Kevin M. A.
Parker
*a and
Michael R.
Mainelli
b
aKKI Associates Ltd, Science and Business Advisor Z/Yen Group, 1 King William Street, London EC4N 7 AF, UK. E-mail: kevin@kkitech.com
bZ/Yen Group, Lord Mayor of the City of London, Mansion House Walbrook, London EC4N 8BH, UK
First published on 24th January 2024
Our estimates show that ‘proven reserves’ of fossil fuels in 2022, would generate an estimated 4777 Gt of CO2 after allowing for non-fuel uses. This quantity already could ‘bust CO2 budgets’ for IPCC RCP2.6, RCP4.5, and RCP6.0 and is approaching the range for RCP8.5. Notwithstanding these results, fossil fuel companies are still exploring and bringing new reserves onstream. We discuss the reasons behind this, and propose some policy options for governments as they address this situation.
Environmental significanceIn 2006 the authors posed a question to a major oil company, and members of the wider financial community. What happens if we burn all the oil you have now? The answers indicated that the world already had sufficient fossil fuel reserves to raise atmospheric carbon to unacceptable levels of 1000 ppm or more. It looked like there was a lot of ‘unburnable carbon’ or ‘stranded assets’ on Oil & Gas company balance sheets. The intention of this paper is improve the 2006 estimates, and carry out new calculations based on reserves in 2022. Is there any evidence that the concept of ‘unburnable carbon’ has in fact led to reduction in fossil fuel reserves? What might happen to IPCC carbon budgets, and atmospheric CO2 levels, if these reserves are in fact very ‘burnable’? Our estimates show that ‘proven reserves’ of fossil fuels in 2022, would generate an estimated 4777 Gt of CO2 after allowing for non-fuel uses. This quantity already could ‘bust CO2 budgets’ for IPCC RCP2.6, RCP4.5, and RCP6.0, and is approaching the range for RCP8.5. Notwithstanding these results, fossil fuel companies are still exploring and bringing new reserves onstream. We discuss the reasons behind this, and propose some policy options for governments as they address this situation. The authors for the paper, Dr Kevin Parker, and Professor Michael Mainelli, are respectively Science Advisor and Founder of the leading city of London think-tank Z/Yen Group. Among other achievements, Professor Mainelli is currently serving as Lord Mayor of the City of London, where he is making the Financing of Green developments as a theme of his year in office. |
The London Accord1 was an agreement by over 100 researchers from 25 organisations around the world to “share investment research with policy-makers and the public” on climate change for the sake of the planet. The organisations included the investment research arms of several global banks, research firms, and academics. The London Accord's thinking began at Z/Yen Group, was conducted under the auspices of the City of London Corporation with the support of the then Lord Mayor, Sir David Lewis, and had significant operational support from British Petroleum (BP).
In December 2007, the London Accord group published a 780 page ‘open source’ research report on the ‘investability’ of climate change.2 Their methodology involved extensive Monte Carlo portfolio modelling as used in financial markets rather than macroeconomics, social cost models, or shadow pricing. The report's conclusion was that carbon prices somewhere above $30 to $60 per tonne of CO2 (in 2007 circa €1 = $1.4, i.e. €21 to €43 per tonne) would provide sufficient opportunities for investment and asset managers to prevent climate change so long as governments restricted emission permits above the $30 to $60 range. Given that, traditional financial services could do much of the work to prevent climate change.
For reference, the then price of carbon allowances on the EU emissions trading system (EU ETS) had hit a peak of almost €30 per tonne in April 2006, but due to oversupply by governments, the price dropped 54% from €29.20 to €13.35 in the last week of April 2006. At the time of the London Accord's report launch, carbon prices were nearly zero.
Perhaps the most referred to reference point during the period of the report was the concentration of carbon dioxide in the atmosphere in parts-per-million (ppm). In 2005–6 it was approximately 370, and the likelihood exceeding 400 ppm around 2015 was seen as a significant, negative, milestone. Later, a more common reference point was ‘degrees warming’ which relies on models. Science advisers on the Intergovernmental Panel on Climate Change estimated an atmospheric CO2 concentration of no more than 450 parts per million for 2 degrees of warming, or 430 ppm for 1.5 degrees. Still later, ‘net zero’ dates and ‘carbon budgets’ have become the more common reference. This change derived from the recognition that global warming is closely tied to cumulative carbon emissions, thus directly linking historic and future reserves exploited over time.
During the London Accord work, researchers queried BP's involvement. Shouldn't the traditional ‘seven sisters’ of oil just stop? Wouldn't BP be better to close itself down than spend time on the project? At the time BP believed that its reserves might only change atmospheric CO2 by 1.5 to 2 ppm, and that the same would be true of the small number of other oil majors. Where would the additional 75+ ppm (450–370) come from to heat the planet by 2 degrees?
In 2006 BP's team and Z/Yen researchers independently came up with some new estimates, largely from the reserves of the national oil companies, of the CO2 levels in ppm from burning the then known fossil fuel reserves (Fig. 1).3
At the time, fracking and shale gas was in its infancy. With fracking and shale gas the numbers could rise much higher. It is fair to say that these numbers were considerably higher than expected at the time.
This led to an obvious question, if it would be impossible to ‘burn’ reserves above 450 ppm or 500 ppm or 550 ppm because society would not allow it, how could the reserves above that level be on corporate balance sheets at current market prices? If one assumed for the sake of example that 450 ppm was the level at which an overheated society would no longer tolerate fossil fuel usage, then 1222 − 450 = 772 that couldn't be burnt. In rough terms at that time, this meant that up to 92% of listed assets had no value.
These discussions were shared with central bankers as a systemic stability issue, especially once they realised the extent to which many pension funds relied on income from fossil fuel giants like BP and Shell.4 This led to the formation of the think-tank Carbon Tracker, who produced their first report in 2011,5 coining the phrases ‘unburnable carbon’ and ‘stranded assets’. The financial community, especially UK pension funds, have widely accepted these concepts.
Meanwhile environmental researchers have realised the threat that large fossil fuel reserves pose to the prospect of limiting climate change. Researchers are attempting to quantify the discrepancies between the carbon budgets for various IPCC scenarios and the CO2 outputs implicit in fossil fuel reserves, pointing out that the latter are much larger than the former.
