Sunteți pe pagina 1din 13

Int J Life Cycle Assess (2017) 22:1802–1814

DOI 10.1007/s11367-016-1232-4

PROMOTING SUSTAINABILITY IN EMERGING ECONOMIES VIA LIFE CYCLE THINKING

Life cycle greenhouse gas emissions of palm oil production


by wet and dry extraction processes in Thailand
Anyanee Bunchai 1 & Oramas Suttinun 1,2 & Aran H-Kittikun 3 &
Charongpun Musikavong 2,4

Received: 30 January 2016 / Accepted: 14 November 2016 / Published online: 23 November 2016
# Springer-Verlag Berlin Heidelberg 2016

Abstract Allocations by mass, economic, and heating values were


Purpose The crude palm oil (CPO) extraction is normally conducted. The trading of GHG emissions from co-
done by a wet extraction process, and wastewater treatment products to GHG emissions from animal feed was assessed.
of the wet process emits high levels of greenhouse gases The GHG emissions or savings from direct LUC (dLUC)
(GHGs). A dry process extracts mixed palm oil (MPO) from and from indirect LUC (iLUC) effects and for the change
palm fruit without using water and has no GHG emissions from wet to dry process were determined.
from wastewater treatment. This work is aimed at determining Results and discussion Palm fruit and firewood were the
the GHG emissions of a dry process and at evaluating GHG major GHG emission sources. Nitrogen recycling on
savings on changing from wet to dry process, including land plantations from fronds significantly affects the GHG
use change (LUC) effects. emissions. With the carbon stocks, the GHG emissions
Methods Life cycle assessment from cradle to gate was allocated by energy value were 550 kg CO2 eq/t MPO.
used. The raw material is palm fruits. The dry process The GHG emissions were affected by −3 to 37% for the
includes primary production, oil room, and utilities. MPO change from wet to dry process. When the plantation
is the main product, while palm cake and fine palm residue area was increased by 1 ha and the palm oil extraction
are co-products sold for animal feed. Case studies were was changed from wet to dry process, and the change
undertaken without and with carbon stocks of firewood included dLUC and iLUC, the GHG savings ranged from
and of nitrogen recycling at plantations from fronds. −0.94 to 5.08 t CO2 eq/ha year. The iLUC was the main
GHG emission source. The GHG saving mostly originat-
ed from the change of extraction process and from the
dLUC effect. Based on the potential use of biodiesel
Responsible editor: Thapat Silalertruksa
production from oil palm, during 2015–2036 in
Electronic supplementary material The online version of this article Thailand, when the extraction process was changed and
(doi:10.1007/s11367-016-1232-4) contains supplementary material, dLUC and iLUC effects were included, the saving in
which is available to authorized users.
GHG emissions was estimated to range from −35,454
to 274,774 t CO2 eq/year.
* Charongpun Musikavong
mcharongpun@eng.psu.ac.th; charongpun@gmail.com Conclusions The change of palm oil extraction process and
the LUC effects could minimize the GHG emissions from
1
Faculty of Environmental Management, Prince of Songkla the palm oil industry. This advantage encourages developing
University, Hatyai 90112, Thailand policies that support the dry extraction process and contrib-
2
Center of Excellence on Hazardous Substance Management (HSM), ute to sustainable developments in palm oil production.
Bangkok 10330, Thailand
3
Department of Industrial Biotechnology, Faculty of Agro-Industry,
Prince of Songkla University, Hatyai 90112, Thailand Keywords Carbon stocks . Direct land use change . Dry
4
Department of Civil Engineering, Faculty of Engineering, Prince of extraction process . Fine palm residue . Indirect land use
Songkla University, Hatyai 90112, Thailand change . Mixed palm oil . Palm oil mill
Int J Life Cycle Assess (2017) 22:1802–1814 1803

1 Introduction does not use water. There is no wastewater and no GHG emis-
sions resulting from wastewater treatment. Small communities
The palm oil industry is an important economic sector in can establish a dry extraction mill at a low investment cost.
Thailand. The total palm plantation area in Thailand in the Thus, the dry extraction process of palm oil could be an alter-
year 2013 was 717,458 ha, and the total production quantity native to mitigate the associated environmental problems and
of fresh fruit bunches (FFBs) was 12.4 million t (OAE 2015). could provide employment to villagers in rural areas.
The life cycle of palm oil production consists of the oil palm In developing a sustainable palm oil industry, there are
plantation, the transportation of the FFBs and palm fruits, the serious concerns about the environmental impacts. This is
extraction of crude palm oil (CPO) or mixed palm oil (MPO), especially so with global warming that is caused by increased
the management of their co-products, the refining of palm oil, levels of GHGs. Previous studies of the GHG emissions from
and the biodiesel production. The extraction of CPO from palm oil industry have emphasized the wet extraction process
FFBs in Thailand is done by a wet extraction process, but (Reijnders and Huijbregts 2008; H-Kittikun et al. 2009;
MPO is extracted from the palm fruit by a dry process. Chuchuoy et al. 2009; Vijaya et al. 2010; Kaewmai et al.
The ever-increasing demand of fossil fuels and the con- 2012; Choo et al. 2011; Kaewmai et al. 2013; Stichnothe
cerns about their sustainability have led to the development et al. 2014). No prior study has addressed the GHG emissions
of alternative sources of energy. Biodiesel is a successfully by palm oil mills using the dry extraction process.
commercialized alternative to mitigate these problems and
concerns. Biodiesel can substitute for or be mixed in with
petroleum diesel. The life cycle of a biodiesel production sys- 2 Methods
tem from palm oil consists of the cultivation of oil palm, palm
oil milling, CPO refining, biodiesel production, and all trans- 2.1 Goal and scope
port activities (Silalertruksa et al. 2012). In the production of
biodiesel, CPO and MPO are used as raw materials. The goal of this work was to determine the GHG emissions in
Some publications report the determined greenhouse gas the conversion of palm fruit to MPO by the dry extraction
(GHG) emissions from the wet extraction process in process. The main sources of these GHG emissions were de-
Thailand (H-Kittikun et al. 2009; Chuchuoy et al. 2009; termined and compared to the wet extraction process. In ad-
Kaewmai et al. 2012; Kaewmai et al. 2013). The wet extrac- dition, direct land use changes (dLUCs) and indirect land use
tion process of CPO generates GHG emissions from certain changes (iLUCs) affect the GHG emissions of wet and dry
activities that include the acquisition of FFBs, the production extraction processes, and these effects were evaluated. The
of fossil fuel chemicals and electricity, the use of fossil fuel, concept of life cycle assessment (LCA) based on a cradle to
and the transportation of all inputs. In addition, the anaerobic gate evaluation was used in this study. The LCA approach has
treatment of wastewater could emit high levels of GHG. been used to determine the GHG emissions for the production
The average GHG emissions from palm oil mills with a wet of palm oil (H-Kittikun et al. 2009; Schmidt 2010; Choo et al.
extraction process, with and without a biogas capture system 2011; Kaewmai et al. 2012; Kaewmai et al. 2013) and palm
in Thailand, and without allocation to co-products were oil-derived biodiesel (Hansen et al. 2012; Silalertruksa and
1039 kg CO2 eq/t CPO in 2009. The major sources of GHG Gheewala 2012; Stichnothe et al. 2014). The functional unit
emissions were the wastewater treatment plants (57.5% of of the dry extraction process was 1 t of MPO.
total GHG emissions) and FFB acquisition (40.6%)
(Kaewmai et al. 2012). Palm oil mills in Thailand produced 2.2 System boundary
GHG emissions of 935 kg CO2 eq/t CPO, on average in 2010.
Wastewater treatment plants with biogas capture produced Figure 1 presents the system boundary of the dry extraction
about 42% of the total GHG emissions, whereas those without process. The raw material is palm fruit. The dry extraction
biogas capture emitted 64% of the total GHG emissions process of MPO from palm fruit includes three major sections:
(Kaewmai et al. 2013). (1) the primary production process, (2) the oil room, and (3)
For the dry extraction process, the palm fruit is separated the utilities. In this study, all GHG emissions from raw mate-
from the FFBs. Then, the fruit is transported to the dry extrac- rial production, transportation, and processing were included.
tion mill, while the empty fruit bunches (EFBs) are brought FFBs are harvested from palm trees, and the palm fruit is
back to oil palm plantations as soil conditioner. MPO is the separated from the EFBs in the plantation or at a collection
main product, and it is used to produce either edible oil or point by workers. The EFBs account for 19.9% by weight of
biodiesel, while palm cake and fine palm residue are co-prod- the FFBs on wet basis (Kaewmai et al. 2012). In the primary
ucts. Currently, all palm cake and fine palm residue from dry production, palm fruit is transported to the palm oil mill by
extraction mills are sold for use in animal feed. Unlike the wet trucks for immediate processing. The fruit is fed to a chamber
extraction process of CPO, the dry extraction process of MPO for cleaning and then exits through a drying chamber heated to
1804 Int J Life Cycle Assess (2017) 22:1802–1814

