Discussion Why UK wind power should not exceed 10 GW H. Sharman, Civil Engineering, 158, No. 4, November 2005 Richard Brookes Many engineers have been arguing in a similar vain in my neck of the woods for a long time. Although with less detailed appreciation and assessment of wind energy than your paper, it has seemed quite obvious to the layman that the variability of wind speed can not support our current needs in any more than a marginal way. Here in southeast England where I am acquainted with engineers in the nuclear business and where we have just had approved a large wind farm, all discussions lead me to the inescapable conclusion that nuclear power, of one form or another, is the only way to reduce carbon emissions. For transport, the substitution of hydrogen fuel, supplied from nuclear power plants, also seems feasible especially if the continued rise in oil and gas prices continues, as it surely will. A complete rethink of all the policies currently being expressed by EU governments seems necessary. Peter Hinson The paper would have had more credibility if it had not been a key feature on antiwind protest groups since 2003 (Country Guardians). Without going into the detail the piece has some fundamental errors. Extrapolating generation issues from a small part of Denmark to the UK is rather like saying that, for example, Yorkshire would have problems if it wasn't linked to the national grid. On page 167 the paper incorrectly claims that ‗wind power is already three to four times more expensive than conventional power‘. The recent Sustainable Development Commission Report Wind power in the UK (May 2005) compares the wholesale price of electricity at 3 p/kWh against onshore wind at 3 .2 p and offshore at 5.5 p. This is not a ‗three to four times‘ difference. There is a fundamental mistake on page 168 in assuming that all CCGT plant will work at reduced load and thereby increase carbon dioxide emissions. The reality is that the dozens of generation stations are called into use on merit order to handle the existing 20 GW daily load variation and most will be working flat-out when in use. But the main flaw is that Sharman, in criticising wind power alone, fails realistically to explore how we achieve the essential 60 % cut in our carbon emissions and avoid dependence on diminishing supplies of fossil fuels. Bob Johnson The paper is generally well reasoned and therefore would be worrying in its conclusions, except that it does raise some doubts about the methodology used. It is not clear from the paper why the conclusions should be based in large part on the comparison of three Danish storms and the lowest summer demand curve for England and Wales, especially when two of the storms occurred in winter. Would it not be more logical to compare average summer and winter wind conditions with average seasonal demand curves? Furthermore, even on the comparison given, the 12 .4GW capacity response to the storms is always well below the minimum summer demand curve and it is not clear from the paper why this justifies the statement ‗it is clear that 12 .4GW of wind ... will impose severe strains on the England and Wales system.‘ The argument appears to rest more on the comparison with the installed capacity per capita in Germany than the response to the Danish storms. Ben Tatham In the Local Government Association magazine First, Jonathon Porritt, chairman of the Sustainable Development Commission UK (SDC), urges councillors to disregard ‗outdated myths‘ and to grant planning applications for wind farms. He also says: ‗Wind power does indeed displace some conventional plant, so there is no need for ―additional‖ or ―dedicated‖ conventional capacity to provide backup for when the wind isn't blowing, as is commonly assumed‘. Would the author please comment on this claim by the chairman of a Government-funded body which is presumably meant to supply the public with reliable data on this important subject? John Richardson I would like to commend the author for giving us a fascinating and realistic insight into the pros and cons of wing energy generation. I very much enjoyed reading his paper which, to my mind, brought a strong note of engineering realism to the highly publicised and politically emotive issue of wind power generation. What came across most vividly was the difficulty of controlling the power balance and delivery from this fickle source of energy—and that fact that the power output is so dependent on wind speed. These facts are not emphasised enough in the public debate. He also highlighted the fact that the best wind sources are some way from the major power demand centres and that this also incurs transmission losses—again another factor that is overlooked. Surely the best way forward for GB Ltd is to continue to ensure we have a sensible mix of core generating capacity properly planned and funded well into the future—and under our national control. It is the responsibility of Government jointly working with the private sector to ensure that the lights do not go out. I share the author‘s concerns about the need to invest in core generating capacity sooner rather than later. I