It is already a consensus that a shift of the global energy matrix towards renewable sources is mandatory. Yet, the role that each specific alternative will play, say, at the year 2050, will be defined along the road, depending on technological developments, political options by stakeholders, economical and social demands. Anyway, in this scenario ethanol will certainly be an important biofuel.

Sugarcane is known to be the most efficient crop for 1G ethanol production, with an energy balance of 9.3 produced/consumed tonne of oil equivalent (toe) [1]. During the 1970’s the Brazilian government initiated the National Ethanol Program (PROALCOOL, in Portuguese, [2]) to decrease national dependence on oil. Since then, the use of 1G ethanol as a vehicle fuel has been consolidated, and presently 86% of the cars sold in this country are flex-fuel, running with any mixture of ethanol and gasoline [3]. In modern facilities, ethanol production is a highly integrated process, with sugarcane bagasse burnt in boilers to supply the industrial plant energy demands, further exporting the surplus of electric energy to the grid.

One of the alternatives for the industrial production of 2G ethanol is the biochemical route, i.e., acid or enzymatic hydrolysis of the biomass followed by fermentation of the resulting sugars. Logistics and transportation of the lignocellulosic raw material may be a bottleneck for 2G ethanol [4]. From this point of view, sugarcane bagasse has an important advantage, since it has already been collected and processed for the extraction of the juice, being immediately available at the plant site. Moreover, sugarcane trash (mostly the leaves) can be transported with the stalk, after small adaptations of the mechanical harvesting – although part of this biomass must be left for covering the fields [5]. Since the process must be energetically self-sufficient, the addition of sugarcane trash as boiler fuel can increase the amount of bagasse available for hydrolysis, therefore enhancing ethanol yields.

Industrial production of 2G ethanol is still not consolidated in large scale. Therefore, an economic analysis is important to indicate if it is the most interesting alternative, specially when compared to selling electric energy (bioelectricity). Nevertheless, this answer is not unique, given the volatility of relative prices: biomass electricity prices in public auctions in Brazil ranged from 85.35 USD/MWh to 53.02 USD/MWh in the last two auctions (Aug/2010 and Aug/2011) [6], a -37.9% variation. The spot market (that buys surplus power, beyond the amount contracted during the auctions) presented an even higher range of prices, between 3.26 USD/MWh and 341.13 USD/MWh over the last nine years (Jan/2003 - Dez/2012) [7]. Ethanol prices are equally volatile, changing from 0.258 USD/L to 0.818 USD/L (for the hydrated fuel) over the same period of nine years [8]. All prices above, used in this study, were calculated in Brazilian reais, brought to December/2012 value (to take into account the inflation in the period), and converted to US dollars using the exchange rate value of 2.077 BRL/USD (dez/2012).

Sugarcane biorefineries have been intensively studied by the recent literature. Seabra et al. (2011) [9] presented economic and environmental analyses of a sugarcane biorefinery. The authors concluded that, although electric energy presented a better economic feasibility, its environmental impact was greater than the one for second generation ethanol. An economic analysis comparing 2G ethanol production with electric energy was also done by Dias et al. (2011) [10]. The authors concluded that, although for the present technology electric energy is a better option, 2G ethanol can compete with it if sugarcane trash is used, provided that new technologies could increase yields. On the other hand, Macrelli et al. (2012) [11] presented results for several sugarcane biorefineries configurations and concluded that 2G ethanol from sugarcane is already competitive with 1G starch-based ethanol in Europe. The advantages of the integrated production of 1G and 2G ethanol production based on sugarcane was highlighted by Dias et al. (2012) [12]. The integrated biorefinery presented higher ethanol production rates, and better economic and environmental performance, when compared to a stand-alone 2G ethanol-from-sugarcane bagasse plant.