Fig. 1 System boundary for


producing mixed palm oil (MPO) Seedling Oil
O il palm
palm cultivation
cultivation
by dry extraction Soil preparation
Fertilizer

Fossil fuel
Cultivation
Plastic bag
Maintaining
Agriculture chemical
Harvesting
Electricity

Palm fruit separation Empty fruit bunch to oil palm plantation

Palm fruit

Palm
Palm oil
oil extraction
extraction

Primary production Oil room

Diesel fuel
Screw press Palm cake

Cleaning
Unfiltered mixed
Firewood palm oil

Drying chamber Screen


Fine palm residue
Electricity
Filter press

Transportation Mixed palm oil

MPO

Where
Where
Input Intermediate
Intermediate product
produ
d ct Co-product Direction

Process
Process Product Transportation

100 °C using firewood. In the oil room, the dry palm fruit is and denitrification in soil generates nitrous oxide emissions.
fed into a single-screw press where the oil is extracted. The The recycled amount of organic nitrogen on an oil palm plan-
outputs from this process are unfiltered MPO and palm cake. tation is between 64.5 and 180 kg nitrogen (N)/ha year, with
The MPO, including the CPO and palm kernel oil (PKO), is an average value of 100 kg N/ha year. When the average value
separated from fine palm residue by screening and filtration. of recycled amount of organic nitrogen was used in GHG
The MPO is collected into a storage tank until it is sold to an estimates, the nitrous oxide emission increased from 3.51
oil refinery. The palm cake and fine palm residue are co- (without recycled organic nitrogen on an plantation) to
products sold for use in animal feed.The GHG emissions for 4.89 kg/ha year (Stichnothe et al. 2014). The nitrogen
the production of palm fruit are not available. The GHG emis- recycling on the plantation was included in the estimate of
sions of palm fruit production, therefore, were calculated by GHG emissions. In addition, sensitivity analysis was conduct-
using the GHG emissions of FFB production, obtained from a ed, for effects of nitrous oxide emissions from fronds on the
study by the German International Cooperation (GIZ) (GIZ plantation on the GHG emissions of total oil (CPO and PKO)
2011a) with system boundary of cradle to gate. The palm fruit and MPO.
was then transported in trucks to the mills. The GHG emis- The utilities section of the dry extraction process includes
sions from the transport of palm fruit from the plantation to the firewood, diesel, and electricity. The firewood is used for dry-
collection point, and from the collection point to the dry ex- ing palm fruit. The emissions from transportation of firewood
traction mills, were included in the calculations. There were into the mill are included in the calculations. In this current
no GHG emissions caused by separating palm fruit from the work, the mill used hevea wood and hevea branches (from
FFBs because this was manually done by workers. rubber plantations or others) as firewood. The GHG emissions
During the harvest of FFBs, fronds are left on the oil palm of production, transportation, and carbon stocks of firewood
plantation. The decomposition of these fronds by nitrification were included in the GHG estimates. The firewood is
Int J Life Cycle Assess (2017) 22:1802–1814 1805

considered as a co-product of a rubber plantation. Diesel is Case (i) without carbon stocks of firewood and sharing the
used by the trucks’ moving inputs, products, co-products, and GHG emissions to the products and co-products.
firewood used in the production process. The GHG emissions Case (ii) without carbon stocks and substituting the GHG
of production, transportation, and combustion processes of the emissions of co-products for the GHG emissions of animal
diesel were counted. The GHG emissions from the electricity feed replaced.
consumed and supplied from the Provincial Electricity Case (iii) without carbon stocks, using fine palm residue
Authority were included. instead of firewood, and sharing the GHG emissions to prod-
Transportation is another section that can generate GHG ucts and co-products.
emissions. For the dry extraction process, the transportation Case (iv) with carbon stocks and sharing the GHG emis-
of all inputs to the mill was taken along in the GHG calcula- sions to products and co-products.
tions. The CO2, CH4, and N2O that are recognized by the Case (v) with carbon stocks and substituting the GHG
Kyoto protocol (United Nations Framework Convention on emissions of co-products for the GHG emissions of animal
Climate Change, UNFCCC 1998) were counted. In general, feed replaced.
the dLUC and iLUC have significant effects on GHG emis- Case (vi) with carbon stocks, using fine palm residue in-
sions. Previous work related to the dLUC and iLUC for palm stead of firewood and sharing the GHG emissions to products
oil production in Thailand was reviewed and considered and co-products.
(Silalertruksa and Gheewala 2012). The GHG emissions (or Case (vii) using LPG instead of firewood and sharing the
savings in emissions, if the effect was a reduction) from the GHG emissions to product and co-products.
dLUC and iLUC effects, and from the switch from a wet to a Case (viii) using LPG instead of firewood and substituting
dry process, were determined at the final stage of assessment. the GHG emissions of co-products for the GHG emissions of
animal feed replaced.
2.3 Methodology for calculating the GHG Case (ix) without carbon stocks, including nitrogen
recycling on oil palm plantation from fronds, sharing the
The GHG calculations and determination of GHG emissions GHG emissions to the products and co-products.
from the dry extraction of MPO followed the LCA approach. Case (x) without carbon stocks and substituting the GHG
The methodologies provided by several organizations, avail- emissions of co-products for the GHG emissions of animal
able for calculating GHG emissions, were reviewed, including feed replaced, including nitrogen recycling on oil palm plan-
Thailand Greenhouse Gas Management Organization (TGO) tation from fronds.
(TGO 2011), Intergovernmental Panel on Climate Change Case (xi) including carbon stocks, including nitrogen
(IPCC) (IPCC 2006a,b,c), and UNFCCC (UNFCCC 2010). recycling on oil palm plantation from fronds and sharing the
Our application target was to calculate the GHG emissions GHG emissions to the products and co-products;
from the conversion of palm fruit to MPO by the dry process. Case (xii) including carbon stocks and substituting the
Case studies were undertaken with and without carbon GHG emissions of co-products for the GHG emissions of
stocks of firewood and recycling of nitrogen from fronds on animal feed replaced, including nitrogen recycling on oil palm
the plantation. At the present time, the mill uses hevea wood plantation from fronds.
and hevea branches as biomass fuel. The hevea branches Replacing the GHG emissions of co-products with those of
come from the felling of rubber trees in the rubber plantation. animal feed, and of firewood by those of fine palm residue and
After the felling of old rubber trees, the cultivators plant new LPG, was calculated in accordance with heating values. It
young rubber trees. The burning of hevea branches can be must be noted that replacement of firewood by fine palm res-
considered carbon neutral, so their carbon stock was not con- idue, if the amount of fine palm residue is insufficient, fire-
sidered. However, when firewood comes from other sources wood will be replaced only partly.
without planting new trees to replace the felled ones, the car- The emission factors used by TGO (2011) and IPCC
bon stocks must be counted. (2006b) were adopted for all factors here (Table S1,
At the present time, the co-products are sold for use in Electronic Supplementary Material). The default value of
animal feed. Shortages of firewood used as the biomass fuel the aboveground biomass of mature rubber was obtained
can occur, and other biomass/fuel sources are sought. The co- from the study of Palm and colleagues (Palm Woomer
produced fine palm residue has a reasonable heating value and et al. 1999) in an IPCC report (IPCC 2006c). The heating
can be used as biomass fuel, but liquefied petroleum gas value and quantity of hevea wood per hectare (Table S2,
(LPG) is normally used for drying. Use of fine palm residue Electronic Supplementary Material) were obtained from
and LPG was proposed to replace the use of firewood. the National Metal and Materials Technology Center
The categories to determine GHG emissions of MPO, palm (MTEC 2013). The emission factor and heating value of
cake, and fine palm residue were divided into 12 cases as the feed for broilers were obtained from the Department
follows: of Live Stock Development (DLD 2015) and TGO (TGO
1806 Int J Life Cycle Assess (2017) 22:1802–1814

