Process simulation and economical evaluation of enzymatic biodiesel production plant

Bioresource Technology 101 (2010) 5266–5274

Contents lists available at ScienceDirect

Bioresource Technology journal homepage: www.elsevier.com/locate/biortech

Process simulation and economical evaluation of enzymatic biodiesel production plant Lene Fjerbaek Sotoft *, Ben-Guang Rong, Knud V. Christensen, Birgir Norddahl Institute of Chemical Engineering, Biotechnology and Environmental Technology, University of Southern Denmark, Campusvej 55, DK-5230 Odense M, Denmark

a r t i c l e

i n f o

Article history: Received 27 November 2009 Received in revised form 25 January 2010 Accepted 26 January 2010 Available online 19 February 2010 Keywords: Enzyme Biodiesel Process simulation Process design Process economy

a b s t r a c t Process simulation and economical evaluation of an enzymatic biodiesel production plant has been carried out. Enzymatic biodiesel production from high quality rapeseed oil and methanol has been investigated for solvent free and cosolvent production processes. Several scenarios have been investigated with different production scales (8 and 200 mio. kg biodiesel/year) and enzyme price. The cosolvent production process is found to be most expensive and is not a viable choice, while the solvent free process is viable for the larger scale production of 200 mio. kg biodiesel/year with the current enzyme price. With the suggested enzyme price of the future, both the small and large scale solvent free production proved viable. The product price was estimated to be 0.73–1.49€/kg biodiesel with the current enzyme price and 0.05–0.75€/kg with the enzyme price of the future for solvent free process. Ó 2010 Elsevier Ltd. All rights reserved.

1. Introduction The major operation in biodiesel production that decides the process route is the transesterification of the vegetable oil or animal fat into fatty acid methyl esters (FAME), the primary product. Typically methanol is used for the transesterification producing methyl esters and glycerin as byproduct (see Eq. (1)). This reaction can be carried out by various forms of catalysts, but so far industrially only chemical homogeneous catalysts are used in large scale. Smaller pilot plants with enzymatic catalysts are reported, but not to a very large scale (Du et al., 2008). In order to evaluate the potential and remaining obstacles of introducing enzymes as the preferred industrial catalyst, analysis of economical as well as environmental impacts of the alternative processes must be evaluated compared to the conventional process. The economy of the traditional production method has been thoroughly analyzed as by Haas et al. (2006), though the raw material price has increased since the publication of Haas et al. (2006). The raw material price based on 0.52 US$/kg soybean oil accounted for 88% of the production costs in a 37.900 m3 biodiesel/year continuous production plant (Haas et al., 2006).

* Corresponding author. Tel.: +45 6550 7443; fax: +45 6550 7354. E-mail address: [email protected] (L.F. Sotoft). 0960-8524/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.biortech.2010.01.130

ð1Þ The industrial production of biodiesel has had a very turbulent lifetime due to the changes in prices of raw materials and fossil fuels as well as regulatory changes and production capacity of biodiesel. All of this affects the process economy on a global scale. When looking at the sustainability of producing biodiesel, this has been questioned in particular with respect to virgin oils as raw materials (Reijnders and Huijbregts, 2008). Nevertheless, if biodiesel is to be produced, an industrial process must be able to produce a product that meets the specifications, i.e., for Europe (EN 14214, 2008). Traditional homogeneous and alternative heterogeneous chemical catalysts and supercritical conditions have been evaluated with regard to process economy for a continuous and batch processes (Table 1). Batch process evaluation has also been carried out by Sakai et al. (2009) and Noordam and Withers (1996) including homogeneous and heterogeneous chemical catalysts and various

5267

L.F. Sotoft et al. / Bioresource Technology 101 (2010) 5266–5274 Table 1 Comparison of previous studies of process economy of biodiesel production.

Sakai et al. (2009) Noordam and Withers (1996)a Marchetti and Errazu (2008) West et al. (2008) Zhang et al. (2003b) Nelson et al. (1994) Bender (1999) You et al. (2008) Haas et al. (2006) Zhang et al. (2003b) Bender (1999) Santana et al. (in press) van Kasteren and Nisworo (2007)a West et al. (2008)a a

Production size (mio. kg/year)

Operation mode

Raw material

Glycerol sales included

Product price (€/kg)

7.26 7.8 36.04 8 8 100 101.20 8–100 33.31 8 1.76 10.5 8–125 8

Batch Batch Continuous Continuous Continuous Continuous Continuous Continuous Continuous Continuous Continuous Continuous Continuous Continuous

Waste cooking oil Canola oil Waste cooking oil Waste cooking oil Waste cooking oil Beef tallow Animal fats Soybean oil Soybean oil Virgin vegetable oil Canola oil Castor oil Waste cooking oil Waste cooking oil

No Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes Yes

0.18–0.19 0.55 0.32–0.33 0.14 0.46 0.25 0.26 0.49–0.62 0.38 0.62 0.33 1.56 0.15 0.66

Supercritical process, no other catalysts.

production ranges as well as two different product purification methods. Batch processes are only relevant compared with existing batch processing plants’ economy, since if biodiesel is to be produced viable as a ‘‘true” bulk chemical then continuous operation is the only realistic option (Seider et al., 2004). Super critical conditions eliminate the need of catalyst and increase reaction rate (Saka and Kusdiana, 2001), but is a very costly production method. For continuous processes, the prices for produced biodiesel were 0.14–0.46 and 0.33–0.62€/kg when using waste cooking oils, etc. and virgin vegetable oils, respectively (see Table 1). Batch production has been estimated to give biodiesel at 0.18–0.19 and 0.55€/kg for waste cooking oil and canola oil, respectively (see Table 1). This is for as well homogeneously as heterogeneously catalyzed processes. Most studies include a glycerol credit with a price/ kg glycerol depending on the quality of the glycerol. Therefore, this study also includes this for comparative purposes. Several processes of theoretical industrial scales have been simulated, but generally lack incorporation of realistic data and actual industrial performance for mass and energy balances. Therefore, some of the simulation results, i.e., water consumption and waste fractions are unrealistically low. As an example, Zhang et al. (2003a) have done a simulation work for four cases, including alkaline catalysis with NaOH of virgin oils and waste cooking oil as well as acid catalysis with either traditional water washing or hexane extraction of methyl esters. The simulation included a liquid–liquid extraction (water washing) of about 1177.20 kg/h biodiesel with 11 kg/h of water, which seems an unrealistic ratio of water to biodiesel when compared to real industrial unit operation in biodiesel production. Waste water from a biodiesel production can be as much as 47.5 kg for production of 100 kg biodiesel (Daka Biodiesel, 2009). The former simulation results leads to an underestimation of the total water consumption. A more realistic mass ratio of water to oil of 1:1 is, i.e., used by Santana et al. (in press) for a process catalyzed by NaOH with castor oil and ethanol as raw materials. The latter simulation gives more reliable results with respect to water consumption and liquid–liquid extractor size and price. Some industrial processes use KOH as catalyst instead of NaOH. Where the cheaper NaOH gives water soluble salts after neutralization with acid, the more expensive KOH gives a precipitate of potassium sulphate and phosphate when neutralizing with sulphuric and phosphoric acid, respectively. The precipitation of salts reduces the total water consumption, because the water can be recirculated to an extent which is impossible when using NaOH. The precipitated salt can furthermore be sold as fertilizer. This is the technology used industrially by BDI – Biodiesel International. Only the use of KOH as by BDI can justify low total water consumption for the whole process due to water recycle, while the use of NaOH (as by Zhang et al. (2003a)) as a catalyst cannot. In both

cases, the water consumption of the liquid–liquid extraction step isolated is of the same magnitude, but the potential water for recycling is very different. None of the references in Table 1 involves biodiesel production catalyzed by enzymes. Enzymes are more expensive and slower reacting than traditional chemical catalysts, but give a much easier and simpler biodiesel purification. A life cycle comparison including biocatalysts has been carried out by Harding et al. (2008), but the analysis need optimization on several points. Water washing is not needed when using enzymes, but still included in the study. The washing step is only used when chemical salt catalyst residues, i.e., sodium ions, must be removed from the biodiesel. This is therefore a redundant step when using enzymes. The prospect of the present paper is to bring biodiesel production with enzymes as catalysts closer to industrial scale application and elucidate the main obstacles that need to be solved for the process to be economically and environmentally sustainable. Therefore, the paper evaluates several important and highly relevant scenarios for enzyme catalyzed biodiesel production processes. Simulations with methanol and solvent-free/co-solvent operations are carried out to investigate how this affect enzyme performance and process design and to elucidate what effect this has on the process economy. That is, too high concentrations of methanol inhibit the enzymes and reduce their lifetime. Therefore, optimal feeding of the methanol is crucial within the process. The process must be operated with maximum 1 mol methanol per

Table 2 Scenarios and data used for process simulation. Scenario Alcohol Cosolvent Enzyme pricea (US$/kg) Data provided by

Productivity (Fjerbaek et al., 2009) (kg biodiesel/kg enzyme)b Yield and reaction time Production size (tons/year) a

1 Methanol No 1000/10 (762.71/ 7.627€) Shimada et al. (1999) 1200

2 Methanol No 1000/10

3 Methanol tert-Butanol 1000/10

4 Methanol tert-Butanol 1000/10

Shimada et al. (1999) 1200

Li et al. (2006)

Li et al. (2006)

