Renewable Energy 157 (2020) 404e414

Contents lists available at ScienceDirect

Renewable Energy
journal homepage: www.elsevier.com/locate/renene

Technical-economic analysis for a green ammonia production plant in

Chile and its subsequent transport to Japan
C. Fúnez Guerra a, *, L. Reyes-Bozo b, **, E. Vyhmeister c, M. Jae

n Caparro

s d, f,

Luis Salazar e, C. Clemente-Jul f
Jose
a
National Hydrogen Center, Prolongacio
n Fernando el Santo Street, 13500, Puertollano, Spain
b
noma de Chile, Av. Pedro de Valdivia 425, Santiago, Chile
Facultad de Ingeniería, Universidad Auto
c
Insight Research Centre for Data Analytics, University College Cork, Cork, Ireland
d
Department of Innovation and New Energies, Enaga
s e Paseo de los Olmos, 19, 28005, Madrid, Spain
e
Departamento de Ingeniería Química, Universidad de Santiago de Chile, Casilla, 10233, Santiago, Chile
f
Department of Energy and Fuels Systems, School of Mining and Energy Engineering, Technical University of Madrid (UPM), Rios Rosas Street 21, Madrid,
Spain

a r t i c l e i n f o a b s t r a c t

Article history: Green ammonia can be produced using fossil fuels or any renewable energy source combined with heat
Received 2 August 2019 or electricity. Chile has one of the highest rates of solar irradiation and also the environmental conditions
Received in revised form that support the development of solar industry. Other renewable sources of energy are the wind and the
30 April 2020
hydraulic dams. These energies could be used to generate ammonia by a standard method such as the
Accepted 9 May 2020
Available online 13 May 2020
Haber-Bosch process, which transforms hydrogen and nitrogen using high temperatures and a catalyst.
The ammonia has several desirable characteristics that suggest its use as a way to store hydrogen. Firstly,
it can be liquified under mild conditions. Secondly, the ammonia has a high weight fraction of hydrogen.
Keywords:
Green ammonia
Thus, it is deemed necessary to study technical and economic variables that may support the use of green
Haber-bosch process ammonia as an energy carrier for hydrogen. The aim of this work is to develop a technical-economic
Ammonia transport analysis about the production of ammonia using hydrogen by means of electrolysis (carried out with
Net present value solar, wind, and hydraulic renewable energies). The aforementioned green ammonia would be produced
Pay-back in Chile and it would be necessary to transport it to Japan. Sensitivity analysis of main parameters (plant
Viability operating hours, size of the ammonia synthesis plant - related to the capacity of the electrolysis plant size
in MW-, electricity price, electrolyzer cost, Haber-Bosch cycle cost, and ammonia sales price) was per-
formed to report what factors are primordial at the moment of carried out techno-economic analyses. An
optimization process that minimize the NPV was run in order to settle the most convenient size of the
electrolyser stack. A net present value (NPV) of base case is V77,414,525 and 7.62 years of pay-back
period were calculated for this green ammonia production plant, which considered hydrogen produc-
tion via electrolysis, Haber-Bosch processes, and trade values of different operational units. The sensi-
tivity analysis has determined that the main variables affecting the NPV are the size of the ammonia
synthesis plant and the electricity price. Furthermore, by considering triangular probability distribution
of specific variables, it was observed that with a 95% of confidence, the NPV would be positive with a
76.1% of occurrence. The optimization process defined that a stack of 164.21 MW would be the most
convenient electrolyser stack dimension, which was estimated by considering all different effects over
the operational expenditure (OPEX), capital expenditure (CAPEX), and investments risk. Therefore, the
hydrogen production in Chile and its transportation to Japan using ammonia as an energy carrier would
be a technically viable and profitable solution, which also has environmental benefits.
© 2020 Elsevier Ltd. All rights reserved.

1. Introduction
* Corresponding author.
** Corresponding author. About a century ago, William Crookes warned that the world
E-mail addresses: carlos.funez@cnh2.es (C. Fúnez Guerra), lorenzo.reyes@ could suffer from starvation [1]. This prediction was based on the
uautonoma.cl (L. Reyes-Bozo).

