BioEnergy Research

, Volume 5, Issue 1, pp 49–60

Current Large-Scale US Biofuel Potential from Microalgae Cultivated in Photobioreactors

Authors

    • Mechanical EngineeringColorado State University
  • Kimberly Catton
    • Civil and Environmental EngineeringColorado State University
  • Nicholas Wagner
    • Mechanical EngineeringColorado State University
  • Thomas H. Bradley
    • Mechanical EngineeringColorado State University
Article

DOI: 10.1007/s12155-011-9165-z

Cite this article as:
Quinn, J.C., Catton, K., Wagner, N. et al. Bioenerg. Res. (2012) 5: 49. doi:10.1007/s12155-011-9165-z

Abstract

Current assessments of the commercial viability and productivity potential of microalgae biofuels have been forced to extrapolate small-scale research data. The resulting analyses are not representative of microalgae cultivation and processing at industrial scale. To more accurately assess the current near-term realizable, large-scale microalgae productivity potential in the USA, this paper presents a model of microalgae growth derived from industrial-scale outdoor photobioreactor growth data. This model is combined with thermal models of the photobioreactor system and 15 years of hourly historical weather data from 864 locations in the USA to more accurately assess the current productivity potential of microalgae. The resulting lipid productivity potential of Nannochloropsis is presented in the form of a map that incorporates various land availability models to illustrate the near-term feasible cultivation locations and corresponding productivity potentials for the USA. The discussion focuses on a comparison of model results with productivity potentials currently reported in literature, an assessment demonstrating the scale of Department of Energy 2030 alternative fuel goals, and a critical comparison of productivity potential in several key regions of the USA.

Keywords

BiofuelsGISMicroalgaeModelProductivity potential

Abbreviations

PAR

Photosynthetic active radiation

PFD

Photon flux density

GIS

Geographic information system

PBR

Photobioreactor

ORP

Open raceway pond

DOE

Department of Energy

NLCD

National Land Cover Database

Nomenclature

cp

Specific heat of water (kJ kg−1 K−1)

Ea

Activation energy carboxylation Rubisco (J mol−1)

Gbottom

Solar energy reaching the bottom (W m−2)

Gn

Solar energy reaching node n (W m−2)

Gsur

Solar energy reaching the surface (W m−2)

hi

Convection coefficient (W m−2 K−1)

hr

Net radiation coefficient with the sky (W m−2 K−1)

k

Thermal conductivity of water (W m−1 K−1)

Ln

Distance between nodes (m)

mn

Total mass represented by node n (kg)

Qi

Energy stored/released by ground (W m−2)

R

Universal gas constant (J K−1 mol−1)

T1

Temperature at node 1 (K)

Tambient

Temperature of the ambient (K)

Tn

Temperature at node n (K)

Tn − 1

Temperature at node n minus 1 (K)

Tn + 1

Temperature at node n plus 1 (K)

Topt

Optimum microaglae growth temperature (K)

Tsky

Temperature of the sky (K)

Tsur

Temperature at the surface (K)

T

Temperature of microalgae culture (K)

t

Time (s)

φT

Temperature efficiency factor

Introduction

Microalgae-based biofuels have several sustainability, economic, and environmental benefits compared to conventional biofuels such as cellulosic ethanol and soy-based biodiesel. Relative to first-generation biofuel feedstocks, microalgae are characterized by higher solar energy yield, year-round cultivation, the use of lower-quality or brackish water, and the use of less- and lower-quality land [14]. Microalgae feedstock cultivation can be integrated with waste streams such as wastewater treatment facilities and CO2 generating processes such as combustion power plants to provide the microalgae with nutrients that enable higher productivities than conventional crops [5]. These advantages have led to an increased interest in microalgae as a third-generation feedstock for the production of biofuels.

