Research Article

Techno-Economic Analysis of Recombinant Endo-β-1,4-Glucanase Production from

Escherichia coli Eg-RK2 Culture using Oil Palm Empty Fruit Bunch

Misri Gozan1,2, Said Zul Amraini2,5, Muhamad Sahlan1,2, Ega Adi Surya2,
Sonia Limoes2, Siswa Setyahadi3, Young Je Yoo4

1Research Centre for Biomedical Engineering, Universitas Indonesia

2Chemical Engineering Department, Universitas Indonesia, Depok – Indonesia
3Center for Bioindustrial Technology, Agency for Assessment and Application of Technology, Serpong – Indonesia
4Chemical and Engineering Department, Seoul National University
5Chemical Engineering Department, Universitas Riau, Pekanbaru - Indonesia

Correspondence: Misri Gozan, Chemical Engineering Department, Universitas Indonesia, Kampus UI

Depok, 16424 Depok, Indonesia.
E-mail: mgozan@che.ui.ac.id

Abstract
A techno-economic analysis of recombinant cellulase production from E. coli Eg-RK2 was
conducted to support the fulfilling of Indonesia’s energy roadmap for ethanol production. The
plant utilizes OPEFB as a primary substrate in cellulase production, with an expected lifetime of
12 years. The plant is assumed to be built in Indonesia and it will fulfill 1% of the total market
demand. The effect of different pretreatment processes (alkaline, steam explosion, and
sequential acid-alkaline) on the profitability parameter was also studied. A simulation using
SuperPro Designer was used to calculate the mass and energy balance based on the kinetic
parameters of E. coli EgRK2. A technology evaluation showed that alkaline pretreatment
provides the highest yield with no known inhibitors formed. The steam explosion pretreatment
offers the lowest rate of lignin and hemicellulose removal, and it is understood to form known
fermentation inhibitors. The NPVs of the alkaline, steam explosion and sequential acid-alkaline
pretreatments are USD 32,121,000, USD -36,841,000, and USD 384,000, respectively, which
means the alkaline pretreatment is economically very feasible for the production of cellulase.

Keywords: Cellulase production; techno-economic; oil palm empty fruit bunch; SuperPro
Designer; recombinant enzyme
Abbreviations: OPEFB, oil palm empty fruit bunch; TCI, total capital investment; NPV, net present
value; IRR, internal rate of return; ROI, return on investment

1. Introduction

1
 Cellulases are currently the third largest industrial enzyme worldwide, because of their
use in ethanol production, denim processing, animal feed additives, and detergents [1, 2].
However, according to Juniarto, as quoted by BPPT in 2013, 99% of the enzymes used for
industries in Indonesia are imported from other countries. Statistics Indonesia states that the
import value of enzyme in Indonesia is US$ 35 billion in 2016 and will continue to increase with
a rate of increase of 7% per annum. This is unfortunate, as there is enormous potential in
Indonesia’s natural resources that should be able to meet those needs.
One example of Indonesia’s natural resource potential is oil palm. The processing of
palm leads to the formation of several by-products and residues that have economic potential.
OPEFB are solid residue and they have high cellulose content, at 46.77% of OPEFB dry weight
[3]. In this case, the high cellulose content of OPEFB is used as a substrate for bacteria
cultivation to produce cellulase. By using OPEFB as an alternative substrate, the production cost
of cellulase on an industrial scale can be suppressed.
Currently, most cellulase enzymes are produced from saprophytic microorganisms [2],
with fungi from T. reseei being one of the most used in industrial applications [4]. However,
industrial applications often require a cost-effective production of enzymes on a large scale,
which is a common bottleneck. An overexpression of individual cellulose enzymes using
recombinant DNA is one of the common solutions [2]. Escherichia coli and Bacillus are two most
common microbes used to express recombinant proteins [2, 4]. The feasibility of recombinant
cellulase production from E. coli will be evaluated in this study using the SuperPro Designer
Simulator.
Prior to simulating the cellulase plant design, kinetic studies of bacteria used in
cultivation are needed to create an accurate simulation. In this research, kinetic studies of E. coli
BPPTCC-EgRK2 growth are examined with the Monod approach to get the Monod constant (Ks)
and maximum specific growth rate (𝜇𝑚 ).

