Alkaline sulfite pretreatment for integrated first and second generation ethanol production: A techno-economic assessment of sugarcane hybrids

Biomass and Bioenergy 119 (2018) 314–321

Contents lists available at ScienceDirect

Biomass and Bioenergy journal homepage: www.elsevier.com/locate/biombioe

Research paper

Alkaline sulfite pretreatment for integrated first and second generation ethanol production: A techno-economic assessment of sugarcane hybrids

T

F.M. Mendesa,∗, M.H. Vasconcelosb,c, M.O.S. Diasd, A. Ferrazb, A.M.F. Milagresb, J.C. Santosb, C.D.F. Jesusa, M.D.B. Watanabea, T.L. Junqueiraa, A. Bonomia a

Laboratório Nacional de Ciência e Tecnologia Do Bioetanol (CTBE), Centro Nacional de Pesquisa Em Energia e Materiais (CNPEM), CEP 13083-970, Campinas, São Paulo, Brazil b Departamento de Biotecnologia, Escola de Engenharia de Lorena, Universidade de São Paulo (EEL/USP), Lorena, São Paulo, Brazil c Instituto Federal de Educação, Ciência e Tecnologia (IFRO), Campus Guajará-Mirim, Rondônia, Brazil d Instituto de Ciência e Tecnologia, Universidade Federal de São Paulo (ICT/UNIFESP), São José Dos Campos, São Paulo, Brazil

A R T I C LE I N FO

A B S T R A C T

Keywords: Sugarcane hybrid Alkaline sulfite pretreatment Biorefinery Simulation Techno-economic analysis

Sugarcane hybrids have been developed through genetic engineering and plant breeding to produce lignocellulosic crops that are more susceptible for second generation biofuels production. Adequate evaluation of these plants requires process development combined with proper economic assessment. In this study, alkaline sulfite pretreatment of sugarcane biomass derived from four selected sugarcane hybrids was assessed for second generation ethanol production integrated with a first generation biorefinery. Process simulation and economic analysis were used to evaluate 32 biorefinery scenarios including different pretreatment conditions (high and low severity), enzymatic hydrolysis time (24 and 72 h) and pentoses destination (fermentation to ethanol or discard). Results indicated that high field productivity and low recalcitrance after pretreatment were critical characteristics for a selected sugarcane hybrid. High sodium sulfite loads were useful to increase ethanol production in the 1G2G biorefinery. However, sodium sulfite cost was relevant in the 1G2G ethanol cost. Sensitivity analysis applied to the best biorefinery scenarios indicated that maximum sodium sulfite prices to reach minimum acceptable rate of return (12%) were US$ 0.66/kg and US$ 0.47/kg for severe and mild pretreatments, respectively.

1. Introduction Computer-aided process simulation and economic evaluation of several biorefinery schemes have been used to assess their technoeconomic viability according to the feedstock or proposed processing route [1-4]. This kind of analysis helps to identify persistent bottlenecks hindering industrial implementation of technological developments reached last years, many of those concerning second generation (2G) ethanol production [2,3]. Techno-economic evaluation of a proposed 2G biorefinery is fundamental to guide researchers and investors. Such type of process simulation requires reliable databases. In the case of sugarcane feedstock, the Brazilian Bioethanol Science and Technology Laboratory (CTBE) developed a robust tool named Virtual Sugarcane Biorefinery (VSB), which allows evaluation of technical, economic, social and environmental impacts of different processing technologies regarding the production of bioethanol, sugar, bioelectricity and other products [2].

∗

Sugarcane bagasse is a particularly valuable source of lignocellulose generated in first-generation (1G) sugar and ethanol mills. Processing of sugarcane bagasse into 2G ethanol could share part of the mill available facilities, such as juice concentration, fermentation, distillation, cogeneration of electricity and storage units, resulting in an integrated 1G2G biorefinery [5]. Considering the current development of 2G ethanol technology, lignocellulose recalcitrance has been overcome by using different pretreatment options and, in spite of the great number of reported works in this topic, this step remains as a relevant bottleneck in the process. For sugarcane bagasse, alkaline-sulfite chemothermomechanical (CTM) pretreatment is efficient to enhance polysaccharide enzymatic hydrolysis rate and sugars yield. The degree of removal of each lignocellulose component during pretreatment depends on the reaction severity, but optimized processes can provide glucose yields as high as 90% after enzymatic digestion of the pretreated material [6,7]. This CTM pretreatment is based on industrially developed CTM pulping processes

Corresponding author. E-mail address: [email protected] (F.M. Mendes).

https://doi.org/10.1016/j.biombioe.2018.10.005 Received 30 March 2018; Received in revised form 28 September 2018; Accepted 3 October 2018 0961-9534/ © 2018 Elsevier Ltd. All rights reserved.

Biomass and Bioenergy 119 (2018) 314–321

F.M. Mendes et al.

