of 18
2354
|
Energy Environ. Sci.,
2016,
9
, 2354--2371
This journal is
©
The Royal Society of Chemistry 2016
Cite this:
Energy Environ. Sci.,
2016,
9
,2354
A comparative technoeconomic analysis of
renewable hydrogen production using solar
energy
Matthew R. Shaner,
ab
Harry A. Atwater,
ac
Nathan S. Lewis*
ab
and
Eric W. McFarland*
d
A technoeconomic analysis of photoelectrochemical (PEC) and photovoltaic-electrolytic (PV-E) solar-
hydrogen production of 10 000 kg H
2
day

1
(3.65 kilotons per year) was performed to assess the
economics of each technology, and to provide a basis for comparison between these technologies as
well as within the broader energy landscape. Two PEC systems, differentiated primarily by the extent of
solar concentration (unconcentrated and 10

concentrated) and two PV-E systems, differentiated by the
degree of grid connectivity (unconnected and grid supplemented), were analyzed. In each case, a base-
case system that used established designs and materials was compared to prospective systems that
might be envisioned and developed in the future with the goal of achieving substantially lower overall
system costs. With identical overall plant efficiencies of 9.8%, the unconcentrated PEC and non-grid
connected PV-E system base-case capital expenses for the rated capacity of 3.65 kilotons H
2
per year
were $205 MM ($293 per m
2
of solar collection area (m
S

2
), $14.7 W
H2,P

1
) and $260 MM ($371 m
S

2
,
$18.8 W
H2,P

1
), respectively. The untaxed, plant-gate levelized costs for the hydrogen product (LCH) were
$11.4 kg

1
and $12.1 kg

1
for the base-case PEC and PV-E systems, respectively. The 10

concentrated
PEC base-case system capital cost was $160 MM ($428 m
S

2
, $11.5 W
H2,P

1
) and for an efficiency of
20% the LCH was $9.2 kg

1
. Likewise, the grid supplemented base-case PV-E system capital cost was
$66 MM ($441 m
S

2
, $11.5 W
H2,P

1
), and with solar-to-hydrogen and grid electrolysis system efficiencies
of 9.8% and 61%, respectively, the LCH was $6.1 kg

1
. As a benchmark, a proton-exchange membrane
(PEM) based grid-connected electrolysis system was analyzed. Assuming a system efficiency of 61% and
a grid electricity cost of $0.07 kWh

1
, the LCH was $5.5 kg

1
. A sensitivity analysis indicated that, relative
to the base-case, increases in the system efficiency could effect the greatest cost reductions for all
systems, due to the areal dependencies of many of the components. The balance-of-systems (BoS)
costs were the largest factor in differentiating the PEC and PV-E systems. No single or combination of
technical advancements based on currently demonstrated technology can provide sufficient cost
reductions to allow solar hydrogen to directly compete on a levelized cost basis with hydrogen
produced from fossil energy. Specifically, a cost of CO
2
greater than
B
$800 (ton CO
2
)

1
was estimated
to be necessary for base-case PEC hydrogen to reach price parity with hydrogen derived from steam
reforming of methane priced at $12 GJ

1
($1.39 (kg H
2
)

