Bioresource Technology 153 (2014) 108–115
Contents lists available at ScienceDirect
Bioresource Technology journal homepage: www.elsevier.com/locate/biortech
Process energy comparison for the production and harvesting of algal biomass as a biofuel feedstock Matthew K. Weschler a, William J. Barr b, Willie F. Harper c, Amy E. Landis b,⇑ a
Department of Civil and Environmental Engineering, University of Pittsburgh, 3700 O’Hara Street, Pittsburgh, PA 15261, United States School of Sustainable Engineering and the Built Environment, Global Institute of Sustainability, Arizona State University, ISTB4 781 E Terrace Road, Tempe, AZ 85287, United States c Department of Systems and Engineering Management, Air Force Institute of Technology, 2950 Hobson Way, WPAFB, OH 45433, United States b
h i g h l i g h t s Energy demand of 122 microalgae biomass production scenarios were compared. Choice of harvesting technology affected energy demand of other phases. Raceway ponds, settling, and chamber ﬁlter press consumed the least energy. Total energy demand for biomass production depends on ﬁnal concentration.
a r t i c l e
i n f o
Article history: Received 31 July 2013 Received in revised form 30 October 2013 Accepted 7 November 2013 Available online 14 November 2013 Keywords: Microalgae Biofuel Harvesting Energy demand
a b s t r a c t Harvesting and drying are often described as the most energy intensive stages of microalgal biofuel production. This study analyzes two cultivation and eleven harvest technologies for the production of microalgae biomass with and without the use of drying. These technologies were combined to form 122 different production scenarios. The results of this study present a calculation methodology and optimization of total energy demand for the production of algal biomass for biofuel production. The energetic interaction between unit processes and total process energy demand are compared for each scenario. Energy requirements are shown to be highly dependent on ﬁnal mass concentration, with thermal drying being the largest energy consumer. Scenarios that omit thermal drying in favor of lipid extraction from wet biomass show the most promise for energy efﬁcient biofuel production. Scenarios which used open ponds for cultivation, followed by settling and membrane ﬁltration were the most energy efﬁcient. Ó 2013 Elsevier Ltd. All rights reserved.
1. Introduction The United States is largely dependent on non-renewable liquid fuels to meet increasing energy demand. In recent years concerns over energy security, fossil fuel depletion and greenhouse gas emissions have lead researchers to investigate the commercialization of renewable fuel sources. To encourage the production of renewable fuels in the United States, policy makers developed the Energy Independence and Security Act (EISA) of 2007 and Renewable Fuel Standard (RFS) to increase vehicle fuel economy, energy savings, and energy security (EISA, 2007). Algal biofuels can contribute to the advanced biofuels volumetric goals set forth by EISA through biomass based biodiesel and ethanol production. Despite the potential of algae biofuels as a renewable energy ⇑ Corresponding author. Address: School of Sustainable Engineering and the Built Environment, Global Institute of Sustainability, Arizona State University, 375 ISTB4 781 E Terrace Road, Tempe, AZ 85287, United States. Tel.: +1 (480) 965 2975. E-mail address: [email protected]
(A.E. Landis). 0960-8524/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.biortech.2013.11.008
source, a number of factors pose challenges to their commercialization. Three major factors include (1) high water demand during algae cultivation, (2) high energy requirements and mineral phosphorus depletion associated with fertilizer consumption, and (3) low energy return on investment (EROI) due to the high-energy requirements associated with the harvesting and drying of the biomass feedstock (Hunter-Cevera et al., 2012). This study will focus on the process energy consumption associated with microalgae biomass feedstock production. Energy return on investments (EROI) between 0.13 and 3.33 have been estimated in the literature for the production of algal biofuel using open pond cultivation systems (Brentner et al., 2011; Clarens et al., 2010; Hunter-Cevera et al., 2012; Sander and Murthy, 2010; Stephenson et al., 2010). This wide range of values is due to differences in the choice of ﬁnal products and production scenarios included in each study. Important factors that affect the EROI include: (1) sources of carbon dioxide (industrially produced, ﬂue gas), (2) sources of nutrients (industrially produced, wastewater), (3) product and coproduct allocation (electricity, nutrient
M.K. Weschler et al. / Bioresource Technology 153 (2014) 108–115
recovery, bioethanol, biogas), (4) extraction and processing (hexane/esteriﬁcation, supercritical methanol, lysing method), and (5) biomass production process (open pond, photobioreactor, harvesting method). Research has shown that the energy consumption for algal biomass production, which includes cultivation, harvesting and drying phases, is a limiting factor for algal biofuel commercialization, and thus warrants detailed analysis (Lardon et al., 2009; Lohrey and Kochergin, 2012; Stephenson et al., 2010; Xu et al., 2011). Cultivation options include photobioreactors (PBR) and raceway ponds (RP). Multiple studies have been conducted to determine the feasibility of commercialization of each mode of cultivation. The two most common PBR designs for algae cultivation are ﬂat plate and tubular PBRs. Tubular PBRs however, are too energy intensive to compete with RPs and ﬂat-plate PBRs (Jorquera et al., 2010). Drying has been shown to be the most energy intensive regardless of the technology selected (Lardon et al., 2009; Lohrey and Kochergin, 2012; Xu et al., 2011). Solar drying has been considered (Show et al., 2013), but the data is limited and how this process will affect lipid recovery and fuel conversion is unknown and is therefore excluded from this study. The high cost of drying has led researchers to consider other fuel conversion methods such as wet lipid extraction and supercritical extraction, (Brentner et al., 2011; Yoo et al., 2012). Harvesting has also been shown to be an energy intensive step of algae biofuel production (Soratana et al., 2012). While the current number of cultivation and drying methods are limited, there are far more options for harvesting. Show et al. (2013) discuss recent advances in harvesting and drying technologies for biofuel production for a large number of processes. They consider sedimentation, air ﬂotation, and electroﬂotation/coagulation technologies. The air ﬂotation and electro techniques are more energy intensive than the sedimentation methods but the