Fig. 1. Block flow diagram for integrated 1G2G process. Process steps in 1G autonomous distilleries (white blocks) and 2G process (gray blocks). Dotted line represents the scenarios in which C5 and C6 fractions were fermented separated from sugarcane juice.

biorefineries, were based on VSB platform and databases [2,3].

currently used in the pulp and paper industry [7]. Along with biomass pretreatment, the development of plants with low recalcitrance helps to decrease biorefinery costs [2,8-10]. For example, sugarcane breeding programs provided hybrids with improved field productivity and varied biomass composition [11-13]. Studies with selected hybrids indicated that high sucrose and fiber contents are useful for 1G ethanol production [10], whereas genotypes with low lignin content require less severe pretreatment for 2G ethanol production [6]. The techno-economic impacts for using these new sugarcane hybrids require integrated evaluation of the 1G2G sugarcane biorefinery, which can be achieved through computer simulation tools. In a previous work, several new sugarcane hybrids were evaluated in the VSB context to estimate production of electricity, sugar and ethanol in a 1G biorefinery model [10]. Only high field productivity, and high sucrose and fiber contents fulfill plant characteristics to provide high internal rate of return for the process. In the present study, selected sugarcane hybrids were used for detailed techno-economic evaluation of an integrated 1G2G biorefinery. Process options included: different chemical loads in the alkaline-sulfite CTM pretreatment; varied enzymatic hydrolysis periods; and C6 fermentation or C5eC6 co-fermentation. These varied scenarios resulted in 32 individual techno-economic runs for processes comparison. All scenarios were evaluated regarding ethanol and electricity production, and the economic perspective considering metrics such as internal rate of return (IRR) and net present value (NPV). Conclusions were drawn to highlight sugarcane hybrid characteristics and processing conditions enabling profitable 1G2G sugarcane biorefineries.

2.2. Process description The biorefinery was modeled for a sugarcane input of 500 metric tons of cane (TC) per hour. An optimized 1G2G ethanol production facility, which uses efficient cogeneration systems with high-pressure boilers (65 bar), ethanol dehydration by adsorption on molecular sieves, and reduced process steam consumption was assumed [3]. Sugarcane bagasse and a fraction of the sugarcane straw (50% of the amount produced in the field) were considered as fuels for production of steam and electricity in combined heat and power (CHP) units, supplying the entire thermal and electrical requirements of the 1G2G processes. Surplus bagasse (the fraction which exceeds the amount required for steam production after using all available straw) was destined for 2G ethanol production, and subjected to pretreatment, hydrolysis of carbohydrate polymers and fermentation of the resulting sugars. In this case, bagasse is pretreated in alkaline-sulfite CTM process [6]. After pretreatment, two fractions are obtained: a solid fraction enriched with cellulose and hemicellulose, and a liquid fraction containing mainly sulfonated lignin. The solid fraction is hydrolyzed by enzymes, where cellulose and hemicellulose are converted into glucose (C6 sugar) and xylose (C5 sugar), respectively. In some scenarios, this liquor is mixed with sugarcane juice in the 1G ethanol process for C6/ C12 fermentation; in the other scenarios, C5 and C6 fractions were fermented in a mixture, separately from the sugarcane juice (C12 sugar). Residual solids from enzymatic hydrolysis containing lignin and unreacted cellulose and hemicellulose are burnt in the combined heat and power (CHP) units as supplementary fuel. After fermentation, alcoholic streams were sent to a series of distillation columns and dehydration processes where anhydrous ethanol (99.6 wt % purity) is obtained. A simplified block flow diagram of the process is shown in Fig. 1. The unit operations of sugars extraction, juice treatment and concentration, fermentation, distillation and dehydration were simulated as described by Bonomi et al. [2]. Process Simulation Diagram (PSD) for pretreatment step was built in the Aspen Plus® software and is shown in Fig. 2. As can be observed, a stream containing surplus sugarcane bagasse (BAGASSE) from combined heat and power generation unit is inserted in the pretreatment

2. Materials and methods 2.1. Raw material and data sources Raw materials described in this work corresponded to four experimental sugarcane hybrids selected from previous work [6,7,10,12]. Data related to chemical compositions [12], field productivity, sucrose and fiber contents [10], and pretreatment and enzymatic digestibility [6] of sugarcane hybrids were obtained from previous reports. Those data can be found in the Supplementary material. Data for mill operational conditions, both 1G and integrated 1G2G sugarcane 315

Biomass and Bioenergy 119 (2018) 314–321

F.M. Mendes et al.

Fig. 2. Process simulation diagram for alkaline-sulfite chemothermomechanical pretreatment step built using Aspen plus®.

2.4. Simulation and economic analysis

reactor (REACTOR), where alkaline-sulfite process is simulated. Outlet stream (PT-BAG-1) from reactor is flashed (PT-FLASH) and the liberated steam energy (STEAM) is used to heat (PT-HEAT) the alkalinesulfite stream (ALKALINE SULFITE) that is also inserted in the pretreatment reactor. Slurry from the flash unit (PT-FLASH) is inserted in a filter, which uses water (PT-H2O) to wash and separate the pretreated bagasse from liquid liquor (LIQUOR) enriched in sulfonated lignin. Pretreated bagasse (PT-BAG-2) is then enzymatically hydrolyzed, while the liquor (LIQUOR) is sent to effluent treatment. PSD was elaborated to estimate mass and energy balances of an industrial process; however, in industry, the pretreatment reactor is coupled to a disk refining that was not represented in Aspen Plus®, although its energy consumption was considered in the simulation of the combined heat and power generation unit. Actually, alkali-sulfite CTM processes are well established at industrial level with machinery available in the market, including chemicals impregnation and disk refining [6]. Non-random two-liquid model (NRTL) was the thermodynamic model employed in most of the simulations. Simulation for steam/ electricity cogeneration system used steam tables and Redlich-KwongSoave and Boston–Mathias alpha function (RKS-BM) [2,4].

