Technical Memorandum

To:		EPA

Title:	Techno-economic analysis of microalgae-derived biofuel production

Author:		Ryan Davis

Date:		November 10, 2009

Abstract
The purpose of this study is to evaluate the economics for a feasible method of producing microalgal-derived oil in the near-future.  Specifically, production via both open-pond and closed photobioreactor (PBR) systems was evaluated for a 10 MMgal/yr oil facility based on assumptions for what can feasibly be achieved in the near-term, represented by "2022" target production.  The scope of this study is based on analysis for the process beginning at CO2 delivery to the facility and ending at lipid production (refining into biodiesel is fairly well-established and is not included here).  Upon completing the base case scenarios, the cost of lipid production to achieve a 10% return was determined to be $11.25/gal for open pond production and $19.50/gal for PBR's.  This cost has potential for significant improvement in the future if better microalgal strains can be identified that would be capable of sustaining high growth rates at a high lipid content.  Given that algal biofuel technology is still in a relatively early stage of development, there is intrinsically a higher degree of uncertainty associated with potential process performance and cost relative to more established technologies.  As such, many assumptions had to be made in this study; the assumptions with the highest economic impact were found to be the optimum amount of nutrients required, the CO2 delivery cost, and the flocculant requirement for harvesting, as well as the material costs for the PBR production system.
Keywords: 	Algae, techno-economic, pond, PBR, biodiesel, oil
Introduction
It is well-established that microalgal-derived biofuels have the potential to make a significant contribution to the US fuel market, due to unique advantages inherent to algae.  While these advantages are outlined elsewhere [1, 2], they include the fact that microalgae have a higher photosynthetic efficiency than terrestrial plants at converting solar energy into biomass and also can possess a very high oil content, resulting in drastically lower land requirements that must be dedicated to biomass growth per unit of oil produced, relative to other oil-based crops.  Furthermore the land used to grow microalgae can be low-value or otherwise not suitable for growing traditional agricultural crops.  Additionally, CO2 from a nearby emissions source can be used to grow the algae and the only source of CO2 from the algae facility itself can be recycled, resulting in a "net zero" emissions process.
In this study, a microalgae facility producing lipids on a basis of 10 MMgal/yr was modeled using Aspen Plus software, in order to get more accurate mass balance information than what is typically assumed in other studies.  Additionally the model was used to estimate the overall amount of power used vs generated throughout the plant.  Using the resulting flow rates, the cost of each process unit was estimated based on vendor information, cost equations, or previous studies to determine the cost of lipid production.  It is important to note up front that algal biofuel technology is still in a very early stage of development, and there are numerous opinions on the optimal configuration and conditions at each stage of the production process; as such, the estimated cost of production varies in other studies from $1 to >$40/gal [3].  There are also various opinions on when the technology will even become feasible on a large scale.  Thus the "2022" startup date should be treated as being representative of what can reasonably be expected in the "current to near-term" time frame.
The open pond and PBR cases will be discussed together, as most of the technology employed in either case is the same downstream of the production step.  To summarize the basic flow scheme chosen for this study, the open pond configuration is as follows (see Figure 1 for a block flow diagram): pure CO2 is concentrated out of flue gas from a nearby power plant and delivered to the facility.  The CO2 material cost used in this study accounts for all upstream operations required to concentrate CO2 out of the flue gas stream (by means such as amine scrubbing).  The CO2 is transferred to the ponds via sumps.  The ponds are simple unlined, paddle wheel-mixed raceway ponds 30 cm deep, providing constant circulation to promote algal growth.  The microalgae grow to a concentration of 0.5 g/L, and are continuously harvested at a rate equal to the growth rate.  Harvesting is accomplished first in a simple settling tank which concentrates the algae to 1% (10 g/L), and then in a secondary step by flocculation using an organic polymer flocculating agent and subsequent collection via dissolved air flotation (DAF), which thickens the material to 10% (100 g/L).  Extraction is accomplished primarily by mechanical methods using homogenizers to lyse the cell, followed by hot oil extraction to purify the oil phase.  The oil, water, and biomass phases separate by gravity in a clarifier and the oil is drawn off, a fraction of which is recycled and heated to provide the "hot oil" solvent.  The water and residual biomass are sent to anaerobic digestion, which produces biogas burned in a gas turbine to aid in supporting power requirements throughout the plant.  The turbine flue gas is recycled to the ponds, while the treated water is discharged.  Makeup water is added to replace this discharge, as well as evaporation from the ponds and a small amount of water entrained in the lipid product.  Nutrients from the anaerobic digestion step are recycled to minimize the amount of fresh nutrients required.
In the PBR case, pure CO2 is delivered to the facility by being injected directly into the feed header, eliminating the need for sumps as in the open pond case.  The PBR system is made up of clear plastic tubes in a parallel configuration, with a degassing/pumping section (achieved by airlift columns) every 80m to strip out accumulated oxygen and provide circulation through the tubular system.  Due to the concentrated nature of PBR culture, the algae are grown to a much higher density of 4 g/L.  The algae are harvested at a rate equal to the growth rate, and the material is again concentrated to 10 g/L in the primary settling step then subsequently to 100 g/L in the DAF unit, similar to the open pond case.  The remaining steps (extraction, purification and spent biomass utilization) are identical to the open pond case, as shown in Figure 1.
                                       
     Figure 1: Block flow diagram of the production scheme evaluated here
Three cases were run for each scenario to evaluate the impact that research advancements will have on the production cost.  These are summarized below and results are presented at the end of this report.
                 Table 1: Algae production analysis scenarios
 
                                   Base case
                                Aggressive case
                                Max growth case

                                   Open pond
                                      PBR
                                   Open pond
                                      PBR
                                   Open pond
                                      PBR
Algae productivity[1]
                                25 [g/m[2]/day]
                              1.25 [kg/m[3]/day]
                                40 [g/m[2]/day]
                               2.0 [kg/m[3]/day]
                                60 [g/m[2]/day]
                               3.0 [kg/m[3]/day]
Cell density [kg/m[3]]
                                      0.5
                                       4
                                      0.5
                                       4
                                      0.5
                                       4
Lipid yield [dry wt%]
                                      25%
                                      25%
                                      50%
                                      50%
                                      60%
                                      60%
CO2 consumed [lb/lb algae]
                                       2
                                       2
                                       2
                                       2
                                       2
                                       2
Operating days/yr
                                      330
                                      330
                                      330
                                      330
                                      330
                                      330
Solar exposure hours/day
                                      12
                                      12
                                      12
                                      12
                                      12
                                      12
   1.  Productivity is on an areal basis (g/m[2]/day) for open ponds and a volumetric basis (kg/m[3]/day) for PBR's
The base case shown above represents a reasonable (though still challenging) near-term target for algae productivity and lipid content, while the "aggressive" case assumes identification of a strain with near-optimal characteristics, and the "max growth" case is provided as a comparison to demonstrate the near-theoretical maximum growth and lipid content based on photosynthetic efficiencies [1].  This will be described in more detail throughout the report.

Process Area Descriptions and Major Assumptions for Model/Cost
As noted, there are numerous possibilities at each stage of the production process, and in many cases no clear winner has yet been established.  Thus the proposed process evaluated here is by no means the only way to achieve the end result, but is one of the more "likely" scenarios to be feasible at this scale, and was chosen by conducting an extensive literature review and building off of previous work conducted within NREL.  With this in mind, an overview of each production stage as well as a summary of the chosen process and underlying assumptions at each stage is described below.
Algae Production: Open Pond
Open pond production is commonly employed today at an industrial level for producing algal-based health food supplements and other higher-value products [1].  The main types of open pond systems are summarized below.
   * Unstirred ponds: Unstirred ponds are the simplest in design, as they are essentially comprised of basic uncovered beds or natural lakes.  As there is no circulation or continuous harvesting, they are limited to a small number of microalgae capable of out-competing other environmental contaminants.  These ponds are generally used for wastewater treatment and β-carotene production [4], but would not be amenable to a continuous large-scale fuel process.
   * Circular ponds: Similar to unstirred ponds, circular ponds have been used for wastewater treatment and algae production destined for various health food applications.  Although they are somewhat more robust than unstirred ponds in terms of sustaining a culture, they are limited in size due to a centrally-pivoted rotating arm which provides mixing.  Additionally, due to their circular shape they are slightly less space-efficient which is an important consideration on this large scale.
   * Raceway ponds: Raceway ponds are generally considered the most efficient and low-cost option for large-scale microalgae production [4].  They consist of two or more "passes" separated by a baffle, with circulation usually provided with a paddle wheel to prevent settling.  Raceway ponds are considered "high-rate" due to a much higher achievable productivity than unstirred ponds, and are more space efficient on large scales due to their physical shape and shallower depth. 
These three pond types are shown below in Figure 2.
                                         