The work carried out by Z/Yen in 2006 was an estimate based on the limited knowledge of the time. The intention of this paper is to improve the original estimates, and carry out new calculations based on reserves in 2022. Key Questions to be addressed include:
• Is there any evidence that the concept of ‘unburnable carbon’ has in fact led to a reduction in fossil fuel reserves?
• What might happen to IPCC carbon budgets, and atmospheric CO2 levels, if these reserves are in fact very ‘burnable’?
• Do the comments around carbon pricing at $30–60 per tonne still apply?
• What are the implications of the current situation for both financial organisations and policy makers?
We leave to other writers the questions around how policy changes such as ‘net zero by 2050’ targets might reduce CO2 emissions. Our focus is instead on the question ‘what is the outcome if these policies are not implemented or are unsuccessful?’
The IPCC have produced a number of climate scenarios, which relate potential climate change to various carbon budgets. They are described in the comprehensive summary for policy makers as part of the 2018 AR5 report7 They described a series of ‘representative concentration pathways’ (RCP2.6, RCP4.5, RCP6.0 and RCP8.5) which specify radiative forcing values between 2.6 and 8.5 W m−2 in the year 2100. These pathways lead to warming in the 2081–2100 period of 1.6 C, 2.4 C, 2.8 C, and 4.3 C respectively (all estimates ± 0.7 C) (Fig. 2).
Carbon budgets are not fixed numbers, but correspond to broad estimates of the probabilities of temperature outcomes, which are being constantly studied and revised. However the overall trend is clear, that budgets are reducing as CO2 emissions continue at increasingly high levels. The Global Carbon Project8 in 2022 commented that ‘The remaining carbon budget for a 50% likelihood to limit global warming to 1.5, 1.7, and 2 °C has, respectively, reduced to 105 GtC (380 GtCO2), 200 GtC (730 GtCO2), and 335 GtC (1230 GtCO2) from the beginning of 2023, equivalent to 9, 18, and 30 years, assuming 2022 emissions levels.’
We will use the numbers listed in the budgets above to compare with the potential CO2 emissions from current, proven fossil fuel reserves.
The IPCC RCP pathways can be visualised in a number of different ways. One approach (Fig. 3) relates the pathways to increasing atmospheric CO2.
Fig. 3 Atmospheric CO2 levels for IPCC scenarios RCP2.6 to RCP8.5 (IPCC AR5 data) (Wikimedia Creative Commons License). |
Another approach (Fig. 4) is to relate the pathways to cumulative anthropogenic CO2 emissions.
Fig. 4 Cumulative CO2 emissions: scenarios RCP2.6 to RCP8.5 (IPCC AR5 SPM Fig. 10). |
The Fossil Fuels registry has pointed out that current reserves would already exceed the carbon budgets for a 1.5 °C rise in temperature, and that the reserves of certain countries, including the US, could exceed IPCC targets by themselves.12,13
In 2015, Greenstone and Stewart calculated the impact of world fossil fuel reserves as potentially producing 5520 GtCO2 emissions, and used this to estimate consequential temperature increases.14,15
In 2016, Heede, (a pioneer in this area), and Oreskes analysed the reserves data of the largest seventy eight companies and state entities involved in oil, gas and coal extraction. They found that the reserves of these organisations would produce over 160% of the 2 °C carbon budget pathway from IPCC AR5.16
In contrast to the above studies, Wang et al. argued in 2016 that total CO2 emissions and consequent climate change would be limited by supply-side effects, notably that ‘recoverable reserves’ (oil that could be produced) were usually considerably less than ‘total reserves’ that appeared in estimates and balance sheets17 They expect fossil fuel production to reach a supply-limited peak around the mid-21st century. Incorporating this in their models limits the maximum atmospheric concentration of CO2 to 610 ppm by 2100, with a corresponding temperature rise of 2.6 °C.
There are considerable variations and hence uncertainties in these estimates, as different countries and companies have different definition of proven reserves. The BP review of World Energy states, for example that ‘The data series for proved oil reserves in this year’s review does not necessarily meet the definitions, guidelines and practices used for determining proved reserves at company level’.18 Coal reserves, even in the USA, are subject to wider variations in estimation methodology, as discussed on Global Energy Monitor19
The variations increase in ‘unconventional’ reserves such as shale oil and gas. A thorough review of these uncertainties was carried out by McGlade in 2012.20 While some producers exaggerate their reserves for political reasons, using proven reserves leads to generally conservative estimates. In addition, changes in technology can move potential reserves to proven reserves, especially in unconventional sources and new regions (such as the offshore Arctic). We discuss this further in our summary below.
‘It was initially believed that the ocean was the main sink of carbon until it became clear that a land sink was needed to close the carbon budget.25It is now admitted that both the land and the ocean play a comparable role, each removing from the atmosphere about 25% of the anthropogenic CO2emissions.4,26,27
However, the obvious question is whether the ocean and the land ecosystems will continue to provide such a service to humanity, removing about half of the CO 2 emitted by human activities. Without these land and ocean sinks, the atmospheric CO 2 increase would be about twice as fast; current atmospheric CO 2 concentration would be already above 500 ppm, inducing a warming of more than 2 K above pre-industrial level (assuming a median estimate of 3 K for the climate sensitivity).
A longer discussion can be found in ESI† to IPCC AR5 (ref. 22) (ref. 4 in Freidlingstein's report above). and in ref. 25–27 cited above. Tans discusses constraints on the CO2 budget, while two papers by Le Quéré discuss trends in CO2 sources & sinks and set out global carbon budget data.23–25
While the concept of carbon budgets does side step, to some extent, the issues around the uncertainty of carbon sinks, we have incorporated some simple calculations into this paper. Essentially our numbers show ‘this is the best case estimate for carbon dioxide concentrations assuming that carbon sinks continue to function as they do at present’.