Mass and energy balances for each scenario were obtained through computer simulations performed using software Aspen Plus®, according to the methodology employed in previous works within the VSB framework [2,4,14]. Simulation of the 2G production process was carried out considering conversion parameters calculated from the results published by LauritoFriend et al. [6]. Table 2 presents a list of the most important parameters used for the 1G2G biorefinery simulations. Production of anhydrous ethanol and surplus electricity was determined for each scenario based on simulation results. These results, along with economic parameters, were used to perform the discounted cash flow analysis to calculate the internal rate of return (IRR), net present value (NPV) and ethanol production cost. Sugarcane and straw production costs (Table 3) were estimated based on biomass productivities, machinery and inputs used for sugarcane production using CanaSoft, an agriculture model within the VSB. These agricultural costs were included in the industrial operating costs by considering a vertically integrated model where biomass production and the industrial

2.3. Definition of scenarios

Table 2 Important parameters used for the 1G2G biorefinery simulations.

Two pretreatment severity levels were evaluated in the 2G process: 5% NaOH and 10% Na2SO3 (high severity) and 2.5% NaOH and 5% Na2SO3 (low severity). For each condition, two reaction times of enzymatic hydrolysis (24 h and 72 h) were evaluated. In the fermentation steps two conditions were evaluated: (1) hexoses (C6 sugars) were fermented in a mixture with sugarcane juice in the 1G process (without pentoses fermentation) and (2) C5 and C6 sugars were fermented in a mixture in the 2G process. Characteristics of each scenario are shown in Table 1. These 8 scenarios were evaluated for each selected sugarcane hybrid (H58, H89, H140, H146), totalizing 32 scenarios. Table 1 Varied scenarios evaluated for the 2G biorefinery using four different sugarcane hybrids. Scenario

Pretreatment

Enzymatic Hydrolysis

Fermentation

I II III IV V VI VII VIII

2.5% NaOH and 5% Na2SO3 2.5% NaOH and 5% Na2SO3 2.5% NaOH and 5% Na2SO3 2.5% NaOH and 5% Na2SO3 5% NaOH and 10% Na2SO3 5% NaOH and 10% Na2SO3 5% NaOH and 10% Na2SO3 5% NaOH and 10% Na2SO3

24 h 24 h 72 h 72 h 24 h 24 h 72 h 72 h

C5 C6 C5 C6 C5 C6 C5 C6

Parameter

Value

Reference

Sugarcane processed (wet basis) Operation period Sugarcane straw produced (dry basis) Fraction of straw recovered from the field Efficiency of juice extraction in the mills Hydrated ethanol purity Anhydrous ethanol purity Boiler pressure 1G Process electric energy consumption Temperature - pretreatment Reaction time - pretreatment Solid content - pretreatment Disk refiner consumption Temperature - enzymatic hydrolysis Reaction time - enzymatic hydrolysis Enzyme dosage (FPU/g of pretreated biomass) Fermentation efficiency (C12/C6) Xylose fermentation efficiency (48 h)

500 TC/ha 200 days/year 140 kg/TC 50% 96% 93 wt% 99.6 wt% 65 bar 30 kWh/TC 120 °C 10 minb 10% 67 kWh/t bagasse 50 °C 24 h and 72 h 10c 90% 80%

[4] [2] [2] [2] [4] [4] [4] [4] [4] [6] [6] [6] [16] [6] [6] [6] [4] [15]

and C6

TC: ton of sugarcane stalks as harvested in the field (wet basis). Reaction time was assumed with basis on industrial operation of disk refiners instead of the two-step lab-scale process adjusted by Laurito-Friend et al. [6]. c Units of enzyme used according to Laurito-Friend et al. [6]. This value was converted in enzyme flow considering the specific activity of commercial enzyme preparation and its protein content. a

and C6

b

and C6 and C6

316

Biomass and Bioenergy 119 (2018) 314–321

F.M. Mendes et al.

will be maintained to allocate the yearly expenses associated with sugarcane, chemical inputs, maintenance, human labor, capital costs and so on. The cost per liter is then obtained by dividing the total yearly expenses by the total amount of ethanol produced in the year. Once the 1G cost is obtained, the last step is to allocate the additional costs from 1G2G process to 2G ethanol [17,18] in order to identify costs exclusively related to 2G ethanol, following the expression Eq. (1) below:

Table 3 Pricesa of sugarcane stalks and straw. Experimental Hybrids

Stalks US$/TCb

Straw US$/ton

H146 H140 H89 H58

32.80 20.30 18.04 20.98

20.08 16.18 15.40 16.39

a b

C1G2G = P1GC1G+P2GC2G

Exchange rate considered US$ 1.00 = R$3.35 (December 2016). TC means ton of sugarcane stalks as harvested in the field (wet basis).