https://doi.org/10.1016/j.renene.2020.05.041
0960-1481/© 2020 Elsevier Ltd. All rights reserved.
 C. Fúnez Guerra et al. / Renewable Energy 157 (2020) 404e414
Nomenclature
CAPEX Capital expenditure V
HeB Haber-Bosch
I Time period Years
Ir Inflation rate %
IRR Internal Rate of Return %
kn Discount rate %
NPV Net Present Value V
OPEX Operational expenditure V
PB Pay-Back period Years
PV Present Value V
R Risk rate %
T Base discount rate %
t Discount rate %
WACC Weighted average cost of capital %
rate which the population was growing and the relatively small
supply of fertilizers available. At that time, Chile had the advantage
of possessing a natural source of fertilizers (in the form of Sodium
Nitrate; saltpeter), which was internationally demanded: approx-
imately, two-thirds of the world requirements were supplied by
Chile [1].
One of the main elements in fertilizers and also a primary
nutrient for the survival of all living organisms is Nitrogen [2]. Even
though Nitrogen is highly abundant in our atmosphere (as dini-
trogen gas, N2), it is almost inaccessible given the highly stable
triple bound formed in this molecule [2]. The potential depletion of
natural resources and the imminent starvation placed on alert to
synthetically fix Nitrogen in order to supply the growing demand of
fertilizers.
Fritz Haber devised a method to fix the nitrogen by combining
atmospheric nitrogen and hydrogen to produce ammonia at rela-
tively high pressures and temperatures (200 Atm at 500
Ce600
C)
[1]. These operational conditions have been extensively investi-
gated and have been reported within the range of 50e200 bar and
650 Ke750 K, with the aid of different catalysts [3]. The process is
nowadays known as the Haber - Bosch (HeB) cycle [4]. The
industrialization of synthetic ammonia has different positive and
negative repercussions (e.g. inexpensive fertilizers, a supply line of
nitrates for explosives, the rise of the population from 1.6 billion to
6.7 billion, the alteration of the global nitrogen cycle or the Chilean
saltpeter industry collapse) [2,5]. The HeB cycle has become so
important at an industrial level that 90% of the produced ammonia
is made by this process [6].
Nowadays, the ammonia can be considered as a multiuse
compound since it is used as a fertilizer, a chemical feedstock, a
clean-burning fuel for transportation, a refrigerant fluid, and a
power generator, among others uses [6,7]. Recently, the ammonia
has also been regarded as a hydrogen storage medium and there-
fore as an energetic vector [7,8]. One of the advantages of consid-
ering ammonia as a fuel or as a hydrogen storage is that the
infrastructure requirements are similar to those used by Liquified
Petroleum Gas (LPG); by taking into account its volumetric energy,
the ammonia could transport almost 50% more energy in current
LPG pipelines [9]. Liquids with a high energy density are indis-
pensable and the current trends of using fuels with non-fossil or-
igins make ammonia an interesting candidate; especially true if its
production involves renewable energies [7].
Industrially, the HeB process requires purified feeds of N2 and
H2. To obtain the Hydrogen, it is necessary to carry out a combi-
nation of processes (catalyzed steam reforming, catalyzed water
405
gas shift reactions to produce more H2 and separation/purification
processes) [6]. Nitrogen is usually obtained by Cryogenic Air Sep-
aration (CAS), which is defined as one of the most efficient,
developed, and economic sound technologies in Nitrogen produc-
tion [9]. After the HeB reaction, the unreacted Hydrogen and Ni-
trogen are recirculated in order to improve the relatively low
conversions that have been obtained (global conversions in the
order of 97%; conversions in the reactor in the order of 15%) [10].
Hydrogen is annually produced at a scale of 60 MT. Approxi-
mately, 95% of the processes involve the use of fossil fuels (gas, oil,
and coal); the rest is primarily produced by water electrolysis [6]. If
it is considered that almost half of the produced hydrogen is used in
the HeB process, then the burden involved in the process is
considerable. The estimates depict that the CO2 equivalents in the
production of ammonia are around 400 MT/year [6]; a modern
ammonia plant can consume 22.1 GJ/T of NH3 as a feedstock with
emissions of 1.2 T of CO2/T of NH3 and 7.2e9.0 GJ/T of NH3 as a fuel
with emissions of 0.4e0.5 T CO2/T of NH3 [11]. Even though
approximately 36% of the CO2 quantity is retained and partially
used as a by-product [11], the production volumes involved in the
HeB process are so high that the emissions are substantial,
regardless of the small improvement in the HeB process efficiency.
Given the tendencies and the unsustainable industrial practices
that have been carried out for decades, leading to a significant
degradation of natural resources and to pollution, the society is
currently taking actions in order to preserve the resources and to
ensure them for the future [12].
As for the HeB process, it is considered to be highly mature with
small room for improvement; it is mentioned that the process is
close to its thermodynamic optimum [11]. On the other hand, other
methods based on electrochemical synthesis (e.g. SSAS, Solid State
Ammonia Synthesis) are expected to reduce their production costs
[9] by reducing operating pressures or modifying the used feed.
Alternatively, an improvement of the process based on the envi-
ronmental aspect (de-carbonization) can be achieved by modifying
the energy sources, the feedstock sources, and/or by improving/
incorporating a CO2 sequestration process [13]. One important
aspect to take into account is that the use of renewable sources is,
by its own nature, a decentralized resource. This would imply that
the demand, which is decentralized as well, could be better covered
by a smaller scale production (islanded HeB systems) [13,14].
Given the reduction costs involved in renewable electricity,