coagulation methods could negatively affect the biomass quality. Centrifugation and ﬁltration can be used to further concentrate the microalgae. Both methods effectively dewater algae to greater than 10% biomass (w/w) and in some cases greater than 20% (w/w). Filtration methods require signiﬁcant maintenance, such as ﬁlter cleaning and replacement. Centrifugation methods are very efﬁcient but energy intensive. Despite the large number of currently available harvesting methods, most studies have only evaluated ﬁve or less harvesting scenarios. A number of studies assess the algal biomass production process in conjunction with biofuel production. Lohrey and Kochergin (2012) considered ﬁve different harvesting technologies using two different production scenarios prior to drying. Lardon et al. (2009) considered two harvesting technologies and one production scenario. The authors explicitly avoided centrifugation, because of its high energy demand. Instead of exploring different harvesting technologies, they compared dry lipid extraction based on the same established process for soybeans and wet lipid extraction to avoid the exorbitant energy consumption associated with thermal drying. Xu et al. (2011) considered three harvesting technologies and one scenario before thermal drying for their dry route analysis and the same harvesting scenario without drying for their wet lipid extraction analysis. Stephenson et al. (2010) selected two harvesting methods and one production scenario with no drying. The goal of this study is to perform a comprehensive process energy analysis of harvesting technologies for potential use in industrial-scale algal biofuel production. In this study we explore the use of multiple technologies and scenarios to reach desired concentrations. We consider cultivation, harvesting, and drying to demonstrate the interdependency of these three phases based on their energy requirements, solids concentrating potential, and biomass recovery efﬁciencies.
2. Methodology A process model was constructed to compare 122 different algal biomass production scenarios using different combinations of technologies. The Supplemental information explains how these technologies were combined to form the production scenarios. The functional unit was deﬁned as 1000 kg algae biomass. Process energy (kWh) inputs were calculated for each unit process. Each of the 122 scenarios was divided into three groups based on the ﬁnal concentration of the algae. Parameters for these groups are summarized in Table 1, and the biomass production methods are listed in Table 2. 2.1. Scope of the analysis Only process energy consumption for biomass production was considered in calculations. Process energy for biomass production includes energy used directly by cultivation, harvest, and drying technology and excludes energy required for raw material extraction, electricity generation and distribution, transportation, infrastructure, maintenance, and ﬁnal waste disposal, as would be included in a traditional cradle-to-grave life cycle analysis (LCA). Energy required for the production of inputs into the microalgae biomass production process, including fertilizers, carbon dioxide, ﬂocculants, and polymer ﬁlters were excluded from the calculation of energy demand. The scope of analysis was limited in this regard to focus solely on the unit processes used for biomass production and avoid the uncertainty associated with upstream and downstream process options. If the energy required for nutrient production was included, for example, the energy demand for cultivation might be overestimated, since up to 73% of the energy required for the production of nutrients can be recovered if anaerobic digestion of the residual algal biomass is chosen as a downstream process option (Brentner et al., 2011). Future LCAs can then use the results presented in Section 3 of this study to model energy requirements for algae biomass production, while determining for themselves the utility of energy recovery using anaerobic digestion in downstream processing. 2.2. Biomass production process The production of algal biomass can be described as a series of unit processes which amplify algal solids concentration and are summarized as follows: (1) cultivation, in which the algal biomass is grown to a dilute concentration 0.1–0.26% (w/w) (2) primary harvest (thickening), in which the concentration is increased to 1.5–10.0% (w/w), (3) secondary harvest (dewatering), in which the solid content is increased to 12.0–27.0% (w/w), (4) thermal drying, in which unbound water is removed from the biomass (Greenwell et al., 2010; Mohn, 1980; Pulz, 2001; Shelef et al., 1984; Uduman et al., 2010). The algae production process is summarized in Fig. 1.
Table 1 Summary of the three groups of scenarios.
Low biomass concentration, wet harvest High biomass concentration, wet harvest Dry harvest
Concentration w/w (%)
Final unit process
# of Scenarios
Primary harvest Secondary harvest Thermal drying
M.K. Weschler et al. / Bioresource Technology 153 (2014) 108–115
Table 2 Summary of variables and values used for analysis. The best-case values from the references listed were used in calculation.
Cultivationc Primary harvestd
Volumetric energy consumption (kWh/m3)
Efﬂuent concentration (w/w%)
Harvesting efﬁciency (%)
Raceway Pond Flat-plate PBR Settling with ﬂocculant Settling without ﬂocculant Micro strainer Electrocoagulation and electroﬂotation Vibrating screen ﬁlter Dissolved air ﬂotation Tangential ﬂow ﬁltration Decanter centrifuge Self-cleaning plate separator Belt ﬁlter press Chamber ﬁlter press Mechanical vapor recompression dryer
RP PBR SF S MS E&E VSF DAF TFF DC SPS BFP CFP MVR
0.3722 5.32 0.12 0.12 0.21 0.31 0.41 1.52 0.22 8.01 1.01 0.51 0.881 7922
0.1%2 0.26%1 3.0%2 3.0%2 1.5%1 5.0%2 6.0%1 10.0%2 10.0%2 22.0%1 12.0%1 18.0%1 27.0%1 90%a
100%a 100%a 99.0%1 75.5%b 89%a 96%1 89%a 99.9%1 89.0%1 95%a 95%a 89%a 89%a 100%a
Primary sources. Secondary sources. a Assumed values. b Values from this study. c Jorquera et al. (2010), Pulz (2001), Richmond and Cheng-Wu (2001) and Wiley et al. (2011). d Collet et al. (2011), Danquah et al. (2009), Edzwald (1993), Greenwell et al. (2010), Henderson et al. (2008), Jun et al. (2001), Mohn (1980), Poelman et al. (1997), Rossignol et al. (1999), Shelef et al. (1984), Uduman et al. (2010) and Wiley et al. (2011). e Mohn (1980). f Van Gemert (2009) and Xu et al. (2011). 2
.1 - .26%
1.5 -10% w/w
12 - 27% w/w
1 g mL1. The third variable was harvest efﬁciency, the mass percentage of algae retained between unit processes. The concentration of the culture mixture was assumed to remain constant, with a dynamic equilibrium existing between rate of dilution, rate of evaporative water loss, growth rate of the algae, and rate of biomass removal for harvesting.
