Techno-Economic Assessment of Nonfossil

Ammonia ProductionPer Tuna,a Christian Hulteberg,a and Serina Ahlgrenb

aDepartment of Chemical Engineering, Faculty of Engineering, Lund University,Lund, Sweden; per.tuna@chemeng.lth.se (for correspondence)bDepartment of Energy and Technology, Swedish University of Agricultural Sciences, Uppsala, Sweden

Published online 4 November 2013 in Wiley Online Library (wileyonlinelibrary.com). DOI 10.1002/ep.11886

The production of nitrogen fertilizers are almost exclu-sively based on fossil feedstocks such as natural gas andcoal. Nitrogen fertilizers are a necessity to maintain the highagricultural production that the world’s population currentlydemands. Ammonia produced from nonfossil-based feed-stocks would enable renewable production of ammonia.Renewable feedstocks are one thing, but perhaps even moreimportant in the future are the security of supply that decen-tralized production enables. In this study, the techno-economic evaluation of production of ammonia from vari-ous renewable feedstocks and for several plant sizes wasinvestigated. The feedstocks included in this study are grid-based electricity produced from wind power, biogas, andwoody biomass. The feedstocks differed in exergy, and tomake a fair comparison, the electric equivalence ratiosmethod was used. The results showed that the energy con-sumption for biogas and electricity is the same at 42 GJ/tonne ammonia. When using the electric equivalence com-parison for the same cases, the results are 26 and 42 GJ/tonne, respectively. Biomass-based production has an energyconsumption of 58 GJ/tonne and 31 GJ/tonne when usingthe electric equivalence comparison, which should be com-pared with the industrial average of 37 GJ (or 21 GJ electricequivalence) per tonne of ammonia. Monte Carlo simula-tions were used to vary the inputs to the process to evaluatethe production cost. The ammonia production cost rangedfrom $680 to 2300/tonne ammonia for the various casesstudied. VC 2013 American Institute of Chemical EngineersEnviron Prog, 33: 1290–1297, 2014

Keywords: ammonia production, nonfossil based, exergyevaluation, production cost

INTRODUCTION

Nitrogen fertilizers are used in agriculture to obtain higheryields of agricultural crops. Large-scale use of synthetic nitro-gen fertilizers began after the Second World War, and it isestimated that about one-third of the protein in humanity’sdiet depends on mineral nitrogen fertilizer [1]. The use ofnitrogen fertilizers is also predicted to increase in the futuredue to population growth, increased consumption of meat,and increased use of biofuels [2, 3].

Nitrogen gas accounts for 78% of the volume of ouratmosphere; however, conversion of nitrogen into a formthat is useful for agriculture costs energy. At present, the pro-

duction of nitrogen fertilizer accounts for 1.2% of global pri-mary energy demand [4]. All commercially producednitrogen fertilizers (e.g., urea and ammonium nitrate) useammonia as raw material. Anhydrous ammonia can also beinjected straight into soils, a method extensively used in theUnited States. Around 79% of the globally produced ammo-nia is used to produce different types of nitrogen fertilizers,3% is directly used as fertilizers, and 10% of the ammonia isused in other sectors. Ammonia is produced by the Haber-Bosch process according to the overall reaction (1).

N2 1 3H2 $ 2 NH3: (1)

Although nitrogen is supplied from ambient air, it is theproduction of the hydrogen that requires the majority of theenergy input. Production of the required hydrogen is mostcommonly based on reforming of natural gas; however, gasi-fication of coal and heavy oil also occurs. In the long run,this is not a sustainable solution for production, as fossilfuels are a finite energy source.

The energy requirement for ammonia production has dra-matically decreased over time from about 55 GJ/metric tonneof ammonia produced in the 1950s to 35 GJ/tonne in the1970s, whereas nowadays, the best plants using natural gasas feedstock need only 28 GJ/tonne [1]. However, there arelarge variations: in China, coal is commonly used, and theestimated average energy use is 59 GJ/metric tonne ofammonia [5]. According to the International Fertilizer Indus-try Association, about 67% of global ammonia production isbased on natural gas, 27% on coal while fuel oil, and naph-tha account for 5% [4]. As a number of old plants are still inoperation, the global average energy requirement was in2008 around 37 GJ/ tonne ammonia (ranging from 27 to 58GJ/tonne NH3) [6].

On a global scale, the production of nitrogen fertilizers iscalculated to represent about 1% of anthropogenic green-house gas (GHG) emissions [4]. On a product level, the pro-duction of nitrogen fertilizers can have a large impact, forexample, when calculating the carbon footprint of food andbiofuel. In a recent study [7], nitrogen fertilizers were foundto represent between 3 and 26% of the total GHG emissionsfrom wheat-based ethanol production. For rapeseed biodie-sel, the nitrogen fertilizer production represented up to 29%of the GHG emissions.