especially in locations with considerable wind and solar resources,
water-electrolysis-based production of NH3 could compete with
steam reforming e based production [6].
Given the possibilities of modifying the current ammonia pro-
duction process by incorporating ambient sound alternatives, some
considerable works involving techno-economic evaluations (based
on sustainability) have been performed [9,14e17]. In fact, as
mentioned in the works of Wang et al. [6] and Allman et al. [14],
different projects, at different stages, produce ammonia (as a fer-
tilizer, a fuel or a chemical product), based on renewable energies
(e.g.: a small-scale solar ammonia facility in Iowa for fertilizer and
fuel; a wind-powered electrolysis operational unit in pilot-scale
ammonia plants in the university of Minnesota; an ammonia fa-
cility in Pilbara, Australia, where H2 is generated by water elec-
trolysis from seawater desalination; and Port Lincoln, Australia,
where H2 would be produced from water electrolysis using
renewable electricity).
Chile, one of the eleven most important suppliers of nitrogen, is
currently performing both: imports and exports nitrogen in
different forms. The saline deposits and brines in Chile are
considerably important as non-metallic minerals. They include
saltpeter, potassium salts, and lithium salts among others [18]. In
2010, the salts of some of the most important nitrogen-based-salts
406 C. Fúnez Guerra et al. / Renewable Energy 157 (2020) 404e414
were Potassium nitrate (MU$ 194.3), Ammonium Nitrate (MU$
63.9) and saltpeter (MU$ 57.1). In the same year, Chile imported
ammonia, anhydrous or in solution, for a total amount of MU$ 128.5
[19]. Given that Chile is endowed with one of the most intense
irradiation of solar energy (irradiation indexes are in the order of
5e12 kWh/m2 from winter to summer [20]), there is a considerable
relevant and constantly growing market (nationally and interna-
tionally), and the considerations of the ammonia as an energetic
vector (additionally to its other uses) [21], the economic opportu-
nity for Chile to install a renewable-energy-based HeB industry to
produce green ammonia could be considerably important. In fact,
some initial works have already been performed regarding imple-
mentation of green ammonia production in Chile [22,23] but there
is still a broad gap for analysis of the possible economic effects
involved. The present work focuses on a technical and economic
evaluation of the installation of a HeB production plant based on
hydrogen obtained by electrolysis using renewable energies (i.e.,
solar, wind and hydraulic) in Chile. Current renewable energy
sources are mainly focused in the use of batteries as energy storage
systems. Batteries are efficient but expensive and therefore the use
of ammonia as high-density low-cost storage system has shown
before to be an affordable methodology for energy storage and
transport [21]. Based on previous considerations, a new economic
niche could be foreseen for Chile, therefore techno-economic an-
alyses based on its most abundant renewable sources could drive
the application to be a reality. The green ammonia would be
exported to Japan as an Asian geographical strategic point. USA is
currently the biggest ammonia importer, but the expected eco-
nomic growth in Asia will make its market considerably attractive
in the future [9].
2. Methods
2.1. Description of the green ammonia production plant
The production of green ammonia considered in this work fo-
cuses on the use of hydrogen obtained from renewable sources
(using the produced energy in electrolytic processes) and nitrogen
obtained from a renewable energy supplied separation-purification
a process in the HeB process. The electrolytic process (polymer
electrolysis system) will be supplied with solar energy. Other
renewable sources could be considered in the analysis, neverthe-
less we focused on solar energy given its abundance in Chile. The
produced ammonia would be storage as liquid (18 bar at ambient
temperature) and then transported by ship to Japan.
The main processes/operational units considered in the
ammonia production plant are based on current technologies that
were quoted from industrial partners. Fig. 1 shows a layout on the
most important processes involved, main energy requirements,
main raw materials required, and the corresponding productions.
The description of these processes are as follows:
 Polymeric electrolyzer: The electrolyzer (SIEMENS brand) is
dimensioned for a total consumption of 160 MW of electricity
(https://new.siemens.com/global/en/products/energy/
renewable-energy/hydrogen-solutions.html). Consumptions
comes from the stack (150 MW), accessories, control, and other
components considered in the total plant balance (10 MW).
Other general components used through the layouts include the
compressors (PDC machines brand), tanks (Calvera and Wor-
thington brand). The electrolyzer consumes 15 L of untreated
water per kg of produced hydrogen. The process yield is in the
order of 2,687 kg of hydrogen per hour and 21,582 kg of oxygen
per hour (both at 35 bar of pressure and purities of 99,999% and
of 99.95%, respectively). the operational unit will produce a
surplus of thermal energy (60 MWh). This surplus will leave the
operational unit as water at a temperature of 65
C (considering
a feeding temperature of 50
C). Although the oxygen and the
remaining heat could be considered as process by-products,
they would not be considered for the economic evaluation
(i.e., the use of these by-products could improve the economic
results).
 Haber-Bosch: The hydrogen generated by the electrolysis pro-
cess would be combined with nitrogen obtained by a cryogenic
system. The generated ammonia would be accumulated in its
liquid state by conditioning it at pressures higher than its vapor
pressure at room temperature (8.8 bar). The HeB process