2011), respectively (Tables S1 and S2, Electronic average, it took 6.07 t of FFBs to generate 1.21 t of EFBs
Supplementary Material). (19.9% by mass of FFB) (Kaewmai et al. 2012). The re-
Allocations in accordance with mass, energy, and econom- maining 80.1% was palm fruit. So, according to these data,
ic values were used to attribute the GHG to the main product 1 t of palm fruit was produced from 1.25 t of FFBs. This is
MPO and to the co-products palm cake and fine palm residue. the value used in our calculations.
This was because further processing of the MPO could con- For the dry extraction process, four scenarios were used for
vert it to food or to fuel in the energy supply chain. A weighted conducting sensitivity analysis. The GHG emissions in cases
average using the annual MPO production of each dry extrac- (i), (iv), (ix), and (xi) were used as baselines. In the case of wet
tion mill to average the GHG emissions was adopted. extraction process, the GHG emission baselines were deter-
For the dry extraction process, the GHG emissions from mined by palm oil mill with biogas capture system, without
the raw material (palm fruit), the diesel fuel, the firewood, (Kaewmai et al. 2012) or with nitrogen recycling on plantation
and the electricity used were summed to the total GHG from fronds. The nitrous oxide emissions by nitrogen
emissions as follows: recycling on plantation from fronds, in the GHG calculation,
were set at 0.6 and 1.8 times of an average value (Stichnothe
et al. 2014).
E Dry extraction ¼ E Palm fruit þ E Fuel þ E Firewood þ EElectricity

The factors are described below. 2.4 Data collection


EDry extraction is the total GHG emissions from the
dry extraction process (kg CO2 eq). The two dry extraction mills that participated in this study are
EPalm fruit is the GHG emissions from the production located in southern Thailand. All the relevant data for the
and transportation of palm fruit calculations were collected continuously. In order to establish
(kg CO2 eq). The nitrogen recycling on oil the reliability of the GHG emission results, as much relevant
plantation from fronds was considered as data were gathered as possible. Several methods of data col-
GHG emissions of the palm fruit. lection were used in this study, such as on-site interviews,
EFuel is GHG emissions from diesel surveys, and questionnaires.
fuel production, transport, and The data used for the calculation of the GHG emissions
combustion (kg CO2 eq) were collected over a 1-year period. The emission factors
EFirewood is the GHG emissions from production, used to convert the quantities of inputs into resulting GHG
transportation, and combustion of firewood emissions were obtained from the IPCC and the TGO.
(kg CO2 eq). The carbon stocks of firewood These factors were expressed as the amount of GHG emit-
are also considered. The firewood is ted per unit. The MPO, palm cake, and fine palm residue
considered a co-product of rubber planta- were collected and analyzed for their lower heating value
tions. The GHG of rubber plantation was (LHV) on a wet basis, using the Leco-AC-500 automatic
distributed to dry rubber, hevea wood, calorimeter.
and hevea branches.
EElectricity is the GHG emissions from electricity
consumption (kg CO2 eq) 3 Results and discussion

The amount of palm fruit used in production was con- 3.1 Dry extraction process
verted to the amount of FFBs and multiplied by their emis-
sion factor, to get the GHG emissions for the production of The inventory list of two dry extraction processes for pro-
palm fruit. This is because there was no prior data for the ducing 1 t of MPO is presented in Table 1. The amount of
GHG emissions from palm fruit production, while the palm fruit required for the production of 1 t of MPO was
GHG emissions of FFB production were available. The 4.5 t. The average oil extraction rate (OER) of this dry
uncertain composition of the palm fruit in FFBs is impor- extraction process was 22.3%, but an average OER of
tant for the calculation of GHG emissions. One ton of 18% has been obtained for mills using the wet extraction
FFBs generated EFBs from 0.19 to 0.24 t (19.1 to 24% process in Thailand (DEDE 2006). However, the OER of
by the mass of the FFBs). The rest (76.0 to 80.9%) was the wet extraction process across 14 mills was 16.5% in
attributed to palm fruit (Kaewmai et al. 2012; Jungniyom 2009, and across a sample of 6 mills, it was 17.0% in 2010
2009; Chavalparit et al. 2006). Conversely, 1 t of the palm (Kaewmai et al. 2012; Kaewmai et al. 2013). The reported
fruit was produced from about 1.24 to 1.35 t of FFBs. With OER of the wet extraction process in Malaysia was 20.5%
regard to the life cycle inventory data of the 14 mills, on (MPOB 2010).
Int J Life Cycle Assess (2017) 22:1802–1814 1807

Table 1 Inventory lists of two mills using dry extraction 3.2 GHG emissions of the dry extraction process and their
Item Unit Consumption/production a main sources

