The IEA's World Energy Outlook 2023

A new energy world by 2030

The International Energy Agency (IEA) released its World Energy Outlook 2023 on 24 October. As with last year, there are three scenarios: STEPS (business as usual); APS (policy in line with commitments); and NZE (net zero). 

Key messages.  These resonate very much with the RMI framing.

1. Renewables are growing fast, with at least 10x as many EV on the road, and renewables supplying half of all power by 2030. Change is ‘unstoppable’.

2. Growth in China is slowing and becoming less energy-intensive. This matters because China has been the key driver of energy demand growth, and there is nobody to replace China in that role.

3. So we will have a dramatic change to the energy system by 2030.  Disruption is baked in.

4. We need to do more to get to 1.5 degrees.  Above all we need to ‘scale up investment in all aspects of a clean energy system’. So, triple renewable capacity, double efficiency and reduce methane emissions by 75%. Also triple clean energy capex in emerging markets and ensure an orderly decline of fossil fuels. 

Other messages, with our comments in italics.

1. Peak fossil fuels.  In every scenario before 2030. Peaking demand for many aspects of the fossil fuel system are behind us. Peak new steel plants (2003), coal plants (2012), ICE cars (2017), fossil heating systems (2019), ICE trucks (2019) and so on. Coal, oil and gas are on plateaus which will be either short or long dependent on the scenario. And ‘technology cost trends point in the direction of increasingly stiff competition for fossil fuels from clean energy technologies across a wide range of market segments’.

2. Peak primary energy demand.  Under APS and NZE we have hit peak primary energy demand.  Because efficiency gains driven by renewable technologies are equal to or higher (3% and 4.1%) than GDP growth to 2030 (3%).

3. The capex shift is well under way.  Fossil fuel capex used to be higher than renewable capex.  In 2023 renewables were 1.8x as much as fossil fuels and by 2030 they will be between 2.5 and 10 times as much.

4. Renewables drive the energy system.  The key driver of change is the speed of renewable deployment, especially solar (lighting the way) and EV. This might seem obvious, but it fits very much into our framing that renewables are the key variable in the system and everything else fits around them.

5. Renewables up, fossils down.  The IEA continue to increase forecasts for renewables and reduce fossil fuel demand. For example, STEPS 2050 renewables supply forecast is 227 EJ, up from 215 EJ last year. Meanwhile STEPS oil demand in 2050 is 186 EJ, down 11 EJ from last year.  The APS scenario has seen a similar shift: APS 2050 renewables supply is 327 EJ, up from 319 EJ last year. And APS 2050 oil demand is 102 EJ, down from 108 EJ last year. 2030 forecasts show a similar shift, with renewables up and total demand down on last year’s forecast.

6. The future gap.  The fossil fuel industry is building for STEPS, the cleantech industry as a whole is building for APS, and solar and EV are on track for NZE. There is of course only one future, so someone will end up with stranded assets. Given the superior economics and politics of renewables, that implies stranded assets for the fossil fuel sector.

7. Peak fossil fuels in China.  Chinese demand for fossil fuels peaks in every scenario by 2025. Because of lower growth, rapid renewable deployment and a shift to a less energy intensive economy. They also look at the (considerable) downside to fossil fuel demand if Chinese growth slows to less than their forecast (4%).

8. End of the golden age of gas.  Gas demand growth has slowed dramatically and demand will either be stagnant or decline. That means some of the huge number of LNG export plants being built will end up facing a gas glut.

9. CCS is only a small part of the solution.  CCS is not growing as fast as other technologies, and it should not be used as an excuse to avoid reducing fossil fuel emissions. As noted before, CCS is a small and shrinking part of the endgame.

10. The risk for solar is to the upside.  Given the massive increase in solar capacity, the rollout of solar may happen faster than in any of their core scenarios.

11. The grounds for hope.  Policy is changing, deployment is happening fast, we have the tools, and new technologies are coming.

12. Areas needing most work:  Deploy capital into the emerging markets, build infrastructure (especially grids), increase resiliency, and work together.

Myth busting.  A few observations on how the IEA helps to defuse some of the incumbent fossil fuel arguments against change.

1. Putin’s war meant faster change.  It has ushered in ‘the beginning of the end’ of the fossil fuel system and ‘brought forward the emergence of a new clean energy economy’.

2. Renewables are a better energy solution than fossil fuels.  Renewables benefit society in a range of ways. They offer energy security, industrial opportunity, jobs (17m net gain by 2030), clean air (APS saves 1.5 million lives pa by 2030 versus STEPS), universal energy access, lower costs, and a safer climate.  

