You are on page 1of 13

Applied Energy 190 (2017) 326–338

Contents lists available at ScienceDirect

Applied Energy
journal homepage: www.elsevier.com/locate/apenergy

Life Cycle Assessment of Power-to-Gas: Approaches, system variations


and their environmental implications
Xiaojin Zhang ⇑, Christian Bauer, Christopher L. Mutel, Kathrin Volkart
Laboratory for Energy Systems Analysis, Paul Scherrer Institute, CH-5232 Villigen PSI, Switzerland

h i g h l i g h t s g r a p h i c a l a b s t r a c t

 Complete Life Cycle Assessment of


Power-to-Gas with system variations.
 System expansion provides more
meaningful results than subdivision.
 For Power-to-Hydrogen, the type of
electricity supply is essential.
 For Power-to-Methane, source of CO2
and the approach applied can make a
difference.
 Sensitivity analysis and impacts other
than climate change were explored.

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

Article history: Electricity generation from intermittent renewable sources is expected to increase rapidly in the next
Received 25 August 2016 decades. Integrating renewable energy requires energy storage during low-demand periods and potential
Received in revised form 21 December 2016 conversion of surplus electricity into other energy carriers. Power-to-Gas (P2G) is a promising technology
Accepted 22 December 2016
due to its potential to provide large-scale and long-term energy storage. However, this technology has
Available online 6 January 2017
many system variations, and their environmental performances need to be evaluated and compared with
conventional technologies before large-scale deployment. In this paper, we investigate the environmental
Keywords:
performance of P2G using Life Cycle Assessment (LCA), and mainly focus on the following three aspects:
Life Cycle Assessment (LCA)
Power-to-Gas (P2G)
(1) discussion of differences as consequence of the approach applied for CO2 Capture and Utilization
Energy storage (CCU); (2) evaluation of technology variations including supply of electricity, alternative system pro-
CO2 Capture and Utilization (CCU) cesses (electrolysis technologies and CO2 sources), product gases (hydrogen and methane), and compar-
Hydrogen ison of these P2G systems with conventional technologies, and (3) investigation of further environmental
Synthetic Natural Gas (SNG) impacts of P2G in addition to the impact of global warming potential. We argue that in case of P2M, sys-
tem expansion provides more meaningful results than subdivision for CCU, since it reflects the added
value of CO2 utilization providing electricity or cement with low GHG intensity. The results of system
variations show that P2G can, depending on electricity supply and CO2 source, reduce GHG emission
compared to conventional gas production technologies, and that P2H has higher potential of emission

Abbreviations: CCU, CO2 Capture and Utilization; CGR, coal gasification and reforming; ENTSO-E, European Network of Transmission System Operators for Electricity; ESI,
Energy System Integration; GWP, global warming potential; ILCD, International Reference Life Cycle Data System; LCA, Life Cycle Assessment; LCI, Life Cycle Inventory; LCIA,
Life Cycle Impact Assessment; P2G, Power-to-Gas; P2H, Power-to-Hydrogen; P2M, Power-to-Methane; SNG, Synthetic Natural Gas; SMR, steam methane reforming.
⇑ Corresponding author.
E-mail address: xiaojin.zhang@psi.ch (X. Zhang).

http://dx.doi.org/10.1016/j.apenergy.2016.12.098
0306-2619/Ó 2016 Elsevier Ltd. All rights reserved.
X. Zhang et al. / Applied Energy 190 (2017) 326–338 327

reduction than P2M. Concerning other impact categories, P2H can have lower impacts than conventional
hydrogen production, while P2M most often has higher impacts than using conventional natural gas.
Ó 2016 Elsevier Ltd. All rights reserved.

1. Introduction studies are focused on P2H [25,26] and do not address P2M, mak-
ing it difficult to compare these two technological pathways under
As penetration of stochastic renewable electricity has grown consistent boundary conditions and assumptions. Second, most
rapidly in the global energy system in recent years, and is expected studies have covered limited system variations, for example by
to grow further in the future, energy storage is becoming more and focusing on biogenic CO2 sources and avoided the complexity of
more important. Renewable energy sources, such as wind and burden allocation issue [27]. Third, all these studies barely explore
solar, have great potential to decrease our dependence on fossil other environmental impacts than climate change for P2G. Last and
fuels and greenhouse gas emissions [1–4], but due to their inter- most importantly, discussion on methodological issues for P2G
mittent nature and challenges concerning grid integration [5,6], environmental assessment in the context of CCU have not been
storage facilities might be required to manage the fluctuation of appropriately addressed. Reiter and Lindorfer have included some
electricity generation, to reduce the curtailment, introduce addi- of the above-mentioned aspects, and estimated the break-even
tional installed capacity of renewable electricity generation [7], point of greenhouse gas intensity for electricity input to electroly-
and eventually facilitate the decarbonization of the electricity sec- sis compared to conventional hydrogen and methane generation
tor [8]. However, the balancing of power supply and demand will technologies [28]. However, their study didn’t consider the appli-
be required at different time durations (season, day, hour) and cation of product gas in an adequate way, resulting in an incom-
scales, resulting in different storage technologies being applied plete lifecycle, and in addition, the multi-functionality of CCU
for various applications [9]. (i.e. generation of CO2 as a co-product in addition to e.g. electricity
and cement, and supply of CO2 as a useful product and feedstock to
1.1. Why Power-to-Gas? methanation) was not taken into account; moreover, P2G infras-
tructure was not considered, which could have important contri-
It can be observed from previous studies [9,10] that many stor- bution when the associated impact of electricity is low. Von der
age technologies are able to provide short-term storage (i.e. rang- Assen et al. [29] identified several pitfalls of considering CCU in
ing from minutes to hours) with limited capacities, but when it previous LCA studies based on the example of methanol synthesis
comes to large-scale seasonal or long-term storage (i.e. up to with CO2 from a coal power plant and direct air capture of CO2, and
months), the options are limited to pumped hydro, compressed recommended a comprehensive framework for applying LCA on
air storage, and P2G [11]. In many countries, the potential of CCU. However, this framework has not yet been consistently
pumped hydro and compressed air storage is limited due to their applied for P2G.
specific requirements on topography, geological locations, and eco- This study therefore aims at filling these gaps. We evaluate and
nomic operating conditions [12]. In addition, compressed energy discuss the differences resulting from the LCA approach applied on
storage may have considerable thermal loss for long-term energy CCU, by comparing the subdivision and system expansion
storage. Thus, P2G is regarded as a promising storage technology approach to account for the multi-functionality of electricity/
to provide large storage capacity, with flexible storage durations cement production with CO2 capture. In addition, we investigate
from minutes to months [13]. P2G with a wide range of alternative system processes (electricity
supplies, electrolysis technologies, CO2 sources), various system
1.2. What is Power-to-Gas? configurations, and operating conditions. Moreover, we quantify
environmental impacts other than climate change in order to pro-
P2G starts with converting electricity to hydrogen through vide a more complete picture concerning the environmental per-
water electrolysis. Hydrogen generated can be injected to natural formance of P2G.
gas network up to a feasible volumetric percentage from 2% to
15% [14,15], or it can be used in other applications that require 2. Methodology
hydrogen as feedstock or fuel. It can also further react with carbon
dioxide to produce Synthetic Natural Gas (SNG), which contains As defined by ISO 14040 [30], LCA is a methodology that quan-
mainly methane. The generated SNG can be dehydrated, com- tifies the potential environmental impacts throughout a product’s
pressed (optional, depending on the applications), conditioned life cycle, from raw material acquisition, production, use, to end-
and eventually supplied to meet natural gas demands, such as of-life treatment, recycling and final disposal. In this study, we
being combusted in a gas boiler, supplied as fuel to power a vehi- carry out process-based, attributional LCA according to ISO
cle, or injected and stored in the existing natural gas network 14044 [31]. System boundaries, subdivision and system expansion
[11,16,17]. Advantages of P2G are its flexibility to convert electric- approaches, functional units, scenarios and Life Cycle Impact
ity to gases for storage, and its utilization of existing natural gas Assessment (LCIA) methods applied are specified below. We use
infrastructure, which reduces the capital investment. the Simapro version 8.0.4.30 [32] as LCA software and version
3.1 of the ecoinvent database as source of background LCI data
1.3. Environmental assessment of P2G: previous research and purpose [33].
of this work
2.1. System boundaries
Despite all these advantages of P2G and most discussions on its
techno-economic aspects [18–24], only limited studies have been The boundary of the foreground system we have analyzed
conducted to assess its life cycle environmental performance, includes the processes visualized in Fig. 1, with detailed descrip-
which results in several research gaps. First, the vast majority of tion of their coverage provided for each process in the Electronic
328 X. Zhang et al. / Applied Energy 190 (2017) 326–338