1
). A comparison with low CO
2
and CO
2
-neutral
energy sources indicated that base-case PEC hydrogen is not currently cost-competitive with
electrolysis using electricity supplied by nuclear power or from fossil-fuels in conjunction with carbon
capture and storage. Solar electricity production and storage using either batteries or PEC hydrogen
technologies are currently an order of magnitude greater in cost than electricity prices with no clear
advantage to either battery or hydrogen storage as of yet. Significant advances in PEC technology
performance and system cost reductions are necessary to enable cost-effective PEC-derived solar
hydrogen for use in scalable grid-storage applications as well as for use as a chemical feedstock
precursor to CO
2
-neutral high energy-density transportation fuels. Hence such applications are an
a
Joint Center for Artificial Photosynthesis, California Institute of Technology, Pasadena, CA, USA. E-mail: nslewis@caltech.edu
b
Division of Chemistry and Chemical Engineering, California Institute of Technology, Pasadena, CA, USA
c
Thomas J. Watson Laboratories of Applied Physics, California Institute of Technology, Pasadena, CA, USA
d
Dow Centre for Sustainable Engineering Innovation, Department of Chemical Engineering, University of Queensland, Australia. E-mail: e.mcfarlan
d@uq.edu.au
Electronic supplementary information (ESI) available. See DOI: 10.1039/c5ee02573g
Received 19th August 2015,
Accepted 5th May 2016
DOI: 10.1039/c5ee02573g
www.rsc.org/ees
Energy &
Environmental
Science
PAPER
Published on 26 May 2016. Downloaded by California Institute of Technology on 30/09/2016 16:28:35.
View Article Online
View Journal
| View Issue
This journal is
©
The Royal Society of Chemistry 2016
Energy Environ. Sci.,
2016,
9
, 2354--2371 |
2355
opportunity for foundational research to contribute to the development of disruptive approaches to
solar fuels generation systems that can offer higher performance at much lower cost than is provided by
current embodiments of solar fuels generators. Efforts to directly reduce CO
2
photoelectrochemically
or electrochemically could potentially produce products with higher value than hydrogen, but many,
as yet unmet, challenges include catalytic efficiency and selectivity, and CO
2
mass transport rates and
feedstock cost. Major breakthroughs are required to obtain viable economic costs for solar hydrogen
production, but the barriers to achieve cost-competitiveness with existing large-scale thermochemical
processes for CO
2
reduction are even greater.
Broader context
Hydrogen and, more broadly, chemical production using solar energy can serve as an energy dense form of decarbonized transportation fuel and reduce t
he
variability of solar electricity production by serving as an energy storage medium. However, to have significant impact, the technological solutio
ns capable of
producing chemicals from solar energy must necessarily be competitive within the economic realities of the marketplace. Rigorous economic competi
tive
analyses, applied after proof-of-concept research and development, can provide critical guidance on a project’s further resource allocation, pri
orities and
trajectory. Our analysis suggests that achieving solar-to-hydrogen system efficiencies of greater than 20% within current embodiments of solar H
2
generators, is
not sufficient to achieve hydrogen production costs competitive with fossil-fuel derived hydrogen. Panel mounting materials, labor and other balanc
eof
systems costs, irrespective of the active materials, amount to hydrogen production cost values in excess of current hydrogen and energy prices. Radi
cally new
materials and system designs that achieve fully installed costs similar to simple material installations such as artificial grass and efficiencies ne
ar
thermodynamic limits are required to achieve the equally dramatic cost reductions needed for solar hydrogen to compete with current generation tech
nologies;
similar if not larger techno-economic challenges hold for CO
2
reduction technologies.
Introduction
Electrolysis using solar energy as a potential commercial source
of hydrogen from water has been pursued for over four decades.
1
Solar-driven water electrolysis has been practiced in two basic
system configurations; (1) photoelectrochemical (PEC) water
splitting, which consists of a single, fully integrated unit that
absorbs sunlight and produces hydrogen and oxygen, and (2)
photovoltaic electrolysis (PV-E), which consists of independent
photovoltaic modules that drive separate electrolyzer units. To
have significant impact on the worldwide supply of energy, these
technological solutions must necessarily be competitive within
the economic realities of the marketplace. Rigorous economic
competitive analyses, applied to these proof-of-concept research
and development technologies, can provide critical guidance
on their further resource allocation, priorities and trajectory.
Accordingly, we describe a technoeconomic evaluation of renew-
able and carbon-free hydrogen production by solar-driven water
splitting. In so doing we build on existing literature by adding:
(i) an updated technoeconomic evaluation of photoelectrochemical
systems based on recent engineering designs and prototypes,
(ii) a complete plant design evaluation and direct comparison of
photoelectrochemical and photovo
ltaic-electrolysis technologies,
(iii) a comparison of solar hydrogen production technoeconomics
to other low-carbon technological options, and (iv) an extension
of the solar hydrogen technoeconomic analysis to solar-driven
CO
2
reduction systems.
The systems analyzed herein include two integrated PEC designs,
as well as grid electrolysis with proton-exchange membrane electro-
lyzers and two PV-E designs using discrete photovoltaic modules
and electrolyzer units. Current and predicted hydrogen production
prices from steam reforming of natural gas (SMR) are reported
as a benchmark. The capital and operating expenses for each
system have been estimated based on technical design specifica-
tions, and allowed calculation of an
estimated plant-gate levelized
cost of hydrogen such that the net present value is zero at the
end of the plant life.
Prior to broader comparisons, an initial comparison between
solar hydrogen production methods has been performed, to
determine the least expensive technology and to suggest future
research needs. Integrated photoelectrochemical hydrogen
production and discrete photovoltaic electrolysis hydrogen
production constitute functionally identical systems and hence
can be compared directly on a cost-basis. The trade-offs involving
construction of a single integrated unit that has potentially
fewer components and directly produces hydrogen, relative to
the increased operational flexibility of the discrete photovoltaic
electrolysis configuration, will therefore ultimately determine
the most economic technology that provides this specific
quality and quantity of energy.
Subsequently, the most economic solar hydrogen source is
compared to steam reforming (SR) of relatively low-cost fossil
hydrocarbons, the dominant current source of molecular hydrogen.
The costs of production of hydrogen by SR are well known at
B
$1.39 kg