Previous assessments of the economics and lifecycle environmental impacts of the latest generation of microalgae cultivation have relied on simulated microalgae growth models, which in general do not have the capability to accurately capture thermal and climatic effects, are not validated, and do not represent industrial-scale facilities [2, 3, 69]. The majority of the geographical evaluations of productivity potential have been based on a conversion of solar irradiance to biomass using a photosynthetic efficiency. This technique does not accurately represent microalgae growth and leads to errors in assessing the large-scale growth potential by not incorporating real effects on growth such as mixing, reactor orientation, temperature, growth dynamics, respiration, diffuse versus direct light, nutrients, and more [10]. Geographical maps have been generated highlighting feasible cultivation sites for large-scale microalgae cultivation based on land availability and slope; however, these studies fail to incorporate detailed growth models for accurate prediction of microalgae productivity potential [11, 12]. Wigmosta et al. incorporated factors such as solar characteristics and temperature to predict biomass based on a solar efficiency open raceway pond (ORP) growth model; however, the hourly weather data, which serve as a primary input to the growth model, are estimated at various US locations through the discretization of stochastically generated synthetic daily weather data which lead to errors in representing growth dynamics. This most recent work does not model lipid accumulation dynamics but assumes a constant lipid percentage [12].

To accurately calculate the productivity potential of microalgae-based biofuels, representative microalgae growth and lipid accumulation models validated with industrial-scale outdoor growth data must be used [13, 14]. This study presents the integration of a photobioreactor (PBR) thermal model used to estimate culture temperature with an industrially representative growth and lipid accumulation model and geographically explicit climate data to investigate the current near-term realizable productivity potential of microalgae in locations that meet the land availability and slope criteria as set forth by previous techno-economic studies [1517]. These models were run with 15 years of hourly meteorological data collected from 864 US locations to evaluate the current biomass and lipid production as a function of geographic location. Productivity results are integrated with graphical information systems (GIS) land availability and slope data to produce a map of the feasible large-scale microalgae production locations in the USA. The discussion focuses on an evaluation of the capability of microalgae to meet 2030 Department of Energy (DOE) alternative fuel goals, a comparison of the modeled productivity results to current values reported in literature, and a critical assessment of states with microalgae biofuel initiatives.

Materials and Methods

The following sections provide details on the photobioreactor thermal model used to predict culture temperature, a basic overview of the temperature dependent microalgae growth model, and the data and criteria for the GIS analysis. A schematic of the simulation architecture for the biomass and lipid productivity based on species, reactor, and metrological data is presented in Fig. 1.
https://static-content.springer.com/image/art%3A10.1007%2Fs12155-011-9165-z/MediaObjects/12155_2011_9165_Fig1_HTML.gif
Fig. 1

Simulation architecture for thermal growth model. Validated growth model requires species and reactor-specific characteristics, water basin temperature, and metrological data to predict biomass and lipid production

The simulation integrates a microalgae growth model which accurately predicts biomass and lipid production based on meteorological data, microalgae species characterization, reactor configuration, and reactor temperature. Results from the microalgae growth model were overlaid with geographical data to generate the dynamic maps presented. The microalgae growth model and PBR thermal model are coded in MatLab® with GIS data reduction performed using ArcGIS.

Photobioreactor Thermal Model

The photobioreactor modeled for this effort (Solix Biosystems Generation 3 photobioreactor) is submerged in a shallow pool of water (water basin) to provide structural and thermal stabilization (detailed descriptions of the system modeled are presented in the supplementary material). To accurately incorporate the effects of temperature on growth, a thermal model of the water basin incorporating radiative, conductive, and convective heat balance is developed to estimate the temperature of the microalgae culture.

Thermal models for shallow pools typically use a single node for thermal calculations assuming no temperature gradients due to continuous circulation through a filtration system and a low surface to volume ratio [18, 19]. The water basin modeled here is not actively mixed and has a surface to volume ratio of 1:3, thus has the potential to have thermal gradients. The water basin was therefore represented by 16 equally spaced vertical nodes, adapted from the methods of Weyer-Geigel [20]. A schematic of the thermal resistance model used to represent the water basin is presented in Fig. 2.
https://static-content.springer.com/image/art%3A10.1007%2Fs12155-011-9165-z/MediaObjects/12155_2011_9165_Fig2_HTML.gif
Fig. 2

Thermal resistance representation for the water basin model used to predict basin temperature for accurate growth modeling