Our major source for 2022 data is the BP Annual Statistical Bulletin of World Energy28 and other sources include the fossil fuel registry,29 the US National Oceanic and Atmospheric Administration,30 and the US Energy Information Administration data sets.31
Our data for 2005 represents the best retrospective sources we can find for that year, sometime published quite a lot afterwards. This can lead to discrepancies with the numbers of our original calculations done in 2006, where we were producing ‘best estimates’ with the data available at the time. In particular, our current calculation of 2005 gas reserves is notably lower than our original estimate, largely due to the distinction between proven and probable reserves. At the end of 2023, we checked whether new estimates of reserves had become available during the writing of this paper in 2022–23 and found no new numbers.
All of our data sources, and the relevant calculations, are available in a downloadable Excel Workbook. We have also written an appendix to this paper going through our step by step calculations in a series of tables (e.g. ‘Table x has the total carbon from quoted reserves, Table y deducts non fuel uses, Table z estimates the impact of 77 years fossil fuel use at current levels’ etc).
Not all hydrocarbon reserves end up as fuel. The US Energy Information Administration estimates that around 7% of oil is converted into non-fuel uses such as industrial solvents, lubricants and bitumen.32 For natural gas, the proportion looks to be slightly lower (3.75%) when comparing current total annual consumption of 4 trillion cubic metres with the 150 billion cubic metres used in non-fuel applications. The largest of these are fertilizer manufacture and methanol production.33
While coal has potential for being a rich source of valuable chemical intermediates, this is still a developing application, albeit with increasing development in China for olefin and glycol production. Figures from IEA resources on non-energy use of coal suggests these amount to around 4.5%.34,35
The relevant equations for converting commonly used fossil fuel quantities to tonnes of carbon dioxide can be summarised as below.
Coal: (reserves ‘short tons’ × 0.90718 = reserves in metric tonnes).
(reserves in metric tonnes × 0.9 = tonnes carbon in anthracite/hard coal).
(reserves in metric tonnes × 0.6 = tonnes in sub-bituminous ‘brown coal’)
(tonnes carbon in coal × 3.67 (44/12) = tonnes CO2 from coal reserves).
(tonnes CO2 from coal reserves × 0.955 = total CO2 allowing for 4.5% non-fuel use).
The calculations for coal are fairly straightforward, with the major uncertainty being the differing proportions of carbon in the various types of coal.
Oil: (reserves in barrels × 0.1364 = reserves in metric tonnes).
(reserves in tonnes × 0.85 = tonnes of carbon in reserves).
(tonnes carbon in reserves × 3.67 = tonnes CO2 from oil reserves).
(tonnes CO2 from oil reserves × 0.93 = total CO2 allowing for 7% non-fuel use).
Crude oil has a number of fairly standardised industry conversion factors owing to the wide variety of Imperial and metric units used by the industry for more than a century. A ‘barrel’ of oil is, these days, an abstract 42 US gallon quantity derived from historic wooden barrels, and the conversion factor of 0.1364 reflects the average density of the most common crude oils. For tar sands, which have a higher density than conventional crude, the conversion factor could be higher (around 0.15 tonnes per barrel).
Gas: (reserves in billion cubic metres × 0.76 = mass of natural gas in 106 tonnes).
(tonnes natural gas × 2.75 = tonnes CO2 from gas combustion).
(tonnes CO2 from gas combustion × 0.965 = total CO2 allowing for 3.7% non-fuel use
or
(reserves in billion cubic metres × 34.121 = energy content in trillion British Thermal Units – ‘btu’)
(Energy content in million btu × 14.43 = kg carbon per million btu).
(kg carbon × 1000 × 3.67 = tonnes carbon dioxide).
The conversion via energy content (btu) looks unusual but is actually widely used in the industry. It allows for variations of gas volume at various temperatures and pressures, and the fact that natural gas is not pure methane but has variable amounts of ethane and higher hydrocarbons. The two methods give results that can be reconciled.
Looking at Table 2 we can see that reserves of all types of fossil fuel, except for ‘Brown Coal’, have increased from 2005 to 2023. This has implications for the potential life of these assets:
• Global coal consumption reached 8 billion (8.0 × 109) tonnes in 2022.36 Proven coal stocks could supply over 130 years of coal consumption at 2022 levels.
• Global Oil consumption reached 36.37 billion barrels pa in 2022.37 Proven oil stocks could supply over 47 years of oil consumption at 2022 levels. Oil sands and shale fracking add an extra 23 years, summing to 70 years.
• Current natural gas consumption is around 4 trillion cubic metres pa,38 so the gas reserves above represent around 46 years supply for conventional reserves and an additional 54 years when shale gas is included.
These numbers suggest that there are already sufficient reserves of fossil fuels to take the world to, or very close to, the year 2100, the endpoint for the various IPCC RCP scenarios.
Of this 4777 Gt CO2 total, over half (3048 Gt CO2) comes from various forms of coal, 1029 Gt comes from traditional and novel oil reserves, and 700 Gt CO2 from gas. It is notable how the advent of new techniques for oil and gas exploitation (tar sands and fracking) have more than doubled the potential impact of these fuels since 2005.
Our estimates are slightly lower than some previous ones, partly due to the use of more recent data, and partly because of taking into account the lower carbon content of ‘Brown Coals’ (sub-bituminous and lignite).
Fig. 6 compares the potential CO2 emissions from Fig. 5 with the IPCC RCP scenarios. The background colours reflect the wide range of uncertainty in the IPCC scenarios mentioned in Fig. 1 above. RCP 2.6 (light green) was estimated as 510 to 1505 GtCO2, RCP 4.5 (mauve) was estimated as 2180 to 3690 Gt CO2, and RCP 6.0 (brown) was 3080 to 4585 Gt CO2.
Potential CO2 from oil reserves alone (1029 Gt) takes the world past the IPCC RCP 2.6 threshold, while adding gas (+700 Gt) takes it close to the start of RCP 4.5. Coal reserves alone (3048 Gt CO2) take the world close to the threshold of RCP 6.0.
Although our re-estimate of 2005 figures is lower than our original 2006 calculation, burning all the reserves present in 2005 would still result in atmospheric CO2 approaching 800 ppm (Fig. 7).
This figure represents the results of the (hopefully) hypothetical experiment – what happens to CO2 levels if all current reserves were combusted in an instantaneous process? While these numbers look alarming, there are a number of factors that could reduce them somewhat.