The 2G biorefinery could be flexible, just as the industry that employs current 1G technology already is, shifting between sugar and ethanol production. This new flexible biorefinery might be able to choose between electric energy and 2G ethanol production. The present study focuses on assessing the economic feasibility of a flexible biorefinery, for an autonomous distillery (not considering the manufacture of sugar), and comparing it to the dedicated 1G + electric energy and 1G + 2G ethanol biorefineries. The process chosen as case study uses enzymatic hydrolysis of the biomass (sugarcane bagasse) and ethanolic fermentation of hexoses and pentoses. Specifically, this study presents a computational applicative that may be a useful tool for the process scheduling of future cane-based biorefineries but, beyond that scope, that may also support decision-making concerning national energy policies. Such computationally robust tool was developed within an equation-oriented simulator (EMSO) [13] and is based on phenomenological modelling, at least for the most important unit operations and reactors that are present in the process.

Process simulation

Two boundary process configurations were considered and compared: an industry producing 1G ethanol and burning all sugarcane bagasse and 50% of the trash produced in the field for power generation in a Rankine cicle (BioEE) and another one using all bagasse surplus (the biorefinery must be energetically self-sufficient) for 2G ethanol production, integrated to the 1G facility (BioEth). The most important information for BioEE and BioEth streams (as enumerated in Figure 1 and Figure 2) is presented in Table 1 and Table 2. 74% of the bagasse could be diverted to 2G ethanol in BioEth. This accounted for an increase in ethanol production of 25.8% when compared to BioEE. At this condition, a specific production yield for 2G ethanol of 120.7 L/tonne of bagasse was obtained, which is a conservative estimate based on current yields (158 L/tonne of lignocellulosic material [12]). The increase in steam consumption for 2G ethanol was entirely fulfilled by the burning of lignin and of non-hydrolyzed cellulose. Since 65% of the cellulose was hydrolyzed, 35% of the material was still able to be separated and used as fuel. Table 3 shows ethanol production (total and specific) for both cases.

Figure 1

Process diagram for first generation ethanol production.

Figure 2

Process diagram for second generation ethanol production.

Economic analysis

The results of the economic analysis are presented in Table 7. Both IRR and Net Present Value (NPV) showed similar results, with BioEth being the best option, followed by the flexible biorefinery. As it can be seen, none of the options presented a positive NPV or, equivalently, an IRR higher than the Minimum Acceptable Rate of Return (MARR), assumed to be 11%/yr, when the present market prices for ethanol and electric energy in Brazil are considered. Nevertheless, as it will be shown later in the sensitivity analysis, an increase in ethanol price of 21.1% can turn all options feasible. Besides, specifically for the BioEth biorefinery, an increase of only 11.5% in this price would assure feasibility.

Unfortunately, a direct comparison of the obtained result with the ones from literature is not a straightforward task. There is a large variability in the economic premisses and in the technical solutions for the biorefineries that can be considered in a techno-economic study. For example, Seabra and Macedo (2011) [9] considered a 2G biorefinery adjacent to the 1G industrial plant (i.e., not sharing utilities and equipment, just importing bagasse), while Dias et al. (2011) [10] and Macrelli et al. (2012) [11] considered that 1G and 2G production plants were integrated (in different degrees). Even when similar processes are considered, the results can be quite different: while the autonomous 1G industrial plant in Dias et al. (2011) [10] obtained an IRR of 15.9 %, in Macrelli et al. (2012) [11] a similar plant obtained an IRR of 32.1 %.

Sensitivity analysis

Since ethanol prices presented an approximately normal distribution, its standard deviation was calculated and a variation equivalent to one standard deviation (20.5%) was used for the sensitivity analysis. On the other hand, electric energy price (spot market) did not present a normal distribution and its influence was better seen when the double of its current price was considered. Therefore, for this case a variation of 40% was considered and, for the other ones that had only punctual data available, a variation of 20% was chosen, thus assuming a percent variation similar to the one in ethanol prices. For Flex sensitivity analysis, the variations described were applied before the simulation of the seasonality effect on the prices.

