Techno-Economic Analysis for the Production of Algal Biomass - Algae …algaebiomass.org/wp-content/gallery/2012-algae-bioma… · · 2017-01-05NREL is a national laboratory of

NRELisanationallaboratoryoftheU.S.DepartmentofEnergy,OfficeofEnergyEfficiencyandRenewableEnergy,operatedby theAllianceforSustainableEnergy,LLC.

Techno-Economic Analysis for the Production of Algal Biomass: Process, Design, and Cost Considerations for Future Commercial Algae Farms

Algae Biomass SummitOctober 24, 2016

Ryan Davis, Jennifer Markham, Christopher Kinchin, Nicholas Grundl

NATIONAL RENEWABLE ENERGY LABORATORY

Intro: 2016 Algal Biomass Design Report

2

• Projects goals to be achieved by 2022 and corresponding economics• Focused on open pond cultivation, given challenges in publicly

available cost/design details for PBRs (and widely varying PBR designs)

• PBR evaluation completed Sept. 2016• Primary value is the use of four independent but credible

sources for design and cost details for pond systems (key step of process)

• This approach shows significantly better agreement on what commercial pond systems should “actually” cost than typical statements made publicly

• Reduces uncertainty in underlying cost estimates, and highlights important economy of scale benefits

• Beyond base case, numerous sensitivity scenarios are considered• CO2 vs flue gas• Lined vs unlined ponds• Productivity vs cost• Alternative strains

• Includes consideration of sustainability metrics including GHG, fossil energy, and water profiles

http://www.nrel.gov/docs/fy16osti/64772.pdf


Background: Large public disparities on algae costs

3

Focusing only on open pond cultivation estimates from literature:• “Today’s” performance claims for algae:

• $280-$2,450/ton biomass, $2-$112/gal biofuels• 7-35 g/m2/day cultivation productivity (@ 330 day/yr uptime)

• “Future” goals:• $280-860/ton biomass, $2-$25/gal biofuels• 15-60 g/m2/day cultivation productivity (@ 330 day/yr uptime)

• Much of this variability may be attributed to differences in several key underlying assumptions –e.g. growth rates, pond system costs

• Given wide lack of agreement on these key metrics, analysis considers two approaches:1) “Top-down”: What does performance + cost “need to be” to hit a given biomass cost goal2) “Bottom-up”: Given a set of defendable assumptions, what is the resulting biomass cost

0 25 50 75 100-500

0

500

1000

1500

2000

2500

3000

-20

0

20

40

60

80

100

120

0 10 20 30 40 50 60 70Productivity(g/m2/day)

Biom

assC

ost($/U.S.Ton

)

AlgalO

ilCo

st($/Ga

llon)

Productivity(U.S.Ton/acre/yr @330days/yr)

AlgalOilCurrentAlgalOilFutureBiomassCurrentBiomassFuture


15 25 35 45 55 65

$0

$50,000

$100,000

$150,000

$200,000

$250,000

$300,000

10 15 20 25 30 35 40 45 50

Productivity(ton/acre/year)

Pond

SystemCap

italCosts($

/WettedAc

re)

Productivity(g/m2/day)

$1000/USDryTon

$700/USDryTon

$550/USDryTon

$430/USDryTon

$300/USDryTon

Approach: “Top-Down” Analysis

4

• Y and X axes – mutually independent variables• Contours = resulting minimum biomass selling price (MBSP)• MBSP reduces for higher productivity or lower pond cost• Likely lower limit for system costs ~$30k/acre (commercial nth

plant) • At this limit $430/ton is possible (@ 30 g/m2/day), but

challenging to reduce costs any further• Even if ponds were “free”, CO2/nutrient/other costs still add

up to $300-$400/ton lower boundary

“Today’s”costs(smallpondswithliner)

Commercialcostgoals(largerunlinedponds)


“Bottom-up Analysis” – Process Schematic

5

A100BIOMASSPRODUCTION

A200INOCULUMSYSTEM

A400MAKEUPWATERDELIVERY+ON-SITECIRCULATIONTO/FROM

DEWATERINGA300

CO2DELIVERY

A500DEWATERING

ALGAEBIOMASSCONVERSIONTO

BIOFUELS(notmodeledhere)

A600STORAGE

PFD-001JULY 2015

OVERALLPROCESS:ALGALPRODUCTIONPROCESS

ALGAE(0.05wt%solids)

AMMONIA

DIAMMONIUMPHOSPHATE

INOCULUMALGAE(0.05wt%solids)