However, the hydrogen needed for ammonia productionby the Haber-Bosch synthesis can also be produced fromrenewable resources. This opens up possibilities for a moreVC 2013 American Institute of Chemical Engineers

Environmental Progress & Sustainable Energy (Vol.33, No.4) DOI 10.1002/ep1290 December 2014

sustainable production of food, feed, fibers, and fuels. Theconcept of renewable-based fertilizer production was alreadyconsidered as an option during the oil crisis in the 1970s and1980s, as a way of reducing the dependency on fossil oil; forexample, techno-economic studies were carried out onelectrolysis-based ammonia production [8, 9], and a plant forproducing ammonia from peat was erected in Finland [10]. Atpresent, ammonia production based on renewables is becom-ing interesting again, as a means to both reduce fossil fueldependency and to reduce GHG emissions. In Minnesota,USA, a plant is currently being commissioned that will produce1 tonne per day ammonia in a Haber-Bosch synthesis reactor;the hydrogen needed for the synthesis is derived from wind-powered electrolysis [11]. Furthermore, several life cycle assess-ment studies have been carried out in Sweden [12–14].

There are several parameters that are of interest to quantifywith respect to production of ammonia from nonfossil sour-ces. For example, the potential amount of ammonia can beproduced from different systems, the investment and produc-tion costs, the energy requirement, and GHG emissions. Someof these have been treated in the previous studies mentioned;however, there is a need for more research especially con-cerning the costs. Furthermore, water consumption andwastewater production have not previously been investigated.

The aim of this study was to perform a techno-economicstudy of a number of promising ammonia production sys-tems based on renewable energy. Different technologies andscales of production were investigated. The studied technol-ogies for hydrogen generation followed by subsequentammonia synthesis were:� Electrolysis of water using renewable electricity� Steam reforming of biogas from anaerobic digestion� Biomass gasification

The heat and mass balances as well as production costswere calculated for each scenario. Monte Carlo simulationshave been used for assessing the sensitivity of the produc-tion costs. The results are expected to be useful for furtherresearch and could also serve as basis for planning for futureproduction facilities.

MATERIALS AND METHODS

Modeling of the Technical SystemsThe different cases were evaluated by flow sheeting cal-

culations from raw material to product. The models for fuelsynthesis were performed in AspenTech’s Aspen Plus 7.3.2,and the gasifier was modeled as an energy/material balance.As the raw material differs, different processing equipmentsare required. Electrolysis requires almost no processingequipment upstream of the ammonia synthesis. Gasificationof woody biomass requires a gasifier, particle and contami-nants removal. Furthermore, biogas and gasification requiresreformers, water–gas shift reactors, CO2 capture equipment,and a methanation reactor to process the gas before theammonia synthesis; overview process flow diagrams for thetechnologies are depicted in Figures 1a–1c.

The electrolysis was modeled in Aspen Plus as a yieldreactor coupled with a calculator block to calculate the nec-essary electric power draw. Since pure hydrogen is producedin the electrolyzer, no downstream purification is necessaryonly the addition of nitrogen. A PSA was modeled as a sepa-rator block to provide enriched nitrogen (95% pure). Theexcess oxygen was reduced to water by hydrogen in a cata-lytic burner. The water in the hydrogen/nitrogen stream wasremoved prior to ammonia synthesis.

Figure 1. (a) Biogas-based ammonia production; (b) biomass gasification-based ammonia production; and (c) electrolyzer-based ammonia production from renewable electricity, for example, wind power.

Environmental Progress & Sustainable Energy (Vol.33, No.4) DOI 10.1002/ep December 2014 1291

A steam reformer (SR) is usually used before an autother-mal reformer (ATR) or secondary reformer when producinghydrogen for ammonia synthesis [15]. The ATR is either oxy-gen enriched or air blown to provide the nitrogen necessaryto reach a H2:N2 ratio of 3:1 prior to synthesis. The reformerswere modeled as Gibbs free energy reactors. The SR is exter-nally heated by burning part of the feed. The burner wasmodeled as a combustion reactor in Aspen Plus, setup toprovide the energy required in the SR. Temperature in theSR was set to 800�C and with a pressure drop of 1 bar. Thesame pressure drop was assumed for the ATR. The ATR wasfed with air through a compressor and with sufficient quan-tity to reach an outlet temperature of 1050�C. The air com-pressor was modeled as a multistage compressor with fourstages with intercooling to 70�C and an isentropic efficiencyof 0.72.

After reforming, the gas contains H2, CO, CO2, N2, andH2O. CO and CO2 are poisons for the ammonia catalyst, andCO contains energy that can be used to convert H2O to H2

in the water–gas shift reaction (2).

