26,27,28] provided an
analytical solution for Fickian diffusion with some
simplifying assumptions such as, the mobility of a
contaminant is limited by diffusion, the specimen has a
uniform initial contaminant concentration and zero
concentration of contaminant at the solidliquid interface.
The Fickian diffusion equation is:
f=
=
(1)
Where f = Cumulative fraction leached of heavy metals. D e
can be calculated from the slope of a n / A 0 versus t n 1/2,
according to the following relationship:
(2)
Where,
From equations (1) and (2),
(3)
To generalize the model equation (3) can be expressed as,
f = A 0 + A 1 t1/2
(4)
Method II: Polynomial Equation

To assess the leaching behavior of heavy metals from
solidified waste, some authors [18,19] provided a semiempirical method to describe the long-term leaching
phenomena of a solidified waste matrix by a polynomial
equation based model. The orthogonal polynomial is one of
the most useful empirical equations [18,19]. Its general form
is:
y(x) =
(5)
Where, A i = Parameter to be determined and

= Function of x.
Here,
is taken as t i/2 and the leaching fraction is given
(6)
To simplify the mathematical treatment, a fifth terms
polynomial of the form is
(7)
Pleacas (2003) [17] adopted up to fourth terms of this model

to observe the leaching behavior of 137Cs from radioactive
waste formation,
f = A 0 + A 1 t1/2 + A 2 t + A 3 t3/2
(8)
some researchers [26,29] used up to third terms of the
polynomial equation based model for immobilization of
radionuclides 137Cs and 60Co in concrete matrix,
f = A 0 + A 1 t1/2 + A 2 t
(9)
Model evaluation
Non-linear models were calibrated with experimental data
generated from leaching (Fe, Cu and Ni) test followed by the
justification of the accuracy using statistical parameters such
as R2, SSE, MSE and RMSE. In general, the values of R2
ranges from 0 to 1, with higher values indicating less error
variance, however typically the values of R2 greater than 0.5
are considered acceptable [30, 31, 32, 33]. SSE, MSE and
RMSE values approaching towards 0 indicate the best fit of
the model profile with the observation [30].
Model calibration
In this study, method II (Polynomial equation based
model) was extended further from fifth to seven terms with
0.5 degree polynomial interval using the least squares
procedure for identifying the best fit profile that can
demonstrate the experimental observation in a better way.
Followings are the model equations extended for the
calibration process:
f = A 0 + A 1 t1/2 + A 2 t + A 3 t3/2 + A 4 t2
[Model 1]
f = A 0 + A 1 t1/2 + A 2 t + A 3 t3/2 + A 4 t2 + A 5 t5/2
[Model 2]
f = A 0 + A 1 t1/2 + A 2 t + A 3 t3/2 + A 4 t2 + A 5 t5/2 + A 6 t3
[Model 3]
(10)
(11)
(12)
The estimation of model parameters were done by following

the non-linear techniques.
RESULTS AND DISCUSSIONS

Experimental results of CFL from SWMB
Heavy metal leaching test results were measured in terms
of cumulative release from waste block to the curing solution.
Here, cumulative release means the total amount of Fe, Cu
and Ni releases from solidified waste mortar block to curing
159
solution up to the each tested curing day like 1, 2, 3, 4, 5, 6,

7, 14, 21, and 28 day without any interruption or replacement
anything. Concentration of Fe, Cu and Ni in curing solution
were measured as mg/l. But they are considered as solid state.
So their releasing amounts were converted into mg/kg. Initial
amount of Fe, Cu and Ni in waste sample means the total
content of entrapped metals presented within the solid waste
block before keeping in curing solution. Metal contents from