(Thyssenkrupp brand) (https://www.thyssenkrupp-industrial-
solutions.com/en/products-and-services/fertilizer-plants/
ammonia-plants-by-uhde) would require 5 MWh of electricity
(principally for the compression and air separation). As feed
streams, the unit would use 2,687 kg of hydrogen per hour (at
35 bar with 99,999% purity) in order to produce 15,000 kg of
liquid ammonia per hour (at 18 bar with 99.95% purity).
 Liquid ammonia storage: The ammonia generated in the HeB
process would be stored in a stainless-steel tank at the gener-
ation pressure (18 bar) (Lapesa brand) (http://www.lapesa.es/
sites/default/files/documentos/glpe.pdf). The tank with a ca-
pacity of 30,000 tons, allows a storage capacity of approximately
2,000 h of operation at full global system productivity.
Finally, the stored liquid ammonia would be shipped from Chile
to Japan (30,000 tons per ship).
2.2. Costs and sensitivity analysis
The economic analysis is based on the capital costs (CAPEX).
Operational costs (OPEX), contingencies, and other cost compo-
nents are obtained from the estimated CAPEX. CAPEX evaluation
followed suppliers’ information plus considerations of bare mod-
ule/Lang factor project evaluations methodologies Capital costs
include hydrogen production bare module, nitrogen facility,
Harber-Bosch ammonia facility, ammonia storage tanks, comple-
mentary equipment, and civil works. Operational costs include
water, electricity, maintenance, staff, and property leasing costs.
The evaluation of the Net Present Value up to year t (NPVt; Equation
(1) e PV is the present value estimated as the monetary income
(sells and amortization) at the year i minus OPEX and taxes (25%) - t
is the discount rate), Internal Rate of Return (IRR e discount rate
that makes the NPV equal to 0) and Pay-Back Period (PB e year at
which the NPV equal to 0 using current discount rate and consid-
erations) was done according to the project evaluation methods
[24e26]. The main European manufacturers’ amounts have been
considered in the estimation of main equipment costs. An 8% in-
ternal rate and a temporary horizon of 25 years were used in the
economic evaluation. The extensive number of years used for
project evaluation is considered to observe trends that are appre-
ciable only after 10 years of project evaluation.
Sensitivity analyses were performed by two methods. First, one
at a time of specific parameters were modified (plant operating
hours, capacity in MW of the electrolysis plant (that could be
considered as an approximation of the synthesis plant size) elec-
tricity price, electrolyzer cost, HeB cycle cost, and the ammonia
sales price) to evaluate the effect of the economic indexes. Second,
Oracle Crystal Ball tool was used to perform sensitivity analysis of
the aggregate effect of the previously mentioned variables. In these
simulations, 10,000 iterations were carried out in which triangular
probability distributions (described below) were used.
 C. Fúnez Guerra et al. / Renewable Energy 157 (2020) 404e414
Fig. 1. Schematic of the renewable ammonia production plant.
X
t X
t
net income
NPVt ¼ PV ¼ (1)
i¼0 i¼0 ð1 þ tÞi
2.3. Optimization process
An optimization process to scale the optimum size of the elec-
trolyzer is carried out independently of the sensitivity analyses. The
optimization is performed based on the considerations that the cost
of the electrolyzer stack tend to be a non-linear function of the total
capacity (MW). The increase in electrolyzer capacity affects other
component size (i.e. CAPEX) and operational cost (i.e. OPEX),
implying a higher investment cost but at the same time a higher
production. Nonetheless, a higher investment implies a higher risk
from investors, therefore an increase in the discount rate is ex-
pected. The electrolyzer cost was estimated based on three ranges
of stack prices (1, 10, and 100 MW) while the discount rate was
chosen based on investment ranges (V 0e50 million, V 50e100
million, V 100e150 million, and V 150e200 million). Other con-
siderations (apart from the numeral 8, 9, and 11 from the following
section) were not modified during the optimization process. The
final discount rate was calculated using Equation (2), where kn is
the discount rate, R is the risk (2.5%, 4%, 6.5%, and 1%), T is the base
discount rate, and Ir is the inflation rate (1.5%).
kn ¼ ð1 þ R þ TÞð1 þ IrÞ  1 (2)
2.4. Considerations for the technical-economic study
The following considerations are taken into account in the
technical-economic study:
1. Yearly operation of 7,500 h per year (~85% of operational
yearly period).
407
2. Electrolyzer efficiencies that could range from 70% up to
76.5% (given a degradation of mV per hour). This implies that
more electric energy to produce the same amount of
hydrogen would be required (i.e. less hydrogen is produced
every year).
3. An electrolysis stack replacement would be needed after
80,000 h or after 10 years of use at full load, which implies an
initial value of 30% the investment cost and would decrease
at a rate of 1% per year from the first to the last year of the
study.
4. The Harber-Bosch cycle operating conditions are fixed at
pressure of 250 bar and temperature of 400
C. A catalytic
beds of iron, potassium, and aluminum oxides is considered
to obtain complete cycle conversion of 97% (i.e. considering
recirculation).
5. According to manufacturer, the catalytic bed needs to be
replaced after 10 years at full load (or 80,000 h of use). The
substitution of the catalyst involves an initial cost of 30% of
the investment cost of HeB process.
6. Oxygen and heat generated in the electrolyzer and the heat
generated in the compressor would not be valued.
7. 30% of the investment on the centralized generation plant
would be made with own resources; the remaining would be
made by a 10-year loan (using the French payment method).
8. The CAPEX and OPEX data of the main equipment being part
of the centralized generation plant are the following:

