2023-Swiss-Planetary Boundaries Assessment of Deep Decarbonisation Options For Building Heating in The European Union

Energy Conversion and Management 278 (2023) 116602
Contents lists available at ScienceDirect
Energy Conversion and Management

journal homepage: www.elsevier.com/locate/enconman
Planetary boundaries assessment of deep decarbonisation options for

building heating in the European Union
Till Weidner , Gonzalo Guillén-Gosálbez *
Institute for Chemical and Bioengineering, Department of Chemistry and Applied Biosciences, ETH Zürich, Vladimir-Prelog-Weg 1, 8093 Zürich, Switzerland
A R T I C L E I N F O A B S T R A C T
Keywords: Building heating is one of the sectors for which multiple decarbonisation options exist and current geopolitical
Heat pumps tensions provide urgency to design adequate regional policies. Heat pumps and hydrogen boilers, alongside
Hydrogen alternative district heating systems, are the most promising alternatives. Although a host of city or country-level
Optimisation
studies exist, it remains controversial what role hydrogen should play for building heating in the European Union
Planetary boundaries
Absolute sustainability
compared with electrification and how blue and green hydrogen differ in terms of costs and environmental
impacts. This work assesses the optimal technology mix for staying within planetary boundaries, and the in
fluence of international cooperation and political restrictions. To perform the analysis, a bottom-up optimisation
model was developed incorporating life cycle assessment constraints and covering production, storage, transport
of energy and carbon dioxide, as well as grid and non-grid connected end-users of heat. It was found that a
building heating system within planetary boundaries is feasible through large-scale electrification via heat
pumps, although at a higher cost than the current system with abatement costs of around 200 €/ton CO2.
Increasing interconnector capacity or onshore wind energy is found to be vital to staying within boundaries. A
strong trade-off for hydrogen was identified, with blue hydrogen being cost-competitive but vastly unsustainable
(when applied to heating) and green hydrogen being 2–3 times more expensive than electrification while still
transgressing several planetary boundaries. The insights from this work indicate that heat pumps and renewable
electricity should be prioritised over hydrogen-based heating in most cases and grid-stability and storage aspects
explored further, while revealing a need for policy instruments to mitigate increased costs for consumers.
1. Introduction 1 (including the United Kingdom), constituting roughly 13.3 % of total