waste block decreased with the contact of curing time at a
diminishing rate and migrated to the curing solution. Tables
2, 3 and 4 represents the cumulative fraction migrated of Fe,
Cu and Ni respectively from waste mortar block during the
measured curing period.
TABLE 2
Experimental CFL values corresponding to Fe
Curing time
Cumulative release
of Fe at each curing
day (mg/l)
Cumulative release of
Fe at each curing day
(mg/kg)
Cumulative amount of Fe
release at each curing day
(a n ) (mg)
Cumulative fraction
leached (CFL) at each
curing day, (f Fe )
0.483
9.283819
0.482999
0.00082643
0.615
11.821012
0.614999
0.00105229
0.706
13.570138
0.705999
0.00120799
0.770
14.800292
0.769999
0.00131750
0.838
16.107330
0.837999
0.00143385
(day), t
Initial amount Fe in
waste sample(A o )
(mg)
584.436267
6
0.882
16.953061
0.881999
0.00150914
0.921
17.702687
0.920999
0.00157587
14
1.084
20.835735
1.083999
0.00185477
21
1.170
22.488755
1.169999
0.00200192
28
1.203
23.123053
1.202999
0.00205839
TABLE 3
Experimental CFL values corresponding to Cu
Curing time
Cu at each curing day
(mg/l)
Cu at each curing day
(mg/kg)
Cumulative amount of
Cu release at each
curing period (a n )
(mg)
Cumulative fraction
curing period, (f Cu )
0.039
0.749625
0.038999
0.00287563
0.049
0.941836
0.048999
0.00361299
0.064
1.230154
0.063999
0.00471903
0.078
1.499250
0.077999
0.00575134
0.089
1.710683
0.088999
0.00656243
(day), t
Initial amount of Cu in
waste sample
(A o )(mg)
13.561877
160
0.099
1.902894
0.098999
0.00729980
0.108
2.075885
0.107999
0.00796342
14
0.132
2.542401
0.132271
0.00975321
21
0.141
2.722523
0.141642
0.01044418
28
0.142
2.734209
0.142250
0.01048899
VOLUME 42, NO. 2
MAY 2016
TABLE 4
Experimental CFL values corresponding to Ni
Curing time
Ni at each curing day
(mg/l)
Ni at each curing day
(mg/kg)
Cumulative amount of
Ni release at each
curing period (a n )
(mg)
Cumulative fraction
curing period, (f Ni )
0.166
3.190712
0.165999
0.00230167
0.219
4.209433
0.218999
0.00303654
0.268
5.151270
0.267999
0.00371596
0.319
6.131549
0.318999
0.00442310
0.366
7.034944
0.365999
0.00507478
(day), t
Initial amount of Ni in
waste sample
(A o )(mg)
72.121042
6
0.390
7.496251
0.389999
0.00540756
0.416
7.996001
0.415999
0.00576806
14
0.488
9.392130
0.488635
0.00677521
21
0.501
9.613212
0.500137
0.00693470
28
0.504
9.682908
0.503763
0.00698497
CFL model calibration and parameter

estimation
The study observed that migration pattern of Fe, Cu and
Ni from the solid waste matrix shows as a function of the
square root of the contact time of leaching. In the whole
tested leaching period i.e. up to 28 days, linearity of metals
releasing behavior was not observed. The study also revealed
that linearity with rapid releasing rate occurs from 1 to 7 days
and 14 to 28 days releasing rate decreases with non-linearity
of leaching contact period. Even the slope of the tested metals
leaching behavioral path at early period is higher than that for
the latter stage. This phenomena associated with the matter of
diffusion. Generally, diffusion rate gradually slow down with
the leaching time elapsed. For this reason, diffusion rate is

not a controlling factor in any of the leaching intervals, if the
diffusion path becomes longer. Figures 1, 2, and 3 represents
the CFL of Fe, Cu and Ni versus curing days respectively
with the two fitted CFL models like diffusion controlled
based model (method I) and polynomial equation based
model (Method II) to the experimental observations. CFL
models were compared with experimental observed data for
test metals like Fe, Cu and Ni to get better understanding the
leaching mechanism based on the curing period. Because of
the simplicity, Method I cannot explain the whole leaching
path i. e. up to 28 days period properly. However, the method
II described the quite good heavy metals migration
phenomena throughout the time tested whereas method I
shows unrealistic leaching pattern.
FIGURE 1
Leaching behavior of Fe with curing time
161
FIGURE 2
Leaching behavior of Cu with curing time
FIGURE 3
Leaching behavior of Ni with curing time
The parameters related to both models for three tested