Figure 3 presents the sensitivity analysis of electric energy selling prices in public auctions on the IRR. As expected, the influence was higher for BioEE, since it has the higher amount of electric energy being sold in public auctions. BioEth and Flex produced the same amount of electric energy to be sold in the auction market. Therefore, both were equally influenced by the variation in auction prices. As it can be seen, the effect of the electric energy price on the biorefineries economic performance was not strong enough to make them feasible. The cash flow provided by the selling of electric energy was one order of magnitude smaller than the one for ethanol.

Figure 3

Impact of electric energy selling prices (annual auctions) on the internal rate of return. All other prices kept unchanged.

Electric energy prices in the spot market only influenced Flex’s IRR (Figure 4). The sensitivity to this parameter was quite small, though. As seen in Table 8, only a few months were dedicated to electric energy production, even when the spot market price was assumed equal to twice its current value. The income from electric energy sold in the spot market was small, compared to the one coming from public auctions. It should be mentioned that an increase in both auction and spot prices can be expected in the near future in Brazil. Thermoelectric plants using natural gas currently complement the production of hydroelectric energy during the dry period, and this is hardly sustainable, both in economic and environmental perspectives. Therefore, an increase in these prices, associated with improvements in the distribution grid, could stimulate investments in cogeneration.

Figure 4

Impact of electric energy selling prices (spot market) on the internal rate of return. The central point corresponds to a price (80.2 USD/MWh) equal to twice the current value. All other prices kept unchanged.

It is worth noticing that while Flex could reproduce almost perfectly the behaviour of BioEth (except for the higher investment cost), the same was not true for BioEE. This is due to the fact that the latter sold all its available electric energy in the auction market, which pays higher prices (in general), while the flexible biorefinery would not do this.

The sensitivity with respect to enzyme prices was not as significant as initially expected (Figure 5). Although enzyme costs played a major role in composing 2G ethanol prices (75% of the total cost), this influence was diluted by the 1G ethanol costs. Therefore, the overall ethanol costs (1G + 2G) changed from 416.3 USD/m³ (for an enzyme price of 2.02 USD/kg) to 392.6 USD/m³ (for an enzyme price of 1.35 USD/kg), a variation of 5.9%, for a 40% reduction in enzyme price.

Figure 5

Impact of enzyme costs on the internal rate of return. All other prices kept unchanged.

Figure 6 shows the impact of ethanol selling prices on the IRR for the current price for electric energy in spot market (a) and for a price equal to twice its current value (b). It is clear that ethanol price presented the higher influence for all biorefineries. In fact, it was the only factor strong enough to make the biorefineries economically feasible, within the range spanned in this study. Accordingly, BioEE, BioEth and Flex became feasible for an ethanol price of 622.1 USD/m³, 572.8 USD/m³and 583.1 USD/m³, respectively. This means an increase in ethanol price of 21.1%, 11.5% and 13.5%, respectively. While in Figure 6 (a) it is clear that BioEth was always superior to the flexible biorefinery, when a higher price in the spot market was considered (Figure 6 (b)) the flexible biorefinery became less influenced by negative variations of the ethanol price. The switch towards electric energy production for the flexible biorefinery is clear in this case. If the current ethanol price were considered, no extra electric energy was produced in any month for the whole period considered (25 years). On the other hand, when ethanol price is on its lower value, the flexible biorefinery switched to electric energy production during 30% of the months, becoming superior to BioEth.

Figure 6

Impact of ethanol selling prices on the internal rate of return. In (a), the current spot energy price (40.1 USD/MWh) is considered, while in (b) a spot energy price of twice this value (80.2 USD/MWh) is used. All other prices kept unchanged.

The non-linear behaviour of the system becomes evident in Figure 6. This is mostly caused by the fact that both tax rates and dividends only apply when there is profit. Therefore, the positive influence of an increase in ethanol prices on the IRR is attenuated by both of them and the second derivative of these curves is negative.

