This is a repository copy of Estimation of global final-stage energy-return-on-investment
for fossil fuels with comparison to renewable energy sources.
White Rose Research Online URL for this paper:
http://eprints.whiterose.ac.uk/148748/
Version: Accepted Version
Article:
Brockway, PE orcid.org/0000-0001-6925-8040, Owen, A orcid.org/0000-0002-3872-9900,
Brand Correa, LI et al. (1 more author) (2019) Estimation of global final-stage energy-
return-on-investment for fossil fuels with comparison to renewable energy sources. Nature
Energy, 4 (7). pp. 612-621. ISSN 2058-7546
https://doi.org/10.1038/s41560-019-0425-z
© The Author(s), under exclusive licence to Springer Nature Limited 2019. This is an
author produced version of an article published in Nature Energy. Uploaded in accordance
with the publisher's self-archiving policy.
eprints@whiterose.ac.uk
https://eprints.whiterose.ac.uk/
Reuse
Items deposited in White Rose Research Online are protected by copyright, with all rights reserved unless
indicated otherwise. They may be downloaded and/or printed for private study, or other acts as permitted by
national copyright laws. The publisher or other rights holders may allow further reproduction and re-use of
the full text version. This is indicated by the licence information on the White Rose Research Online record
for the item.
Takedown
If you consider content in White Rose Research Online to be in breach of UK law, please notify us by
emailing eprints@whiterose.ac.uk including the URL of the record and the reason for the withdrawal request.
1
Estimation of global final stage energy-return-on-investment for fossil fuels with 1
comparison to renewable energy sources 2
Paul E. Brockway
a,
*, Anne Owen
a
, Lina Brand-Correa
a
, Lukas Hardt
a
3
a
Sustainability Research Institute, School of Earth and Environment, University of Leeds, Leeds LS2 9JT, UK 4
* Corresponding author: p.e.brockway@leeds.ac.uk. Tel.: +44-113-343-2846 5
Fossil fuels under many scenarios remain the dominant energy source to at least 2050. However, harder-6
to-reach fossil fuels require more energy to extract and hence are coming at an increasing ‘energy cost’. 7
Associated declines in fossil fuel energy-return-on-investment ratios at first appear of little concern, given 8
published estimates for oil, coal and gas sources are typically above 25:1. However, such ratios are 9
measured at the primary energy stage, but should be estimated instead at the final energy stage (e.g. 10
electricity, petrol) where energy enters the economy. Here, we calculate global time-series (1995-2011) 11
energy-return-on-investment ratios for fossil fuels at both primary and final energy stages. We concur 12
with common primary-stage estimates (~30:1), but find very low ratios at the final stage: around 6:1, and 13
declining. This implies fossil fuel energy-return-on-investment ratios may be much nearer to those of 14
renewables and could decline precipitously in the near future. 15
16
The field of net energy analysis first came to prominence during the 1970s oil crises
1–4
as a means of assessing 17
how much energy is delivered to society. Various metrics have emerged
5
including energy profit ratio, energy 18
gain, energy payback, and the most well-known ‘energy-return-on-investment’ (EROI). Kunz et al.
6
define EROI 19
in its simplest form as a ratio which “divides the total energy output by the energy input”. Several factors have 20
contributed to increasing attention being paid to the EROI research field. First, there are concerns over 21
declining EROI ratios of fossil fuels – which under many scenarios remain the dominant energy source to at 22
least 2050
7
– due to depletion of finite reserves
8,9
. Second, the estimated EROI ratios for renewable energy 23
sources are often contentious, vary greatly depending on adopted methodology, and are commonly estimated 24
as lower than fossil fuels
10
. Concerns follow that the renewables-led energy transition required to meet climate 25
targets
11
may have adverse socio-economic impacts
12
. Third, EROI as a topic has become more accessible 26
through the readily-visualised concepts of a ‘net energy cliff’
9
– where available net energy declines 27
precipitously below EROI ratios of 5:1 - and a minimum threshold level of societal-level EROI
13,14
. 28
However, much of the increased attention is confined to academic circles. One reason may be that fossil fuel 29
EROI is commonly estimated at the primary (energy source) stage, where EROI ratios (i.e. for oil, coal and gas) 30
are high, typically over 25:1
8,15
. Such ratios suggest to modellers and policy makers that EROI ratios won’t fall 31
below a threshold of concern until well into the renewables transition
12
. However, this is a misleading 32
perception, as instead, fossil fuel EROI should be estimated at the final (energy carrier) stage (e.g. electricity, 33
gas, and petrol), where energy enters the economy. This enables a fairer comparison to renewables-based 34
EROI estimates, and the platform for improved energy and climate policy. 35
We build on recent EROI research
10,15,16
to provide an estimate of global fossil-fuel based EROI at a final energy 36
stage, which better matches that of renewables-based EROI. We combine national-level International Energy 37
Agency (IEA) energy data with a multi-regional input-output (MRIO) approach to include a wider boundary of 38
direct energy production sectors and associated indirect (supply-chain) energy impacts, including trade. To 39
enable comparison to existing methods and EROI ratios, we estimate global fossil fuel EROI for both primary 40
(
) and final (
) energy stages, and also provide time-series estimates for the 1995-2011 41
period. Our results indicate that by 2011 global ratios for
(~6:1) are much lower than
42
(~30:1), and both are declining. Two implications follow. First, EROI of fossil fuels may be much nearer to 43
renewables than commonly supposed, meaning a global renewables transition may not be as biophysically 44
troublesome as previously thought. Second, the low and declining
ratios for fossil fuels provides an 45
immediate concern, and also implies we are much nearer a ‘net energy cliff’ than previously thought, where 46
the non-linearity of EROI means low ratios (below 5:1) quickly restrict available net energy to society.
