Key takeaways from Stockholm Institute paper on “lock-in” from fossil fuel infrastructure

Sunk costs for coal are less:  Not surprisingly, of the three fossil fuels, it is coal for which production would need to be scaled back the most in a low-carbon scenario, both as a share of production (34%) and in absolute carbon terms (about 5 Gt CO2). At the same time, the analysis indicates that investments in coal produc­tion may also be the easiest to “unlock”. As indicated by the areas of the coloured (red, green and blue) bars and light grey bars, coal resources are far less capital-intensive (less than 5 USD/t CO2) than oil or gas, for which new fields require investments of 30 USD/t CO2 or more. This indicates that sunk costs for infrastructure, and creditor concerns, may contribute less to lock-in for coal.

More lock-in risk in oil and gas investments: Coal is also far less rent-intensive on average, with most deposits yielding rents of less than 10 USD/t CO2, while rents to oil and gas production average 50 USD/t CO2 or more. This suggests that carbon pricing – or normal fluctuations in resource prices – could have a greater effect on coal than on oil or gas production. With combined capital and produc­tion costs that are far closer to expected prices than those for oil and gas, coal mines are at far greater risk of being rendered uneconomic by carbon pricing. Thus, at least based on economic considerations, investments in coal production may create less “lock-in” risk than investments in oil or gas production.

Lock-in risks for coal are still high but planners may want to look at oil and gas before coal. Of course, social and political considerations – as well as local differences in project economics, including rents – might change the outlook. For example, other research has suggested that coal production is more labour-intensive than oil or gas, and that coal production interests have already been among the most powerful opponents of climate policies in both the U.S. and the EU. Furthermore, coal-fired power production (the demand side of coal markets) still presents a significant lock-in risk.22 That said, planners concerned about carbon lock-in risks from fossil fuel supply investments may want to look at oil and gas before coal.

Oil is both the most capital-intensive and most rent-intensive fossil fuel, with average capital intensity of 44 USD/t CO2 (16 USD/bbl) and rent intensity of 200 USD/t CO2 (74 USD/bbl), when assessed across all barrels produced. This high rent intensity suggests that, for many oil deposits, carbon pricing would be less likely to affect production as substantially. The capital intensity of oil production ranges from 4 to 41 USD per barrel (11 to 112 USD/t CO2), with significant investment required especially for higher-cost (currently not producing) offshore resources unlikely to be developed in a cost-efficient low-carbon scenario. Near-term investment in these resources could be substantial, creating momentum for future over-production.

Investment in future oil production most likely to increase carbon lock-in. For the barrels over-produced in 2030, our analysis shows a capital intensity of 97 USD/t CO2 and rent intensity of 55 USD/t CO2 – or 153 USD/t CO2 combined (see Table 1), sug­gesting that those resources would be well-insulated from future price fluctuations or carbon pricing, once capital is invested. Overall, our analysis suggests that among investments in fossil fuels, those in oil production, especially in higher-cost, yet-to-produce resources, are most likely to increase carbon lock-in.

Such investment should be cut sharply. Looking deeper at offshore oil, our analysis indicates that production from yet-to-be made investments in this infra­structure would need be cut by half in 2030 in the IEA’s 450 Scenario, relative to BAU. The Americas (North and South) represent the greatest source of over-production (nearly half). Capital investments such as these may deserve special scrutiny because, once oil platforms and other major fuel extraction infrastructure are in place, the marginal cost of producing each unit of resource drops to 50 USD/barrel or less (the operating cost – i.e. the black or grey portion of each bar in Figures 2–4). This insulates the resource from likely expected variations in fuel price, whether due to climate policy or normal market fluctuations.

Rent intensities for natural gas production can also be substantial (averaging 93 USD/t CO2 or 5 USD/MBtu), though large variations in regional gas prices complicate the assessment of rents for gas, and the average values indicated in Figure 4 may not apply for regions where natural gas prices are very low (e.g. North America, with prices as low as 3 USD/MBtu) or very high (e.g. liquefied natural gas in East Asia, with prices as high as 15 USD/MBtu, as indicated in Figure 4). As with oil, yet-to-produce offshore gas resources averaging 144 USD/t CO2 (7.4 USD/MBtu). They also are set to over-produce by the greatest quantity: 0.6 Gt CO2 in 2030.