4250

4250

>96% and 48 h 8000

>96% and 48 h 200,000

95% and 12 h

95% and 12 h

8000

200,000

The enzyme prices equal 762.71 and 7.627€/kg, respectively. Productivity calculated from 95% yield, 54 cycles and 4 wt.% enzyme out of reaction mixture. b

5268

L.F. Sotoft et al. / Bioresource Technology 101 (2010) 5266–5274

mole triglyceride and addition of methanol several times to reach full conversion. A way to minimize methanol inhibition and increase enzyme lifetime and mass transfer is to use a cosolvent as tert-butanol together with the raw materials oil and methanol. Simulations with tert-butanol as a cosolvent is also included to determine what effect the increase in process volume, the need of solvent recovery and higher enzyme life time has to the overall process economy. Finally, this paper includes two scenarios with the current price (762.71€/kg enzyme, Fjerbaek et al., 2009) and a more attractive price of enzymes in the future (7.627€/kg enzyme, Mittelbach, 2005) as well as two scenarios with respect to enzyme life time without solvent (1200 kg biodiesel/kg enzyme) and with solvent (4250 kg biodiesel/kg enzyme), see Table 2. 2. Process design As discussed earlier, it is the catalyst employed in the transesterification that determines the process routes for the biodiesel production. To illustrate the differences of process structures when exchanging traditional homogeneous chemical catalysts with heterogeneous biocatalysts, the process diagram for a traditional biodiesel production with KOH and downstream processing is shown as Fig. 1. The production of biodiesel by the process is carried out in both batch and continuous mode industrially. Alcohol and catalyst are mixed initially before entering a tank reactor together with oil. Alcohol is added in stoichiometric surplus. After reaction to a specified extent, the product mixture is sent to a distillation tower for removal and recycle of excess methanol. The bottom product is then washed with water in a liquid–liquid extraction tower. Polar substances as glycerol, salts and residual substances are then neutralized with acid in the water phase and separated by distillation into water for recycle, glycerol and solid fertilizer for sale plus a waste fraction. The fertilizer is only produced in plants using KOH as catalyst and H3PO4 for neutralization. The non-polar product phase is purified by distillation to meet specifications. If a high quality raw material is used, only water is removed as the top product, while the bottom product is the biodiesel product. If lower quality raw materials like used cooking oils or animal fats are used, then the biodiesel also needs distillation and is taken out as an additional top product. This step depends on reaction yield and product specifications. Modification of the design is carried out for raw materials containing much free fatty acids. In this case, an esterification step catalyzed by acids is introduced before the caustic transesterification reaction.

For enzymatic industrial biodiesel production, the process design is very different from the traditional setup. Enzymes and oil are mixed in the reactor after which alcohol is added. Due to inhibition of alcohol to the enzymes, the alcohol is added stepwise in stoichiometric deficit, i.e., three stepwise additions of 1 mol alcohol per mole triglyceride. The downstream processing is then a filter to retain and recirculate the enzymes followed by a decanter to separate the non-polar and polar phases. Excess methanol is again removed by distillation from the polar phase and recycled. The polar glycerol as the bottom product has a very high purity, since it is not mixed with water. The non-polar phase is distilled due to the lower yield of the enzymatic process at the time being compared to the traditionally catalyzed processes in order to meet specifications. The perfect scenario for enzymes would be no methanol inhibition and 100% yield, which would make the product distillation step dispensable and considerable reduce energy requirements. Methanol inhibition can be diminished using tert-butanol as a cosolvent. This increases enzyme lifetime and reaction rate, but introduces an additional unit operation due to solvent recovery by distillation. An advantage though is that the inhibition is diminished, so that all methanol can be added in stoichiometric amounts in one reactor. This greatly simplifies the reaction system. Overall, it is important to notice that the use of enzymes gives some favorable changes as to the required number of unit operations, energy consumption, waste production and glycerol product quality. Table 3 shows the needed unit operations or equipment when producing biodiesel with traditional catalysts and enzymes. The latter both with and without a cosolvent. Three reactors and decanters in series are required for the solvent free enzymatic process compared to one reactor and decanter for as well the traditional as the cosolvent enzymatic process. Distillation for methanol recovery and biodiesel purification is needed for all three process types, while no washing is required for any enzymatic process. The difference between the solvent free and cosolvent enzymatic processes is the need for solvent recovery by distillation for the latter process type. These requirements will influence the flow sheet, process simulation and economical analysis for each process type. In the following section, a more detailed description of the simulations carried out is presented. 3. Process simulation Several scenarios are simulated for investigation, all aiming at industrial scale enzymatic biodiesel productions regarding process design and economy. Two production sizes are chosen for evaluation. A yearly production of 8 mio. kg biodiesel is chosen for

Fig. 1. Process flow diagram for traditional biodiesel production with homogeneous KOH catalyst.