                  Figure 2: Schematic of open pond types [4]
Given the brief comparison described above, raceway ponds were chosen for this system.  The basic design of the pond system is based on a compilation of multiple detailed analyses by Benemann and Oswald [5], and essentially targets the lowest-cost design possible due to the relatively low value of the final product (fuel).  Feed (makeup water, nutrients and CO - 2) would be added just after the paddle wheel, and a fraction of the material would be harvested just before the circulation returns back to the paddle wheel.  The pond characteristics (which feed into the cost estimate) are summarized below:
   * Unlined ponds: even a simple plastic liner could double the capital cost of the ponds [6], thus only pond bottom compaction is included, to control percolation and silt suspension
   * Site is on low-value, flat, relatively bare land: minimize cost of leveling and site preparation 
   * Earthen levees with a geotextile liner to prevent erosion
   * 30 cm depth: This is an important parameter as it affects system volume and water usage.  Typical depth is 15-30 cm, with 30 cm optimal to limit CO2 outgassing [7]
   * Mixing by paddle wheels requiring 18 kwh/ha/day of power [B&O]
These characteristics formed the basis of the Benemann and Oswald cost estimate for the ponds.  The B&O study estimated capital and operating costs for every stage of production, and related the resulting cost to "$/hectare" of growth ponds for simplicity; the majority of other subsequent economic analyses used this basis also at every stage, but with different parameters such a method of cost estimating loses its accuracy for downstream units.  Thus this basis was avoided for most costs not related to the growth ponds, and original cost estimates were used instead for this study.  However, for the actual growth ponds and associated equipment, a "$/ha of pond" basis is directly applicable.  Given this basis, the specific cost estimate from B&O for site preparation and pond materials was $6,000/ha, with another $5,000/ha for paddle wheels (note these prices are in 1996 dollars).
Equally important as the physical design of the pond system is the resulting algae yield and lipid content that could reasonably be expected of such a system.  The "base case" representative of near-term potential between now and 2022 was set at a growth rate of 25 g/m[2]/day based on areal pond size, with 25 wt% lipid content.  This was based on NREL's recent Report to Congress summarizing the current state of the technology, which states that the same lipid content and 20 g/m[2]/day is a "reasonable (though still challenging) near-term target" [1].  Other studies have achieved 15-25 g/m[2]/day, but not on an ongoing/sustained basis [8].  Thus it is reasonable to assume that 25 g/m[2]/day at a sustained rate of 330 operating days/year is feasible within a 10-year timeframe.  The assumption of 330 operating days/year also implies that a site location is chosen which receives high year-round solar exposure; two such sites which could sustain this high (90%) operating factor are Brownsville, TX or Tarpon Springs, FL [9], as an example that such an aggressive target is plausible.  In addition, a solar exposure of 12 hours/day was assumed.  While all downstream operations are assumed to operate 24 hr/day, algae undergo growth only during daylight hours and CO2 would be fed only during this time (although the ponds or PBR's would be continuously mixed 24 hr/day to prevent settling).  
Two other cases were also studied as shown in Table 1, representing feasible longer-term research advancements in strain improvement and the theoretical maximum that could possibly be achieved based on photosynthetic efficiency limitations [1].  The open pond case was assumed to reach a cell density of 0.5 g/L, held constant across each scenario [10].  Additionally, in all cases a CO2 requirement of 2.0 g/g algal biomass was assumed [1], a stoichiometric ratio independent of production method.  This value corresponds well to another study which uses a ratio of 1.83 assuming 100% CO2 uptake efficiency [11], thus a ratio of 2.0 allows for 90% uptake efficiency which is typical.
Algae Production: Closed PBR
The other means to produce microalgae is by way of a closed photobioreactor (PBR) system.  While this option offers many distinct advantages over an open pond configuration, it is generally stated that the cost can be an order of magnitude larger than for open-pond systems [5] and as such there have been very few rigorous techno-economic studies done for PBR production.  A brief comparison between open pond and PBR production is given below [4, 12]:

                                   Open Pond
   * + Lower capital and maintenance cost
   * + Easy to scale up (add more ponds)
   * + Easy to integrate with wastewater treatment
   * + Readily available technology
   * - High water evaporation losses
   * - Low cell density = high land requirement and harvesting cost
   * - Susceptible to invasive species (limits the number of viable strains and can derail economics if too many re-starts are required)
                                      PBR
   * + Higher cell density = lower land use and downstream processing cost
   * + Higher volumetric productivity (higher surface area/volume ratio)
   * + Reduced contamination risk (not open to environment)
   * + Low evaporative losses (but temperature control can be a problem)
   * - Higher capital and maintenance cost
   * - Requires degassing to limit O2 accumulation (which inhibits growth)
   * - Not demonstrated on large-scale


Although a closed PBR system mitigates contamination risk and reduces harvesting cost through a much higher cell concentration, the capital cost is usually expected to outweigh these benefits.  While there are many possible PBR configurations, they can broadly be classified into two groups:
   * Flat plate: Flat plate PBR's are made up of a clear plastic containment system, such as a rectangular box or series of plastic bags housing the microalgal culture.  They usually are designed to allow for a short light path to maximize light exposure.  Flat plate PBR's can have lower capital cost than tubular reactors due to their more simple design, but may not be as readily scalable [13].
   * Tubular: Tubular PBR's have historically been used the most frequently for pilot scale PBR studies [4].  They consist of clear (plastic or glass) tubes that carry a circulation of culture between degassing and harvesting.  They can be arranged in a number of different configurations (vertical, horizontal or inclined) to maximize efficient land use or sun exposure.  Although productivities are comparable, some studies have found tubular PBR's to have higher productivity than flat plate reactors [4].   Additionally, they are generally considered more feasible for large scale use as they are modular and accommodate higher flows simply by adding more parallel passes [2, 13].
These PBR types are depicted in Figure 3.
                                       
     Figure 3: Depiction of flat plate (left) and tubular (right) PBR [14]
Unlike raceway ponds for the open pond case, no clear winner has been universally established yet for the PBR type.  Based on the above comparison (notably the fact that a tubular system may be more amenable to large-scale production), a horizontal tubular PBR was chosen here.  In this system, the culture would be circulated through the tubes, with degassing columns located throughout the system to strip out oxygen that accumulates as a product of photosynthesis.  CO2 would be injected immediately following the degassing columns.  The system forms a closed-loop recycle scheme with a fraction of the circulation drawn off for harvesting.  This flow configuration is shown below in Figure 4.
                                       