EU27 + 1 GHG emissions [3]. Besides emissions, recent geopolitical
The current environmental footprint of the European Union (EU) is tensions also put in question the current heating system with its supply
well above what could be considered sustainable, with a substantial dependencies. Hence, there are urgent calls to change the way heating
contribution of greenhouse gas (GHG) emissions from energy provision needs are managed. Part of this energy transition is often in conflict with
(~74 % according to Eurostat), and in particular the burning of fossil stakeholder interests; energy providers are interested in maintaining
fuels. The decarbonisation of the power sector through the massive their business model, e.g. in the case of natural gas providers and dis
expansion of renewable electricity generation is on the horizon, with tribution grid operators [4], while household consumers demand low-
ambitious targets being defined by the European Green Deal [1]. Among cost solutions [5], and non-government organisations and climate ac
the other major contributing end uses of energy are transportation and tivists demand a strong focus on rapid decarbonisation [6]. Due to the
heat. The latter includes heating of residential and commercial buildings strong influence of administrative and legislative conditions on the
and industrial process heat, with the residential heating sector alone design of the heating sector, policy makers shape the transition the most
generating around 0.5 Gt CO2 eq. emissions annually [2] in the EU27 + and need to balance those sometimes contradicting interests. Thus, they
Abbreviations: BAU, Business as usual; BECCS, Bioenergy with carbon capture and storage; CCGT, Combined cycle gas turbine; CCS, Carbon capture and storage;
COP, Coefficient of performance; DT, Dispatchable technologies; EU, European Union; GAMS, General Algebraic Modeling System; GHG, Greenhouse gas; GWP,
Global warming potential; HRE, Heat Roadmap Europe; LCA, Life cycle assessment; LCI, Life cycle inventory; LCIA, Life cycle impact analysis; PEM, Proton exchange
membrane; SMR, Steam methane reforming; VRE, Variable renewable energy; WTL, Weighted transgression limit.
* Corresponding author.
E-mail address: [email protected] (G. Guillén-Gosálbez).
https://doi.org/10.1016/j.enconman.2022.116602
Received 3 May 2022; Received in revised form 14 December 2022; Accepted 15 December 2022
Available online 31 January 2023
0196-8904/© 2022 The Author(s). Published by Elsevier Ltd. This is an open access article under the CC BY license (http://creativecommons.org/licenses/by/4.0/).
T. Weidner and G. Guillén-Gosálbez Energy Conversion and Management 278 (2023) 116602
often rely on expert-driven energy systems models to make more of detail for hydrogen transport and distribution networks also varied
informed decisions. widely, from its cost exogenously assumed [21] to specific modelling
approaches to evaluate them [11]. The choice of energy storage tech
1.1. Building heating in international energy systems models nologies when electrification was an option also differed, as often only
one energy carrier was evaluated (hydrogen in most of the UK studies) or
The need to change the energy system has led to the development of a range of technological choices was modelled and optimised for [17].
a range of integrated assessment models that consider economic, policy, None of the studies considered energy storage through compressed
and technological aspects. However, the main focus in the past has been gases, such as air or CO2. The studies above indicate a knowledge gap in
on power systems [7], while heating has received much less attention. EU studies, as it remains yet unclear how a fully hydrogen-fuelled
Nevertheless, the building heating sector is of particular relevance due heating sector, including non-grid connected consumers, would
to the existence of a variety of potential energy carriers and heating compare in terms of costs and impacts to the current system and fully
solutions, and the greatly differing energy demand across the year, electrified potential systems.
potentially misaligned with renewable energy supply, in particular solar
PV. In addition, the political union of the member states could allow for 1.3. Environmental sustainability consideration related to heating
intensive cooperation and trading to exploit the regional strengths
regarding renewable energy generation, underground gas storage, and Another aspect that remains largely unexplored is related to the
seasonal heating fluctuations of individual countries [8]. Heating has environmental evaluation of heating systems. Sustainability studies
thus recently received more attention and continent-wide models have have usually focused on GHG reduction targets, either by following a life
been developed to study the sector, either isolated [9] or coupled with cycle assessment (LCA) approach [23] or by imposing a price on CO2 to
the power and transport sector, e.g. PyPSA-Eur-Sec [9] and JRC-EU- obtain the desired target [12]. Still, many of the above-mentioned
TIMES [8]. These studies have investigated how future heating sys optimisation studies related to heating do not specifically consider
tems could meet certain emission mitigation targets by combining a CO2 emissions alongside costs. In addition, even after 2015, reduction
range of strategies, including electrification, combustion of conven targets in official assessments have not always been in line with the Paris
tional fuels (e.g. biomass, fossil fuels) and utilisation of alternative en Agreement, for example the Heat Roadmap Europe project [24]. Many
ergy carriers (e.g. hydrogen, synthetic fuels). Most models focus on cost- of the integrated assessment models considering the heating sector still
optimisation but other aspects are of relevance as well: the choice of rely on a significant amount of fossil fuel combustion in their cost-
energy carrier for seasonal storage and heating appliances [11], decar optimised scenarios and it remains unclear if the results respect the
bonisation policy options [12], and the potential synergies through Earth’s ecological capacity, delimited by the planetary boundaries.
sector coupling and international collaboration. The latter showed These planetary limits define what level of human-induced pressures on
promising results for reducing costs (up to 20 % through cooperation the environment can be considered sustainable, i.e. not risking tipping
[10]), the required storage capacity (up to 50 % reduction for a points and irreversible damage in a range of earth system processes [25].
modelling study in the UK [13]) as well as the final energy needs (34 % There is thus a gap in the literature related to the evaluation of other
reduction for a Danish case study [14]). Nevertheless, this aspect is often impact categories beyond global warming for mixed energy carrier
underexplored when combined with environmental assessments or studies.
when considering the whole EU. Besides international collaboration,
studies differ in the choice and favourability of different low-carbon 1.4. Novelties and objectives
energy options, mainly hydrogen and electrified heating. The large-
scale use of hydrogen has so far been neglected in most academic en In a recent review in Joule covering 32 studies using hydrogen in
ergy scenarios [15] and existing EU-wide studies on heating usually heating [26], only one reports environmental impacts [27] but the scope
exclude it [16,17] or only consider very small shares of it [18]. is restricted to the UK and it does not include any energy systems model
or optimization. Although on consumer cost modelling one study for the
1.2. Hydrogen use for heating in country-level studies EU exists [28], it does not model green/blue hydrogen infrastructure
and trade on a country-level basis, but introduces a simplified weighted
One reason for the underrepresentation of hydrogen may be the average cost study across all EU member states based on current elec
significantly higher electricity consumption of heating with electrolytic tricity consumption (so no time period dynamics). A study combining
green hydrogen relative to direct electrified heating with heat pumps the previously described aspects, namely, investigating the effect of
[13]. However, other considerations such as seasonal storage, the extensive international trade (i.e. collaboration) and comparing
transport network, and end-user equipment play a crucial in role hydrogen and electrified heating in an EU-wide energy systems analysis
determining which option is most favourable. This has led to a range of that also includes a holistic sustainability assessment, is so far missing in
studies considering hydrogen for heating in the United Kingdom, which the literature. This work thus aims to explore these aspects around deep
has an extensive natural gas grid potentially suitable for retrofitting. decarbonisation in the EU27 + 1 heating sector devising a bottom-up
Purely relying on hydrogen, Sunny et al. [19] found that blue hydrogen optimisation model incorporating life cycle assessment (LCA). The
from natural gas in combination with biomass energy with carbon latter is carried out as an absolute sustainability assessment [29], where
capture and storage (BECCS) is cheaper than green hydrogen to deliver the impacts relative to the planet’s carrying capacity are assessed and
carbon–neutral heating. Others have compared direct electrification the system is designed to minimise the transgression of critical bio
with hydrogen-based heating for the UK, either by comparing two sce physical limits regulating the Earth’s resilience [30]. Besides combining
narios, for instance with sector coupling [20] and storage considerations the three previously mentioned aspects in a multi-criteria optimisation
[13], or applying cost-based optimisation, combined with district model, this study’s novelties also include the consideration of rural users
heating [21], inter-seasonal storage [11] and supply chain optimization and end-user equipment, the modeling of the CO2 network encompass
[22]. The scenario-based assessments showed almost twice the elec ing carbon capture and storage (CCS) technologies, pipeline and storage
tricity requirements and higher costs for the hydrogen-only scenarios infrastructure, and the inclusion of compressed air energy storage,
(around 20 %). Nevertheless, hydrogen-based heating was chosen to alongside gaseous hydrogen storage options. The work reveals that
varying degrees by the optimisation studies (between 0 and 50 %, staying within planetary boundaries is only possible through electrifi
depending on the region and assumptions). Seasonal hydrogen storage cation via heat pumps but also that such a transition would come at high
was prominent in all studies but its contribution to overall system costs system-level costs and require some level of collaboration. Further,
varied significantly, between around 5 % [11] and 60 % [19]. The level hydrogen heating is shown to be either economically unfavourable in
2
the case of green hydrogen or environmentally unfavourable in the case deployed low-carbon electricity generation infrastructure. It includes
of blue hydrogen. All those aspects provide crucial insights into the long- variable renewable energy sources (VRE), i.e. photovoltaic, onshore and
term planning of the EU-wide energy transition in a post-carbon world. offshore wind, as well as dispatchable technologies (DT), including