Mill 1 Mill 2 Averageb In the dry extraction process, without the carbon stocks of
firewood in the calculation, the average values of the total
Palm fruit t 4.55 4.36 4.49
GHG emissions were 537 kg CO2 eq (weighted average) be-
Electricity used kWh 98.2 84.0 93.6
fore the allocation of the extraction of 1 t of MPO in cases (i)
Firewood t 0.27 1.04 0.52
and (ii) (Fig. 2 and Table S3, Electronic Supplementary
Diesel fuel l 3.41 2.44 3.09
Material). The GHG value before allocation was used to rep-
MPO t 1.00 1.00 1.00
resent the total amount of GHG emissions from the dry ex-
Palm cake t 0.54 0.81 0.63 traction process and was also used in this current work to
Fine palm residue t 0.38 0.35 0.37 determine a breakdown of the GHG emissions. The acquisi-
The amounts correspond to producing 1 t of MPO tion of palm fruit was the major source of GHG emissions, at
a
The values are expressed per functional unit
about 448 kg CO2 eq/t MPO. This emitted on average 83.4%
b
Weighted average by MPO produced
of the total GHG emissions (Table S4, Electronic
Supplementary Material). The other sources of GHG emis-
sions were the electricity used (9.78%) and the production
To make an equitable comparison between the OER of dry and transportation of firewood (4.99%). The GHG emissions
extraction process and wet extraction process, it is important from diesel fuel used were 1.82% of the total.
to distinguish between the extraction rates of CPO and PKO. When fine palm residue biomass was used as the fuel for
Kaewmai et al. (2012) stated that the PKO can be either pro- drying in case (iii), the amount of firewood was reduced and
duced on-site by a palm oil mill using a screw press followed its GHG emissions dropped to 8 kg CO eq/t MPO. The GHG
by filtration or off-site in specialized mills in Thailand. The emissions of the palm fruit, electricity, and diesel fuel were
reported PKO yield on-site was 0.0116 t per 1 t FFBs. similar to those of cases (i) and (ii). The total GHG emissions
Stichnothe et al. (2014) reported typical mass flow values of obtained were 518 kg CO2 eq/t MPO, and this was the lowest
CPO production. One ton of FFBs can be used to produce across all the cases studied. The acquisition of palm fruit was
0.075 t of palm kernel, which yields 0.038 t PKO (i.e., 3.8% again the major source of GHG emissions contributing 86.5%.
by weight of FFBs). When the OER of PKO—from the stud- When the carbon stocks of firewood were included, the
ies of Kaewmai et al. (2013) and Stichnothe et al. (2014)— aboveground biomass of mature rubber was converted to car-
was used to calculate the total OER of CPO + PKO, this total bon stocks of firewood. The burning of this firewood emitted
OER of wet extraction process ranged from 17.7 to 21.8%. carbon dioxide. The GHG emissions per 1 t of MPO extracted
The OER of the dry extraction process was slightly higher were 871 kg CO2 eq, before the allocation (case (iv)) and
than that of the wet extraction process. This is because only substituting the GHG emissions of co-products for the GHG
palm fruit was used as raw material in the dry extraction pro- emissions of animal feed replaced (case (v)). The major
cess, whereas the wet extraction process used FFBs composed sources of GHG were the acquisition of palm fruit (51.4%)
mainly of palm fruit and EFBs. The co-products correspond- and the carbon stocks combined with production and trans-
ing to 1 t of extracted MPO were palm cake (0.63 t) and fine portation of firewood (41.5%). When fine palm residue was
palm residue (0.37 t). The average values of other inputs used as fuel for drying in case (vi), the amount of firewood
employed in the production of 1 t MPO were firewood used decreased. Then, the GHG of firewood was reduced to
0.52 t, diesel fuel 3.09 L, and electricity 93.6 kWh. 102 kg CO eq/t MPO, and total GHG emissions of 612 kg
The average percentages of outputs per input mass of palm CO2 eq/t MPO were obtained.
fruit, on a wet basis, were 22.3% of MPO, 14.0% of palm LPG can be used as fuel on drying palm fruit, instead of
cake, and 8.24% of fine palm residue, for the dry extraction using firewood. However, the GHG emissions before the al-
process. These outputs do not add up to 100% of input, be- location and substituting the GHG emissions of co-products
cause moisture is lost during drying. The average LHVs of for the GHG emissions of animal feed replaced were 1125 kg
MPO, palm cake, and fine palm residue, determined with an CO2 eq/t MPO. The use of LPG generated the highest GHG
automatic calorimeter, were 39,048, 19,165, and 27,063 MJ/t, emissions of all the cases considered. This mostly originated
respectively. An LHV of 39,212 MJ/t was reported earlier for from the production, transportation, and combustion of LPG
CPO by Kaewmai et al. (2012). The LHVs of the MPO and (54.7% of total GHG emissions) and the acquisition of palm
CPO were similar. The average price of CPO in 2010 was 917 fruit (39.8%). The GHG from the production, transportation,
US dollar (USD)/t (Department of Industrial Trade, DIT and combustion of LPG was 615 kg CO eq/t MPO.
2016). The average prices of MPO, palm cake, and fine palm Without the carbon stocks of firewood, including the annu-
residue were 820, 131, and 74 USD/t, respectively. al nitrogen recycling on oil palm plantation varied between 60
1808 Int J Life Cycle Assess (2017) 22:1802–1814

Cases (i-ii)
Diesel Fuel Cases (iii)
Cases (iv-v)
Cases (vi)
LPG Cases (vii-viii)
Cases (ix -x)
Cases (xi-xii)
Firewood

Electricity

Palm fruit

Total GHG
emission

0 100 200 300 400 500 600 700 800 900 1000 1100 1200

kgCO2eq/t MPO
Fig. 2 Breakdown of GHG emissions from dry extraction process, before allocation (cases (i), (iii), (iv), (vi), (vii), (ix), (xi)) and before reduction of
GHG emissions from animal feed production replaced by co-products (cases (ii), (v), (viii), (x), (xii))

and 180 kg/ha, the GHG emissions increased from the case firewood (36.0%). Nitrogen recycling increased GHG emis-
without nitrogen recycling (case (i)), namely from 15 to 44% sions from no recycling (case (iv)) by 15%. The sensitivity
(Fig. 1S, Electronic Supplementary Material). Using the aver- analysis (Fig. 2S, Electronic Supplementary Material) shows
age value 100 kg N/ha year, the total GHG emissions were that when the annual nitrogen recycling on oil palm plantation
668 kg CO2 eq per 1 t MPO in cases (ix) and (x). The increase was varied from 60 to 180% in case (xi), the GHG emissions
of GHG emissions from case (i) by 24.4% was significant. were affected by −5 to 10%.
The acquisition of palm fruit was the major source of GHG For the wet extraction process with biogas capture, the GHG
emissions at about 578 kg CO2 eq/t MPO. This emitted on emissions from acquisition of raw material, chemical input and
average 86.5% of the total GHG emissions (Table S4, disposal of chemical packaging waste, energy input, and waste-
Electronic Supplementary Material). Figure 2S (Electronic water management were included. In the wastewater manage-
Supplementary Material) presents the sensitivity analysis ment, a biogas capture system was used for treating the palm oil
when the nitrogen recycling on plantation is included (case mill effluent. The GHG emissions from wastewater treatment,
(ix)). When the annual nitrogen recycling on oil palm planta- sludge treatment, wastewater discharge, fugitive emissions, and
tion was varied from 60 to 180% of the average level, the flaring were included in the calculation. The wet extraction
GHG emissions were affected by −8 to 16%. process with biogas capture before allocation emitted 1039 kg
When the carbon stocks of firewood and nitrogen recycling CO2 eq/t CPO of GHG. The major sources of GHG emissions
on plantation were included and the annual nitrogen recycling were the wastewater treatment plant and the acquisition of
varied between 60 and 180 kg/ha, the GHG emissions in- FFBs. Their GHG emissions accounted for 49.4 and 48.7%
creased from the case without nitrogen recycling but with of the total, respectively (Kaewmai et al. 2012). When the
included carbon stocks (case (iv)), from 9 to 27%. At the GHG emissions and amount of on-site PKO were included in
average nitrogen recycling rate of 100 kg N/ha year, the the calculation, the wet extraction process with biogas capture
GHG emissions per 1 t of MPO extracted were 1002 kg emitted GHG of 975 kg CO2 e/t total oil (CPO and PKO).
CO2 eq, before the allocation (case (xi)) and substituting the When the annual nitrogen recycling was included and varied
GHG emissions of co-products for the GHG emissions of between 60 and 180 kg/ha, the GHG emissions increased from
animal feed replaced (case (xii)). The major sources of GHG the case without nitrogen recycling by 9 to 26%. At the average
were the acquisition of palm fruit (57.8%) and the carbon 100 kg N/ha year, the total GHG emissions were 1115 kg CO2
stocks combined with production and transportation of eq per 1 t total oil. The sensitivity analysis in Fig. 2S (Electronic
Int J Life Cycle Assess (2017) 22:1802–1814 1809