This analysis notably reduces the contribution of coal to the carbon budget, as a sizeable portion of coal reserves might only be consumed after 2100. Nonetheless, the potential emissions from 77 years consuming current reserves are still sufficient to exceed the IPCC RCP2.6 and RCP4.5 budgets, and to land within the range for RCP6.0 (Fig. 9).
Fig. 10 shows the possible changes to CO2 levels following fossil fuel consumption continuing at current rates until 2100 with carbon sinks maintaining their efficiency. Essentially these numbers show that ‘business as usual’ carbon dioxide emissions are likely to produce atmospheric CO2 levels in 2100 exceeding, at best, 620 ppm. If carbon sinks become less effective, we could see that rise to around 860 ppm.
The graphic below relates the numbers from Fig. 10 to the IPCC scenarios previously shown in Fig. 3 above (Fig. 11).
Fig. 11 How current reserves impact potential CO2 levels in IPCC scenarios RCP2.6 to RCP8.5 (IPCC AR5 data). |
Even if carbon sinks continue to function as at present, there is sufficient CO2 from proven reserves to maintain current rates of fossil fuel consumption, taking the world past RCP 4.5 and towards RCP 6.0.
Fig. 12 relates the potential CO2 emissions of current reserves to the IPCC RCP pathways mentioned in Fig. 4 above. The RCP pathways generate estimates of cumulative CO2 by adding the numbers quoted in the paragraph above to the historic emissions from 1870–2010 (for IPCC AR5) and/or 2019 (for IPCC AR6). The latter source quotes cumulative historic emissions as 2400 Gt CO2. Adding the 3400+ Gt from exploiting current reserves at their current rate, takes the total to 5800 Gt CO2 by 2100. As Fig. 12 shows, this is very much in the range of RCP 6.0 – a temperature anomaly over 2.5 C. These numbers provide support for the position of the International Energy Agency that ‘no new long-lead time oil and gas projects are needed. Neither are new coal mines, mine extensions, or unabated coal plants’.
Fig. 12 Cumulative CO2 emissions: scenarios RCP2.6 to RCP8.5 (IPCC AR5 SPM Fig. 10). |
Firstly, oil companies tend to have conservative accounting policies – usually recording only those reserves which have a very high certainty (>90%) of being successfully exploited.42
Secondly, advances in technology will often extend the actual amount extracted from Oil and Gas fields. This is particularly notable for large fields producing over an extended period of time, where companies are incentivised to exploit infrastructure in place.
A well-documented example is BP's large Forties field in the North Sea. In 1970 when the discovery of the field was announced, it was estimated that the recoverable amount of oil was 1.8 billion barrels,43 and production was expected to stop in the 1990's. In 2003 the field was sold to a smaller company, (Apache Oil) after BP estimated just 144 million barrels remained. By 2010, the field had produced 2.64 billion barrels.44 The field is still producing oil in 2023,45 with estimated production by Apache approaching 300 million barrels.46
We also note the number of new discoveries made during the preparation of this paper – not yet classified as proven reserves, but significant in size. The ‘Carbon Bomb’ report mentioned above (from the Guardian Newspaper) identified 195 oil and gas projects that could each add a billion tonnes of CO2 to the atmosphere.42 These projects are not included in our calculations, as they do not yet appear in authoritative sources like the BP Guide to World Energy (who have not included reserves estimates in the last two surveys).
We might expect that proven coal reserves will continue to rise, given the wide distribution of coal around the world (with notable untapped reserves in Russian Siberia). However it is also likely that some coal will be ‘left in the ground’ as the operating cost advantages of other technologies increase. Some thermal power stations designed for coal can be converted to run on natural gas or biomass,51 increasing their efficiency as well as reducing their carbon footprint. It is also feasible, though economically challenging, to mothball coal mines.52 Individual countries can make telling contributions, such as the recent decision of India to pause coal generating capacity for five years53
However, the widespread availability of coal deposits around the world means there is always a temptation for users to switch back to coal when expedient or cheap. The international community could discourage that ‘switching-back’ by measures such as:
• Technology sharing of renewable electrical generation methods such as solar PV.
• Converting coal to gas powered generation. In the appendix (Table 16) we show calculations for a 1000 MW coal powered plant that would emit 5 million tonnes CO2 at 60% capacity factor. A gas powered 1000 MW plant would emit less than 2 million tonnes CO2 at the same capacity factor (not only does gas have a lower carbon content, combined cycle gas generators have a better energy efficiency than coal plants54). Additionally, gas generation can be used with non-fossil fuel biogas sources.
• Subsidizing the decommissioning and removal of coal power stations to remove the temptation to switch back.55
• Building renewable electrical generating plants on old coal plant premises, to take advantage of existing electrical infrastructure, generators, and grid connections. A long-term example of this is the UK's new STEP fusion reactor at an old coal power site in Nottinghamshire – ultimately using traditional steam turbines and generators to connect to the grid. See the graphic on Bay-Fusion's blog page.56
Encouragingly, a 2022 analysis of Chinese coal station construction posits that much of the new building is carried out by provincial authorities seeking short term economic stimulus. The capacity factor of these power plants is below 50%, many are loss-making and liable to closure in the not too distant future.57 Coal may indeed be the big problem, but hopefully not an insuperable one.
It is perhaps not surprising that much current thinking calls for governments to impose/enforce rather higher prices. The IMF report mentioned above calls for carbon to be priced around $75 per tonne CO2. A Reuters poll of climate economists in 2012 suggested that prices of $100 per tonne CO2 were required for countries to meet their ‘net-zero by 2050’ targets.59
Our analysis prompts the obvious question – if coal is potentially responsible for over half of carbon emitted up to 2100, how can we get rid of it? The 1000 MW power station mentioned in the discussion on coal above might more than halve its emissions by converting to gas. Table 16 in the appendix shows how the economics of the plant might change. Using medium term prices (pre Covid and Ukraine) such a plant would save 16% on its annual fuel cost without a carbon tax – perhaps not enough to defray the costs of switching? At $20 per tonne this rises to 35% saving, at $50 per tonne 44%, and at $75 per tonne 49%.