metals are presented in Table 5. Model parameter estimation
was done following non-linear techniques. Both model
parameters were estimated along with their confidence
interval.
In order to evaluate both models statistically, SSE, MSE
and RMSE values are shown in Table 6. Based on the

comparison criteria for tested metals, it was observed that R2
value was found greater in method II than method I. On the
contrary, the value of SSE, MSE and RMSE were also lower
in method II than method I. So, the study ranked the method
II as better fitted model to describe whole the cumulative
TABLE 5
Estimated model parameters corresponding to Method I and Method II
Parameter
162
Method I (Diffusion equation based model)
Method II (Polynomial equation based model)
Fe
Cu
Ni
Fe
Cu
Ni
A0
7.26469 x 10-4
1.61897 x 10-3
2.19506 x 10-3
1.51962 x10-4
4.33445 x10-3
2.36038x10-3
A1
2.79532 x 10-4
1.95796 x 10-3
1.02501 x 10-3
7.63789 x 10-4
5.80512x 10-3
2.45471x 10-3
A2
-----
-----
-----
9.37377x10-5
5.48206x10-3
3.11289x10-3
A3
-----
-----
-----
2.80517 10-6
1.31871x10-3
8.15089x10-4
A4
-----
-----
-----
9.62576x10-8
1.00514x10-4
6.53543x10-5
VOLUME 42, NO. 2
MAY 2016
TABLE 6
Statistical evaluation of model profiles corresponding Method I and Method II
Parameter
Fe
Cu
Ni
Method I
Method II
Method I
Method II
Method I
Method II
0.9353
0.9997
0.9410
0.9925
0.8309
0.9984
SSE
9.50x10 -8
3.49 x10 -10
4.22 x10 -6
5.05 x10 -7
3.76 x10 -6
3.74 x10 -8
MSE
1.18 x10 -8
6.98 x10 -11
5.27 x10 -7
1.01 x10 -7
4.70 x10 -7
7.49 x10 -9
RMSE
1.08 x10 -4
8.36 x10 -6
7.26 x10 -4
3.18 x10 -4
6.85 x10 -4
8.65 x10 -5
R2
fraction leached behavior for Fe, Cu and Ni as a function of

leaching time from solidified waste samples.
CFL model extension

Calibration of CFL model extension graphs are shown in
the Figures 4, 5 and 6, which represents the well fitted model
profile with the experimental observation data for Fe, Cu and
Ni respectively. In Figure 4, although three extended models
for Fe concentrations shows almost similar leaching pattern

with the experimental observations, based on the statistical
analysis Model 1 described better leaching phenomena,
whereas the Model 2 and Model 3 were found to be well
calibrated with the Cu (Figure 5) and Ni (Figure 6) release
data from solid waste matrix.
Based on the value of R2, SSE and RMSE, extended
calibrated leachate model options of Fe, Cu and Ni were
evaluated for the best fitness to their experimental
CFL of Fe (fFe)
0.0025
0.002
Observation data
Model 1
Model 2
Model 3
0.0015
0.001
0.0005
0
0
10
15
20
Curing days, t(d)
25
30
FIGURE 4
Calibration results corresponding to the migration of Fe
FIGURE 5
Calibration results corresponding to the migration of Cu
163
CFL of Ni (fNi)
0.008
0.007
0.006
0.005
0.004
0.003
0.002
0.001
0
Observation data
Model 1
Model 2
Model 3
10
15
20
25
Curing days, t(d)
30
FIGURE 6
Calibration results corresponding to the migration of Ni
observation data, which are shown in Tables 7, 8, and 9

respectively. Statistical analyzing values of three parameters,
the study calculated that Model 2 showed highest R2 value
and lowest SSE, MSE & RMSE values than Model 1 and
Model 3 for Fe (Table 7) and Cu (Table 8) respectively.
Similarly, highest R2 value and lowest SSE, MSE & RMSE
values were observed for Model 3 the than other Model 1 and
Model 2 for Ni (Table 9) indicating the best model option for
experimental data interpretation.
The parameter estimation outcomes of well fitted model
equations 2 for Fe and Cu are presented in the Table 10 and
11 respectively, whereas the Table 12 represents the
parameter estimation outcomes of best fitted model 3 with
their 95% confidence interval of lower and higher bound.
Validation of the extended CFL model