47
2
Global fossil fuel EROI based on a final energy stage
48
Cleveland et al.’s
17
landmark study in the 1980s estimated EROI in the United States for fossil fuels at the ‘well 49
head’ (oil and gas) and ‘mine mouth’ (coal). Since then, many fossil fuel EROI studies have been published
15
, 50
though largely these remain at the primary energy stage (as coal, oil, gas). The most common exceptions are 51
fossil fuel based EROI estimates of electricity, which are at the final energy stage. However, their 52
methodologies (and hence estimates) vary, with some (e.g. ref.
10,15
) taking primary stage EROI estimates and 53
applying direct (thermal) loss factors in conversion to electricity, while others (e.g. ref.
18
) use LCA-based 54
methods to include both thermal losses and supply-chain energy investment. 55
At the same time, an increasing number of studies are estimating EROI ratios for modern renewables, 56
particularly electricity generated from photovoltaics (PV) and wind turbines – which are seen as two energy 57
technologies pivotal
11
for reductions in global greenhouse gas emissions. 58
A summary of the different EROI estimates for these two different energy sources (fossil fuels and renewables) 59
at primary and final energy conversion stages is given in Table 1: 60
Table 1: Comparison of EROI ratio estimates for different energy sources/carriers and conversion stages 61
Energy source /
carrier
Published estimates of EROI ratios (X:1)
at different energy conversion stages
Primary energy stage
(EROI
PRIM
)
Final energy stage
(EROI
FIN
)
Reference
Coal
40 – 55 (mine mouth)
80 (mine mouth)
Hall et al.
15
Court and Fizaine
19
Oil
15 (well head)
18 (well head)
20 (well head)
4-5 (refined oil fuels)
Court and Fizaine
19
Gagnon et al.
8
Hall et al.
15
Brandt
20
Gas
18 (well head)
20 (well head)
75 (well head)
Gagnon et al.
8
Hall et al.
15
Court and Fizaine
19
Electricity (gas)
6
Δ
8
Δ
11⁰ – 14⁰
Hall et al.
15
King and Van Den Bergh
10
Raugei and Leccisi
18
Electricity (coal)
4⁰
13
Δ
– 18
Δ
17
Δ
Raugei and Leccisi
18
Hall et al.
15
King and Van Den Bergh
10
Electricity (PV)
19* – 38*
6⁰ – 12⁰
10
Δ
4⁰ - 20⁰
Raugei et al.
21
Hall et al.
15
Leccisi et al.
22
Electricity (Wind)
14⁰ – 26⁰
15⁰ – 30⁰
Kubiszewski et al.
23
Raugei and Leccisi
18
* ‘Primary energy equivalent’ value by Raugei et al.