Since there are many uncertainties regarding the investment costs on the second generation ethanol production process, a sensitivity analysis was also performed for this parameter. Maximum and minimum values of twice and half the base investment cost for the 2G ethanol process were considered. As seen in Figure 7, the impact of the investment cost on the IRR is small. The decrease in IRR between the maximum and minimum investment costs was only of -0.72% and -0.68% for BioEth and Flex, respectively. Therefore, it is expected that the uncertainty of this value will not invalidate the analyses performed.

Figure 7

Impact of the investment cost on second generation ethanol on the internal rate of return. All other prices kept unchanged.

It is worth highlighting that the 2G process is coupled to the 1G process, which is responsible for the low investment cost of the former, when compared to literature [15], because costs of fermentation, distillation and combined heat and power stages are included in our 1G cost, which is scaled proportionally to the combined process flows.

A flexible sugarcane biorefinery (Flex) was simulated and compared to dedicated first generation + cogeneration (BioEE) and first + second generation ethanol (BioEth) biorefineries. The flexible one presented an inferior economic performance in all cases for 2012 market prices. Nevertheless, if an increase in electric energy prices in the spot market were considered, the flexible biorefinery could be the best option.

In general, all biorefineries were not economically feasible for 2012 selling prices and costs. This conclusion was actually validated in practice by recent governmental actions (April/2013) which aimed to improve the competitiveness of the ethanol industry. Additionally, of all parameters considered in the sensitivity analysis, ethanol prices were the only ones that could make the biorefineries economically viable, within the studied range of values. In fact, an increase in ethanol price of 21.1% would be sufficient to make feasible all biorefineries. Particularly for BioEth, a 11.5% increase in ethanol prices would be enough for viability.

Enzyme prices, on the other hand, were less significant than it could be expected. This is due to the fact that 2G ethanol costs were diluted by 1G ethanol’s, produced in higher volumetric rates. Therefore, the overall ethanol production cost in the integrated plant was not greatly influenced by enzyme prices.

Finally, it is obvious that the quantitative results presented here are dependent on the economic scenario proposed and on the assumed process yields and energy demands, but the presented methodology is general.

Process implemented

EMSO Software

EMSO [13] was the software chosen as the platform for the simulations in this study. It is an equation-oriented, general purpose process simulator with its own modelling language [21]. Besides the several models for the main process pieces of equipments, the software also allows the user to implement his/hers own models. The software has several numeric solvers for solution of algebraic and differential-algebraic systems, and users can plug in their own numerical routines (in C/C++ or FORTRAN). Physical and thermodynamic properties can be added to the package database by the user whenever needed.

