Falling costs of electric vehicle and solar technology could halt growth in global demand for oil and coal from 2020, finds a new report co-authored by the Grantham Institute at Imperial College London and the Carbon Tracker Initiative launched on Thursday. The future of natural gas is highly uncertain in this new scenario analysis.
Carbon Tracker Initiative (CTI), the London-based NGO which invented the concept of “stranded assets”, warns that “big energy companies are seriously underestimating low-carbon advances with a business-as-usual (BAU) approach, and that stranding of fossil fuel assets is likely as the low- carbon transition gathers pace.”
This warning in itself is not new, but CTI’s latest report – Expect the Unexpected: the Disruptive Power of Low-Carbon Technology – is based on new research. It “explores how plausible advances in solar PV and EVs alongside efforts to reach international climate targets could impact on future fossil fuel demand. It models a range of scenarios using the latest data and market projections for future cost reductions in PV and EVs, with varying levels of global climate policy effort and energy demand”.
In the graph below you can see the building blocks for the scenarios that were explored:
According to CTI, the ‘Cancun-Original Cost’ scenario is a typical “business-as-usual” scenario used by energy companies, from which they then extrapolate into the future. But the authors argue that this sets up a misleading base case, because the present already looks very different. (All the scenarios can be viewed with an interactive online tool here: www.carbontracker.org/expect-the-unexpected-dashboard)
Instead, the CTI study uses ‘NDC_PV_EV’ as a starting point scenario, i.e. an assumption that the Paris pledges (NDCs or nationally determined contributions) are met and that costs of solar PV and EVs (electric vehicles) continue to decline. “While the outcomes in this scenario are not guaranteed, this pathway is a credible reflection of the low carbon transition as indicated by current technology trends and policy commitments, and should be used as the new starting point for any scenario analysis”, say the authors.
Gas is almost wiped out in the power sector, but nuclear power remains very strong
Note that even this scenario does not achieve a 2 degrees C outcome, but would limit global warming to 2.4-2.7 degrees C by 2100, which is “significantly lower than business-as-usual scenarios to 4 degrees C and over often used by the energy industry”.
The “new starting scenario” used by CTI finds that:
- “Solar PV could supply 23% of global power generation in 2040 and 29% by 2050, entirely phasing out coal and leaving natural gas with just a 1% market share. By contrast, ExxonMobil sees all renewables supplying just 11% of global power generation by 2040.
- EVs could make up a third of the road transport market by 2035, more than half the market by 2040 and more than two thirds of market share by 2050. BP’s 2017 outlook expects EVs to make up just 6% of the market in 2035.
- Coal demand could peak in 2020 and fall to half 2012 levels by 2050. Oil demand could be flat from 2020 to 2030 then fall steadily to 2050. Most major oil and gas companies do not expect coal to peak before 2030 and none see peak oil demand occurring before 2040.”
CTI warns investors that “fossil fuels may lose 10% of market share to PV and EVs within a single decade — this may not sound much but it can be the beginning of the end once demand starts to decline. A 10% loss of power market share caused the collapse of the US coal mining industry and Europe’s five major utilities lost more than €100 billion in value from 2008 to 2013 because they were unprepared for an 8% growth in renewable power, of which solar PV was a big part.”
Power sector impacts
If we take a closer look at the power sector, we see that the impact of solar PV on fossil fuel alternatives will be devastating by 2050 under the study’s starting point scenario:
Interestingly, in 2050 in the Strong_PV scenario, gas is almost wiped out in the power sector, but nuclear power remains very strong. Note that the model is a “least-cost optimising model”, so it assumes that technologies will be employed on the basis of cost. It does assume certain “technology growth constraints”, i.e. if a certain technology, say wind power, is the cheapest, the model will put a limit on its deployment, since in the real world no single technology would be used exclusively. The assumptions behind the model can be found on the internet here: www.carbontracker.org/expect-the-unexpected-technical-report
According to CTI, “To grow to 60% of global power capacity by 2050 will require a huge build-out of solar PV, much of which occurs between 2030 and 2040. Scenario ‘NDC_PV’ sees solar PV capacity grow by over 5000GW in this 10 year period, for example. Such a rapid deployment of solar PV reflects a shift that could arise when solar PV becomes materially cheaper than alternative power options, and utility and consumer preferences change accordingly. In such a scenario of significant change, the mass stranding of downstream fossil fuel assets is highly likely.”