nuclear fission, and pulverised coal power plants, and combined cycle
2. Methods: gas turbines (CCGT) with CCS. Interactions with the incumbent power
sector were not explicitly modelled; the additional and new electricity
This section highlights the components and boundaries of the system demand for building heating was considered on top of currently existing
under investigation (2.1), then sheds light on the case study (2.2) and electricity demand, requiring additional infrastructure. Biomass uti
model setup (2.3), followed by details about the cost (2.4) and impact lisation for electricity or hydrogen production was neglected as many
(2.5) assessments, and the optimisation (2.6). Lastly, the different sce previous studies indicate that the regional potential should be used for
narios are explained (2.7). The main steps are listed in Fig. 1. timber [32] or BECCS [33] and is thus attributed to other sectors.
Hydrogen could be produced by proton exchange membrane (PEM)
electrolysis, called e-hydrogen here, or by steam methane reforming
2.1. System boundaries and problem statement (SMR) with CCS, also called blue hydrogen. It could also be converted to
electricity in fuel cells, thus enabling hydrogen to act as a seasonal
The components of the system, summarised in a superstructure storage carrier without necessarily requiring hydrogen heating end-user
encompassing all possible alternatives to meet the heating needs, are equipment. The compression of hydrogen for transportation and distri
shown in Fig. 2. This work considers the building heating sector of the bution (70 bar, accounting for distance-dependent losses), storage (170
EU27 + 1 (+1 is the United Kingdom). The heating demand as thermal bar in salt caverns and 250 bar in pressurised tanks) and truck transport
energy includes residential space heating, hot water heating, and service (250 bar) was considered, both through compressor capacity infra
sector building heating, while excluding industrial and process heating. structure and additional electricity consumption. Included are the
The heat demand was assessed for three different types of end-users electricity consumption of both pathways and the natural gas con
according to [2]; a) connected to the gas grid, b) connected to a dis sumption of SMR.
trict heating grid and c) individual users. The latter is further separated The transport of the two energy carriers to the consumers was
into currently fossil fuel or biomass boiler-based (“individual”) and users incorporated by simplified country-level representations (see section
with existing electrical heating, typically electric radiators, to discern 2.6). For electricity, high voltage grid infrastructure was modelled for
costs for electrification requirements (lower for users already heated inter-connecters between countries and transmission between regions
electrically). The split between end-use types was assumed to be con within a country (accounting for already existing capacity), medium
stant, mainly due to no observed changes in the last 20 years [2] and voltage grids for distribution within a region, as well as low voltage
stable heat density patterns (e.g. cities versus rural with different ra household level connections (e.g. fortified wiring). For hydrogen, in
tionales for grid access). The thermal demand could either be met by ternational transport and domestic transmission pipelines were
electrification (via heat pumps) or hydrogen (via hydrogen boilers). modelled alongside gas network distribution (retrofitted and new-built)
Electric resistive heaters were excluded since calculations determined for grid-connected buildings and pressurised road transport of hydrogen
inferior economic and environmental performance compared to heat (delivered by hydrogen-fuelled trucks) for individual homes. Additional
pumps. Other types of low-carbon heating, such as geothermal or solar local piping was considered for all cases. Geological storage was an
thermal energy [31] that could likely play an increasing role in some option for seasonal storage of hydrogen (salt caverns) and a requirement
locations, were excluded due to their limited universal applicability. for CO2 captured in gas and coal power plants and the SMR (in depleted
The supply of electricity was modelled by considering additionally
Fig. 1. Outline of the methodological approach for the EU27 + 1 building heating assessment according to planetary boundaries. The superstructure is shown in
Fig. 2 and is mathematically described by energy and mass balances, technical constraints and costs. The life cycle inventories include infrastructure and operational
input elements. The planetary boundaries were downscaled to the population of the EU27 + 1 and the share of today’s heating sector emissions compared with those
of the global economy. The transgression level of the downscaled boundaries was minimised in the impact-based optimisation. Icon credit: Prettycons, Smashicons,
smallikeart.
3
Fig. 2. Superstructure of the proposed system to satisfy building heating demand in the EU27 + 1. Country-specific capacity limits are defined for photovoltaic (PV),
on– and offshore wind, nuclear, salt cavern storage and storage for captured CO2 from fossil-fuel electricity production, and the steam methane reforming process.
oil and gas fields). Overground storage was considered an option for Choosing a reference year in the somewhat distant future would allow
electricity storage (as compressed air) and hydrogen (in pressurised for a transition based on technical (e.g. technology maturity) aspects.
tanks). All options were available for intra-monthly and seasonal stor A bottom-up model with exogenous demand assumptions for a year
age. Batteries were only available for short-term storage. Included are in the future is seen as adequate here, as heating demand is less affected
storage efficiency losses and compression requirements, excluded are by price and income growth than other sectors [34]. The reference year
long-term storage losses and competition for geological storage with was further divided into 12 time periods, representing the typical days of
other sectors. each month. Each time period was characterised by an average daily
End-user equipment converting electricity or hydrogen into heat was heating demand, specific capacity factors for renewable energy gener
defined depending on the end-user type; gas grid-connected buildings ation, and coefficient of performance (COP) of heat pumps. Intra-day
were assumed to either use an electric air source heat pump or a and weekly fluctuations of supply and demand were not modelled
hydrogen boiler. District heating was delivered through hot water net explicitly but were rather approximated by differentiating between
works and household heat exchangers and was assumed to be generated average and hourly peak demand and by grid balancing constraints,
by large-scale commercial units semi-centrally, either by ground-source including positive effects of international trade, a higher share of dis
heat pumps or hydrogen boilers. Buildings not connected to either grid patchable technologies and battery storage capacity (see Equation (1) in
could also employ heat pumps or hydrogen boilers heating but the latter 2.3).
required an additional storage tank for gaseous hydrogen delivered by The hydrogen pipeline network was assumed to have internal storage
trucks. Included are piping and wiring to connect the boilers or heat capacity [19] for intra-day fluctuations and hydrogen production was
pumps and excluded is the building piping for radiator or floor-based considered flexible enough to increase production during peak hours,
heating systems. with a minimum monthly production of 10 % through the more flexible
Given the system described above, economic, cost data, technical electrolysis [19]. Cost reductions for recently developed technologies
parameters and heating demand, the goal of the analysis is to find the due to broader commercialisation (technological learning curves) are
optimal combination of technologies that meet the heating requirements assumed for most technologies and infrastructure items. Their capacity-
at minimum cost and environmental impact, as described later in the dependent capital, fixed and variable operating costs are taken from
article. Further details about system components and assumptions are projections for 2040. Reductions in heat demand for 2040 compared
described in section 4 in the Supplementary Information (SI). with 2015 due to thermal insulation are assumed based on historical
evidence and projections. These reductions are slightly lower than what
2.2. Case study: Reference year, projections and spatial resolution is assumed as the maximum potential in other studies (e.g. [23].
The chosen spatial resolution was the country level, which allows
This work studies the economic and carrying capacity aspects of assessing the location-specific energy supply and demand throughout
decarbonised heating systems to provide a meaningful picture of a po the year and potential balancing through storage and trade. Average
tential future. The focus is on the most suitable technologies to achieve a values for natural resource availability and heating demand, area and
sustainable low-carbon transition but without defining the temporal population density-dependent transmission infrastructure, and country-
pathways for realising such a transition. Thus, this study is not a capacity specific geological storage capacity provide country-level characteristics
expansion study but considers the potential energy mix to satisfy heat without unnecessarily complicating the modelling process. The country
demand in a typical year in the future, i.e. 2040. As such, this is a is typically the political and administrative unit in which energy policy
greenfield study from the perspective of energy provision but assumes is drafted and implemented and is hence ideal for directional recom
no change to existing building archetypes (e.g. end-use types) and mendations. Naturally, no recommendations for a very detailed sub-
connecting infrastructure. The greenfield approach is justified as none of national energy system design can be derived from this work. The
the technologies considered are currently deployed on a large scale. benefits of international cooperation are nevertheless highlighted and
4
can inform European Union policymaking. (LOSSHV ) and the population-weighted distance between the countries
(DISTrans
k,kk ). The option to curtail was introduced to reduce computational
complexity and reflect current practice with large-scale renewable en
2.3. Model set-up ergy infrastructure. Note that only countries with existing trading ca
pacity would be able to trade (represented by the matrix ITMk,kk ).
The model takes the form of a linear programming formulation, ∑( ( ))
which is implemented in GAMS 35.2 [35]. The LP formulation leads to egen
k,t + etrans
kk,k,t 1 − LOSS
HV
DISTransHV
k,kk + estore,out
k,t + eFC
k,t
lower computational requirements compared to mixed-integer and non- ∑
kk∈IT
∑
linear models. We note that the modelling of minimum capacities con = etrans
k,kk,t + euse store,in
k,t,i + ek,t + ePEM
k,t + ecomp
k,t + eSMR batt
k,t + ek,t (1 − μBATT)
straints and economies of scale would require including binary vari kk∈IT i∈I
ables, leading to an MILP formulation. However, numerical examples + ecurtail