Separate sets of leachate data obtained from the laboratory
experiments were used for the validation of proposed
extended CFL model. Figures 7 and 8 represents the
comparison between the model profile (Model 2) and
experimental observation corresponding to Fe and Cu
respectively, whereas the validation of Model 3 with
experimental CFL values of Ni is shown in Figure 9. The
accuracy of the models with respective experimental
objectives was justified through statistical analysis (i. e. R2,
SSE and RMSE).
The R2 value of proposed model for Fe, Cu and Ni with
TABLE 7
Statistical evaluation of extended polynomial models based on Fe release data
Parameter
Model 1
Model 2
Model 3
0.999762239
0.999772536
0.999779152
SSE
3.49 x10 -10
3.34E-10 x10 -8
3.44E-10 x10 -8
MSE
8.98E-11 x10 -11
6.35E-11 x10 -8
1.08E-10 x10 -8
RMSE
9.36E-06 x10 -6
8.14E-06 x10 -8
1.04E-05 x10 -8
R2
TABLE 8
Statistical evaluation of extended polynomial models based on Cu release data
Parameter
Model 1
Model 2
Model 3
0.976990931
0.997386508
0.996820593
SSE
1.64E-06 x10 -8
1.77E-07 x10 -8
2.16E-07 x10 -8
MSE
3.29E-07 x10 -8
4.44E-08 x10 -8
7.21E-08 x10 -8
RMSE
5.74 E-04 x10 -8
2.10E-04 x10 -8
2.68E-04 x10 -8
R2
164
VOLUME 42, NO. 2
MAY 2016
TABLE 9
Statistical evaluation of extended polynomial models based on Ni release data
Parameter
Model 1
Model 2
Model 3
0.989715227
0.995557907
0.997498333
SSE
2.28E-07 x10 -8
9.88E-08 x10 -8
6.17E-08 x10 -8
MSE
4.57E-08 x10 -8
2.47E-08 x10 -8
2.05E-08 x10 -8
RMSE
2.13E-04 x10 -8
1.57E-04 x10 -8
1.43E-04 x10 -8
R2
TABLE 10
Estimated Parameter derived from Model 2 for Fe with 95% confidence interval
Parameter
Estimate
95% Confidence Interval

Lower Bound
Upper Bound
A0
3.8300 x10 -4
-6.957 x10 -5
8.3557 x10 -4
A1
2.4742 x10 -4
-7.2466 x10 -4
1.2195 x10 -3
A2
3.2530 x10 -4
-4.4093 x10 -4
1.0915 x10 -3
A3
-1.534 x10 -4
-4.3399 x10 -4
1.2706 x10 -4
A4
2.714 x10 -5
-2.097 x10 -5
7.5250 x10 -5
A5
-1.760 x10 -6
-4.870 x10 -6
1.360 x10 -6
TABLE 11
Estimated Parameter derived from Model 2 for Cu with 95% confidence interval
Parameter
Estimate

Lower Bound
Upper Bound
A0
1.1579 x10 -2
8.94428 x10 -3
1.4214 x10 -2
A1
-2.1997x10 -2
-2.76574 x10 -2
-1.633 x10 -2
A2
18.622 x10 -2
1.41608 x10 -2
2.3083x10 -2
A3
-6.2190 x10 -3
-7.85235 x10 -3
-4.585 x10 -3
A4
9.4866 x10 -4
6.6854 x10 -4
1.2287 x10 -3
A5
-5.5070 x10 -5
-7.3210 x10 -5
-3.692 x10 -5
their experimental leachate data were found to be close to 1