TED ¼ Fig. 1. Algae Biomass Production Phases. The system boundary for this study includes the four phases associated with cultivating and concentrating algal biomass to be used as a feedstock for biofuel production. The secondary harvest and drying phases may be omitted if alternative downstream production methods, such as wet lipid extraction, are implemented.
Following the production process shown in Fig. 1, the biomass may then serve as the feedstock for the production of biofuels and their associated co-products. Many technologies exist for the synthesis of algae biofuels, including the extraction of oil from the algae cells followed by transesteriﬁcation or hydrogenation of algal lipids to produce biodiesel or green diesel (Brentner et al., 2011; Xu et al., 2011). Following cultivation, any unit process shown in Fig. 1 can be omitted, depending on the ﬁnal concentration of algae desired. The level of concentration of the algae after biomass production may impact the efﬁciency of downstream lipid extraction and fuel conversion. A low concentration of algae (10–30% w/w) saves energy in biomass production by omitting the energy intensive drying step required for a high concentration of algae (80–90% w/w), but increases the energy required for cell disruption and solvent evaporation during the subsequent lipid extraction step (Xu et al., 2011).
x¼4 X EDx
EDx ¼ Hx V x Ex
TED x EDx Hx
Total energy demand (kWh/1000 kg dry microalgae) Unit process, where 1 = cultivation, 2 = primary harvest, 3 = secondary harvest, and 4 = thermal drying Energy demand for unit process x (kWh/1000 kg dry microalgae) Multiplier for additional biomass for unit process x, to take into account biomass lost during downstream processing (unitless) Volume of slurry (for x = 1, 2, and 3) or water (for x = 4) in unit process x (m3) Volumetric energy consumption for unit process x (kWh m3)
Hx was calculated using Eq. S.1 in the Supplemental data and Vx was calculated using Eqs. S.2 (for x = 1, 2, and 3) and S.3 (for x = 4).
2.3. Energy consumption 2.4. Technologies Energy demand (ED) of unit processes and total energy demand (TED) were calculated using Eqs. (1) and (2). Three variables were used for calculating energy demand. The ﬁrst variable was volumetric energy consumption, measured as the energy required by each technology to process 1 m3 of algal slurry. The second variable was efﬂuent concentration from each unit process, as a percentage of total mass (w/w %). The density of dry algae was assumed to be
Data used as inputs to the calculations for energy demand were derived from the peer reviewed literature and the settling test described in the Supplemental data. Since values were derived from multiple unafﬁliated sources in the literature, several inconsistencies exist between the parameters of each study, including algae species, geographic location, source of water, and method of
M.K. Weschler et al. / Bioresource Technology 153 (2014) 108–115
ﬂocculation used prior to harvesting. These variations in study parameters will impact the calculation of energy demand in several ways. The efﬂuent concentration and harvest efﬁciency of all harvest methods depend on the size, density, and surface charge of the microalgae species. The pH of source water and method of ﬂocculation will affect the cohesive properties of the algae cells and ease of separation. Variations in temperature and solar irradiance between geographic locations will impact the biological growth and rheological characteristics of the algal slurry, thus impacting energy requirements for biomass production. Two methods of cultivation, seven primary harvest technologies, four secondary harvest technologies, and one thermal drying method were chosen for analysis. These methods and the data values used to calculate energy demand are summarized in Table 2. While attempts were made to use primary source data, secondary data was used if the original source could not be traced. In the absence of any source information, reasonable assumptions were made based on the performance of similar technologies. These assumptions were used to estimate the harvest efﬁciency of the micro strainer (MS), vibrating screen ﬁlter (VSF), decanter centrifuge (DC), self-cleaning plate separator (SPS), belt ﬁlter press (BFP), and chamber ﬁlter press (CFP), and are discussed further in Section 2.4.2. Many variables used in this analysis were reported as a range of values in the literature. In order to remain consistent in analysis, the best case values were used, including the lowest volumetric energy, the highest efﬂuent concentration, and the highest harvest efﬁciency. This study therefore represents a best case scenario. The authors do not conclude that this necessarily represents the most realistic scenario, given the often high degree of uncertainty with the performance of certain technologies. An alternative analysis is provided in the Supplemental data based on sensitivity and the range of values reported in the literature for each technology (Collet et al., 2011; Danquah et al., 2009; Henderson et al., 2008; Richmond and Cheng-Wu, 2001; Rossignol et al., 1999). 