Polymeric electrolyzer. CAPEX: 450 V/kW of stack power.
OPEX (considering only maintenance cost): 1.5% of the
initial investment cost. Stack replacement: 80,000 h or 10
years.

HeB cycle including the air fractionation plant and the
ammonia storage at 18 bar. CAPEX: 425 V/kW of electro-
lyzer power. OPEX (considering only maintenance cost):
1.5% of the initial investment cost. Stack replacement:
80,000 h or 10 years.

Cost of services and civil works correspond to 10% of the
cost of the main equipment sum.
408 C. Fúnez Guerra et al. / Renewable Energy 157 (2020) 404e414
9. The investment cost of the ammonia production plant is V
144,375,000.
10. Electricity costs: 18 V/MWh.
11. Staff cost: 800,000 V/year.
12. Gross Water Cost: 2.5 V/m3.
13. Land leasing cost: 200,000 V/year.
14. Liquid ammonia transport from Chile to Japan of 50 V/t
15. Weighted Average Cost of Capital (WACC): 8%.
16. Inflation rate (Ir): 1.5% per year.
17. Ammonia sale cost of 450 V/t
3. Results
3.1. Results of the technical-economic green ammonia production
and transport
The technical-economic green ammonia production and trans-
port results, focus mainly in economical indicators like Net Present
Value, Internal Rate of Return and Return period of the investment
are the following:
 Net Present Value (V): 77,414,525.
 Internal Rate of Return (%): 17.
 Return period of the investment (years): 7.62.
Fig. 2 shows the Net Present Value of each of the years consid-
ered in the evaluation. The NPV25, estimated as the sum of each
present value up to the year of consideration, is equal to V
77,414,525. In the same figure it can be observed that an amorti-
zation is made between the year 7 and 8 (Return Period of the In-
vestment estimated as 7.62). It is observed a small decrease in the
NPVt around year 11 and year 21 which is related to the replace-
ment of the electrolysis stack and catalysts from Haber-Bosch
process (i.e. an increase in maintenance costs). In the same way,
123,400 t/year of liquid renewable ammonia are produced annually
and the emission of 210,000 t/year of carbon dioxide is avoided.
3.2. Sensitivity studies of one variable at a time
It is crucial to carry out a sensitivity analysis of the key factors in
Fig. 2. Net Present Value (V) VS time (years).
order to evaluate the limiting conditions in the decision making.
The parameters considered in the analysis are:
(1) PB variation based on ammonia sales price in Japan port;
(2) NPV variation based on the electricity price;
(3) NPV variation based on ammonia facility size;
(4) NPV variation based on operating hours;
(5) PB variation based on HeB cycle cost; and
(6) PB variation based on electrolyzer cost.
During each of the sensitivity analysis all the variables other
than the one under evaluation were kept constant.
3.2.1. PB variation based on the ammonia sales price in Japan port
The sales price (V/t) of the renewable ammonia in Japanese
ports changed between 350 V/t and 850 V/t. As may be seen in
Fig. 3, the return periods of the investment are longer than 10 years
for a sales price of renewable ammonia in Japanese ports lower
than 400 V/t, making it difficult for companies to make a decision
to invest.
3.2.2. NPV variation based on the electricity price
The price of electricity changed from 10 V/MWh to 30V/MWh.
As seen in Fig. 4, a linear effect between the NPV and the electricity
cost was obtained. Furthermore, the ammonia generation plant
would not be economically viable for electricity costs above 26
V/MWh; in other words, electricity prices lower than 26 V/MWh
are needed in order to obtain a positive Net Present Value.
3.2.3. NPV variation based on capacity in MW of the electrolysis
plant
The size of the capacity in MW of the electrolysis plant changed
from 5 MW to 235 MW. As shown in Fig. 5, a higher NPV is obtained
as the size of the electrolyzer capacity is increased. This trend
(linear) is shown because the study considers the scale factor
related to the investment costs (CAPEX) and also because, after
increasing the electrolyzer capacity to a certain point, the costs
would be almost linear (the equipment cost is not linear but the
energy and water consumption is linear; the effect of the energy
consumption is the most important and, for this reason, the
 C. Fúnez Guerra et al. / Renewable Energy 157 (2020) 404e414 409

Fig. 3. Pay back (years) VS ammonia sales price in Japan (V/t).

Fig. 4. Net Present Value (V) VS electricity cost (V/MWh).

behaviour is almost linear). In the same way, it could be considered
that the NPV would be less than zero for renewable ammonia
generation facilities with an equivalent electrolyzer capacity lower
than 50 MW (i.e., the facility would not be economically viable).
3.2.4. NPV variation based on operating hours
The hydrogen generation plant operating hours changed from
1,000 h/year to 8,000 h/year. Fig. 6 shows that at higher annual
operating hours, the higher the Net Present Value. Likewise, it can
be seen that it is necessary for the renewable ammonia generation
plant of 150 MW electrolysis stack to operate at least 3,700 h per
year in order to avoid economic losses.
3.2.5. PB variation based on Haber-Bosch cycle cost
The investment cost of the HeB cycle measured in V/kW of the
required electrolysis stack changed from 300 V/kW to 800 V/kW.
As shown in Fig. 7, a lower investment cost of the HeB cycle implies
a shorter investment period of return. The cost of the HeB cycle is
used to make the sensitivity analysis since it is the equipment,
together with the electrolyzer, that has a higher cost and, therefore,
one of the biggest impacts from the economic perspective. In the
same way, it can be observed that it is necessary for the costs of the
HeB cycle to be lower than 630 V/kW to have a Pay-Back under 10
years.
3.2.6. PB variation based on electrolyzer cost
The electrolyzer investment costs varied from 300 V/kW to 900
410 C. Fúnez Guerra et al. / Renewable Energy 157 (2020) 404e414

Fig. 5. Net Present Value (V) VS ammonia facility size (electrolyzer MW capacity).