2. Materials and methods

2.1. Kinetic Study of E. coli BPPTCC-EgRK2
2.1.1. Materials
The microorganism used in this study was recombinant E. coli EgRK2 harboring the
endoglucanase producing genes 16S rDNA isolated from B. Amyloliquefaciens [5]. The culture
was obtained from BPPT (Indonesia’s Agency for Assessment and Application of Technology),
Serpong, Indonesia. The materials used for kinetic studies are listed as follow: Yeast extract;
peptone water, NaCl and D-glucose were purchased at Merck, Indonesia. The main fermentation
was done in Erlenmeyer flask with a capacity of 500 ml. The seed fermentation was done in a
Erlenmeyer flask with a capacity of 25 mL.

2
2.1.2. Culture conditions
Growth and culture conditions were referred to the previous research [6]. 1-2 loopful of E. coli
Eg-RK2 was inoculated into a 100 ml of LB Erlenmeyer flask containing 25 mL of Luria Bertani
medium made from (g/L): yeast extract, 5; peptone, 10; and NaCl, 10. The pH of each medium
was adjusted to 7.0 with 0.1N NaOH. Sterilization was done for 15 min at 121 °C, 1.2 atm. This
seed culture was grown for 6 hours at 37oC with agitation at 150 rpm. The glucose condensed
solution (25 mg glucose + 20 mL H2O) was sterilized for 15 min at 121 °C and was added to the
production medium before inoculation. After 6 h, the seed culture was fed into the 225 mL
production medium, so that the production medium contains 10% inoculum. The cultures were
incubated at 150 rpm at 37oC. Several samples were withdrawn from each culture.

2.1.3. Analytical procedures

Sample was measured every one hour to check the cell growth and glucose concentration. Cell
growth was monitored by measuring the optical density of sample at 600 nm (OD600). The
glucose concentration, Cs (g/L), was assayed using the 3,5-dinitrosalicylic acid method [7]. The
specific growth rate (µ), Monod constant (Ks), and maximum specific growth rate were solved
using the MATLAB 7.6 optimization tool (Mathworks, Natick, MA, USA).

2.2. Process description

The production of recombinant endo-β-1,4-glucanase is based on a kinetic study from this and a
previous study [6] and the process scheme is shown in Fig. 1. OPEFB was used as a raw material
for cellulase enzyme production in this study. OPEFB was collected from a local palm oil mill,
and its characteristics are shown in Table 1. The OPEFB was pretreated with three different
procedures, alkaline [8, 9], steam explosion [10], and sequential acid-alkaline [11]. The
pretreated OPEFB was fermented using E. coli EgRK2. The microbes were obtained from
Indonesia’s Agency for the Assessment and Application of Technology.
Fermentation was conducted in two seed fermenters and one production fermenter operating
in staggered mode. All of the fermenters were regarded as completely mixed reactors. The
seeding fermenter-to-fermenter ratio was 1:40, with fermentation times of 3 h, 4 h, and 10 h,
respectively. The optimum production conditions were obtained from a previous study at pH 7,
agitation at 150 rpm, and a temperature of 37oC, and this condition is known to produce 3.5
IU/mL cellulase [5, 6]. The known cellulase concentration was used to estimate the scale-up
enzyme production yield. The fermentation reaction was:
C6H12O6 + 0.69 C10H14N2O5 + 1.56 NH3 + 0.085 O2 → 12.25 CH1.77O0.49N0.24 + 2.33 H2O + 0.64 CO2

3
The following assumption is used for the mass balance: 1) nitrogen sources come from fish
powder, represented as thymidine; 2) cellulase is an intracellular product and is represented as
E. coli; 3) glucose, cellulase and cell concentration were calculated based on the experimental
data.
The resulting cell was separated from its medium using centrifugation to concentrate
the cell before it is disrupted. The cell was disrupted to release its cellulase using a high-
pressure homogenizer at 800 bar with three passes. Cellulase was then separated from cell
debris using centrifugation. Crude cellulase was then stabilized using maltodextrin [12] and
spray dried to produce a crude enzyme powder. The spray-dried product was expected to have
an activity loss of up to 37%.