1
or $0.042 kWh

1
($3 (MM BTU natural gas)

1
),
which is less than current US average electricity prices.
2
In the
absence of a price applied to CO
2
production, or other policy-
driven mandates such as a renewable fuels standard, all hydrogen
production technologies will compete in the marketplace directly
against fossil fuels for energy production and storage. Because
photovoltaic electricity production currently is more expensive in
most locations than levelized electricity prices of $0.07 kWh

1
,
the more complicated task of solar hydrogen production by
stand-alone or grid assisted PV-electrolysis is not expected to be
economically favored relative to fossil-fuel-derived energy or
hydrogen. Given the length of energy system transitions being
Paper
Energy & Environmental Science
Published on 26 May 2016. Downloaded by California Institute of Technology on 30/09/2016 16:28:35.
View Article Online
2356
|
Energy Environ. Sci.,
2016,
9
, 2354--2371
This journal is
©
The Royal Society of Chemistry 2016
generally 40–60 years or more,
3
under this scenario, fossil fuels are
thus expected to continue to dominate over any solar hydrogen
system throughout at least the first half of this century.
However, solar hydrogen technologies constitute a carbon-
neutral source of energy production and storage, and thus
provide a differentiated quality of energy that may eventually
be valued in the marketplace. Hence we have also compared the
cost of solar hydrogen to other carbon neutral or low carbon
sources of hydrogen that could play a role in a carbon-constrained
energy market. Nuclear fission-ba
sed grid electrolysis and biomass
reforming are two of the main alte
rnative technical approaches,
though biomass-derived energy is potentially limited in scale due
to land area constraints. Another potential low-carbon technology
option is fossil-fuel-derived grid electrolysis in conjunction
with carbon capture and storage (CCS).
We have also compared the cost of solar hydrogen to other
approaches that can provide similar functionality as a part of a
low-carbon energy system. Carbon-neutral energy production
and storage technologies, such as electricity derived from either
nuclear fission or solar electricity, in conjunction with battery
storage, pumped hydroelectricity, or compressed air-based energy
storage, provide alternative technological options relative to the
use of solar hydrogen in the grid storage and, in some cases, the
transportation sectors. These technologies mainly compete with
the electrolysis unit, and all of the approaches will have different
operational efficiencies as well as mutually different capital and
operating expenses. Many of these existing technologies have a
first-to-market advantage, while PEC-derived hydrogen remains
at a fundamental research level.
Solar hydrogen technologies
In each of the PEC and PV-E system configurations, solar
photons are absorbed in semiconducting materials that have
at least one junction that converts photogenerated electron/
hole pairs into incipient electrical energy. The photogenerated
electrons and holes are collected asymmetrically at the two
electrodes and are transferred to electrocatalysts or electro-
catalytic sites to perform the respective hydrogen- and oxygen-
evolution reactions. The ions that are generated at one electrode
surface must be transported through a membrane and/or
electrolyte to complete the electrochemical circuit, and must
react to form the complementary product without an explosion
hazard being present. The products are collected separately, or
alternatively must be separated subsequent to collection and
then processed for final use.
Numerous types of photoelectrochemical cells have been
demonstrated at the laboratory scale, with solar-to-hydrogen
(STH) efficiencies as high as 12.4% for a cell possessing at least
one semiconductor–liquid junction
4
and 18% for a cell con-
structed from semiconductors that only contain buried semi-
conductor junctions.
5–7
Many small-scale demonstrations of
photovoltaic-based electrolysis systems, and models optimizing
their behavior, have been described, with differing levels
of complexity of the connection between the photovoltaic
modules and electrolyzers leading to differing operational
flexibility and ultimately to different system efficiencies.
8–10
In general, the efficiency of a PV-E system is the product of the
individual efficiencies of the photovoltaic module, the power
electronics and the electrolyzer unit.