Supplementary Information (ESI) Table 2. Boxes with dashed lines recommendations from von der Assen et al. [29], and for the first
in Fig. 1 represent alternative process options. Due to data avail- time applies it to the P2M systems, with consideration of various
ability, only alternative process options in electrolysis and CO2 cap- CO2 sources, system configurations and multiple environmental
ture are considered in this study, while thermo-chemical impact categories. We apply and compare two approaches: subdi-
methanation is the only option considered for methanation. In vision and system expansion (visualized in Fig. 2 using the exam-
P2M, consumption of SNG in mobility is considered; in P2H, the ple of capturing CO2 from power generation using fossil fuel):
application is not specified and we set the system boundary at
the point of compressed hydrogen production. Specific applica-  Subdivision: from the perspective of power/cement plant or
tions of hydrogen, such as consumption in stationary fuel cells or power-to-gas system operators, the environmental perfor-
in mobility are out of scope of this analysis. Comparison of the mance of their own product is of primary interest. The multi-
environmental performance of P2M with conventional natural functional process of electricity/cement and feedstock CO2 pro-
gas technologies requires adjustment of system boundaries, duction needs to be subdivided into two separate production
depending on the approach used for CCU. activities, so that the environmental burdens associated with
the specific products can be quantified. When subdivision is
2.2. Dealing with CO2 capture and utilization in power-to-methane: applied in this study, all inputs and outputs at the power plant
subdivision and system expansion or cement factory are categorized based on physical causalities,
and exclusively assigned to one of the joint products: either
CO2 required for methanation in P2M can be obtained from var- electricity/cement or feedstock CO2, depending on the purpose
ious sources. In case of CO2 capture at power plants and cement of their use or their origin. For example, in case of CO2 captured
production, these processes are multi-functional due to the com- from power plants, additional SO2 emissions removal (to avoid
bined production of CO2 and electricity or cement [29]. Besides reaction with the absorbent used for CO2 capture) and a fraction
the multi-functionality (MF) in the CO2 capture process, the utiliza- of the electricity generated by the plant are required for CO2
tion of CO2 as feedstock in methanation needs to be taken into capture. Such inputs are assigned to feedstock CO2 because
account in a consistent way in LCA. Von der Assen et al. discussed these are ‘‘caused” by CO2 capture. When the environmental
different accounting approaches and system boundary issues of performance of only the P2G system needs to be evaluated, this
CO2 Capture and Utilization, using the example of methanol syn- approach is applied to avoid allocation, as recommended by ISO
thesis [29,34], and pointed out that, when joint functional units 14044-2006 (Section 4.3.4.2) [31].
for MF processes are impractical, or the product-specific environ-  System expansion: from a more general perspective of, for
mental performance has to be evaluated, allocation of environmen- example, policy regulator or society as a whole, the total envi-
tal burdens based on economic value can be applied; however, ronmental impacts of the complete and expanded system,
system expansion is recommended if joint functional units are pos- including both electricity/cement production with CO2 capture
sible, because it avoids the arbitrary choices of allocation criteria. and power-to-gas, are decisive. These impacts need to be eval-
They also objected the avoided-burden approach, as it does not uated and compared with the conventional production of equiv-
distribute environmental benefits among all functions of the alent amounts of electricity/cement without CO2 capture, as
system. In general, our analysis follows the framework and well as the conventional production and use of natural gas

Fig. 1. System boundary of Power-to-Gas with alternative process options.


X. Zhang et al. / Applied Energy 190 (2017) 326–338 329

Fig. 2. Power-to-Gas system with CO2 from electricity production with CO2 capture: product systems and system boundaries in case of subdivision and system expansion.

when P2M is not in place. System expansion does not require the source of CO2 capture, and product or service provided by the
the split of process between electricity/cement production P2G system. Electricity input is the only common dimension for
and captured feedstock CO2. Instead, electricity/cement produc- expanded systems with different CO2 sources and thus different
tion with CO2 capture and the subsequent P2M system are con- quantities of multiple products. The other reason to define the
sidered as one expanded system. This approach is also functional unit this way is that P2G is supposed to be an energy
recommended by von der Assen et al. [29]. storage technology to integrate surplus electricity from renewable
sources; the functional unit of 1 kWh of electricity input represents
Note that although the term ‘‘allocation” was often used as the most useful basis for comparing the environmental benefits
counterpart for ‘‘system expansion” in LCA [29,30], it is not used and drawbacks of storing 1 kWh of surplus electricity via different
in this study. Instead, ‘‘subdivision” is used, because we subdivided storage technologies [35].
process exchanges based on their causalities, whereas allocation is In both subdivision and system expansion approach, the com-
usually performed based on properties of the products, such as bustion of SNG has to be included for P2M, to take into account
energy contents, economic values, etc. the complete cycle of CO2 from the point that it enters the system
In this study, results based on both subdivision and system until it is emitted into the atmosphere after SNG being combusted.
expansion approaches are compared with the results of conven-
tional technology scenarios without P2M. The difference in results 2.4. Scenarios
between these two approaches is further explored using the exam-
ple of a hard coal power plant with post-combustion CO2 capture Different scenarios were developed to investigate the impact of
in Section 3.1. system configurations and technology options on the environmen-
tal performance of P2G. The system variations considered in the
2.3. Reference products and functional units scenarios include:

P2G can produce hydrogen or SNG, which can be used in differ-  Electricity supply to electrolysis
ent ways as alternatives for various products and services (Fig. 1). – Electricity from renewable energy sources (power from wind
Therefore, multiple reference products and functional units need to turbines and solar photovoltaics) in Switzerland, considering
be considered in this study: for P2H, we analyze 1 MJ of hydrogen the ranges of location-specific annual yield and lifetime
generated; for P2M where SNG is used as fuel in mobility, we ana- – Swiss grid supply (based on consumption mix in
lyze 1 km of distance travelled by a compressed natural gas (CNG) Switzerland)
fueled mid-size passenger vehicle when we apply the subdivision – European average grid supply (ENTSO-E mix)
approach for CCU. In these cases, reference products are the same  Electrolysis:
as the functional units. – Electrolysis technology: Alkaline electrolysis and Polymer
The functional unit of 1 kWh of electricity input to electrolysis Electrolyte Membrane (PEM) electrolysis.
is chosen when the system expansion approach is applied in order  CO2 capture technology and CO2 sources: CO2 capture consid-
to account for the multiple reference products of the expanded sys- ered in this study is based on previous work of one of the co-
tem (as shown in Fig. 2), namely electricity or cement produced at authors [36] and includes CO2 captured from power plants with
330 X. Zhang et al. / Applied Energy 190 (2017) 326–338