This model assumes that all nodes receive solar energy with a distribution based on the absorption characteristics of water incorporating losses at the air water interface based on Snell’s law and solar incident angles. The basic heat balance equation applied to each node is:
$$ {E_{\text{in}}} - {E_{\text{out}}} + {E_{\text{generated}}} = {E_{\text{stored}}} $$
(1)
Negligible heat is generated in the PBR; the heat balance equation for the temperature of an internal node, n, is:
$$ {G_n} - \frac{k}{{{L_n}}}\left( {{T_n} - {T_{{n - 1}}}} \right) - \frac{k}{{{L_n}}}\left( {{T_n} - {T_{{n + 1}}}} \right) = {m_n}{c_{\text{p}}}\frac{{\Delta {T_n}}}{{\Delta t}} $$
(2)
where Gn is the solar radiation reaching node n, k is the thermal conductivity of water, Ln is the distance between nodes, mn is the mass representative of node n, cp is the specific heat of water, and T is the temperature. The surface node (Eq. 3) and ground node (Eq. 4) temperatures are calculated similarly to Eq. 2 but represent half the mass of an internal node with Gsur and Gbottom being the solar radiation at the surface and bottom, respectively, hr being net radiation coefficient with the sky, Qi being the stored or released energy from the ground, and hi being the convection coefficient.
$$ {G_{\text{sur}}} - \frac{k}{{{L_n}}}\left( {{T_{\text{sur}}} - {T_1}} \right) - {h_r}\left( {{T_{\text{sur}}} - {T_{\text{sky}}}} \right) - {h_i}\left( {{T_{\text{sur}}} - {T_{\text{ambient}}}} \right) = \frac{{{m_n}}}{2}{c_{\text{p}}}\frac{{\Delta {T_{\text{sur}}}}}{{\Delta t}} $$
(3)
$$ {G_{\text{bottom}}} - \frac{k}{{{L_n}}}\left( {{T_n} - {T_{{n - 1}}}} \right) - {Q_i} = \frac{{{m_n}}}{2}{c_{\text{p}}}\frac{{\Delta {T_n}}}{{\Delta t}} $$
(4)

Heat loss with the ground and the walls is assumed to be through conduction (hi) with a heat transfer coefficient of 10 W m−2 based on typical soil characteristics [18]. The Sartori equation was used to calculate evaporative losses based on a large-scale facility [21]. This thermal resistance network between the nodes of the pool and between the pool and the ambient is solved incrementally at each time step to predict the heat flux between nodes and determine the basin temperature. When an inverted temperature gradient occurs, it is assumed that due to a density gradient the two adjacent nodes will mix with the resulting temperatures being the average of the two mixed nodes [22].

The water basin model uses inputs of solar radiation, dry-bulb temperature, dew-point temperature, wind speed, wind direction, cloud cover, and atmospheric pressure to calculate the heat balance and temperature of the water basin. The water basin temperature, which is assumed to be equivalent to the culture temperature, is then used as an input to the microalgae growth model. The cultivation is assumed to shut down when the water basin freezes.

Microalgae Growth Model

A time-resolved, microalgae growth model based on a PBR architecture, incorporating 21 species- and reactor-specific characteristics to represent biomass growth and lipid accumulation based on real-world climactic and thermal conditions, was adapted to this study [14]. When provided primary inputs of solar photosynthetic active radiation (PAR) and microalgae temperature (water basin temperature), the model has been shown to accurately represent the biomass and lipid production of Nannochloropsis oculata cultivated in Solix Generation 3 photobioreactors.