The current costs of photovoltaic installations and compo-
nents are well known, with national- and state-level monitoring
of the total installed costs for residential, commercial and utility-
scale photovoltaic systems performed extensively throughout the
United States and parts of Europe.
11,12
Commercial electrolyzer
costs, including proton-exchange membrane (PEM) and alkaline
electrolyzers, are also known, with published values verified by
system manufacturers.
13
Many configurations are possible for a photovoltaic electro-
lysis system, each having different systems economics. One of
two configurations analyzed herein consists of a photovoltaic
array interfaced directly to a PEM electrolyzer. The electrolyzer
units have been sized to accept all, or most, of the maximum
instantaneous power produced by the photovoltaic array. This
design results in a capacity factor for the electrolyzer equal to
that of the photovoltaic array (
B
20%). The second configu-
ration analyzed includes a grid connection to supplement the
electrical power supplied by the photovoltaic array, such that
the electrolyzers are able to operate at their maximum capacity
factor (97%), with the photovoltaics being sized such that their
maximum instantaneous power matches the capacity of the
electrolyzers. Another system not investigated herein, but that
could provide an economic opportunity, is a H
2
and electricity
co-generation system that consists of an overcapacity of
the photovoltaic component as compared to the electrolysis
component, similar to current photovoltaic installations that are
limited by the capacity of the inverter.
11
This type of configuration
would yield a slight increase in the
capacity factor of the electro-
lyzer, as is demonstrated by recent photovoltaic installations,
11
and could generate added revenue from sale of the excess
electricity during times of peak solar flux.
The key active components of PEC-based systems are currently
the subject of intense research and development. Many potential
configurations exist, including non-concentrating and concentrat-
ing planar semiconductor desi
gns (Type 3 & 4, respectively),
14
as
well as slurry systems that utili
ze particulate semiconductors
suspended in a solution to absorb light and effect hydrogen and
oxygen evolution (Type 1 & 2).
14,15
TheType3&4technologiescan,
and have, made use of existing knowledge from the photovoltaic
industry, and are thus further in development than Type 1 & 2
technologies. Accordingly, the costs of PEC systems are less well
understood as compared to PV-E
systems, because no commercial
PEC systems have been constructed and operated to date. To obtain
reasonable estimates and guide res
earch, technoeconomic analyses
have been performed for these Type 1–4 system configurations and
technology options.
14,15
The predicted levelized cost of hydrogen
(LCH) is lowest for the less-developed Type 1 & 2 systems,
14
albeit
with far more unknowns and thus more associated technological
as well as market risk, relative to the Type 3 & 4 technologies. We
update and build on these analyses herein by focusing on recent
PEC system engineering designs, broadening the scope of compar-
ison to discrete PV-E systems and other technological options, and
extending the analyses to CO
2
reduction concepts.
Energy & Environmental Science
Paper
Published on 26 May 2016. Downloaded by California Institute of Technology on 30/09/2016 16:28:35.
View Article Online
This journal is
©
The Royal Society of Chemistry 2016
Energy Environ. Sci.,
2016,
9
, 2354--2371 |
2357
Methods
Capital cost analysis
Table 1 lists the base-case design specification and financial
parameters that were used to evaluate the capital costs for
each 3.65 kiloton per year (10 000 kg per day, 13.8 MW of H
2
,
5.1 MW
e
given current MW-scale storage and fuel cell efficiencies)
system. All capital costs and results were inflation-adjusted to
2014 dollars.
14
Systems
Fig. 1 displays schematically the five types of systems that
were evaluated herein. The first two systems are photoelectro-
chemical in nature, with the first consisting of a louvered
design having slats of a semiconductor and catalyst oriented
towards the sun and slats of a membrane oriented perpendi-
cular to the sun, all held within a chassis that allows light
penetration while holding the aqueous electrolyte.
17
This
system is similar to the Type 3 system that was the subject of
a previous technoeconomic analysis.
14
The second PEC system
considered herein is similar to the first, but includes 10