2.4.1. Cultivation Two cultivation methods were considered in this study that represent the two general approaches to produce algae, the raceway pond (RP) and ﬂat-plate photobioreactor (PBR). Volumetric energy consumption values are based on cultivation of the microalga Nannoschlorpsis sp., including energy consumed for gas/liquid circulation, mixing, and liquid cooling (Jorquera et al., 2010). For both cultivation methods, it was assumed no biomass would be lost (100% harvest efﬁciency). 2.4.2. Primary and secondary harvest Values for energy consumption for primary and secondary harvest were taken from technology studies and reviews in the literature. Descriptions and details of each of these technologies can be found in Table 2 and were primarily from Show et al. (2013), Uduman et al. (2010), Shelef et al. (1984) or Mohn (1988). Settling with and without the use of ﬂocculants were evaluated (S and SF, respectively). The best case value for volumetric energy consumption for the two settling methods were based on settling using a low lamella separator (Uduman et al., 2010). It was assumed that volumetric energy consumption and efﬂuent concentration values would be the same with and without ﬂocculant use. The best case value for harvest efﬁciency for settling with ﬂocculant use is based on the removal of Synedra acus sp. from reservoir water using 2.16 mg l1as Al (Al2(SO4)3 and 0.25 mg l1 cationic polymer C-599A (Jun et al., 2001). The harvest efﬁciency for settling without ﬂocculant use was determined using the settling test described in detail in the Supplemental data. Chlorella vulgaris sp. was grown under controlled conditions in glass jar PBRs before being harvested and allowed to settle undisturbed for 180 min. The
initial and ﬁnal optical densities of the algal culture were measured and used to calculate the harvest efﬁciency using Eq. S.4 in the Supplemental data. The energy consumption values for the MS, VSF, DC, SPS, BFP, and CFP were taken from the studies of Mohn (1980), using the alga species Coelastrum proboscideum. For effective use, the DC and BFP require a pre-concentration of 1.5% and 4% (w/w), respectively (Mohn, 1980, 1988; Show et al., 2013). Thus, in order to use either secondary harvesting technology, a primary harvest method must be used that produces the required concentration. The tangential ﬂow ﬁltration (TFF) method can achieve a harvest efﬁciency of up to 89% for the removal of mixed algae species from reservoir water using a 0.45 lm pore-size membrane (Petrusevski et al., 1995). To remain consistent, the MF, VSF, BFP, and CFP ﬁltration methods were assumed to have the same harvest efﬁciency. In practice, the harvest efﬁciencies of each of these ﬁltration methods will depend on the species of algae being harvested. Small algae species approaching bacterial dimensions are likely to pass through the ﬁlter pores as part of the permeate (Molina Grima et al., 2003). The harvest efﬁciency of centrifuge methods is dependent on the physical characteristics of the algae species, dimensions of the ﬁlter bowl, and magnitude of centrifugal acceleration (Molina Grima et al., 2003). An industrial supercentrifuge can achieve a harvest efﬁciency of 95% for N. oculata sp. using an applied acceleration factor of 13,000 g (Heasman et al., 2000). The DC and SPS centrifuge types were assumed to have this same harvest efﬁciency. The best case efﬂuent concentration for electrocoagulation and electroﬂotation (E&E) included the use of alum ﬂocculation (Shelef et al., 1984). The volumetric energy consumption and harvest efﬁciency for E&E are based on the elimination of mixed algae strains from water using 6 cathodes and 3 anodes, a distance between cathodes and anodes of 26.5 cm, voltage of 26.5 V, current of 1.0 A and exposure time of 75 min (Fon Sing et al., 2011; Poelman et al., 1997). The values for volumetric energy consumption using dissolved air ﬂotation (DAF) range from 1.5 to 20 kWh/m3, the largest difference of any technology used in this study (Wiley et al., 2011). The harvest efﬁciency was based on the removal of Chorella vulgaris sp. from reservoir water using a saturator gauge pressure of 483 kPa and 5 min ﬂocculation period prior to ﬂotation (Edzwald, 1993).
2.4.3. Thermal drying The energy consumed for the thermal drying process is dependent on the initial volume of water to be dried and the ﬁnal concentration to be achieved. In general, the energy required for drying is much higher than the latent heat of water, due to inherent inefﬁciency (Xu et al., 2011). The energy efﬁcient CarverGreenﬁeld Mechanical Vapor Recompression (MVR) dryer type was used in this study (Van Gemert, 2009).
3. Results and discussion 3.1. Best case scenarios Three groups of best case production scenarios were analyzed, which are summarized in Table 1: (1) low biomass concentration with wet harvest, where the ﬁnal unit process was primary harvest (see Fig. 1 for system boundaries), (2) high biomass concentration with wet harvest, where the ﬁnal unit process was secondary harvest, and (3) high biomass concentration with dry harvest, where the ﬁnal unit process was thermal drying. Total energy consumptions for each process in the three groups were compared.