ALGA

EPR

ODU

CT(20wt%

solids)

ALGA

EPR

ODU

CT

(20wt%

solids)

DIAMMONIUMPHOSPHATE

WAT

ER

RECYCLEWATER

AMMONIA

CO2

CO2

CO2(Fromoutsideoffacility) MAKEUPWATER

RECYCLEWATER


Biomass Production: Process Considerations

6

2022 goals:• Productivity: targeting 25 g/m2/day (AFDW annual avg)

• External reviewer agreement that >25 is or must be achievable by 2022 to demonstrate sufficient progress over today’s benchmarks

• Best performance published to date = 23 g/m2/day (+ 40% lipids) (Huntley/Cellana), 8-21 g/m2/day April-October (White/Sapphire)

• Composition: mid-harvest/high-carbohydrate Scenedesmus (HCSD), 27% FAME lipids• Scenedesmus selected given detailed compositional data, commercial relevance• Composition + productivity = ~3.9% PE to biomass (from full-spectrum irradiance), vs ~14% max

• Seasonal variability: 3:1 (max vs min seasonal growth)• Key challenge unique to algae – adds design constraints for downstream conversion facility• Most recent basis from PNNL BAT model = ~5:1 average for Gulf Coast• May be reduced either through strain engineering or seasonal strain rotation• Current ATP3 data ~3-4:1 average of all sites, <2.5:1 for Florida (“representative” Gulf Coast site)

• Evaporation: Based on prior harmonization modeling work (Gulf Coast average)

Metric Summer Fall Winter SpringAnnualAverage

BiomassProductivity(g/m2/dayAFDW) 35.0 24.9 11.7 28.5 25ProductivityVarianceversusSummerPeak NA(1:1) 1.4:1 3.0:1 1.2:1 NA

PondEvaporation(cm/day) 0.090 0.035 0.035 0.189 0.087Blowdown(MML/day) 7.3 2.8 2.7 12.4 6.3


Pond Design Scenarios

7

NREL solicited 4 separate inputs on 8 pond designs/costs:

• Key aspect of this work – address common conceptions that commercial algae pond costs are too scattered, uncertain to “really” establish with any certainty

• Ponds grouped into 100-acre “modules”, in turn constituting a 5,000 acre facility based on cultivation area (~7-9k total farm footprint)

• Continuous cultivation at fixed 0.5 g/L AFDW harvest density• Freshwater scenario, includes blowdown to control salt/inorganics

à All pond designs are based on unlined ponds with native clay soils• Plastic liners only used on berms or pond turns (2-25% of pond area)• Full pond liners considered as sensitivity (strongly influence total costs)

TypicalSumpLocation(variesbydesign)

1%Elevationchange

Weirevery2ndchannel

PaddlewheelStation

CirculationPump

CirculationPump

0.1%Slope

Paddlewheelraceway(typ) GAIgravityflow+pump Leidos serpentinepond

Source 2acre 10acre 50 acreLeidos (engineering firm) R R S

MicroBio (expert consultants) R R

Harris Group(engineeringfirm) RGAI(commercialdeveloper) G G

R=paddlewheelraceway

S=gravity-flowserpentine

G=GAIdesign(gravityracewaywithpump)


Pond Cost Estimates

8

a Additionaldatapoints(notincludedinfullTEA)addedtothisplottofurtherdemonstratecostalignmentbypondsize.

b Bealcostsbasedonextrapolatingfrompublishedcostsforfullylinedpondtoaminimally-lineddesign.IfafullylinedpondwereusedfortheBealcase,totalinstalledcostwouldbe$114,000/acre.

c GAIcasesincludeelectricalcostsunder“otherpondcosts”.

• Pond costs show reasonable agreement based on “small”, “medium”, or “large” size groupings

• More strongly a function of scale –highlights economy of scale advantages for building larger ponds >2-3 acres

• Largest cost drivers = paddlewheels + concrete (“other” category), piping, civil

• Economies of scale are possible for piping (individual feed/harvest lines), paddlewheels, electrical

• No notable scale advantages for civil


$576$649

$452$491

$545$475 $491

$419 $392

$0

$100

$200

$300

$400

$500

$600

$700

Algalbiomasss

ellin

gprice($/ton

AFD

W)