optical
concentration and pressurized gas production of 10 atm from
the PEC module. This system is similar to a Type 4 system that
has been evaluated previously. In both PEC systems, H
2
gas is
collected
via
polyvinylchloride (PVC) piping that has been sized
to balance pipe usage against pumping losses.
14
The next two systems considered herein consist of photo-
voltaic modules connected through DC power electronics to
discrete electrolysis units. One system, referred to as PV-E,
relies solely on solar energy for hydrogen production. In this
system, the electrolyzers are connected to the photovoltaics
with or without a DC–DC converter, and are sized to match the
maximum output of the photovoltaics. The second system,
referred to as GSPV-E, includes a grid connection and sized
the electrolysis units based on their maximum capacity factor
such that grid electricity supplements the photovoltaic electri-
city whenever the photovoltaic modules are not operating at
their maximum capacity.
The last system, grid electrolysi
s, which served as a benchmark
by which to measure the above fo
ur systems, is the predominant
Table 1
Operating and financial parameters used for all systems analyzed
Parameter
Value
Hydrogen production rate
14
10 000 kg day

1
Plant lifetime
13
20 years
Hydrogen plant gate pressure
13
450 psi
Solar capacity factor (2008–2013 average)
16
0.204
Discount rate (
r
)
12%
Inflation rate
14
1.9%
$ basis year
2014
Fig. 1
(a) Block diagram depicting the power flow through a PEC plant. The cell specifics for the Type 3 and 4 systems are shown in the insets. (b) Block
diagram of the power flow through photovoltaic electrolysis (PV-E), grid assisted photovoltaic electrolysis (GSPV-E) and grid electrolysis plant
s.
Paper
Energy & Environmental Science
Published on 26 May 2016. Downloaded by California Institute of Technology on 30/09/2016 16:28:35.
View Article Online
2358
|
Energy Environ. Sci.,
2016,
9
, 2354--2371
This journal is
©
The Royal Society of Chemistry 2016
currently practiced technique for hydrogen production from
electricity.
A final general scenario is mapped out over a range of capital
expenses and STH full plant efficiency values, to demonstrate
their relationship to the LCH, as well as to describe the
performance and economic values that must be met for solar
hydrogen to be economically competitive with existing and
developing technologies.
Technoeconomic assumptions
All economic assumptions are base
d on values taken from the U.S.
market. In general, material and equipment capital expenses
are transferrable globally, but installation labor and other
soft balance of system costs such as customer acquisition and
permitting can be more location dependent. Weighted average
capital costs for utility scale photo
voltaic installations in 2013–2014
were
B
$2.3 W
p

1
in the U.S. with only Europe and China having
lower costs at
B
$1.9 W
p

1
and
B
$1.6 W
p

1
, respectively.
18
Such
differences are likely due to soft balance of systems cost differences
as is the case for residential systems, but the magnitude of the
differences is significantly smaller for utility-scale installation.
19
These capital cost differences for utility-scale systems are roughly
offset by the higher capacity factors in the U.S., suggesting that
the conclusions discussed herein remain valid irrespective of
the location dependent cost differences and are representative
of the state-of-the-art costs globally.
18
Photovoltaic electrolysis (PV-E) system.
Table 2 shows the
system specific technical parameters and capital expenses for
the PV-E system. Values for non-subsidized, single crystalline Si
photovoltaic module costs are taken from very recent wholesale
prices; these costs include the cells along with the encapsula-
tion and electronics necessary for operation and stability for
20+ years.
20
Wiring, panel mounting material and other hard-
ware balance-of-system (BoS) costs are taken from very recent
utility-scale photovolt
aic installation costs.
22
A direct connection
was assumed between the photovolt
aic modules and electrolyzers,
because the efficiency loss due to non-optimal operation is
similar to the efficiency losses incurred with a DC–DC converter
which can provide optimal operation at all times but incurs
additional costs for the converter unit.
23,24
The assumed elec-
trolyzer unit costs are identical to those for assumed for the
grid electrolysis system evaluated below.
13
Photovoltaic instal-
lation labor and other soft BoS costs are taken from very recent
utility-scale PV installation costs.
21,22
The base-case system STH efficiency was assumed to be 9.76%,
whichistheproductofthephot
ovoltaic module efficiency of 16%
and the electrolyzer pl
antefficiencyof61%.
13,21,22
Replacement
expenses for the electrolyzer were assumed to be identical to
that of the grid electrolysis system, whereas the photovoltaics
were assumed to last the lifetime of the plant.
13
Grid supplemented photovoltaic electrolysis (GSPV-E) system.
Table 3 shows the assumed GSPV-E system specific technical
parameters, capital expenses and electricity price. All costs are
identical to the PV-E system, except that grid electricity operating
costs and the capital costs of an AC–DC rectifier and DC–DC
converter are included for proper electrical control.
13,26
The
electrolysis units were sized based on their maximum capacity
factor (0.97), and the photovoltaics were sized such that their
maximum, instantaneous power output (at 1000 W m

2
)
matched the electrolysis capacity (1.8

10
5
m
2
). The electrolyzer
stack cost per solar collection area is the same as for the PV-E
system because both systems are sized to match the electrolyzer
to the maximum instantaneous power output of the photovoltaic
array. Thus, the area of solar collection determines the number
of electrolyzers or
vice versa
, and the electrolyzer cost per photo-
voltaic area remains constant for the PV-E and GSPV-E systems,
aside from the slightly sub-unity electrolyzer capacity factor for
the GSPV-E system. This set of assumptions resulted in 21% of
the hydrogen produced by the solar energy input and 79% of the
Table 2
PV-E system technical parameters and capital and operating
expenses
System specific technical parameters
STH efficiency
9.76%
Electrolyzer efficiency
61%
Electrolyzer and PV capacity factor
0.204
Photovoltaic efficiency
16%
Photovoltaic area (m
S
2
)
7.5