5269

L.F. Sotoft et al. / Bioresource Technology 101 (2010) 5266–5274 Table 3 Unit operations/equipment needed (major) for different biodiesel production process types. Process type Catalyst Solvent

Traditional Homogeneous chemical No

Enzyme solventfree Heterogeneous enzyme No

Enzyme cosolvent Heterogeneous enzyme tert-Butanol

Yes Yes Yes Yes No

Yes Yes Yes Yes No

Yes Yes Yes Yes Yes

Process Tank reactor and number Decanter and number Tank reactor esterification if high in FFA Decanter/centrifuge after esterification if high in FFA

Yes, 1 Yes, 1 Yes Yes

Yes, 3 Yes, 3 No No

Yes, 1 Yes, 1 No No

Purification Distillation – methanol recovery Washing (L–L extraction) Distillation – water recovery Neutralization Drying – for technical grade glycerol Distillation – for pharma. grade glycerol Salt wash and drying Distillation – FAME Distillation – solvent recovery

Yes Yes Yes Yes Yes Yes Yes In some cases No

Yes No No No No No No Yes, at present No

Yes No No No No No No Yes, at present Yes

Storage Storage Storage Storage Storage Storage

tank tank tank tank tank

biodiesel glycerol MeOH oil tert-butanol

Fig. 2. Simulation process and STream data of the process for 1000 kg/h solvent free production of biodiesel with enzymes in three continuous stirred tank reactors in series.

5270

L.F. Sotoft et al. / Bioresource Technology 101 (2010) 5266–5274

comparison reasons to existing simulations, see Table 1. A significantly larger production scale of 200 mio. kg biodiesel/year is also chosen in order to compare with a Danish biodiesel producer expanding to this yearly production capacity. Four scenarios are selected as relevant for simulation and investigation. They combine variations in production size and cosolvent/ solvent free production. Each scenario has further been analyzed with respect to two enzyme prices; the current market price and a potential significantly lower price in the future (Mittelbach, 2005). The scenarios are presented in Table 2. The background to the experimental reaction data and enzyme performance is taken from existing literature, while the current prices of enzymes are taken from Fjerbaek et al. (2009). The model substrate is based on oleic acid as the overall fatty acid structure, where rapeseed oil is modeled as trioleate and biodiesel as methyl oleate. The substrate is chosen not to contain water or free fatty acids. The raw material is assumed to have 0.3‰ nonreacting

material based on information about the nonreacting fraction from an existing rapeseed biodiesel plant (Emmelev Molle, 2009). The process simulations are carried out in Aspen Plus 2006.5 and Aspen Icarus Process Evaluator 2006.5 together with the method of Peters et al. (2003). A UNIFAC-DMD thermodynamic model is used for estimation of activity coefficients, as it has been found to provide good fit between estimated and measured methanol–biodiesel and methanol–glycerol vapor–liquid equilibrium data (Kuramochi et al., 2009). In Aspen Plus 2006.5, flowsheets are constructed and simulations are carried out with satisfying results for streams and unit operations. The flowsheets for solvent free and cosolvent simulations with stream data for 8 mio. kg biodiesel/year can be seen as Figs. 2 and 3. Solvent free production must be carried out in less than stoichiometric ratio of oil and methanol preferably three times 1 mol methanol per mole of oil. Therefore, the setup is constructed as

Fig. 3. Simulation process and STream data of the process for 1000 kg/h production of biodiesel with enzymes and cosolvent tert-butanol in one continuous stirred tank reactor.

5271

L.F. Sotoft et al. / Bioresource Technology 101 (2010) 5266–5274

three continuous stirred tank reactors in series (reactors 1–3). After each reactor, a decanter is needed to remove formed glycerol to avoid potential inhibition of enzymes by glycerol and increase the overall yield (decanters 1–3). The biodiesel product is distilled (Estcol) to ensure high product quality by removal of unreacted material as the remnant and biodiesel as the distillate of 99.8 wt.% methyl oleate. The polar phase containing methanol and glycerol is separated in a distillation column (Glycol). The methanol is recycled back as a feed stream, and the glycerol with a purity of 99 wt.% can be sold as byproduct. No purge stream of methanol has been found necessary, since there is no knowledge of any accumulation of polar substances in this phase.

The cosolvent production process is setup with one continuous stirred tank reactor (reactor), since the presence of tert-butanol diminishes the inhibition of the enzymes if methanol is present in a stoichiometric ratio. The solvent and unreacted methanol is then removed from the product phase by distillation (Solcol), where the distillate (ptertbut) is split into a recycle and a purge stream. A decanter is needed after the solvent recovery column for separation of the polar glycerol phase from the product and unreacted oil phase (decanter). The biodiesel phase is then distilled (Estcol) to ensure high product quality by removal of unreacted material as the remnant and biodiesel as the distillate of 99.7 wt.% methyl oleate.