OSBL

Dewatering

Ponds+Inoculum

FixedOPEXCosts

OtherVariableOPEX

Nutrients

CO2

TEA Results: Base Case

9

Facilitysize 5,000acres(2,023ha)wettedcultivationareaCO2 demand 417,700ton/yrOn-linetime 7,920h/yr (330days/yr,i.e.,90%on-linefactor)Biomassproductionrate 0.19MMton/yr (AFDW)Biomassyield 37.5ton/acre/yr (84.1tonne/ha/yr AFDW)Totalinstalledequipmentcost $238MMTotalcapitalinvestment(TCI) $390MMTCIperannualtonbiomass $2,080MinimumBiomassSellingPrice $491/tonAFDWContributionfromcultivationsystem $278/tonContributionfromCO2 +nutrients $112/tonContributionfromremainder $101/ton

• MBSP results follow same trend as pond costs (largest driver on MBSP)

• Strong economy of scale advantages for pond design: $122/ton average premium for 2 vs 10 acre ponds

• $85/ton savings to move from 10 to 50 acre ponds, but becomes more speculative at such large scales

• For purposes of selecting a single MBSP value, average of the four 10-acre cases was used

TEADetails(averageof10-acrecases):


Sensitivity Analysis

10

Key drivers:• Productivity: dictates

economics, critical to achieve >25 g/m2/day

• Liners: adding full pond liners = >$120/ton MBSP penalty ($0.85/GGE MFSP impact on conversion costs)

• Farm size: 1,000 acres = $100/ton MBSP penalty ($70 labor cost + $30 capex)

$300

$400

$500

$600

$700

$800

$900

10 20 30 40 50

Algalbiomasss

ellin

gprice($/ton

)


• CO2 cost/sourcing• Price for purchased CO2 (flue gas CCS) $0-100/tonne = +$100/ton MBSP• Additional scenarios considered for flue gas: 15 km flue gas transport infeasible• Flue gas co-located with power plant: possible to reduce MBSP ~$45/ton, but logistical challenges for pond delivery


Summary and Concluding Remarks

11

• Algal biomass costs are tied strongly to productivity + cost of ponds, followed by CO2 + nutrients

• To achieve economically viable MBSP, critical to:a) Increase productivity and strain robustnessb) Maximize economy of scale benefits using >10-acre pondsc) Maximize farm size to >5,000 acresd) Demonstrate pond operability without pond liners

• “Bottom-up” modeling targets a 2022 base case MBSP of $491/ton AFDW

• Updated conversion models project 2022 targets near $5-6/GGE for this cost (CAP + HTL)

• Possible to reduce biomass costs to ~$430/ton, but achieving $3/GGE will require fundamental shift towards coproducts

• CAP pathway is well-suited for coproduct opportunities: non-destructive isolation of sugar/lipid/protein constituents

• Coproducts are a key focus of our TEA work moving forward


Questions?

Jennifer Markham Chris Kinchin Nick GrundlEric TanPhil PienkosLieve LaurensNick NagleBob McCormickJake KrugerMary Biddy

Dave Humbird, DWH ConsultingSue Jones, PNNLEd Frank, ANLJohn McGowen/Valerie Harmon, ATP3

Bill Crump, LeidosDavid Hazlebeck, GAIIan Woertz, Tryg Lundquist, John Benemann, MicroBio EngineeringJohn Lukas, Danielle Sexton, Harris GroupDesign report peer reviewers

12

Acknowledgements

FundingforthisworkwasprovidedbytheBioenergyTechnologiesOfficeintheDepartmentofEnergy'sOfficeofEnergyEfficiencyandRenewableEnergy.WethankDanielFishman,Christy

Sterner,andAlisonGossEngofthatprogramfortheirsupportandinput.

NREL, Sept, 2010, Pic #18229


Backup Slides

13


Facility layout – 5,000 acre farm

14

5,000acrefacilitybasedoncultivationarea(7-9kacretotalfootprint=~12sq.mi.)