10
5
m
2
Number of PEM stacks (500 kg day

1
stack

1
)99
Capital expenses
Component
2014 $ m
S

2
Electrolyzer stacks
13
65
Photovoltaic modules
20
96
Wiring
21,22
16
Other electrolyzer hard BoS
13
61
Panel mounting materials
21,22
29
PV installation labor
21,22
29
Electrolyzer installation labor
13
19
Other PV soft BoS
21,22
56
Table 3
Grid-assisted PV-E system technical parameters and capital and
operating expenses
System specific technical parameters
STH efficiency
9.76%
Electrolyzer efficiency
61%
Electrolyzer capacity factor
0.97
Photovoltaic efficiency
16%
Photovoltaic area (m
S
2
)
1.8

10
5
m
2
Number of PEM stacks (500 kg day

1
stack

1
)21
Capital expenses
Component
2014 $ m
S

2
Electrolyzer stack
13
64
Photovoltaic modules
20
96
Wiring
21,22
16
DC–DC converter
51
AC–DC rectifier
13
30
Other electrolyzer hard BoS
13
61
Panel mounting materials
21,22
29
Photovoltaic installation labor
21,22
29
Electrolyzer installation labor
13
19
Other soft BoS
21,22
56
Operating and maintenance expenses
Electricity
25
$0.07 kWh

1
Energy & Environmental Science
Paper
Published on 26 May 2016. Downloaded by California Institute of Technology on 30/09/2016 16:28:35.
View Article Online
This journal is
©
The Royal Society of Chemistry 2016
Energy Environ. Sci.,
2016,
9
, 2354--2371 |
2359
hydrogen resulting from the inpu
tofgridpower.Thereplacement
expenses for the electrolyzer were assumed to be identical to those
assumed for the grid electrolysi
s system, and the photovoltaics
were assumed to last the lifetime of the plant. Implicit in the
electricity price is the cost of the existing transmission and
distribution system; if however ne
w transmission and distribution
is required for such a GSPV-E system due to remote siting of the
solar electricity installation, the electricity price could be signifi-
cantly higher than the base-case assumption.
Type 3 PEC system.
Table 4 shows the assumed Type 3 PEC
system specific technical parameters, capital expenses and elec-
tricity price. The semiconducto
r component cost was assumed to
be identical to the current Si photovoltaic cell cost ($0.38 W
p

1
)
and is distinct from the photovoltaic module cost for the PV-E and
GSPV-E systems (Tables 2 and 3) because these costs only encom-
pass the photoelectrode semicon
ductor material and fabrication
costs and do not include the module material and assembly costs.
Costs for junction formation and front contact metallization,
which comprise approximately 20%, or $0.08 W
p

1
, of the
cell cost, were excluded because PEC systems can utilize
semiconductor–liquid junctions.
20,29
This assumes that a tandem
and/or triple-junction stacked structure can be fabricated at costs
equivalent to Si cell fabrication today, with a solar-to-electric
efficiency equivalent to 16% and with current and voltage char-
acteristics optimized for the electrolysis current and voltage load
characteristics. Because this assumption has yet to be realized
commercially, three high-photovoltage (
4
650 mV) Si photo-
voltaic cells could be arranged side-by-side and wired electrically
in series, to produce the necessary voltage while still achieving
the efficiency metrics assumed; multiple architectures for such a
device have been outlined and/or demonstrated previously. Fig. 2
depicts one possible architecture with no major cost differences
expected between different side-by-side system designs.
31,32
The
semiconductor cell cost would increase by $13 m

2
(to $61 m

2
)
relative to the Si cell cost of $48 m

2
($0.3 W
p

1
) as specified in
Table 4 to include junction formati
on and front contact metalliza-
tion costs; the overall PV efficiency would remain identical being
equivalent to that of an individual cell. The major cost differentiator
between these two architectural options, stacked tandem or triple-
junction cell
versus
side-by-side, is the semiconductor cell costs;
all other components are expected to be identical. Thus, at
present the side-by-side design is expected to be the lowest cost
option commercially and the challenge for stacked cells is to
outcompete single junction Si cells.
Including the PEC chassis material, PEC module labor and
AR coated glass window would result in a component similar to
a PV module with costs (not including membrane or catalyst
costs) of $96 m