Table 4 Equipment costs and corrected total capital investment. Scenario

1

2

3

4

Part Storage Storage Storage Storage Storage Storage

8,000,000 kg/year Solvent free Cost (€) 102,700 25,000 24,200 80,200 –

200,000,000 kg/year Solvent free Cost (€) 1,029,400 155,400 155,400 669,200 –

8,000,000 kg/year cosolvent Cost (€) 108,900 24,900 24,000 80,200 10,200

200,000,000 kg/year cosolvent Cost (€) 1,122,900 155,400 150,500 669,200 21,900

Process P1 P2 P3 P4 P5 P6 P7 H1 Reactor 1 Reactor 2 Reactor 3 Decante 1 Decante 2 Decante 3

4340 4120 4120 4120 3670 4340 4340 77,200 260,010 237,240 237,240 13,050 13,050 13,050

4230 4340 4340 4340 7400 7300 7300 32,000 1,614,510 1,512,810 1,480,320 19,800 19,800 18,810

4340 4120 4000 9000 – – – 52,300 328,320 – – 32,850 – –

7300 4560 4120 8300 – – – 21,800 2,306,250 – – 284,580 – –

Purification P8 P9 P10 P11 P12 P13 P14 H2 H3 H4 H5 Estcol-tower Estcol-cond Estcol-cond acc Estcol-reflux pump Estcol-reb Glycol-tower Glycol-cond Glycol-cond acc Glycol-reflux pump Glycol-reb Solcol-tower Solcol-cond Solcol-cond acc Solcol-reflux pump Solcol-reb

4340 4120 4120 4120 4120 4120 4000 52,300 – – – 72,800 15,900 7020 4340 83,500 18,700 20,700 2205 4000 20,400 – – – – –

7200 3560 4230 4230 4230 4560 4120 21,000 – – – 504,700 29,700 17,010 8300 1,090,000 28,200 20,700 2205 4120 34,300 – – – – –

4340 4340 4120 4120 4000 4120 – 53,500 53,500 95,500 24,300 72,400 18,400 7020 4340 31,700 18,000 20,700 2205 4000 20,600 54,500 16,200 43,200 4450 34,300

8800 7200 7200 4560 4230 4340 – 21,700 20,900 30,900 63,900 495,700 24,200 18,450 8500 382,700 24,000 16,200 2205 4000 31,400 383,800 77,300 366,120 10,000 590,700

Total plant equipment cost Total direct cost Fixed capital investment Total capital investment (TCI)

1,442,795 2,861,600 8,143,873 8,551,067

8,531,865 11,891,200 18,337,200 19,110,157

1,286,985 3,100,300 9,703,210 10,188,370

7,419,315 13,592,900 25,868,023 27,161,424

tank tank tank tank tank

biodiesel glycerol MeOH oil tert-butanol

5272

L.F. Sotoft et al. / Bioresource Technology 101 (2010) 5266–5274

Again, the polar phase containing methanol, tert-butanol and glycerol is separated in a distillation column (Glycol). The methanol and tert-butanol is recycled back, and the glycerol with a purity of 99.99 wt.% can be sold. Purge of methanol has again been found unnecessary. The simulation results are now available for the economical evaluation. 4. Economical evaluation Based on the flow sheets and simulations above, an equipment size and cost estimation of the process equipment and corrected total capital investment (TCI) have been carried out for the four scenarios in Aspen Icarus Process Evaluator 2006.5. TCI includes working capital. This is then combined into an economical plant evaluation. The scenarios include 8 and 200 mio. kg biodiesel production plant a year with either cosolvent or solvent free with an enzyme price of 762.71 and 7.627€/kg. Consumption of enzymes is calculated based on expected productivity, see Table 2. The economical evaluations are carried out for a continuous biodiesel production plant. The storage capacity is 2 weeks for raw materials and 3 weeks of production for products based on Table 5 Operational data and revenue values for this study.

a b

Item

Cost

Enzyme Methanol Rapeseed oil tert-Butanol Biodiesel Glycerol (92 wt.%) Cooling water Electricity Number of shifts Working weeks/year Pricing Working capital Operating and laboratory charges Overhead Taxes General and administrative expenses (G&A) Depreciation model Plant type Location

762.71€/kg (7.627€/kg) 0.295€/kg 0.607€/kg 1.70€/kg 1.112€/kga 0.8667€/kgb 1.34€/m3 0.26€/kWh 3 52 weeks €, 0.76€/US$ 5% 15%/period 5% 15%/period 8%/period 10%/year in 10 years Bare field project Europe

Based on market price for biodiesel January 2009. The price is taken from Zhang et al. (2003b).