k,t ∀k
showed that the MILP relaxation tends to provide similar solutions and ∈ K, t ∈ T
insights as the LP model solved in this work, so we decided to avoid the
(1)
use of binary variables in the formulation. Moreover, nonlinearities
were also avoided by assuming simple linear approximations of the Akin to electricity, the energy balance for hydrogen, shown in
capacity limitations. Hence, connections between countries, generation Equation (2), included hydrogen generation from electrolysis and the
and storage infrastructure, transmission and distribution infrastructure, SMR process, the balance of imports (htrans
kk,k,t ) and exports (hk,kk,t ) and
trans
and end-user equipment are modelled linearly by capacity or size. storage (hstore
k,t ), and the consumption for building heating (hk,t
totaluse
), fuel
Economies of scale for large infrastructure are implicitly accounted for use,truck
cell conversion (hFC
k,t ) and delivery trucks (hk,t ). Only one variable was
by selecting cost parameters for large units. For end-user equipment,
required for hydrogen storage since no direct loss of energy content was
typical household sizes were assumed to calculate the kW-based costs
assumed to occur, although energy is required to compress the gas.
and impacts.
∑ ∑
Concerning the model size, it comprises six electricity generation hgen,e + hgen,NG htrans htrans totaluse
+ hFC store use,truck
k,t k,t + kk,k,t = k,kk,t + hk,t k,t + hk,t + hk,t ∀k
technologies, 12 time periods or months, four end-use types, 28 coun kk∈IT kk∈IT
tries, 16 impact categories, and 39 elements (mainly infrastructure). ∈ K, t ∈ T
Several model parameters were defined as described below. (2)
3-dimensional parameters: The capacity factor of VRE (by technol
ogy, country, and month), heating demand (by end user, country, and The amount of electricity stored in each month and country (estorelvl
k,t )
month), and electricity conversion efficiency (by end-user, country, and was determined by Equation (3), from the storage level of the previous
month). k,t− 1 ), the amount sent to storage minus the efficiency losses
month (estorelvl
2-dimensional parameters: The possible grid connections between (LOSSStore,e ) and the withdrawal amount [11]. The balance was similar
countries, the distances between countries, and the impacts associated for the two hydrogen storage options, salt caverns, and compressed gas
with each infrastructure item for different impact categories. tanks, excluding the losses.
1-dimensional parameters: Available capacity for VRE, the fuel ef ( )
ficiency of power plants, and the hours per month; for countries, the estorelvl
k,t = estorelvl store,in
k,t− 1 + ek,t 1 − LOSSStore,e − estore,out
k,t ∀k ∈ K, t ∈ T (3)
typical distances between urban agglomerations, area, and population
Different applications required compression of gases, resulting in an
density and geological storage capacities; for end-use technologies, the
overall electricity consumption Equation (4) for compression. The
conversion efficiency for hydrogen, the ratio of peak to average demand,
hydrogen has to be compressed for storage (ecomp,store ), international
the capacity-based distribution network length, as well as cost and im k,t
pacts for all infrastructure items. (ecomp,trans

k,t ) and regional (ecomp,reg
k,t ) transport and truck delivery (ecomp,truck
k,t ),
A list of all parameters is provided in Supplementary Table 1 and while CO2 also requires compression for transport and storage (ecomp,CO2
k,t ).
their values are in Supplementary Tables 3 to 6.
The equations are grouped into energy balances, other constraints, ecomp
k,t = ecomp,store
k,t + ecomp,trans
k,t + ecomp,reg
k,t + ecomp,truck
k,t + ecomp,CO2
k,t ∀k ∈ K, t ∈ T
and cost and impact assessment equations. The most relevant balances (4)
and constraints are explained here and all other equations are described
Equation (5) shows how the electricity required for international
in section 3 in the SI. All variables are listed in Supplementary Table 2.
transport of hydrogen was calculated. The amount exported (htrans
k,kk,t ) from
2.3.1. Energy balances country k to country kk was multiplied by the distance between those
The equations used in this work are bespoke to the problem formu countries (DISTRANS
k,kk ), an equivalent distance for a pressure drop of the
lation but most are based on general mass and energy conservation laws gas from 70 to 30 bar (LOSSPipe ) and the compression requirements to
as well as efficiency and capacity limitations formulas. Specifically, the compress hydrogen from 30 to 70 bar (COMP3070 ).
energy provision, capacity constraints, and emissions and costs-related ( )
equations are based on those used in the RAPID model [36]. For elec
∑ DISTRANS
(5)
comp,trans trans k,kk 3070
ek,t = hk,kk,t COMP ∀k ∈ K, t ∈ T
gen LOSSPipe
tricity, the amount of power generated by technologies (ek,t ), plus the kk∈IT
total amount of electricity imported (etrans

kk,k,t ) minus distance-dependent The satisfaction of the building heating demand is guaranteed
losses (LOSSHV ), the withdrawals from storage and electricity from through Equation (6). Heating demand by end-user for each country and
hydrogen fuel cells (eFC
k,t ) have to match the electricity consumption in month (DEMHeat
k,t,i ) can be covered by hydrogen (hk,t,i ) or electricity (ek,t,i ),
use use
the same country k and month t, as shown in Equation (1). The elec both converted to heat using a conversion factor (CONFek,t,i and CONFhi ,
tricity consumption includes the amount exported (etrans
k,kk,t ), the amount where the former also depends on the average temperature in each
used for heating (euse store,in
k,t,i ), the input to storage (ek,t ), the amount used to country and month). Losses in the regional electricity transmission are
produce hydrogen from electrolysis (ePEM accounted for by the average weighted distance (DISREG ) within a
k,t ), the amount required for
k
hydrogen compression (ecomp SMR country and a loss factor (LOSSHV ). Losses in the hydrogen network are
k,t ) and the SMR process (ek,t ), the losses
considered in the energy balance for the compressors.
(μBATT) from short-term battery storage (ebatt
k,t ) and any curtailment
(ecurtail
k,t ). Losses due to transport are accounted for using a loss factor
5
( )
euse e
k,t,i CONFk,t,i 1 − LOSSHV DISREG
k + huse h Heat
k,t,i CONFi = DEMk,t,i ∀i ∈ I, k ∈ K, t
profiles. Such spatially diversified deployment prevents brown-outs and
black-outs during doldrums for system designs where the internationally
∈T
traded electricity is high and from a few locations [37].
(6)
∑( )1
Only CO2 and water use were calculated on a mass basis. Equation 7 captech tech
j=wind on,k + capj=wind off ,k ≤ captech tech
j=wind on,kk + capj=wind off ,kk ∀k ∈ K
5
describes the balance for CO2 captured (co2cap
k,t ), consisting of the product
kk∈K
(11)
of the amount of hydrogen produced via the SMR process (hGen,NG
k,t )
multiplied with its CO2 emission intensity (CO2SMR ) and capture effi Equation (12) ensures that all captured CO2 (co2cap
k,t ) equals the total
ciency (CAPEFFSMR , 90 %), the amount of electricity produced from amount stored in the EU + 1 (co2store
k,t ), while Equation (13) ensures that
either coal (etech

j=CoalCCS,k,t ) or gas (ej=CCGTCCS,k,t ) power plants and their
tech for each country and month, the total amount of CO2 captured (co2cap
k,t )
respective emission intensities (CO2 and CO2 ) and capture ef and imported (co2trans ) from all countries k equals to the amount stored
′
COAL CCGT
k ,k,t
′
ficiencies (CAPEFFCOAL , 87 % and CAPEFFCCGT , 86 %). (co2store

k,t )
and exported (co2Trans
′ ) from all countries k .
′
k,k ,t
∑ cap ∑
co2cap gen,NG
k,t = hk,t CO2SMR CAPEFFSMR + etech
j=CoalCCS,k,t CO2
COAL
CAPEFFCOAL co2k,t = co2Store ∀t ∈ T (12)
k,t
+ etech
j=CCGTCCS,k,t CO2
CCGT
CAPEFFCCGT ∀k k∈K k∈K
∈ K, t ∈ T ∑ ∑
co2cap
k,t + co2Trans
k’ ,k,t = co2Trans Store
k,k’ ,t + co2k,t ∀k ∈ K, t ∈ T (13)
(7) k’∈IT k’ ∈IT
2.3.2. Technical constraints