where: P1G: percentage share of first generation on the total 1G2G ethanol production (%); P2G: percentage share of second generation on the total 1G2G ethanol production (%); C1G: first generation ethanol production cost ($/liter); C2G: second generation ethanol production cost ($/liter); C1G2G: integrated first and second-generation ethanol production cost ($/liter)

Table 4 Basic parameters used in the techno-economic analysis. Parameter

Valuea

Reference

Project lifetime Construction and start-up Depreciation (linear) Tax rate (income and social contributions) Minimum acceptable rate of return (MARR) NaOH costb Na2SO3 costb Enzyme cost Electricity (producer price) Anhydrous ethanol (producer price)

25 years 2 years 10% per year 34% 12% per year US$0.62/kg US$2.19/kg US$ 0.13/L US$ 57.89/MWh U$$ 0.50/L

[17] [17] [17] [18] [17] [19] [20] [18] [21,22] [23]

a b

(1)

Considering the significant impact of additional chemicals to the 2G process, a parametric sensitivity analysis was performed to evaluate to what extent the cost of sodium sulfite and sodium hydroxide would influence the IRR values under selected scenarios. The sensitivity analysis was performed separately for sodium sulfite and sodium hydroxide costs.

Exchange rate considered US$ 1.00 = R$3.35 (December 2016). Parametric sensitivity analysis was applied to chemicals cost.

facility is controlled by the same company. In order to calculate the yearly revenues and other industrial operating costs (e.g., chemicals, labor, maintenance), the outputs obtained from mass and energy balances were used in the cash flow analysis. These values, along with investment costs (calculated based on VSB databank and Jones et al. [16] for mechanical refining equipment), were the basis for determining the cash flow. The IRR, NPV and ethanol production cost for each scenario were calculated and compared, taking into account the economic assumptions summarized in Table 4 [2]. The reference date for the economic analysis is December 2016. The ethanol production costs were calculated based on the allocation procedure detailed in Bonomi et al. [2]. Firstly, a separate scenario of a 1G biorefinery whose main products are 1G ethanol and bioelectricity is assessed. Such 1G process is simulated to represent 1G2G scenario before the introduction of the 2G process. All operating and capital expenses of the 1G plant are then allocated according to the ethanol and electricity's share on the total revenues. This proportion

3. Results and discussion 3.1. Technical results Results for the process simulation regarding anhydrous ethanol and surplus electricity production in 1G2G biorefineries are displayed in Tables 5 and 6. Table 5 summarizes data for mild pretreatment, whereas Table 6 presents data for severe pretreatment conditions. Independently on the sugarcane hybrid, the highest ethanol production levels were obtained in the severe pretreatment scenarios (Table 6), which agrees with production of less recalcitrant lignocellulose, resulting in higher 2G sugar yields. The highest ethanol production level (124.8 L/TC) was achieved with H89 pretreated at the severe condition, the longest enzymatic hydrolysis period (72 h) and C5/C6 fermentation. In contrast, the lowest ethanol production (82.5 L/TC) was achieved with the most recalcitrant hybrid (H140) [6], pretreated at the mild condition, the shortest enzymatic hydrolysis period (24 h) and C6/C12

Table 5 Simulation results for the production of 1G2G ethanol and surplus electricity in the scenarios with low severity pretreatment (2.5% NaOH and 5% Na2SO3). Sugarcane Hybrids

H58

Reaction time -Enzymatic hydrolysis

24 h 72 h 24 h 72 h H89 24 h 72 h 24 h 72 h H140 24 h 72 h 24 h 72 h H146 24 h 72 h 24 h 72 h Average values for mild pretreatment a

Liquor fermentation

Anhydrous Ethanol (L/ TC)

Anhydrous Ethanol (L/ haa)

Surplus electricity (kWh/TC)

Surplus electricity (kWh/haa)

C6 C6 C5 C5 C6 C6 C5 C5 C6 C6 C5 C5 C6 C6 C5 C5

86.5 86.8 105.9 106.8 96.3 105.9 109.5 116.3 82.5 84.2 104.4 106.7 93.7 94.5 106.6 107.0 99.6

7344 7376 8992 9068 9752 10725 11084 11774 7261 7410 9192 9394 4929 4975 5609 5631 8157

115.0 116.0 104.1 105.6 108.6 93.0 109.5 104.4 112.7 109.4 111.3 108.5 74.9 74.4 78.4 77.9 100.2

9766 9850 8841 8969 10995 9418 11081 10572 9925 9635 9795 9555 3943 3915 4123 4098 8405

and C6 and C6

and C6 and C6

and C6 and C6

and C6 and C6

ha: hectare. 317

Biomass and Bioenergy 119 (2018) 314–321

F.M. Mendes et al.