CO 1 H2 $ CO2 1 H2: (2)

The water–gas shift reaction is an equilibrium reactionthat is pushed toward H2 at low temperature. To achievemaximum H2 content in the gas, two water–gas shift reactorsare required, a high temperature and a low temperature. Thereactors were modeled in Aspen Plus as equilibrium reactorsoperating adiabatically with outlet temperatures of 450 and220�C, respectively [15].

After the second water–gas shift reactor, the gas containsonly trace amounts of CO, and most of the carbon in the gasis available as CO2. About 97% of the CO2 is removed in aPSA modeled in Aspen as a separator block. As both CO andCO2 are poisons for the ammonia catalyst, both componentsneed to be removed. Removal of carbon oxides is achievedby methanation (Reactions 3 and 4) in a methanation reactor[16].

3H2 1 CO $ CH4 1 H2O: (3)

4H2 1 CO2 $ CH4 1 2H2O: (4)

The methanation reactor was modeled as an adiabaticequilibrium reactor operating with an inlet temperature of200�C. Complete removal of carbon oxides are achievedwhile consuming about 6% of the H2 in the gas.

Compression of the H2/N2 mixture was performed in amultistage compressor with 70�C intercooling between eachstep and an isentropic efficiency of 0.72. Recirculation com-pression was performed in a single-stage compressor with anisentropic efficiency of 0.72.

Ammonia synthesis was modeled as three adiabatic reac-tors with recirculation. Pressure drop was set to 3 bar, andinlet temperature was set to 427�C for each reactor. Usableexcess heat, for example, for use in a district heating grid,was set to be available at levels down to 70�C [16].

Gasification was not modeled in Aspen Plus, instead thematerial and energy balance for the Carbona/Andritz gasifierwas used. The inlet stream in the gasifier model flow sheetwas the producer gas from the gasifier minus particulates.The producer gas composition is listed in Table 1.

A total of five different scenarios were evaluated: 1 and 3MW of electric power input for electrolysis, 5 and 10 MW ofbiogas, and 50 MW of biomass input to the gasifier. The sizeof the gasifier was chosen to give enough economies-of-scaleeffect, a reasonable biomass collection area and to fit the tech-nology with the highest technology-readiness level. For thesmaller scale of the studied systems, a steam cycle for internalelectricity production would be too costly. Therefore, electric-ity needed for compression and utilities, for the gasifier andbiogas routes, were assumed to be purchased externally.

Energy AssessmentWhen using mixed energy carriers, first law of thermody-

namics efficiency calculations are helpful but not necessarilyan appropriate indicator of the “best” system as it do nottake into account the exergy of the various energy sources.The electricity equivalents method is therefore used in thisstudy to represent the overall exergy of the system, and theyare calculated by using power generation efficiencies [18],using the best available technology to the authors’ knowl-edge. This allows for a comparison on an equal basis for allsystems. To exemplify, the energy input needed for produc-ing one tonne of ammonia would be calculated by determin-ing the amount of electricity equivalents that are fed to thesystem, in the form of biomass, biogas, and electricity, andthus yield a comparable number based on the exergycontent.

Economic AssessmentThe investment cost assessment has been performed by

using factor methods from various sources. The heat andmass balances determined from the Aspen simulations havebeen used to do detail dimensioning of the process equip-ment. To the total bare-module cost, 18% contingency and30% auxiliary have been added giving the overall investmentcost; the estimate is believed to be within 630%. Thereafter,the cost of production of the ammonia is determined. In

Table 1. Gasifier producer gas composition [17]

Component Composition (vol %)

CO 27.3H2 28.8CO2 16.9O2 0.0H2O 22.7CH4 3.8N2 0.0Ar 0.0C2H2 0.0C2H4 1.1C2H6 0.0C3H6 0.0C6H6 0.2C7H8 0.0H2S (vppm) 150NH3 (vppm) 2000Tars (mg/Nm3) 500

Table 2. Assumptions for financial modeling

Parameter Value Reference

Annual time-on-stream (h) 8000 –Economic plant life (years) 15 –Investment cost electrolyzer

(per kWh)$577 [19]

Electricity consumptionper kWh/(Nm3 H2/h)in electrolyzer

4.25 [19]

Contingency (%) 18 [19]Auxiliary (%) 30 [19]Vessel material Stainless steel –CEPCI 2011 585.7 –

Environmental Progress & Sustainable Energy (Vol.33, No.4) DOI 10.1002/ep1292 December 2014

Table 2, some key parameters for the financial modeling arelisted; for more information on detail-dimensioning, pleasesee Refs. 19 and 20.