2.4. Cost assessment
The capacity deployed for electricity generation (captech
j,k ) in each
country must be lower or equal to the available capacity (technical The compact form for the cost-based objective function inspired by
j,k ) minus the capacity
potential) for the respective technology (¯CAPAVA (Galán-Martín et al., 2021) is shown below in Equation (14). Here, the
allocated to serving the power sector in the future (CAPPWR
j,k ), as shown in
capacity (capf ,k ) of each infrastructure item f (generation, storage,
Equation in 8 (Galán-Martín et al., 2021). transport, distribution, and end-user equipment) is multiplied for each
country by the capital expenditure (CAPEXf ) and a capital recovery
captech ¯AVA PWR
j,k ≤ CAPj,k − CAPj,k ∀j ∈ J, k ∈ K (8) factor (CRFf ), and added to the capacity-dependent operational ex
penses (OPEXf ) and the input cost (ICf ) considering their annual use
The electricity generated (captech
j,k ) by each renewable technology j in
(Inputsf,k ) for each country k contributes to total cost (TC).
month t and country k averaged by the hours of the month (MHt ) cannot
∑∑
be larger than the capacity deployed in that country (captech
j,k ) multiplied minTC = CAPEXf capf ,k CRF f + OPEXf capf ,k + ICf Inputsf ,k (14)
by the capacity factor (CAPFGen
j,k,t , <1) as shown in Equation (9). The latter
f ∈F k∈K
depends on sunshine and wind speed characteristics in each country and The capital recovery factor (CRFf ) as shown in Equation (15) (Galán-
month for renewable technologies. Dispatchable technologies have an Martín et al., 2021) annualises the infrastructure investment using the
annual capacity factor and a minimum capacity factor for each month. equipment lifetime (LTf ) and the weighted average cost of capital
(WACC) of 7 % (Galán-Martín et al., 2021).
etech
(9)
j,k,t
≤ captech Gen
j,k CAPFj,k,t ∀j ∈ VRE, k ∈ K, t ∈ T
MHt WACC(1 + WACC)LTf
CRFf = ∀f ∈ F (15)
The electricity balancing constraint is introduced through Equation (1 + WACC)LTf − 1
(10) for each country k and month t, approximating required measures The optimisation model minimises either the total annualised system
to balance supply and demand mismatches and fluctuations. A minimum cost, using weighted average capital cost and equipment lifetime, or the
amount of dispatchable energy (etechj∈DT,k,t ) is enforced through a back-up weighted transgression level (see below). The assumed costs for all el
production factor (BUPE). The requirements for dispatchable energy ements, their variability, and potential consideration of mutual use and
are reduced through imports (etrans
k ,k,t
′ ), battery (ebatt
k,t ) or compressed air existing infrastructure are detailed in Supplementary Table 7.
energy storage (estore,out
k,t ), and electricity generation from large-scale
hydrogen fuel cells (eFCk,t ). The different factors describe the approxi 2.5. Impact assessment
mated stabilising potentials of those measures, being lower for inter
national transport [13] and higher for battery utilisation [20]. Note that This study considers the absolute sustainability of different scenarios
the resolution of the study also implicitly caters for wind turbines being for building heating in the EU27 + 1. Life cycle analysis is employed
dispersed throughout each country with regional trade for balancing. based on ISO 14040, including a scope and goal definition (see 2.1),
( ) collection of life cycle inventories (LCI), the life cycle impact assessment
∑
tech
∑
use
∑
Trans batt store,out FC (LCIA), and an interpretation of results. The latter is based on the rela
ej,k,t ≥ ek,i − ek’ ,k,t 0.5 − ek,t 2.7 − ek,t − ek,t BUPE ∀k
j∈DTj i∈I k’ ∈IT
tive sustainability between scenarios and the scenario’s absolute sus
tainability relative to the planetary carrying capacity or safe operating
∈ K, t ∈ T
space [25]. The concept employed here is based on linking the LCA
(10) results with the control variables of the planetary boundary framework
As an additional balancing constraint, Equation (11) states that the [40], hence the life cycle impacts on the nine planetary boundaries are
total capacity of newly deployed onshore (captech
j=wind on,k ) and offshore
assessed. Fig. 1 illustrates the devised method.
(captech In LCA phase one, the functional unit is defined as the final heating
j=wind off,k ) wind power installed in a single country may not be larger
demand of the EU27 + 1 (monthly profile). The system boundaries cover
than 1/5 (heuristic value) of the total wind power capacity installed in
the supply chain impacts of infrastructure and fuel from outside the
the 28 countries. This constraint only refers to the additional electricity
EU27 + 1 (cradle-to-grave).
generation for the heating sector and introduces spatially diversified
With regards to LCA phase two, the LCI was constructed based on
capacity deployment to smoothen regional wind energy generation
existing Ecoinvent (3.7) entries and some tailored entries for novel
6
technologies (e.g. PEM electrolysis, CCS, geological storage) obtained Weighting the impact categories in any way is already a normative
from the literature. Further details and sources are listed in Supple decision [45] but equal weighting is deemed appropriate here to find
mentary Table 8. Of the 39 activities considered in this study, five are scenarios that are absolutely sustainable in all categories.
consumables (fuel and water) and 33 are infrastructure items such as
solar panels, power plants, pipelines, and end-user equipment. For 2.6. Optimisation model
renewable electricity generation, the impacts per unit of capacity
(considering the full life cycle of infrastructure) required to generate the The overall bi-criteria optimisation problem can be expressed in
required output, rather than the impacts per produced unit of electricity, compact form as follows [46]:
were chosen in this work (operational impacts were assumed to be
negligible). This is to reduce inaccuracies resulting from relying on fixed (M)min{ψ k (x, y) }
x,y
capacity factor assumptions in Ecoinvent entries compared with flexible
capacity factors used in this work (both for renewable and dispatchable s.t.h(x, y) = 0
technologies).
Concerning LCA phase three, the activities (LCIf ) were multiplied by g(x, y) ≤ 0
the respective characterisation factors (CFp ) related to the planetary
boundary categories according to Ryberg et al. [38] as shown in Equa x ∈ Rn , y ∈ {0, 1}
tion (16) with the calculations for biosphere integrity being done ac Where Equations S51-S58 and S62-S63 are in the form h(x, y) = 0
cording to Galan-Martın et al. [39]. and Equations S27-S50, S59-S61 and S64 in the form g(x, y) ≤ 0.
IMPACTf,p = LCIf CFp ∀f ∈ F, p ∈ P (16) The problem can be formulated as a bi-criteria LP that seeks to
minimise the average transgression level and total cost simultaneously,
Each element’s impact values in each of the EF categories were where model M can be solved with the epsilon constraint method [47].
extracted from SimaPro 9.2, either in units of length (km), capacity The approach was simplified by first minimising WTL, the least-impact
(MW, MWp or tonne/h), energy content (MWh), mass (tonnes) or solution, which may or may not show a TLp value in one boundary
weight-distance (tkm). Total impacts (TIp ) in 9 different planetary above 1. Since atmospheric CO2 was always the category with the
boundary categories p are determined similarly to the cost (see Equation highest impact, a value above 1 in this category meant the system can’t
(14)), using the lifetime (LTf ) to annualise infrastructure impacts be made absolutely sustainable under scenario assumptions. If this was
(IMPACTf,p ) as shown in Equation (17). For consumable elements such the case, the total cost (TC) was simply minimised, generating the least-
as fuel or input, the IMPACTf parameter with element-specific units is cost solution. If the impact in all PBs in the least-impact solution laid
multiplied by the annual consumption (Inputsf,k ). within the limit (i.e. below 1), the least-cost solution was obtained and
∑∑(IMPACTf,p ) an additional third problem was solved that minimises the cost of
TI p = capf ,k + IMPACTinputs
f,p Inputsf ,k ∀p ∈ P (17) operating within the safe operating space. This can be done using a hard
LTf
f ∈F k∈K constraint but it was found that this was numerically challenging.
The impacts of all elements were incorporated in the model through Instead, a penalty was added by introducing a positive slack variable (sp ,
a 2-dimensional parameter matrix. In the LCIA phase, the impact values see [48] that was larger than zero when one of the impact categories was
in each category are expressed relative to the allocated safe operating above 1 as shown in Equation (21), thus introducing a penalty (π) when
space (aSOSp ) of the planetary boundaries [25], resulting in a trans the TLp was exceeding the boundary as shown in Equation (22).
gression level (TLp ) for each category as shown in Equation (18), akin to TLp ≤ 1 + sp ∀p ∈ P (21)
[40].
∑
TIp minTC + sp π (22)
TLp = ∀p ∈ P (18) p∈P
aSOSp
The global safe operating space was downscaled to the European 2.7. Scenarios and cases
Union as shown in Equation (19), first by an egalitarian approach, i.e.
the population (POPEU+1 ) as a share of the global population (POPworld ), The business as usual (BAU) scenario was based on the current
around 5.6 %. It was further downscaled by a grandfathering or historic heating provision patterns of the EU but assumed the same demand
rights approach [41], in this case by the share of GHG emissions of the reduction due to thermal insulation. District heating is further divided
heating sector (GHGheating ) compared with the whole economy into combined heat and power plants by type; coal, oil, natural gas, and
(GHGEU+1 ), around 17.7 %. This resulted in roughly 1 % of the safe biomass plants (according to Eurostat). The cost of delivered electricity,
operating space allocated to the building heating sector in the EU27 + 1. heating oil, natural gas (European Commision data portal) and district
heat [49] was defined according to household retail prices, which
POPEU27+1 GHGheating
aSOSp = SOSp ∀p ∈ P (19) already include generation or import cost, transmission, distribution,
POPworld GHGEU27+1
and grid balancing costs. For coal and biomass to individual consumers,
There are inherent normative issues around downscaling [42] but it a fixed market price was chosen and costs for transportation including
was still chosen here to enable a constrained economic optimisation. import, regional and final distribution were accounted for using average
Another approach to downscaling is by household consumption expen distances. The diesel use and the number of delivery trucks required
diture share. It was found that the allocated SOS would be similar as, on were calculated akin to the method for gaseous hydrogen delivery. The
average, the EU27 + 1 households spend 22 % on energy [43], of which user cost for heating equipment was based on the number of users,
around 78.4 % is for space and water heating (resulting in 17.3 %). which was derived from national population and average household
Finally, a single impact metric (weighted transgression level, WTL), size, and typical boiler equipment size. Existing Ecoinvent 3.7 LCI data
is derived from the sum of the transgression level (TLp ) in each category were used to estimate the impact of heat provision per kWh basis
divided by the number of categories (|PB|), i.e. equal weighting is (Supplementary Table 10), additionally accounting for the driven ton
applied as described in Weidner et al. [44] and shown in Equation (20). ne⋅km for distribution to individual consumers. Data sources and pa
∑ 1 rameters are listed in Supplementary Table 9.
minWTL = TLp (20) The Heat Roadmap Europe 4 (HRE) scenario [24] is used as a com
|PB|
p∈P
parison and to shed light on the sustainability of previously proposed
7
system configurations. The split between energy carriers for 2040 is Section 4.
defined as the average between 2030 and 2050 (assuming linear
change). The method is similar to the BAU; household retails prices were 3.1. Planetary boundaries, CO2 emissions and economic performance
chosen for electricity, natural gas, and district heat as the EU average
(since the actual split by country is not reported), and for biomass and Fig. 3 shows the transgression levels for each planetary boundary
coal, market prices were used and transport requirements accounted for. downscaled to the building heating sector of the EU27 + 1. Analysing
For the main network model, the approach mentioned above was the BAU solution, it shows an average TL of 5.2, and attains a maximum