information from an existing biodiesel production plant. In the sizing of the unit operations, the enzyme volume and handling is not included due to the uncertainty of this. Only the price, reaction time and yield of them are included in this study. Equipment cost for each operational unit and TCI can be seen in Table 4. The total capital investment (TCI) is lower for solvent free production than with the cosolvent. The plant equipment cost alone is lower for the cosolvent production, but the installation costs of the solvent recovery column is higher than equipment and installation costs of the extra reactors and decanters needed when running solvent free. The TCI increases with plant size, see Fig. 4, but the lowest TCI is for solvent free production plants in the investigated range compared to the use of cosolvent. The use of cosolvent requires extra distillation facilities that are more expensive than the extra number of reactors and decanters needed for solvent free operation. It has previously been documented for biodiesel by Zhang et al. (2003b), that plant size has a great influence on the outcome of an economical biodiesel process evaluation. This study shows that TCI increases with production size, but not linearly and consequently the larger the plant the more potentially cost-effective is the plant. This is valid for solvent free and cosolvent operation alike. The total production costs and payback period as well as product price are calculated based on the simulation results. The total production costs are the sum of raw materials including enzymes at 762.71 or 7.627€/kg, utilities, labor, maintenance, supervision, operating charges, plant overhead and G&A. Operational data, raw material prices and revenue values can be seen in Table 5. The price of tert-butanol is supplied by a European industrial solvent supplier. The glycerol price is a significant factor in the overall process economy. Though a higher quality of glycerol can be obtained with enzymes, because the purification does not involve water washing, the price is set relatively low (the price of 92 wt.% glycerol). This is done in order to not overestimate the price and to lower the influence of the glycerol price on the viability of a plant. Total product price (Peters et al., 2003) is then calculated by the following equation:

Product price ¼

Manufacturing costs Production rate

ð2Þ

Manufacturing costs ¼ TPC þ Fixed costs  Byproduct sales

ð3Þ

Fixed costs ¼ Taxes þ Depreciation þ Rate of return  TCI

ð4Þ

Table 6 Total production cost of biodiesel and payback period calculation for enzymatic biodiesel production plant. Scenario Enzyme cost (€/kg) Direct production costs Raw materials Utilities Labor, maintenance, supervision Operating Charges Plant Overhead G&A Total production cost (TPC) Fixed costs Depreciation Taxes Byproduct sales Product price (€/kg)a Payback periodb a b

1 762.71

2 762.71

3 762.71

4 762.71

1 7.627

2 7.627

3 7.627

4 7.627

16,900,439 984,191 862,000 114,000 43,100 1,512,298 20,416,028

407,708,992 2,602,900 1,409,000 114,000 70,450 32,952,427 444,857,769

12,815,552 11,002,632 822,700 114,000 41,135 1,983,682 26,779,701

318,650,433 267,388,536 1,238,000 114,000 61,900 46,996,229 634,449,098

11,862,495 984,191 862,000 114,000 43,100 1,109,263 14,975,049

281,869,015 2,602,900 1,409,000 114,000 70,450 22,772,509 307,428,874

11,390,692 11,002,632 822,700 114,000 41,135 1,869,693 25,240,852

283,107,950 267,388,536 1,238,000 114,000 61,900 44,152,831 596,063,217

855,108 1,282,660 12,308,202 1.49 (3.03) N/A

1,734,161 2,601,242 307,789,665 0.73 (2.27) 0.25

1,018,837 1,528,256 12,303,633 2.38 (3.92) N/A

2,716,142 4,074,214 307,294,295 1.70 (3.23) N/A

855,107 1,282,660 12,308,202 0.75 (2.35) 3.59

1,925,406 2,888,109 307,769,240 0.05 (1.59) 0.09

1,018,837 1,528,256 12,303,633 2.19 (3.72) N/A

2,716,142 4,074,214 307.294.295 1.50 (3.04) N/A

Price without parenthesis is with byproduct sales, while price in paranthesis is without byproduct sales. Based on market price for biodiesel and glycerol January 2009 and enzyme price of 762.71€/kg and 7.627€/kg.

L.F. Sotoft et al. / Bioresource Technology 101 (2010) 5266–5274

Fig. 4. Total capital investment in M€ as a function of plant size.

The results are presented in Table 6 for an enzyme price of 762.71 and 7.627€/kg together with an analysis of payback period. Analysis of payback period (Peters et al., 2003) is calculated by Eqs. (4), (2), and (3)

Payback period ¼

TCI Biodiesel sales  Manufacturing costs

ð5Þ

First, results for scenarios with the current enzyme price. Only the large solvent free plant is cost-effective with a very short payback period of 0.25 year (based on a product price of 1.12€/kg). If a minimum product price on the other hand is calculated, the minimum product price is 0.73€/kg biodiesel. The scenarios of cosolvent operation and small scale solvent free are not cost-effective and have a minimum product price from 1.49 to 2.38€/kg. The cosolvent operation is largely made uneconomical due to the large utility (energy) requirements. At an enzyme cost of 7.627€/kg, the annual manufacturing cost of biodiesel and calculation of payback period for enzymatic biodiesel production plant is very different. The enzyme price is seen greatly to influence the economic viability and product price for the scenarios. Neither cosolvent scenarios are cost-effective, while the solvent free operation in large scale is very viable with a payback period of 0.09 year and a minimum product price of 0.05€/kg. The extremely low product price is due to the earnings on byproduct sales. The small scale solvent free scenario has a low minimum product price of 0.75€/kg, which is lower than the estimated market price, but the payback period is just in the upper end of what is viable. Generally, viable processes must