The production cost is determined and analyzed usingMonte Carlo simulations. The simulations constitute 10,000cases and have been performed using a Beta-PERT distribu-tion function that has a density function [21]:

f ðxÞ5 xv21ð12xÞw21

Bðv;wÞ

" #0 � x � 1 ; 0 otherwise

where B(v,w) is the beta-function:

Bðv;wÞ �ð10

tv21ð12tÞw21dt;

and the distribution function:

FðxÞ5 Bxðv;wÞBðv;wÞ 0 � x � 1; 0 otherwise;

where Bx(v,w) is the incomplete beta-function:

Bðv;wÞ �ðx0

tv21ð12tÞw21dt

Typically, sampling from the beta-distribution requiredminimum and maximum values and two shape parameters(v,w) and a scale parameter k which is set to 4 giving themost likely values more influence than the extreme values.The mean l is calculated as:

l5ðxmin 1xmax 1k3xmost likelyÞ

k12

and used to calculate the v and w shape parameters:

v5ðl2xmin Þð23xmost likely2xmin 2xmax Þ

ðxmost likely2lÞðxmax 2xmin Þ;

w5vðxmax 2lÞðl2xmin Þ

:

A number of initial values were chosen, and these valueswere used as most likely in the Beta-PERT distribution. Foreach case examined, the minimum production cost, the max-imum production cost, the mean production cost, and theproduction cost distribution were determined. The maxi-mum, minimum, and most likely values are listed in Table 3.

RESULTS

Heat and Mass BalancesThe simulation results for energy and water consumption

for the different ammonia systems are summarized in Table 4.All systems require electric power for compression and aux-

iliaries. Biomass gasification requires more energy input thanthe other, and the reason is the costly pretreatment require-ment of biomass gasification and the oxygen production. Oxy-gen production and compression are the most energy intenseparts of the gasification system: 4 MW of electric power wasrequired for oxygen production and 3.5 MW for compression.

Biogas as an energy source is a promising alternative tothe natural gas used for conventional ammonia synthesis. Thesimulations results point to a lower electric power demandper tonne than the other two alternatives. The biogas systemuses only air for the reformers but still required some 1.0 MWof electric power for compression at the 10-MW scale.

When using electric power as energy source, roughly2000 metric tonnes can be produced by the 3-MW electricinput, as can be seen in Table 5, at an input of about 43 GJel

equiv/tonne. At the same time, some 0.46 MW of heat is avail-able for district heating. For the 10-MW electric input case,almost 6800 metric tonnes of ammonia can be produced peryear at an input of about 43 GJel equiv/tonne and making1.52 MW of district heat available. The 3-MW electric inputscales more or less linearly with the 10-MW input. There isno water knockout for electrolysis, and as a result, no waste-water is produced. The water consumption for electrolysisammonia is roughly 2 tonnes of water per tonne ammoniaproduced.

At 10 MW (5.8 MWel equiv) of biogas input, the systemrequires 1 MW of electric power which totals 6.8 MWel equiv. Theoverall production of ammonia in this case is 7480 tonnes peryear at 26 GJel equiv/tonne; this scale, 3.0 MW of heat, is availablefor district heating. The water consumption for the biogas sys-tem is higher than the other two systems at 2.8 tonnes water pertonne ammonia produced. Wastewater production from the bio-gas systems are roughly 2 tonnes per produced tonne of ammo-nia. The 5-MW biogas input system requires 0.5 MW of electric

Table 3. Input values for the Monte Carlo simulations, converted to 2011 prices using producer price index for chemical andallied products [22]

Cost Minimum Most likely Maximum Reference

Biomass ($/MWh) 33.9 67.8 101.7 [19]Electricity ($/MWh) 40.1 80.1 120.2 [19]Biogas ($/MWh) 50.8 101.7 152.5 [23, 24]Water ($/tonne) 1 2 3 [25]Wastewater ($/tonne) 10 20 30 Estimated from [25]Investment cost 230% Grass-root cost 130% [19, 20]Interest rate (%) 5 8 12 —

Table 4. Energy and water inputs for the studied cases

Case

Chemicalenergy

input (MW)

Electricinput(MW)

Waterconsumption

(tonne/yr)

3 MW electrolysis — 3 403010 MW electrolysis — 10 13,4005 (2.9 el equiv)

MW biogas5 0.5 10,500

10 (5.8 el equiv)MW biogas

10 1.0 20,900

50 (23 el equiv)MW biomass

50 7.8 60,400

Environmental Progress & Sustainable Energy (Vol.33, No.4) DOI 10.1002/ep December 2014 1293

power giving a total of 3.4 MWel equiv input and produces 3750tonnes of ammonia at 26 GJel equiv/tonne.