Pondsdividedinto100-acreplots;eachplotincludescirculationpipelines+

primarydewatering

Terracedfacilitydesignovergradual1%slopewithcentraldewatering,

inoculum,conversionprocessingon-site

• 5,000 acre facility based on cultivation area (~7-9k acre total footprint)• Ponds divided into 100-acre plots; each plot includes circulation pipelines and primary

dewatering• Graded over gradual 1% continuous land slope = “terraced” design allowing for downhill

gravity circulation to central dewatering + downstream conversion (but requires uphill pumping of clarified water from central dewatering)

• Continuous cultivation/harvesting at a fixed 0.5 g/L AFDW harvest density from ponds• Freshwater base case avoids introducing subjectivity for proximity/cost of saline water sourcing

and brine disposal (consistent with prior harmonization models)• Blowdown still included to mitigate salt/inorganics <4,000 mg/L – taken off primary dewatering

recycle line (lowest algae concentration point = minimize biomass losses)


Inoculum system

15

• Inoculum system based on increasingly larger volume steps: PBR –covered lined ponds – open lined ponds

• Each step grows inoculum from 0.1 to 0.5 g/L based on the same seasonal productivities as main ponds

• Final stage inoculates production ponds at 0.1 g/L• Inoculum system sized to require inoculation once every 20 days during

peak summer season• Equivalent to 5% of facility ponds requiring re-inoculation each dayà Key nth plant assumption – robust strains withstanding frequent culture crashes

H2O+CO2+Nutrients

SeedTrain(fromlab)

Photobioreactor

CoveredPond LinedPond

H2OEvaporationLoss

ToCultivationPonds

H2O+CO2

+Nutrients

H2O+CO2

+Nutrients


Dewatering

16

• Primary dewatering occurs within the 100-acre modules to avoid circulating large volumes of water over entire facility

• Concentrates biomass from 0.5 g/L (0.05 wt% AFDW) to 10 g/L (1%) = 95% reduction in volume throughput

• Achieved using low-cost in-ground gravity settlers• Lowest-cost dewatering option, critical for economically processing tremendous harvested culture

volumes• Demonstrated at large scale at Cellana [Huntley et al] and WWT facilities in CA [MicroBio]• Highly strain-specific, but Scenedesmus is likely to settle well – assumed 4 hr settling time, 90% recovery

• Secondary dewatering = hollow fiber membranes• Demonstrated at large scale over sustained timeframe by GAI• Cost, performance based on inputs from GAI• Concentrates biomass to 130 g/L (13% AFDW) at >99% recovery

• Final dewatering = centrifugation• Established technology, standard for algal biomass concentration• Cost, performance based on inputs from engineering contractor (vendor quote)• Concentrates biomass to 200 g/L (20% AFDW) at 97% recovery

Fromponds0.5g/L0.05wt%

10g/L1.0wt%

Biomasstoupgrading

200g/L20.0wt%

Recycletoponds0.4g/L0.04wt%

130g/L13.0wt%

Recirculationtoponds0.1g/L0.01wt%

Blowdown0.1g/L0.01wt%

Settlersarelocatedin100-acrepondmodules

Membranesandcentrifugesarelocatedinthecentraldewateringfacility

Settlers Centrifuges

Membranes


Other design considerations

17

• CO2• Sourcing via off-site flue gas carbon capture• Priced at $45/tonne delivered to facility gate (supercritical)

• Consistent with average future CCS price projections in literature, DOE target of $40/tonne by 2020-2025

• Additional costs for on-site storage and delivery to ponds• Bulk flue gas scenarios considered in sensitivity analysis

• Nutrients• Set based on stoichiometric biomass composition at harvest, plus 20% excess

allowance• No recycle credits are taken on front-end model, to remain agnostic to back end

conversion pathway; any recycle credits should be assigned to reduce $/gal MFSP instead

• Water circulation• Maintains consistency with harmonization models to source freshwater via

nearby ground water resource, ~0.8 mile pipeline distance to facility gate• On-site circulation accomplished with aqueducts for “downhill” circulation to

central dewatering, pipelines for “uphill” return of clarified effluent back to pond modules

• Storage• Model also includes major storage tanks• Dewatered biomass storage assumed to incur 1% loss to degradation – should

be processed as quickly as possible through downstream conversion


Scale impacts for farm size

18

Significant economy of scale penalties <5,000 acre farm size• MBSP = $100/ton @ 1,000 acres,

$200/ton @ 500 acres• $70/ton labor, $30/ton capex

• MFSP = $2-3/GGE @ 1,000 acres, $5-6/GGE @ 500 acres

• Driven by scale more than biomass cost

• Also equipment operability concerns i.e. upgrading (min boundary = 1,000 bbl/day which is still very small)

• Central upgrading possible, but may lose ability to recycle nutrients (critical for LCA)