Table 6 Simulation results for the production of 1G2G ethanol and surplus electricity in the scenarios with high severity pretreatment (5% NaOH and 10% Na2SO3). Sugarcane Hybrids

Reaction time –Enzymatic hydrolysis

H58

24 h 72 h 24 h 72 h H89 24 h 72 h 24 h 72 h H140 24 h 72 h 24 h 72 h H146 24 h 72 h 24 h 72 h Average values for severe pretreatment

Liquor fermentation

Anhydrous Ethanol (L/ TC)

Anhydrous Ethanol (L/ha)

Surplus electricity (kWh/ TC)

Surplus electricity (kWh/ ha)

C6 C6 C5 C5 C6 C6 C5 C5 C6 C6 C5 C5 C6 C6 C5 C5

98.1 101.7 113.8 117.9 106.8 110.5 120.3 124.8 92.7 95.6 111.7 115.6 99.8 99.2 109.9 109.8 108.0

8330 8634 9662 10016 10809 11187 12183 12629 8161 8419 9830 10176 5250 5221 5780 5775 8879

83.1 80.1 90.4 88.1 80.8 78.5 91.0 89.4 87.8 84.6 93.0 91.2 66.1 66.0 72.9 72.9 82.2

7056 6800 7674 7483 8184 7950 9209 9054 7726 7452 8190 8030 3480 3475 3838 3835 6840

and C6 and C6

and C6 and C6

and C6 and C6

and C6 and C6

pretreatment, reinforcing alkaline-sulfite CTM pretreatment costs as a process bottleneck (Fig. 3). Several of the positive IRR values were achieved for H58 (5 scenarios) and H89 (5 scenarios) that are low recalcitrant sugarcane hybrids [6,12], which can take advantage of mild pretreatment. Another relevant information was that among 12 scenarios with positive IRR, 8 considered C5eC6 co-fermentation, suggesting positive contributions to the IRR of C5 fermentation. Better understanding of factors influencing IRR and NPV values was obtained when 1G, 2G, and 1G2G ethanol production costs were evaluated separately (Fig. 4). Breakdown of 1G2G ethanol production costs also helped to discriminate individual economical contributions to the final ethanol costs in each evaluated scenario (Fig. 5). Fig. 4 data highlight that, independently on the evaluated scenario, 2G ethanol production costs are significantly higher than costs calculated for 1G and 1G2G integrated biorefineries. The ethanol production costs achieved for 1G biorefineries varied from US$ 0.337 to US$ 0.477 per liter, depending on the sugarcane hybrid. H89 and H146 presented the lowest and the highest 1G ethanol production cost, respectively, which was already explained by varied sugarcane field productivity and sucrose contents as previously reported [10]. In contrast, 2G ethanol production costs varied from US$ 0.889 to US$ 3.461 per liter, which is in agreement with previous evaluations regarding 2G ethanol production [23,28]. Independent on the sugarcane hybrid, a general trend for 2G ethanol production costs was that the lowest costs were achieved at severe pretreatment and using C5C6 co-fermentation (Fig. 4). Severe pretreatment, despite requiring high chemicals load, provide less recalcitrant substrates with high sugars yield and, consequently, more ethanol. C5C6 co-fermentation always provided the lowest values for 2G and also 1G2G ethanol production costs. This observation reinforces the importance of co-fermentation of C5C6 for ethanol production, as indicated by several authors [24-27]. For example, Macrelli et al. [27] evaluated different configurations for a 1G2G integrated ethanol biorefinery. Simultaneous saccharification and fermentation (SSF) and time-separated hydrolysis and fermentation (tSHF) were investigated, as well as different uses for pentoses (biodigestion and fermentation to ethanol). The authors observed that tSHF configuration and pentoses fermentation to ethanol showed higher potential for reduction on minimum ethanol selling price. In another work, Zhang et al. [26] observed that the yield of ethanol fermentation can be increased 25% with the use of the xylose to produce ethanol. Regarding sugarcane hybrids, the low recalcitrant materials (H58 and H89) [6,12] presented the lowest 2G ethanol production costs. In contrast, the most recalcitrant hybrid (H140) presented the highest 2G ethanol production cost (Fig. 4).

fermentation. Surplus electricity followed an inverse behavior of ethanol production mainly because mild pretreatment and short enzymatic hydrolysis periods resulted in less 2G sugar yields [6], leading to higher residual solids amounts that were directed to heat and power generation (Table 5). Furthermore, the conditions without C5 fermentation provided less ethanol and, consequently, lower energy consumption for distillation, providing more steam for the turbines, which increases electricity production. Comparison between scenarios without C5 fermentation and scenarios with C5eC6 co-fermentation showed, for both pretreatment conditions, higher increase in ethanol production for H140 hybrid, followed by H58, with lower increases observed for H89 and H146. This behavior can be related to the higher hemicellulose and fiber content of hybrids H140 and H58, since higher hemicellulose content results in higher xylose availability for conversion to ethanol, although xylan hydrolysis yields are also important. H146 hybrid is an outlier because, despite its high hemicellulose content, its low fiber fraction limits total ethanol production. Comparison of ethanol production levels from 1G [10] and 1G2G (Tables 5 and 6) biorefineries clearly shows increased ethanol levels in 1G2G processes (either comparing ethanol production in L/TC or L/ha). The overall increase in ethanol production depended on the pretreatment severity as well as on enzymatic hydrolysis and C5eC6 co-fermentation steps. In summary, for severe pretreatment, 72 h enzymatic hydrolysis period, and C5eC6 co-fermentation, ethanol production levels increased by 33%, 31%, 35% and 19% for H58, H89, H140 and H146, respectively, when 1G alone and 1G2G biorefineries were compared. 3.2. Techno-economic analysis Despite increased ethanol production by integrating 1G2G biorefineries, techno-economic analysis of the integrated processes indicated IRR values lower than MARR (12%) in all evaluated scenarios (Fig. 3). In some cases, IRR values were not calculated, since revenues were lower than expenses. The highest IRR value (7.04%) was achieved for the less recalcitrant hybrid H89 [6,12] pretreated at the mild condition, suggesting pretreatment costs as a bottleneck for economical feasibility. Even at the highest IRR, a negative NPV was calculated (for a 12% discount rate), which demonstrated that under considered scenarios, limited economical viability is achieved. Deeper evaluation of the variables affecting IRR values helped to discriminate processes bottlenecks. For example, among the 32 evaluated scenarios, only 12 presented positive IRR values: 5 with H58, 5 with H89 and 2 with H146. Eight of these scenarios corresponded to low severity 318