            Figure 4: Basic flow scheme for airlift-driven PBR [15]
The specific tubular PBR characteristics for the design assumed here are summarized as follows:
   * Tube ID = 8 cm: The diameter is restricted to <10 cm to ensure adequate light penetration, and various other studies assume 6 cm for a smaller-scale plant [2]
   * Tube section length = 80 m: The maximum tube length in a single section between degassing zones is 80 m, as too much oxygen accumulates at longer lengths which inhibits cell growth [11].  Given these dimensions for a single tube, the total number of tubes required can be calculated based on the total system volume (which is set by the volumetric productivity). 
   * 1 hectare land use = 500 m[3] culture volume: This is the translation between culture volume and areal footprint of the tubular system, specific to one study [16].  This also corresponds closely to a second study [11], thus the same footprint was again assumed here.
   * Circulation via airlift columns: Airlift columns work by pumping air into the bottom of a riser column, which imparts a lower bulk density to the fluid and causes the fluid to rise.  At the top of the column the liquid spills over into a downcomer, which in turn provides head to circulate the fluid.  This eliminates any shear force associated with mechanical pumps that could damage or kill the cells.  In addition, this doubles as a degassing column (which is required anyway), by allowing the oxygen entrained in the liquid to disengage and exit out of the top with the bubbled air [2].  As mentioned previously, an airlift column would be required every 80 m to degas the accumulated oxygen.
   * Temperature controlled via sprinkler system: The temperature can rise over the length of the tube and must be controlled to 20-30 °C [2].  Various methods have been proposed to control the temperature; the more popular methods are either to submerse the tubes in a pool of water or use a conventional sprinkler system to spray the tube surface (evaporative cooling).  While the latter may only be feasible in dry climates, it would be far cheaper and is the option chosen by various other PBR studies [16, 17].
Given this basis for the physical configuration of the system, the PBR costs could then be estimated.  As noted, there have been very few thorough economic analyses done for PBR systems, thus costs were either scaled off of a basis if available or estimated using engineering judgment.  Consequently there is a higher degree of uncertainty around capital costs for this system, and this is captured in the sensitivity analysis later in this report.  
The largest cost item was the material cost for the tubes themselves.  While there are various materials that could be feasible for the tubes, one option commonly assumed is clear plastic LDPE with a UV-protectant coating [16].  Given that LDPE is not frequently used for rigid piping (especially in bulk), this cost was approximated using bulk HDPE pipe at $1.50/ft [18].  This assumption can be justified in that HDPE and LDPE resins have a very similar price [19], and any marginal extra cost associated with UV coating would likely be offset by the fact that the quoted price was for a slightly larger pipe size.  A 1.5x installation factor was applied to this base estimate to account for modular installation as well as manifolds and valves [16].
The next cost assumption made was the cost for airlift columns.  These columns are typically used only for special applications and are custom-built for the purpose; consequently a bulk cost was not available.  Other studies have assumed a cost number for this item, ranging from 20-200% of tube material cost [20, 16].  Given the fact that there would be far fewer airlift columns relative to the number of 80m tube sections (roughly one million tubes) and that the airlift columns should also be fairly simple in design, the low-end estimate seems much more reasonable here, 20% of tube material cost.  In the Aspen model, a power requirement assumption for air compression was also required; the assumptions here were to introduce air into 20 ft tall columns at a 1:1 volume ratio of air:liquid.
The cost and water use associated with sprinkler cooling was estimated based on previous TE studies [16, 17].  The capital cost used in the prior study was for agricultural sprinklers at $2,500/ha of PBR area; while this may not fully cover every inch of pipe, it is a reasonable first-pass assumption to provide enough cooling to ensure proper performance.  The water requirement used in Aspen to estimate circulation pump power was 200 L/min/ha.  An evaporative water loss at 1% of this circulation was assumed, which requires fresh water makeup at the same rate.
The final cost item is a simple plastic liner on the ground.  To maximize light utilization, the ground beneath the PBR system is usually painted white or covered with a white liner [11].  A liner was chosen here to also provide simple drainage, and the cost was estimated at $3,200/ha [16].
Again, a critical element impacting the overall production cost is the specific microalgal growth that can be expected of the given system described above.  Productivity has been shown to range from 1-1.5 g/L/day [21], with a common value for maximum productivity around 1.25 g/L/day [22].  As these numbers were again a short-term maximum rather than a sustained result, it was assumed here that a reasonable near-term "base case" target would be 1.25 g/L/day sustained over 330 days/year.  The same "base case" algae strain is assumed as for open ponds, with 25% lipid content.  Similar to the open pond scenario, an "aggressive" and "max growth" case was also evaluated in a sensitivity study to evaluate potential improvements in PBR production cost, with parameters shown in Table 1.  The algal productivity for these additional cases was originally only given for the open pond scenario, thus various assumptions were made to estimate reasonable corresponding values for the PBR scenario.  The "aggressive" case was estimated by assuming the same percentage improvement in productivity as was given for open ponds, while the "max growth" case was calculated by scaling the "max" open pond productivity by the ratio of (surface area/volume) for PBR tubes vs ponds given the specific design of each.  All PBR cases were assumed to achieve a cell density of 4 g/L [11]; this is an 8x improvement over open ponds, and results in drastically lower water use and consequently downstream harvesting cost.
Water Delivery
Water usage is a critical sustainability metric in microalgal biomass production.  Given that the algae grow in a very dilute medium, a large-scale production facility would process vast amounts of water.  Even if water losses in the system were kept to a relative minimum, a substantial amount of makeup water could still be required; thus water source and means of delivery is key, especially in a dry climate.  To that end, a decision was made for the base case to assume low-value brackish or saline water is pumped from an underground source rather than drawing from fresh water resources (the latter case is considered in a sensitivity study).  Although relatively little is known about saline/brackish groundwater resources, there are many areas in the US with relatively easy access to such water, and the ability of certain microalgal strains to grow in brackish or saline water means that there is no need to further strain freshwater resources [1].  This option of pumping water from underground was analyzed by Benemann & Oswald [5], and the same capital cost is assumed here; namely $4,200/ha (1996$) for open ponds, which translates roughly to $72/(m[3]/day) on a flow rate basis for the PBR case.  This includes pumps and all infrastructure required to deliver the water from underground, although a specific depth was not specified.  The associated pump power requirement used in Aspen was back-calculated to be 37.5 kw/(MM m[3]/yr) of water [5].  Being pumped from an underground aquifer, the brackish water has no additional cost beyond capital and operating cost of delivery.
Overall makeup water demand was determined for the open pond case by assuming a 0.3 cm/day rate of evaporation [1] and 95% process water recycle.  95% water recycle implies 5% water lost in addition to evaporative losses; of the 5% loss, a small fraction is water entrained in the lipid product and the remainder is assumed for mass balance purposes to be blow down to treatment (anaerobic digestion), to control contaminant buildup in the system.  For the PBR case, being a closed system there is no evaporative loss of process water, thus brackish water makeup is only required to replace the 5% draw-off.  If the water has a high salinity, using the same saline water rather than fresh water to make up for blowdown would result in salt accumulation in the system; however, it is assumed here that the brackish water has sufficiently low salinity in keeping with the original source study [5] (the cost difference is small either way).  Finally, as noted above there is a separate sprinkler water circulation for PBR cooling; this water must be fresh (non-saline) water to prevent scaling on the outside of the tubes.  The total sprinkler water circulation was set to 200 L/min/ha [17], of which 1% was assumed to be lost to evaporation and must be replaced with fresh water (priced at $0.0013/gal [23]).
CO2 Delivery
Another important consideration is CO2 delivery.  Again for this item, the analysis conducted by Benemann & Oswald is still the most detailed to date, and the same cost basis is used in this analysis [5].  There were two options considered in the B&O study for CO2 delivery, one scenario assumed using the total flue gas material while the other assumed that pure CO2 is concentrated out of the flue gas.  In the "flue gas" case, the flue gas from a nearby power plant or other emissions source is piped 2.5 km (roughly 1.5 mi) to the facility at low pressure.  A cost of $5,000/ha of ponds is given which includes pipes and blowers.  Additionally, to ensure efficient CO2 uptake and minimize outgassing in the case of open ponds, the gas is transferred to the liquid by way of a 1.5 m deep sump with a baffle and CO2 sparger, costing an additional $5,000/ha (again note all prices from B&O are in 1996$).  The associated power requirement to transport the flue gas this distance is given as 15,400 kwh/ha/yr.   
The second option considered in the B&O study is a "pure CO2" case, where instead of transporting the entire flue gas material, the CO2 is scrubbed out and transported under pressure to the facility.  As the piping and valving system for this case would be more simple than for the flue gas case, a cost of $300/ha is quoted for supply and distribution.  Additionally, for open ponds the gas again must be efficiently transferred by way of a sump, costing slightly less than the flue gas case at $4,000/ha.  The B&O study did not include a power estimate for the "pure CO2" case in a similar manner as the "flue gas" case; instead, all operating costs associated with concentrating CO2 out of flue gas (i.e. by a method such as amine scrubbing) as well as transporting the CO2 are captured in an overall CO2 material cost of $40/tonne, which although a rough value corresponds well to other estimates [5, 49].   
For both the open pond and PBR cases, the pure CO2 scenario was chosen as the "base case" for several reasons.  For the PBR case, the choice was obvious given the system cost; the tube cost was already found to be extremely high with just liquid + CO2, so to alternatively use bulk flue gas at a typical 15% CO2 content the amount of tubes required and thus the cost would be prohibitive.  Additionally, utilizing flue gas in general implies that the microalgal culture can tolerate the nitrogen compounds (NOx), sulfur compounds (SOx), and other minor contaminants.  While there is some speculation that the NOx in flue gas could be used to meet the nitrogen requirements for the microalgal culture, this has not been proven and additionally the use of bulk flue gas could result in pH imbalances [12, 24].  For these reasons as well as to provide a consistent basis for comparison between the cases, the pure CO2 option was used for the "base case" in both scenarios; the flue gas option is also considered for the open pond case in the sensitivity analysis.
As noted, the total capital cost for CO2 delivery and distribution for the open pond case is $4,300/ha [5].  For the PBR case, the delivery portion was retained ($300/ha), with no cost included for sumps as the gas could be directly injected into a pipe header since it is a closed system.  While the cost basis is price/ha of growth ponds for an open pond scenario, as noted above the CO2 requirement per unit algae growth is constant for PBR and open ponds at 2.0 g/g biomass, and the biomass growth rate is the same between the two scenarios for the base case (~158,500 tonne microalgae/yr to produce 10 MM gal/yr lipids); thus the CO2 rate is equal and the open pond land area can be applied to this cost for PBR CO2 delivery.  
The operating cost for the pure CO2 case is based on the $40/tonne estimate described above; however, this basis was slightly adjusted to account for power.  Although the $40/tonne cost implicitly includes all power required to capture and transport the CO2 from the flue gas [5], the actual quantity of power could be substantial.  As the plant power balance is an important consideration in the context of this study, it was decided that the amount of power associated with the CO2 capture step should be explicitly included in the overall power balance in the Aspen model to provide a more representative estimate.  A basis of 250 kwh/ton CO2 was assumed, based on a preliminary estimate of flue gas carbon capture efficiency provided by the EPA [25].  This translates to $16.14/tonne CO - 2 as an electricity cost, which is reasonable at roughly 40% of the total material cost basis.  Thus to ensure that this item was not double-counted, the resulting power was added to the total plant power balance, and $16.14/tonne was subtracted from the cost basis, resulting in an adjusted operating cost of $23.86/tonne CO2.
Harvesting
The next step after production is to harvest and concentrate the biomass to bring the throughput down to a manageable range for extraction.  While there are many options to accomplish this task, it is critical that the selected option be low cost and scalable to process large volumes of water.  The most feasible options for harvesting are summarized below.
   * Bioflocculation: Bioflocculation is a phenomenon where small microalgal colonies spontaneously flocculate into larger colonies, which promotes some degree of settling and concentration (up to 4% solids concentration has been observed [1]).  While this is not sufficient for final concentration, the process would be the simplest and lowest cost option, as the required equipment would essentially consist of settling tanks.  For this reason, most open pond studies assume bioflocculation as a primary separation method [1, 5].
   * Filtration: Filtration has frequently been studied as a means to concentrate microalgal cells, either under pressure or vacuum.  The process consists of particles either flowing through a membrane or being retained on the feed side depending on size.  While filtration of larger microalgal species has been proven on the lab scale to achieve concentration factors up to 180x [22], large-scale continuous filtration is still inefficient, due to the spherical and gelatinous nature of cells and their propensity to clog filter membranes [1, 12, 26]. 
   * Centrifugation: Centrifugation separates solid and liquid phases using centrifugal force.  It is an established technology and is an effective method to concentrate most microalgal species, achieving up to 150x concentration factor [22].  However, it is very energy-intensive and has fairly high capital cost, so its use would be limited to secondary recovery to limit the flow rate processed [5, 12].
   * Flocculation: Flocculation consists of neutralizing repulsive negative charges on the microalgal cells to promote cell agglomeration, using a cationic agent as a coagulant.  The coagulant can be a multivalent salt such as alum, although large doses are required and the agent is not biodegradable which limits downstream processing options.  Alternatively, organic polymers can be used, which in addition to neutralizing the cells act as a bridge to physically link flocs together [12].  After the coagulant is added, the mixture is gently agitated using a paddle wheel or similar device, and the solution is transferred to a settling basin, where the agglomerated flocs either settle by gravity or are forced to the surface using flotation methods.  While the equipment is relatively low-cost, the largest cost can be the material cost of coagulant which can become significant depending on the amount of material processed.
Based on this information, it was decided that at least the open pond case would require primary concentration using bioflocculation.  This is a simple and low-cost process which is important given the large amount of water processed at this stage (roughly 150,000 GPM).  It is a somewhat speculative assumption that this process will work as intended, as the "natural settling" phenomenon is not well-understood and has only been observed on a small scale [1].  However, the general conclusion is that this is the only economically viable option for primary settling on a large scale [1, 5].  Therefore an important aspect of future research is to better understand bioflocculation and how it can be achieved on a large scale using different microalgal species.  For costing purposes, an equation was used for clarifier settling basins by Wang et al. [27], which is based on a compilation done by the EPA of various wastewater clarifiers in the US, using data regression to establish a cost equation based on throughput.  In the Aspen model a conservative assumption was made that the solids are only concentrated from 0.5 g/L to 10 g/L (1%).  For the PBR case, primary concentration is not as critical as for open ponds given that the culture is more concentrated at 4 g/L; however, it was decided that primary settling would also be used for the PBR case given that the simple settling tank has lower capital and operating cost than secondary concentration.  The same outlet concentration of 10 g/L was assumed as the open pond case.
For secondary harvesting, flocculation followed by flotation was chosen for both open pond and PBR cases.  Although a centrifuge is another viable option here, some experts in industry conclude that it will not be the most practical option at this scale due to cost and power use [28, 29].  Still, for comparison a centrifuge is also examined in the sensitivity study.  Chitosan (an organic polymer) was chosen as the coagulant as it is fairly well-studied, relatively low cost, and biodegrades downstream in anaerobic digestion, thus does not build up in the system or cause toxicity concerns [12, 30].  One note to be aware of is that chitosan has a relatively high C:N ratio which could exacerbate a nitrogen imbalance in the digester, in which case the feed to the digester would have to be adjusted for proper operation [31].  
Upon formation of flocs, the agglomerated cells would be collected by skimming the top layer of a dissolved air flotation (DAF) tank.  DAF was chosen over sedimentation because it can achieve higher solid concentrations and also requires residence times on the order of minutes rather than hours, greatly reducing equipment size and cost [32].  DAF essentially works by pressurizing the liquid feed along with air, sending the mixture to a retention tank to allow the air to dissolve, then reducing the pressure across a valve which releases the dissolved air as very small bubbles.  These bubbles attach to the flocs as they rise and bring them to the surface where they are skimmed off the top.  For an algae process, a recycle scheme is optimum as the pressurizing pump is located on the clarified water line rather than the feed, thus the cells are not subjected to the pump's shearing action.  This configuration is shown below in Figure 5.
                                       