(Figures 7, 8 and 9), indicating the validity of the models.
Besides, calculated SSE and RMSE values were found to
reasonable for Fe, Cu and Ni (Figures 7, 8 and 9).
CONCLUSIONS
Two different non-linear methods such as diffusion
equation based on plane source model and polynomial

equation based model were evaluated along with their
parameter estimation on the basis of Fe, Cu and Ni leaching
mechanism from solidified waste mortar block. The analysis
of the results showed that polynomial equation is better than
diffusion approach for explaining the metal release
mechanism from solidified matrix. The polynomial equation
was extended further for three different longer terms in order
to achieve better calibration results using the least squares
165
TABLE 12
Estimated Parameter derived from Model 2 for Ni with 95% confidence interval
Parameter
Estimate

Lower Bound
Upper Bound
A0
5.3273 x10 -3
-2.30906 x10 -3
2.2963 x10 -2
A1
-8.9665x10 -3
-5.59481 x10 -2
3.8015 x10 -2
A2
8.2393 x10 -3
-4.00892 x10 -2
5.6567 x10 -2
A3
-2.6241 x10 -3
-2.73519 x10 -2
2.2103 x10 -2
A4
3.4194 x10 -4
-6.33784 x10 -3
7.0217 x10 -3
A5
-1.073 x10 -5
-9.2041 x10 -4
8.9895 x10 -4
A6
-7.4 x10 -7
-4.987 x10 -5
4.838 x10 -5
FIGURE 7
Validation of proposed CFL model with Fe release concentration
FIGURE 8
Validation of proposed CFL model with Cu release concentration
procedure. This study shows that Fe and Cu migration
characteristics follow the polynomial equation having 2.5
degree, whereas 3 degree polynomial equation is found to be
well-calibrated with experimental data corresponding to the
release of Ni from solid waste block. The proposed leachate
models were justified against statistical parameters along
with the validation of the model profiles with experimental
166
observation and found to be satisfactory for demonstrating

the respective heavy metals release phenomena.
ACKNOWLEDGEMENT
The authors would like to thank to the authority of Dhaka
VOLUME 42, NO. 2
MAY 2016
CFL of Ni (f Fe)
0.008
R2 = 0.94697
SSE = 1.51x10 -6
0.006
Constant parameter:
- Temperature
- pH of curing water
- W/C ratio
Experimental
observation data
Model profile
0.004
0.002
0
0
15
20
10
Curing days, t(d)
25
30
FIGURE 9
Validation of proposed CFL model with Ni release concentration
South and North City Corporations for the research grant to

collect solid waste sample from Matuail and Aminbazar
landfill sites.
NOMENCLATURE
f = Cumulative fraction leached of contaminant for each
leaching period
a 0 = Initial amount of contaminant in sample (mg)
V = Volume of sample (cm3)
S = Exposed surface area of sample (cm2)
D e = Diffusion co-efficient (cm2 day-1)
m = Slope of straight line (day-1)
t = Duration of leachant renewal period (day)
A 0 , A 1 , A 2 ,.. , A n = Parameter of model equation
f Fe = Cumulative fraction leached of Fe
f Cu = Cumulative fraction leached of Cu
a n = Cumulative leached of contaminant for each leaching
period (mg)
f Ni = Cumulative fraction leached of Ni
SWMB = Solidified waste mortar block
CFL = Cumulative fraction leached
OPC = Ordinary Portland cement
ANS = American Nuclear Society
R2 = Coefficient of determination
RMSE = Root mean square error
SSE = Sum of square errors
MSE = Mean square error
S/S = Solidification and stabilization
REFERENCES
[1] F.R.M. Omar, M.H. Isa, S.R.M. Kutty, A. Malakahmad,
M.N. Adlan, Solidification and stabilization of waste