2.3. Economic analysis

The mass and energy balance were calculated from SuperPro Designer 9.0 based on the
described process. The same simulation was also used to estimate various economic parameters
using data, such as equipment purchase cost, direct cost, and indirect cost. The resulting data
would be used to assess the feasibility of recombinant cellulase production from OPEFB using E.
coli EgRK2.

4
Fig. 1. A) Alkaline pretreatment, B) steam explosion pretreatment, C) sequential acid-alkaline pretreatment and D) recombinant endo-β-1,4-

glucanase production scheme.

5
2.3.1. TCI
Two costs were used to evaluate the economic parameters of the plant, total capital
investment (TCI), and annual operating cost. TCI was calculated based on the main equipment
purchase cost, and the equipment cost was estimated using an equation from Seider et al. [13]
and a graph from Petrides [14]. The additional direct/indirect costs were estimated from the
equipment purchase cost [14].

2.3.2. Operating Cost

The annual operating cost was estimated from the raw materials cost, labor-dependent
cost, facility-dependent cost, utilities, and depreciation cost. Depreciation was calculated using a
declining balance method for 12 years. The revenue used for the economic evaluation comes
from cellulase sales, where the selling price for cellulase was USD 250/kg or 1.8×10-5 USD/IU.
3. Results
3.1. Kinetic study of cellulase production

2.5 0.110

2
0.085

Glucose concentration (g/L)

1.5
OD600

0.060
1

0.035
0.5

0 0.010
0 1 2 3 4 5 6 7 8 9 10
Time (h)
OD600 Glucose concentration

Fig 2. Concentration of cell and glucose during cell culture. Glucose concentration curve (solid
square) and cell concentration curve (solid circle)

Kinetic parameters of microbial growth were evaluated in this study. The growth curve
shown in Fig. 2 was used to calculate Monod’s kinetic parameter of cell growth, μmax and Ks. It
was observed that the microbes have a relatively short lag phase period, from 0 h to 1 h.
Parameters were estimated with optimization tool on Matlab 7.6 and it was found that the μmax
and Ks value were 0.792 h-1 and 0.165 g.L-1.

6
3.2. Technology comparison of cellulase production
To evaluate the most suitable cellulase production process, the yield, batch duration,
number of units, pretreatment performance, and economic parameters were compared. As
quoted from Indonesia’s Ministry of Energy and Mineral Resources, the alkaline pretreatment
process was able to fulfill 1% of Indonesia’s market share using 8.5 tons of OPEFB as a
substrate, smaller than the acid-alkaline and steam explosion pretreatments, which required 19
tons and 29 tons of OPEFB, respectively. The alkaline pretreatment also gives the highest
product yield, 0.03, compared with acid-alkaline and steam explosion, which give yields of 0.021
and 0.008, respectively.
As for delignification efficiency, the sequential acid-alkaline pretreatment gives the
highest lignin removal rate, removing 90% of lignin from OPEFB [11]. Steam explosion did not
remove any lignin from OPEFB but offers high hemicellulose removal, which increases its
permeability and increases enzymatic digestibility [10]. However, this process also produces
fermentation inhibitors, such as levulinic acid (0.11 g/L), formic acid (0.21 g/L), hidroxymethyl
furfural (0.71 g/L), and furfural (0.23 g/L) [10]. Similar results were also achieved by another
study by Duangwang in 2016 [15]. The performances of each simulation are shown in Table 1.

3.2.1. TCI and operating cost

The TCI was calculated based on the purchase and installation cost of each piece of
equipment, piping, instrumentation, facility-dependent cost, engineering, and construction fee
[16]. Table 1 shows the TCI and annual operating cost of each simulation. The main equipment
purchasing cost of each piece of equipment was USD 5,213,000, USD 7,948,000, and USD
12,908,000 for the alkaline, steam explosion, and acid-alkaline pretreatments, respectively.
The comparison of each equipment cost was shown in Table 2. The annual operating cost of
recombinant cellulase production is highly influenced by three different costs: raw materials
cost, labor-dependent cost, and facility-dependent cost. These costs were simulated using
SuperPro Designer and they are shown in Table 1. The steam explosion pretreatment had the
highest operating cost, at USD 36,156,000, with 44.48% spent on labor-dependent costs.