                    Figure 5: DAF process with recycle [33]
This process shown in Figure 5 was modeled in Aspen using the following assumptions:
   * 10% solids concentration: This represents a 10x concentration factor and is based on one source stating 8-10% could be achievable [5].  Although published data is scarce on actual results for a DAF system concentrating algae using chitosan, this appears to be a reasonable target.
   * Pressurizing to 5 atm: Typical value given by a vendor [34] 
   * Recycle ratio = 50% of incoming flow: Typical recycle value [33]
   * Air/solid ratio = 0.01 mL/mg: Typical value used in wastewater [33]
   * Chitosan required = 40 mg/L solution: This value is specific to the species being cultivated, but is an average of several studies' conclusions of 15-80 mg/L required, with 40 mg/L fairly common [30, 35]; this range is examined in the sensitivity analysis.
   * Power requirement for agitation is the same on a per-flowrate-basis as the paddle wheel power use given for open ponds by Benemann & Oswald [5]: This is assumed, but should be fairly reasonable as a paddle wheel is also used in this application to agitate the coagulant mixture.
Finally, the cost for the DAF unit was estimated in a similar manner to the clarifier cost equation, where a cost based on flowrate was established based on a compilation of DAF units in wastewater applications throughout the US [27].  The material cost for the chitosan was set at $4.84/lb [36].



Extraction
After the cells are concentrated, the lipid stored inside the cell is extracted and purified to produce the oil product.  The extraction step is commonly regarded as the most speculative in terms of large-scale feasibility, but also among the most costly [5]; thus extraction will be a critical area of research going forward to achieve practical algal lipid production.  Existing technology commonly employed to extract seed oils (e.g. from soy beans, canola, etc.) does not lend itself well to algae oil extraction [1].  Plant-based oils are extracted by crushing the seeds then using mechanical presses and/or hexane extraction to recover the oil.  The process is not as simple for microalgae, given that a cell is very small and gelatinous with a thick cell wall protecting the lipids, and even after concentration the solids content is still only 10-20%.  Some of the more common methods discussed for algal oil extraction are summarized below.
   * Solvent extraction: As noted, solvent extraction is a common process used to extract seed oils.  The process involves using an organic solvent such as hexane to exploit the miscibility of the oil phase into the solvent.  The oil and solvent are then separated downstream and the solvent is recycled.  While this method is plausible for use in an algae application, it would only be effective if the suspension was dried to a low water content [1].  This would introduce significant energy costs, both to dry the material as well as to separate the solvent off downstream using distillation.  This has led some in industry to conclude that this method will not be economical on a large scale [29].
   * Supercritical Fluid Extraction: Supercritical CO -  - 2 could eliminate the problem of downstream distillation to separate the solvent from oil.  Supercritical CO2 has properties of liquid and gas, allowing it to penetrate the cell wall and act as an organic solvent without forming a single liquid phase requiring subsequent separation [1].  However, its use has largely been limited to removing caffeine from coffee, and is currently too expensive and power-intensive to be feasible for a lower-value product such as fuel.  However, this would be an ideal application for the technology if further advancements are made in the future.
   * Mechanical Extraction: Mechanical extraction employs some type of device which disrupts the cell wall.  Even if not used as the only extraction technique, some form of mechanical extraction would likely be required to weaken the cell wall for a subsequent step such as solvent extraction [22].  Mechanical devices could include crushers and presses similar to what is used in seed oil extraction (but only after dewatering) or alternatively could employ sonication, homogenization, or bead milling to treat the entire wet slurry [1].  Sonication consists of using high-frequency sound waves to create cavitation bubbles which rupture the cell; however, this is a new technology which needs further development before it becomes feasible.  Pressure homogenization uses rapid pressure swings to lyse the cell, and is commercially available on a fairly large scale [37].  Bead milling involves mixing the slurry with small beads to lyse the cell using shear forces; this could create problems separating the beads back out, and also is only available on a relatively small scale [26].
Given these options, it was decided that mechanical extraction using pressure homogenizers would be the primary method to extract the oil from the biomass.  Solvent extraction, although proven, would create unnecessary additional drying and separation steps, and is not seen as the appropriate solution at large scale as described above.  Homogenizers are commonly used in the dairy industry and are also being introduced in the wastewater industry to lyse bacteria [38], thus the technology is established and has been shown to work for a similar application.  Furthermore, of the wet processing devices they are available at the closest scale to what is required here.
As described above, the goal of homogenization would be to lyse the cell and release the contents.  Once the oil is released, it would be purified and removed using a simpler form of solvent extraction described by Benemann & Oswald [5].  The process as envisioned would utilize hot oil as the water-immiscible solvent to extract the lipid phase, similar to existing technology used for extracting beta-carotene.  However, this would take it a step further and use the lipid product itself as the oil solvent, recycling and heating a portion of the purified product and sending it back to the mixture.  It is important to note that although this is a novel technology as applied to lipid extraction and hasn't been proven on a large-scale, the underlying fundamentals are well established for beta-carotene purification.  One difference from the Benemann study is that after the recycled oil is introduced to the mixture, the Benemann study goes on to separate the phases (oil, water, and spent biomass) using a 3-phase centrifuge, which would have a high capital cost and power requirement.  Since the water phase will now contain high amounts of dissolved proteins and carbohydrates that were released from the cells, it should not be recycled back to the process anyway, but should be treated as wastewater.  Thus it is not important to separate the water and solid phases, and a simple clarifier settling tank would suffice [39].  This is somewhat similar to the concept presented by OriginOil, where the cells are ruptured using electromagnetism then the various phases simply settle out without mechanical aid [40].
The underlying assumptions behind the extraction step are summarized below:
   * 90% cell rupturing efficiency: While a maximum cell rupture of 85% is suggested to be currently possible [37], it is assumed here that with advancements in technology as well as homogenizers designed specifically for algae processing, this can be increased to 90%
   * Temperature rise across homogenizer = 25 °C: Per vendor [41].  This is advantageous in that the extractability of the oil phase increases with temperature [5]
   * Homogenizer power requirement = 200 kwH/tonne solids processed: Per vendor [42]
   * Lipid recycle ratio for oil extraction = 5:1  -  Conservative estimate based on envisioned recycle ratio of up to 6:1 [26]
   * Recycle stream heated to 65 °C: Hot oil solvent is envisioned to be heated to 60-70 °C [5].  This would be accomplished by cross-exchange with the turbine flue gas stream
   * 5% lipid loss into the water phase: The recovery of beta-carotene using hot oil extraction is >98% [5], so it is reasonable to assume 5% of the lipid is left entrained in the water/solid phase sent to anaerobic digestion.  Additionally, the same quantity of water is assumed to be entrained in the oil phase as a conservative estimate for the water balance.
Given these assumptions, the unit was modeled in Aspen and resulting costs were calculated.  The maximum capacity for the homogenizers was quoted by a vendor at 17 GPM/unit; at a flow rate of roughly 800 GPM through the homogenizers (for both open pond and PBR cases), this equates to 47 homogenizer units.  The total installed cost for each unit was quoted at $1MM/unit, thus the homogenizer system would cost close to $47MM as economies of scale cannot be assumed here (each unit is already operating at maximum capacity).  The clarifier settling tank was costed in the same manner as the open pond primary settling tank, using a flowrate-based cost equation [27].  The resulting streams leaving the extraction clarifier are the oil product and the contaminated water + spent biomass solids.  Although the oil product may require an upgrading step such as degumming to remove phospholipids prior to biodiesel conversion (similar to traditional seed oil biodiesel processing), this step was not included as there is no information in the literature regarding the process logistics specific to algal-derived oil, and the cost would be marginal relative to the other units.
Spent Biomass Utilization
After the cell is ruptured and the oil product is drawn off, the remaining spent biomass material can be utilized in several ways to provide a byproduct credit.  As noted previously, the water will contain high levels of dissolved protein and carbohydrates after the cell is ruptured, and as such it should not be returned to the system.  This is the blowdown stream described above in the "Water Delivery" section.  Thus both the water and solid phases are drawn off and sent to a "byproduct utilization" stage, of which there are two main options commonly discussed:
   * Animal feed: The protein and nutrient content of the microalgae makes the remaining biomass amenable to formulation into specialty animal feeds.  Although bulk animal feed may have too low of a value for sufficient co-product credit, certain nutrients such as carotenoids have a higher value and could be used in specialty feeds [1].  However, some experts in industry caution against this option, as extensive testing for toxicity and other properties would be required and there is no guarantee that the product could be sold into this market [43].
   * Anaerobic digestion: The spent biomass can also be sent to anaerobic digestion to produce biogas, which can in turn be burned in a gas turbine to offset the power requirements for the plant.  This also improves sustainability as it enables the plant to be more self-sufficient and either requires less power from the grid or none at all, depending on plant power requirements.  This has an added benefit of providing partial treatment to the blowdown water as well.
Given these options, anaerobic digestion and power generation is utilized in this study as power usage is another important sustainability metric.  The anaerobic digestion block in the biochemical ethanol Aspen model was assumed to be applicable here [23], so the stream data from the algae model was exported to the ethanol model digester inlet stream, and the digester block was modified to include properties of the new algae components.  For the purposes of anaerobic digestion calculations, the microalgal cell was assumed to be composed of 25% lipid with the remainder split evenly between protein and carbohydrate as a simple first-pass approximation [1].  Of the carbohydrate, the composition was split evenly between glucose and cellulose (for the wall structure); although in reality the cellulose content may be lower [44], this assumption is conservative in that cellulose does not degrade in the digester model and thus does not contribute to the biogas yield.  The output from the digester model was exported back to the algae model, and the stream sent to a flash drum to separate the gas (burned in the turbine) from the liquid (containing nutrients to be recycled and purified water to be discharged).
Process assumptions behind the digestion and gas turbine units are summarized below.
   * Nutrients are required for anaerobic digestion, and the required amount calculated by the ethanol model is still valid for this process.  The nutrients for digestion are composed of "primarily caustic with some phosphoric acid, urea and micronutrients" [23].
   * The remaining nutrients after anaerobic digestion are recycled to the microalgal production stage, to mitigate the amount of fresh nutrients required
   * The water exiting digestion is treated further (described later in "other" capital costs), then discharged from the system
   * Thermal efficiency = 33%: The generated biogas is combusted in a gas turbine with a stoichiometric amount of air.  The resulting thermal efficiency to generate power (work extracted/higher heating value) was assumed to be 33% [5]
   * The turbine flue gas is recycled to the production stage 
After modeling the units in Aspen, capital cost was calculated.  All costs for equipment in anaerobic digestion were calculated in the same manner as is done in the biochemical ethanol study [23], namely by scaling a previous cost estimate by flow rate according to the equation:
New Cost= Base CostNew RateBase Ratef
The gas turbine capital cost was taken from a vendor quote in Benemann & Oswald, stated as $1,000/kw in 1995$ [5].
Other Capital Costs
 In addition to the main process units described above, there would be additional systems and miscellaneous equipment required to support the process which carry additional capital cost, as described below.
Inoculum Production
A culture inoculum would be required for startup as well as to restart the growth process after the system is shut down for cleaning, culture crashes etc.  This is a cost which isn't usually included in most studies; some studies justify this by assuming that the strain is robust enough to not require inoculation [12].  However, this is a step which would very likely be required and it is important to at least estimate what the cost might be.  For this study, a system described by Benemann & Oswald is assumed [5].  The configuration as envisioned would consist of a series of steps cascading from a small lab-scale photobioreactor up to a medium-size covered pond before the culture is released to the full scale open ponds.  The analysis concluded that it would be feasible to build such a system at $9/m[2] and requiring no more space than 10% of the open pond area.  Another conclusion was that at such a small scale, capital cost would dominate and associated operating costs could be neglected.  For the PBR case, it was assumed that only half of the resulting cost for the open pond case would be required, as the final stages using covered ponds would not be necessary for PBR's.