Biomass and Bioenergy 119 (2018) 314–321

F.M. Mendes et al.

Fig. 3. Internal rate of return (IRR) and Net present values (NPV) for each sugarcane biorefinery scenario. Mild and severe pretreatments corresponds to 2.5% NaOH/ 5% Na2SO3 and 5% NaOH/10% Na2SO3, respectively. (Bars) IRR 1G2G values; (circles) NPV values.

pretreatment, 72 h of enzymatic hydrolysis, and C5eC6 co-fermentation). The maximum and the minimum prices of these chemicals in the sensitivity analysis referred to Brazil [19,20] and USA [29] markets, respectively. Comparing Na2SO3 and NaOH prices, sulfite was the variable with the highest impact the IRR values (Fig. 6). Fig. 6A showed that variation in NaOH price from US$0.39 to US$0.63 had a low influence on the IRR because its lower amount in pretreatment and its lower price compared to Na2SO3. On the other hand, Na2SO3 price had strong influence on IRR values, especially for the more severe pretreatment (5% NaOH and 10% Na2SO3) (Fig. 6B). When the minimum Na2SO3 price is considered, the higher levels of ethanol production under severe pretreatment conditions can compensate for overall process costs, since IRR values were higher than MARR of 12% per year. If these values could be a long-term price in a cash flow exercise, the biorefinery would become economically viable using a sugarcane with the characteristics of the hybrid H89. A global evaluation of the process simulation and their economic

Breakdown of 1G2G ethanol production costs (Fig. 5) reveals that sugarcane and pretreatment chemicals presented significant and variable contributions to the overall process costs. Sugarcane costs (presented in Table 3) are mainly related to the field productivity of each hybrid as well as their fiber and sucrose contents as already discussed in a previous work [10]. Costs of pretreatment chemicals, especially Na2SO3, were significant in the overall process costs. The original price considered for Na2SO3 (US$2.19/kg) was based on current Brazilian market [20]. However, significantly lower costs for Na2SO3 (US$0.34/ kg) have been recently reported with basis on the North American market [29]. This is relevant because pretreatment with high sulfite loading led to higher ethanol production (Table 6); however, the requirement of a high chemical load along with high sulfite price had a large influence on ethanol production cost (Fig. 5). A parametric sensitivity study was performed to evaluate the impact of Na2SO3 and NaOH prices on IRR values of the best-identified scenarios (hybrid H89, different NaOH and Na2SO3 loadings in

Fig. 4. Ethanol production costs in each sugarcane biorefinery scenario allocated by 1G, 2G and 1G2G. Mild and severe pretreatments corresponds to 2.5% NaOH/5% Na2SO3 and 5% NaOH/10% Na2SO3, respectively. 319

Biomass and Bioenergy 119 (2018) 314–321

F.M. Mendes et al.

Fig. 5. 1G2G ethanol cost breakdown for the different sugarcane biorefinery scenarios. Mild and severe pretreatments corresponds to 2.5% NaOH/5% Na2SO3 and 5% NaOH/10% Na2SO3, respectively.

marketable product [31]. 4. Conclusions The techno-economic analysis performed in this study presented the main bottlenecks for 1G2G ethanol production from sugarcane biomass using alkaline sulfite pretreatment. Severe pretreatment proved efficient to increase ethanol production but required high chemicals load, decreasing IRR values for the proposed biorefinery. Considering diverse sugarcane hybrids, the most important characteristics of sugarcane to achieve better economic results included high field productivity (meaning low sugarcane cost) and low recalcitrance after pretreatment (meaning high sugar and ethanol yields). Sodium sulfite used in the pretreatment step represented a significant fraction of the final 1G2G cost. Sensitivity analysis helped to evaluate the effect of sodium sulfite costs on the biorefinery IRR. For the sugarcane hybrid with the best characteristics (H89), processed in a scenario with 72 h of enzymatic hydrolysis and C5 and C6 co-fermentation, maximum sodium sulfite prices to reach MARR (12% IRR) were US$ 0.66/kg and US$ 0.47/kg for severe and mild pretreatments, respectively. Acknowledgements This work was supported by Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP, contract number 2014/06923-6). F.M. Mendes thanks FAPESP for her post-doctoral fellowship (2015/107560). This study was also financed in part by the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior - Brasil (CAPES) Finance Code 001.