Fig. 6. Net Present Value (V) VS operating hours (hours).

V/kW. As it is shown in Fig. 8, the lower the electrolyzer investment
cost, the shorter the return period of the investment. The electro-
lyzer cost is used to make the sensitivity study since it is the
equipment, together with the HeB cycle, that has a higher cost and,
therefore, it has one of the biggest impacts from the economic point
of view. In the same way, it is also emerged that it is necessary for
the costs of the polymer electrolysis technology to be lower than
620 V/kW in order to have a PB under 10 years. Likewise, it is
observed that the return period of the investment for electrolysis
costs higher than 800 V/kW increases considerably, which in-
dicates that it is the maximum price that could be reached by the
electrolysis equipment without significantly penalizing the eco-
nomic profitability of the renewable ammonia generation plant.
The problem was solved as a nonlinear function in the fmincon
solver. Parameters were kept as: Interior point as algorithm and 1e-
3 as optimality tolerance. The upper lower limit and upper limit of
the stack power were set as 0 and 500 MW, respectively. As result it
was obtained an optimum of 164.21 MW, which is relatively close
to the stack specified on the sensitivity analyses. This stack pro-
duced a NPV of V 88,221,025 (approximately an increase of a 10% by
increasing the stack power in 14.21 MW). A considerable increase
over the pre-specified stack. These results implies that the meth-
odology of investment should take considered during techno-
economic analyses and could provide considerable benefits by
small modifications of the productions sizes.

3.2.8. Multiparametric sensitivity studies

3.2.7. Optimization results In this work, the Oracle Crystal Ball tool [27e29] was used to
The optimization problem was solved by using Matlab R2019b. analyse the effect of the different parameters on the NPV: plant
 C. Fúnez Guerra et al. / Renewable Energy 157 (2020) 404e414 411

Fig. 7. Haber-Bosch cost (V/kW stack electrolyzer needed) VS pay-back (years).

Fig. 8. Electrolyzer cost (V/kW) VS pay-back (years).

operating hours, size of the ammonia synthesis plant (in relation
with the capacity in MW of the electrolysis plant size), electricity
price, electrolyzer cost, HeB cycle cost and ammonia sales price.
Each of the values with a given range were handled with a trian-
gular distribution where the maximum value is reported as the
most probable one.
 Operating hours. Between 1,000 and 8,400, the most probable
value of 7,500 h.
 Size of the ammonia synthesis plant, referred to the capacity of
the electrolysis stack in MW. Between 0 and 250 MW, the most
probable value of 150 MW.
 Cost of electricity in V/MWh. Between 0 and 70 V/MWh. Most
probable value of 18 V/MWh.
 Cost of the electrolyzer in V/kW. Between 300 and 900 V/kW.
Most probable value of 450 V/kW.
 Cost of the HeB cycle in V/kW of the electrolysis plant. Between
300 and 800 V/kW. Most probable value of 425 V/kW.
 Ammonia sales price including the cost of transportation to
Japan in V/t. Between 200 and 600. Most probable value of 450
V/Tn.
From the base case being considered (NPV of V 77,414,525)
some simulations were done in order to estimate the success
probability of the project. Fig. 9 shows that, with a 95% of
412 C. Fúnez Guerra et al. / Renewable Energy 157 (2020) 404e414
Fig. 9. Probability distribution of the Net Present Value (NPV) sensitivity analysis.
Table 1
Analysis of the sensitivity to Net Present Value.
Assumptions
Electricity Price (V/MWh)
Haber-Bosch cycle cost (V/kW)
Plant operating hours (h)
Electrolyzer Cost (V/kW)
Size of the ammonia synthesis plant (MW)
Ammonia sales price (V)
confidence, the NPV would be higher than zero for 18.10% of the
combinatorial cases.
Table 1 shows the most important variables that generate con-
tributions to the NPV distribution obtained; as observed the elec-
tricity price explains 96.37% of the NPV variance and therefore it
has the biggest effect in the NPV. (i.e. Net Present Value is nega-
tively correlated with the electricity price).
Further analysis were performed by increasing the plant oper-
ating hours (increased up to 8,000 h) and modifying the electricity
price (V/MWh) distribution (considered in the range of 0e30 with
18 the most probable value). The new Monte Carlos run show, with
a 95% of confidence, that a NPV higher than zero is obtained for
76.1% of the studied cases (see Fig. 10). For this simulation, a NPV of
88,777,680 (V) and a 7.22-year PB are obtained for. Under this
scenario, the ammonia synthesis plant explains 46.37% of the NPV
variance and the electricity price explaining 42.70% of the NPV
variance (Table 2). Other variables, such as plant operation hours,
electrolyzer cost and Haber-Bosch cycle cost explains less than
5.00% of the NPV variance.
4. Conclusions
The most relevant contributions that can be summarized from
the present work, and could bias implementation of a production
factory of ammonia in Chile based on renewable energies are that
(1) sales price of the renewable ammonia in Japanese ports must be
lower than 400 V/t in order to maintain an acceptable Pay-Back
period (PB) (that is, under 10 years). Values above 400 V/t offer a
PB that companies would not be willing to assume in the short
term. (2) The electricity cost is a key parameter (renewable elec-
tricity prices higher than 26 V/MWh would make the economic
Contribution to variance (%) Rank correlation
96.37 0.92
1.35 0.11
0.97 0.09
0.92 0.09
0.38 0.06
0.00 0.00
aspect of the project implementation negative). (3) The number of
operating hours is a parameter that has a relevant impact on the
profitability of the ammonia production plant, being necessary at
least 3,700 h per year to obtain positive economic profitability. (4)
As expected, the plant scale factor has a positive influence on the
economic profitability. (5) Electrolyzer cost values higher than 800
V/kW have a very significant impact in the return period of the
investment; this indicates that it is essential to have electrolyzer
acquisition costs as low as possible. (6) There are plant secondary
products (oxygen and the residual heats) that may generate addi-
tional income, making evaluations to have higher probability of
success. (7) In the base case, the Net Present Value is equal to V
77,414,525 and the amortization is made between the year 7 and 8
(Return Period of the Investment estimated as 7.62). In the same
way, 123,400 t/year of liquid renewable ammonia are produced
annually and the emission of 210,000 t/year of carbon dioxide is
avoided. (8) The multiparametric sensitivity analysis shows that the
electricity cost explain 96.37% of the NPV variance. When is
increased the plant operation hours and the distribution of elec-
tricity price is modified, the more important variables to NVP
variance are the size of the ammonia synthesis plant and electricity
cost. (9) The risk of investment should be considered at the
moment of performing techno-economic analyses and assess their
effect by optimization could produce extra benefits. An optimiza-
tion performed over the electrolyzer stack net power showed that
about a 10% of income could be increased by small modifications in
the stack size. (10) The main limitations of this study are a) this
study considers a power purchase agreement with fix price for all
the study period and sometimes it is so complex to obtain and b)
this study don’t consider the ammonia market volume and the
availability to introduce the ammonia produced in this facility.
 C. Fúnez Guerra et al. / Renewable Energy 157 (2020) 404e414 413