The last studied case is the 50-MW input biomass gasifica-tion plant. When comparing the production volume and theenergy requirements for the gasification route with those forbiogas and electrolysis, it is clear that the gasification requiresmore energy per produced tonne ammonia than the othertechnologies (58 GJ when compared with 42 GJ per tonneammonia). However, when comparing the energy consump-tion on an electricity-equivalent basis, the biomass gasifiershows quite low numbers: 31 GJ/tonne. The water consump-tion for the gasification plant was lower than that for the bio-gas plant at 2.1 tonne water per produced tonne ammonia.The 50-MW gasification plant produces 14.6 MW of heat thatcould be used for district heating. Wastewater production forthe gasification plant was the same as for biogas at 2 tonnesof wastewater per tonne of produced ammonia.

Economic EvaluationBased on the heat and mass balances presented above, a

detailed design of the process equipment may be performed.This has been done for all five cases and results in an overallgrass-root investment cost; the results from the analysis arereported in Table 6. With respect to the overall investmentcost, the biomass gasification plant is the one with the high-est overall investment cost with M$117. However, whenlooking at the relative investment cost per produced tonneper annum, it ranks as number 3 of the selected cases. Thehighest investment cost equipment (installed) is the syngascompressor (k$22,632), CO2 separation (k$14,066), and dryer(k$10,101). The biogas-based ammonia synthesis show thelowest relative investment costs with k$3.7 per producedtonne per annum in the 5-MW case and k$3.1 per producedtonne per annum in the 10-MW case. The overall investment

is dominated by the CO2 separation (21% and 20% in the 5-MW and 10-MW cases, respectively) and the syngas compres-sor (17% and 22% in the 5-MW and 10-MW cases, respec-tively). The investment cost of the electrolyzer systems aredominated by the electrolyzers and syngas compressors (23%and 23% in the 3-MW case and 26% and 22% in the 10-MWcase). This clearly shows that there is an economy-of-scaleeffect in the syngas compressor (and the rest of the processequipment for that matter), which is not mirrored in the elec-trolysis investment.

Given the production costs, there is quite a large span inthe production costs between the various technologies. Thelowest production cost is found for the biomass gasificationcase with a mean value of $970/tonne. In this case, the var-iance between the lowest and highest cases is $666/tonne,which is the lowest variance of the cases investigated. In thecase of biogas-based production, despite the relatively lowinvestment cost, the ammonia production cost becomes quitehigh due to the high biogas cost. The mean production costis $1671/tonne in the 5-MW case and $1594/tonne in the 10-MW case. This also shows that there is little influence of thecost-of-production with the increase in scale, which is due tothe high percentage of operating cost for the biogas cases(81% and 77% for 5-MW and 10-MW cases, respectively); thehigh operating cost also indicates that if feedstock produc-tion comes down in price, significantly lower costs may beachieved in ammonia synthesis. The same is true in the elec-trolysis case, where about 30–40% is fixed costs. However,the production cost is even higher in this case when com-pared with the biogas case with averages of $1725/tonneand $1640/tonne for 3-MW and 10-MW cases, respectively.The variance in the production cost over the 10,000 MonteCarlo simulations is also larger in the cases using electrolysiswhen compared with the biogas cases; the variances aresummarized in Figure 2 for all the cases.

Table 5. NH3 production, district heating, and wastewater for the studied cases

CaseNH3 production

(tonne/yr)District

heat (MW)Wastewater(tonne/yr)

Net energyinput (GJ)

Net energyinput (GJel equiv)

3 MW electrolysis 2030 0.46 — 42.6 42.610 MW electrolysis 6760 1.52 — 42.6 42.65 (2.9 el equiv) MW biogas 3730 1.5 7500 42.3 26.110 (5.8 el equiv) MW biogas 7480 3.0 14,900 42.2 26.150 (23 el equiv) MW biomass 28,700 14.6 58,700 58.0 31.1Natural gas industrial standard 25 16.1

Table 6. Results from the financial analysis

CaseInvestment

cost (k$)

Relativeinvestment

cost [k$/(tonneNH3/yr)]

Distributionoperatingcost/fixedcost (%)

Meanproduction

cost($/tonne)

Maximumproduction

cost($/tonne)

Minimumproduction

cost($/tonne)

3 MW electrolysis 10,182 5.0 60/40 1725 2392 107810 MW electrolysis 29,034 4.3 68/32 1640 2328 10155 (2.9 el equiv)

MW biogas13,860 3.7 77/23 1671 2297 1087

10 (5.8 el equiv)MW biogas

22,923 3.1 81/19 1594 2199 1028

50 (23 el equiv)MW biomass

117,311 4.1 50/50 970 1342 676

Investment cost is given as calculated grass-root cost.