different fuels via post-combustion capture technology, as well potential environmental impacts are covered by considering
as from cement plant with different energy supplies. In addi- ranges of the following parameters (Table 1):
tion, direct atmospheric capture with waste heat utilization is
considered.
2.6.2. Electrolysis
Electrolysis that has been considered in this study represents
These P2G systems are compared to conventional technologies
low-temperature electrolysis (PEM and alkaline electrolysis)
of hydrogen and natural gas production and use. The conventional
occurring in the range of less than 100 °C under the pressure of less
hydrogen production technologies considered are steam methane
than 30 bars. The energy consumption (in kWh) per unit volume
reforming (SMR) and coal gasification and reforming (CGR) [37].
(1 Nm3) of hydrogen varies depending on the system configura-
In P2M, when subdivision for CCU is applied, the environmental
tions (e.g. electrolysis technology, system operations, etc.), and is
burdens of P2M are compared with the conventional supply and
scaled based on different system sizes for PEM electrolyzers. Scal-
use of natural gas for mobility in Switzerland [33]. When system
ing in alkaline electrolysis is not considered in this study due to
expansion for CCU is applied, the environmental burdens of the
limited data availability.
expanded system are compared with a reference system including
For alkaline electrolysis, the LCI data for hydrogen generation
both production of cement/electricity without CO2 capture [36]
are based on Chapter 2.3.7 in Wokaun et al. [37] with slight mod-
and conventional natural gas supply and use.
ification on unit energy consumption per Nm3 of hydrogen gener-
Further analysis is performed to understand the sensitivity of
ation, for consistency of data source used in alkaline and PEM
system parameters, such as lifetime, system sizes, electrolysis
electrolysis, representing a facility around 300 kW in power rating,
operation point and system electricity consumption, which can
which can generate up to 60 Nm3 of hydrogen per hour. Original
be found in the ESI Section 6.
data were also modified to exclude high pressure compression
and storage, as well as hydrogen dispensing in all the P2M scenar-
2.5. Life Cycle Impact Assessment (LCIA)
ios, as hydrogen produced is used in the methanation process
downstream without being compressed to high pressure. Based
The Institute for Environment and Sustainability in the Center
on Table 2 in Grond et al. [16], 5.2 and 4.9 kWh of electricity is con-
of European Union Joint Research performed a review study on a
sumed on average per Nm3 of hydrogen generation from alkaline
range of midpoint LCIA methods for different impact categories
and PEM electrolysis system respectively. Water consumption is
in 2011, under the framework of International Reference Life Cycle
assumed to be constant at 1.1 kg per Nm3 of hydrogen production
Data System (ILCD), and it provided recommendations on global
in all scenarios, which is based on the foreground data collected
models and characterization factors that can be used for the impact
[42].
assessment in LCA [38]. The resulting ‘‘ILCD 2011 Midpoints” as
Material inputs used for BOP of PEM electrolyzer are derived
implemented in Simapro 8.0.4.30 [32] were used in this study. A
from PEM fuel cell data in ecoinvent version 3.1 [33], whereas
table listing the impact categories included in this study, the rec-
the stack material inputs are based on expert judgement [43]. Scal-
ommended impact assessment methods by ILCD, and the unit of
ing of electrolyzer is performed based on the approach by Remer
characterization factors can be found in the ESI Section 4, Table 4.
et al. [44]; more detailed equations and assumptions can be found
in the ESI Section 1.
2.6. Life Cycle Inventory (LCI) inputs and assumptions
In the P2H scenarios, the hydrogen generated is compressed
before supply. When hydrogen is used as a vehicle fuel, the pres-
The data used in this study includes both background data from
sure after compression is usually at 350 bar or 700 bar. Based on
ecoinvent version 3.1 [33], and foreground data collected by con-
the DOE technology validation project by Gardiner [45], the aver-
sultation with the field experts, project partners and from
age energy consumption for compression of hydrogen is assumed
literature.
to be 3.1 kWh/kg, which is the average energy demand of hydrogen
compression to 350 bar (1.7 kWh/kg) and to 700 bar (6.4 kWh/kg).
2.6.1. Electricity generation
Data for electricity generation used as supply for electrolysis are
all based on ecoinvent version 3.1 [33]. Four types of electricity 2.6.3. CO2 Capture and Utilization (CCU)
supply are used in the scenarios, namely electricity from wind tur- The atmosphere, cement and coal power production are consid-
bines, photovoltaics, Swiss (CH) grid, and average grid of European ered as potential CO2 sources in this study. Transportation of CO2
Network of Transmission System Operators for Electricity (ENTSO- from the sources to P2G system is not considered, due to its high
E). Since the analysis focuses on the Swiss context, renewable variability in real operation. Instead, we assume that P2G is oper-
energy generated from wind and photovoltaics used in this study ated in proximity of power or cement plants, which will most
correspond to the resources and performances in Switzerland. likely be beneficial from an economic perspective.
The environmental impact of electricity generation with wind tur- Data for CO2 captured from the atmosphere is based on the tech-
bines and PV will differ depending on the annual electricity gener- nology of Climeworks [46,47]. Since the specification of the sorbent
ation affected by location-specific factors (such as wind condition, is not disclosed, generic organic chemicals are used in the LCI, which
solar irradiance) and technology parameters, as well as the lifetime requires future refinement when more information is available. It is
of the turbines and PV panels. Assuming that the same technolo- assumed that the capture process operates at 95 °C, and to capture
gies are applied in different locations of Switzerland, ranges of 1 kg of CO2, 0.875 kWh of waste heat from electrolysis and methana-

Table 1
Performance ranges of solar PV and wind electricity production in Switzerland.

Renewable electricity type Parameters Lower value Reference average value according to ecoinvent [33] Higher value
Electricity from solar PV Annual yield (kWh/kWp year) 850 [39] 922 1500 [39]
Lifetime (years) 20 [40] 30 30 [40]
Electricity from wind turbines Full-load hour (h/year) 1000 [28] 1675 2600 [28]
Lifetime (years) 10 [41] 20 30 [41]
X. Zhang et al. / Applied Energy 190 (2017) 326–338 331

tion can be utilized, which results in reduced electricity demand of However, based on sustainable forestry in Switzerland (i.e. har-
0.37 kWh per kg CO2 captured. vesting only at the natural growth rate), this is supposed to result
Data for wood fuel production is from ecoinvent [33], which in negligible net CO2 emissions, as opposed to other climatic
uses a simplified model, and does not consider, for example, regions and clear-cutting [48].
potential carbon emissions resulting from the change of land-use.

CO2 capture from atmosphere

electricity for capture facility,


CO2 from CO2 capture material and energy CO2 emissions
(a) atmosphere consumpon

CO2 Power-to- Fuel


Capture Methane Combuson
CO2 CO2 CH4
captured

CO2 captured from cement producon


clinker
produced
electricity for CO2 emissions
capture facility,
CO2 CO2 capture material and energy
(b) release consumpon

Clay and CO2 Power-to- Fuel


carbon Lime Stone
Clinker Producon
Capture Methane Combuson
CO2 CO2 CH4
emissions captured

CO2 captured from biomass-based fuel power producon

electricity sent
electricity for CO2
to the grid
capture
CO2 from electricity for capture facility,
(c)
atmosphere other demand material and energy CO2 emissions
CO2 at plant consumpon
release
Electricity CO2 Power-to- Fuel
Biomass
Producon Capture Methane Combuson
CO2 CO2 CH4
emissions captured

CO2 captured from fossil fuel power producon


electricity sent to
the grid electricity for CO2
capture

electricity for capture facility, CO2 emissions


(d) other demand material and energy
CO2 at plant consumpon
release
Fossil Electricity CO2 Power-to- Fuel
carbon Fuel Producon Capture Methane Combuson
CO2 CO2 CH4
emissions captured

indirect CO2 from fossil or underground sources indirect CO2 from biogenic sources
direct CO2 from fossil or underground sources direct CO2 from biogenic sources or atmosphere
electricity facility, material and other energy consumpon

Fig. 3. System boundaries and carbon flows for CO2 capture from (a) atmosphere, (b) cement production, (c) power generation from biomass, (d) power generation from fossil
fuel (coal or gas).
332 X. Zhang et al. / Applied Energy 190 (2017) 326–338

g of CO2 eq /MJ of Hydrogen from Electrolysis


300 Others

with Different Electricity Supply


Electricity
250

200

150

100

50

0
PEM Alkaline PEM Alkaline PEM Alkaline PEM Alkaline Steam Coal
Electrolysis Electrolysis Electrolysis Electrolysis Electrolysis Electrolysis Electrolysis Electrolysis Methane Gasificaon
(Wind) (Wind) (PV) (PV) (CH Mix) (CH Mix) (ENTSO-E (ENTSO-E Reforming Reforming
Supply) Supply)
P2H Convenonal

Fig. 4. Life cycle GHG emissions of hydrogen production: Power-to-Hydrogen vs. conventional hydrogen production (100 kW electrolyzer; based on lower heating value
(LHV) of hydrogen of 10.8 MJ/Nm3 [52].) category ‘‘others” include contributions from inputs other than electricity supply to electrolysis, such as facilities, water, and
chemicals required for electrolysis; error bars represent variation in wind and PV power generation performance.