2
, identical to the PV module areal cost. Thus,
any capital cost differences between the Type 3 and PV-E systems
will be due to balance of system costs or any material differences
for the electrolysis portion of the system.
This near-term demonstration system serves as a baseline
for comparison with photoelectrochemical approaches on a
technoeconomic basis. Platinum (Pt) and iridium oxide (IrO
x
)
catalysts were assumed, a worst-case cost scenario because of
the high spot prices for both materials. The $8 m
S

2
cost of the
catalyst for a specified solar collection area (m
S
2
) is assumed to
be identical to that of the PV-E system because state-of-the-art
PEM electrolyzer catalyst loadings,
B
1mgcm

2
of Pt (466 nm
thick) and
B
2mgcm

2
of IrO
2
(1.7
m
m thick)
33
for 1–10 A cm

2
operating current density, correspond to similar total catalyst
mass loadings as state-of-the-art photoelectrochemical catalyst
loadings, 1–10
m
gcm

2
(0.5–5 nm thick) of Pt and 2–20
m
gcm

2
of IrO
2
(1.7–17 nm thick) for 10 mA cm

2
of operating current
density in a PEC system.
26
A Nafion PEM was assumed to serve as the ionically con-
ductive, gas impermeable membrane, with costs of $2000 kg

1
estimated based on current production volume prices for a
5mil(127
m
m) thick membrane.
30
Based on the photoelectro-
chemical cell design, the membrane area required is 10% of the
solar collection area.
34
A polypropylene chassis having a 1 cm
thickness and an area equal to the PEC area was assumed, with a
raw material price of $1.5 per kg.
28
The chassis was assumed to
be manufactured
via
injection molding, where the raw materials
Table 4
Type 3 PEC system technical parameters and capital and oper-
ating expenses
System specific technical parameters
STH efficiency
9.76%
PEC Area (m
S
2
)
7.6

10
5
m
2
Capital expenses
Component
2014 $ m
S

2
Window (AR coated glass)
27
5
Chassis (polypropylene)
28
33
Semiconductors (c-Si, 16% S-E)
20,29
48
Catalyst (Pt, IrO
x
)
26
8
Membrane (Nafion, 5 mil)
30
50
PEC cell assembly labor
13
10
Compressors (2 stage)
14
16
Water condenser
14
0.3
Heat exchangers
14
0.4
Piping (PVC)
14
3.4
Control systems
14
5.4
Panel mounting materials
21,22
29
Installation labor
21,22
29
Other BOS
21,22
56
Fig. 2
One possible architecture (not to scale) for a series connected
side-by-side triple junction Si PV cell structure directly integrated into the
chassis of an electrolysis unit designed for the Si device output. Such a
structure would be a single unit that could be installed like a traditional PV
panel, identical to the base case Type 3 design, with gas collection as
opposed to electrical connections.
Paper
Energy & Environmental Science
Published on 26 May 2016. Downloaded by California Institute of Technology on 30/09/2016 16:28:35.
View Article Online
2360
|
Energy Environ. Sci.,
2016,
9
, 2354--2371
This journal is
©
The Royal Society of Chemistry 2016
cost is approximately 43% of the total manufactured chassis
cost.
35
The window was assumed to be made from high quality,
anti-reflective glass used by the photovoltaic industry that is
compatible with acidic media.
27
Replacing the back of the
polypropylene chassis with glass would decrease the materials cost
of the PEC module. However, the cost differential is relatively small
as compared to the total capital cost, and increases in handling
related costs due to the different mechanical properties of glass
versus
polypropylene could nullify the
material cost differential. If,
however, a measurable difference in the base-case capital costs
assumed herein exists, the impac
t of these differences on the LCH
can be assessed using the analysis summarized below (in Fig. 10).
These cell materials were assumed to be resistant to degradation by
sunlight over the lifetime of the device, and any mechanical issues
related to thermal mismatches between materials were assumed
to be solved for the quoted costs of the base case. The photo-
electrochemical module assembly labor was assumed to be equal
to the electrolyzer assembly cost on a $ W

1
basis because
both systems entail assembly of the chassis and active compo-
nents (membrane electrode assembly for an electrolyzer and
membrane and photoelectrode(s) for a PEC device).
13
This is a
reasonable estimate given publicly available data, but is likely
an optimistic lower bound because the PEC system areal power
density (W m