Cuts more than stated here are likely to be necessary for deep decarbonization. These findings depend on the year chosen (here, 2030) and the scenarios used (here, the IEA’s New Policies and 450 scenarios). Thus, our analysis is just one possible outcome of such an exercise. Low-carbon scenarios that foresee greater reduction in oil consumption, for example, might suggest the need to further scale back capital-intensive oil investments. Indeed, lower oil price scenarios in Rystad Energy’s assess­ment may foreshadow what might occur under even deeper low-carbon scenarios, as they also lead to a substantial scale-back of capital investment in onshore tight oil production, especially from not yet producing assets.23

Similarly, if an analysis year well beyond 2030 were chosen, some low-carbon scenarios might foresee significant avail­ability of carbon capture and storage (CCS) facilities, thereby enabling higher levels of coal production, and thus less over-production relative to BAU. That said, our assessment of over-production of coal, oil, and gas in a 2030 time frame is broadly consistent with a recent meta-analysis of fossil fuel production in a low-carbon economy.24

Policy implications and conclusions

This paper presents a generalized approach for assessing carbon lock-in risk from investments in fossil fuel extraction, building on common approaches to energy scenario analysis and fossil fuel resource analysis. Using this approach and its three metrics – over-production, capital intensity and rent intensity – policy-makers can assess the consistency of plans for developing new fossil fuel resources, or infrastructure to support them, with climate protection objectives.

Our application of this approach at the global scale suggests that rents for coal extraction are low enough that, in principle, scaling down coal extraction may be within reach of climate policy – e.g. through carbon pricing at the point of extraction or through financial incentives.25 Indeed, others have pro­posed policy mechanisms, such as supply-side cap-and-trade, designed to transition away from coal.26

In contrast, oil extraction is relatively profitable and, in many cases, capital-intensive. This suggests that strong financial interests may pose substantial barriers and tend to keep capital-intensive oil resources in production, even if later policy efforts (including carbon pricing) were to call for a transition away from oil.

The capital-intensive nature of new, unconventional and offshore oil developments, as identified here, suggests that near-term investments may bring resources online that will be especially difficult to unlock. Furthermore, some research­ers have suggested that resource owners may deliberately speed up investment and production in the near term, while carbon prices are low or non-existent, so they can lock in and insulate resources against the loss of rents due to the even­tuality of steeply increasing carbon prices.27 Policy-makers concerned about carbon lock-in risks, but also eager to ensure that near-term energy needs are met, may want to try to steer investment towards less capital-intensive oil reserves.

More broadly, our analysis highlights the importance of identifying the potential for fossil fuel “over-production” and the capital and rent intensities associated with those resources. Policy-makers could then tailor policy measures to fit the capital and rent intensity of each type of resource. Where rent intensity is low, financial measures (such as carbon pricing and subsidy reform) may be particularly effec­tive. For resources that are both rent- and capital-intensive, non-financial measures, such as quotas or limits on extraction (implemented through permitting decisions, for example), might be more effective. Further research is needed to better understand which approaches are most effective, and how they might be combined.

Of course, carbon lock-in risk is just one of many factors that policy-makers may consider in regulating the development of fossil fuel resources. Countries with substantial fossil fuel resources may have only a small subset of the high-risk resources assessed here. Or they may already be deeply “entangled”, heavily dependent on fossil fuel extraction – or be counting on it for their future energy supply and economic development.

Applying this approach at the regional or national level is likely to raise questions about accounting and equity that policy-makers have yet to resolve. For example, some countries extract (and generate rents from) fossil fuels that are exported to other jurisdictions, where they release CO2 emissions that are not generally attributed to the countries of origin.28 By limiting extraction, such countries would forgo economic rents without getting “credit” for any emis­sions avoided. (Global CO2 emissions would be avoided to the extent that the forgone production was not matched by production increases in other countries.)