Activated sludge from petroleum refinery. International
Conference on Environment, 2008.
[2] USEPA, Office of water, Citizens Guide To GroundWater Protection. EPA 440/6-90-004, 1990, 5p.
[3] G. Bhalla, P.K. Swamee, A. Kumar, A. Bansal,
Assessment of groundwater quality near municipal solid
waste landfill by an Aggregate Index Method.
International Journal of Environmental Sciences,

Volume 2 (2012), pp. 14921503.
[4] E.O. Longe, M.R. Balogun, Groundwater quality
assessment near a municipal landfill Lagos, Nigeria,
Department of Civil and Environmental Engineering,
University of Lagos, Nigeria. Research Journal of
Applied Sciences, Engineering and Technology, Volume
2(2010) pp. 3944.
[5] G.E. Voglar, D. Lestan, Efficiency modeling of
solidification/stabilization of multi-metal contaminated
industrial soil using cement and additives, Journal of
Hazardous Materials, Volume 192(2011) pp. 753762.
[6] D. Lee, D. Chung, J. Hwang, H. Choi, Fixation and of
Lead contaminants in Pb-doped solidified waste forms.
Environmental engineering research, Korean Society of
Environmental Engineers, Volume 12(2007) pp. 101
108.
[7] J.R. Conner, Chemical Fixation and Solidification of
Hazardous Wastes, Van Nostrand Rein hold, New York,
1990.
[8] D.M. Montgomery, C.J. Sollars, R. Perry, Cement based
solidification for the safe disposal of heavy metal
contaminated sewage sludge. Waste Management &
Research, Volume 6(1998) pp. 217226.
[9] C. Bobirica, L. Bobirica, R. Stanescu, I. Constantinescu,
Leaching behavior of cement-based solidified wastes
containing hexavalent chromium. Scientific BulletinUniversity Politehnica of Bucharest, Volume 72(2010)
121128.
[10] C. Shi, R. Spence, Designing of cement-based formula
for solidification/stabilization of hazardous, radioactive,
and mixed waste. Critical Reviews in Environmental
Science and Technology, Volume 34(2004) pp. 391417.
[11] A.K. Karamalidis, E.A. Voudrias, Leaching behavior of
metals released from cement-stabilized/solidified
Refinery oily sludge by means of sequential toxicity
characteristics leaching procedure, Journal of
Environmental Engineering, Volume 134(2008) pp. 493
504.
[12] H. A. Van der Sloot, Developments in evaluating
environmental impact from utilization of bulk inert
wastes using laboratory leaching tests and field
verification, Waste Management, Volume 16(1996) pp.
167
6581.
[24] A.M.
[13] USEPA, Test methods for evaluating solid waste,
physical/chemical methods. Atomic absorption methods

7000A, EPA publication No. SW-846, Washington, D.
C., 1992a.
[14] I. Plecas, S. Dimovic, Mathematical Modelling of
Immobilization of Radionuclides 137Cs and 60Co in
Concrete Matrix. The Open Waste Management
Journal, Volume 2(2009) pp. 4346.
[15] R. Malviya, R. Chaudhary, Evaluation of leaching
characteristics and environmental compatibility of
solidied/stabilized industrial waste. Journal of
Material Cycles and Waste Management, Volume
8(2006) pp. 7887.
[16] T.S. Singh, K.K. Pant, Solidification/stabilization of
arsenic containing solid wastes using Portland cement,
fly ash and polymeric materials, Journal of Hazardous
Materials, Volume 131(2006) pp. 2936.
[17] I.B. Plecas, Comparison of mathematical interpretation
in radioactive waste leaching studies. Journal of
Radioanalytical and Nuclear Chemistry, Volume
258(2003) pp. 435437.
[18] I. Plecas, S. Dimovic, Mathematical modeling of
transport phenomena in concrete matrix. Scientific
Journal of Physics, Chemistry and Technology, Volume
9(2011) pp. 2127.
[19] I. Plecas, S. Dimovic, Immobilization of industrial
waste in cement-bentonite clay matrix. Bulletin of
Material Science, Volume 27(2004) pp. 175178.
[20] H.W. Godbee, D.S. Joy, Assessment of the loss of
radioactive isotopes from waste solids to the
environment. Part I: Background and theory. Chemical
Technology Division, U. S. Atomic Energy Commission,
1974.
[21] K.Y. Cheng, P.L. Bishop, Developing a kinetic leaching
model for solidified/stabilized hazardous wastes.
Journal of Hazardous Materials, Volume 24(1990) pp.
213224.
[22] B. Batchelor, Leach models: theory and application.
Journal of Hazardous Materials, Volume 24(1990) pp.
255266.
[23] B. Batchelor, Leach Models for Contaminants
Immobilized by pH-Dependent Mechanisms, Environ.
Sci. Technol., Volume 32(1998) pp. 17211726.
168
El-Kamash, A.M. El-Dakaroury, H.F. Aly,