Land Use
After calculating the land use required for the growth ponds themselves, the total plant land use was estimated assuming open ponds require 2/3 of the total land for the plant [5].  This was decreased to 50% for the PBR case, given that PBR's are more space-efficient.  The land cost used was $3,000/acre assuming low-value land not intended for crop production.
Nutrient Delivery
The capital cost for nutrient delivery equipment for the open pond case was set at $1,000/ha in 1996$ [5].  For the base case scenario, this translates to $96/(kg/hr) of algae produced, a basis which can then be applied to the PBR case.
Secondary Waste Treatment
As mentioned previously, after the blowdown water containing dissolved proteins and carbohydrates is treated in anaerobic digestion, a second line item for waste treatment was also included in the Benemann & Oswald analysis [5] and is included here.  The cost stated for this process was $1,000/ha (1996$), which for the PBR case translates to $0.33/m[3] of total production system volume (which is the basis for the amount of blowdown ultimately leaving the system). 
Electrical Supply
A separate line item was included in the Benemann & Oswald study for electrical supply infrastructure [5].  This was originally set at $2,000/ha (1996$) but is decreased here to $1,000/ha given that much of the new equipment costs used here already include all electrical elements.  This is an appropriate basis for both the open pond and PBR case.
General Machinery
The final "miscellaneous" capital cost item included in Benemann & Oswald was for general machinery [5].  While it was not stated what this would include, it is assumed here that this includes pumps to circulate the material throughout the plant; relative to the large units in each processing step, the circulation pumps would carry a marginal cost.  This cost was set at $500/ha (1996$).
Operating Costs
Some of the operating costs have been described previously, but all costs are summarized as follows:
Power
All power requirements throughout the plant are met using electricity.  This is supplied from the gas turbine, and any additional electricity if required would be purchased from the grid.  Power cost was updated to 2007 at $0.06/kwh per the thermochemical ethanol design report [45], and the same price was used for a power credit if the power generated exceeded the total power required.  As described previously, all units downstream of the growth stage were assumed to operate 24 hours/day and electricity use was calculated as such.  For the growth step (ponds or PBR's), nutrients and CO2 would be fed only 12 hr/day corresponding to the daylight hours when algae are grown (thus the energy use associated with CO2 capture described above was based on 12 hr/day operation).  Therefore, given that power (an instantaneous rate of energy) is used at a different rate during the day than at night, it is more appropriate to report the total amount of energy used annually, as shown in the "Results" section.
Nutrients
Nutrients for microalgal cultivation constitute a significant portion of operating cost.  Many other studies assume that for a first-pass analysis, nutrients can be approximated as "fertilizer" [46].  However, the optimal amount of fertilizer (or other nutrient medium) varies widely across different studies, even for the same strain [12].  The minimum (stoichiometric) ratio of nutrients to microalgal cells was calculated assuming cell N content of 5.5% and P content of 1.1% of dry weight [12], met by using urea (46% N) [50] and diammonium phosphate (18% N, 20% P) [51].  This results in a total nutrient requirement of 1:6.5 ton/ton biomass growth, at a composition of 64% urea and 36% DAP.  This ratio was used as the "base case" here, and the variation is studied in the sensitivity analysis.  As noted previously, nutrients from digestion are recycled, thus the amount of fresh nutrients required is lower.  
The associated price for each fertilizer type was extrapolated to the year 2022 according to EIA projected oil prices [47] as these prices are correlated.  The resulting blended fertilizer cost is $819/ton [48], to which an additional 20% was added to allow for additional micronutrients such as iron and vitamins [12].  It is important to note that there are other options for nutrient sources.  For example, it has been suggested that the NOx in the flue gas could potentially provide the required nitrogen, but this has not been proven and could present additional pH problems [12], thus this was not examined further.  Additionally, the plant could be set up to treat wastewater, where some nutrients would already be present and thus fresh nutrient requirements would be either mitigated or eliminated.
CO2
As described previously, both cases assume CO2 purified from the flue gas at an adjusted cost of $23.86/tonne, which accounts for upstream operating costs (excluding power) to purify the CO2 from flue gas.  The power associated with this step was estimated at 250 kwh/ton CO2 [25].  The sensitivity of the oil production cost on this CO2 price is evaluated in the sensitivity analysis.  Additionally, an alternate option of piping the total flue gas stream to the facility is considered in the sensitivity analysis, and in this case the flue gas is assumed to be free beyond the power required to get it there.
Coagulant
Chitosan is used as the harvesting coagulant in DAF for reasons described above, and is required at 40 mg/L and $4.84/lb.  As both PBR and open pond production utilizes primary settling to concentrate the biomass to 1% prior to the DAF unit, the same amount of chitosan is required for either case.
Waste Disposal
A marginal operating cost associated with miscellaneous waste disposal was included in the Benemann & Oswald study [5] at a rate of $1,000/ha (1996$).  This is assumed here for open ponds, and translated to PBR's at $0.48/(m[3] system volume)/yr.
Water
As described previously, for the base case brackish water is pumped from underground and assumed to be free beyond the associated power requirement.  Thus the only water cost is for the fresh makeup water to replace evaporative losses in the cooling system for the PBR.  Fresh water price was based on the biochemical ethanol design report at $0.0013/gal [23].  An alternate option of purchasing all required water as a utility at this rate is considered in the sensitivity analysis.
Labor and Overhead
The number of operators required to run the plant was estimated as follows:
   * Open ponds: 50 ponds/operator (results in 26 operators)
   * PBR: 25 hectare/operator (results in 31 operators)
   * Primary settling: 1,000 man-hours/10,000 ft[2] of settler tank area [27] (15 operators for open pond, 3 for PBR)
   * DAF: 2,800 man-hours/1,000 ft[2] of DAF area [27] (5 operators for open pond or PBR)
   * Homogenizers: 4 operators required (assumed)
   * Product clarifier tank: 1,000 man-hours/10,000 ft[2] of settler tank area [27] (1 operator)
   * Digester/turbine: 4 operators required (assumed)
All other employees were set at the same number and pay rate as the biochemical and thermochemical ethanol reports [23, 45].  The overhead rate was set at 60% of labor cost, per the biochemical model.
Maintenance, Taxes, Insurance
For the open pond case, maintenance was set at 2% of the installed equipment cost per the biochemical model.  Insurance and taxes were set at 1.5% of the total installed cost per the same basis.  For the PBR case, given that the capital cost is much larger, using this same basis would result in unrealistically high rates.  Thus twice the resulting cost as open ponds was assumed, given that maintenance for PBR's would be higher than for open pond.
Financial Parameters
The remaining financial parameters assumed for this study are as follows:
   * Indirect capital cost: rates based on the thermochemical and biochemical models [23, 45]:
         # Open pond
               # Site development: 9% of installed cost of process equipment
               # Warehouse: 1.5% of total installed equipment cost
               # Prorateable  costs: 10% of total installed cost
               # Field expenses: 10% of total installed cost
               # Home office and construction: 25% of total installed cost
               # Contingency: 10% (given the unknowns associated with algae production)
               # Other costs: 10% of total capital investment (see Aden et al report for a description of these items)
               # Working capital: 25% of operating costs [5]
         # PBR: Set all indirect capital expenses to same value as open pond case, as these items would all be fairly similar and using the same basis for PBR's would be unrealistic
   * 10% return on investment
   * 20-year plant life
   * 35% tax rate
   * MACRS depreciation
Results
Base Cases 
Given the assumptions described above, the resulting oil production cost is shown in Figure 6 for all three scenarios presented in Table 1 (note TAG= triacylglyceride, the oil constituent).  For consistency with NREL's other current studies, all costs are in 2007 dollars.
                                       