carried out, where a least-cost (“Cost”) and least-impact (“WTL”) solu TL in the category atmospheric CO2 concentration (climate change). The
tion was calculated for each scenario. If the WTL minimisation led to HRE4 scenario is still vastly unsustainable with only marginally less
results within planetary boundaries, the cost optimisation was also transgression in climate change-related boundaries (recall that HRE4 is
carried out using the slack variables to obtain a solution within plane similar to the BAU but uses more gas instead of oil and coal).
tary boundaries (“Cost+”). Besides the standard scenario, the main The alternative solutions greatly reduce the overall impacts of the
alternative scenarios to investigate the decarbonised heat system model proposed system compared with the BAU and HRE4 scenario. It was
are a) non-collaboration between states (no trade of energy carriers or found that it would be possible for the EU27 + 1 building heating sector
shared usage of geological storage capacity for H2 and CO2, “no-col”), b) to stay within planetary boundaries (i.e. all TL values are ≤ 1). The share
a political or popular blockade for additional onshore wind and nuclear of hydrogen used for heating was zero when optimising for impact and
energy, also called capacity constraints (“no-N/O”), d) combining a) and very small when optimising for costs with a penalty for transgression (a
b) representing political gridlock (“PG”), and e) hydrogen use only maximum of 4.0 % in the “Cost+” scenario) but higher when purely
(“Hyd”). optimising for cost (up to 24.4 % in the “Cost” scenario). It was not
possible to stay within planetary boundaries when relying only on
3. Results hydrogen (“Hyd” scenarios), even when using green hydrogen. Large-
scale electrification via heat pumps was thus identified as the only
This section reports the outcome from modelling the building heat absolutely sustainable option. Another condition to stay within plane
ing system in the EU27 + 1, including an overview of impacts and costs, tary boundaries is that limited political roadblocks exist regarding either
details about costs by scenario and end-user, and finally details about international trade or onshore wind power and nuclear energy, as their
technical characteristics of the system. Explanations of results are pro combination (“PG” scenarios) would transgress the climate change-
vided in this section, while their wider implications are discussed in related planetary boundaries. The use of predominantly blue hydrogen
Fig. 3. Overview of system-level results. The x-axis shows the transgression levels for nine planetary boundaries, total CO2 equivalent emissions, the impact objective
as the weighted transgression level, the total annualised system costs and the abatement or mitigation costs per ton of CO2 derived from the difference in costs and
GHG emissions compared with the BAU. The y-axis shows the different scenarios, which are optimised either by least cost (Cost), impact (WTL), and, when all
transgression levels < 1, for the WTL solution also by boundary-constrained least-cost (Cost + ). Scenarios have differing restrictions, which are noted after the
optimisation type description; “no-col” does not allow for trade, “no-N/O” does not allow for nuclear or onshore wind, “PG” combines the two previous restrictions
and “Hyd” allows only hydrogen for heating. Colour legends for the five different x-axis metrics are shown on the right.
8
(Cost_Hyd) reduces emissions compared with the BAU but still greatly doubles total system costs compared with the standard “Cost” scenario.
surpasses many boundaries. Note that the assumptions for life-cycle CO2 There are several reasons for this. Significantly more electricity gener
emissions from blue hydrogen production (3.78 kg CO2/kg H2) in this ation capacity is needed to satisfy the heating demand when green
work are on the lower spectrum of reported values [50,51]. Besides hydrogen is used (“WTL_Hyd”) compared with direct electrification (see
emissions, this scenario requires 0.68 Gt of CO2 storage capacity annu Fig. 6a) due to conversion and efficiency losses. Although transmission
ally. Although this is a fraction of the estimated 120 Gt storage capacity and interconnector costs are similar for the electrification vs hydrogen-
in the EU [52], this stored amount adds up every year, and there are only scenarios, the end-user cost is significantly higher due to the
numerous competing uses for the storage capacity (required CO2 in the hydrogen storage tanks required for individual buildings not connected
“Cost+” scenario is ~ 7 times lower). No significant burden-shifting to to the grid. This greatly increases total end-user cost; the average costs of
impact categories not related to climate change is observed, except for hydrogen produced and delivered are 2.93 and 3.62 €/kg in “Eco_Hyd”
freshwater use in the “Cost_Hyd” scenario due to the water used in the and 4.33 and 5.28 €/kg for “WTL_Hyd”. Due to widely differing as
steam methane reforming process. sumptions related to equipment and infrastructure costs represented by
A breakdown of impact drivers is shown in Supplementary Fig. 1 in the error bars, numerical results have to be considered with adequate
the SI. Renewable energy infrastructure is the impact driver in all sce uncertainty.
narios except for “Cost_Hyd”, where blue hydrogen production through The restriction of international trade (“no-col” scenarios) led to re
SMR with carbon capture is primarily responsible for the shown impacts. quirements for energy storage, with both CEAS, and thus slightly higher
From the GHG emission reductions and total system costs, the overall costs. This effect was even more pronounced when optimising for
abatement costs for climate change were calculated. The “Cost+” sce the least TL. It can be derived from this that energy storage reduces the
nario is at ~ 210 €/ton (i.e. the carbon price required for cost parity). impact of electricity production at a higher total system cost. The cost
The higher costs for hydrogen-only scenarios, paired with smaller increase due to prohibiting onshore wind and nuclear (with a shift to
emission reductions, result in a very high abatement cost between more expensive offshore wind) was smaller compared with trade re
around 1450 and 1600 €/ton for hydrogen-only scenarios. strictions whether optimising for cost or impact. In general, optimising
for impact did increase the total system cost to some extent but less than
3.2. Breakdown of costs by scenario and end-user type expected, especially for the political gridlock scenarios.
The cost per kWh of heating varies between the different end-user
Fig. 4 shows the breakdown of the total annualised costs. The pro types as shown in Fig. 5 for a selection of the scenarios. For electrifi
posed decarbonised system (recall that the alternative systems need to cation, end-user types in denser urban environments not connected to
meet the 0.068 Gt/year constraint on CO2 emissions, ~1% of the global district heating have the least cost, while non-grid connected (often
boundary) is significantly more expensive than the current system rural) buildings have slightly higher costs. It was found that the high
(BAU) for all scenarios. The main driver of the total cost is the end-user investment cost of ground-source heat pumps used for district heating
equipment (heat pumps). The HRE scenario is slightly cheaper results in higher costs, not making up for the efficiency gains compared
compared with the BAU case due to the switch from oil and coal to less with air-source heat pumps. For hydrogen, there is a sharp increase in
expensive gas boilers. Relying only on hydrogen for heating more than cost for non-grid connected individual buildings due to storage tank
Fig. 4. Total system cost for building heating in the EU27 + 1 broken down by technologies and scenarios. Error bars describe the minimum and maximum cost
depending on differing cost assumptions of all the technologies and equipment considered.
9
Fig. 5. Total cost per unit of heating demand met for the different end-user types; connected to the gas (GRID) or district heating (DH) grids and individual buildings
(IND). Three different scenarios are shown, electrification-only (E-…) optimised by least cost, hydrogen-only (H-…) optimised by either least cost or impact (WTL).
Note that the electrification-only scenario differs from the “Cost” scenario shown in previous figures as hydrogen use was set to zero (24.4%, as shown in Fig. 6 for the
H2 end use share in the Cost scenario).
requirements. The hydrogen distribution by trucks to those houses was storage. The share of hydrogen from water electrolysis increases without
not more expensive per unit of heating delivered, mainly due to the collaboration (most relevant for the “WTL_Hyd” scenario), as not all
relatively high storage capacity for each house requiring fewer trips. countries can store CO2 captured during the SMR process for blue
Considering only grid-connected buildings (district heating or gas), the hydrogen. The generation capacity obtained in this study was consid
least-cost scenario for hydrogen heating – over 75 % from blue hydrogen ered additional to the power sector.
– is cost-competitive with electrification. Note, however, that this sce Supplementary Table 11 compares the additional capacity to that
nario is far from being absolutely sustainable as shown in Fig. 3. For projected for the power sector in 2040. In particular, for offshore wind
grid-connected buildings the results also reveal a contrasting picture turbines in the WTL case and all alternative scenarios, the additional
between electrification and hydrogen; a large share of the total system requirements present a multiple of the projected capacity, raising
cost has to be borne directly and initially by the end-user through questions about political feasibility. Further details about the monthly
equipment investments in the electrification scenario. Contrastingly for breakdown of electricity generation and the installed capacity per
the hydrogen-only scenarios, the majority of costs occur during country for all scenarios are described in section 5 of the Supplementary
hydrogen production and are only indirectly borne by the end-user Information (Supplementary Figs. 2-6). Country-level results are shown
through the energy bill. in Supplementary Figures 12 and 13.
4. Discussion
3.3. Detailed choice of technology and system configuration
The findings from this study have a range of implications discussed in
Fig. 6a shows the deployed primary generation (renewable elec
this section, followed by comments about the sensitivity of the results
tricity and SMR) capacity for all cases and scenarios and Fig. 6b selected
and some limitations. A comparison with other studies is provided in
system properties. The main technology used was wind energy and the
section 5 in the SI.
main difference when optimising for impact (WTL) compared with cost
was a larger share of offshore wind, the increase in nuclear power, and a
much smaller share of GasCCS. Solar PV capacity was higher when no 4.1. Results interpretation
collaboration was feasible and an even larger share was needed in the
scenario with both restrictions. Different shares of GasCCS and nuclear Hydrogen is too expensive and beyond the safe operating space:
power were chosen to provide the stabilising load. Hydrogen heating Direct electrification via heat pumps was strongly favourable for all
played a limited role (only for a small share of gas and district heating countries and irrespective of restrictions on capacity or collaboration.
grid-connected end-users) in the cost optimised scenarios with penalties The issues of hydrogen-based heating identified in this work are a) the
for transgression and no role in impact optimised scenarios (except strong trade-off between costs and impacts (“Cost_Hyd” versus
when it was forced in “Hyd” scenarios). The required electricity gener “WTL_Hyd”), which is primarily due to the choice of production method
ation capacity quadrupled for the hydrogen-only case when green (blue hydrogen or green hydrogen), and b) the high cost for non-grid
hydrogen was used, while SMRCCS capacity reduces this need in the cost connected users. Predominantly using blue hydrogen for grid-
optimised hydrogen-only scenario (the generation capacity of renew connected users is cost-competitive with electrification but is far from
ables is still higher than in all electrification scenarios). being absolutely sustainable when considering the occupation of safe
Focusing on Fig. 6b and the two standard scenarios, international operating space of the planetary boundaries, in particular climate