The importance of fossil fuel extraction to some lower-income countries’ development should also be carefully con­sidered. Their policy-makers may rightfully note that many other countries have based economic development on fossil fuel energy. Thus, the application of this analytical approach at the regional and national scales would need to consider this concern, as well as possible relationships between the location (and forgone rents) of fossil fuels left in the ground and the financial responsibility for climate change mitigation.

Other researchers have suggested that policies to limit fossil fuel supply, such as supply-side caps, can increase the effi­ciency and effectiveness of demand-side measures to reduce CO2 emissions as well.29 Additional research is needed to clarify how supply-side policies can complement demand-side policies. This framework can contribute to that research by helping to shed light on the types of fossil fuel resource investments likeliest to create carbon “lock-in”, and thus help policy-makers to develop well-targeted and effective supply-side climate strategies.


1 Global Commission on the Economy and Climate (2014). Better Growth, Better Climate: The New Climate Economy Report. The Global Report. Washington, DC.

2 IEA (2015). Energy Technology Perspectives 2015: Mobilising Innovation to Accelerate Climate Action. International Energy Agency, Paris.

3 For an in-depth discussion, see: Unruh, G.C. (2000). Understanding carbon lock-in. Energy Policy, 28(12). 817–30. DOI:10.1016/S0301-4215(00)00070-7.

4 Davis, S.J., and Socolow, R.H. (2014). Commitment accounting of CO2 emissions. Environmental Research Letters, 9(8). 084018. DOI:10.1088/1748-9326/9/8/084018.

Bertram, C., Johnson, N., Luderer, G., Riahi, K., Isaac, M., and Eom, J. (2015). Carbon lock-in through capital stock inertia associated with weak near-term climate policies. Technological Forecasting and Social Change, 90, Part A. 62–72. DOI:10.1016/j.techfore.2013.10.001.

5 Erickson, P., Kartha, S., Lazarus, M., and Tempest, K. (forthcoming). Assessing Carbon Lock-In. Environmental Research Letters.

Guivarch, C. and Hallegatte, S. (2011). Existing infrastructure and the 2°C target. Climatic Change, 109(3-4). 801–5. DOI:10.1007/s10584-011-0268-5.

IEA (2013). Redrawing the Energy-Climate Map: World Energy Outlook Special Report. International Energy Agency, Paris.

6 IEA (2013). Redrawing the Energy-Climate Map.

7 Gurría, A. (2013). The Climate Challenge: Achieving Zero Emissions. Lecture by the OECD Secretary-General. London, 9 October 2013.

See also the Stop Arctic Ocean Drilling Act of 2015, introduced by U.S. Senator Jeff Merkley and others in July 2015:

8 Collier, P. and Venables, A.J. (2014). Closing coal: economic and moral incentives. Oxford Review of Economic Policy, 30(3). 492–512. DOI:10.1093/oxrep/gru024.  See also Frumhoff, P. C., Heede, R. and Oreskes, N. (2015). The climate responsibilities of industrial carbon producers. Climatic Change, online 23 July. DOI:10.1007/s10584-015-1472-5.

Faehn, T., Hagem, C., Lindholt, L., Maeland, S. and Rosendahl, K.E. (2013). Climate Policies in a Fossil Fuel Producing Country: Demand versus Supply Side Policies. Statistics Norway, Research Department, Discussion Paper 747, Oslo.

9 Vivid Economics (2009). G20 Low Carbon Competitiveness. The Climate Institute and E3G.

10 Leaton, J., Ranger, N., Ward, B., Sussams, L. and Brown, M. (2013). Unburnable Carbon 2013: Wasted Capital and Stranded Assets. Carbon Tracker and Grantham Research Institute on Climate Change and the Environment, London School of Economics, London.