For this study, photosynthetic productivity is a non-linear function of PAR. The model captures and incorporates two of the three regions of the photosynthetic productivity curve typically used to describe photosynthetic behavior: (a) a photon flux density (PFD) dominated region where the photosynthetic rate increases linearly with increasing light intensity and (b) a light saturation photosynthetic region characterized by constant photosynthetic rate with increasing PFD. The model does not include a photo-inhibition region characterized by a decrease in photosynthetic rate with increasing PFD [2325]. Photo-inhibition typically occurs at high light intensities which are not achieved in the system modeled [23, 26]. Consistent with data presented in Henley et al. and similar studies, Fig. 3 illustrates that the initial slope of the photosynthetic curve is not affected by the temperature; only the overall maximum photosynthetic rate is affected [23, 24, 2729].
https://static-content.springer.com/image/art%3A10.1007%2Fs12155-011-9165-z/MediaObjects/12155_2011_9165_Fig3_HTML.gif
Fig. 3

Temperature efficiency (φT) as a function of temperature based on Eqs. 5 and 6 (left). Photosynthetic response to temperature efficiency (right). Light saturation level illustrated in red is 200 μmol m−2 s−1

The temperature dependence of photosynthesis is described by the effect of temperature on ribulose-biphosphate carboxylase (Rubisco) activity. The model presented by Alexandrov and Yamagata [30] relates thermodynamic concepts, such as activation energy, to the typical bell shape of the enzyme activity temperature curve illustrated in Eqs. 5 and 6 and has been adapted to this model. This model assumes that temperature only affects the light-saturated photosynthesis rate and respiration rate and not the initial slope of the photosynthesis-irradiance curve as illustrated in Fig. 3 [31].
$$ {\varphi_T} = \frac{{2 \cdot f(T)}}{{\left( {1 + {f^2}(T)} \right)}} $$
(5)
$$ f(T) = {e^{{\frac{{{E_{\text{a}}}}}{{R \cdot {\text{Topt}}}} - \frac{{{E_{\text{a}}}}}{{R \cdot T}}}}} $$
(6)

The efficiency factor for temperature (φT) is a dimensionless number between 0 and 1. At the optimum growth temperature, φT = 1, and for temperatures higher or lower than the optimum temperature, 0 < φT < 1 according to Eq. 6 illustrated in Fig. 3. More details and experimental results illustrating the effectiveness of capturing the effect of temperature using Eqs. 5 and 6 are presented in the supplementary material.

The microalgae growth model incorporates light, temperature, and nutrients as primary factors to model carbon fixation based on the activity of the Rubisco enzyme while considering respiration losses and energy required for nutrient uptake. All model assumptions, validation, and details on the biological growth model including details on nutrient efficiency factors are presented in Quinn et al. [14].

The growth model was operated for this study with a time-based harvest schedule, where cultures are inoculated at 1 g L−1 at hour 0 and harvest of the culture occurs at 160 h or a density of 3 g L−1 (whichever occurs first). This harvest model is representative of the function of the research and development facility used for model validation. Additionally, the productivity potentials presented in this study are on a per photosynthetic area basis and do not include land required for large-scale cultivation infrastructure such as materials storage, transportation, or process infrastructure. The lipid potentials reported are in terms of total lipids produced and do not include potential losses from extraction or transesterification. It is important to note that the results presented do not include extraction efficiencies or land required for facility infrastructure.

Historical Weather Data

Hourly weather data from 1991 to 2005 from 864 US locations were input to the photobioreactor thermal model and the microalgae growth model [32]. The water basin model outputs a basin temperature that was then input into the growth model with the corresponding solar characteristics to predict biomass and lipid yields on an hourly basis over the 15 years simulated. The lipid yields for the resulting 864 locations are presented in terms of their 15-year annual average.

Geographic Information System

The output of the microalgae growth model is the productivity potential of microalgae installations at 864 locations in the USA. A GIS map was generated by overlaying the productivity data with land use data to evaluate feasible microalgae cultivation sites in the USA.

Maps were generated by first interpolating the microalgae oil productivity over the USA using an inverse distance weighting technique in ArcMap [33]. Next, the National Land Cover Database (NLCD) from the Multi-Resolution Land Characteristics Consortium was used to evaluate feasible locations for microalgae production based on land classifications. The following NLCD land cover classifications were considered as available land for cultivation for the baseline scenario: barren, scrubland, shrubland, and grassland/herbaceous [10, 34]. The baseline scenario avoids high value agricultural land, urban areas, sensitive wetlands, open water, and forested land. The land cover data set was overlaid with federal land data to exclude microalgae cultivation from wilderness areas, national parks, federal research areas, national forests, and national recreation areas. Included as possible cultivation locations were Bureau of Reclamation land and Department of Energy sites [35]. Results from a sensitivity to land use criteria were performed by including potential microalgae production on forest land and pasture land.