trade was chosen over energy storage to balance national supply and change. Producing hydrogen through renewable electricity and PEM
demand, and impact optimised scenarios relied much more on it. Energy electrolysis (green hydrogen) reduces impacts to almost acceptable
storage is chosen only when trade was not possible, which also restricted levels but is vastly uneconomical. The main explanation for higher costs
some countries’ ability to deploy GasCCS due to a lack of geological and impacts is the overall efficiency of hydrogen boilers (around 75 %
10
Fig. 6. a: Deployed primary generation capacity by technology for each scenario, Fig. 6b: Properties of the building heating system obtained for each scenario. The
first column describes the amount of heating demand covered by hydrogen, the second the share of hydrogen produced from electricity (green and purple) compared
with the total, the third column the total amount of electricity sent to storage (in kWh) compared with the total generation, and the fourth puts the amount of
electricity traded (in kWh) in perspective considering the total amount of electricity used in the system.
based on HHV, including production) compared to heat pumps (between political gridlock scenario. Interestingly, restricting onshore wind
200 and 400 %). This was known before [53] but this study revealed deployment due to concerns such as preserving landscape beauty ne
how much more costs and impacts are generated through purely cessitates more interconnectors, which are also typically built over
hydrogen-based heating when considering a larger union of nations. ground and often not considered aesthetic. This study did not find a
Collaboration and political support for technologies recom justification for deploying large-scale hydrogen infrastructure for sea
mended: Large-scale international trade of electricity combined with sonal storage for the building heating sector because seasonal storage
renewable electricity generation in the most favourable locations (e.g. was either not required or CEAS was considered more favourable due to
the North Sea for offshore wind and solar PV in Italy and Spain) has been a higher round trip efficiency and lower costs. High levels of cooperation
identified as a promising option to reduce overall system costs. The and social acceptance of certain technologies could also make large-
possibility to send CO2 through other countries in pipelines also had a scale battery storage obsolete for the heating sector.
considerable effect as many countries depend on CO2 export to ramp up Electrified heating for not exceeding the planetary boundaries:
electricity production in winter with dispatchable power such as GasCCS Previously stated projections by the EU (HRE4) are far from being
plants. Balancing supply and demand through a combination of trade absolutely sustainable in the most pressing categories, calling for more
and dispatchable generation means there was almost no need for energy ambitious proposals that consider planetary boundaries. A building
storage. Thus, expanding interconnector capacity and deploying GasCCS heating system design that would be mainly built on the set-up shown
and some nuclear might be cheaper than deploying large-scale energy for the Cost + scenarios would stay within all (downscaled) planetary
storage technology while staying within planetary boundaries. The non- boundaries, while providing substantial environmental co-benefits on
necessity of energy storage also applies when environmental impacts are top of climate change mitigation. Staying within allocated boundaries
even further reduced by replacing more GasCCS with nuclear power and for the most urgent impact categories was possible via heat pump-based
onshore wind with offshore wind and solar PV, provided countries are heating even without any collaboration or further expansion of onshore
still allowed to trade electricity and CO2. Regions like the EU can sub wind energy or nuclear (however, more GasCCS capacity is then
stantially benefit from working together on energy policy and engaging required). Electrification of building heating comes with many co-
with their citizens, as illustrated through the counterexample of the benefits as virtually all impact categories are improved substantially –
11
except freshwater use. Not all options can be considered (absolutely) distribution) was assumed to be used for electrical heating via heat
sustainable as even the use of expensive green hydrogen would violate at pumps. This percentage might also differ between countries and,
least two boundaries. depending on the intra-day demand spikes, be slightly higher or lower.
Financial support and targeted policies required: The higher A comparison with findings from other literature is discussed in
costs of a decarbonised and absolutely sustainable building heating section 6 in the SI and further limitations of the work are described in
system will likely be borne mainly by the consumer as heat pumps are section 7 in the SI.
significantly more expensive than boilers. This means energy policy
should a) accessibly communicate the systemic benefits of heat pumps, 5. Conclusion
b) incorporate incentives for consumers to switch and ease the financial
burden, e.g. through subsidies, and c) increase research & development The decarbonisation of the building heating sector of the EU27 + 1
efforts to make heat pumps cheaper and more efficient. In addition, any was assessed via a bottom-up greenfield model with monthly time pe
(geo-)political constraints would increase the cost of energy provision riods throughout a reference year in 2040, considering seasonal imbal
and thus also increase consumer prices. To prevent this, politicians ances and country-specific differences. The study’s main novelty is the
should not just aim to further collaboration but also aim to increase combination of an i) optimisation-based comparison of energy carriers,
acceptance of interconnectors, onshore wind farms, and potentially ii) an investigation of the effects of collaboration, and iii) a holistic
nuclear energy within their own populace (the issue of radioactive waste sustainability assessment in relation to the planetary boundaries. Con
and its treatment is not captured in this study). Further, richer countries trasting previous building heating studies, this work also included CCS
in the EU27 + 1 will likely have to help financing the transition in poorer technologies and the required CO2 transport and storage network, the
countries. Lastly, market-based mechanisms alone are likely insufficient assessment of large-scale hydrogen use for the EU27 + 1, and the
as the required CO2 price of around 210 €/tonne for cost parity, in the consideration of building-level equipment and inclusion of non-grid
best case, will likely not be reached well before 2040. connected buildings.
It was found that the building heating sector in the EU27 + 1 can be
4.2. Outcome dependencies of parametric assumptions: made sustainable in absolute terms through an expansion of renewable
electricity, in particular wind energy, combined with heat pumps. This
Heat pumps performance: The COP of air-to-water heat pumps was approach, however, may increase the cost by 38 % compared with the
calculated based on empirical equations derived for current state-of-the- current system. Further, it was shown that collaboration changes the
art heat pumps and was in heating-intensive months between 2 and 3. design of optimal systems and reduces costs and impact noticeably.
Other studies [17] assume higher values between 3 and 4 as efficiencies Conversely, large-scale hydrogen-based heating was not recommended
will increase with future developments (however, no empirical equa as there is a strong trade-off between costs, favouring blue hydrogen,
tions are available so far, and using a fixed value was not considered and impacts, with blue hydrogen being only slightly better than the
adequate for this study). This would lower the system cost significantly current vastly unsustainable system. Green hydrogen was shown to be
as the total electricity generation requirements drop and the required uneconomical due to a large amount of renewable energy infrastructure
capacity of heat pumps drops almost proportional to an increase in required, and the cost driver for hydrogen was generally high-end user
efficiency. costs for local hydrogen storage. It was also showed that building
Grid balancing options: The balancing of supply and demand heating based on green hydrogen transgresses several global warming-
through a range of energy storage and dispatchable technologies, as well related planetary boundaries. The need for energy storage can be sub
as imported electricity, was represented by Equation (1). The results are stantially reduced through international trade and GasCCS power plants,
somewhat dependent on the coefficients used, such as the back-up and the remaining storage requirements can be met predominantly
production coefficient (0.33, main influence) and the factors used for through CEAS. This work gives limited guidance on more granular as
batteries (2.7, only relevant in the PG scenario) and other types of pects, such as grid balancing or the locally optimum designs of heating
storage (1) and import (0.5). The latter were conservatively chosen and systems. Instead, it provides a ballpark estimate for policy-makers and
thus capacities might be slightly lower. Introducing demand-side mea energy professionals about expected costs and environmental impacts
sures might reduce the need for such balancing efforts, while persistent from developing the technologies covered here. Hence, further analyses
drops in wind speed throughout Europe that might occur around once in should be carried out to perform a detailed design guided by general
a decade might necessitate additional seasonal storage. insights like the ones discussed here.
Peak factor: The end-user and transport capacity was sized based on Potential further work building on these insights could include
the ratio between average daily heating demand and hourly peak de exploring synergies with the power sector through seasonal storage and
mand (~2), at which the capacity was defined. This ratio could be curtailment reduction, a more detailed intra-day grid balancing and
slightly higher in less well-controlled systems and slightly lower with stabilisation study for the proposed setup, and further investigation of
demand management. The effect would be a proportional change in end- more affordable hydrogen-based solutions for non-grid connected
user and transport costs. buildings. Future modelling studies considering international building
Heating demand: The total heating demand per country in the heating systems should include a range of storage options, such as CAES,
future is not certain. Historical trends show an increase in floor area and consider CO2 networks and potential geopolitical constraints on
a decrease in heating intensity per floor area. If demand drops further geological storage, and consider non-grid-connected users. Another
than assumed here (24 % for residential and 23 % for service buildings), avenue for research could cover policy instruments and consumer buy-in
due to stagnating or decreasing population levels and very ambitious for heat pump deployment. The complementary role of other renewable
building insulation improvements, total system costs, and the trans heat sources, such as solar thermal and geothermal energy in combi
gression level will be proportionally affected. nation with an expansion of district heating should also be explored.
Transmission and distribution networks: The output of the cost- As current projections were identified as highly unsustainable, a
minimised cases is sensitive to the cost of the sub-regional electricity strong need to plan and start implementing more sustainable building
distribution network and the assumptions regarding the shared use of heating systems is implied. The scenarios presented in this study provide
infrastructure with the power sector. The ratio of overhead to under an indication for policy-makers and energy systems planners about
ground cables of the EU in total was chosen, the latter being more than which pathways are most sensible to be explored further.
2.5 times as expensive; in reality, the percentage differs between
countries but should not change the results drastically. For the shared
use, 25 % of the infrastructure (regional transmission and sub-regional
12
6. Code availability statement [14] Ashfaq A, Kamali ZH, Agha MH, Arshid H. Heat coupling of the pan-European vs.
regional electrical grid with excess renewable energy. Energy 2017;122:363–77.
https://doi.org/10.1016/J.ENERGY.2017.01.084.