Leaching kinetics of 137Cs and 60 Co radionuclides fixed
in cement and cement-based materials, Cement and
Concrete Research, Volume 32(2002) pp. 17971803.
[25] I. Plecas, R. Pavlovic, S. Pavlovic, Leaching behavior
of 60Co and137Cs from spent ion exchange resins in
cementbentonite clay matrix, Journal of Nuclear
Materials, Volume 327(2004) pp. 171174.
137
60
[26] I. Plecas, S. Dimovic, Immobilization of Cs and Co
in concrete matrix. Part 2: Mathematical modeling of
transport phenomena, Annals of Nuclear Energy, Volume
32(2005) pp. 15091515.
[27] R.O.A. Rahman, A.A. Zaki, A.M. El-Kamash,
Modeling the long-term leaching behavior of 137Cs,
60
Co, and 152,154Eu radionuclides from cementclay
matrices, Journal of Hazardous Materials, Volume
145(2007) pp. 372380.
[28] I. Plecas, S. Dimovic, Curing Time on the Fraction of
137
Cs from Cement-Ion Exchange Resins-Bentinite Clay
Composition. Scientific Journal of Physics, Chemistry
and Technology, Volume 8(2010) pp. 3943.
[29] I.B. Plecas, Mathematical modelling of immobilization
of radionuclides137Cs and 60Co in concrete matrix,
Progress in Nuclear Energy, Volume 52(2010) pp. 685
688.
[30] D.N. Moriasi, J.G. Arnold, M.W.V. Liew, R.L. Bingner,
R.D. Harmel, T.L. Veith, Model evaluation guidelines
for systematic quantification of accuracy in watershed
simulations. American Society of Agricultural and
Biological Engineers, Volume 50(2007) pp. 885900.
[31] M.N. Islam, An Introduction to Statistics and
Probability, 3rd edition (2004), Book World, Simple
regression and correlation, pp. 253256.
[32] C. Santhi, J.G. Arnold, J.R. Williams, W.A. Dugas, R.
Srinivasan, L.M. Hauck, Validation of the SWAT model
on a large river basin with point and nonpoint sources.
J. American Water Resources Assoc., Volume 37(2001)
pp. 1169-1188.
[33] A.W. Alansi, M.S.M. Amin, G. Abdul, H.Z.M. Shafri, W.
Aimrun, Validation of SWAT model for stream ow
simulation and forecasting in Upper Bernam humid
tropical river basin, Malaysia. Hydrology and Earth
System Sciences Discussions, Volume 6(2009) pp. 7581
7609.
VOLUME 42, NO. 2
MAY 2016

The Journal of Solid Waste Technology and Management

Încărcat de

Informații document

Titlu original

Drepturi de autor

Formate disponibile

Partajați acest document

Partajați sau inserați document

Opțiuni de partajare

Vi se pare util acest document?

Este necorespunzător acest conținut?