Fig. 6. Sensitivity analyses for the impact of NaOH (A) and Na2SO3 (B) prices on the internal rate of return (IRR) for two selected scenarios for biorefining sugarcane hybrid H89. Scenarios included mild and severe pretreatments, 72 h of enzymatic hydrolysis, and C5 and C6 co-fermentation. The maximum and the minimum prices of NaOH and Na2SO3 used in this analysis referred to Brazil [19,20] and USA [29] markets, respectively.

Appendix A. Supplementary data analysis indicated that for integrated 1G2G ethanol production, the best sugarcane characteristics were identified in the H89. Regarding process conditions, severe pretreatment, long enzymatic hydrolysis periods and C5C6 co-fermentation were useful to provide more ethanol per ton of sugarcane. However, the economic viability would be attained only in scenarios in which Na2SO3 prices approximates to US$0.34/kg. With this regard, an interesting possibility would include in situ Na2SO3 production from SO2 and NaOH [30], aiming to reduce production costs and, consequently, increase IRR values. Another possible way to add value to the currently proposed biorefinery would include recovery of sulfonated lignin produced during the pretreatment step, which is a

Supplementary data to this article can be found online at https:// doi.org/10.1016/j.biombioe.2018.10.005. References [1] S. Farzad, M.A. Mandegari1, M. Guo, K.F. Haigh, N. Shah, J.F. Görgens, Multiproduct biorefineries from lignocelluloses: a pathway to revitalization of the sugar industry? Biotechnol. Biofuels 10 (2017) 8. [2] A. Bonomi, O. Cavallet, M.P.D. Cunha, M.A.P. Lima, Virtual biorefinery: an optimization strategy for renewable carbon valorization, Green Energy and Technology, Springer International Publishing, Switzerland, 2016.

320

Biomass and Bioenergy 119 (2018) 314–321

F.M. Mendes et al.

[17] M.D.B. Watanabe, L.G. Pereira, M.F. Chagas, M.P. Cunha, C.D.F. Jesus, A. Souza, E.C. Rivera, R. Maciel Filho, O. Cavalett, A. Bonomi, Sustainability assessment methodologies, in: A. Bonomi, O. Cavalett, M.P. Cunha, M.A.P. Lima (Eds.), Virtual Biorefinery-an Optimization Strategy for Renewable Carbon Valorization, Springer International Publishing, Basel, 2016, pp. 155–188. [18] T.L. Junqueira, M.F. Chagas, V.L.R. Gouveia, M.C.A.F. Rezende, M.D.B. Watanabe, C.D.F. Jesus, O. Cavalett, A.Y. Milanez, A. Bonomi, Techno-economic analysis and climate change impacts of sugarcane biorefineries considering different time horizons, Biotechnol. Biofuels 10 (50) (2017) 1–12. [19] QUÍMICA E DERIVADOS. < http://www.quimica.com.br/category/revista/ > . Accessed on 18 May 2017. [20] MDIC -Ministério do Desenvolvimento, Indústria e Comércio Exterior. Álcool Combustível. ALICEWEB, http://www.desenvolvimento.gov.br/sitio/interna/ interna.php?area=2&menu=999, (2017) , Accessed date: 5 May 2017. [21] CCEE - Câmara de comercialização de energia elétrica, Energia de Reserva, http:// www.ccee.org.br/portal/faces/oquefazemos_menu_lateral/energia_reserva?_ afrLoop=3451532015059570#%40%3F_afrLoop%3D3451532015059570%26_ adf.ctrl-state%3Do1ezahuld_218, (2017) , Accessed date: 5 May 2017. [22] MME - Minas e Energia, (2013) http://www.mme.gov.br/programas/leiloes_de_ energia/menu/leiloes_realizados. Accessed on 05 May 2017. [23] CEPEA, The CEPEA registers. Center for advanced studies on applied economics (CEPEA), http://www.cepea.esalq.usp.br, (2018) , Accessed date: 10 March 2017. [24] Y. Chen, Development and application of co-culture for ethanol production by cofermentation of glucose and xylose: a systematic review, J. Ind. Microbiol. Biotechnol. 38 (2011) 581–597. [25] Z. Losordo, J. John McBride, J.V. Rooyen, K. Wenger, D. Willies, A. Froehlich, I. Macedo, L. Lynd, Cost competitive second-generation ethanol production from hemicellulose in a Brazilian sugarcane biorefinery, Biofuels, Bioprod. Bioref. 10 (2016) 589–602. [26] Y. Zhang, C. Wnag, L. Wang, R. Yang, P. Hou, J. Liu, Direct bioethanol production from wheat straw using xylose/glucose cofermentation by coculture of two recombinant yeasts, J. Ind. Microbiol. Biotechnol. 44 (2017) 453–464. [27] S. Macrelli, M. Galbe, O. Wallberg, Effects of production and market factors on ethanol profitability for an integrated first and second generation ethanol plant using the whole sugarcane as feedstock, Biotechnol. Biofuels 7 (26) (2014). [28] E. Smullen, J. Finnan, D. Dowling, P. Mulcahy, Bioconversion of switchgrass: identification of a leading pretreatment option based on yield, cost and environmental impact, Renew. Energy 111 (2017) 638–645. [29] C.A. de Assis, C. Houtman, R. Phillips, E.T. Bilek, O.J. Rojas, L. Pal, M.S. Peresin, H. Jameel, R. Gonzalez, Conversion economics of forest biomaterials: risk and financial analysis of CNC manufacturing, Biofuels, Bioprod. Biorefin. 11 (2017) 682. [30] A. Cavka, C. Martín, B. Alriksson, M. Mörtsell, L.J. Jönsson, Techno-economic evaluation of conditioning with sodium sulfite for bioethanol production from softwood, Bioresour. Technol. 196 (2015) 129–135. [31] T. Aro, P. Fatehi, Production and application of lignosulfonates and sulfonatedlignin, ChemSusChem 10 (2017) 1861–1877.