5273

have a payback period below 2 years for high-risk ventures, and >4 years for payback are not considered viable (Seider et al., 2004). Due to uncertainties and the new, but unstable market of biofuels, this kind of venture must be considered as high-risk. Therefore, the economical potential of the smaller scale plant must be considered carefully. The results are though very promising, and enzymatic biodiesel seems closer to be a realistically viable industrial process. With this study, the use of cosolvents and enzymes cannot be considered as a possible industrial viable production procedure, but the focus of research must alone be to improve solvent free processes. The scenarios in this paper are based on a price for a high value raw material; rape seed oil. When comparing the calculated prices with the literature studies shown in Table 1, it can be seen that the price of 0.05–2.38€/kg depending on production scale and enzyme price is in the same range as the literature values of 0.14–0.62€/kg for continuous processes. Comparison of the product price for production of 8 mio. kg biodiesel/year will show that the literature studies can produce biodiesel for 0.55–0.62€/kg with high quality raw materials and traditional catalysts, while this study shows that it can be produced solvent free at 0.75–1.49€/kg with enzymes. Enzymes are thus more expensive to use, but if improvements in enzyme lifetime and yield can be made together with significant, documented improvements in the environmental impact of the enzymatic production process then enzymatic biodiesel production is closer to become reality. A question also to be answered by this study is how the enzymes do influence the productions costs. To examine the distribution of the production costs, Fig. 5 presents the manufacturing costs allocated to the different costs minus revenue from byproduct sales for production of 8 and 200 mio. kg biodiesel/year. For all operations, the major cost factor is raw materials followed by byproduct sales and utilities. The enzyme price has a significant influence on the cost of the solvent free operation, but not to the same extent when it comes to the cosolvent operation due to the higher productivity of the enzymes in the latter. Byproduct sales are very high and important for the process economy, but are the same for all scenarios. Cost of utilities is much higher for the cosolvent than for the solvent free production for all scenarios, see Fig. 5. This outweighs the improved enzyme performance, and therefore the manufacturing costs are higher for cosolvent than solvent free production regardless of production scale and enzyme cost. A more detailed analysis of the raw material costs shows how the cost of oil, methanol and enzymes affect this group of costs, see Fig. 6.

Fig. 5. Distribution of manufacturing costs for production of 8 and 200 mio. kg biodiesel/year.

5274

L.F. Sotoft et al. / Bioresource Technology 101 (2010) 5266–5274

References

Fig. 6. Distribution of raw material costs for solvent free and cosolvent enzymatic biodiesel production.

It can be seen that for solvent free operation that costs of raw materials is 50% enzymes, 47% oil and 3% methanol, while the influence of enzyme cost is lower for cosolvent operation due to the improved enzyme performance. For cosolvent operation, the oil makes up 73% of the raw material costs, while the enzymes only make up for 22%. Four percentage of the costs can be contributed to the methanol, while the tert-butanol only accounts for 1% due to recycle. In summary this study documents the possibility of using enzymes for a viable solvent free biodiesel production. The result depends highly on production scale and on raw material price for as well oil as enzymes. The energy consumption of the cosolvent production makes it too expensive compared to the solvent free process. The present study and the study by Zhang et al. (2003b) show that production capacity is very important when planning biodiesel production plants, also for the use of other catalysts than enzymes. Too small scale often makes the plants cost too high to make productions viable. The study illustrates the importance of raw material and product prices to the plants viability and that the plant economy is very influenced by them making planning and profitability analysis very complicated. This though holds for all biodiesel plants, and is not limited to enzyme catalyzed plants.

5. Conclusion An enzyme catalyzed biodiesel production plant is simulated and economically evaluated for production of 8 and 200 mio. kg/ year biodiesel from rapeseed oil and methanol. The product price for solvent free production is estimated to 0.73–1.49€/kg biodiesel with a price of 762.71€/kg enzyme and 0.05–0.75€/kg biodiesel with a price of 7.63€/kg enzyme. Biodiesel can be produced with enzymes and cosolvent to a price of 1.50–2.38€/kg biodiesel. Solvent free enzyme biodiesel production process is viable and shows most promise, while the use of a cosolvent together with enzymes is not viable.

Acknowledgement The work was supported by The Danish Council for Strategic Research.