Slope restrictions were set based on recommendations from the literature. A survey of the literature illustrates that there is a debate regarding minimum acceptable slope requirements for large-scale microalgae cultivation. Benemann et al., Lansford et al., and Muhs et al. define a requirement for the slope to be 2% or less for economic reasons considering the construction of open raceway ponds [1517]. The DOE algae road map defines an acceptable slope of 5% or less [36]. The baseline scenario for this study assumes that a 2% slope or less is required for microalgae cultivation. GIS slope data were extracted from the Shuttle Radar Topography Mission 90-m resolution digital elevation data and overlaid so that land with greater than 2% slope is excluded from cultivation [37]. A sensitivity to slope criteria is presented and includes evaluating available land based on 1%, 2%, and 5% slope limits.

Results

The results from this work are divided into three sections: (a) a map illustrating the current realizable productivity potential of microalgae in the continental USA and Hawaii, (b) an integration of the productivity results with GIS land availability and slope criteria to generate a dynamic map based on the baseline land availability scenario (<2% slope and barren land), and (c) a sensitivity analysis to GIS slope and land availability.

Current US Productivity Potential

The modeled near-term realizable lipid productivity potential for microalgae cultivated in a photobioreactor is presented in Fig. 4. Regions with high productivity potential (20–27 m3 ha−1 year−1) are characterized by high degree of solar irradiance and year-round temperatures above freezing, which include the Southwest, West Texas, Hawaii, and Florida. Low productivity regions (8–14 m3 ha−1 year−1) are characterized by lower temperatures corresponding to shorter growing seasons and lower levels of solar irradiance and include the Northeast and Mountain West.
https://static-content.springer.com/image/art%3A10.1007%2Fs12155-011-9165-z/MediaObjects/12155_2011_9165_Fig4_HTML.gif
Fig. 4

Modeled microalgae lipid productivity potential in the USA

Current Feasible Cultivation Locations

The lipid productivity results are combined with baseline GIS slope and land availability data to generate the baseline dynamic map describing locations of feasible microalgae production, presented in Fig. 5. The baseline scenario defines cultivatable land as barren land (barren, shrubland, scrubland, and grassland) with less than a 2% slope.
https://static-content.springer.com/image/art%3A10.1007%2Fs12155-011-9165-z/MediaObjects/12155_2011_9165_Fig5_HTML.gif
Fig. 5

Dynamic map of near-term realizable large-scale microalgae lipid production in cubic meters per hectare per year. Areas that do not meet the land availability and slope criteria defined for the baseline scenario are shown in gray

In contrast to Fig. 4, Fig. 5 shows that land availability considerations remove several otherwise promising regions of the USA from consideration for microalgae biofuels production. For example, forest cover in the coast of the Gulf of Mexico and agricultural cover in California exclude low-impact microalgae production from much of these regions. The Southwest and Texas can be highlighted as high-productivity regions with a large amount of available land. The total land available in the continental USA including Hawaii based on the baseline GIS restrictions is 75 million ha corresponding to 10 billion barrels of lipid production.

GIS Land and Slope Sensitivity

A sensitivity analysis was conducted to investigate the effect of slope restrictions on the land availability and microalgae productivity potential on forested and forested and pasture land. A comparison of the land availability criteria and slope restrictions on the total available land and lipid production potential in the USA is presented in Table 1.
Table 1

Land availability and lipid production potential based on the three land availability criteria: barren (defined as the baseline scenario which includes barren, shrubland scrubland, and grassland), barren and forested (baseline with forested land), and barren, forested, and pasture (baseline with forested and pasture land) with slope criteria of <1%, <2%, and <5% for each of the land availability scenarios

GIS land classifications available for microalgae cultivation

GIS slope classifications available for microalgae cultivation (%)

Total US land available for microalgae cultivation (106 ha)

Total US microalgae-based lipid productivity (109 bbl)