M.K. Weschler et al. / Bioresource Technology 153 (2014) 108–115
3.1.1. Low biomass concentration with wet harvest The ﬁnal biomass concentration in the ‘low biomass concentration with wet harvest’ group of scenarios ranges from 3% to 10% (w/w). All 14 scenarios in this group are depicted in Fig. 2. Although energy consumptions for these scenarios are low compared to the high biomass concentration groups of scenarios (shown in Figs. 3 and 4), downstream lipid extraction efﬁciency may be hampered by excess water, due to reduced contact between algal cells in the wet suspension and hydrophobic extraction solvents (Yoo et al., 2012). To compensate for losses of lipids during extraction, more microalgae would need to be cultivated and harvested during biomass production, thus increasing the total energy demand. The concentration of algae leaving cultivation has an impact on the energy consumed during primary harvest, since lower concentrations correspond to a higher slurry volume necessary to achieve one functional unit. The PBR achieves 2.6 times higher efﬂuent concentration than the RP, reducing energy consumption during primary harvest by 62%. This does not offset the high initial energy demand of the PBR, however, which consumes 14 times as much energy as the RP per volume of output. The primary harvest efﬁciency has an impact on the energy consumed during cultivation, since higher biomass loss corresponds to a higher culture volume necessary to achieve one functional unit. Energy consumption during cultivation is more sensitive to changes in primary harvest efﬁciency when PBRs are used, as opposed to RPs. A 1% decrease in primary harvest efﬁciency corresponds to approximately 4 and 20 kWh increase in energy consumption per functional unit for the RP and PBR, respectively. Assuming an algal higher heating value (HHV) of 20 MJ/kg (Illman et al., 2000), and taking only process energy consumption into consideration, the ratio of embodied energy of the algae to process energy consumed for PBR scenarios all fall within 2–3, compared to ratios of 3–12 for RP scenarios. Improvements to the caloriﬁc content of algae and energy efﬁciency for PBRs would increase these ratios. For the production of 10 GJ of algal biodiesel using a ﬂat-plate PBR for cultivation, Brentner et al. (2011) estimated a total energy demand (TED) of 10.8 GJ. The ratio of useable energy from biodiesel to TED results in an energy return on investment (EROI) of 0.93. 3.1.2. High biomass concentration with wet harvest The ﬁnal biomass concentration in the ‘high biomass concentration with wet harvest’ group of scenarios shown in Fig. 3 ranges from 12% to 27% (w/w). Energy consumption for this group of scenarios is slightly higher than that of the low concentration, wet harvest group, but signiﬁcantly less than that of the dry harvest
kWh/1000 kg biomass
2500 2000 1500 1000
group. The ratio of embodied energy of the algae to process energy consumed for the ‘high biomass concentration with wet harvest’ group of scenarios fall between 1 and 3 for PBR scenarios and 2 and 11 for RP scenarios. For the production of green diesel and hydrogen gas using a RP for cultivation and centrifuge and mechanical dehydration for harvesting, Xu et al. (2011) estimated an EROI of 1.37. Eleven of the 54 scenarios from this group are shown in Fig. 3. These 11 scenarios were chosen to elucidate the variation in energetic interactions between different unit processes. Many scenarios exhibited similar trends to those shown in Fig. 3, and were therefore excluded from analysis. The results for all 54 scenarios are summarized in the Supplemental data in Fig. S5. Primary harvest is used to reduce the volume of slurry to be processed during secondary harvest, since primary harvest technologies are often less energy intensive than secondary harvest technologies. This introduces one more step in the production chain, however, which decreases harvest efﬁciency. A trade-off therefore exists between reducing the energy consumed during secondary harvest by reducing the slurry volume, and increasing the energy consumed during cultivation by decreasing the harvest efﬁciency. The interaction of unit processes may determine whether the inclusion of primary harvest is energetically beneﬁcial for the production of algal biomass. When the raceway pond (RP) is used for cultivation, for example, the scenario which uses settling without ﬂocculation (S) for primary harvest before using the chamber ﬁlter press (CFP) for secondary harvest consumes a total of 51% less energy than the same scenario which does not include primary harvest. Conversely, when the ﬂat-plate photobioreactor (PBR) is used for cultivation, the scenario which uses S for primary harvest and CFP for secondary harvest consumes 10% more energy than the same scenario which does not include primary harvest. When primary harvest efﬁciency is improved with the use of ﬂocculants for settling (SF), the two scenarios which use RP and PBR for cultivation consume 60% and 11% less energy, respectively. As discussed in Section 2.1, the upstream energy required for ﬂocculant production was not included in this study. The energetic beneﬁt of improving harvest efﬁciency through the use of ﬂocculants could be diminished if this parameter were included in the calculation of total energy demand. Other studies have shown, however, that the production of ﬂocculants typically represent less than 4% of the energy required for cultivation, and thus would have little impact on the energy savings from improved harvest efﬁciency (Brentner et al., 2011; Xu et al., 2011). It is energetically beneﬁcial, in some cases, to consume more energy during primary harvest if a higher output concentration can be achieved. The vibrating screen ﬁlter (VSF), for example,
Key: RP = Raceway Pond PBR = Flat-panel Photobioreactor SF = Settling with Flocculent S = Settling without Flocculent MS = Micro strainer E&E = Electrocoagulation and Electroflotation VSF = Vibrating Screen Filter DAF = Dissolved Air Flotation TFF = Tangential Flow Filtration
Fig. 2. Process energy demand for ‘low biomass concentration with wet harvest’ group of scenarios.
M.K. Weschler et al. / Bioresource Technology 153 (2014) 108–115 3500
Key: RP = Raceway Pond PBR = Flat-panel Photobioreactor N = No Primary Harvest SF = Settling with Flocculent S = Settling without Flocculent MS = Micro strainer VSF = Vibrating Screen Filter CFP = Chamber Filter Press DC = Decanter Centrifuge
kWh/1000 kg biomass
2500 2000 1500 1000 500 0
Fig. 3. Process energy demand for ‘high biomass concentration with wet harvest’ group of scenarios.
kWh/ 1000 kg biomass
8000 7000 6000 5000 4000 3000 2000 1000 0
Key: RP = Raceway Pond PBR = Flat-panel Photobioreactor N = No Primary Harvest SF = Settling with Flocculent S = Settling without Flocculent MS = Micro strainer E&E = Electrocoagulation and Electroflotation TFF = Tangential Flow Filtration SPS = Self-cleaning Plate Separator DC = Decanter Centrifuge CFP = Chamber Filter Press BFP = Belt Filter Press MVR = Mechanical Vapor Recompression Dryer
Fig. 4. Process energy demand for ‘high biomass concentration with dry harvest’ group of scenarios.