Environmental Progress & Sustainable Energy (Vol.33, No.4) DOI 10.1002/ep1294 December 2014

DISCUSSION

As natural gas at present is the dominating feedstock, theammonia price is related to the natural gas price, which inturn is also related to the oil price [26]. Abram and Forster[27] suggest that feedstock makes up around 90% of ammo-nia production costs. However, it can also be argued thatvariations in grain prices also influence the nitrogen prices(as previously mentioned, 79% of globally produced ammo-nia is used for fertilizer production); the higher the cropvalue, the more willing the farmer is to pay for the fertilizer.According to the fertilizer producer Yara [28], �50% of thevariations in the urea price can be explained by grain pricefluctuations. The driver for grain prices is in turn dependenton many factors, for example, global income growth, bio-fuels mandates, high petroleum prices, droughts, and declin-ing storage reserves [29]. This makes any predictions on thefuture evolvement on ammonia and nitrogen fertilizer pricesvery challenging.

The US market price for ammonia was on average $976per metric tonne ammonia in January 2013 [30], equivalent to$1185 per metric tonne pure nitrogen. In 2008, the ammoniaprices peaked and then went down, but have steadily beenrising during the last years and are now at around the samelevel as 2008 [31]. This means that in the studied scenarios,only the gasification case has a mean production cost in thesame vicinity as natural gas-based ammonia. However, in thefuture, it can be expected that natural gas will be moreexpensive, which could make nitrogen based on renewablesmore competitive. Another aspect that should be mentionedis the fact that the costs reported above are free on boardGulf Coast and does not include transport. A more localizedproduction may tolerate a slightly higher cost of productiondue to lower transport costs as well as security of supplyaspects. However, the future competiveness of ammoniafrom nonfossil sources is also dependent on the develop-ment of food prices (as previously mentioned, higher grain

Figure 2. The production cost distribution for the investigated cases. The parabolic (blue) lines indicate the distribution ofresults from the Monte Carlo simulation, and the other (red) lines show the accumulative values. [Color figure can be viewedin the online issue, which is available at wileyonlinelibrary.com.]

Environmental Progress & Sustainable Energy (Vol.33, No.4) DOI 10.1002/ep December 2014 1295

prices gives larger willingness to pay for inputs inagriculture).

Looking at the energy input required for producing onetonne of ammonia, there are significant differences betweenthe technologies. When only comparing on energy inputbasis, the electrolysis paths and the biogas pathways haveabout the same energy input (42 GJ/tonne ammonia). How-ever, when changing the basis of comparison to electricityequivalents, better reflecting the exergy use, the numbers dif-fer significantly (42 GJel equiv to 26 GJel equiv per tonne ofammonia). The highest energy consumption is when usingbiomass gasification, most likely because a more difficultconversion of the solid feedstock, with 58 GJ/tonne or 31GJel equiv per tonne of ammonia. The energy consumption inthe investigated cases is higher than the industrial average(37 GJ or 21 GJel equiv per tonne of ammonia), both withrespect to the energy and exergy use and much higher thanthe current best available technology (28 GJ or 16 GJel equiv

per tonne of ammonia). This is most likely due to both thechosen method of production and the economy-of-scaleeffects.

There are significant differences in the above-investigatedtechnologies, and these are clearly reflected in the varyingdemand of fresh water, wastewater, energy input, and suita-ble scale of production. An individual assessment will haveto be performed from case to case when considering whichtechnology to choose based on the water supply, energy sit-uation, ability to get rid of wastewater, and so forth to endup with the best solution for each case. An argument for theproduction of ammonia and nitrogen fertilizers is the poten-tial to reduce fossil energy use and GHG emissions fromagriculture. In a study by Ahlgren et al. [13], the productionof nitrogen fertilizers from biomass via electrolysis and gasifi-cation was studied in a life cycle assessment. The resultsshowed that using this bio-based nitrogen in rape seed pro-duction could lower the emissions from cultivation with upto 46%. This could further have implications for the GHGbalance of biodiesel in which the cultivation emissions oftenconstitute a major part. As of today, there are no governmen-tal incentives for producing renewable fertilizers; however,there may be a spill-over effect from the production of bio-fuels. As the requirements for CO2 emission reduction areincreasing for qualifying as a biofuel, the ability to pay a pre-mium for a low-emitting fertilizer will increase.

In many countries, ammonia is not used directly as a fer-tilizer. Instead, ammonia is converted to ammonium nitratesor urea and applied to the fields as a solid granulate. Thismeans that the production costs calculated in this study isnot directly comparable with fertilizer prices in these regions.It is however interesting to note that the price trend forammonium nitrate and urea seems to follow the ammoniaprice. One region where such granulate is used as a fertilizeris Sweden, where prices in January 2013 was around $2000per metric tonne nitrogen, in the form of ammonium nitrate[32] (to be compared with $1185 for ammonia). As the granu-late price is higher than the ammonia price, an interestingcontinuation of this work would be to study the cost of con-version of ammonia to ammonium nitrate or urea. Anotherinteresting continuation would be to look into the infrastruc-ture in Sweden, to see if it would be possible to introduceanhydrous ammonia or ammonia dissolved in water (ammo-nia hydroxide). This would save costs for further conversionsof the ammonia and also save emissions of nitrous oxidesthat are formed during ammonium nitrate production. InDenmark, it is more common to use anhydrous ammonia ornitrogen solutions, approximately representing 3 and 10% ofthe total nitrogen fertilizer use in 2010 [33].