Barren

<5

150

21

Barren

<2

75

10

Barren

<1

32

4

Barren and forested

<5

208

28

Barren and forested

<2

98

13

Barren and forested

<1

40

5

Barren, forested, and pasture

<5

246

33

Barren, forested, and pasture

<2

118

16

Barren, forested, and pasture

<1

49

6

These results show that microalgae productivity potential is sensitive to assumptions regarding both minimum slope and land cover restrictions. Changes in land type excluded from cultivation are shown to affect the US productivity potential by a factor of 8. A table summarizing the productivity and land availability for individual states is presented in the supplementary material which includes the baseline land assumptions, baseline plus forested land, and baseline plus forested and pasture land at slope restrictions of <1%, <2%, and <5%.

Discussion

Based on these results, we can now (a) examine microalgae’s potential to meet DOE 2030 goals incorporating geographical relevant growth modeling and GIS land availability, (b) compare the current productivity potential of microalgae based on this study to values presented in literature, and (c) investigate current productivities and land availability in states that are dedicating resources to microalgae biofuels cultivation.

Microalgae Productivity Potential Relative to DOE Alternative Fuel Goals

The Energy Policy Act of 1992 directed the US Department of Energy to evaluate the goal of replacing 30% (∼1 billion barrels) of the transportation fuel consumed in the US by 2030 with replacement fuels [38]. Based on this goal, the scalability of current feedstocks for biofuels can be critically evaluated. Microalgae-based feedstocks are purported to be the most scalable of the biofuel feedstocks currently being considered for meeting DOE goals [39]. Current scalability assessments for corn and soybeans require 385% and 148% of the current available farm land in the USA, respectively, to meet the 2030 DOE alternative fuel goals [8].

The modeling effort presented here represents the current near-term realizable productivity potential of microalgae incorporating geographically relevant growth characteristics based on a photobioreactor architecture. Assuming a packing factor of 0.8 [3], an oil extraction and conversion efficiency of 0.8 [9] and an average lipid productivity of 18 m3 ha−1 year−1, 13.1 million ha of land would be required to meet the DOE 2030 alternative transportation fuel goal of 1 billion barrels of fuel [38]. The results from this study show that (using baseline GIS land and slope criteria) the USA has 60 million ha of land feasible for microalgae cultivation, which can potentially produce 13.3 billion barrels of oil or 13.3 times the DOE 2030 alternative transportation fuel goal. Under the most conservative land use (barren) and slope restrictions (<1%), microalgae biofuels in the USA can surpass DOE goals by a factor of 2.56. An analysis of productivity potential on a state level shows that the lipid productivity potential of some key states is comparable in magnitude to DOE 2030 alternative fuel goals, Fig. 6. The results from this study show that microalgae can scale to DOE alternative fuel goals based on near-term realizable productivity potentials determined though growth modeling which incorporates geographically relevant weather and land availability data.
https://static-content.springer.com/image/art%3A10.1007%2Fs12155-011-9165-z/MediaObjects/12155_2011_9165_Fig6_HTML.gif
Fig. 6

Productivity potential of the top ten producing states in the USA for nine different GIS land and slope criteria. B barren (includes barren land, shrubland, scrubland, and grassland), F barren and forested land, P barren, forested, and pasture land. DOE 2030 alternative fuel goal highlighted by dashed line. Results include land use packing factor of 0.8 and an extraction and conversion efficiency of 0.8

Current Productivity Potential of Microalgae Compared to Literature

Many studies of microalgae have reported productivity potentials that are orders of magnitude higher than tradition terrestrial crops [4, 8, 40]. The majority of the productivity potentials reported are calculated by the linear scaling of small-scale laboratory-based growth and lipid production data. This type of analysis has been shown to not be representative of the true current near-term productivity potential of microalgae at large-scale. The simplistic scaling of small-scale laboratory-based data leads to erroneous assumptions about industrial growth facility function and is a source of uncertainty in microalgae biofuels economic and sustainability assessments [14].