This simple analysis shows that prices in the $50–$75 per tonne band might induce switching out of coal and consequent reduction in emissions. A more sophisticated analysis has been carried out by Stanford University in a 2022 report on ‘carbon arbitrage’, the process of investing to ‘go short on coal’.60 They put the ‘social cost of carbon’ and the corresponding benefit of removing it at $75 per tonne.
(a) Demand for energy: Fossil fuels, such as oil, gas, and coal, still account for a significant portion of global energy consumption. As the global population continues to grow, and developing economies increase their energy needs, there remains a demand for these energy sources. Even with the emergence of renewable energy sources, fossil fuels still offer a lucrative market with established infrastructure and customer base.
(b) Scepticism around ‘stranded assets’ and Environmental, Social, & Governance (ESG) factors in general. While some investors have accepted the concept of ‘stranded assets and ‘unburnable carbon’, this has not translated into significant share price impact according to a 2015 analysis by Griffin et al..61 Bebbington62 (2020) amplified this in a series of interviews with Oil industry stakeholders and observers – stranded assets were ‘accepted as a concept’ but one that had little importance until ‘information was demanded by stock markets and/or governments’.
(c) Profitability and price spikes: Fossil fuel exploration and production can be highly profitable for companies. Companies aim to maximize their profits by meeting the existing demand. In particular, companies with lower cost crude can take profitable advantage of price volatility even in an overall declining market. In contrast to coal, reducing oil production capacity by shutting down entire fields is ‘lumpy’ and there is little option to mothball oil wells.63 This is likely to lead to price spikes in an otherwise declining oil price64,65 A recent article in Reuters noted that companies claim to ‘invest to maintain price stability’66 – although one might be sceptical of this claim given the large profits made by companies during the recent oil and gas price rises caused by the war in Ukraine. An economically rational approach might be to develop a new field and run it at less than full production capacity until prices spike. This approach effectively generates a ‘real option’ value to expanding or maintaining oil reserves, which has been recognised by industry economists since the 1990's.67 One could imagine these valuations being used in compensation cases if/when legislation forces oil fields to be abandoned (see next point).
(d) The value of ‘booked reserves’: Companies instinctively explore and develop new reserves as their current reserves deplete. Given that firms are often valued by reserve capacity, adding to it cheaply (by acquiring rights to undeveloped reserves) looks good for shareholders. One might summarise the attitude as ‘if it can be brought onstream quickly and cheaply, then explore and drill’. Furthermore, once a reserve has been quantified and valued, it can be the subject of compensation claims in Investor-State Dispute Settlement (ISDS) cases.68 ISDS clauses embedded in the international energy charter treaty have allowed Oil and Gas companies to pursue over 200 court cases against countries seeking to implement fossil fuel reduction policies.
(e) Uncertain transition and new technologies: While there is a global push towards transitioning to renewable energy, the transition process takes time. Fossil fuel companies may continue exploration as they anticipate a gradual shift in energy sources, allowing them to adapt their business models and investments accordingly.69 Many Oil and Gas companies are pursuing technological solutions that promise the exploitation of fossil fuels in a low/lower carbon way. Examples include the production of blue/green hydrogen,70 carbon capture,71 and ammonia production.72
Left alone, fossil fuel industries are likely to carry on a large amount of business as usual, interspersed with a rather gradual response to changes they see taking decades to come to fruition. Investment in oil and gas fields continues unabated.73 While they recognise that governments may reach international agreements and enact policies to restrict fossil fuel use, their actions show that ‘they don't believe it will happen any time soon’. The long-standing industry discount rate (10%)74 to assess new energy investments is still being applied in 202375 despite the substantial rise in interest rates over the last few years. Given that discount rates are a combination of ‘risk-free’ bond interest rates and an industry-related ‘risk premium’, this can only imply that they do not see increased risks from climate change policies. What are the policy options for governments in this situation?
(1) Clear and ambitious regulations: Governments should establish clear and ambitious regulations that set out the requirements and timeline for achieving net zero carbon targets. These regulations should provide a stable and predictable policy framework, giving investors and companies the confidence to transition towards cleaner energy sources. The breadth of the challenge is illustrated by the lengthy, comprehensive, report76 by the Energy Transitions Commission (ETC), which discusses the need to cut coal use by >80%, gas use by >55% and oil use by >75% by 2050. ETC does not expect carbon capture or removal to play a significant role in reduction, and instead call for reduced investment in fossil fuel supply. In some notable cases,77 countries have actually announced that no further fossil fuel exploration will be allowed in their territories.
(2) Carbon pricing: More governments could introduce or strengthen carbon pricing mechanisms like the EU ETS. These mechanisms put a price on carbon emissions, incentivizing companies to reduce their carbon footprint and invest in cleaner technologies. Carbon pricing provides economic signals that align with net zero goals and encourages investors to support low-carbon projects.
(3) Renewable energy incentives: Governments can provide incentives and subsidies for renewable energy projects, making them more attractive for investors. These incentives can include tax credits, feed-in tariffs, grants, or low-interest loans. By promoting the growth of renewable energy, governments signal their commitment to decarbonization and create investment opportunities in the sector.
(4) Support for research and development: Governments can invest in research and development (R&D) initiatives focused on clean energy technologies. By funding R&D projects and offering grants or tax credits for innovation, governments can stimulate the development of new solutions and attract private investment in clean technologies.
(5) Green infrastructure investment: Governments can prioritize investments in green infrastructure projects, such as renewable energy installations, public transportation systems, and energy-efficient buildings. By demonstrating a commitment to sustainable infrastructure, governments can create a conducive environment for investors and companies to align their activities with net zero goals.
(6) ESG disclosure and reporting: Governments can mandate enhanced environmental, social, and governance (ESG) disclosure and reporting requirements for companies. This ensures that investors have access to transparent and standardized information about companies' climate-related risks and opportunities. Improved disclosure can facilitate informed decision-making and encourage investors to support companies that align with net zero targets. Investors generally welcome clear rules around ESG and become frustrated by regulatory instability78,79
(7) Collaborative engagement: Governments can engage in collaborative dialogues with investors, financial institutions, and fossil fuel companies to discuss the transition to a low-carbon economy. By involving stakeholders in the policy-making process, governments can address concerns, build consensus, and foster cooperation towards achieving net zero goals.