Supplementary Material) shows that when the annual nitrogen CO2 eq, respectively, for case (ix). Savings in GHG emissions
recycling on oil palm plantation was varied from 60 to 180% of 219 kg CO2 eq/t MPO were found in case (x).
around the average, the GHG emissions varied by −5 to 10%. It When the carbon stocks of firewood were included, the
can be stated that nitrogen recycling significantly affected the GHG emissions of MPO were 429, 550, and 768 kg CO2
GHG emissions, with both dry extraction and wet extraction. eq, based on allocation by mass, energy, or economic value,
When the carbon stocks of firewood were not included, the in case (iv). Attribution by economic value allocated the
GHG emissions from the dry process in cases (i) and (ii), (iii), highest amount of GHG emissions to MPO. The GHG emis-
and (ix) and (x) were 45, 47, and 31% lower than from the wet sions increased by 38% when carbon stocks were included.
extraction process, respectively. When the carbon stocks of On substituting the feed for broilers with the co-products, the
firewood were included, the GHG emissions from the dry ex- GHG emissions were reduced by 15 kg CO2 eq/t MPO. When
traction process in cases (iv) and (v) were 15% lower than from nitrogen recycling was included, the GHG emissions of MPO
wet extraction, respectively. On using the fine palm residue as allocated by mass, energy, and economic values were 495,
biomass fuel for drying, case (vi), the dry extraction emitted 634, and 884 kg CO2 eq, respectively, in case (xi). GHG
37% less GHG than the wet extraction. The reduced GHG emissions of 116 kg CO2 eq/t MPO were found for case
emissions seem to favor dry extraction over wet extraction. (xii). If the MPO is used to produce biodiesel or used in the
Without carbon stocks and included nitrogen recycling on energy sector, the GHG emissions should be allocated by
plantation from fronds, the GHG emissions from the dry pro- LHV. If the MPO is used for edible oils and cosmetics, then
cess in cases (ix) and (x) were 31% lower than from the wet price is the first priority followed by mass. This is because it is
extraction process. When carbon stocks and nitrogen unfair to share or attribute GHG on mass basis to MPO and its
recycling were included, the GHG emissions from dry extrac- low value co-products.
tion process were 3% higher than from the wet extraction The average GHG emissions by economic value allocation
process. In the case of lack of firewood, LPG can be used as of palm cake in case (i) were 76 kg CO2 eq/(t palm cake).
the fuel in the drying process. However, when LPG was used When carbon stocks were included (case (iv)), the GHG emis-
in cases (vii) and (viii), the GHG emissions increased by 29% sions shared among products and co-products by economic
in comparison to case (iv). In addition, the GHG emissions in value were 123 kg CO2 eq/(t palm cake). In case (i), the pro-
cases (vii) and (viii) were 15% higher than for wet extraction, duction of a ton of fine palm residue emitted an average GHG
when LPG was used. The mill should stock sufficient fire- amount of 42 kg CO2 eq, when weighted by economic value.
wood or use fine palm residue and avoid using LPG. In case (iv) with carbon stocks included, the GHG emissions
allocated by economic value were 67 kg CO2 eq (Table S5).
3.3 GHG emissions of MPO, palm cake, and fine palm The allocation by mass showed the highest GHG emissions
residue for palm cake and fine palm residue, followed by attributions
based on energy and economic values. For the palm cake and
The average GHG emissions of MPO production for the cases fine palm residue, the contribution by economic value is the
(i)–(xii) are presented in Fig. 3 and in Table S5 (Electronic first priority when they are used in animal feeds.
Supplementary Material). At present, the mill uses firewood
as fuel for drying and the co-products are sold for use in 3.4 Recommendations for GHG mitigation
animal feeds. The GHG emissions were not assessed for the
cases where LPG was used with fine palm residue. A comparison of the dry and the wet palm oil extraction
Without carbon stocks, the GHG emissions for producing a processes, for the inputs and the outputs, is shown in
ton of MPO allocated by mass, energy, and economic values Table S6 (Electronic Supplementary Material). The palm
were 267, 341, and 473 kg CO2 eq, respectively, in case (i). fruit was the major input in the dry extraction process,
The GHG emissions of the feed for broiler were used to sub- whereas the FFBs were the major input in the wet extraction
stitute GHG emissions of palm cake and fine palm residue in process. The MPO and the CPO were the products from the
case (ii). The amounts of palm cake and fine palm residue dry and the wet extraction processes, respectively. For the
were converted to amounts of feed for broiler on the basis of co-products, the dry extraction process produced palm cake
heating value. Then, the amount of feed for broiler was mul- and fine palm residue. The wet extraction process had palm
tiplied by its emission factor to get GHG emissions replaced kernels and shells as co-products. In some mills, the palm
by the GHG emissions of palm cake and fine palm residue. In kernel was treated with a screw press and filtered to get
this case, the high levels of GHG emissions were reduced. The PKO and palm kernel meal (Kaewmai et al. 2012). In terms
savings in GHG emissions would be 350 kg CO2 eq/t MPO of the wastes, the dry extraction process did not generate
produced. When the nitrogen recycling on the plantation from solid waste or wastewater. The wet extraction process gen-
fronds was included, the GHG emissions of MPO allocated by erated EFBs, decanter cake, and wastewater. These wastes
mass, energy, and economic values were 333, 425, and 589 kg require good system management.
1810 Int J Life Cycle Assess (2017) 22:1802–1814

1200

1000

Mass Energy Price


800

600
kg COeq/to m MPO

400

200

0
Case (i) Case (iii) Case (iv) Case (vi) Case (vii) Case (ix) (Case (xi) Case (ii) Case (v) Case (viii) Case (x) Case (xii)

-200

-400

-600
Fig. 3 Allocation of GHG emissions from producing MPO by dry extraction

A comparison of the GHG emissions from the alternative nitrogen recycling on plantation from fronds, without carbon
processes is summarized in Table 2. For the dry extraction stocks, the palm fruit was the major source of GHG emissions.
process, the major sources of the GHG emissions were in With carbon stocks, the major GHG emissions were the palm
obtaining the palm fruit and in the electric consumption, when fruit acquisition and firewood acquisition and combustion.
the carbon stocks of firewood were not included. In the case The major sources of GHG emissions in the wet extraction
where carbon stocks were included, obtaining the palm fruit process were the wastewater treatment process (49.4%) and
and the production, transportation, and combustion of fire- the FFB acquisition (48.7) (Kaewmai et al. 2012). When the
wood were the major sources of GHG emissions. Including GHG emissions of kaolin production and transportation were

Table 2 Comparison of GHG


emissions from dry and wet Resource Percent of GHG emissions (%)
extraction processes
Dry extraction process Wet extraction
processa

Cases (i) Cases (iv) Cases (ix) Cases (xi) Case Case
and (ii) and (v) and (x) and (xii) (a) (b)