Bender, M., 1999. Economic feasibility review for community-scale farmer cooperatives for biodiesel. Bioresource Technology 70, 81–87. Daka Biodiesel, 2009. Green Accounting 2007/2008 Daka Biodiesel a.m.b.a. Daka Biodiesel, Løsning. Du, W. et al., 2008. Perspectives for biotechnological production of biodiesel and impacts. Applied Microbiology and Biotechnology 79, 331–337. Emmelev Molle, 2009. Green Accounting 2008. Emmelev Molle, Otterup. EN 14214, 2008. EN14214:2008 Automotive Fuels – Fatty Acid Methyl Esters (FAME) for Diesel Engines – Requirements and Test Methods. European Committee for Standardization, Brussels. Fjerbaek, L., Christensen, K.V., Norddahl, B., 2009. A review of current state of biodiesel production using enzymatic transesterification. Biotechnology and Bioengineering 102, 1298–1315. Haas, M.J., McAloon, A.J., Yee, W.C., Foglia, T.A., 2006. A process model to estimate biodiesel production costs. Bioresource Technology 97, 671–678. Harding, K.G., Dennis, J.S., von Blottnitz, H., Harrison, S.T.L., 2008. A life-cycle comparison between inorganic and biological catalysis for the production of biodiesel. Journal of Cleaner Production 16, 1368–1378. Kuramochi, H. et al., 2009. Application of UNIFAC models for prediction of vapor– liquid and liquid–liquid equilibria relevant to separation and purification processes of crude biodiesel fuel. Fuel 88, 1472–1477. Li, L. et al., 2006. Lipase-catalyzed transesterification of rapeseed oils for biodiesel production with a novel organic solvent as the reaction medium. Journal of Molecular Catalysis B: Enzymatic 43, 58–62. Marchetti, J.M., Errazu, A.F., 2008. Technoeconomic study of supercritical biodiesel production plant. Energy Conversion and Management 49, 2160–2164. Mittelbach, M., 2005. Biodiesel: production technologies and perspectives. In: Proceedings of First International Conference on Renewable Resources and Biorefineries. Ghent. Nelson, R.G., Howell, S.A., Weber, J.A., 1994. Potential feedstock supply and costs for biodiesel production. In: Bioenergy ‘94 Proceedings of the Sixth National Bioenergy Conference. Reno/Sparks. Noordam, M., Withers, R., 1996. Producing biodiesel from canola in the inland northwest: an economic feasibility study. Idaho Agricultural Experiment Station Bulletin 785, 12. Peters, M.S., Timmerhaus, K., West, R.E., 2003. Plant Design and Economics for Chemical Engineers, fifth ed. McGraw-Hill, New York. Reijnders, L., Huijbregts, M.A.J., 2008. Biogenic greenhouse gas emissions linked to the life cycles of biodiesel derived from European rapeseed and Brazilian soybeans. Journal of Cleaner Production 16, 1943–1948. Saka, S., Kusdiana, D., 2001. Biodiesel fuel from rapeseed oil as prepared in supercritical methanol. Fuel 80, 225–231. Sakai, T., Kawashima, A., Koshikawa, T., 2009. Economic assessment of batch biodiesel production processes using homogeneous and heterogeneous alkali catalysts. Bioresource Technology 100, 3268–3276. Santana G.C.S., Martins, P.F., de Lima da Silva, N., Batistella, C.B., Filho, R.M., Macial, M.R.W., in press. Simulation and cost estimate for biodiesel production using castor oil. Chemical Engineering Research and Design (Corrected Proof). Seider, W.D., Seader, J.D., Lewin, D.R., 2004. Product and Process Design Principles, second ed. John Wiley & Sons, Inc., New York. Shimada, Y., Watanabe, Y., Samukawa, T., Sugihara, A., Noda, H., Fukuda, H., Tominaga, Y., 1999. Conversion of vegetable oil to biodiesel using immobilized Candida antarctica Lipase. Journal of American Oil Chemists’ Society 76, 789– 793. van Kasteren, J.M.N., Nisworo, A.P., 2007. A process model to estimate the cost of industrial scale biodiesel production from waste cooking oil by supercritical transesterification. Resources, Conservation and Recycling 50, 442–458. West, A.H., Posarac, D., Ellis, N., 2008. Assessment of four biodiesel production processes using HYSYS. Plant. Bioresource Technology 99, 6587–6601. You, Y.D. et al., 2008. Economic cost analysis of biodiesel production: case in soybean oil. Energy Fuels 22, 182–189. Zhang, Y., Dube, M.A., Mclean, D.D., Kates, M., 2003a. Biodiesel production from waste cooking oil: 1. Process design and technological assessment. Bioresource Technology 89, 1–16. Zhang, Y., Dube, M.A., Mclean, D.D., Kates, M., 2003b. Biodiesel production from waste cooking oil: 2. Economic assessment and sensitivity analysis. Bioresource Technology 90, 229–240.