consumes twice as much energy per volume of output than the micro strainer (MS). The VSF produces a 4 times higher concentration, however, which reduces energy consumption during secondary harvest. When the RP is used for cultivation and the decanter centrifuge (DC) is used for secondary harvest, the process which uses the VSF consumes 16% less energy than the same process which uses the MS, as shown in Fig. 3. 3.1.3. High biomass concentration with dry harvest The ﬁnal biomass concentration of the ‘high biomass concentration with dry harvest’ group of scenarios was assumed to be 90%, the approximate solids content of soybeans before lipid extraction for the production of biodiesel (Sazdanoff, 2006). A sample of 14 scenarios is shown in Fig. 4, to represent the variation in energetic interaction of unit processes. Scenarios which exhibited repetitive trends to those shown in Fig. 4 were excluded from analysis. The results for all 54 scenarios are summarized in the Supplemental data in Fig. S6. Volumetric energy consumption for thermal drying is high compared to technologies used in primary and secondary harvest, due to the high latent heat of water, large concentration gap between secondary harvest and drying, and dryer inefﬁciency. Energy consumption for this group of scenarios ranged from 41% to 91% of the total process energy. Of the 54 scenarios which use drying in this study, 3 processes have ratios of embodied energy of algae to total energy demand
greater than 2, 24 processes have ratios between 1 and 2, and 27 processes have ratios lower than 1, assuming a HHV for algae of 20 MJ/kg (Illman et al., 2000). For the production of 104 MJ of biodiesel using a RP for cultivation, settling with ﬂocculants for primary harvest, and rotary press for secondary harvest followed by thermal drying, Lardon et al. (2009) estimated a TED of 106 MJ. The ratio of useable energy from biodiesel to TED results in an energy return on investment (EROI) of 0.98. Without signiﬁcant improvements to the thermal drying process and a substantial increase in the embodied caloriﬁc content of algae, scenarios which use dry harvest will not be energy efﬁcient. Wet algae lipid extraction and conversion techniques, such as the use of osmotic shock (Yoo et al., 2012) and supercritical methanol (Brentner et al., 2011), have therefore been suggested as alternatives to drying methods. 3.2. Comparison to other studies Several theoretical life cycle assessments of algal biofuel production have included comparable technologies to those analyzed in this study (Brentner et al., 2011; Clarens et al., 2010; Lardon et al., 2009; Sander and Murthy, 2010; Stephenson et al., 2010). Direct comparisons from those studies to this study are difﬁcult to make, however, because they differ in their system boundaries and report their results normalized to mass or energy content of biodiesel, as opposed to biomass. Xu et al. (2011) performed an
M.K. Weschler et al. / Bioresource Technology 153 (2014) 108–115
energy analysis for the production of biodiesel and green diesel using algae biomass as a feedstock, and reported energy inputs to microalgae cultivation, harvest, and drying normalized to 1000 kg of algal biomass. The authors chose four biomass production scenarios for analysis. One of those scenarios included: raceway pond (RP) for cultivation, disc stack centrifuge with preﬂocculation for primary harvest, rotary pressure ﬁlter for secondary harvest, and Delta Dryer for thermal drying. A comparable scenario from this study included: RP for cultivation, settling with ﬂocculant (SF) for primary harvest, chamber ﬁlter press (CFP) for secondary harvest, and mechanical vapor recompression (MVR) dryer for thermal drying. Xu et al. (2011) assumed 50% solids concentration as an optimistic near-future value from the rotary ﬁlter press for secondary harvest, compared to 27% from the CFP from this study. The subsequent energy savings during drying were partially offset by using a less energy efﬁcient dryer, and including the energy required for the production of urea and single superphosphate fertilizers without considering recycling of nutrients. Total energy demand for the production of algal biomass was 2200 kWh/1000 kg, compared to 2600 kWh/1000 kg from this study, a 17% difference. If the authors had not included the energy required for the production of fertilizers, total energy demand would have decreased to 1800 kWh/1000 kg, a 36% difference from this study. These differences highlight the sensitivity of the total energy demand to differences in technology parameters between studies. In addition to differences in choice of technology parameters, the wide range of values reported in the literature for volumetric energy consumption, efﬂuent concentration, and harvest efﬁciency cast additional uncertainty onto the expected total energy demand and energy return on investment for the production of algal biomass. The Supplemental information provides a sensitivity analysis based on the range of values found in literature. Scenarios, which used technologies with a large range of values reported in the literature (high uncertainty), showed the most sensitivity to this alteration. Scenarios that used tangential ﬂow ﬁltration (TFF) technology for primary harvest increased in total energy demand by more than 450%, while at the lower level of uncertainty, scenarios which used raceway ponds, settling with and without, ﬂocculants and chamber ﬁlter press for secondary harvest increased in total energy demand by less than 35%. 3.3. Important considerations Compared with other algae production studies, the model presented in this study was more extensive in terms of number of harvest options, but more focused in terms of scope, by targeting the energy intensive stages of algal biofuel production, i.e. the biomass production processes. This methodology presented multiple advantages compared to a traditional cradle-to-grave life cycle analysis (LCA) for the production of algal biofuel. By limiting the system boundaries to only include the cultivation, harvest, and drying unit processes, focus was placed on evaluating technologies at the bottleneck for algae biofuel systems: the energy required for cultivation, harvesting, and drying. To maintain this focus while remaining consistent in analysis, energy requirements for the production of algal biofuel up- and downstream of the system boundary shown in Fig. 1 were excluded from the calculation of total energy demand. These exclusions, described in detail in the methods, include the energy required to produce ancillary inputs for the production of algal biomass, such as nutrients, ﬂocculants, and membrane ﬁlter cloths. The total energy demands of the algal biomass production scenarios analyzed in this study can form the building blocks of more informed and comprehensive LCAs to determine the optimal algae-based biofuel production scenario using holistic parameters such as energy return on investment