21
, estimated by dividing EROI
FIN
value for PV (6-12) by the EU-27 62
electric grid efficiency
= 0.31. 63
Δ
includes power plant / transformational conversion efficiencies only 64
⁰ includes power plant / transformational conversion efficiencies AND supply chain energy investments 65
66
Table 1 reveals the divergence between the modal EROI estimates: primary-stage fossil fuels are much higher 67
(typically 20:1-80:1) than final-stage renewable electricity (typically 5:1-20:1). This creates a potentially 68
misleading perception to modellers of high fossil fuel EROI and low renewables EROI. Raugei
24
warns such 69
apples-to-oranges comparisons are flawed, as they “compare [energy] carriers that cannot be put to similar 70
end-use”. As energy-economy models are now starting to include EROI within their analytical framework, this 71
has the potential to lock-in bias towards fossil fuels. For example, Sers and Victor
12
suggest an ‘energy-72
emissions trap’ is approaching, as “Reducing emissions will necessitate the transition from relatively high EROI 73
dispatchable fossil fuels to […] relatively low EROI intermittent renewables”. 74
3
Sers and Victor’s reference to ‘intermittent’ renewables highlights an important point: estimation of EROI for 75
renewables is a much newer field with a host of interwoven issues ongoing for the mainly life cycle analysis 76
(LCA) based methodology, including capital investment, payback times, and intermittency. 77
Whilst resolving such renewables-EROI issues are worthy and should continue, we suggest the heavy focus on 78
them has distracted from the equally pressing need to move fossil fuel EROI to the final energy calculation 79
stage. This is important for two key reasons. First, incumbent fossil fuels remain important: the 80
Intergovernmental Panel on Climate Change (IPCC) future scenarios assume they remain as the dominant 81
source of energy to at least 2050
25
. Second, EROI increasingly is being included in energy-economy models as 82
noted earlier, to help study future energy transitions
10,26
and their macroeconomic impacts
12,27,28
. As energy 83
enters the productive economy at the final energy stage, EROI ratios at the same final energy stage are thus 84
also in the correct format for inclusion in energy-economy models. 85
Matching fossil-fuel EROI estimates to the same (final) energy stage and (economy-wide or global) scale as 86
energy-economy models is therefore important. Currently, economy-wide and global fossil-fuel EROI 87
estimates – excluding electricity as seen in Table 1 – remain at a primary energy stage, using either a site-level 88
or price based approach. Lambert et al.
9
provide an example of the site-level approach, collating sample 89
studies from different countries at the ‘mine mouth’ and ‘well head’. The price-based approach typically 90
involves using energy prices and/or expenditure data to estimate direct and indirect (including capital) energy 91
investment
29
: Gagnon et al.
8
, Court and Fizaine
19
and Guilford et al.
30
provide examples. King et al.
31
provide 92
another route, using total energy expenditure to estimate aggregate EROI. 93
Analytical approach 94
We build on previous work by Brand-Correa et al.
16
, developing an input-output based approach to estimate 95
global fossil fuel EROI at final and primary energy stages for the period 1995-2011. Two key advances underpin 96
the method. First, we use International Energy Agency (IEA) extended energy balances time-series data
32
. This 97
provides us with country-level data for fossil fuel energy produced at both a primary and final energy stage. It 98
also allows access to the direct energy use for energy production sectors at both primary energy (e.g. coal 99
mines) and final energy (e.g. oil refineries, coke production, coal gasification) stages. Second, we use EXIOBASE 100
– a large global multi-regional input-output (MRIO) database
33
- to estimate the indirect ‘supply-chain’ energy 101
associated with production of fossil fuel energy at both primary and final energy stage, including trade. 102
We adopt a net EROI (
) calculation basis, as given in equation (1), where net energy output is equal to 103
the gross (or total) energy output, minus the energy input. This aligns firstly with other net energy 104
research
6,10,34
, which focuses on the energy that enters the productive economy, and secondly with our final 105
energy data which is already in net energy terms. (Note, net EROI = gross EROI – 1 as shown in Methods. 106
Therefore our results are applicable to both definitions, and we remove the Net suffix hereafter). 107
(1)
Our conceptual boundaries for calculating global
and
are set out in Figure 1: 108
4
109
Figure 1: Conceptual framework for global fossil fuels energy-return-on-investment estimation.
is aggregate 110
fossil fuel EROI at the primary energy conversion stage.
denotes aggregate fossil fuel EROI at the primary energy 111
conversion stage. Direct energy invested is the energy consumed in production, transformation and distribution of 112
energy. Indirect energy invested is the supply-chain embodied energy in products that are used in production, 113
transformation and distribution of energy.
114
Referring to Figure 1, we calculate the components of
and
via equations (2) and (3), which 115
adapt equation (1) to the primary and final energy stages: 116
(2)
=
(3)
The calculations proceed in three steps, in turn calculating net energy produced; and the direct energy (
) 117
and indirect energy (
) invested in energy production. These are outlined next, with further detail provided 118
in Methods (including modelling limitations). 119
From equation (2), the net energy produced for
is equal to the total gross production of primary 120
energy
, minus the direct energy used by the energy extraction industries (
. We obtain 121
values for
from the ‘production’ energy data from the IEA extended world energy balances
32
. In IEA 122
terms, this is a sub-category of ‘total primary energy supply’, and represents the primary energy extracted in 123
each country/region, including exports, but excluding imports. As we use the EXIOBASE MRIO database in our 124
analysis, we do not need to account for imported energy, as these flows will be reflected in the MRIO 125
transaction matrices. 126
Next, in equation (3), we directly obtain the net energy produced at the final energy stage via the ‘Total Final 127
Consumption’ (TFC) provided in the IEA extended world energy balances
32
. This is equal to the gross final 128
energy (for fossil fuels) used by each country/region
), minus the direct energy
associated 129
with the production of the final energy. 130