2
) is roughly two orders of magnitude lower than
the electrolyzers, requiring significantly larger areas of PEC
components to be assembled and/or seamless integration of
the materials to allow for fabrication integrally and/or with
minimal labor.
The water delivery and gas collection, processing and control
system costs were taken from previous work,
14
but the compres-
sors were assumed to provide a higher compression ratio of
B
5.5 : 1
versus
4.5:1 in the reference case evaluated previously.
Polyvinyl chloride (PVC) piping was assumed in the base case due
to the sufficiently low hydrogen permeability and embrittlement
of PVC at the modest hydrogen collection pressures present in
both PEC systems.
26,36
These assumptions result in gas proces-
sing and water delivery unit costs ($ m
S

2
) that are roughly half
the cost of the units used to perform the same tasks in the PV-E
design. Confidence is higher on the PV-E hard BoS costs due to
the commercial maturity of each of the individual systems, while
the PEC system costs have only undergone a high-level engineer-
ing design because no known systems have received design
certifications and permitting nor been constructed.
14
Thus the
potential for significant changes to the PEC system hard BoS
costs exists, with the values assumed herein likely representing
an optimistic cost scenario. The panel mounting materials,
installation labor and other soft balance of systems (BoS) costs
were taken directly from utility-scale photovoltaic panel installa-
tions on a $ m
S

2
basis.
21,22
The installation was assumed to be
sited in areas that historically on a decade-scale have little chance
of experiencing a hard freeze, specifically in plant hardiness
zones 8 and above (
i.e.
where citrus trees are planted and thrive);
consequently, additional costs associated with heating to avoid
any liquid water from freezing were not included.
37
The active components (semico
nductors, catalyst, membrane)
were assumed to be replaced every 7 years, based on expected
component lifetimes from the electrolyzer industry, though no
complete PEC cell that performs unassisted water splitting has
yet been demonstrated to be stable for more than one week.
13,17
The installation cost for replacement components was taken to be
15%ofthecomponentcost.Allo
ther components were assumed
to need no replacement over the system lifetime. An annual
operating and maintenance cost o
f3.2%oftheinstalledcapital
was taken from the PEM electrolyzer industry.
13
All of the other
components (DI water production, initial charge of acid or base,
etc.
) were not considered independently because previous studies
have found these costs to be insignificant relative to the other
capital and operating cost contributions.
14,15
The plant efficiency was assumed to be identical to that of
the PV-E system, 9.76%. This efficiency is consistent with a
photovoltaic component efficiency of 16%, an electrolysis and
electrochemical cell efficiency of 70% (1.75 V) and a gas
collection and processing effici
ency of 87%. A maximum practical
PEC efficiency of 25% was estimated using the product of the
maximum predicted efficiency of
the PEC cell (28.7%, radiative
recombination-limited photovoltai
cs and state-of-the-art catalysts)
38
and a gas collection and processing efficiency of 87%.
Type 4 system.
Table 5 lists the system specific technical
parameters and capital expenses assumed for the Type 4 base-
case system design. A medium-range predicted cost for a high-
efficiency tandem-junction photovoltaic cell was assumed
at $5.8 W
p

1
, commensurate with state-of-the-art III–V photo-
voltaic fabrication methods at present.
39
Pt and IrO
x
catalysts
were taken to be the same for the same solar collection areal
cost as the Type 3 system; this assumption is consistent with a
situation in which a 10

increase in the catalyst thickness
offsets the 10

decrease in the area of the semiconductor.
The Type 4 chassis was assumed to be twice as thick (2 cm) as
that in the Type 3 system, to withstand the higher hydrogen
Table 5
Type 4 PEC system technical parameters and capital and oper-
ating expenses
System specific technical parameters
STH efficiency
20%
Capacity factor
0.186
PEC area (m
S
2
)
3.7

10
5
m
2
Capital expenses
Component
2014 $ m
S

2
Window (AR coated glass)
27
0.5
Chassis (polypropylene)
28,35
6.6
Semiconductors (InGaP/GaAs)
39
175
Catalyst (Pt, IrO
x
)
26
8
Membrane (Nafion, 5 mil)
30
5
Tracker hardware
40
44.8
Concentrators (parabolic)
41
48
Compressor (1 stage)
14
14.5
Water condenser
14
0.2
Heat exchanger
14
0.4
Piping (PVC)
14
1.6
Control systems
14
8.9
Panel mounting materials
21,22
29
Installation labor
21,22
29
Other BOS
21,22
59
Energy & Environmental Science
Paper
Published on 26 May 2016. Downloaded by California Institute of Technology on 30/09/2016 16:28:35.
View Article Online