Figure 6: Cost to produce 10 MM gal/yr oil at each growth scenario for open pond and PBR production
The oil production cost for the open pond case ranges from $11.25/gal in the current case to $3.99/gal for the max growth case; this corresponds well with other studies [3].  The cost for PBR production is roughly twice the cost for the open pond case.  To put these costs into a larger context, it is important to note that "max" in this case merely means the maximum algae growth and oil content (based on efficiency limits) applied to this specific configuration and associated assumptions.  Thus while it results in the lowest costs for this study, it does not imply that these are the absolute lowest costs that can ever be achieved as there are numerous other process options and even unforeseen additional improvements that could be realized as the technology develops.  
The results from Figure 6 show that cost of land is marginal in all cases, while for the open pond case the majority of the cost is split fairly evenly between capital and operating costs.  The PBR case is dominated by capital cost of the tubular production system, although it also has a higher operating cost than open ponds; this is mainly due to higher maintenance, tax, and insurance cost (a reflection of higher capital cost).  The specific cost breakdown, as well as a summary of land and resource requirements, is shown further below in Table 2.
Table 2 demonstrates the advantage of PBR's over open ponds with respect to land and water use, with PBR's requiring <10% of the water demand for the open pond scenario (more than half of which is lost to evaporation in the base case), and roughly 50% of the open pond total land requirement.  While the CO2 requirement per unit weight of microalgae is constant, the required rate of algae production decreases in the more aggressive cases due to higher lipid content, thus the CO2 demand also decreases.  The overall flue gas rate required to meet the pure CO - 2 demand can be calculated for a given flue gas composition.  At a bulk vol composition of 15% CO2, 6% O2, and 79% N2 representative of one particular flue gas [24], the flue gas rate required to meet the CO - 2 demand shown here would be approximately 1.3  -  0.7  -  0.6 MM ton/yr, for each growth case respectively.  Note that the CO2 demand shown in Table 2 accounts for recycling the turbine flue gas to the production stage; in this respect, it could be considered that the plant itself has "zero" air emissions, and the only other stream leaving the plant (beyond the lipid product) is treated water at the "blowdown" rate shown in Table 2.
As noted previously, the fertilizer nutrient requirement was calculated at a minimum stoichiometric ratio of 1 ton/6.5 tons biomass grown, which is further decreased by recycling the nutrients from anaerobic digestion.  This implies that the nutrients from anaerobic digestion can provide the same cell growth rate as the actual fertilizer nutrients.  Thus it is important to note that the fertilizer requirement shown in Table 2 is an optimistic assumption due to these two factors (although this can be justified given that the nutrient cost calculated on the basis presented by Benemann & Oswald [5] would be even lower).  Even under these optimistic assumptions, the nutrient cost is still a significant portion of the total operating costs (roughly 40%).  The cost impact of adjusting the nutrient requirements is presented in the sensitivity analysis.
In all cases evaluated, there was a net electricity import required after considering all power demands in the plant (including CO2 capture) beyond what can be met with the gas turbine, thus there was an associated cost to purchase additional electricity from the grid.  This stands in contrast to other studies which typically show a net power export [5], however these studies do not consider the CO2 capture step in the power balance, which in this study was found to be the single largest power draw in the plant; if CO2 capture were not explicitly included here, all cases would show a net power export.  It is important to note that additional assumptions had to be made in the Aspen models (which otherwise had no effect on the mass balances) in order to estimate power usage as outlined in the process area descriptions above.  Thus the electricity requirement presented in Table 2 is more accurate as a qualitative comparison (e.g., the open pond case requires more power due to the higher amount of water being processed) rather than on an absolute quantitative basis.  
Table 2: Summary of results for microalgal biomass production at each growth scenario
Basis = 10 MM gal/yr lipid production
                                   Base case
                                Aggressive case
                                Max growth case

                                   Open pond
                                      PBR
                                   Open pond
                                      PBR
                                   Open pond
                                      PBR
Yield
                                       
                                       
                                       
                                       
                                       
                                       
Lipid yield [gal/ton algae produced] 
                                      57
                                      57
                                      114
                                      114
                                      138
                                      138
 
                                       
                                       
                                       
                                       
                                       
                                       
Land Use:
                                       
                                       
                                       
                                       
                                       
                                       
Pond/PBR land size [acre]
                                     4,743
                                     1,897
                                     1,482
                                      593
                                      823
                                      329
Total plant land required [acre]
                                     7,079
                                     3,795
                                     2,212
                                     1,186
                                     1,229
                                      659
 
                                       
                                       
                                       
                                       
                                       
                                       
Resource Assessment:
                                       
                                       
                                       
                                       
                                       
                                       
Net water demand [MM gal/yr]
                                     9,740
                                      720
                                     3,830
                                      320
                                     2,710
                                      250
Net water demand [gal/gal lipid]
                                      974
                                      72
                                      383
                                      32
                                      271
                                      25
  -Water evaporated [gal/gal lipid]
                                      556
                                      19
                                      174
                                       6
                                      96
                                       3
  -Water blowdown to treatment/discharge [gal/gal lipid]
                                      418
                                      52
                                      209
                                      26
                                      174
                                      22
 
                                       
                                       
                                       
                                       
                                       
                                       
CO2 consumed [lb/lb biomass produced]
                                       2
                                       2
                                       2
                                       2
                                       2
                                       2
Net CO2 used from offsite flue gas [ton/yr][1]
                                    290,000
                                    290,000
                                    150,000
                                    150,000
                                    120,000
                                    120,000
 
                                       
                                       
                                       
                                       
                                       
                                       
Nutrients (fertilizer) required for algae [ton/yr][2]
                                    23,920
                                    23,880
                                    12,000
                                    11,980
                                    10,010
                                    10,000
Nutrients required for anaerobic digester [ton/yr][3]
                                     2,960
                                     3,000
                                     1,440
                                     1,460
                                     1,190
                                     1,200
Chitosan required for flocculation [ton/yr]
                                      690
                                      690
                                      350
                                      350
                                      290
                                      290
 
                                       
                                       
                                       
                                       
                                       
                                       
Net annual electricity imported [MM kwh/yr][4]
                                     60.2
                                     35.7
                                     27.0
                                     18.8
                                     19.8
                                     16.0
 
                                       
                                       
                                       
                                       
                                       
                                       
Production Cost:
                                       
                                       
                                       
                                       
                                       
                                       
Overall lipid production cost [$/gal]
                                    $11.25
                                    $19.49
                                     $5.11
                                    $10.07
                                     $3.99
                                     $6.62
  -Land cost [$/gal]
                                     $0.38
                                     $0.21
                                     $0.12
                                     $0.06
                                     $0.07
                                     $0.04
  -Capital cost [$/gal]
                                     $5.10
                                    $13.21
                                     $2.16
                                     $6.75
                                     $1.63
                                     $4.07
  -Operating cost [$/gal]
                                     $5.76
                                     $6.07
                                     $2.83
                                     $3.26
                                     $2.29
                                     $2.51
   1.           After recycling flue gas from gas turbine
   2.           Net requirement after recycling digester nutrients.  Fertilizer composition approx. 64% urea, 36% DAP [12]
   3.           Primarily caustic with some phosphoric acid, urea, and micronutrients [23]
   4.           After supplementing with power from gas turbine; includes CO2 capture step.  Imported power is purchased from grid
Sensitivity Analysis
Given that there are still many uncertainties in developing a microalgal oil production technology and thus a number of assumptions had to be made, the potential cost impact of such assumptions can be captured in a sensitivity analysis.  Individual parameters were varied from the established base case over a reasonable range, and the results are shown in Figures 7-8.
                                       
                   Figure 7: Open pond sensitivity analysis
Figure 7 shows that for open ponds, the amount of nutrients required has the highest impact on production cost of the variables evaluated.  As noted previously, the total nutrient requirement was set at a minimum ratio of 1 ton/6.5 tons biomass produced, with fertilizer being used as the primary nutrient source after recycling the nutrients from anaerobic digestion (which was a minimal amount at roughly 10% of the total nutrient requirement).  However, to avoid nutrient limitation, nutrients may be added in excess and the optimum amount varies widely [12].  For example, another study states that it is reasonable to assume that an "average" microalgal strain will require fertilizer at a ratio closer to 1:3 [46].  At such a high ratio, the study acknowledged that the associated nutrient cost alone could cause a project to become infeasible.  This high cost sensitivity is reflected in the figure above.  Even though the 1/6.5 ratio assumed here appears to be the minimum stoichiometric amount and could thus be optimistic, it is a reasonable first-pass estimate as further justified by the fact that the nutrient cost calculated on the basis of the Benemann & Oswald study [5] would actually be on the opposite end of the scale, roughly 5x lower than even this cost calculated here (shown as the "low" limit above).  This wide lack of agreement in the literature highlights that this is an important aspect of future research.
The rest of the parameters evaluated are fairly self-explanatory, but can be summarized briefly.  The option of delivering flue gas rather than concentrated CO - 2 was evaluated with the process differences highlighted previously, and the resulting cost impact is shown to be fairly low, roughly $0.50/gal improvement.  The water source was also examined, with a sensitivity around purchasing water as a utility at the rate outlined in Aden et al [23] and eliminating the somewhat substantial power requirement associated with pumping water from deep underground.  This increased the production cost which further demonstrates that this is not the optimum choice for water source.  Next, the amount of chitosan coagulant required was studied, as the optimum amount is not well-published.  The few studies that were done have found that the amount varies by strain but falls mostly between 15-80 mg/L , thus this was the range evaluated here.  
The last three parameters shown in Figure 7 are related to capital cost.  As mentioned, the optimum technology for the extraction step is still fairly speculative, and homogenizers are only one feasible option.  The base cost assumes that the entire material must be homogenized to achieve 90% cell disruption and subsequent lipid release.  However, if it were possible to achieve the same results by only processing half of the material (analogous to only "weakening" the cell wall), the high capital and power cost of homogenization could be reduced, resulting in nearly $1/gal improvement.  Next, the Benemann & Oswald study [5] assumed the same extraction scheme with the exception that a 3-phase centrifuge was used for phase separation instead of the simple clarifier assumed here.  The additional capital and power use results in a marginally higher production cost.  Finally, the cost impact of the inoculum system was investigated.  As noted, many studies do not include inoculation cost, and additionally the cost assumed here was based on very rough assumptions.  However, the cost impact is not extremely large, as eliminating this unit would result in roughly $0.75/gal improvement.
Figure 8 shows the results of the sensitivity analysis for the PBR system:
                                       