Fig. 3 shows that CO2 captured from biomass combustion or the lower CO2 emissions due to capture. The captured CO2 – further
atmosphere is balanced with the CO2 emitted back into the atmo- utilized in P2M – is no longer regarded as elementary flow, but
sphere during SNG combustion. The climate change impact assess- as feedstock and economic product flow in methanation [29]. This
ment method ILCD recommends is based on IPCC 2007 [49], in means all the burdens (e.g., energy and material required for cap-
which CO2 taken from the atmosphere and biogenic emissions ture) and benefits (e.g., negative emissions of NOx and SO2) result-
are considered with zero global warming potential (GWP). ing from CO2 capture are accounted for and assigned to feedstock
The LCI for CO2 capture from power plants and cement plants CO2.
are based on Volkart et al. [36], using the technology data from
the 2025 scenario, with extra emission removal of SO2 and NOx
required for the capture considered. Both the burdens and the 2.6.4. Methanation
resulting emission reductions in SO2 and NOx are accounted in Thermo-chemical methanation is assumed to be carried out
the CO2 capture process. CO2 cleaning processes such as N2 and between 300 °C and 550 °C; hydrogen produced from electrolysis
H2S removal might be required for methanation, but are not con- and CO2 captured from other processes are supplied at 20 °C to
sidered due to lack of data. In terms of capture technology, post- 25 °C. The heat required for the input hydrogen and CO2 is
combustion capture is considered, which separates CO2 from nitro- assumed to be met by the waste heat released from methanation.
gen in the flue gas after the fuel is combusted. There are other CO2 Energy required for compression of gas inflows to supplement the
capture technologies available from the same references, namely pressure loss in this reactor, as well as for the cooling of the reac-
oxy-fuel combustion capture and pre-combustion capture [36]. tor, is not considered in this study due to insufficient data. The effi-
For simplification, only results with CO2 from post-combustion ciency of methanation is assumed to be around 80%, and an
capture are discussed in detail, while the results with the other average demand of 9.3E-6 kg of Ni-based catalyst is needed per
two capture technologies are included in the ESI. The data for Nm3 of SNG generation according to Table 4 in Grond et al. [16],
CO2 capture from cement production refers to a cement plant in which corresponds to a catalyst lifetime of 2–3 years. The Ni-
Switzerland with inventory data based on Kellenberger et al. based catalyst consists of 19% nickel and 81% aluminum alloy by
[50], which represents the state-of-art cement production in Eur- mass. The energy requirement and facility needed in catalyst man-
ope. In post-combustion capture, according to Volkart [36], the ufacturing, as well as the treatment and disposal of catalyst mate-
emissions of SO2, NOx and PM are assumed to be further reduced rials are not considered in this study due to limited data. The
by 95%, 2.5% and 50%, respectively, due to CO2 capture, in both methanation reactor is assumed to have a lifetime of 20 years
power plant and cement plant. [16], and it is assumed to be made of 300 kg stainless steel (corre-
In Fig. 3, system processes associated with the CO2 flows out- sponding to a system with three-step methanation that could gen-
side and within the P2G system are illustrated for different CO2 erate 1 Nm3 of SNG per hour). The product gas from methanation is
sources. It shows that when CO2 is captured from fossil fuel com- SNG, and is assumed to have equivalent composition and energy
bustion and cement production, the CO2 emitted during SNG com- content as conventional natural gas.
bustion originates from fossil fuel or other underground sources Depending on the application, the produced SNG is injected into
(e.g., limestone in case of cement production), and these emissions the natural gas network (distribution network up to 5 bars) with-
are not compensated upstream in terms of global warming poten- out any compression, or is dehydrated and compressed to 200 bars
tial, as it is for CO2 from biogenic origin. CO2 capture at cement and before it is supplied for mobility purpose. Dehydration of gas is
fossil fuel power production partially shifts CO2 emissions from assumed to be carried out by two-phase heat exchange from
cement and power plants to the P2G system. 285 °C to 12 °C and from 12 °C to 5 °C. Energy demand for product
Note that when CO2 is captured from power plants with fossil gas compression from 10 bar to 200 bar is assumed to be 0.39 kWh
and biomass fuel, the additional electricity consumed for the cap- per Nm3 of SNG produced based on Edwards et al. [51]. In P2M for
ture process (modeled as internal consumption) is associated with mobility, it is assumed that one km travelled requires 0.077 Nm3 of
X. Zhang et al. / Applied Energy 190 (2017) 326–338 333

Electrolysis (with PV supply) CO2 Capture


Methanaon and Fuel Processing Conv. Natural Gas Producon
Fuel Combuson SNG Dispense
Road Vehicle Maintenance
Vehicle Construcon Life Cycle Emissions
500
450
g of CO2 eq/km travelled

400 390 392 385


350
125 125 125
300
282
250 7 9 3 33
249 245
200 125 125
150
8 4
100
50
0
atmosphere wood power natural gas power hard coal power Cement plant Convenonal CH
capture, with plant plant plant with EU grid and Natural Gas
waste heat waste heat supply Supply

Fig. 5. Life cycle GHG emissions of Power-to-Methane for mobility: Power-to-Methane vs. conventional Swiss natural gas supply using the subdivision approach for CCU.
Swiss natural gas supply and vehicle according to ecoinvent version 3.1 [33]; electricity for electrolysis (100 kW PEM electrolyzer) from PV in Switzerland; error bars
represent variation in PV power generation performance.

