Gordon Hughes: Response To Goodall & Lynas
“The assertion that wind turbines don’t reduce carbon emissions is a myth, according to conclusive statistical data obtained from National Grid and analysed here in the Guardian for the first time,” Chris Goodall and Mark Lynas claimed in a recent Guardian blog.
Professor Gordon Hughes responds:
1. It is unfortunate that the polarisation of positions renewable energy has coarsened arguments about the impact of wind and solar power on electricity systems. This has been apparent in a series of postings on, inter alia, various blogs hosted by The Guardian and associated publications. Some weeks ago Maria McCaffery, Chief Executive of RenewableUK, published some absurd statements about one of my reports for the GWPF. It is the job of lobby groups to defend the indefensible, so there is no reason to give such comments more weight than they deserve. However, on September 26th Chris Goodall and Mark Lynas (Goodall-Lynas) published a blog purporting to provide empirical evidence disproving suggestions that wind power might not reduce CO2 emissions by as much as the advocates of wind development claim. While their logic and evidence are only marginally stronger than the usual claims, it is worth examining the details of the argument in order to clarify the central points at issue in this debate.
2. The Goodall-Lynas claim rests upon setting up and then knocking down a straw man. It is a variant of the one-for-one substitution calculation that is used to support claims that a proposed wind farm will “save” X tons of CO2. The essence of such claims is that 1 MWh of electricity generated by wind will directly replace 1 (or somewhat less) MWh generated by power plants using coal, gas or oil. This is a straw man because, to the best of my knowledge, all of those who question the contribution of wind power to reducing CO2 emissions argue that this is an incomplete and very one-sided description of the impact of wind power on the electricity system as a whole. The evidence produced by Goodall-Lynas is little better because they have made no attempt to take account of these broader impacts. Further, they have chosen to represent my analysis of the impact of the UK’s policies in 2020 as if it is relevant today, which I have explicitly ruled out.
3. There is no dispute that the amount of electricity generated by wind farms cannot be controlled in the same way as the electricity generated by a gas plant. If the amount of wind is too small or too large, wind plants do not operate either at all or at full output regardless of whether demand for electricity is low or high. It is how this intermittency affects the design and operation of the electricity system as a whole which lies at the heart of the dispute.
4. In fact, the issue is rather more complicated than this statement implies because the variability of wind output has different dimensions and, thus, induces different responses. The total output from wind farms can vary significantly in the very short term – for example, from one 5 minute period to the next. It is impossible to start up or shut down thermal plants so quickly in order meet demand, so that such fluctuations in wind output must be offset by drawing upon plants that are already running but which are not feeding electricity into the grid – called spinning reserve. Even with no wind power a margin of spinning reserve is required to maintain the frequency and voltage of electricity supplies and to cope with unexpected failures or fluctuations in demand. If the level of wind generation is a small proportion of total demand, short term fluctuations in wind output can easily be accommodated by drawing upon the margin of spinning reserve. This is the case in the UK in 2012 as wind output is typically less than 10% of total demand.
5. This is the first point where the Goodall-Lynas evidence is incomplete. It relies upon data about the plants which are supplying electricity to the grid. It takes no account of the CO2 emissions of plants that are operating as spinning reserve. For simplicity, let us suppose that all spinning reserve is provided by gas combined cycle plants (CCGTs). If changes in wind output are balanced by changes in the level of spinning reserve, then the total amount of gas that is burned – and, thus, CO2 emissions – is completely independent of change in wind output. In terms of the Goodall-Lynas evidence, higher levels of wind generation displace gas generation one-for-one. But, there is absolutely no saving in CO2 emissions because the gas plants carry on running as before but they are just feeding less electricity into the grid. The reason for the error is that their figures take no account of what is happening in the parts of the electricity system that they have ignored.
6. There is another problem with the Goodall-Lynas evidence, even as a description of short term responses to wind variability. They state that their data covers “the last 3 months”. To be precise – and allowing time for data processing and writing the article – let us assume that this is June 16th to Sept 15th. This is the worst possible period to carry out any analysis of such data. Both electricity demand and wind output are at their seasonal troughs in June, July & August, so that there is a huge margin of spare capacity. The analysis tells us little about how the electricity system will respond when the margins of spare capacity are smaller and over periods that require adjustments in the composition of available and operating capacity.
7. Once we look beyond very short term variations in wind output a different set of issues come to the fore. These are discussed at length in my GWPF report titled Why is wind power so expensive? so I will only provide a brief summary here.