(8) ‘Just transition’ plans: A just transition approach helps mitigate resistance to the transition and ensures a fair and equitable distribution of the benefits of a low-carbon economy. A focused example of such a plan has been promulgated in the North Sea by employee groups working alongside environmentalists.80 A much broader outline from McKinsey81 looks at aggregate shifts needed for net-zero economies, on energy, land-use, industries and individuals. It comments that cumulative spending of about $275 trillion on physical assets would be required to achieve this. Governments should develop comprehensive just transition plans that support workers and communities affected by the shift away from fossil fuels. These plans should include measures for retraining and reskilling workers, creating alternative job opportunities, and providing support to impacted regions.
(9) Government ‘skin in the game’: Governments ask investors to make twenty-five year investment decisions on renewable energy or hydrogen transportation, yet reserve the right to change their minds on policies overnight. A 2011 article about the UK Governments then new ‘Green Investment Bank’ made the following comment ‘Before it risks significant investment in policy-supported new markets, the private sector understandably wants the government to underwrite policy and regulatory risk which is within its control (but not within the private sectors control). In other words, it wants the government to commit to its own policy by putting some skin in the game’.82
One proposal that has recently borne fruit is ‘policy performance bonds’ or ‘sovereign sustainability-linked bonds’ (SSLB's). These are financial instruments that hold governments accountable for achieving their policy objectives. The idea is that governments would issue bonds tied to specific ‘Sustainability Performance Targets’ (SPTs), such as carbon emissions reduction targets, or renewable energy deployment. These bonds would pay a higher yield if the government successfully meets its targets, but would incur a penalty if it fails to do so. The bonds would be tradable in financial markets, allowing investors to speculate on the government's performance. The purpose of policy performance bonds is to create financial incentives for governments to prioritize and actively pursue their policy objectives.83 SPTs could be set to address some of the factors encouraging continued invest in fossil fuel extraction mentioned in section 5.3 above. First promulgated (by the authors) in 2009, it took until 2022 for the first SSLB's to be issued by the Governments of Chile and Uruguay84 leading to interest from a wider range of countries and institutions. Chile's Government has committed to targets that include decarbonising 60% of its electricity production by 2032, while Uruguay has set a challenging gross greenhouse gas reduction target for 2025. The issues combined longer maturity dates and lower interest rates than traditional bonds, yet were immediately oversubscribed by a wide range of investors, and achieved high credit ratings.85
• Q. Is there any evidence that the concept of ‘unburnable carbon’ has in fact lead to reduction in fossil fuel reserves?
• A. Apart from ‘Brown Coal’, there is no evidence from our results of any reserve reduction. Potential carbon emissions from fossil fuels show a 46% rise in since 2005.
• Q. What might happen to IPCC carbon budgets, and atmospheric CO2 levels, if these reserves are in fact very ‘burnable’?
• A. Current reserves consumed at current rates are sufficient to add 3400 Gt CO2by 2100, taking the world a long way towards IPCC path RCP 6.0. Atmospheric CO2 would exceed 620 ppm v/v by 2100.
• Q. Do the comments around carbon pricing at $30–60 per tonne still apply?
• A. Current expert opinion is that carbon prices around and over $75 per tonne are required to achieve net zero by 2050. We estimate coal to gas switching in thermal power stations would be prompted as carbon prices approach $50 per tonne.
• Q. What are the implications of the current situation for both financial organisations and policy makers?
• A. Policy makers, helped by private and public sector investors, need to develop laws, regulations, and incentives, that steer fossil fuel companies away from ‘business as usual’ approaches to exploration and development. It seems unlikely that fossil fuel companies will change direction without these incentives.
We hope this paper will promote further thoughts, both around the reserves data discussed above, and the policy options and constraints for governments addressing these issues.
Parameter | Values | Data sources |
---|---|---|
a Trenbeth K, and Smith L, 2005 The Mass of the Atmosphere: A Constraint on Global Analyses J.Climate, 6, 864-875. b PPM converter for Gases https://www.lenntech.com/calculators/ppm/converter-parts-per-million.htm (accessed May 2023). | ||
Global proved coal reserves/tonnes | 1.07411 × 1012 | BP |
Global proved oil reserves/barrels | 1.7348 × 1012 | BP |
Oil sands reserves/barrels | 4.232 × 1011 | BP |
Proven natural gas reserves/109 m3 | 1.881 × 105 | ΕΙΑ |
Shale tight oil reserves/109 barrels | 418.9 | ΕΙΑ |
Shale gas reserves/ft3 | 7576.6 | ΕΙΑ |
CO2/C ratio | 3.67 | |
Proportion of carbon in coal | 0.9 | ΕΙΑ |
Proportion of carbon in sub-bituminous coal & lignite | 0.6 | ΕΙΑ |
Mass of Earth's atmosphere/metric tonnes | 5.148 × 1015 | Trenbeth & Smitha |
Density of crude oil (tonne/barrel) | 0.1364 | BP |
Proportion of carbon in oil (w/w) | 0.85 | Britannica |
Proportion of carbon in natural gas (kg/106 btu) | 14.43 | US EPA |
2022 proportion of CO2 in atmosphere/ppm | 417 | NOAA |
Oil reserves used non-combustion purposes (USA) | 7% | EIA |
Conversion of wt/wt to vol/vol | 0.667 | Lenntechb |
Parameter | 2022 | 2005 |
---|---|---|
‘Hard coal’ anthracite and bituminous (billion metric tonnes) | 754 | 478 |
‘Brown coal’ sub-bituminous and lignite (billion metric tonnes) | 321 | 430 |
Oil (billion barrels) | 1735 | 1201 |
Natural gas (trillion cubic metres) | 188 | 156 |
Oil sands (billion barrels) | 423 | 84.7 |
Shale oil (billion barrels) | 419 | 0 |
Shale gas (trillion cubic metres) | 214 | 15 |
Parameter | 2022 values | 2005 values |
---|---|---|
Mass (coal)/metric tonnes | 7.536 × 1011 (ref. 18) | 4.788 × 1011 |
Mass (carbon)/metric tonnes | 6.782 × 1011 | 4.070 × 1011 |
Mass (CO2)/metric tonnes | 2.487 × 1012 | 1.492 × 1012 |
Fraction of CO2/total atmosphere | 4.831 × 10−4 | 3.069 × 10−4 |
Increase of CO2/ppm by wt | 483 | 307 |
Increase of CO2/ppm v/v | 322 | 205 |
Parameter | 2022 values | 2005 values |
---|---|---|
Mass (coal)/metric tonnes | 3.205 × 1011 | 4.303 × 1011 |
Mass (carbon)/metric tonnes | 1.923 × 1011 | 2.582 × 1011 |
Mass (CO2)/metric tonnes | 7.050 × 1011 | 9.467 × 1011 |
Fraction of CO2/total atmosphere | 1.370 × 10−4 | 1.84 × 10−4 |
Increase of CO2/ppm by wt | 137 | 184 |
Increase of CO2/ppm v/v | 91 | 123 |
The ‘Proven Global Reserves’ of coal have increased by nearly 40% since 2005, despite coal being consumed at increasing amounts throughout the period. Global coal consumption reached 8 billion (8.0 × 109) tonnes in 2022.36 Proven coal stocks of ‘hard’ and ‘brown’ together coal could supply over 130 years of coal consumption at 2022 levels.