Fig. 10. Effect of sensitivity electricity price in the NPV.

Table 2
Effect of plant operation hours and electricity price on the Net Present Value sensitivity.

Assumptions Contribution to variance (%) Rank correlation

Size of the ammonia synthesis plant (MW) 46.37 0.65

Electricity Price (V/MWh) 42.70 0.62
Plant operating hours (h) 4.94 0.21
Electrolyzer Cost (V/kW) 3.04 0.17
Haber-Bosch cycle cost (V/kW) 2.94 0.16
Ammonia sales price (V) ~0.00 ~0.00

Declaration of competing interest
The authors declare that they have no known competing
financial interests or personal relationships that could have
appeared to influence the work reported in this paper.
CRediT authorship contribution statement
C. Fúnez Guerra: Conceptualization, Methodology, calculations.
L. Reyes-Bozo: Data curation, Writing - original draft. E. Vyh-
meister: Supervision. M. Jae
n Caparro

s: Visualization, Investiga-

Luis Salazar: Software, Validation. C. Clemente-Jul:
tion. Jose
Writing - review & editing.
Acknowledgments

noma de Chile
Grants to Dr. L. Reyes-Bozo from Universidad Auto
(“Apoyo a Eventos Científicos Nacionales o Internacionales” and
“Proyecto VRIP EI 04e2020” both from Vicerrectoría de Inves-