To give a reference point, around 170,000 tonnes offossil-based nitrogen in straight and compound fertilizers areused every year in Swedish agriculture, and the arable land

is 2.6 Mha [34]. The average application of nitrogen is 65 kgN/ha; however, the amount of nitrogen varies greatlybetween fields and crops. During the cropping seasonNovember 2010, the average application to winter wheat was149 kg N/ha. In the studied scenarios, between 2030 and28,700 tonne ammonia was assumed to be produced peryear. This is enough nitrogen to replace some 1–17% of theammonia used in Sweden or to fertilize between 11,000 and158,600 hectares of winter wheat depending on choice oftechnology. When assessing the risk of constructing a plantfor ammonia production, however, the relatively high abso-lute investment cost in the gasification case may be prohibit-ing for investment. In that case, the more modularelectrolyzer-based production comes across as more attrac-tive, despite the higher production cost. However, this is aconsideration that has to be done on a case-to-case basis,and an individual risk assessment should be the basis of theinvestment case.

CONCLUSIONS

Based on the investigation presented above, it can beconcluded that production of ammonia from nonfossil sour-ces is possible but not competitive with fossil-based produc-tion with respect to cost, perhaps with the exemption ofbiomass gasification. This can be explained by economy-of-scale effects as well as the lower feedstock cost. There arealso significant differences in the exergy use for producingone tonne of ammonia depending on what method is cho-sen, with biogas-based production showing the lowest valuesand electrolysis the highest; however, both values are higherthan the current industrial average. There is however otherbenefits to nonfossil-based production of ammonia, such assecurity of supply and lower transportation costs. Whichtechnology to choose of the three investigated cases has tobe decided on a case-to-case basis, weighting risk andreward for each case given the local conditions.

LITERATURE CITED

1. Smil, V. (2004). Enriching the earth: Fritz Haber, CarlBosch, and the transformation of world food production,MIT Press, Sabon, United States.

2. Erisman, J.W., Sutton, M. A., Galloway, J., Climot, Z.,Winningwater, W. F. (2008). How a century of ammoniasynthesis changed the world, Nature Geoscience, 1, 636–639.

3. Smeets, E., & Faaij, A. (2005). Future demand for fertil-izers from bioenergy crop production. Paper presented atthe 73rd IFA Annual Conference, Kuala Lumpur,Malaysia.

4. IFA, I.F.I.A. (2009). Fertilizers and climate change. Enhanc-ing agricultural productivity and reducing emissions,http://www.fertilizer.org/ifacontent/download/26110/374422/version/1/file/2009_tech_energy_efficiency.pdf.

5. Kahrl, F., Li, Y., Su, Y., Tennigkeit, T., Wilks, A., Xu, J.(2010). Greenhouse gas emissions from nitrogen fertilizeruse in China, Environmental Science & Policy, 13, 688–694.

6. IFA, I.F.I.A. (2009). Energy efficiency and CO2 emissionsin ammonia production 2008–2009 Summary Report.http://www.fertilizer.org/ifacontent/download/22830/327295/version/4/file/2009_climate_change_brief.pdf.

7. B€orjesson, P., & Tufvesson, L.M. (2011). Agriculturalcrop-based biofuels—Resource efficiency and environ-mental performance including direct land use changes,Journal of Cleaner Production, 19, 108–120.

8. Dubey, M. (1978). Technical and economic feasibility ofmaking fertilizer from wind energy, water, and air. In:Sun, Mankind’s future source of energy. New Delhi,

Environmental Progress & Sustainable Energy (Vol.33, No.4) DOI 10.1002/ep1296 December 2014

India: International Solar Energy Society Congress. Editor:Volume 3.: Winter, F. de;Cox, M.

9. Grundt, T., & Christiansen, K. (1982). Hydrogen by waterelectrolysis as basis for small scale ammonia production.A comparison with hydrocarbon based technologies,International Journal of Hydrogen Energy, 7, 247–257.

10. Koljonen, J., & Sodervall, T. (1991). Synthesis gas frompeat; A Finnish peat ammonia venture, Energy Sources,13, 95–103.

11. Dispatch, B. (2012). Ag and Energy Expo explores pro-ducing fertilizer from wind. Available from: http://brainerddispatch.com/news/2012-02-26/ag-energy-expo-explores-producing-fertilizer-wind, [Access date: 2013-02-13].