(EROI), global warming potential (GWP), and economic return on investment. In addition, focusing on the energetic bottleneck for algae production enables the identiﬁcation and improvement of energy efﬁcient strategies for cultivation, harvest, and drying. Using the scenarios developed in this study, future LCAs can take many other factors into consideration, such as the energy and material inputs and outputs associated with: construction and maintenance requirements for cultivation and harvest systems, production of nutrients either artiﬁcially produced or from wastewater treatment plants, production of carbon dioxide either synthetically produced or from carbon emitting production facilities, the production and distribution of electricity and natural gas, the production and use of ancillary inputs, such as ﬂocculants and membrane ﬁlter replacements, the supply of either potable or non-potable water, land availability, and spatial proximity to sources of inputs to production. Ultimately, technologies used for the production of biofuel from algae should be chosen using an integrated engineering approach. If microalgae growth is coupled to wastewater treatment, for example, improvements in the energetic performance of biomass harvesting technology should be balanced against the objectives of the wastewater treatment facility. A higher energy demand may be beneﬁcial, if it contributes to improved efﬂuent water quality. Based solely on process energy consumption, almost all scenarios that use raceway ponds (RPs) for cultivation require less energy than processes that use ﬂat-plate photobioreactors (PBRs). The most energy efﬁcient scenarios for all three groups analyzed in this study included RPs for cultivation followed by settling with ﬂocculants for primary harvest. RPs face signiﬁcant drawbacks, however, including: threat of contamination, high evaporative losses, and diffusion of carbon dioxide into the air (Pulz, 2001). Settling basins used for primary harvest are inexpensive and energy efﬁcient. Speed, reliability, and efﬁciency of settling basins, however, are dependent on algae species. Some species may have speciﬁc gravities less than water, making settling impossible (Uduman et al., 2010). Flocculants improve settling efﬁciency, but add additional cost to the system and may affect the quality of downstream co-products, such as biodiesel, nutraceuticals or animal feed. The chamber ﬁlter press is energy efﬁcient and produces a high concentration output, reducing total energy demand. Filter screens, however, may be prone to blockage if small algae species are used and require periodic replacement, adding additional cost to harvesting (Uduman et al., 2010). 4. Conclusion The choice of biomass harvest technology will have an impact not only on the energy required for primary and secondary harvest, but on the energy requirements for cultivation, due to variations in harvest efﬁciency, and drying, due to variations in efﬂuent concentration. Thus, a systems approach must be used to determine the optimal harvest method for algal biomass production. Based solely on process energy demand, scenarios presented in this study which used raceway ponds for cultivation, followed by settling with ﬂocculants for primary harvest and the chamber ﬁlter press for secondary harvest were the most energy efﬁcient. Acknowledgements The authors gratefully acknowledge funding from the Undergraduate Summer Research Program at the University of Pittsburgh’s Mascaro Center for Sustainable Innovation, the Undergraduate Research Internship Program at the University of Pittsburgh’s University Honors College, and from the National
M.K. Weschler et al. / Bioresource Technology 153 (2014) 108–115
Science Foundation Award Nos 0932606/1241697, 124697, and 1039406.
Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at http://dx.doi.org/10.1016/j.biortech.2013. 11.008.
References Brentner, L.B., Eckelman, M.J., Zimmerman, J.B., 2011. Combinatorial life cycle assessment to inform process design of industrial production of algal biodiesel. Environmental Science & Technology 45 (16), 7060–7067. Clarens, A.F., Resurreccion, E.P., White, M.A., Colosi, L.M., 2010. Environmental life cycle comparison of algae to other bioenergy feedstocks. Environmental Science & Technology 44 (5), 1813–1819. Collet, P., Helias, A., Lardon, L., Ras, M., Goy, R.-A., Steyer, J.-P., 2011. Life-cycle assessment of microalgae culture coupled to biogas production. Bioresource Technology 102 (1), 207–214. Danquah, M.K., Ang, L., Uduman, N., Moheimani, N., Forde, G.M., 2009. Dewatering of microalgal culture for biodiesel production: exploring polymer ﬂocculation and tangential ﬂow ﬁltration. Journal of Chemical Technology and Biotechnology 84 (7), 1078–1083. Edzwald, J.K., 1993. Algae, bubbles, coagulants, and dissolved air ﬂotation. Water Science and Technology 27 (10), 67–81. EISA, 2007. United States Congress. Energy independence and security act of 2007. Public Law (110–140). Fon Sing, S., Isdepsky, A., Borowitzka, M., Moheimani, N., 2011. Production of biofuels from microalgae. Mitigation and Adaptation Strategies for Global Change, 1–26. Greenwell, H.C., Laurens, L.M.L., Shields, R.J., Lovitt, R.W., Flynn, K.J., 2010. Placing microalgae on the biofuels priority list: a review of the technological challenges. Journal of the Royal Society, Interface 7 (46), 703–726. Heasman, M., Diemar, J., O’Connor, W., Sushames, T., Foulkes, L., 2000. Development of extended shelf-life microalgae concentrate diets harvested by centrifugation for bivalve molluscs – a summary. Aquaculture Research 31 (8–9), 637–659. Henderson, R., Parsons, S.A., Jefferson, B., 2008. The impact of algal properties and pre-oxidation on solid–liquid separation of algae. Water Research 42 (8–9), 1827–1845. Hunter-Cevera, J.C., Boussiba, S., Cuello, J.L., Duke, C.S., Efroymson, R.A., Golden, S.S., Holmgren, J., Johnson, D.L., Jones, M.E., Smith, V.H., Steger, C., Stephanopoulos, G.N., Walker, L.P., Williams, E., Zimba, P.V., 2012. Sustainable Development of Algal Biofuels in the United States. The National Academies Press, Washington DC. Illman, A.M., Scragg, A.H., Shales, S.W., 2000. Increase in Chlorella strains caloriﬁc values when grown in low nitrogen medium. Enzyme and Microbial Technology 27 (8), 631–635. Jorquera, O., Kiperstok, A., Sales, E.A., EmbiruÁu, M., Ghirardi, M.L., 2010. Comparative energy life-cycle analyses of microalgal biomass production in open ponds and photobioreactors. Bioresource Technology 101 (4), 1406–1413. Jun, H., Lee, Y., Lee, B., Knappe, D., 2001. Effectiveness of coagulants and coagulant aids for the removal of ﬁlter-clogging Synedra. Aqua 50, 135–148.