SNG according to the fuel consumption of a medium-size passen- cycle GHG emissions due to its higher average efficiency. Using
ger car consuming natural gas [33]. electricity with very low GHG intensity reduces life cycle GHG
emissions by more than 90% compared to the conventional produc-
3. Results and discussion tion of hydrogen from fossil fuel. Using the average power supply
from the Swiss grid (consumption mix, to a large extent supplied
The climate change impact of P2G is investigated in Sections 3.1 by hydro and nuclear) still reduces life cycle GHG emissions by a
and 3.2, while the other impact categories according to ILCD 2011 factor of about two compared with conventional production,
are discussed in Section 3.3. Results based on subdivision (Sec- whereas electrolysis powered by the average European grid mix
tion 3.1.1) and system expansion (Section 3.1.2; applicable only (with a high share of fossil fuels) results in higher emissions.
to P2M) are presented and compared in Section 3.1.3. When sys- Similarly, life cycle GHG emissions of P2M for mobility using
tem expansion is applied, the amounts of reference products per the subdivision approach are shown in Fig. 5. In this figure, elec-
functional unit are always listed in tables above the results in the trolysis technology and the electricity supplied to electrolysis
figures. Within the P2G system, the climate change impacts of (from photovoltaics) are the same for all scenarios, and only CO2
P2H and P2M are compared (Section 3.2). Sensitivity analyses on sources are varied. CO2 is generated using post-combustion cap-
the impact of important system parameters on life cycle GHG ture at power plants and cement plants (complete numerical
emissions are performed (ESI Section 6), including electrolysis sys- results for all technologies are included in the ESI). Using CO2 cap-
tem electricity demand, scale of electrolyzer, operation point of tured from the wood power plant and from the atmosphere results
electrolyzer (load density), and electrolyzer lifetime. in the lowest life cycle GHG emissions. Although CO2 capture from
atmosphere is relatively energy intensive [53], due to its assumed
3.1. Subdivision versus system expansion utilization of low-temperature waste heat and the origin of CO2
from the atmosphere, the associated life cycle GHG emissions are
3.1.1. Subdivision: Power-to-Gas vs. conventional gas production only slightly higher than those of the system with CO2 captured
Life cycle GHG emissions of P2H and conventional hydrogen from a wood power plant. For non-biogenic CO2 sources, using
production are shown in Fig. 4. Alkaline and PEM electrolysis are SNG from P2G with electrolysis supplied by PV power can generate
considered and compared to hydrogen generation from SMR of higher emissions than using conventional natural gas as vehicle
natural gas as well as CGR. The variability of renewable electricity fuel, given the electrolysis performance assumed. For biogenic
generation from wind and solar PV in Switzerland, and the result- CO2 sources, whether (zero impact) biogenic GHG emissions from
ing range on system life cycle GHG emissions is represented by the fuel combustion are able to compensate the emissions difference
error bars. It shows that PV exhibits a higher variability than wind between conventional natural gas production and fuel production
in absolute terms. Electrolysis supplied by wind power can achieve from P2G, determines the performance of P2M for mobility com-
a higher reduction of emissions because of the lower GHG emis- pared to vehicles fueled by conventional natural gas.
sions per kWh from wind (ranging from 8 to 62 g CO2 eq/kWh,
with a reference value of 19 g CO2 eq/kWh) than PV (ranging from 3.1.2. System expansion: Power-to-Methane with different sources of
50 to 132 g CO2 eq/kWh, with a reference value of 81 g CO2 eq/ CO2 vs. conventional gas production
kWh). The contribution of electricity input to electrolysis is domi- The results above based on subdivision for CCU represent the
nant concerning climate change impacts of the P2G system. How- emissions of the P2G system only and do not reflect the CO2 emis-
ever, its contribution is comparable to electrolysis facility, if the sion mitigation achieved by CO2 capture at the power/cement
GHG emissions of input electricity are lower than about 20 g of plant. In order to have a fair comparison between P2G and the con-
CO2 eq/kWh. With the same electricity supply, PEM electrolysis ventional technologies, the system needs to be expanded to
performs slightly better than alkaline electrolysis in terms of life include both P2G system and electricity/cement production
334 X. Zhang et al. / Applied Energy 190 (2017) 326–338

(Co-)products per kWh of electricity input to electrolysis


664m distance travelled
0.135kWh 0.278kWh 0.056 kWh
n.a 0.122kg cement
electricity electricity electricity
Mobility and Convenonal Scenario, in g of CO2 eq/kWh
Emission Reducon Comparing Scenario with P2M for
Detailed breakdown by process and comparison with
120 results from subdivision approach is illustrated in Figure 7
100
87.8 88.3
of Electricity Input to Electrolysis

80 84.6
78.3
73.5
60

40 41.6

20 21.8 25.0 25.5


15.5 10.7
0
-20 -21.2
-40

-60

-80
-100
Wind PV Wind PV Wind PV Wind PV Wind PV Wind PV
Atmosphere Cement plant w Cement plant w Hard coal power Natural gas Wood power
Capture grid & waste heat grid supply from plant power plant plant
supply hard coal

Fig. 6. Life Cycle GHG emission reduction comparing P2M for mobility with CO2 captured from different sources and conventional scenarios using system expansion
approach; functional unit: 1 kWh electricity input to electrolysis; electricity for PEM electrolysis in P2G is supplied by PV or wind in Switzerland. Conventional technologies
represent scenarios without P2M, i.e. distance travelled with a conventional natural gas and electricity/cement production without CO2 capture.

(Fig. 2) with its reduced GHG intensity. System expansion allows Taking into account the potential variability of PV and wind power
for comparing the environmental performance of P2G with con- in terms of their GHG intensities, the advantage of wind power
ventional energy carriers from a societal perspective. supply over PV is extended from 63 g CO2-eq, to a range between
Life cycle GHG emission reductions due to P2G in comparison 11 and 124 g CO2-eq emission reduction per kWh of electricity
with natural gas vehicles and electricity/cement production are input to P2G.
shown in Fig. 6. For comparison, only source of CO2 for methana- Fig. 6 also shows that, even by considering the worst wind
tion is varied, with the same electrolysis and methanation pro- power performance, system GHG reduction can always be
cesses in all scenarios (complete numerical results for CO2 achieved, which is not the case when the electrolyzer is supplied
sources and other impacts are included in the ESI Table 6–9). Elec- by electricity from PV.
tricity supplies from wind and solar PV are considered with possi- P2G with CO2 captured from a wood power plant can achieve
ble ranges of associated GHG emissions. Since system expansion is the highest reduction of GHG emissions; a cement plant using
applied, the functional unit is 1 kWh of electricity input to electrol- waste heat for CO2 capture and direct air capture of CO2 achieve
ysis, and there are two joint products from the expanded system: almost the same emission reductions. CO2 provided by cement
the co-product at the CO2 source (i.e. cement or electricity) and plant with heat for CO2 capture generated from hard coal combus-
distance travelled by the SNG vehicle. The amounts of H2, SNG tion results in the lowest potential reduction (actually, considering
and vehicle distance travelled corresponding to the functional unit the emission range of PV electricity supply, it could even increase
are equal for all scenarios independently of the source of CO2. the system emissions). Both CO2 captured from a wood power
However, due to the difference in capture technology and charac- plant and direct air capture are preferred sources of CO2 because
teristics of CO2 sources, the amounts of co-products are different of the biogenic origin of carbon; however, direct air capture must
in the scenarios (ESI Table 2). This has the following implication rely on clean electricity supply and utilization of waste heat in
in the interpretation of results: (1) comparison based on the abso- order to achieve high GHG emission reductions. In general, good
lute emissions of P2M systems with different CO2 sources are not CO2 sources provide comparatively high reductions of GHG emis-
meaningful, as they have different co-products; (2) The preferred sion at the source of capture, being determined by both the emis-
source of CO2 is determined by the highest absolute emission sion reduction per unit of co-product, and the amount of co-
reduction achieved, comparing the expanded system with electric- product produced.
ity or cement production without CO2 capture and distance trav- Comparing the variation of GHG emission reduction caused by
elled by the CNG vehicle powered by conventional natural gas. different CO2 sources and electricity supplies shows that – without
Reduction of overall system GHG emissions mainly depends on considering the uncertainties within each individual CO2 capturing
the electricity supplied to electrolysis and the source of CO2. P2G process – electricity supply causes higher variation on system GHG
with electrolysis using wind power can achieve higher life cycle emissions than different CO2 sources.
GHG emission reduction than with supply from PV. Based on the In the best case (both in terms of CO2 source and electricity sup-
reference performance of Swiss PV and wind (without considering ply, being wood combustion and high performance wind power),
variability), the reduction with wind power is 63 g CO2-eq higher system GHG emissions can be reduced by 99 g CO2-eq/kWh elec-
than the reduction with PV supply per kWh of electricity input. tricity input compared to the conventional scenario. In the worst
X. Zhang et al. / Applied Energy 190 (2017) 326–338 335