12. Ahlgren, S., Bernesson, S., Nordberg, A., Hansson, P. A.(2010). Nitrogen fertiliser production based on biogas—Energy input, environmental impact and land use, Biore-source Technology, 101, 7181–7184.

13. Ahlgren, S., Baky, A., Bernesson, S., Nordberg, A.,Hansson, P. A. (2012). Consequential life cycle assess-ment of nitrogen fertilisers based on biomass—A Swedishperspective, Climate Change, 2, 80–101.

14. Ahlgren, S., Baky A. Bernesson, S. Nordberg A. Hansson,P. A., (2008). Ammonium nitrate fertiliser productionbased on biomass—Environmental effects from a lifecycle perspective, Bioresource Technology, 99, 8034–8041.

15. Satterfield, C.N. (1996). Heterogeneous catalysis in indus-trial practice (2nd Edition, p. 554), Malabar: Krieger Pub-lishing Company.

16. Jennings, J.R. (1991). Catalytic ammonia synthesis: Funda-mentals and practice, Plenum Press, New York, UnitedStates.

17. Lundgren, J., Ekbom, T., Hulteberg, P. C., Larsson, M.,Grid, C. E., Nilsson, L., Tuna, P. (2013). Methanol produc-tion from steel-work off-gases and biomass based synthe-sis gas. Applied Energy, 112, 431–439.

18. Larson, E.D., Consonni, S., Katofsky, R. E., Iisa, K., Fredrick Jr.,W. J. (2006), A cost-benefit assessment of gasification-basedbiorefining in the kraft pulp and paper industry, Princeton:Princeton Environmental Institute, p. 152.

19. Hulteberg, P.C., & Karlsson, H.T. (2009). A study of com-bined biomass gasification and electrolysis for hydrogenproduction, International Journal of Hydrogen Energy,34, 772–782.

20. Hulteberg, P.C. (2007). Hydrogen, an energy carrier ofthe future, Lund: Chemical Engineering, Lund University.

21. Vose, D. (2008). Risk analysis: A quantitative guide,Wiley, West Chichester, West Sussex, England.

22. Bureau of Labor Statistics (2013). Producer Price Index -PPI. [cited March 15, 2013]. Available from: http://data.-bls.gov/, [Access date: 2013-03-15].

23. Urban, W., Lohmann, H., & Girod, K. (2009). Technolo-gien und Kosten der Biogasaufbereitung und Einspeisungin das Erdgasnetz, Oberhausen: Fraunhofer-Institut f€urUmwelt-, Sicherheits- und Energietechnik UMSICHT.

24. Berglund, P., Bohman, M., Svensson, M., Benjaminsson, J.(2012). Teknisk och ekonomisk utv€ardering av lantbruks-baserad fordonsgasproduktion, Malm€o: Swedich GasTechnical Centre.

25. Seider, W.D., Seader, J.D., & Lewin, D.R. (1999). Processdesign principles: Synthesis, analysis, and evaluation,Wiley, New York, United States.

26. Pettersen, I., Hval, J. N., Vasaasen, A., Alnes, P. K. (2010).Globalt marked med nasjonale særpreg—Utredning omkonkurransen i de nordiske mineralgj�dselmarkeder,ISBN: 978-82-7077-765-5, ISSN: 0805-7028, Oslo, Norway,Publisher: Norsk Institutt for landbruks€okonomisk for-skning (NILF).

27. Abram, A., & Forster, D.L. (2005). A primer on ammonia,nitrogen fertilizers, and natural gas markets, Departmentof AED Economics, The Ohio State University, UnitedStates.

28. Yara (2012). Yara fertilizer industry handbook. p. 91.29. Wright, B.D. (2012). International grain reserves and

other instruments to address volatility in grain markets.World Bank Research Observer, 27, 222–260.

30. U.S. Department of Agriculture (2013). Agricultural mar-keting service. Market News. Available from: http://www.ams.usda.gov/mnreports/gx_gr210.txt, [Accessdate: 2013-02-13].

31. Fertecon (2013). Ammonia report, weekly review of theammonia market [cited January 3, 2013]. Available from:http://fertecon.agra-net.com/home/ammonia, Accessdate: 2013-01-03.

32. ATL (2013). Fertilizer prices 2013-01-14. [cited February12, 2013]. Available from: http://www.atl.nu/noterin-gar, [Access date: 2013-02-12].

33. International Fertilizer Industry Association Statistics Data-base (2013). Available from: http://www.fertilizer.org/HomePage/STATISTICS, [Access date: 2013-02-12].

34. Wahlstedt, G. (2012). Sales of fertilisers for agriculturaland horticultural purposes in 2010/11. Final statistics,Statistics Sweden, Stockholm, Sweden.

Environmental Progress & Sustainable Energy (Vol.33, No.4) DOI 10.1002/ep December 2014 1297