In a paper published in the journal Joule, researchers at Imperial College London (ICL) claim that studies that predict whole systems can run on near-100% renewable power by 2050 may be flawed as they do not sufficiently account for the reliability of the supply.
Using data for the UK, the team tested a model for 100% power generation using only wind, water and solar (WWS) power by 2050. The ICL researchers found that the lack of firm and dispatchable “backup” energy systems, such as nuclear or power plants equipped with carbon capture systems (CCS), means the power supply would fail often enough that the system would be deemed inoperable. They found that even if they added a small amount of backup nuclear and biomass energy, creating a 77% WWS capacity system, around 9% of the annual UK demand could remain unmet, leading to considerable power outages and economic damage.
Lead author Clara Heuberger, a PhD student from the Centre for Environmental Policy at Imperial, said: “Mathematical models that neglect operability issues can mislead decision makers and the public, potentially delaying the actual transition to a low carbon economy. Research that proposes ‘optimal’ pathways for renewables must be upfront about their limitations if policymakers are to make truly informed decisions.”
Co-author Niall Mac Dowell, also from the CEP, and director of the Clean Fossil and Bioenergy Research Group, said: “A speedy transition to a decarbonised energy system is vital if the ambitions of the 2015 Paris Agreement are to be realised. However, the focus should be on maximising the rate of decarbonisation, rather than the deployment of a particular technology, or focusing exclusively on renewable power. Nuclear, sustainable bioenergy, low-carbon hydrogen, and carbon capture and storage are vital elements of a portfolio of technologies that can deliver this low carbon future in an economically viable and reliable manner. Finally, these system transitions must be socially viable. If a specific scenario relies on a combination of hypothetical and potentially socially challenging adaptation measures, in addition to disruptive technology breakthroughs, this begins to feel like wishful thinking.”
The study made use of Mark Jacobson’s 100% WWS generic global scenario, but the ICL team had to modify it for their UK version so that they could test its reliability using a system optimisation model. That would not let them retain the wave and tidal inputs that Jacobson had included (2.5% & 1.8% respectively), so proxies were used. The high level of solar PV (~40%) was also not viable in the test model. The final result was that high levels of curtailment were identified (33%), though it was noted that “this could be a loss or an opportunity for processes using this excess power” i.e. power to gas conversion of surpluses, but this wasn’t followed up. It was also noted that “further, demand-side management, which is not included in this model, could alleviate curtailment levels”. There were significant shortfalls, but it might be that the omissions above could go some way to avoiding them.
That’s not to say that full balancing is easy, or that every model does it right, and earlier more extensive ICL work had pointed out the issues. However, the assertion in this new study that nuclear and fossil CCS are vital may raise some eyebrows. Nuclear seems unlikely to be able to make a major contribution to grid balancing (see my earlier post) since it is too inflexible and CCS, with many projects around the world abandoned, seems increasing unlikely to play a major role – an issue I will be exploring shortly in a series of posts.
The promotion of nuclear is, of course, still relentless, despite all the problems. Some claim that nuclear power can scale up quickly enough to meet the global climate threat, while renewables cannot. However, according to US energy guru Amory Lovins, “global and national data show the opposite”. He and his co-authors identify a range of errors, biases and misinterpretations in the studies claiming to prove that nuclear growth has been faster than for renewables. Some focus on programmes in individual countries on a per capita basis. For example, looking at a paper by Cao, Hansen et al., Lovins et al. say “Swedish nuclear power (which in 1976–86 grew 4.4× to 70 TWh/y) is shown as scaling 55× faster than Chinese wind-power (which in 2004-14 grew 124× to 158 TWh/y) – because Sweden’s population averaged 1/158th of China’s. Conversely, China’s unique addition in less than a decade (through 2016) of 25% of global solar photovoltaic (PV) and 35% of global wind-power capacity is shown as the slowest national achievement – an odd description of the nation that in 2016 added over 40% of new global renewable electric capacity, because it’s divided by 1.4 billion Chinese”.
Looking at the global data in absolute terms gives a very different picture. Even comparing specific programmes in specific countries, renewables expansion has been faster relatively than nuclear expansion, as the paper shows in the case of France (nuclear) and Germany (renewables). The former took 30 years to get going, the latter 9 years – as did China’s renewables programme. The timescale comparisons are sometimes even used against renewables. Nuclear has been getting support for many decades, and although that led to growth early on, it has now stalled. Renewables have been ignored until recently. So decadal comparisons are not very helpful since they dilute the impact of renewable’s recent expansion, and even in some cases, by using earlier cut-off dates, ignore it, and the recent nuclear stasis. Similar issues emerge in relation to cost estimates and project completion rates. Nuclear scores poorly on both with, the paper claims, few credible signs of improvement, whereas the costs of renewables are clearly falling rapidly and implementation rates are rising.
There is some room for debate on the cost claims. Variable renewables need balancing, which adds to the cost, and grid integration problems have led to wasteful curtailment of output, notably in China. However, these are not fundamental problems: upgraded smart grid systems with storage and supergrid links can balance variable supply and demand, making the overall system more efficient and cheaper to run than the existing inflexible system, while avoiding the high cost of using nuclear and fossil fuels – points made by Lovins in an earlier study.
In this new paper, Lovins et al. end by saying that they hope their exposition of “how up-start, granular, mass-produced technologies can overtake a powerful centralized incumbent may illuminate whether the pace of global decarbonization must inevitably be constrained by incumbents’ inertias, could be sped by insurgents’ ambitions, or perhaps both”.
Clearly they think renewables can expand fast, which goes against an earlier study by the equally authoritative Vaclav Smil, who concluded that “replacing the current global energy system relying overwhelmingly on fossil fuels by biofuels and by electricity generated intermittently from renewable sources will be necessarily a prolonged, multidecadal process”.
That led to some challenges, including from Ben Sovacool at the Science Policy Research Unit at the University of Sussex, UK, drawing on some historical examples of rapid change, and from a range of other academics.
But perhaps the best challenge comes from the actual progress being made: see the new REN21 annual review, which I look at in my next post.