case (CO2 supply from a cement plant with heat for CO2 capture from P2M, used as fuel in a passenger vehicle, are higher than
from hard coal combustion, and with worst performance PV those of conventional natural gas. System expansion instead takes
power), system GHG emissions increase by 72 g CO2-eq/kWh elec- into account the mitigation by CO2 capture in electricity produc-
tricity input compared to the conventional scenario. tion, resulting in lower overall emissions for the system including
both P2M and electricity production. The comparison between
3.1.3. Comparison of system expansion and subdivision subdivision and system expansion for CCU in P2G shows that only
Although both system expansion and subdivision have shown system expansion provides meaningful results from an overall per-
biomass combustion with CO2 capture as the preferred CO2 source spective, since subdivision does not reflect the reduced emissions
among all the options, the ranking of the technology options and of electricity (or cement) production with CO2 capture.
the conclusions for some of them are quite different when different
approaches are applied. As opposed to subdivision, applying sys-
tem expansion shows that P2M systems with CO2 from cement, 3.2. Power-to-Hydrogen vs. Power-to-Methane
coal and natural gas power plants can reduce overall GHG emis-
sions. In addition, the system expansion results shows that using When excess electricity is generated by intermittent renewable
CO2 from a cement plant can achieve slightly higher emission sources and P2G is chosen as a storage (or conversion) technology,
reduction than direct air capture of CO2 with waste heat the environmental performances of P2H vs. P2M need to be evalu-
utilization. ated. Fig. 8 shows the potential reduction of GHG emissions, if elec-
In order to further investigate the differences between subdivi- tricity from wind power or PV is fed into P2G system, and H2 from
sion and system expansion, a more detailed breakdown of life cycle P2H replaces conventional hydrogen production from fossil fuels,
GHG emissions by process is shown using CO2 from a hard coal while SNG from P2M replaces natural gas as vehicle fuel. When
power plant with post-combustion CO2 capture (Fig. 7). The func- P2H is powered by wind or PV, the GHG emissions of hydrogen
tional unit is 1 kWh of electricity input to electrolysis, supplied by production are much lower than those from conventional tech-
solar PV or wind power. In case of subdivision, GHG emissions nologies using fossil fuel reforming. Previous studies indicated that
associated to CO2 capture is shown, whereas in system expansion the storage and transportation of hydrogen is a challenge [54,55],
approach, GHG emissions associated with CO2 capture is not sepa- but this can be avoided if hydrogen is produced at the fueling sta-
rated, and included as part of joint production of electricity and tion. The potential emission mitigation P2M could bring is much
CO2 (in the category of electricity production w/wo CO2 capture lower. It is known from Fig. 6 that, by applying system expansion
in the figure). The GHG emissions of SNG compression, transport, and thus considering the emission mitigation at the source of
supply and combustion are equal to those of conventional natural CO2 supply, the scenario with the highest life cycle GHG reduction
gas, and the GHG emissions associated with vehicle and road achieved is the system with CO2 captured from a wood power
infrastructure are equal in all scenarios. Differences in GHG emis- plant with wind electricity supply (Fig. 6). However, even with this
sions between scenarios are mainly from the emission difference best-performing P2M system, when comparing it to vehicle dis-
of fuel production and joint accounting of electricity production tance driven by consuming conventional natural gas, the climate
and CO2 capture. change mitigation is less than P2H per 1 kWh of electricity input
In case of subdividing burdens between electricity production to electrolysis. P2H can also always bring emission mitigation in
and captured feedstock CO2, the life cycle GHG emissions of SNG compared to conventional hydrogen production, even when the

(Co-)products per kWh of electricity input to electrolysis


664m distance travelled
0.135kWh electricity
g CO2 eq/kWh of Electricity Input to Electrolysis

Electrolysis CO2 Capture


Methanaon and Fuel Processing SNG/NG Compression
Conv. Natural Gas Producon Electricity Producon w/wo CO2 capture
400 Fuel Combuson SNG Dispense
Road Vehicle Maintenance
350
300 corresponding emission 21
250 reducons in Figure 6 83
83 20 114
200 20
6 6 21 32 32
150
100
50
0
Wind, CO2 from PV, CO2 from Convenonal Wind, CO2 PV, CO2 from Convenonal
post-combuson post-combuson natural gas captured from post-combuson natural gas
of hard coal of hard coal supply post-combuson of hard coal supply & Hard
power plant power plant of hard coal power plant & Coal Power Plant
power plant & Electricity Electricity
Electricity producon with Producon
producon with CO2 capture without CO2
CO2 capture capture
Subdivision (P2G) System Expansion (P2G +Electricity Producon with
CO2 capture)

Fig. 7. Power-to-Methane for Mobility – Subdivision vs. System Expansion: life cycle GHG Emissions of P2G using subdivision (left) and system expansion (right); CO2
captured from a hard coal power plant by post-combustion capture. Electrolysis in P2G is supplied by PV or wind power; error bars indicate variability of GHG intensities of
electricity from PV and wind in Switzerland. SNG (in P2G scenarios) and NG (in conventional technology scenarios) are used for vehicle operation.
336 X. Zhang et al. / Applied Energy 190 (2017) 326–338

Emission Reducon Comparing Scenario with P2G


and Convenonal Scenario, g of CO2 eq/kWh of
(Co-)products per kWh of electricity input to electrolysis
Product: 0.204 Nm3 hydrogen Products: 664m distance
450

Electricity Input to Electrolysis


0.056 kWh electricity
400
350
300
250
200
150
100
50
0
-50 Compared to Compared to Compare to Compared to CO2 from wood CO2 from wood
SMR CGR SMR CGR power plant power plant
Compared to Compared to
Convenonal Convenonal
P2H with PV Supply P2H with Wind Supply P2M for P2M for
Mobility with Mobility with
PV Supply Wind Supply

Fig. 8. Life cycle GHG emission reduction comparing P2H and P2M for mobility with conventional scenarios (fossil fuel pathways): with PV and wind electricity supply to
100 kW PEM electrolyzer; error bars indicate variability of GHG intensity of electricity from PV and wind in Switzerland.

emissions of renewable electricity production are not ideal, which non-cancer”, ‘‘freshwater eco-toxicity” and ‘‘resource depletion”
is not the case of P2M. (mineral, fossil and renewable). In all these four impact categories,
P2M for mobility scenarios perform worse than the scenarios of
expanded systems with conventional technologies (natural gas
3.3. LCIA – other impact categories supply and cement/electricity production without CO2 capture),
even when considering performance variability and different types
The LCIA characterization results using ILCD midpoint indica- of renewable electricity production that feeds into electrolysis (ESI
tors of selected scenarios in P2H and P2M for mobility are shown Table 6–9). This is because methanation reduces the efficiency of
in ESI Table 5–9, while the normalized results are shown in ESI the system and the capture of CO2 requires additional energy and
Table 10–12. Since normalized impacts help to quantitatively materials. When electricity supply to electrolysis is switched from
relate absolute impact with the annual impacts of an average per- PV to wind power (ESI Table 11 vs. ESI Table 12), the differences in
son, we refer to it to first narrow down the discussion to the rela- most impact categories between P2M and conventional scenarios
tively important impact categories, and then to the other impact are reduced, but P2M still has higher impacts.
categories. The LCIA results consider high and low performances In general, the LCIA results of other impact categories besides
of renewable electricity production, based on the assumptions in climate change show that P2H may cause lower impacts than con-
Section 2.6.1. ventional technologies, while P2M most often does not perform
In P2H, normalized results (ESI Table 10) are substantially better than conventional technologies. However, it should also be
higher in two impact categories than for the other impact cate- considered that the methodological uncertainties in water
gories: ‘‘human toxicity cancer” and ‘‘freshwater eco-toxicity”. resource depletion, freshwater toxicity and human toxicity are rel-
Since steam methane reforming (SMR) is the dominant technology atively high in the ILCD 2011 mid-point indicators, according to
for conventional hydrogen production, we discuss P2H vs. H2 from Table 1 in Hausschild et al. [38].
SMR. The results in ESI Table 5 show that except for P2H with PEM
electrolyzer and best performance of renewable electricity produc-
3.4. Sensitivity analysis
tion, hydrogen production from SMR has lower human toxicity
cancer impacts than all the other P2H scenarios. This is because
Results of a sensitivity analysis are provided in the Electronic
of the relatively low human toxicity cancer impact associated with
supplementary information (ESI 6) and shows that the electricity
natural gas supply compared to P2H, in which relatively high
consumption for electrolysis is the most sensitive parameter con-
impacts are associated with material consumption such as steel
cerning life cycle GHG emissions as well as for most other impact
and aluminum in the infrastructure of wind and PV power plants,
categories. This is followed by the system operation point (density
and the consumption of copper and fossil fuel when electricity is
on a unit area (W/cm2)), while the sensitivity of electrolyzer size
supplied by the Swiss consumption mix. Concerning freshwater
and lifetime are much less influential.
eco-toxicity, conventional production of hydrogen has lower
impacts than all the P2H scenarios, mainly due to the high fresh-
water eco-toxicity impact of copper mining: consumption of cop- 4. Conclusions and outlook
per in the infrastructure for electricity supply needed by
electrolysis is higher than in the natural gas supply chain. How- By considering different P2G technology and system variations,
ever, in the other impact categories, P2H may have lower impacts our analysis provides quantitative reference to support the tech-
than SMR. nology selection for the system design of P2G from the environ-
Among the normalized LCIA results of P2M using the system mental perspective. The discussion of CCU related LCA
expansion approach (ESI Table 11, ESI Table 12), results are rela- methodology – subdivision vs. system expansion – and its applica-
tively high in four impact categories indicating comparatively tion to P2G reveals its complexity and also shortcomings of previ-
higher importance: ‘‘human toxicity cancer” and ‘‘human toxicity ous studies and provides a benchmark for further analysis.
X. Zhang et al. / Applied Energy 190 (2017) 326–338 337