The CO2 produced in burning all current coal reserves, is estimated as 3192 Gigatonnes, enough to take the world past the RCP2.6 target towards RCP4.5 (Table 5).
Parameter | 2022 values | 2005 values |
---|---|---|
Volume (oil)/barrels | 1.735 × 1012 | 1.201 × 1012 |
Mass (oil)/metric tonnes | 2.366 × 1011 | 1.638 × 1011 |
Mass (carbon)/metric tonnes | 2.011 × 1011 | 1.392 × 1011 |
Mass (CO2)/metric tonnes | 7.375 × 1011 | 5.104 × 1011 |
Fraction of CO2/total atmosphere | 1.433 × 10−4 | 9.915 × 10−5 |
Increase of CO2/ppm by wt | 143 | 99 |
Increase of CO2/ppm v/v | 96 | 66 |
There is no sign of ‘peak oil’ in this data, as reserves have increased by over 40%, despite the increasing consumption of oil. These reserves, representing about 47 years supply at current rates of consumption, are again sufficient to take the world past the RCP2.6 budget (Table 6).
Parameter | 2022 values | 2005 values |
---|---|---|
Volume (gas/109 m3) | 1.881 × 105 | 1.557 × 105 |
Energy (gas)/1012 btu | 6.418 × 106 | 5.313 × 106 |
Mass (carbon)/kg | 9.261 × 1013 | 7.666 × 1013 |
Mass (CO2)/metric tonnes | 3.396 × 1010 | 2.811 × 1010 |
Fraction of CO2/total atmosphere | 6.596 × 10−5 | 5.460 × 10−5 |
Increase of CO2/ppm by wt | 66 | 55 |
Increase of CO2/ppm v/v | 44 | 36 |
Proven natural gas from ‘conventional’ (e.g. non fracking) reserves have increased by around 21%, again despite the switch to gas fired electricity generation in multiple countries. Current natural gas consumption is around 4 trillion cubic metres pa,1 so the gas reserves above represent around 46 years supply (Table 7).
Parameter | 2022 values | 2005 values |
---|---|---|
Barrels | 4.232 × 1011 | 1.80 × 1011 |
Mass (oil)/metric tonnes | 5.77 × 1010 | 2.57 × 1010 |
Proportion carbon | 0.9 | 0.9 |
Mass (carbon)/metric tonnes | 5.20 × 1010 | 2.31 × 1010 |
Mass (CO2)/metric tonnes | 1.905 × 1011 | 8.486 × 1010 |
Fraction of CO2/total atmosphere | 3.49472 × 10−5 | 1.56 × 10−5 |
Increase of CO2/ppm by wt | 35 | 15.57 |
Increase of CO2/ppm v/v | 23 | 11 |
The increasing demand for hydrocarbon fuels has led to new resources being discovered and exploited. Proven reserves from oil sands (sometimes termed tar sands) such as those found in Athabasca, have increased by over 220% since 2005. This represents an incremental 11.6 years supply of oil at current rates of demand (Table 8).
Parameter | 2022 values | 2005 values |
---|---|---|
V (oil)/barrels | 4.189 × 1011 | 0 |
m (oil)/metric tonnes | 5.7 × 1010 | 0 |
Mass (carbon)/metric tonnes | 4.86 × 1010 | 0 |
Mass (CO2)/metric tonnes | 1.781 × 1011 | 0 |
Fraction of CO2/total atmosphere | 3.4592 × 10−5 | 0 |
Increase of CO2/ppm by wt | 35 | 0 |
Increase of CO2/ppm v/v | 23 | 0 |
As well as oil sands, significant amounts of oil are now being extracted from shale via fracking. In 2005 this was still a new technology and no proven reserves were recorded in our analysis. 2022 reserves are similar to those for oil sands, providing an incremental 11.5 years at current demands (Table 9).
Parameter | 2022 values | 2005 values |
---|---|---|
Volume (gas/m3) | 2.14543 × 1014 | 1.46114 × 1013 |
Energy (gas)/1012 btu | 7320435.186 | 498554.0422 |
Mass (carbon)/metric tonnes | 1.05634 × 1011 | 7.19 × 109 |
Mass (CO2)/metric tonnes | 3.873 × 1011 | 2.638 × 1010 |
Fraction of CO2/total atmosphere | 7.52378 × 10−5 | 5.124 × 10−6 |
Increase of CO2/ppm by wt | 75 | 5 |
Increase of CO2/ppm v/v | 50 | 3 |
‘Unconventional’ (e.g. from fracking) shale gas reserves have increased by nearly 15 times since 2005. Proven reserves are now higher than those in conventional natural gas fields, and provide an estimated 53 years at current consumption levels.