[3] M.O.S. Dias, M.P. Cunha, C.D.F. Jesus, G.J.M. Rocha, J.G.C. Pradella, C.E.V. Rossell, R. Maciel Filho, A. Bonomi, Second generation ethanol in Brazil: can it compete with electricity production? Bioresour, Technol. 102 (2011) 8964–8997. [4] M.O.S. Dias, T.L. Junqueira, C.D.F. Jesus, C.E.V. Rossell, R. Maciel Filho, A. Bonomi, Improving second generation ethanol production through optimization of first generation production process from sugarcane, Energy 43 (2012) 246–252. [5] M.O.S. Dias, A.V. Ensinas, S.A. Nebra, R. Maciel Filho, C.E.V. Rossell, M.R.W. Maciel, Production of bioethanol and other bio-based materials from sugarcane bagasse: integration to conventional bioethanol production process, Chem. Eng. Res. Des. 87 (2009) 1206–1216. [6] D.F. Laurito-Friend, F.M. Mendes, F.M. Reinoso, A. Ferraz, A.M.F. Milagres, Sugarcane hybrids with original low lignin contents and high field productivity are useful to reach high glucose yields from bagasse, Biomass Bioenergy 75 (2015) 65–74. [7] F.M. Mendes, G. Siqueira, W. Carvalho, A. Ferraz, A.M.F. Milagres, Enzymatic hydrolysis of chemithermomecanically pretreated sugarcane bagasse and two experimental samples with reduced initial lignin, Biotechnol. Prog. 27 (2011) 395–401. [8] CGEE, Química Verde No Brasil: 2010-2030. Brasília, DF, Centro de Gestão e Estudos Estratégicos, 2010. [9] H.R. Ghatak, Biorefineries from the perspective of sustainability: feedstocks, products, and processes, Renew. Sustain. Energy Rev. 15 (2011) 4042–4052. [10] F.M. Mendes, M.O.S. Dias, A. Ferraz, A.M.F. Milagres, J.C. Santos, A. Bonomi, Techno-economic impacts of varied compositional profiles of sugarcane experimental hybrids on a biorefinery producing sugar, ethanol and electricity, Chem. Eng. Res. Des. 125 (2017) 72–78. [11] M.E. Loureiro, M.H.P. Barbosa, F.J.P. Lopes, F.O. Silvério, Sugarcane breeding and selection for more efficient biomass conversion in cellulosic ethanol, in: M.S. Buckeridge, G.H. Goldman (Eds.), Routes to Cellulosic Ethanol, Springer, New York, 2011, pp. 199–239. [12] F. Masarin, D.B. Gurpilhares, D.C.F. Baffa, M.H.P. Barbosa, W. Carvalho, A. Ferraz, A.M.F. Milagres, Chemical composition and enzymatic digestibility of sugarcane clones selected for varied lignin contents, Biotechnol. Biofuels 4 (2011) 55. [13] T.H.F. Costa, F. Masarin, T.O. Bonifácio, A.M.F. Milagres, A. Ferraz, The enzymatic recalcitrance of internodes of sugar cane hybrids with contrasting lignin contents, Ind. Crop. Prod. 51 (2013) 202–211. [14] O. Cavalett, T.L. Junqueira, M.O.S. Dias, C.D.F. Jesus, P.E. Mantellatto, M.P. Cunha, H.C.J. Franco, T.F. Cardoso, R. Maciel Filho, C.E.V. Rossell, A. Bonomi, Environmental and economic assessment of sugarcane first generation biorefineries in Brazil, Clean Technol, Environ. Pol. 14 (3) (2012) 399–410. [15] A.M. Petersen, M.C. Aneke, J.F. Görgens, Techno-economic comparison of ethanol and electricity coproduction schemes from sugarcane residues at existing sugar mills in Southern Africa, Biotechnol. Biofuels 7 (105) (2014). [16] B.W. Jones, R. Venditti, S. Park, H. Jameel, Comparison of lab, pilot, and industrial scale low consistency mechanical refining for improvements in enzymatic digestibility of pretreated hardwood, Bioresour. Technol. 167 (2014) 514–520.

321