7
Table 1. Simulation summary of different pretreatment method effect on project profitablility.

Pretreatment Method
Parameter Alkaline Steam Explosion Acid-Alkaline
(CHEMEX) [8, 9] [10] [11]
OPEFB feed/batch (Ton) 8.5 29 19
Product/batch (kg) 255 231.36 394.84
Number of batches/year 585 589 428
Yield p/s 0.03 0.008 0.021
Number of units 18 20 21
Fermentation inhibitor - Yes -
Removal efficiency
-Cellulose 19% 30% 41%
-Hemicellulose 67% 82% 86%
-Lignin 80% 0% 90%
Batch duration (h) 81.1 174.08 85.71
Minimum cycle time (h) 13.42 12 18.33
TCI (USD) 35,165,000 53,335,000 85,493,000
Operating Cost (USD) 23,408,000 36,156,000 34,848,000
IRR 31.64% N/A 7.11%

NPV (9.6%) (USD) 32,121,000 - 36,841,000 384,000

ROI 37.26 8.02 14.06

Payback period 3.29 12.48 7.11

8
 Table 2. Summary of estimated equipment cost

Alkaline Pretreatment Steam Explosion Pretreatment Acid-Alkaline Pretreatment

No Unit
Specification F.O.B. Price (USD) Specification F.O.B. Price (USD) Specification F.O.B. Price (USD)
1 Crusher 11.33 MT/h 31,000 29 MT/h 82,000 19 MT/h 49,000
2 Hammer mill 8.5 MT/h 32,000 29 MT/h 71,000 19 MT/h 59,000
3 Conveyor 150 ft 75,000 150 ft 75,000 150 ft 75,000
4 Alkaline delignification 46 m3 154,000 181.98 m3 329,000
5 Neutralization tank 46 m3 68,000
6 Steam explosion tank 29.63 m3 235,000
7 Gas cyclone 15.7 ft3/min 3,000
8 Condenser 103.94 ft2 16,000
9 Washing tank 76.68 m3 68,000
10 Acid pretreatment 214.6 m3 333,000
11 Rotary vacuum filter 684 ft2 558,000 456 ft2 312,000 912 ft2 653,000
12 Media preparation 73.18 m3 107,000 92.26 m3 123,000 73.38 m3 116,000
13 Autoclave 73.18 m3 148,000 92.26 m3 312,000 73.38 m3 277,000
14 Seed fermenter 1 42 L 2,000 52 L 2,000 41 L 2,000
15 Seed fermenter 2 2.05 m3 20,000 2.59 m3 22,000 2.06 m3 20,000
16 Production fermenter 82.27 m3 179,000 104 m3 225,000 82.52 m3 180,000
17 Plate and frame filtration 30.32 m2 153,000 54.13 m2 206,000 20.01 m2 123,000
18 Centrifuge 1 15,954 L/h 588,000 16,771.98 L/h 604,000 5,560 L/h 440,000
19 High pressure homogenizer 4,973 L/h 71,000 4,705 L/h 69,000 5,590.5 L/h 123,000
20 Centrifuge 2 13,191 L/h 313,000 4,681 L/h 303,000 4,021 L/h 269,000
21 Stabilization tank 7.3 m3 6,000 6.4 m3 6,000 6.5 m3 6,000
22 Spray dryer 4.93 m3 771,000 4.3 m3 716,000 4.79 m3 777,000

9
3.2.2. Profitability parameters
The plant’s revenue came solely from cellulase sales. The annual cellulase production
was 154 tons (1% market share), with a revenue of USD 37,300,000. The product was sold at
USD 250/kg, or 1.8×10-5 USD/IU. The price was determined from the prices of other similar
products available on the market. Net present value (NPV) was the key to determining the
feasibility of a project. Simulations show that the alkaline pretreatment gives the highest NPV, at
USD 32,121,000, calculated at a discounted rate of 9.6% and a project lifetime of 15 years. The
profitability parameters of other processes can be seen in Table 1.