n y Postgrado).
tigacio
References
[1] David M. Kiefer, Chemistry Chronicles, American Chemical Society, 2001.
Available online: http://pubs.acs.org/subscribe/archive/tcaw/10/i02/html/
02chemch.html.
[2] A. Bernhard, The nitrogen cycle: processes, players, and human impact, Nat.
Edu. Knowl. 3 (10) (2010) 25.
[3] A. Vojvodic, A.J. Medford, F. Studt, F. Abild-Pedersen, T.S. Khan, T. Bligaard,
J.K. Norskov, Exploring the limits: a low-pressure, low-temperature Haber-
Bosch process, Chem. Phys. Lett. 598 (2014) 108e112.
[4] C. Philibert, Renewable Energy for Industry. From Green Energy to Green
Materials and Fuels. OECD/IEA, International Energy Agency, 2017.
[5] S.K. Ritter, The Haber-Bosch reaction: an early chemical impact on sustain-
ability, Chem. Eng. News 86 (33) (2008). Available online: https://cen.acs.org/
articles/86/i33/Haber-Bosch-Reaction-Early-Chemical.html.
[6] L. Wang, M. Xia, H. Wang, K. Huang, C. Qian, C.T. Maravelias, G.A. Ozin,
Greening ammonia toward the solar ammonia refinery, Joule 2 (2018) 1e20,
https://doi.org/10.1016/j.joule.2018.04.017.
[7] K. Kohse-Ho €inghaus, Clean combustion: chemistry and diagnostics for a sys-
tems approach in transportation and energy conversion, Prog. Energ Combust.
65 (2018) 1e5.
[8] D. Miura, T. Tezuka, A comparative study of ammonia energy systems as a
future energy carrier, with particular reference to vehicle use in Japan, Energy
68 (2014) 428e436.
[9] Y. Bicer, I. Dincer, C. Zamfirescu, G. Vezina, F. Raso, Comparative life cycle
assessment of various ammonia production methods, J. Clean. Prod. 135
(2016) 1379e1395.
[10] M. Appl, Ammonia. Ullmann’s Encyclopedia of Industrial Chemistry, Wiley-
VCH Verlag GmbH & Co KGaA, 2006.
[11] J. Benner, M. van Lieshout, H. Croezen, Identifying Breakthrough Technologies
for the Production of Basic Chemicals. A Long Term View on the Sustainable
Production of Ammonia, Olefins and Aromatics in the European Region, Delft
Report, CE Delft, 2012. Available online: https://www.cedelft.eu/en/
publications/download/1258.
[12] S. Neckel, The sustainability society: a sociological perspective, Cult. Pract. Eur.
2 (2) (2017) 46e52.
[13] L. Ye, R. Nayak-Luke, R. Ban ~ ares-Alca
ntara, E. Tsang, Reaction: “Green”
ammonia production, Inside Chem. 3 (5) (2017) 712e714.
[14] A. Allman, P. Daoutidis, D. Tiffany, S. Kelley, A framework for ammonia supply
chain optimization incorporating conventional and renewable generation,
AIChE J. 63 (10) (2017) 4390e4402.
[15] Y. Bicer, I. Dincer, Life cycle assessment of ammonia utilization in city trans-
portation and power generation, J. Clean. Prod. 170 (2018) 1594e1601.
[16] R. Michalsky, B.J. Parman, V. Amanor-Boadu, P.H. Pfromm, Solar thermo-
chemical production of ammonia form water, air and sunlight: thermody-
namic and economic analyses, Energy 42 (2012) 251e260.
[17] M.A. Shipman, M.D. Symes, Recent progress towards the electrosynthesis of
ammonia from sustainable resources, Catal. Today 286 (2017) 57e68.
[18] L.A. Cisternas, E.D. Ga
lvez, Chile’s Mining and Chemicals Industries, CEP
Magazine, American Institute of Chemical Engineers (AIChE), 2014. Available
online: energycongress.org/sites/default/files/cep/20140646_2.pdf.
[19] United nations commodity trade statistics database, UN Comtrade. Available
online: https://comtrade.un.org/db/mr/daCommoditiesResults.aspx?
px¼H2&cc¼2814. (Accessed 12 June 2018).
[20] E. Vyhmeister, C. Aleixendri Mun ~ oz, J.M. Bermúdez Miquel, J. Pina Moya,
C. Fúnez Guerra, L. Rodríguez Mayor, A. Godoy Faúndez, P. Higueras,
C. Clemente-Jul, H. Valde
s-Gonz
alez, L. Reyes-Bozo, A combined photovoltaic
and novel renewable energy system: an optimized techno-economic analysis
414 C. Fúnez Guerra et al. / Renewable Energy 157 (2020) 404e414

for mining industry applications, J. Clean. Prod. 149 (2017) 999e1010. Prod. Econ. 42 (1995) 79e96.
[21] F. Baldi, A. Azzi, F. Marechal, From renewable energy to ship fuel: ammonia as [26] V. Bartosova

, P. Majer
k, D. Hraskova
ca
, Taking risk into account in the eval-
an energy vector and mean for energy storage, Comput. Aid. Chem. Eng. 46 uation of economic efficiency of investment projects: traditional methods,
(2019) 1747e1752. Proc. Econ. Financ. 24 (2015) 68e75.
[22] J. Armijo, C. Philibert, Flecible production of green hydrogen and ammonia [27] Oracle (Epm Information Development Team), Oracle® crystal Ball enterprise
from variable solar and wind energy: case study of Chile and Argentina, Int. J. performance management system, release 11.1.2.4. User’s guide, Available on
Hydrogen Energy 45 (3) (2020) 1541e1558. line: http://www.oracle.com/technetwork/middleware/crystalball/
[23] O. Siddiqui, I. Dincer, A new solar energy system for ammonia production and documentation/index.html, 2015.
utilization in fuel cells, Energy Convers. Manag. 208 (15) (2020) 112590. [28] J. Swan, Sensitivity analysis and scenarios, in: J. Swan (Ed.), Practical Financial
[24] Alexandre C. Dimian, Costin S. Bildea, Anton A. Kiss, Economic evaluation of Modelling, the Development and Audit of Cash Flow Models, third ed.,
projects, in: Alexandre C. Dimian, Costin S. Bildea, Anton A. Kiss (Eds.), Elsevier Ltd., 2016, pp. 231e254.
Computer Aided Chemical Engineering, vol. 35, Elsevier, 2014, pp. 717e755. [29] L. Papada, D. Kaliampakos, A Stochastic Model for energy poverty analysis,
[25] D.S. Remer, A.P. Nieto, A compendium and comparison of 25 project evalua- Energy Pol. 116 (2018) 153e164.
tion techniques. Part 1: net present value and rate of return methods, Int. J.