The uncertainty in the current productivity potential of microalgae has led to a wide range of values being reported. Values reported in literature surveyed range from 8.2 m3 ha−1 year−1 reported by Scott et al. [40] to 136.9 reported by both Chisti [8] and Mata et al. [4] with Clarens et al. [41], Davis et al. [9], Schenk et al. [1], Lardon et al. [6], Campbell et al. [7], Sheehan et al. [42], Huntley and Redalje [43], Chisti [44, 45], Rodolfi et al. [46], Hirano et al. [47], Wijffels and Barbosa [2], Williams and Laurens [48], Sawayama et al. [49], and Batan et al. [3] all reporting values between these extremes as shown in Table 2.
Table 2

Table comparing reported productivity potentials (some calculations performed for comparison purposes with assumptions detailed in notes column) from various sources

Source

Oil yield (m3 ha−1 year−1)

Article type

Purpose of scaling

Notes

Scott et al.a [40]

8.2

Review

Microalgae potential

46% oil

Clarens et al.a [41]

11.8

Research model

LCA modeling effort

9.45 g m−2 day−1, 30% oil assumed, ORP

Schenk et al.a [1]

12.0

Review

Microalgae potential

10 g m−2 day−1, 30% TAG

Clarens et al.a [41]

16.1

Research model

Economic modeling effort

12.9 g m−2 day−1, 30% oil assumed, ORP

Lardon et al.a [6]

18.0

Research model

LCA modeling effort

24.75 g m−2 day−1, 17.5% oil, ORP

Campbell et al.a [7]

18.7

Research model

LCA modeling effort

15 g m−2 day−1, 30% oil assumed, ORP

Sheehan et al.a [42]

21.2

Research model

Economic modeling effort

17 g m−2 day−1, 30% oil, ORP

Davis et al.[9]

23.4

Research model

Economic modeling effort

25 g m−2 day−1, 25% oil, ρ = 880 kg m−3, ORP

Huntley and Redaljea [43]

30.7

Research model

Economic modeling effort

18.5 g m−2 day−1, 40% oil, PBR an ORP

Lardon et al.b [6]

30.8

Research model

LCA modeling effort

19.25 g m−2 day−1, 38.5% oil, ORP

Chistia [44]

31.1

Letter response

Microalgae potential

25 g m−2 day−1 20% oil, ρ = 880 kg m−3, ORP

Rodolfi et al. [46]

34.1

Research model

Economic modeling effort

Location with 20 MJ m−2 day−1

Campbell et al.b [7]

37.3

Research model

LCA modeling effort

30 g m−2 day−1, 30% oil assumed, ORP

Hirano et al.a [47]

37.4

Research model

NER modeling effort

30 g m−2 day−1, 30% oil assumed, ORP

Wijffels and Barbosa [2]

40.0

Perspective

Microalgae potential

3% solar efficiency, 50% oil

Scott et al.b [40]

40.0

Review

Microalgae potential

50% oil

Williams and Laurensb [48]

40.7

Review

Economic modeling effort

28 g m−2 day−1, 35% oil

Yeang et al.a [50]

46.0

Opinion

Microalgae potential

 

Chistib [44]

51.9

Letter response

Microalgae potential

25 g m−2 day−1, 50% oil, ρ = 880 kg m−3, ORP

Sawayama et al.a [49]

51.9

Research model

Modeling effort

15 g m−2 day−1, 20.5% oil,

Batan et al. [3]

51.9

Research model

LCA modeling effort

25 g m−2 day−1, 50% oil, PBR

Chistia [8]

58.7

Review

Microalgae potential

30% oil, PBR

Mata et al.a [4]

58.7

Review

Microalgae potential

30% oil

Sheehan et al.b [42]

74.7

Research model

Economic modeling effort

60 g m−2 day−1, 30% oil, ORP

Chisti [45]

98.4

Opinion

Microalgae potential

1.535 kg m−3 day−1, 30% oil

Schenk et al.b [1]

98.5

Review

Microalgae potential

50 g m−2 day−1, 30% TAG

Huntley and Redaljeb [43]

99.6

Research model

Economic modeling effort

60 g m−2 day−1, 40% oil, PBR and ORP

Chistib [8]

136.9

Review

Microalgae potential

70% oil PBR

Mata et al.b [4]

136.9

Review

Microalgae potential

70% oil

Some authors reported a range of productivity potential

aThe low values are repeated in this table

bThe high values are repeated in this table

By comparison to the data in Table 2, the peak annual productivity potential in the southwestern USA as modeled in this study is 5.7 times lower than the highest value reported in the literature surveyed and is lower than the median productivity reported in the literature surveyed by a factor of 2. The reduced productivity can be attributed to the high level of detail present in this model but not present in other models. The modeling effort presented utilizes real climatic data with a validated large-scale outdoor microalgae photobioreactor growth and lipid accumulation model to more accurately assess productivity potential.