8. To maintain secure reserve margins, each MW of wind generating capacity has to be backed by approximately 1 MW of generating plant which can be run on demand – called dispatchable plant (and interconnectors don’t count in such calculations). The ideal backup for wind power is hydro, but the UK has very limited opportunities to develop more hydro plants. At the moment, the UK electricity has large amounts of old coal, gas and oil plants that provide backup for the limited amounts of wind and solar capacity. That situation will change as the amount of installed wind capacity increases and older fossil fuel plants are decommissioned. Hence, the problem of the nature and performance of backup to wind power only starts to arise after 2015 and becomes acute by 2020.
9. The argument then rests on a comparison between two options for meeting projected electricity demand in 2020. One option is to continue with the government’s strategy of promoting renewable energy, so that a substantial fraction of total demand in 2020 is met from wind power. My calculations suggest a requirement for 36 GW of wind capacity in 2020, while current official projections suggest that wind capacity would be 28-29 GW. The difference is partly a consequence of downward revisions in official projections of electricity demand in 2020. More importantly the official estimates are based on what are, in my view, much too optimistic views of the load factor for wind plants in 2020. Nonetheless, on either view this is a large increase in wind capacity relative to the current figure of about 7.5 GW and it will require a large expansion in dispatchable plant to back up intermittent wind energy.
10. The second option is to construct a sufficient number of CCGTs to meet the demand that would be met from wind generation and associated backup. This is what the market would deliver without subsidies and targets for renewable energy, even after allowing for a carbon price that is higher than the floor level for 2020 stated in official policies. Since it would be the least cost way of meeting both electricity demand and CO2 emission targets in 2020, this option must be treated as the reference or baseline scenario. Hence, all comparisons of CO2 emissions and costs refer to changes in 2020 relative to the baseline scenario. This is the correct way of assessing the future impact of policies.
11. In this framework, the reduction in CO2 emissions achieved by the government’s current policies is much smaller than is implied by one-for-one substitution, though the exact outcome depends upon how the gap left by intermittent wind generation is filled. In Germany, for example, it seems likely that the gap may be filled by coal plants and the result will almost certainly be an increase in CO2 emissions relative to a baseline of gas but no wind and coal. In the UK it is more likely that the gap will be filled by gas plants. The problem, which is rarely understood by those who don’t go into the details of system operation, is that the typical load factor for backup plants will be very low – well below 20%. This is not unusual in systems that are dominated by hydro power. Investors do not build CCGTs to meet such demand. That is why I have argued that the backup plant will be dominated by open cycle gas plants, which have lower construction costs and lower thermal efficiency than CCGTs.
12. The outcome of the comparison between the two options is that the overall reduction in CO2 emissions due to large investments in wind power is likely to be very small. It could be negative – i.e. with higher CO2 emissions in 2020 relative to the baseline option – if old and relatively inefficient coal and/or gas plants are retained as backup to the large increase in wind capacity.
13. The Goodall-Lynas evidence is entirely irrelevant to these arguments. The possibility that increasing the amount of wind power may not reduce CO2 emissions refers to a comparison between alternative system configurations in 2020, when the planned level of wind capacity is 4-5 times higher than in 2012. It has nothing to do with short term changes in wind output for a fixed composition of capacity in one short period of time, whether in 2012 or 2020. Sadly, but inherent in the nature of the problem, there can be no appeal to direct evidence from a single country over time to test the argument. It is possible to draw upon comparisons of investment and operating decisions in countries with different profiles of generating capacity and reliance upon non-dispatchable sources of energy.
14. A final note on civility. After my GWPF report on the economics of wind power, Mark Lynas contacted me by email with a substantial number of requests for elucidation and additional data. I replied promptly and at considerable length. He is entitled to take a different view of the evidence and to reach different conclusions about the impact of further investment in wind power on future emissions of CO2. However, it is neither courteous nor constructive in the broader context to create a straw man that is supposed to represent my position when I have provided detailed analysis and arguments that are clearly different. It is an elementary precept of both journalism and academic enquiry to check whether the views presented are accurate. No attempt has been made to carry out such checks in this case.
15. This is not a matter of offended academic amour-propre. If energy and environmental policies are to be based on solid evidence rather than just who can shout loudest or has the most money for lobbying, then the impacts of alternative policies must be assessed by careful analysis of complex hypotheses and relevant data. Attacking the arguments of those with different views on grounds that are either silly or spurious is rarely effective and undermines the prospect of constructive dialogue. Outsiders may draw the inference that if the proponents of one view have to resort to foolish or misleading arguments, then this may be because they do not have any sound evidence or arguments to support their case.