AlgalFarmSize(CultivationAcres) 5,000 1,000 500Algalbiomasstoconversion(AFDWton/day) 568 114 57Totalvolume flowtoconversion(MGD) 0.68 0.14 0.07CAPoilyieldtoupgrading(bbl/day) 1,060 212 107Biomasssellingprice(MBSP, $/tonAFDW) $491 $593 $691CAPpathwayMFSP($/GGE) $5.89 $8.04 $10.47HTLpathwayMFSP($/GGE)– perSueJones,PNNL $4.77 $7.74 $10.85NumberofCAPfacilitiestosupport5BGY 228 1,141 2,283NumberofHTLfacilitiestosupport5BGY 172 860 1,720

$5.89

$8.04

$10.47

$0

$2

$4

$6

$8

$10

$12

5,000 1,000 500

CAPMFSP($/G

GE)

FarmSize(Acres)

MFSPimpactduetobiomasscost

MFSPimpactduetoscale

BaseMFSPat5,000acrefarmsize


TEA Details: Algal Biomass Design Case

19


Sensitivities – Liners + Productivity

20

$667

$813

$537

$651 $644 $635 $617 $584 $552

$0

$100

$200

$300

$400

$500

$600

$700

$800

$900

Algalbiomasss

ellin

gprice($/ton

AFD

W)

OSBL

Dewatering

FullPondLiners

Ponds+Inoculum

FixedOPEXCosts

OtherVariableOPEX

Nutrients

CO2

$300

$400

$500

$600

$700

$800

$900

10 20 30 40 50

Algalbiomasss

ellin

gprice($/ton

AFD

W)


Full liner costs contribute almost the same amount as pond + inoculum costs – significant incentive to prioritize locations based on soil characteristics

• Biomass cost follows similar asymptotic curves as found in prior TEA – very strong cost sensitivity <25 g/m2/day

• Above 35 g/m2/day, other costs start dominating (CO2 + nutrients contribute >$100/ton in base case)


Additional Sensitivity Scenarios

21

• CO2: carbon capture vs bulk flue gas1) Bulk flue gas pipeline 15 km from source: requires more

power to transport the needed CO2 rate than the power generated to produce that amount of CO2

– Also translates to ~$49/tonne (vs $45/tonne target for purified CO2)

2) Flue gas co-location with algae facility (no significant off-site transport): $447/ton (~$45/ton MBSP savings) – But significant logistical/practicality questions regarding the use of multiple large ductwork pipelines routed around facility

• Alternative strains• Considered 9 total strain scenarios for tradeoffs in biomass

composition vs nutrient demands• Early-growth/high-protein biomass added up to $80/ton to

MBSP to sustain high N/P levels in biomass (*does not include N/P recycle considerations from downstream)

Fluegassource

60"

60"60"

48"

Centrif.Blower

IDFan

• Alternative dewatering scenarios1) Replace membranes with DAF

• Added substantial cost due to flocculant2) Replace membranes with EC

• Appears competitive with membranes, but requires large-scale demonstration

3) Replace membranes/centrifuge with filter press

• Potential to reduce MBSP by ~$15/ton but requires large-scale demonstration and may require a flocculant (would add to cost)


Financial Assumptions: Algal Biomass Design Case

22

• Model maintains the use of standard financial assumptions employed for other (biorefinery conversion) cases

• Exceptions:• Indirect capital cost factors: treated separately for cultivation, dewatering, and

OSBL operations based on best expectations for how such costs may factor into fixed capital investment (FCI)

• Labor: adjusted labor FTE categories and rates to more reasonably reflect algae farm (versus standard rates employed for a biorefinery)• Labor costs scale inversely with pond size (fewer total ponds required when each

pond is larger size = fewer ponds to service and maintain)

Plantlife 30yearsDiscountrate(IRR) 10%Generalplantdepreciation 200%decliningbalance(DB)Generalplantrecoveryperiod 7yearsFederaltaxrate 35%Financing 40%equityLoan terms 10-year loan at 8% APRConstructionperiod 3yearsFirst12months’expenditures 8%Next12months’expenditures 60%Last12months’expenditures 32%

Workingcapital 5%offixedcapitalinvestmentStart-uptime 6monthsRevenuesduringstart-up 50%Variablecostsincurredduringstart-up 75%Fixedcostsincurredduringstart-up 100%


Algal biomass design case: indirect capital cost allocations

23


Algal biomass design case: capital cost details

24


Algal biomass design case: labor details

25

Techno-Economic Analysis for the Production of Algal Biomass - Algae …algaebiomass.org/wp-content/gallery/2012-algae-bioma… · · 2017-01-05NREL is a national laboratory of

Documents