                      Figure 8: PBR sensitivity analysis
As shown in the figure above, the single largest cost item in the PBR system is the cost of the tubes themselves, as this cost has a very high economic impact.  As described previously, the base cost of $1.50/ft is for a similar-size HDPE pipe, as an approximation to the specialized UV-coated LDPE material.  While this is a fairly low price per foot, the system would be composed of nearly one million 8cm ID x 80m pipes to accommodate the design specifications listed above in the "PBR" section, thus this cost becomes very large.  The +50% range results in a +25% overall production cost impact, implying that the cost of the tubes alone accounts for nearly half of the production cost.  Further compounding the large uncertainty here is that the price quoted in other studies varies even more than this 50% range.  For example, one study used a cost equivalent to $1.45/m[2] of tube surface area which would result in a $10/gal cost reduction relative to the baseline here [16], while another study assumed $150/m[2] on an areal basis, which would increase the cost by $7.50/gal [12]; furthermore the latter study stated that such a cost basis was even quite optimistic.  Such wide variations highlight the challenge in determining a credible production cost for PBR's at this stage.
The next cost item evaluated was the cost of the airlift columns.  As these are generally specialized and custom to the application, a cost estimate was not available for these columns.  Thus the cost was assumed to be 20% of the tube cost given that these would be fairly simple in design and far fewer would be required than the number of 80-meter tubes.  It would seem feasible for this cost to range anywhere between 5-50% of the tube cost, and over this range the cost impact is not overwhelming.
Many PBR studies assume the harvesting would be done in one step using centrifuges since the biomass is more concentrated for PBR's than open ponds [22].  However, for consistency the DAF option was assumed in both cases for this study.  As shown in the sensitivity, the use of a centrifuge instead of DAF appears on a first pass to result in higher production cost, although only marginally.
Another important parameter is the CO2 delivery cost.  As described previously, the base cost assumed was $40/tonne [5] (adjusted internally to $23.86/tonne after removing CO - 2 capture energy), and varying this parameter over a reasonable range from $5-$75/tonne has a cost impact of roughly +$1/gal (the same degree of cost impact is equally applicable to open ponds).  Thus the production cost is somewhat sensitive to CO2 price, although not as strongly as some of the other parameters evaluated.  The remaining operating cost items evaluated in the PBR sensitivity were discussed above for open ponds.
A final note should be mentioned regarding economies of scale.  Although 10 MM gal/yr oil production is not excessively large, the total system throughput is considerably higher due to the dilute nature of the process.  Thus even at the scale evaluated here, the associated flow rates are much higher than the largest wastewater treatment plants in the US [42].  Consequently, all of the equipment would already be modular and multiple units would be required to accommodate the flow rate in excess of any given unit's maximum capacity.  Therefore, it is not expected that economies of scale would play a significant role in reducing the production cost at larger plant scales (for example doubling capacity to 20 MM gal/yr).  However, as noted above if the equipment itself could be improved and tailored specifically to microalgae production, significant cost improvements could result (i.e. developing a low-cost material that could be produced in bulk for PBR production, modifying homogenizers to handle more flow etc).
 

Concluding Remarks
The overall lipid cost shows that in the near-term, the economics of a microalgal biofuel facility may only be feasible if production also targets higher-value coproducts which are produced commercially today.  However, the production cost could improve substantially with identification of algae species capable of maintaining high sustained growth rates with a high lipid content; thus this is one major area important for future research.  Additional critical areas as outlined previously include reaching a consensus on the optimum nutrient requirements, understanding the mechanisms behind bioflocculation to prove its large-scale feasibility, and developing an effective low-cost method for lipid extraction.  Additionally, there is a large potential for cost reduction associated with development of a low-cost material suitable for use in a PBR system. 
The intent of this report was to provide a first-pass analysis of the near-term feasibility of microalgal lipid production, by performing a ground-up estimate of material balances using Aspen modeling as well as equipment costing by individual unit.  This is a more detailed approach than has been taken in most prior studies, but still carries a relatively high degree of uncertainty in both the process steps as well as costing methods, intrinsic to the early state of the technology (most of the processes have only been demonstrated for algae applications on a lab or small scale).  The results presented here are specific to one set of processing options and assumptions, among many other feasible solutions.  However, both the assumptions and results correspond well to other studies [3, 7].

References
[1]  U.S. Department of Energy, "Report to Congress: Microalgae Feedstocks for Biofuels Production" (EISA 2007  -  Section 228).
[2]  Chisti, Yusuf.  "Biodiesel from Microalgae Beats Bioethanol."  Trends in Biotechnology (Opinion) 26 (2007) 126-131.
[3]  Sun, A. et al., "Techno-Economic Analysis of Algae Biofuel Deployment."  Presented at the 2009 Algae Biomass Summit, San Diego, CA (October 2009).
[4]  Shen, Y. et al., "Microalgae Mass Production Methods."  Transactions of the ASABE  52 (2009) 1275-1287.
[5]  Benemann, J. et al., "Systems and Economic Analysis of Microalgae Ponds for Conversion of CO2 to Biomass."  Final Report to the Department of Energy, Pittsburgh Energy Technology Center (1996) DOE/PC/93204-T5
[6]  Craggs, Rupert.  "Wastewater Treatment High Rate Algal Ponds for Biofuel Production: Pilot- and Large- (hectare) scale Research in New Zealand."  Presented at the 2009 Algae Biomass Summit, San Diego, CA (October 2009).
[7]  Benemann, John.  "Growth and Productivity of Algae Biomass."  Presented at the 2009 Algae Biomass Summit, San Diego, CA (October 2009).
[8]  Barclay, Bill.  "Challenges for Algae Biofuel Production: Perspective of an Algae Entrepreneur."  Presented at the 2009 Algae Biomass Summit, San Diego, CA (October 2009).
[9]  Jorquera, O. et al., "Comparative Energy Life-Cycle Analyses of Microalgal Biomass Production in Open Ponds and Photobioreactors."  Bioresource Technology 101 (to be published in 2010) 1406-1413.
[10]  Carlsson, A. et al., "Micro- and Macro-Algae: Utility for Industrial Applications."  Outputs from the EPOBIO Project (2007), ISBN 978-1-872691-29-9
[11]  Chisti, Yusuf.  "Biodiesel from Microalgae."  Biotechnology Advances 25 (2007) 294-306.
[12]  Abayomi, A. et al., "Microalgae Technologies & Processes for Biofuels/ Bioenergy Production in British Columbia: Current Technology, Suitability, and Barriers to Implementation."  Final report submitted to the British Columbia Innovation Council (2009).
[13]  Molina Grima, E. et al., "Photobioreactors: Light Regime, Mass Transfer, and Scaleup."  Journal of Biotechnology 70 (1999) 231-247.
[14]  Willson, Bryan.  "Solix Technology Overview" powerpoint (2009), received by email September 2009.
[15]  Molina, E. et al., "Tubular Photobioreactor Design for Algal Cultures."  Journal of Biotechnology 92 (2001) 113-131.
[16]  Tapie, P. et al., "Microalgae Production: Technical and Economic Evaluations."  Biotechnology and Bioengineering 32 (1988) 873-885.
[17]  Tredici, M. et al., "A Tubular Integral Gas Exchange Photobioreactor for Biological Hydrogen Production: Preliminary Cost Analysis."  BioHydrogen (1998) 391-401.
[18]  Personal communication with Secor (pipe vendor), October 2009.
[19]  From SRI report on polyethylene resin prices (LDPE and HDPE), http://www.sriconsulting.com/CEH/Public/Reports/580.1310/; http://www.sriconsulting.com/CEH/Public/Reports/580.1340/
[20]  Techno-economic analysis conducted by NMSU researchers: http://ias.newmexicoconsortium.org/news2/operations/webdocs/DARPAA-Tech-Transition-Starbuck-020808.doc.pdf
[21]  Eriksen, Niels.  "The Technology of Microalgal Culturing."  Biotechnology Letters 30 (2008) 1525-1536.
[22]  Molina Grima, E. et al., "Recovery of Microalgal Biomass and Metabolites: Process Options and Economics."  Biotechnology Advances 20 (2003) 491-515.
[23]  Aden, A. et al., "Lignocellulosic Biomass to Ethanol Process Design and Economics Utilizing Co-Current Dilute Acid Prehydrolysis and Enzymatic Hydrolysis for Corn Stover" (2002), http://www.nrel.gov/docs/fy02osti/32438.pdf
[24]  Stepan, D. et al., "Subtask 2.3  -  Carbon Dioxide Sequestering Using Microalgal Systems."  Final Report to the Department of Energy (2002), Cooperative Agreement No. DE-FC26-98FT40320
[25]  Personal communication with Dallas Burkholder and William Stevens (EPA), November 2009.
[26]  Hsi, K. et al., "Biofuels from Algae: a Techno-Economic Analysis from NREL."  MIT practice school internal NREL document, May 2009.
[27]  Wang, L. et al.  Handbook of Environmental Engineering: Biosolids Treatment Processes.  New Jersey: Humana Press, 2007.
[28]  Brune, David.  "Algal Biomass Production for Greenhouse Gas Reductions."  Presented at the 2009 Algae Biomass Summit, San Diego, CA (October 2009).
[29]  Butler, Joel.  Presented at the 2009 Algae Biomass Summit, San Diego, CA (October 2009).
[30]  Divakaran, R. et al., "Flocculation of Algae using Chitosan."  Journal of Applied Phycology 14 (2002) 419-422.
[31]  Personal communication with Phil Pienkos (NREL), November 2009.
[32]  Shelef, G. et al., "Microalgae Harvesting and Processing: A Literature Review."  SERI subcontract report prepared for the Department of Energy (1984), contract no. DE-AC02-83CH10093
[33]  Armenante, Piero.  "Flotation."  PowerPoint notes, http://cpe.njit.edu/dlnotes/CHE685/Cls05-1.pdf
[34]  DAF vendor website, http://www.envirowise.gov.uk/uk/Dissolved-air-flotation-DAF-application-and-design-.html
[35]  Diemar, M. et al., "Development of Extended Shelf-Life Microalgae Concentrate Diets Harvested by Centrifugation for Bivalve Molluscs  -  A Summary."  Aquaculture Research 31 (2000) 637-659.
[36]  Bough, W. et al., "Treatment of Food-Processing Wastes with Chitosan and Nutritional Evaluation of Coagulated By-Products."  Proceedings of the First International Conference on Chitin/Chitosan (1978) 218-230.
[37]  Homogenization article from Pro Scientific website, http://www.proscientific.com/Homogenizing.shtml
[38]  Hetherington, M. et al., "Emerging Technologies for Biosolids Management."  U.S. EPA document EPA 832-R-06-005 (2006), publicly available online at: http://www.docstoc.com/docs/585772/Emerging-Technologies-for-Wastewater-Biosolids-Management
[39]  Merkle, Peter.  "Recovery of Biodiesel Precursors from Heterotrophic Microalga Chlorella protothecoides."  Sandia whitepaper, http://www.sandia.gov/news/publications/white-papers/Separation%20of%20Heterotrophic%20Biomass%20Working%20Paper%20APR.pdf
[40]  From OriginOil website, http://www.originoil.com/technology/low-cost-oil-extraction.html
[41]  From homogenizer vendor website, http://www.microsludge.com/microsludge/in_detail.php
[42]  Personal communication with Microsludge (homogenizer vendor), September 2009.
[43]  Willett, Elizabeth.  "Mars Symbioscience."  Presented at the 2009 Algae Biomass Summit, San Diego, CA (October 2009).
[44]  Personal communication with Al Darzins (NREL), November 2009.
[45]  Phillips, S. et al., "Thermochemical Ethanol via Indirect Gasification and Mixed Alcohol Synthesis of Lignocellulosic Biomass" (2007), http://www.nrel.gov/docs/fy07osti/41168.pdf
[46]  Hassannia, Jeff.  "Algae Biofuels Economic Viability: A Project-Based Perspective."  Article posted online: http://www.biofuelreview.com/content/view/1897/1
[47]  Energy Information Administration price forecasts, http://www.eia.doe.gov/oiaf/forecasting.html
[48]  UT Extension, Farm Management Newsletter, Spring 2009.  Posted online at:  http://www.utextension.utk.edu/managecamp/spring09.pdf
[49]  IEA Greenhouse Gas R&D Programme, "Leading Options for the Capture of CO2 Emissions at Power Stations."  Report Number PH3/14, February 2000.
[50]  Incitec Pivot, Nitrogen Factsheet, March 2005.  Posted online at: http://www.incitecpivot.com.au/zone_files/FertFacts/NitrogenFS.pdf
[51]  Incitec Pivot, Phosphorus Factsheet, December 2003.  Posted online at: http://www.incitecpivot.com.au/zone_files/FertFacts/PhosphorusFS.pdf