Raw material 83.4 51.4 67.0 44.7 48.5 42.8


Nitrogen recycling on oil – – 19.6 13.1 – 11.8
plantation from fronds
Chemicals used – – – – 0.85 0.75
Firewood 4.99 41.5 4.01 36.0 – –
Fossil fuels 1.82 1.16 1.46 0.97 1.07 0.94
Electricity 9.78 6.03 7.86 5.24 0.42 0.37
Wastewater – – – – 49.2 43.4
a
Adopted from Kaewmai et al. (2012) for mills with biogas capture system: Case (a) without nitrogen recycling
on oil plantation from fronds, and case (b) with nitrogen recycling on oil plantation from fronds
Int J Life Cycle Assess (2017) 22:1802–1814 1811

included in the calculation, the wastewater treatment process Crop displacement was used for the calculation of potential
(49.2%) and the FFB acquisition (48.5%) were the major iLUC effects on GHG emissions. The grassland and forestland
GHG emitted sources. Including GHG emissions from kaolin were displaced for growing rubber and cassava. The potential
and nitrogen recycling on plantation from fronds, the GHG iLUC effects of the scenarios were determined as (i) grassland
emissions of FFB acquisition and the wastewater treatment to rubber in the northeast, (ii) forestland to rubber in the north-
process accounted for 54.6 and 43.4% of the total GHG emis- east, (iii) grassland to cassava in the northeast, (iv) forestland
sions, respectively. It can be stated that obtaining the FFBs and to cassava in the northeast, and (iv) grassland to cassava in
palm fruit were the major sources of GHG emissions regard- Vietnam (Silalertruksa and Gheewala 2012).
less of the process. In case the environmental policy makers Our current study used the results of Silalertruksa and
wish to address the GHG emissions from palm oil production, Gheewala (2012) in the calculation of GHG emission effects
the dominant portion of these emissions come from the oil from dLUC and iLUC. The dLUC and iLUC related to forest-
palm plantation. land conversion were not included in the calculations of GHG
GIZ has developed GHG calculations and optimization savings, in our study. This is because, in Thailand, setting up
guidelines for oil palm plantations in Thailand (GIZ plantations for economic crops in forestland is illegal
2011a,b). The GHG emissions from chemical fertilizer pro- (National reserved forest Act 1964). At present, the central
duction, transportation, and the utilities accounted for 80% of and local governments strongly enforce the law and require
the total GHG emissions in the plantations. Reducing the registration, in order to preserve the forestland. The GHG
GHG emissions from the oil palm plantations could be savings (t CO2 eq/ha year) gained by changing from wet ex-
achieved by increasing the quantity and quality of the pro- traction process to dry extraction process, including dLUC
duced FFBs. The principles of Good Agricultural Practice and iLUC effects, were evaluated. This was based on five
(GAP) advocated by the Department of Agriculture (DOA scenarios: (1) rubber to oil palm in the south and grassland
2012) must be used by growers of oil palms. In GAP, appro- to rubber in the northeast, (2) cassava to oil palm in the north-
priate fertilizer and soil analysis are recommended for use in east or the east and grassland to cassava in the northeast, (3)
oil palm plantations. In addition, the nitrogen recycling on cassava to oil palm in the northeast or the east and grassland to
plantation from fronds significantly increases the GHG emis- cassava in Vietnam, (4) paddy field to oil palm in the south,
sions (Stichnothe et al. 2014). Therefore, it must be considered and (5) land set-aside for oil palm in the south.
in the calculation of GHG emissions and in the management From 2009 to 2013, the average FFB production was
of palm oil plantation. 17.6 t/ha year across the oil palm plantations in Thailand
The main cause of GHG emissions from oil palm planta- (OAE 2015). The estimates of MPO and CPO extraction were
tions was N fertilizer (approximately 80%) (GIZ 2011a). based on the amount of FFBs produced in 1 ha. An OER of
Therefore, organic fertilizer use of the EFBs was recommend- 22.3% for the dry extraction process (this study) and of 16.5%
ed, to partly replace chemical fertilizers (GIZ 2011b). Soil of CPO and 1.16% of PKO for the wet extraction process
analyses must be done initially and later to monitor the oper- (Kaewmai et al. 2012) were used. The GHG emissions before
ation of a plantation and to select the most appropriate fertil- allocation to product and co-products were used as represen-
izers. In practice, these guidelines should be used to educate tative of the total GHG emissions, for dry or wet extraction.
oil palm growers in both small-scale and large-scale planta- The GHG emissions and savings from converting a wet ex-
tions (GIZ 2011b; Uniliver 2003). traction process to a dry extraction process are presented in
Among the specific major sources of GHG emissions from Table 3 for palm oil production.
the dry extraction process, the electricity supplied by the When a wet extraction process was used, the GHG emis-
Provincial Electricity Authority was considered an important sions before allocation without (with) nitrogen recycling on
source. Further development should focus on the technology oil plantation from fronds were 3.03 (3.44) t CO2 eq/ha year.
needed for using fine palm residue as biomass fuel to generate With the dry extraction process, we assessed four cases: case
electricity, as 20.43 kWh/t (palm fruits) is used by the mill. (i) without carbon stocks, case (iv) including carbon stocks,
Silalertruksa and Gheewala (2012) evaluated the GHG case (ix) without carbon stocks and including nitrogen
emissions from palm biodiesel in Thailand. The future avail- recycling from fronds, and case (xi) including carbon stocks
abilities of FFB and CPO as feedstock were estimated, to help and nitrogen recycling from fronds. The GHG emissions
satisfy the future demands of food and fuel. The dLUC and were 1.69, 2.73, 2.10, and 3.15 t CO2 eq/ha year, respec-
iLUC were included in the GHG calculations. Five dLUC tively (Table 3). Eventually, the GHG savings were 1.34,
scenarios in Thailand were examined including (i) rubber to 0.30, 1.34, and 0.30 t CO2 eq/ha year, respectively, from
oil palm in the south, (ii) cassava to oil palm in the northeast or conversion of the wet extraction process to the dry extrac-
the east, (iii) paddy field to oil palm in the south, (iv) land set tion process in the cases (i), (iv), (ix), and (xi). By way of
aside (such as grassland) for oil palm in the south, and (v) comparison, the GHG savings between cases (i) and (ix) and
forestland to oil palm in the south or the northeast. between cases (iv) and (xi) were similar.
1812 Int J Life Cycle Assess (2017) 22:1802–1814

Table 3 GHG emissions and savings from conversion of wet to dry palm oil extraction

Cases GHG emissions GHG saving for using dry


(t CO eq/ha year) process instead of wet
process (t CO eq/ha year)

Wet extraction process 3.03 –


Wet extraction process included nitrogen recycling on oil plantation from fronds 3.44 –
Dry extraction process (GHG before allocation)
Case (i) without carbon stocks 1.69 −1.34a
Case (ii) without carbon stocks and substituting the GHG emissions of co-products for the GHG −1.10 –
emissions of animal feed replaced
Case (iii) without carbon stocks, using fine palm residue instead of firewood 1.62 −1.41a
Case (iv) included carbon stocks 2.73 −0.30a
Case (v) included carbon stocks, substituting the GHG emissions of co-products for the GHG −0.05 –
emissions of animal feed replaced
Case (vi) included carbon stocks, using fine palm residue instead of firewood 1.92 −1.11a
Case (vii) LPG was used in drying process 3.53 +0.50a
Case (viii) LPG was used in drying process, substituting the GHG emissions of co-products for 0.75
the GHG emissions of animal feed replaced
Case (ix) without carbon stocks, including nitrogen recycling on oil palm plantation from fronds 2.10 −1.34b
Case (x) without carbon stocks, substituting the GHG emissions of co-products for the GHG −0.69 –
emissions of animal feed replaced, including nitrogen recycling on oil palm plantation from fronds
Case (xi) including carbon stocks, including nitrogen recycling on oil palm plantation from fronds 3.15 −0.30b
Case (xii) including carbon stocks, substituting the GHG emissions of co-products for the GHG 0.36 –
emissions of animal feed replaced, including nitrogen recycling on oil palm plantation from fronds