3. Renewables are a better option for a fragmented world.  ‘Opportunities to produce low-emissions energy cost competitively are far more widely distributed around the world than is the case for fossil fuels. It would therefore be natural to assume that, as a largely domestic source of energy, renewables would fare better in a more fragmented international context’.

4. Emerging markets will not drive fossil fuel demand growth.  They are not large enough to outweigh falling demand in China and the OECD, and they too are adopting renewable technologies. Meanwhile ‘solar PV is poised to accelerate clean energy transitions around the world’.

5. A renewable system is cheaper than a fossil fuel one.  The total cost is lower and the cost per household is lower in developed markets. The extra money spent on capex is rapidly repaid by lower opex on expensive fossil fuels.

6. Renewables give options to Africa.  80-85% of the growth of new generation in Africa to 2030 is likely to be renewables. Cleantech enables the provision of clean cooking and electricity to those without.

7. We can do this thing.  80% of the solution by 2030 is well known and feasible: to triple renewables, double efficiency and reduce methane losses by 75%.

8. Industry and emerging markets will not drive growth in gas demand.  Because the two largest areas of demand (power and buildings) have seen a peak.

9. There are solutions to renewable variability.  They note that as renewables increase their deployment, so we need much more flexibility. And this will come mainly from batteries and demand side response.

10. Barriers to change are not insoluble.  The transition is described as ‘unstoppable’. There are issues, as with any rapid transition, but they have solutions.

11. The fossil fuel industry is not underinvesting.  They are investing for STEPS at the same time that STEPS expected fossil fuel demand in 2030 keeps falling. A favourite trope of fossil fuel promoters is to argue that we are underinvesting in fossil fuels. The IEA busts this myth.

12. Grid capex for a rapid energy transition is only 20% more than BAU.  The STEPS scenario requires 2030 grid capex of $565bn pa whilst NZE needs $680bn. So this implies annual capex of 20% more under the NZE, which seems feasible for such a central issue in the energy transition.

13. Importers can replace lost fossil fuel revenue.  As fossil fuel demand falls, so many worry that this revenue source will impact government budgets. The IEA notes that for importers fossil fuel consumption taxes can be matched by a fall in fossil fuel subsidies and an increase in carbon pricing. 

14. OPEC and the US oil companies are wrong to think oil demand will keep rising.  This question was addressed directly by Dr Birol at the press conference, who suggested that it might be wise for them to stop talking to each other and get out more. And talk to people in the solar, heat pump and EV industries who are driving change. The mystery in this debate is why anyone thinks that OPEC would have an objective view on the future of oil demand; incumbents always talk their book and miss technology shifts.

Critique

1. Scenario choice.  Whilst the central scenario is presented as STEPS, we believe that STEPS should now be seen as an outlier (normative) scenario because it requires a massive slowdown in the ongoing transition. It is more credible to take the APS as a central scenario and model a future between the APS and the NZE scenarios.

2. Upgrade NZE.  It would be good to integrate NZE as a more key scenario. Rather than just referring to it from time to time.

3. GDP growth the same.  As we keep noting, it is simply wrong to assume similar levels of GDP growth under scenarios with and without dangerous levels of global warming.

4. Unambitious renewable forecasts this decade.  The forecasts for renewable growth under STEPS and APS are quite low. New solar additions in 2030 for example are around 500 GW pa in STEPS and 600 GW under APS, when capacity build-out and S curves suggest they are likely to be 800 GW or higher. Expectations are for EV under STEPS in China to be 65% of sales in 2030, when S-curve modelling implies they will be over 90% of sales.

5. Limited renewable cost differentiation.  Forecast renewable costs are expected to be similar under the three scenarios. This is unambitious; the existence of learning curves suggests that the more renewables we deploy, the lower the cost, as analysed in detail by Doyne Farmer and the Oxford INET team. For example, US solar LCOE in 2050 is the same under all three scenarios; given that deployment of solar is 2x that of STEPS in the NZE scenario, we would expect costs to be around 20% lower.

6. Unambitious on renewable costs and growth after 2030.  Solar cost falls after 2030 drop to just 1% per annum. After 2030, solar growth in STEPS drops to 8% pa, and after 2040, solar growth under all three scenarios drops to 3-4% pa. The result of this is that the scenarios are conservative on the share of electricity in 2050 final demand (30-53%) and have room for very large amounts of biomass (up to 100 EJ in 2050), CCS (up to 6,000 mt in 2050), and need high CO2 costs (up to $250 per t in developed markets in 2050 under NZE). The share of electricity in final demand in 2050 rose by two percentage points under both STEPS and APS, and over time is likely to continue to be increased.

 The link to the report: https://www.iea.org/reports/world-energy-outlook-2023