11 Gurría (2013). The Climate Challenge: Achieving Zero Emissions.

12 A distinction is frequently made between fossil fuel “reserves” – volumes that can be produced economically with current technology – and “resources” – volumes that are not yet fully characterized or require more advanced technologies to extract cost-effectively. Here we use a broader definition of “resources”, to mean ultimately recoverable resources, including both those categories. Our usage is in line with: McGlade, C. and Ekins, P. (2015). The geographical distribution of fossil fuels unused when limiting global warming to 2°C. Nature, 517(7533). 187–90. DOI:10.1038/nature14016.

13 McGlade and Ekins (2015). The geographical distribution of fossil fuels unused when limiting global warming to 2°C.

Leaton et al. (2013). Unburnable Carbon 2013.

14 IEA (2014). World Energy Outlook 2014. International Energy Agency, Paris.

For a quick overview of the scenarios, see: Note that the IEA also has a Current Policies Scenario that is a more classic BAU, assuming no changes from existing policies.

In line with the concept of carbon lock-in, we focus solely on CO2 emissions from fossil fuel combustion, though we must note that other GHGs, notably methane, are released in the course of extracting and processing fossil fuels. Such emissions can often be reduced through cost-effective improvements in extraction and processing operations, however, and may thus be less “locked-in”.

See: U.S. EPA (2012). Global Anthropogenic Non-CO2 GHG Emissions: 1990–2030. EPA 430-R-12-006. U.S. Environmental Protection Agency, Washington, DC.

15 Dixit, A. (1989). Entry and exit decisions under uncertainty. Journal of Political Economy, 97(3). 620–38.

Leis, J. (2015). What the Recent Oil Price Shock Teaches about Managing Uncertainty. Bain Brief. Bain & Company.

16 Rystad Energy (2015). UCube, Version 1.18.


17 Leaton et al. (2014). Carbon Supply Cost Curves. We make the simplifying assumption that the difference between the “breakeven coal price” and “cash cost” in this source is solely capital investment.

18 Gurría (2013). The Climate Challenge: Achieving Zero Emissions.

19 Beblawi, H. (1987). The Rentier State in the Arab World. Arab Studies Quarterly, 9(4). 383–98.

20 Rystad Energy (2015). UCube, Version 1.18.

21 Leaton et al. (2014). Carbon Supply Cost Curves.

22 Bertram et al. (2015). Carbon lock-in through capital stock inertia associated with weak near-term climate policies.

Erickson et al. (forthcoming). Assessing Carbon Lock-In.

23 Rystad Energy (2015). UCube, Version 1.18.

24 McGlade and Ekins (2015). The geographical distribution of fossil fuels unused when limiting global warming to 2°C.

25 Bauer, N., Mouratiadou, I., Luderer, G., Baumstark, L., Brecha, R. J., Edenhofer, O. and Kriegler, E. (2013). Global fossil energy markets and climate change mitigation – an analysis with REMIND. Climatic Change, online 22 October. DOI:10.1007/s10584-013-0901-6.

26 Collier and Venables (2014). Closing coal: economic and moral incentives.

27 Sinn, H.-W. (2012). The Green Paradox: A Supply-Side Approach to Global Warming. The MIT Press, Cambridge, MA, US.

28 Davis, S.J., Peters, G.P. and Caldeira, K. (2011). The supply chain of CO2 emissions. Proceedings of the National Academy of Sciences, 108(45). 18554–59. DOI:10.1073/pnas.1107409108.

Peters, G.P., Davis, S.J. and Andrew, R. (2012). A synthesis of carbon in international trade. Biogeosciences, 9(8). 3247–76. DOI:10.5194/bg-9-3247-2012.

29 Collier and Venables (2014). Closing coal: economic and moral incentives.

Faehn et al. (2013). Climate Policies in a Fossil Fuel Producing Country.

© Flickr / Maersk Drilling

The Maersk Intrepid, the first of four jack-up rigs built for ‘ultra-harsh’ environments, at the Keppel FELS shipyard in Singapore, before being mobilized to drill wells on the Martin Linge field development in the Norwegian North Sea.

Published by:

Stockholm Environment Institute

1402 Third Avenue, Suite 900

Seattle, WA

98101, USA

+1 206 547 4000

Author contact:

Peter Erickson,

Media contact:

Marion Davis,