The GHG emissions before allocation from palm oil mill using wet extraction were adopted from Kaewmai et al. (2012)
a
The GHG saving was the difference between GHG of wet extraction process and dry extraction process (without nitrogen recycling on oil plantation
from fronds)
b
The GHG saving was the difference between GHG of wet extraction process and dry extraction process (with nitrogen recycling on oil plantation from
fronds)

The net GHG savings were obtained by combining the the conversion from wet to dry at existing conditions. Without
conversion from wet to dry extraction with the effects of the scenario (iii), the conversion of wet to dry extraction in sce-
dLUC and iLUC on GHG emissions and are shown in Table 4. narios (i), (ii), (iv), and (v) and cases (i) and (ix) accounted for
Without carbon stocks and including nitrogen recycling in the 7.3–47.2% of the net GHG savings, respectively. Both with
plantation, the net GHG saving ranged from 0.10 to 5.08 t and without carbon stocks for scenario (iii), the GHG savings
CO2 eq/ha year. The lowest net GHG savings were obtained mostly derived from the process conversion.
for scenario (iii) dLUC of cassava to oil palm and iLUC of It can be stated that the highest GHG savings with process
grassland to cassava in Vietnam and conversion from wet to conversion were for scenario (i) with rubber to oil palm
dry at existing conditions. The greatest net GHG savings (dLUC) and grassland to rubber in the northeast (iLUC).
could obtained in scenario (i) dLUC of rubber to oil palm The lowest savings were scenario (iii), cassava to oil palm
and iLUC of grassland to rubber in the northeast and conver- (dLUC) and grassland to cassava in Vietnam (iLUC).
sion from wet to dry at existing conditions. The conversion of Ministry of Energy promulgated the Alternative Energy
wet to dry extraction in scenarios (i), (ii), (iv), and (v) and Development Plan: AEDP 2015 in Thailand from 2015 to
cases (i) and (ix) accounted for 26.4–80.3% of the net GHG 2036. The target of biodiesel production in 2036 is 14.0 mil-
savings, respectively. lion l per day. To support this plan, the area of oil palm plan-
When carbon stocks and nitrogen recycling on plantation tations in Thailand must increase from 0.72 million ha in 2015
from fronds were included, the net GHG saving ranged from to 1.63 million ha in 2036 (MOE 2015). In 2015–2036, when
0.63 to 4.04 t CO2 eq/ha year. The lowest net GHG savings the wet extraction process with OER from 17.9 to 20.1% is
were obtained in scenario (ii) dLUC of cassava to oil palm and employed, the area should increase by 45,500 ha per year. The
iLUC of grassland to cassava in the northeast and the conver- PKO can be produced on-site at palm oil mill with OER
sion from wet to dry at existing conditions. The greatest net 1.16% per ton FFBs (Kaewmai et al. 2012). The OER of
GHG savings were obtained in scenario (i) dLUC of rubber to PKO was assumed to increase at the same rate with that of
oil palm and iLUC of grassland to rubber in the northeast and CPO. The total oil recovery rate was obtained by summing
Int J Life Cycle Assess (2017) 22:1802–1814 1813

Table 4 GHG emissions and


savings for five scenarios with Scenarios GHG saving (t CO2 eq/ha year)
dLUC and iLUC accounted for,
on conversion from wet to dry dLUC iLUC dLUC iLUC Net
extraction of palm oil
DP: case (i) DP: case (iv)
and (ix) to WP and (xi) to WP

(i) Rubber to oil palm GL to rubber in the −2.15 −1.59 −5.08 −4.04
northeast
(ii) Cassava to oil palm GL to cassava in the −2.68 2.35 −1.67 −0.63
northeast
(iii) Cassava to oil palm GL to cassava in −2.68 3.92 −0.10 0.94
Vietnam
(iv) Paddy field to oil palm – −1.71 – −3.05 −2.01
(v) Set-aside land to oil palm – −1.52 – −2.86 −1.82

Case (i) without carbon stocks. Case (iv) included carbon stocks. Case (ix) without carbon stocks, including
nitrogen recycling on oil palm plantation from fronds. Case (xi) including carbon stocks, including nitrogen
recycling on oil palm plantation from fronds. Case (xii) including carbon stocks and substituting the GHG
emissions of co-products for the GHG emissions of animal feed replaced, including nitrogen recycling on oil
palm plantation from fronds. The GHG emission factors for dLUC and iLUC were obtained from the study of
Silalertruksa and Gheewala (2012). The GHG emissions from palm oil mill with wet extraction process were
obtained from Kaewmai et al. (2012)
dLUC direct land use change, iLUC indirect land use change, WP wet extraction process, DP dry extraction
process, GL grassland

those of CPO and PKO and was used to estimate the increases conversion of wet to dry process. The iLUC was the main
in plantation area and in GHG emissions (Table S7, Electronic GHG emission source. The GHG saving mostly originated
Supplementary Material). The OER oil yield of the dry extrac- from the change of extraction process and dLUC.
tion process was 17.8% as weight fraction of the FFBs. Based The change of palm oil extraction process and the land use
on known increase of OAE from that of the wet process, the change effects could minimize the GHG emissions of the palm
OERs of dry process were assumed to be 18.8% for 2017 and oil industry. The advantages of dry extraction include not only
2019 and 19.8% for 2026 and 2036 (Table S7, Electronic reduction of GHG emissions but also reduction of investment
Supplementary Material). costs, as well as suitability for small-scale palm oil extraction
The 45,500 ha per year increase in area of oil palm planta- at community level. These advantages encourage developing
tions, and corresponding estimated yields of FFBs, CPO, MPO, policies supporting the dry extraction process and contributing
and PKO shown in Table S7 (Electronic Supplementary to sustainable developments in palm oil production.
Material), was used to calculate GHG savings. It must be noted
here that the total GHG emissions (before allocation to product
and co-products) were used as representative of the total GHG 4 Conclusions
emissions from the utilization of dry and wet extraction pro-
cesses. The yield of FFBs was assumed to decrease during the GHG emissions of dry extraction process and GHG savings
period from 2015 to 2036 (MOE 2015). The yields in CPO, from changing from wet to dry process, including land use
MPO, and PKO production were assumed to increase. change effects, were determined. The LCA done was based on
Without carbon stocks and including nitrogen recycling in a cradle to gate evaluation. Case studies were undertaken with
the plantation, the GHG saving ranged from 18,522 to and without the carbon stocks of firewood or the nitrogen
274,774 t CO2 eq/year. When carbon stocks and nitrogen recycling on plantations from fronds. The emissions were al-
recycling in the plantation were included, the scenario (iii) located to the various products, alternatively by mass, eco-
cassava to oil palm as dLUC and grassland to cassava in nomic value, or heating value. The GHG emissions of co-
Vietnam as iLUC, and wet to dry process at existing condi- products were also replaced by the GHG emissions of animal
tions, emitted GHGs between 18,902 and 35,454 t CO2 eq/ feed. Palm fruit and firewood were the major GHG emission
year. With the other scenarios, the GHG savings ranged from sources. The nitrogen recycling on plantation from fronds sig-
45,045 to 220,768 t CO 2 eq/year (Table S8, Electronic nificantly affected the GHG emissions. With the carbon
Supplementary Material). Without carbon stocks, the change stocks, the GHG emissions allocated by energy value were
from wet to dry process can save higher amount of GHG. The 550 kg CO2 eq/t MPO. The GHG emissions were affected
major GHG savings were from land use changes and from by −3 to 37% for the change from wet to dry process. When
1814 Int J Life Cycle Assess (2017) 22:1802–1814