Current US Biofuels Initiatives

The high reported productivity potential of microalgae compared to first-generation feedstocks has led to the development of many local-level and state-level initiatives in the USA to support research and pilot-scale microalgae production in an effort to meet alternative energy goals. The geographically resolved microalgae productivity potential calculated in this study now allows us to critically evaluate the ability of these programs to meet their near-term-stated goals.

For instance, Hawaii has historically been considered a prime location for large-scale microalgae cultivation based on its climate and energy prices and is currently the location of a variety of microalgae research activities. General Atomics is currently cultivating microalgae on Kauai expanding the facility to a size of 11 ha [51]. Phycal is constructing a 12-ha pilot plant on Oahu [52]. Cyanotech is currently cultivating microalgae for high valued products on a 36-ha facility [53]. In conjunction with these pilot plant facility installations, Hawaii has solicited and dedicated federal and state funds to demonstrate microalgae cultivation is commercially viable in an effort to meet renewable portfolio goals. The state of Hawaii currently leads the USA in petroleum dependence with over 90% of its energy being derived from fossil fuels [54]. In an effort to decrease dependence on imported liquid fuels, Hawaii has an ambitious renewable portfolio goal of 70% all energy for electricity and transportation (28.56 million barrels) be derived from renewable energy [55, 56]. This study shows that the total productivity potential of microalgae in Hawaii is less than 0.5 million barrels (based on the baseline scenario of barren land and <2% slope), an order of magnitude less than the Hawaii (state level) 2030 renewable portfolio goals.

This result is notable because it suggests that the development of microalgae as a large-scale fuel source is dependent on cultivation of algae in less than ideal geographic locations and conditions. Pilot plant installations in Hawaii do not develop the knowledge base required to cultivate microalgae at significant scale in locations such as the southwestern USA. The environmental and fuel displacement benefits of microalgae biofuels are dependent on not just productivity but productivity and area.

Strictly based on land availability (baseline scenario), the top five states in the USA in descending order are Texas, New Mexico, Montana, Arizona, and Nevada. Texas has an immense amount of land resource with relatively high productivity, 26 million ha of available land corresponding to 29.5 million barrels annually. New Mexico, Arizona, and Hawaii have been the focus of discussion in regard to large-scale production of biofuels from microalgae; however, the results presented show Texas has the same production capacity as all three of these states combined. Although other states may outperform Texas in terms of productivity, an improved understanding of land availability makes Texas very appealing in terms of total production potential.

The results presented are based on hourly meteorological data collected from 864 US locations over a period of 15 years and used to evaluate the current near-term realizable microalgae lipid productivity potential in the USA based on a bulk growth model validated with outdoor large-scale PBR data. The productivity results have been combined with GIS slope and land availability data to generate a dynamic map cataloging the lipid productivity in the USA for large-scale microalgae biofuels production. A comparison of modeled results to productivity potentials reported in literature illustrates that the majority of the studies surveyed overestimate the near-term realizable microalgae productivity potential. Productivity results show that microalgae can scale to 2030 DOE alternative fuel goals. This study presents detailed results on microalgae facility locations, productivity potentials, and land requirements necessary to critically evaluate microalgae as a feedstock for biofuels.

Acknowledgments

We gratefully acknowledge support from Solix Biosystems and thermal modeling support form Kristina M. Weyer.

Supplementary material

12155_2011_9165_MOESM1_ESM.pdf (1.1 mb)
ESM 1(PDF 1.14 mb)

Copyright information

© Springer Science+Business Media, LLC. 2011