We compared the environmental performance of P2G with con- Electricity (SCCER-Hae) from 2014 to 2016 and supported by the
ventional hydrogen and natural gas production and utilization as Energy System Integration (ESI) platform of PSI. We acknowledge
vehicle fuel, applying both subdivision and system expansion for the funding provided by the SCCER-Hae and the ESI platform,
joint electricity/cement and feedstock CO2 production in case of and would like to thank our colleagues Felix Büchi, Tilman Schild-
P2M. We found that within the P2G system, the type of electricity hauer and Sinan Levent Teske, who provided insight and expertise
and source of CO2 dominate the life cycle GHG emissions and the that greatly assisted this research.
reduction potential compared to conventional, non-renewable
sources: only with renewable electricity supply that has low
GHG-intensity and feedstock CO2 from particular, preferably bio- Appendix A. Supplementary material
genic sources including direct air capture, life cycle GHG emissions
of P2G can be lower than those of conventional technologies. When Supplementary data associated with this article can be found, in
the supply of electricity is largely based on fossil fuels (e.g. average the online version, at http://dx.doi.org/10.1016/j.apenergy.2016.
EU supply), P2G causes higher life cycle GHG emissions than con- 12.098.
ventional technologies. Comparing the variation of GHG emission
reduction caused by different CO2 sources and electricity supply
options shows that, electricity supply causes higher variation of References
system GHG emissions than different CO2 sources.
Our results using two different approaches of dealing with CCU [1] Beaudin M, Zareipour H, Schellenberglabe A, Rosehart W. Energy storage for
in P2M – system expansion and subdivision of joint production of mitigating the variability of renewable electricity sources: an updated review.
Energy Sustain Develop 2010;14:302–14.
electricity/cement and CO2 – show that in case of fossil sources of [2] Evans A, Strezov V, Evans TJ. Assessment of utility energy storage options for
CO2, only applying system expansion allows for meaningful con- increased renewable energy penetration. Renew Sustain Energy Rev
clusions: subdivision does not account for the reduced GHG inten- 2012;16:4141–7.
[3] Denholm P, Ela E, Kirby B, Milligan M. The role of energy storage with
sity of cement and electricity production with CO2 capture and renewable electricity generation; 2010.
therefore, P2M performs worse than producing and using conven- [4] Edenhofer O, Pichs-Madruga R, Sokona Y, Farahani E, Kadner S, Seyboth K,
tional natural gas. System expansion reveals the full potential of et al., editors. Climate change 2014 mitigation of climate change. contribution
of working group III to the fifth assessment report of the intergovernmental
P2M in terms of GHG reduction and shows that P2M based on fossil panel on climate change. Intergovernmental Panel on Climate Change; 2014.
CO2 sources can – although with smaller benefits than biogenic [5] Luo G, Li Y, Tang W, Wei X. Wind curtailment of China’s wind power operation:
CO2 sources – contribute to climate change mitigation. Comparing evolution, causes and solutions. Renew Sustain Energy Rev 2016;53:1190–201.
[6] Solomon AA, Kammen DM, Callaway D. The role of large-scale energy storage
P2H and P2M shows that P2H replacing hydrogen from fossil
design and dispatch in the power grid: a study of very high grid penetration of
sources can more effectively contribute to reduction of GHG emis- variable renewable resources. Appl Energy 2014;134:75–89.
sions than SNG replacing natural gas as vehicle fuel. However, tak- [7] Johnson JX, De Kleine R, Keoleian GA. Assessment of energy storage for
ing into account the existing natural gas infrastructure and the transmission-constrained wind. Appl Energy 2014;124:377–88.
[8] de Sisternes FJ, Jenkins JD, Botterud A. The value of energy storage in
large range of options for using natural gas, large-scale conversion decarbonizing the electricity sector. Appl Energy 2016;175:368–79.
of electricity to SNG might be more realistic than large-scale P2H. [9] Luo X, Wang J, Dooner M, Clarke J. Overview of current development in
In addition to impacts on climate change, the LCIA results for electrical energy storage technologies and the application potential in power
system operation. Appl Energy 2015;137:511–36.
the ILCD midpoint indicators show that P2H could perform better [10] Chen H, Cong TN, Yang W, Tan C, Li Y, Ding Y. Progress in electrical energy
than hydrogen from fossil resources with renewable electricity storage system: a critical review. Prog Nat Sci 2009;19:291–312.
supply, while using SNG from P2M as vehicle fuel, compared to [11] Hofstetter D, Battke B, Cox B, Hughes J. Power-to-Gas in Switzerland - demand,
regulation, economics, technical potential. Erdgas Zürich, Elektrizitätswerk der
natural gas, most often have higher impacts. Stadt Zürich (ewz), Swissgrid AG, Swiss Gas Industry Association (VSG),
Although many scenarios are explored in the study, there are Electrochaea, Swiss Federal Institute of Technology Zurich; 2014.
still some questions that have not been addressed and require fur- [12] Rehman S, Al-Hadhrami LM, Alam MM. Pumped hydro energy storage system:
a technological review. Renew Sustain Energy Rev 2015;44:586–98.
ther analysis. For example, further analysis should be carried out [13] Lehner M, Tichler R, Steinmüller H, Koppe M. Power-to-gas technology and
for technological variations in methanation, as soon as more data business model. Austria: Springer; 2014.
concerning heat generation, external energy demand and catalysts [14] Garmsiri S, Rosen MA, Smith GR. Integration of wind energy, hydrogen and
natural gas pipeline systems to meet community and transportation energy
will be available. Also the utilization of hydrogen, e.g. in fuel cell
needs: a parametric study. Sustainability 2014;5:2506–26.
vehicles, needs to be addressed and compared to vehicles using [15] Altfeld K, Pinchbeck D. Admissible hydrogen concentrations in natural gas
other fuel types. In addition, the environmental performance of systems. DIV Deutscher Industrieverlag GmbH; 2013.
P2G used as electricity storage should be compared against other [16] Grond L, Schulze P, Holstein J. Systems analyses power to gas deliverable 1:
technology review. Part of TKI project TKIG01038 – systems analyses Power-
large-scale seasonal storage technologies, such as pumped hydro to-Gas pathways. Groningen: DNV KEMA Energy & Sustainability; 2013.
storage and compressed air energy storage. [17] Bünger U, Landinger H, Pschorr-Schoberer E, Schmidt P, Weindorf W, Jöhrens J,
Even if our analysis shows that only applying system expansion et al. Power-to-Gas (PtG) in transport - status quo and perspectives for
development. Study in the context of the scientific supervision, support and
for joint production processes in P2G systems provides meaningful guidance of the BMVBS in the sectors transport and mobility with a specific
results from the societal perspective, further approaches for alloca- focus on fuels and propulsion technologies, as well as energy and
tion based on physical properties should be explored, primarily climate. Munich, Heidelberg, Leipzig, Berlin: Federal Ministry of Transport
and Digital Infrastructure (BMVI); 2014.
according to product price or revenue. Furthermore, our analysis [18] Guandalini G, Campanari S, Romano MC. Power-to-gas plants and gas turbines
exclusively addresses the environmental aspects, but economic for improved wind energy dispatchability: energy and economic assessment.
and social issues need to be considered in parallel for decision Appl Energy 2015;147:117–30.
[19] Qadrdan M, Abeysekera M, Chaudry M, Wu J, Jenkins N. Role of power-to-gas
making. Multi-Criteria Decision Analysis (MCDA) can be performed in an integrated gas and electricity system in Great Britain. Int J Hydrogen
on top of these evaluations to establish a more comprehensive Energy 2015;40:5763–75.
understanding of Power-to-Gas under various boundary conditions [20] Schiebahn S, Grube T, Robinius M, Tietze V, Kumar B, Stolten D. Power to gas:
technological overview, systems analysis and economic assessment for a case
and decision criteria.
study in Germany. Int J Hydrogen Energy 2015;40:4285–94.
[21] Zoss T, Dace E, Blumberga D. Modeling a power-to-renewable methane system
Acknowledgement for an assessment of power grid balancing options in the Baltic States’ region.
Appl Energy 2016;170:278–85.
[22] Bensmann B, Hanke-Rauschenbach R, Müller-Syring G, Henel M, Sundmacher
The work was carried out as part of the activities of The Swiss K. Optimal configuration and pressure levels of electrolyzer plants in context
Competence Center for Energy Research for Storage of Heat and of power-to-gas applications. Appl Energy 2016;167:107–24.
338 X. Zhang et al. / Applied Energy 190 (2017) 326–338