3.3. Sensitivity Analysis

Project feasibility is also evaluated based on its sensitivity to key variables, such as
selling price, raw materials cost, and labor cost. A project that is highly affected by external
factors is considered more risky. Figure 3 shows the effect of changes in labor cost, raw
materials cost, selling price, and overall operating cost on NPV value. It was found that the
cellulase selling price has the highest influence on NPV. Both the labor and raw materials costs
have similar effects on the project’s NPV. However, the fluctuating costs of these two
parameters do not affect the project as much as the selling price, showing less risk related to
these parameters.

70

60
NPV (Milion USD)

50

40

30

20

10

-
-25% -20% -15% -10% -5% 0% 5% 10% 15% 20% 25%
Price Sensitivity

Cellulase Price Raw Material Cost

Labor Cost Overall Operating Cost

Fig. 3. Sensitivity analysis of recombinant cellulase production from OPEFB

10
4. Discussion
4.1. Kinetic parameters of E. coli EgRK2
It was observed that E. coli EgRK2 has a relatively short lag phase period; from 0 h to 1
h. E. coli EgRK2 also has higher growth rate than the wild type, Bacillus sp. RK2 [6]. This result is
consistent with previous study that show increase in cellulase production from the wild type
[6]. The overexpressed cellulase was produced intracellularly, resulting in linear correlation
between cell concentration and cellulase activity, shown by the same increasing trend in both
Monod parameters and Michaelis-Menten parameters [6]. Higher growth rate value results in
shorter fermentation time, resulting in higher productivity which overcomes the downside of
enzyme produce intracellularly. The kinetic data was used to model the fermentation process
and scaled up using SuperPro Designer simulator.
It shows that E. coli EgRK2 has much higher growth rate than other cellulase
microorganism producers. This result is expectable due to much higher growth rate in bacteria,
while the other cellulase producers are mostly fungi [17]. To our knowledge, literature working
with Monod equation in cellulase producing E. coli is novel.

4.2. Techno-economic analysis

The result shows that both steam explosion pretreatment process and sequential acid-
alkaline process had higher capital investment then the alkaline process. A high capital
investment in steam explosion and sequential acid-alkaline pretreatments was due to its
number of units. Steam pretreatment has 20 different units, with nine sacharification tanks, one
fermenter, and one stabilization tank standing by in staggered mode. Similarly, sequential acid-
alkaline pretreatment uses 21 units with multiple rotary vacuum filtration units, resulting in a
higher capital investment in equipment and installation costs. Moreover, both the steam
explosion and sequential acid-alkaline pretreatments require more space for their equipments,
affecting the land and building costs of the plant.
The data also show that steam explosion process gives the highest operating cost. The
cost is spent mostly on labor-dependent cost, affected by the high number of units installed and
longer process time. The steam explosion has a 5 days saccharification process in 10 different
saccharification tanks, resulting in higher number of operator needed. Sequential acid-alkaline
process also give higher operating cost then alkaline pretreatment due to higher energy demand
for both acid and alkaline processes in longer period of time than the alkaline process.
From the revenue wise, the alkaline pretreatment also give the highest NPV value,
showing the investment value in present worth manners. An economically feasible project
should give positive NPV value. Sequential acid-alkaline pretreatment gives a positive NPV value
but has a payback period more than half of the plant operating time with IRR value smaller than

11
the discounted rate used in NPV calculation. High annual operating cost in steam explosion
pretreatment and sequential acid-alkaline pretreatment are reducing the annual profit,
resulting in lower rate of return for the investors. A longer project lifetime or lower operating
cost are required to make both steam explosion pretreatment and sequential acid-alkaline
pretreatment more economically feasible.
This study shows that alkaline pretreatment is the most feasible to produce
recombinant cellulase from OPEFB. It has the lowest capital investment and annual operating
costs, and offer highest investment return, with NPV, IRR, and payback period of 32,121,000
USD, 31.64% and 3.29 years, respectively. These are financially promising figures as it offers
higher investment return than Indonesia’s market rate of return.
Sensitivity analysis shows that the cellulase selling price has the highest influence on
NPV, affecting this value more than other parameters, as it is the only revenue stream at the
plant. However, as this project was intended to fulfill the Indonesian government’s renewable
energy target, a steady demand is expected in the near future. Further assessment in cellulase
purification and stabilization is recommended to find a more cost effective production method
in the future. Financial figures might be different for other regions. Nevertheless, this study
offers an advisable background.