The GAMS code and the accompanying Excel input file to reproduce [15] Quarton CJ, Tlili O, Welder L, Mansilla C, Blanco H, Heinrichs H, et al. The curious
the results can be obtained from the authors upon reasonable request. case of the conflicting roles of hydrogen in global energy scenarios. Sustainable
Energy Fuels 2020;4(1):80–95.
[16] Hoggett, R. (2020). Hydrogen, heat, and the decarbonisation of energy systems (Issue
CRediT authorship contribution statement September).
[17] Victoria M, Zhu K, Brown T, Andresen GB, Greiner M. The role of storage
technologies throughout the decarbonisation of the sector-coupled European
Till Weidner: Conceptualization, Methodology, Software, Visuali
energy system. Energ Conver Manage 2019;201(June):111977. https://doi.org/
zation, Writing – original draft. Gonzalo Guillén-Gosálbez: Supervi 10.1016/j.enconman.2019.111977.
sion, Validation, Writing – review & editing. [18] Blanco H, Nijs W, Ruf J, Faaij A. Potential for hydrogen and Power-to-Liquid in a
low-carbon EU energy system using cost optimization. Appl Energy 2018;232
(June):617–39. https://doi.org/10.1016/j.apenergy.2018.09.216.
Declaration of Competing Interest [19] Sunny N, Dowell NM, Shah N. What is needed to deliver carbon-neutral heat using
hydrogen and CCS? Energ Environ Sci 2020;13(11):4204–24. https://doi.org/
10.1039/D0EE02016H.
The authors declare that they have no known competing financial [20] Chaudry M, Jayasuriya L, Blainey S, Lovric M, Hall JW, Russell T, et al. The
interests or personal relationships that could have appeared to influence implications of ambitious decarbonisation of heat and road transport for Britain’s
net zero carbon energy systems. Appl Energy 2022;305:117905. https://doi.org/
the work reported in this paper. 10.1016/J.APENERGY.2021.117905.
[21] Jalil-Vega FA, Hawkes AD. Spatially Resolved Optimization for Studying the Role
Data availability of Hydrogen for Heat Decarbonization Pathways. ACS Sustain Chem Eng 2018;6
(5):5835–42. https://doi.org/10.1021/ACSSUSCHEMENG.7B03970.
[22] Quarton CJ, Samsatli S. The value of hydrogen and carbon capture, storage and
Data will be made available on request. utilisation in decarbonising energy: Insights from integrated value chain
optimisation. Appl Energy 2020;257:113936. https://doi.org/10.1016/J.
APENERGY.2019.113936.
Acknowledgements [23] Baumgärtner N, Deutz S, Reinert C, Nolzen N, Kuepper LE, Hennen M, et al. Life-
Cycle Assessment of Sector-Coupled National Energy Systems: Environmental
This publication was created as part of NCCR Catalysis (grant Impacts of Electricity, Heat, and Transportation in Germany Till 2050. Front
Energy Res 2021:27. https://doi.org/10.3389/FENRG.2021.621502.
180544), a National Centre of Competence in Research funded by the
[24] Nijs, W., Castelló, P. R., & González, I. H. (2017). Heat Roadmap Europe: Baseline
Swiss National Science Foundation. scenario of the total energy system up to 2050. 695989.
[25] Steffen W, Richardson K, Rockström J, Cornell SE, Fetzer I, Bennett EM, et al.
Planetary boundaries: Guiding human development on a changing planet. Science
Appendix A. Supplementary data 2015;347(6223). https://doi.org/10.1126/science.1259855.
[26] Rosenow J. Is heating homes with hydrogen all but a pipe dream? Joule 2022;6
Supplementary data to this article can be found online at https://doi. (10):2225–8.
[27] Slorach PC, Stamford L. Net zero in the heating sector: Technological options and
org/10.1016/j.enconman.2022.116602.
environmental sustainability from now to 2050. Energ Conver Manage 2021;230:
113838. https://doi.org/10.1016/J.ENCONMAN.2021.113838.
References [28] Baldino, C., O’Malley, J., Searle, S., & Christensen, A. (2021). Hydrogen for heating?
Decarbonization options for households in the European Union in 2050. www.theicct.
org.
[1] European Commission. (2019). The European Green Deal, COM(2019) 640 final.
[29] Bjørn A, Hauschild MZ. Absolute versus Relative Environmental Sustainability.
https://ec.europa.eu/info/sites/default/files/european-green-deal-
J Ind Ecol 2013;17(2):321–32. https://doi.org/10.1111/J.1530-9290.2012.00520.
communication_en.pdf.
X.
[2] Bertelsen N, Mathiesen BV. EU-28 Residential Heat Supply and Consumption:
[30] Li Y, Lan S, Pérez-Ramírez J, Wang X. Achieving a low-carbon future through the
Historical Development and Status. Energies 2020;13(8):1894. https://doi.org/
energy–chemical nexus in China. Sustainable Energy Fuels 2020;4(12):6141–55.
10.3390/EN13081894.
https://doi.org/10.1039/D0SE01337D.
[3] Climate Watch. (n.d.). Historical GHG Emissions. Retrieved October 20, 2021, from
[31] Tian Z, Zhang S, Deng J, Fan J, Huang J, Kong W, et al. Large-scale solar district
climatewatchdata.org/ghg-emissions.
heating plants in Danish smart thermal grid: Developments and recent trends.
[4] Szabo J. Energy transition or transformation? Power and politics in the European
Energ Conver Manage 2019;189(October 2018):67–80. https://doi.org/10.1016/j.
natural gas industry’s trasformismo. Energy Research and Social. Science 2022;84
enconman.2019.03.071.
(June 2021):102391. https://doi.org/10.1016/j.erss.2021.102391.
[32] Patrizio P, Fajardy M, Bui M, Dowell NM. CO2 mitigation or removal: The optimal
[5] Holzmann A, Schmid E. Consumer behaviour in the residential heating sector in
uses of biomass in energy system decarbonization. IScience 2021;24(7):102765.
Austria: Findings from a bottom-up modelling approach. Energ Buildings 2018;
https://doi.org/10.1016/J.ISCI.2021.102765.
158:486–93. https://doi.org/10.1016/j.enbuild.2017.10.036.
[33] Calvo-Serrano R, Guo M, Pozo C, Galán-Martín Á, Guillén-Gosálbez G. Biomass
[6] Fisher DR, Nasrin S. Climate activism and its effects. Wiley Interdiscip Rev Clim
Conversion into Fuels, Chemicals, or Electricity? A Network-Based Life Cycle
Chang 2021;12(1):1–11. https://doi.org/10.1002/wcc.683.
Optimization Approach Applied to the European Union. ACS Sustain Chem Eng
[7] Hall LMH, Buckley AR. A review of energy systems models in the UK: Prevalent
2019;7(12):10570–82. https://doi.org/10.1021/ACSSUSCHEMENG.9B01115.
usage and categorisation. Appl Energy 2016;169:607–28. https://doi.org/
[34] Ó Broin E, Nässén J, Johnsson F. The influence of price and non-price effects on
10.1016/J.APENERGY.2016.02.044.
demand for heating in the EU residential sector. Energy 2015;81:146–58.
[8] Ruggiero F, Vlad LB, Vlăsceanu G, Gavriș A, Vasile D. Energy Cooperation – The
[35] Bussieck MR, Meeraus A. General Algebraic Modeling System (GAMS). Springer,
Strength of the EU’s Economic Development. Amfiteatru Economic Journal 2015;
Boston, MA 2004. https://doi.org/10.1007/978-1-4613-0215-5_8.
17(40):1080–94. https://www.econstor.eu/handle/10419/168966.
[36] Galán-Martín, Á., Vázquez, D., Cobo, S., Mac Dowell, N., Caballero, J. A., &
[9] Kozarcanin S, Hanna R, Staffell I, Gross R, Andresen GB. Impact of climate change
Guillén-Gosálbez, G. (2021). Delaying carbon dioxide removal in the European
on the cost-optimal mix of decentralised heat pump and gas boiler technologies in
Union puts climate targets at risk. Nature Communications 2021 12:1, 12(1), 1–12.
Europe. Energy Policy 2020;140:111386. https://doi.org/10.1016/J.
https://doi.org/10.1038/s41467-021-26680-3.
ENPOL.2020.111386.
[37] Monforti F, Gaetani M, Vignati E. How synchronous is wind energy production
[10] Brown T, Schlachtberger D, Kies A, Schramm S, Greiner M. Synergies of sector
among European countries? Renew Sustain Energy Rev 2016;59:1622–38. https://
coupling and transmission reinforcement in a cost-optimised, highly renewable
doi.org/10.1016/J.RSER.2015.12.318.
European energy system. Energy 2018;160:720–39. https://doi.org/10.1016/J.
[38] Ryberg MW, Owsianiak M, Richardson K, Hauschild MZ. Development of a life-
ENERGY.2018.06.222.
cycle impact assessment methodology linked to the Planetary Boundaries
[11] Samsatli S, Samsatli NJ. The role of renewable hydrogen and inter-seasonal storage
framework. Ecol Ind 2018;88(November 2017):250–62. https://doi.org/10.1016/
in decarbonising heat – Comprehensive optimisation of future renewable energy
j.ecolind.2017.12.065.
value chains. Appl Energy 2019;233–234:854–93. https://doi.org/10.1016/J.
[39] Galan-Martın A, Tulus V, Dıaz I, Pozo C, Perez-Ramırez J, Guillén-Gosálbez G.
APENERGY.2018.09.159.
Sustainability footprints of a renewable carbon transition for the petrochemical
[12] Zhu K, Victoria M, Brown T, Andresen GB, Greiner M. Impact of CO2 prices on the
sector within planetary boundaries. One Earth 2021;4:565–83. https://doi.org/
design of a highly decarbonised coupled electricity and heating system in Europe.
10.1016/j.oneear.2021.04.001.
Appl Energy 2019;236:622–34. https://doi.org/10.1016/J.
[40] Tulus V, Pérez-Ramírez J, Guillén-Gosálbez G. Planetary metrics for the absolute
APENERGY.2018.12.016.
environmental sustainability assessment of chemicals. Green Chem 2021;23(24):
[13] Gallo Cassarino T, Barrett M. Meeting UK heat demands in zero emission renewable
9881–93. https://doi.org/10.1039/D1GC02623B.
energy systems using storage and interconnectors. Appl Energy 2022;306:118051.
https://doi.org/10.1016/J.APENERGY.2021.118051.
13
[41] Lucas E, Guo M, Guillén-Gosálbez G. Optimising diets to reach absolute planetary [47] Guillen-Gosalbez G, Caballero JA, Esteller LJ, Gadalla M. Application of life cycle
environmental sustainability through consumers. Sustainable Production and assessment to the structural optimization of process flowsheets. Computer Aided
Consumption 2021;28:877–92. Chemical Engineering 2007;24:1163–8. https://doi.org/10.1016/S1570-7946(07)
[42] Ryberg MW, Andersen MM, Owsianiak M, Hauschild MZ. Downscaling the 80218-5.
planetary boundaries in absolute environmental sustainability assessments – A [48] Ehrenstein M, Galán-Martín Á, Tulus V, Guillén-Gosálbez G. Optimising fuel supply
review. J Clean Prod 2020;276:123287. chains within planetary boundaries: A case study of hydrogen for road transport in
[43] OECD. (2021). HC1.1 Housing-related expenditure of households. http://oe.cd/ahd. the UK. Appl Energy 2020;276(June):115486. https://doi.org/10.1016/j.
[44] Weidner T, Galán-Martín Á, Ryberg MW, Guillén-Gosálbez G. Energy systems apenergy.2020.115486.
modeling and optimization for absolute environmental sustainability: current [49] Werner, S. (2016). European District Heating Price Series.
landscape and opportunities. Comput Chem Eng 2022;164:107883. https://doi. [50] Parkinson B, Balcombe P, Speirs JF, Hawkes AD, Hellgardt K. Levelized cost of CO
org/10.1016/j.compchemeng.2022.107883. 2 mitigation from hydrogen production routes. Energ Environ Sci 2019;12(1):
[45] Hjalsted AW, Laurent A, Andersen MM, Olsen KH, Ryberg M, Hauschild M. Sharing 19–40.
the safe operating space: Exploring ethical allocation principles to operationalize [51] Bauer C, Treyer K, Antonini C, Bergerson J, Gazzani M, Gencer E, et al. On the
the planetary boundaries and assess absolute sustainability at individual and climate impacts of blue hydrogen production. Sustainable Energy Fuels 2021;6(1):
industrial sector levels. J Ind Ecol 2021;25(1):6–19. https://doi.org/10.1111/ 66–75. https://doi.org/10.1039/D1SE01508G.
jiec.13050. [52] Pozo C, Galán-Martín Á, Reiner DM, Mac Dowell N, Guillén-Gosálbez G. Equity in
[46] Guillén-Gosálbez G, Grossmann IE. Optimal design and planning of sustainable allocating carbon dioxide removal quotas. Nat Clim Chang 2020;10(7):640–6.
chemical supply chains under uncertainty. AIChE J 2009;55(1):99–121. https:// https://doi.org/10.1038/s41558-020-0802-4.
doi.org/10.1002/AIC.11662. [53] Trask, A., Hanna, R., & Rhodes, A. (2022). The future of home heating. https://www.
mcscharitablefoundation.org/.
14

2023-Swiss-Planetary Boundaries Assessment of Deep Decarbonisation Options For Building Heating in The European Union

Uploaded by

Copyright:

Available Formats

2023-Swiss-Planetary Boundaries Assessment of Deep Decarbonisation Options For Building Heating in The European Union

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

2023-Swiss-Planetary Boundaries Assessment of Deep Decarbonisation Options For Building Heating in The European Union

Uploaded by

Copyright:

Available Formats

Energy Conversion and Management 278 (2023) 116602

Contents lists available at ScienceDirect

Energy Conversion and Management

Planetary boundaries assessment of deep decarbonisation options for

1. Introduction 1 (including the United Kingdom), constituting roughly 13.3 % of total

ables, leading to an MILP formulation. However, numerical examples + ecurtail

pacts for all infrastructure items. (ecomp,trans

total amount of electricity imported (etrans

either coal (etech

ficiencies (CAPEFFCOAL , 87 % and CAPEFFCCGT , 86 %). (co2store

2.3.2. Technical constraints

You might also like