[23] Giglio E, Lanzini A, Santarelli M, Leone P. Synthetic natural gas via integrated [39] Borgna L, Geissbühler C, Häberlin H, Kämpfer M, Zwahlen U. Photovoltaik-
high-temperature electrolysis and methanation: part I—Energy performance. J Systemtchnik (PVSYSTE) Schlussbericht; 2007.
Energy Storage 2015;1:22–37. [40] Flowers ME, Smith MK, Parsekian AW, Boyuk DS, McGrath JK, Yates L. Climate
[24] Giglio E, Lanzini A, Santarelli M, Leone P. Synthetic natural gas via integrated impacts on the cost of solar energy. Energy Policy 2016;94:264–73.
high-temperature electrolysis and methanation: part II—Economic analysis. J [41] Raadal HL, Vold BI, Myhr A, Nygaard TA. GHG emissions and energy
Energy Storage 2015;2:64–79. performance of offshore wind power. Renew Energy 2014;66:314–24.
[25] Burkhardt J, Patyk A, Tanguy P, Retzke C. Hydrogen mobility from wind energy [42] Cianelli C. Personal Communication on the HSR Power-to-Gas Pilot Plant;
– a life cycle assessment focusing on the fuel supply. Appl Energy 2014.
2016;181:54–64. [43] Büchi F. Personal communication on the energy consumption of PEM
[26] Bhandari R, Trudewind CA, Zapp P. Life cycle assessment of hydrogen electrolysis system; 2014.
production via electrolysis – a review. J Cleaner Prod 2014;85:151–63. [44] Remer DD, Chai LH. Estimate costs of scaled-up process plants. Chem Eng
[27] Collet P, Flottes E, Favre A, Raynal L, Pierre H, Capela S, et al. Techno-economic 1990:138–75.
and Life Cycle Assessment of methane production via biogas upgrading and [45] Gardiner M. Energy requirements for hydrogen gas compression and
power to gas technology. Appl Energy. liquefaction as related to vehicle storage needs; 2009.
[28] Reiter G, Lindorfer J. Global warming potential of hydrogen and methane [46] Timofte A. A Life Cycle Assessment of the Climeworks technology for direct air
production from renewable electricity via power-to-gas technology. Int J Life capture of carbon dioxide, a semester project in the ecological systems design
Cycle Assess 2015;20:477–89. group. Institute of Environmental Engineering of ETH Zurich; 2013.
[29] von der Assen N, Jung J, Bardow A. Life-cycle assessment of carbon dioxide [47] Goeppert A, Czaun M, Surya Prakash GK, Olah GA. Air as the renewable carbon
capture and utilization: avoiding the pitfalls. Energy Environ Sci source of the future: an overview of CO2 capture from the atmosphere. Energy
2013;6:2721–34. Environ Sci 2012;5:7833–53.
[30] ISO. 14040 Environmental management - Life cycle assessment - Principles [48] Searchinger T, Heimlich R, Houghton RA, Dong F, Elobeid A, Fabiosa J, et al. Use
and framework. 2nd ed. Switzerland 2006. of U.S. croplands for biofuels increases greenhouse gases through emissions
[31] ISO. 14044 Environmental Management - Life Cycle Assesment - Requirements from land-use change. Science 2008;319:1238–40.
and Guidelines; 2006. [49] IPCC. Climate change 2007: The physical science basis. contribution of working
[32] PRé. SimaPro 8.0.4.30 Multi user. PRé Consultants bv, Stationsplein 121, 3818 group I to the fourth assessment report of the intergovernmental panel on
LE Amersfoort, The Netherlands 2014. climate change: Cambridge University Press, Cambridge, United Kingdom and
[33] ecoinvent. ecoinvent version 3.1, allocation, cut-off by classification system New York, NY, USA; 2007.
model. July 8 ed.; 2014. [50] Kellenberger D, Althaus H-J, Jungbluth N, Künniger T. Life cycle inventories of
[34] von der Assen N, Voll P, Peters M, Bardow A. Life cycle assessment of CO2 building products. Final report ecoinvent data v2.0 No. 7.: Swiss Centre for Life
capture and utilization: a tutorial review. Chem Soc Rev 2014;43:7982–94. Cycle Inventories, Dübendorf, CH.; 2007.
[35] Sternberg A, Bardow A. Power-to-What? - Environmental assessment of [51] Edwards R, Larivé J-F, Beziat J-C. Well-to-wheels analysis of future automotive
energy storage systems. Energy Environ Sci 2015;8:389–400. fuels and powertrains in the European context JRC scientific and technical
[36] Volkart K, Bauer C, Boulet C. Life cycle assessment of carbon capture and reports. Luxembourg: European Commission Joint Research Centre Institute
storage in power generation and industry in Europe. Int J Greenhouse Gas for Energy and Transport; 2011.
Control 2013;16:91–106. [52] Waldheim L. Heating value of gases from biomass gasification. Report
[37] Wokaun A, Wilhelm E, Schenler W, Simons A, Bauer C, Bond S, et al. Transition prepared for: IEA bioenergy agreement, Task 20 – thermal gasification of
to hydrogen - pathways toward clean transportation. New York: Cambridge biomass; 2001.
University Press; 2011. [53] Keith DW, Ha-Duong M, Stolaroff JK. Climate strategy with CO2 capture from
[38] Hausschild M, Goedkoop M, Guinee J, Heijungs R, Huijbregts M, Jolliet O, et al. the air. Climatic Change 2006;74:17–45.
Recommendations for Life Cycle Impact Assessment in the European context - [54] Sharma S, Ghoshal SK. Hydrogen the future transportation fuel: From
based on existing environmental impact assessment models and factors production to applications. Renew Sustain Energy Rev 2015;43:1151–8.
(International Reference Life Cycle Data System - ILCD handbook). 1st ed., [55] Parfomak PW. Energy Storage for power grids and electric transportation: a
European Commission-Joint Research Centre - Institute for Environment and technology assessment. In: Congressional research service report for congress
Sustainability; 2011. prepared for memebers and commitees of congress; 2012.

You might also like