Appendix

Capital and operating cost breakdowns for each growth case are summarized below.

                   Table 3: Open pond capital cost breakdown
 
                                     Base
                                  Aggressive
                                  Max growth
Ponds
                                                                    $15,865,290
                                                                     $4,957,903
                                                                     $2,754,391
Mixing (paddle wheels)
                                                                    $13,221,075
                                                                     $4,131,586
                                                                     $2,295,325
CO2 delivery + sumps
                                                                    $10,611,925
                                                                     $3,961,923
                                                                     $2,566,174
Primary harvesting (settling)
                                                                    $30,265,124
                                                                    $14,119,617
                                                                    $11,547,902
Secondary harvesting (DAF)
                                                                     $1,352,781
                                                                       $621,301
                                                                       $501,157
Cell rupturing
                                                                    $47,000,000
                                                                    $24,000,000
                                                                    $20,000,000
Clarifier tank
                                                                       $222,920
                                                                        $50,953
                                                                        $42,007
Water/Nutrient/Waste/Electrical Supply
                                                                    $19,038,348
                                                                     $5,949,484
                                                                     $3,305,269
Land Costs
                                                                    $21,238,326
                                                                     $6,636,977
                                                                     $3,687,209
Anaerobic Digestion
                                                                     $6,370,177
                                                                     $4,445,692
                                                                     $4,700,790
Gen-Set
                                                                    $13,171,225
                                                                     $6,399,650
                                                                     $5,267,786
Inoculum production system
                                                                    $23,797,935
                                                                     $7,436,855
                                                                     $4,131,586
General machinery
                                                                     $1,322,107
                                                                       $413,159
                                                                       $229,533
Initial Water Charge
                                                                             $0
                                                                             $0
                                                                             $0
TOTAL INSTALLED DEPRECIABLE CAPITAL
                                                                   $182,238,906
                                                                    $76,488,122
                                                                    $57,341,920
TOTAL INSTALLED NON-DEPRECIABLE CAPITAL
                                                                    $21,238,326
                                                                     $6,636,977
                                                                     $3,687,209
TOTAL INSTALLED CAPITAL
                                                                   $203,477,232
                                                                    $83,125,099
                                                                    $61,029,129



                  Table 4: Open pond operating cost breakdown
 
                                     Base
                                  Aggressive
                                  Max growth
Power
                                                                     $3,524,525
                                                                     $1,583,830
                                                                     $1,158,392
Power Flue Gas Supply
                                                                             $0
                                                                             $0
                                                                             $0
Nutrients (fertilizer) + wastewater nutrients
                                                                    $24,763,242
                                                                    $12,405,622
                                                                    $10,345,256
CO2
                                                                     $6,344,268
                                                                     $3,190,159
                                                                     $2,559,819
Flocculant
                                                                     $6,678,039
                                                                     $3,381,286
                                                                     $2,789,561
Waste Disposal
                                                                     $2,748,351
                                                                       $858,860
                                                                       $477,144
Water
                                                                             $0
                                                                             $0
                                                                             $0
Labor and Overhead
                                                                     $6,966,277
                                                                     $4,092,688
                                                                     $3,517,970
Maintenance, tax, ins
                                                                     $6,618,268
                                                                     $2,782,828
                                                                     $2,089,100
Total Algal Pond Op Costs ($/yr)
                                                                    $57,642,969
                                                                    $28,295,272
                                                                    $22,937,243



                      Table 5: PBR capital cost breakdown
 
                                     Base
                                  Aggressive
                                  Max growth
Photobioreactor equipment
                                                                   $642,527,429
                                                                   $200,715,037
                                                                   $111,558,210
Mixing (paddle wheels)
                                                                             $0
                                                                             $0
                                                                             $0
CO2 delivery + sumps
                                                                       $740,367
                                                                       $276,413
                                                                       $179,035
Primary harvesting (settling)
                                                                     $3,703,417
                                                                     $1,727,705
                                                                     $1,413,742
Secondary harvesting (DAF)
                                                                     $1,359,727
                                                                       $616,989
                                                                       $501,200
Cell rupturing
                                                                    $47,000,000
                                                                    $24,000,000
                                                                    $20,000,000
Clarifier tank
                                                                       $222,920
                                                                        $51,062
                                                                        $42,132
Water/Nutrient/Waste/Electrical Supply
                                                                     $4,474,680
                                                                     $2,005,971
                                                                     $1,564,527
Land Costs
                                                                    $11,383,743
                                                                     $3,557,420
                                                                     $1,976,344
Anaerobic Digestion
                                                                     $3,989,533
                                                                     $3,558,729
                                                                     $3,237,034
Gen-Set
                                                                    $13,350,757
                                                                     $6,488,982
                                                                     $5,342,252
Inoculum production system
                                                                    $11,898,967
                                                                     $3,718,428
                                                                     $2,065,793
General machinery
                                                                       $528,843
                                                                       $165,263
                                                                        $91,813
Initial Water Charge
                                                                             $0
                                                                             $0
                                                                             $0
TOTAL INSTALLED DEPRECIABLE CAPITAL
                                                                   $729,796,640
                                                                   $243,324,580
                                                                   $145,995,739
TOTAL INSTALLED NON- DEPRECIABLE CAPITAL
                                                                    $11,383,743
                                                                     $3,557,420
                                                                     $1,976,344
TOTAL INSTALLED CAPITAL
                                                                   $741,180,383
                                                                   $246,881,999
                                                                   $147,972,083



                     Table 6: PBR operating cost breakdown
 
                                     Base
                                  Aggressive
                                  Max growth
Power
                                                                     $2,088,003
                                                                     $1,102,799
                                                                       $937,955
Power Flue Gas Supply
                                                                             $0
                                                                             $0
                                                                             $0
Nutrients (fertilizer) + wastewater nutrients
                                                                    $24,740,772
                                                                    $12,392,695
                                                                    $10,335,425
CO2
                                                                     $6,268,379
                                                                     $3,152,476
                                                                     $2,633,460
Flocculant
                                                                     $6,762,571
                                                                     $3,381,286
                                                                     $2,789,561
Waste Disposal
                                                                       $183,223
                                                                        $57,257
                                                                        $31,810
Water
                                                                       $125,122
                                                                        $24,111
                                                                        $13,395
Labor and Overhead
                                                                     $6,295,773
                                                                     $3,613,756
                                                                     $3,065,162
Maintenance, tax, ins
                                                                    $14,257,046
                                                                     $8,896,888
                                                                     $5,337,682
Total PBR Op Costs ($/yr)
                                                                    $60,720,890
                                                                    $32,621,268
                                                                    $25,144,449