With regard to gas, if readers from the gas industry are alarmed at the predictions above, it should be noted, as the report puts it, that “The black and white nature of least-cost optimising models means they can flip-flop between favoured technologies depending on the cost-competitiveness landscape in that decade. This happens with natural gas-fired power generation in our scenarios …”
In fact, even when assuming higher energy demand levels, driven through higher global economic growth assumptions, it appears that overall fossil fuel demand would still plateau from 2030 onwards
“After a small increase in demand to 2020, all our scenarios see natural gas-fired generation declining to 2030. In scenarios with original solar PV costs, this is due to the widespread take-up of biomass as a power source; in lower cost scenarios, solar PV picks up some demand. By 2040, scenario ‘NDC_Orig’ has stringent enough climate policy for coal demand to collapse, but not to overly penalise natural gas – hence a temporary spike in generation as previously idled plants are brought back online. In scenarios with lower solar PV costs, natural gas-fired power generation continues to decline through to 2050 as it is out-competed. …”
The important point to note here is that the outcome for gas is not inevitably bad, but very uncertain: “The volatility seen in gas generation here could be interpreted to reflect how marginal natural gas options are. This is already the situation in some markets such as Europe where the uncertainty over future carbon, commodity and power prices make it almost impossible for commercial operators to decide to invest in new gas generation.
The signals being sent from some governments regarding the phase out of coal create some opportunities for gas as a transition fuel but as the model shows this may only be a temporary respite, as gas is outcompeted by alternatives soon after, resulting in cripplingly low utilisation rates and asset stranding. Although no carbon budget is in place in these scenarios, this would constitute a further problem for those considering gas as a transition option – if ever tightening emissions constraints are to be met, then new gas plants would have a very limited lifetime.”
EVs and oil demand
The scenarios for EVs show similar dramatic differences in their outcomes, depending on the assumptions of cost and policies:
What does this mean for oil demand? The “scenario ‘NDC_EV’, i.e. the scenario assuming an NDC consistent level of climate policy action combined with lower EV costs, sees 16.4 million barrels of oil per day (Mbd) being displaced annually by 2040 due to EV penetration in the road transport sector”, notes the report. “By 2050 this figure is 24.6 mbd in ‘NDC_EV’. This level of oil displacement due to EVs is a little above that from BNEF’s [Bloomberg New Energy Finance] 2016 New Energy Outlook (NEO). Both are significantly above the IEA’s 2016 New Policies Scenario (NPS) [the central scenario of the International Energy Agency, editor] and even the IEA’s 2°C (450) scenario that shows EVs could displace 6 mbd of oil demand by 2040.
The fossil fuel industry is equally conservative about the potential for EVs to displace demand for oil – BP’s 2017 energy outlook sees 1.2mbd being displaced by switching to EVs by 2035 and this is after a significance increase from the 2016 outlook that put this figure at 0.7mbd.”
Oil demand at this moment is around 90 mbd, but the report notes that “the recent 2014-15 oil price collapse was as a result of a 2% shift in the supply-demand balance, roughly 2mbd. The IEA’s New Policies Scenario does not see 2 mbd of oil being displaced by EVs before 2040. BNEF sees this displacement threshold being surpassed by 2028, while ‘NDC_EV’ sees 2 mbd of oil displaced as soon as 2025.”
Peak fossil fuel demand
When could peak fossil fuel demand occur, in CTI’s scenarios? Quite soon, as can be seen in this figure:
According to CTI, “The chart shows exactly how significant the 2015 Paris Agreement could be – climate policy effort consistent with the NDCs sees total fossil fuel demand peaking in 2030 even with original technology cost assumptions (‘NDC_Orig’). Applying lower cost solar PV and EV assumptions instead results in an 11% reduction in fossil fuel demand against this scenario by 2050 (‘NDC_PV_EV’). Reduced energy demand serves to accentuate the decline after the 2030 peak even more. Any policy effort beyond the NDCs – ‘Strong_PV_EV_Low’ for example – could result in global fossil fuel demand peaking in 2020 and subsequent CO2 emissions taking a much lower trajectory.”