Economic analysis

Authors’ original submitted files for images

1. Macedo IC, Seabra JEA, Silva JEAR: Green house gases emissions in the production and use of ethanol from sugarcane in Brazil: The 2005/2006 averages and a prediction for 2020. Biomass Bioenergy 2008,32: 582-595. 10.1016/j.biombioe.2007.12.006View Article
2. Zanin GM, Santana CC, Bon EPS, Giordano RLC, Moraes FF, Andrietta SR, Neto CCC, Macedo IC, Fo DL, Ramos LP, Fontana J: Brazilian bioethanol program. Appl Biochem Biotechnol 2000, 84: 1147-1163.View Article
3. National Association of Motor Vehicles (ANFAVEA): Brazilian automotive industry yearbook. Tech. rep., São Paulo, 2012
4. Gnansounou E: Production and use of lignocellulosic bioethanol in Europe: Current situation and perspectives. Bioresour Technol 2010, 101: 4842-4850. 10.1016/j.biortech.2010.02.002View Article
5. Hassuani SJ, Leal MRLV, Macedo IC: Biomass power generation: Sugarcane bagasse and trash. Piracicaba: United Nations Development Programme and Sugarcane Technology Centre; 2005.
6. Electric Energy National Agency – ANEEL (in portuguese) http://​www.​aneel.​gov.​br
7. Electric Energy Commercialization Chamber – CCEE (acronym in portuguese) http://​www.​ccee.​org.​br/​
8. Center of Advanced Studies in Applied Economy – CEPEA/ESALQ/USP (in portuguese) http://​www.​cepea.​esalq.​usp.​br
9. Seabra JE, Macedo IC: Comparative analysis for power generation and ethanol production from sugarcane residual biomass in Brazil. Energy Policy 2011, 39: 421-428. http://​www.​sciencedirect.​com/​science/​article/​pii/​S030142151000770​6 10.1016/j.enpol.2010.10.019View Article
10. Dias MO, Cunha MP, Jesus CD, Rocha GJ, Pradella JGC, Rossell CE, Maciel Filho R, Bonomi A: Second generation ethanol in Brazil: Can it compete with electricity production? Bioresour Technol 2011,102(19):8964-8971. 10.1016/j.biortech.2011.06.098View Article
11. Macrelli S, Mogensen J, Zacchi G, et al.: Techno-economic evaluation of 2 nd generation bioethanol production from sugar cane bagasse and leaves integrated with the sugar-based ethanol process.Biotechnol Biofuels 2012, 5: 22. 10.1186/1754-6834-5-22View Article
12. Dias M, Junqueira T, Cavalett O, Cunha MP, Jesus C, Rossell C, Filho R, Bonomi A: Integrated versus stand-alone second generation ethanol production from sugarcane bagasse and trash. Bioresour Technol 2012,103: 152-161. 10.1016/j.biortech.2011.09.120View Article
13. Soares RP, Secchi AR: EMSO: A new environment for modelling, simulation and optimisation. Comput Aided Chem Eng 2003, 14: 947-952.View Article
14. Agblevor F, Rejai B, Wang D, Wiselogel A, Chum H: blueInfluence of storage conditions on the production of hydrocarbons from herbaceous biomass. Biomass Bioenergy 1994, 7: 213-222. 10.1016/0961-9534(94)00063-YView Article
15. Eggeman T, Elander RT: Process and economic analysis of pretreatment technologies. Bioresource Technol2005,96(18):2019-2025. 10.1016/j.biortech.2005.01.017View Article
16. Furlan FF, Costa CBB, Fonseca GC, Soares RP, Secchi AR, Cruz AJG, Giordano RC: Assessing the production of first and second generation bioethanol from sugarcane through the integration of global optimization and process detailed modeling. Comput Chem Eng 2012, 43: 1-9.View Article
17. Canilha L, Chandel AK, Suzane dos Santos Milessi T, Antunes FAF, Luiz da Costa Freitas W, das Graças Almeida Felipe M, da Silva SS: Bioconversion of sugarcane biomass into ethanol: An overview about composition, pretreatment methods, detoxification of hydrolysates, enzymatic saccharification, and ethanol fermentation. J Biomed Biotechnol 2012, 2012: 1-15.View Article
18. Wooley R, Putsche V: Development of an ASPEN PLUS physical property database for biofuels components. NREL, Tech. rep., Report MP-425-20685 1996, 38
19. Silva C, Zangirolami T, Rodrigues J, Matugi K, Giordano R, Giordano R: An innovative biocatalyst for production of ethanol from xylose in a continuous bioreactor. Enzyme Microb Technol 2012, 50: 35-42. 10.1016/j.enzmictec.2011.09.005View Article
20. Pereira LTC, Pereira LTC, Teixeira RSS, Bon EPS, Freitas SP: Sugarcane bagasse enzymatic hydrolysis: rheological data as criteria for impeller selection. J Ind Microbiol Biotechnol 2011,38(8):901-907. 10.1007/s10295-010-0857-8View Article
21. Rodrigues R, Soares RP, Secchi AR: Teaching chemical reaction engineering using EMSO simulator. Comput Appl Eng Educ 2010,18(4):607-618. 10.1002/cae.20255View Article