Lardon, L., Helias, A., Sialve, B., Steyer, J.-P., Bernard, O., 2009. Life-cycle assessment of biodiesel production from microalgae. Environmental Science & Technology 43 (17), 6475–6481. Lohrey, C., Kochergin, V., 2012. Biodiesel production from microalgae: Co-location with sugar mills. Bioresource Technology 108, 76–82. Mohn, F.H. (Ed.), 1980. Experiences and strategies in the recovery of biomass from mass cultures of microalgaeShelef, G., Soeder, C.J. (Eds.). Algae Biomass Production and Use. Elsevier/North-Holland, Amsterdam. Mohn, F.H., 1988. Harvesting of micro-algal biomass. In: Borowitzka, M.A., Borowitzka, L.J. (Eds.), Micro-Algal Biotechnology. Cambridge University Press, Cambridge [Cambridgeshire]/New York, pp. 395–413. Molina Grima, E., Belarbi, E.H., Acién Fernández, F.G., Robles Medina, A., Chisti, Y., 2003. Recovery of microalgal biomass and metabolites: process options and economics. Biotechnology Advances 20 (7–8), 491–515. Petrusevski, B., Bolier, G., Van Breemen, A.N., Alaerts, G.J., 1995. Tangential ﬂow ﬁltration: a method to concentrate freshwater algae. Water Research 29 (5), 1419–1424. Poelman, E., De Pauw, N., Jeurissen, B., 1997. Potential of electrolytic ﬂocculation for recovery of micro-algae. Resources, Conservation and Recycling 19 (1), 1–10. Pulz, 2001. Photobioreactors: production systems for phototrophic microorganisms. Applied Microbiology and Biotechnology 57 (3), 287–293. Richmond, A., Cheng-Wu, Z., 2001. Optimization of a ﬂat plate glass reactor for mass production of Nannochloropsis sp. outdoors. Journal of Biotechnology 85 (3), 259–269. Rossignol, N., Vandanjon, L., Jaouen, P., Quéméneur, F., 1999. Membrane technology for the continuous separation microalgae/culture medium: compared performances of cross-ﬂow microﬁltration and ultraﬁltration. Aquacultural Engineering 20 (3), 191–208. Sander, K., Murthy, G., 2010. Life cycle analysis of algae biodiesel. The International Journal of Life Cycle Assessment 15 (7), 704–714. Sazdanoff, N., 2006. Modeling and Simulation of The Algae to Biodiesel Fuel Cycle. Ohio State University Thesis, pp. 1–67. Shelef, G., Sukenik, A., Green, M., 1984. Microalgae harvesting and processing: a literature review. Prepared for the US department of Energy, Contract No. 0 EAC02-83CH 10093. Show, K.-Y., Lee, D.-J., Chang, J.-S., 2013. Algal biomass dehydration. Bioresource Technology 135, 720–729. Soratana, K., Harper Jr, W.F., Landis, A.E., 2012. Microalgal biodiesel and the renewable fuel standard’s greenhouse gas requirement. Energy Policy 46, 498– 510. Stephenson, A.L., Kazamia, E., Dennis, J.S., Howe, C.J., Scott, S.A., Smith, A.G., 2010. Life-cycle assessment of potential algal biodiesel production in the United Kingdom: a comparison of raceways and air-lift tubular bioreactors. Energy & Fuels 24 (7), 4062–4077. Uduman, N., Qi, Y., Danquah, M.K., Forde, G.M., Hoadley, A., 2010. Dewatering of microalgal cultures: a major bottleneck to algae-based fuels. Journal of Renewable and Sustainable Energy 2 (1), 012701–12715. Van Gemert, G.W., 2009. The Delta dryer: theoretical and technological development of an energy-efﬁcient dryer for sludge, 154 pages. Wiley, P.E., Campbell, J.E., McKuin, B., 2011. Production of biodiesel and biogas from algae: a review of process train options. Water Environment Research 83 (4), 326–338. Xu, L., Brilman, D.W.F., Withag, J.A.M., Brem, G., Kersten, S., 2011. Assessment of a dry and a wet route for the production of biofuels from microalgae: energy balance analysis. Bioresource Technology 102 (8), 5113–5122. Yoo, G., Park, W.-K., Kim, C.W., Choi, Y.-E., Yang, J.-W., 2012. Direct lipid extraction from wet Chlamydomonas reinhardtii biomass using osmotic shock. Bioresource Technology 123, 717–722.