The report adds that “In fact, even when assuming higher energy demand levels, driven through higher global economic growth assumptions (‘NDC_PV_EV_High’), it appears that overall fossil fuel demand would still plateau from 2030 onwards. It is worth noting, however, that recent OECD economic growth forecasts suggest the world economy is on a growth path to 2050 which is somewhere between the levels assumed in our medium and lower energy demand scenarios, a far cry from the level in the higher demand scenario.”
What the energy industry has to say about future coal, oil and gas demand is very different from what CTI has to say
Gas is again the most uncertain of the three fossil fuels. According to the report, “Integrated assessment models … tend to position natural gas as a transition fuel when no CO2 emissions constraints are put in place, like in our scenarios. This means the decline of coal is to a degree substituted with natural gas, such as in scenario ‘NDC_Orig’… When lower solar PV costs are applied, gas demand growth is lower – refer scenario ‘NDC_PV_EV’. In such a scenario, however, lower power sector usage of gas can also reduce the cost of using it in industry or heating buildings, which may rise to compensate overall gas demand somewhat. Lower energy demand reduces natural gas demand growth across all sectors, but it is only in our most bullish ‘Strong_PV_EV_Low’ scenario that we see natural gas demand peak in 2030 and fall thereafter.” For the gas industry, “bullish” would of course be “bearish”.
You won’t be surprised to hear that, as the report puts it, “what the energy industry has to say about future coal, oil and gas demand” is very different from what CTI has to say. As far as CTI is concerned, energy companies such as BP and ExxonMobil should stop employing business as usual” scenarios, update their model assumptions and be more transparent about the assumptions behind their models.
You can read the full report here.
[adrotate banner=”82″]
Are Hansen says
“In scenarios with original solar PV costs […]”
That is meaningless. We know that PVs will fall in price, it has been going on for a long time already, and shows no signs of abating. And that is with the same, silicon based tech. With tech breakthroughs (like upcoming perovskite, graphene, …) the price will drop even faster.
2040 and 2050 is far away in time. The conversion efficiency of PVs will be way higher and prices will be way lower
Ferdinand Engelbeen says
Not only the price of PV will be determinative, the price and global capacity of storage will be even more important. The difference in yield for solar between summer and winter is about a tenfold. That means that one either need an enormous storage or other means of generation in winter, when demand is highest.
Another point, in all PV scenario’s, nuclear delivers a significant part of the (base)load. Tell that to the European Commission, which doesn’t allow nuclear as part of a CO2-free solution…
Further:
Note that even this scenario does not achieve a 2 degrees C outcome, but would limit global warming to 2.4-2.7 degrees C by 2100
Sorry, that is based on climate models, which – to say the least – are not performing that good. Real world increase since 1950, until now, will end around 1.2-2.2 degrees C for a bussiness as usual scenario, if the trend continues as is observed.
That may be food for another, more fundamental discussion… See e.g.:
http://journals.ametsoc.org/doi/abs/10.1175/JCLI-D-12-00473.1
Phillip Kopp says
There seems to be a missing element. The EIA in the us associates an additional cost to renewable energies related to complexities in control and infrastructure for T&d. In fact the more renewable you put into the system the less competitive nuclear becomes. We already see everywhere the nuclear generators closing on cost basis. You are only left with coal and gas to fill the gaps, until the renewable is so prevalent it’s on the base margin, which means you will see a lot of spilling and dramatically increase the unit cost. We will have to store all the energy long before gas or coal becomes totally uncompetitive. It doesn’t seem like there is any realistic near term technology so the 2050 picture is HIGHLY theoretical, based on a storage technology which today does not yet exist or even science envisioned. While the scenario of higher renewable and ev is likely, the analysis here is fundamentally flawed and missing all kind of components related to margin and infrastructure.