the plantation area was increased by 1 ha and the palm oil H-Kittikun A, Cherabu M, Maliwan C (2009) Preliminary evaluation of
greenhouse gas production from oil palm plantation and crude palm
extraction was changed from wet process to dry process and
oil extraction in Thailand. Hatyai Journal 7(2):97–106
the change included dLUC and iLUC effects, GHG savings IPCC (2006a) Guidelines for national greenhouse gas inventories.
ranged from −0.94 to 5.08 t CO2 eq/ha year. The iLUC was Volume 5, the institute for global environmental strategies (IGES):
the main GHG emission source. The GHG savings mostly Hayama, Japan
IPCC (2006b) Guidelines for national greenhouse gas inventories.
originated from the change of extraction process and the
Volume 2 Energy, the institute for global environmental strategies
dLUC effect. Using a forecast of oil palm-based biodiesel (IGES): Hayama, Japan
production during 2015–2036 in Thailand, when the palm IPCC (2006c) Guidelines for national greenhouse gas inventories.
oil extraction process was changed and the change included Volume 4, Agriculture, Forestry and Other Land Use.
Intergovernmental Panel on Climate Change, Hatayama, Japan
dLUC and iLUC effects, the savings in GHG emissions
Jungniyom T (2009) Zero-waste process in oil palm extraction industries.
ranged from −35,454 to 274,774 t CO2 eq/year. Hatyai Journal 6(2):159–164
Kaewmai R, H-Kittikun A, Musikaving C (2012) Greenhouse gas emis-
Acknowledgements This research was funded by the project on sion of palm oil mills in Thailand. Int J Greenh Gas Con 11:141–151
Sustainable Palm Oil Production for Bioenergy, which is jointly imple- Kaewmai R, H-Kittikun A, Suksaroj C, Musikavong C (2013)
mented by OAE and GIZ. The project’s activities were commissioned by Alternative technologies for the reductin of greenhouse gas emis-
the Federal Ministry for the Environment, Nature Conservation, and sions from palm oil mill in Thailand. Environ Sci Technol 47(21):
Nuclear Safety (BMU). The authors would like to express sincere appre- 12417–12425
ciation to the Graduate School, Prince of Songkla University (PSU) and MOE (2015) Alternative Energy Development Plan: AEDP 2015.
the palm oil mills. The authors would like to thank the Research and Ministry of Energy, Bangkok
Development Office, PSU and Associate Professor Dr. Seppo Karrila, MPOB (2010) Overview of the malaysian oil palm industry 2010.
Faculty of Science and Industrial Technology, PSU, for English Official Portal of Malaysian Palm Oil Board, Kuala Lumpur
correction. MTEC (2013) Thai National Database of the saw mill and drying of
hevea wood. National Metal and Materials Technology Center
(MTEC), Pathumthani
National reserved forest Act. B.E, 2507 (1964) BHUMIBOL
References ADULYADEJ, REX, dated April 16, B.E.2507
OAE (2015) Agricultural statistic of Thailand. Office of Agriculture
Chavalparit O, Rulens WH, Mol APJ, Khaodhair S (2006) Options for Economics, Ministry of Agriculture and Cooperative, Bangkok
environmental sustainability for the crude palm oil industry in Palm Woomer CA., Alegre J, Arevalo L, Castilla C, Cordeiro DG, Feigl
Thailand through enhancement of industrial ecosystem. Environ B, Hairiah K, Kotto-Same J, Mendes A, Maukam A, Murdiyarso D,
Dev Sustain 8(2):271–287 Njomgang R, Parton WJ, Ricse A, Rodrigues V, Sitompus SM, Van
Choo YM, Muhamad H, Hashim Z, Subramanium V, Puah CW, Tan Y Noordwijk M (1999). Carbon sequestration and trace gas emissions
(2011) Determination of GHG contributions by subsystems in the in slash-and-burn and alternative land-uses in the Humid Tropics.
oil palm chain using the LCA approach. Int J Life Cycle Assess 16: ACB Climate Change Working Group. Final Report Phase II.
669–681 Nairobi, Kenya
Chuchuoy K, Paengjuntuek W, Usubharatana P, Phungrassami H (2009) Reijnders L, Huijbregts MAJ (2008) Palm oil and the emission of carbon-
Preliminary study of Thailand cabon reduction label: a case study of based greenhouse gases. J Clean Prod 16:477–482
crude palm oil production. Eur J Scientific Res 34(2):252–259 Schmidt JH (2010) Comparative life cycle assessment of rapeseed oil and
palm oil. Int J Life Cycle Assess 12:183–197
DEDE (2006) Best practice guide for eco-fficiency in palm oil industry.
Silalertruksa T, Bonnet S, Gheewala SH (2012) Life cycle sosting
Department of Alternative Energy and Development Energy,
and externalties of palm oil biodiesel in Thailand. J Clean Prod
Bangkok, Thailand
28:225–232
DIT (2016) Monthly prices of oil palm and palm oil. Department of
Silalertruksa T, Gheewala SH (2012) Food, fuel, and climate change is
Industrial Trade, Ministry of Commerce, Nonthaburi, Thailand.
palm-based biodiesel a sustainable option for Thailand. J Ind Ecol
http://agri.dit.go.th/web_dit_sec4/home/index.aspx. Accessed 13
16(4):541–551
Oct 2016
Stichnothe H, Schuchardt F, Rahutomo S (2014) European renewable
DLD (2015) Nutrition Knowledge of Animal Feeds. Department of Live energy directive: critical analysis of important default values and
Stock Development. http://nutrition.dld.go.th/Nutrition_ methods for calculating greenhouse gas (GHG) emissions of palm
Knowlage/requirement/requirement_meat_chicken.htm/. Accessed oil biodiesel. Int J Life Cycle Assess 19:1294–1304
22 May 2015 TGO (2011) Guideline for assessment of the carbon footprint of products.
DOA (2012) Good agricultural practice. Department of Agriculture, Thailand Greenhouse Gas Management Organization (Public
Ministry of Agriculture and Cooperatives, Bangkok, Thailand Organization), Bangkok
GIZ (2011a) GHG emissions calculation guideline for Thai palm oil UNFCC (1998) Kyoto Protocol to the United Nations Framework
industry. Under the project of sustainable palm oil production for Convention on Climate Change. United Nation Framework
bioenergy. OAE and GIZ, Bangkok Convention on Climate Change, Bonn, Germany
GIZ (2011b) GHG emissions optimization guideline for life cycle of the UNFCC (2010) Project Search. United Nations Framework Conversion
palm oil industry. Under the project of developing GHG calculation on Climate Change. Bonn, Germany
methodology for the Thai palm oil industry. Office of OAE and GIZ, Uniliver (2003) Sustainable palm oil: good agricultural practice guide-
Bangkok lines. Ghana
Hansen SB, Olsen SI, Ujang Z (2012) Greenhouse gas reduction through Vijaya S, Ma AN, Choo YM (2010) Capturing biogas: a means to reduce
enhanced use of residues in the life cycle of Malaysian palm oil green house gas emissions for the production of crude palm oil. Am
derived biodiesel. Bioresour Technol 104:358–366 J Geosc 1:1–6

S-ar putea să vă placă și