Acknowledgments
The authors are grateful for the financial support from Universitas Indonesia under
PITTA 2017 grant. The authors also gratefully acknowledge Center of Bioindustrial Technology,
Agency for Assessment and Application of Technology (BPPT), Indonesia for providing the E.coli
BPPTCC-EgRK2.

Conflict of interest
The authors declare no financial or commercial conflict of interest.

5 References
[1] Wilson, D. B. Cellulases and biofuels. Current Opinion in Biotechnology, 2009, 20(3), 295-
299.
[2] Juturu, V. & Jin, C. W. Microbial Cellulases: Engineering, production and applications.
Renewable and Sustainable Energy Reviews, 2014, 33, 188-203.
[3] Piarpuzán, D., Quintero, J. A., Cardona, C. A. Empty fruit bunches from oil palm as a
potential raw material for fuel ethanol production, Biomass and Bioenergy, 2011, 35(3), 1130-
1137.
[4] Hasunuma, T. et al. A review of enzymes and microbes for lignocellulosic biorefinery and
the possibility of their application to consolidated bioprocessing technology. Bioresource
Technology, 2013, 513-522.
[5] Setiawan, R., 2014. Kloning dan Ekspresi Endo-B-1,4-Glukanase dari Isolat Lokal BPPTCC-
RK2 (in Bahasa). Master Thesis. Depok: Universitas Indonesia.

12
[6] Amraini S.Z, Ariyani, LP., Setyahadi, S., Hermansyah, H., Gozan, M., et al. Production and
Characterization of Cellulase from E. coli EgRK2 Recombinant based on Oil Palm Empty Fruit
Bunches. Biotechnology and Bioprocess Engineering, 2017.
[7] Miller, G. L. Use of dinitrosalicylic acid reagent for determination of reducing sugar.
Analytical Chemistry, 1959, 426-428.
[8] Sudiyani, Y. et al. Utilization of biomass waste empty fruit bunch fiber of palm oil for
bioethanol production using pilot - scale unit. Energy Procedia, 2013, 32, 31-38.
[9] Jeon, H. et al., 2014. Production of anhydrous ethanol using oil palm empty fruit bunch in a
pilot plant. Biomass & Bioenergy, 67, 99-107.
[10] Medina, J. D. C. et al. Steam explosion pre-treatment of oil palm empty fruit bunches (EFB)
using autocatalytic hydrolysis: A biorefinery approach. Bioresource Technology, 2016, 199, 173-
180.
[11] Kim, S., Park, J. M., Seo, J.-W. & Kim, C. H. Sequential acid/alkali-pretreatment of empty
palm fruit bunch fiber. Bioresource Technology, 2012, 229-233.
[12] Belghith, H., Chaabouni, E., Gargouri, A. Biostoning of denims by Penicillium occitanis
(Pol6) cellulases. Journal Biotechnol, 2001. 89(2-3), 257-262.
[13] Seider, W. D., Seader, J. D. & Lewin, D. R. Product and Process Design Principles: Synthesis,
Analysis, and Evaluation. John Wiley and Sons, Inc., New York, 2003.
[14] Petrides, D. Bioprocess Design and Economics, in: Bioseparations Science and Engineering,
Oxford University Press, Oxford, 2002, 441-510.
[15] Duangwang, S. et al. Pilot-scale steam explosion for xylose production from oil palm
empty fruit bunches and the use of xylose for ethanol production. Bioresource Technology,
2016, 203, 252-258.
[16] Han, W. et al. Techno-economic analysis of dark fermentative hydrogen production from
molasses in a continuous mixed immobilized sludge reactor. Journal of Cleaner Production,
2016, 567-572.
[17] Kirchman, D.L. Microbial growth biomass production and controls, in: Processes in
Microbial Ecology. Oxford University Press Inc., New York, 2012, 99-116.

13