Parameter | 2022 values | 2005 values |
---|---|---|
‘Hard’ coal | 2487 | 1580 |
‘Brown’ coal | 705 | 947 |
Oil | 737 | 510 |
Natural gas | 340 | 281 |
Oil sands | 190 | 85 |
Shale oil | 178 | 0 |
Shale gas | 387 | 26 |
Total | 5025 | 3429 |
Except for ‘brown coal, fossil fuel reserves have increased since 2005, with coal overall still much the largest contributor to potential carbon dioxide increase. Coal has decreased as a portion of the total emissions potential, mainly due to the notable increase from shale gas (Table 11).
Parameter | 2022 values | 2005 values |
---|---|---|
‘Hard’ coal | 322 | 205 |
‘Brown’ coal | 91 | 123 |
Oil | 96 | 66 |
Natural gas | 44 | 36 |
Oil sands | 23 | 11 |
Shale oil | 23 | 0 |
Shale gas | 50 | 3 |
Sub-totals | 650 | 444 |
Starting concentration in each year | 417 | 370 |
‘Burn it all’ total | 1067 | 814 |
Parameter | Non fuel usage | 2022 values | 2005 values |
---|---|---|---|
‘Hard’ coal | 4.5% | 2375 | 1509 |
‘Brown’ coal | 4.5% | 673 | 904 |
Oil | 7% | 686 | 475 |
Natural gas | 3.75% | 327 | 271 |
Oil sands | 7% | 177 | 79 |
Shale oil | 7% | 166 | 0 |
Shale gas | 3.75% | 373 | 25 |
Total | 4777 | 3263 |
Parameter | Non fuel usage | 2022 values | 2005 values |
---|---|---|---|
‘Hard’ coal | 4.5% | 308 | 196 |
‘Brown’ coal | 4.5% | 87 | 117 |
Oil | 7% | 89 | 62 |
Natural gas | 3.75% | 42 | 35 |
Oil sands | 7% | 22 | 10 |
Shale oil | 7% | 21 | 0 |
Shale gas | 3.75% | 48 | 3 |
Sub-totals | 618 | 422 | |
Total adding starting concentrations from Table 11 | 1035 | 792 |
Parameter | ‘Burn it all’ emissions | Years of reserves at current consumption | ‘Burn 77 years’ emissions at current consumption) |
---|---|---|---|
Coal | 3048 | 130 | 1805 |
Oil | |||
Conventional oil | 686 | 47 | |
Oil sands | 177 | 12 | |
Oil shale | 166 | 11 | |
All oil sub-total | 1029 | 70 | 1132 |
Gas | |||
Conventional gas | 327 | 47 | |
Shale gas | 373 | 54 | |
All gas sub-total | 700 | 101 | 504 |
Total | 4777 | 3441 | |
Potential CO2 increase in atmosphere ppm (v/v) | 619 | 446 | |
Starting concentration 2023 | 417 | 417 | |
Total atmospheric concentration | 1036 ppm | 863 ppm |
This analysis notably reduces the contribution of coal to the carbon budget, as a sizeable portion of coal reserves might only be consumed after 2100. Nonetheless, the potential emissions from current reserves are still sufficient to exceed the IPCC RCP 2.6 and RCP4.5 budgets, and to land within the range for RCP6.0. The last three lines of the table calculate the potential rise in CO2 concentration if 77 years of fossil fuel reserves were consumed in one fast ‘burn it all’ process. They are not a prediction of what concentrations might be like in 2100 if the reserves were consumed steadily over the period to 2100. What, if anything, can we say about that scenario? Table 15 addresses this question, using the assumption that carbon sinks continue to function as they do now.
Year | CO2 emissions (tonnes) |
---|---|
2005 | 2.96 × 1010 |
2006 | 3.06 × 1010 |
2007 | 3.15 × 1010 |
2008 | 3.21 × 1010 |
2009 | 3.16 × 1010 |
2010 | 3.34 × 1010 |
2011 | 3.45 × 1010 |
2012 | 3.50 × 1010 |
2013 | 3.53 × 1010 |
2014 | 3.56 × 1010 |
2015 | 3.56 × 1010 |
2016 | 3.55 × 1010 |
2017 | 3.61 × 1010 |
2018 | 3.68 × 1010 |
2019 | 3.71 × 1010 |
2020 | 3.53 × 1010 |
2021 | 3.71 × 1010 |
Total | 5.83 × 10 11 |
Potential CO2 increase in atmosphere 2005–21 (ppm v/v) | 72.5 |
Actual increase in atmosphere (ppm v/v) | 38 |
Proportion CO2 absorbed by carbon sinks | 47.3% |
Potential increase in CO2 2023–2100 (from Table 14, ppm v/v) | 456 |
Potential increase if sinks function as now (ppm v/v) | 240 |
Current CO2 concentration in atmosphere (ppm v/v) | 417 |
Potential total concentration 2100 (ppm v/v) | 657 |
Finally, Table 16 summarises some calculations illustrating the reduction in carbon dioxide emission and fuel costs by switching from a hypothetical coal power station to a natural gas plant producing the same energy.
Parameter | Values | Coal | Gas |
---|---|---|---|
Power station | 1000 MW | ||
Annual energy output at 60% capacity factor | 5256 GWh | ||
Energy from 1 tonne (MWh) | 8.33 | 13.63 | |
Mass of fuel needed at 100% efficiency (tonnes) | 630,720 | 385,534 | |
Typical efficiency | 37% | 55% | |
Realistic mass fuel required (tonnes) | 1704649 | 700971 | |
Proportion of carbon | 80% | 75% | |
Carbon emitted (tonnes) | 1363719 | 525728 | |
CO2 emitted (tonnes) | 5000303 | 1927670 | |
Average fuel cost 2017–2020 | $84 per tonne | $170 per tonne | |
Realistic fuel cost | $143 m | $121 m | |
Carbon price additional cost | $20 per tonne CO2 | $100 m | $38 m |
$50 per tonne CO2 | $250 m | $96 m | |
$75 per tonne CO2 | $375 m | $144 m | |
Total fuel cost at $75 per tonne | $518 m | $265 m |
Footnote |
† Electronic supplementary information (ESI) available. See DOI: https://doi.org/10.1039/d3ea00107e |
This journal is © The Royal Society of Chemistry 2024 |