Drepturi de autor:

Formate disponibile

The Journal of Solid Waste Technology and Management

Încărcat de

Drepturi de autor:

Formate disponibile

THE JOURNAL OF

SOLID WASTE TECHNOLOGY

THE JOURNAL OF SOLID WASTE TECHNOLOGY AND MANAGEMENT

SENIOR ASSOCIATE EDITOR:

Ilona Srvri Horvth

HYDROGEOCHEMISTRY AND STABLE ISOTOPES OF A SOLID WASTE DISPOSAL AREA

Romina Sanci, Hctor Osvaldo Panarello, Fernanda Cravero

WASTE-TO-ENERGY CHARACTERIZATION AND MODELING: A CASE STUDY AT THE US

P.A. Caton, J. Schmidt, C. Adsit, E. Bermudez, C. Chase, T. Kerner

COMPOSTING OF RUMEN CONTENT WASTE USING ANAEROBIC-ANOXIC-OXIC (A2/O)

Rhenny Ratnawati, Yulinah Trihadiningrum, Sri Rachmania Juliastuti

OPPORTUNITY OF BIOCHEMICAL PROCESS FOR PHOSPHOGYPSUM UTILIZATION

EVALUATION AND QUANTIFICATION OF HOUSEHOLD WASTES SENT FOR RECYCLING

Vanessa A. Mantovani, Sandro D. Mancini, Thalita R. Barros, Jos L. Ferraz

IMPERIAL SMELTING FURNACE SLAG AS FINE AGGREGATE IN CEMENT CONCRETE

USE OF RECYCLED AGGREGATE CONCRETE IN PAVEMENT CONSTRUCTION

Roz-Ud-Din Nassar, Parviz Soroushian

A MULTI-CRITERIA EVALUATION OF THE METHODS FOR RECYCLING SCRAP TIRES

Shoou-Yuh Chang, Frank Gronwald

ASSESSMENT ON NON-LINEAR MODELS FOR DEMONSTRATING HEAVY METALS

M. Aminul Haque, M. A. Hoque

HYDROGEOCHEMISTRY AND STABLE ISOTOPES OF A SOLID

and biological processes that act upon the waste. Leachate

groundwater quality of the area and surrounding areas due to

processes between leachate and groundwater that could affect

JOURNAL OF SOLID WASTE TECHNOLOGY AND MANAGEMENT

VOLUME 42, NO. 2

mineral horizon obscured by organic matter, Bt o Bk to

(AgNO 3 ). Dissolved organic carbon (DOC) was determined

final water due to their geographical location (downstream)

JOURNAL OF SOLID WASTE TECHNOLOGY AND MANAGEMENT

VOLUME 42, NO. 2

P samples are neutral. On the other hand, data represented on

not found in those piezometers and Fe was only found in P10,

ages found in the models reflect the influence of leachate on

JOURNAL OF SOLID WASTE TECHNOLOGY AND MANAGEMENT

VOLUME 42, NO. 2

Mixing process (%)

Mixing process (%)

Initial water LX1

constant 13C-DOC values between 25.9 and -23.9 and

municipal site was minimal, with water depths that ranged

2 m) would correspond to the unsaturated zone as well as part

higher concentrations of Al, Cu, Fe, Mn and Zn, SP12 of Cu,

Deep level (cm)

JOURNAL OF SOLID WASTE TECHNOLOGY AND MANAGEMENT

VOLUME 42, NO. 2

surface waters sampled during March, 2010 (Boujon, 2014)

JOURNAL OF SOLID WASTE TECHNOLOGY AND MANAGEMENT

VOLUME 42, NO. 2

could be the result of the above-mentioned mixing, but this is

at P and TZ1 piezometers are higher than at M9w, especially

When there are acid inputs, pH decreases but, since

JOURNAL OF SOLID WASTE TECHNOLOGY AND MANAGEMENT

VOLUME 42, NO. 2

leachate and are thought to be related to production of

gradable organic compounds as volatile fatty acids. But in the

y 2H values suggest not only a meteoric origin for the

JOURNAL OF SOLID WASTE TECHNOLOGY AND MANAGEMENT

VOLUME 42, NO. 2

Soil Survey Staff (2014) Keys to Soil Taxonomy, 12th ed.

C in CO 2 soil gas from a landfill: first results from a

JOURNAL OF SOLID WASTE TECHNOLOGY AND MANAGEMENT