understand the French have an 80 % core nuclear generating capacity already. The UK is at 15 % and falling. I know there is a lot of resistance to the nuclear option but nature has bound up such huge amounts of limitless energy into the nuclear forces that surely this carbon-dioxide-free source must be seriously considered for the future. Technology must have moved on significantly from the Calder Hall days such that environmental risks from modern generating plants are very much lower now. Such plants sensibly placed should be part of our future energy supply picture such that the core power generating capability of the country is secure well into the future. I agree that variable output renewable sources should play their part but the author makes the point very clearly that if this proportion gets too high then control of the whole power supply networks will become impossible and inefficient to operate with increased risk to supplies. Politicians please note! Mark Petterson I support many of the general conclusions of the paper, namely that the UK needs as diverse an energy supply as possible wind power can make a significant contribution the UK power sector faces a huge replanting activity in the near future the challenges for our engineering profession are significant. To these I would add that the regulatory climate for facilitating the additional transmission infrastructure needs reviewing to ensure that the new generation capacity can access the market. Decades of spare capacity in our gas and electricity transmission sectors have allowed security of supply to take a back seat to other consumer issues, but the Office of Gas and Electricity Markets (Ofgem) now needs to correct this to avert a crisis and to allow National Grid to make the necessary strategic investments. Having gas and power available on demand is the primary issue for all consumers. I would, however, disagree with two issues raised by Mr Sharman. The capacity factors shown in Table 1 are wrong as the significant new capacity added during each ROC period artificially lowers the annual average in this over-simplified analysis as a full year‘s output is not included from these new sites. Furthermore, National Grid balances the system in real time and does not rely on the day-ahead forecast as presumed by the paper. The days of the ‗pool‘ are long gone. Forecasting accuracies for aggregate wind output are extremely good several hours before each gate closure and the market then works to correct any looming imbalance. These facts are reflected in the more tolerant attitude that National Grid, the system operator, is showing towards the potential market penetration of wind power than readers of the paper might presume. Therefore, although many good points were highlighted in the paper, the central one and the title of the paper is incorrect. Robert Freer It is interesting that the detailed information shown in Figs 2 and 13 is openly available in Denmark but similar information for turbines in this country is apparently not published. It seems to me that developers in this country who accept public money in grants or subsidies should be invited to make public similar information about the performance of their machines. It is helpful to have in Table 1 a list of actual load factors (as distinct from estimated values) based on the issue of Renewables Obligation Certificates (ROCs) as published by Ofgem. David Milborrow Around 1980, a senior director of the Central Electricity Generating Board (CEGB) presented a paper at the World Energy Conference in which he stated that there was no necessity to associate storage plant with wind power up to a wind capacity of at least 20 % of system peak demand. That message was refined in subsequent CEGB studies in the 1980s and again in 1990. I managed the latter study, but the renewables branch did not do the work (we were probably seen as wild-eyed enthusiasts); that was carried out by hard-nosed engineers in the operations department. In parallel, detailed analyses had been carried out at Imperial College and Reading University. Most of this work probably overestimated the magnitude of the wind power fluctuations, since these were simulated from wind measurements. Wind turbines are good averaging devices. The conclusion from every study was the same: assimilation of wind power will incur very small extra costs, and it may need to be curtailed on a very few occasions, once the energy penetration level exceeds about 10 %. National Grid has recently stated, ‗if there is a limit to the amount of wind that can be accommodated, that limit is likely to be determined by economic/market considerations‘. So I was very surprised to see this paper. If every single one of the previous studies has got it wrong, where have they gone wrong? I did not find the answer, nor did I find any reference to the enormous amount of work that has been carried out on the topic over the past 25 years. Author’s reply to Richard Brookes My point was and is that an island system, like UK, should not wind down a diversely energized, firm generating capacity, such as still exists in 2005 and replace this with a mixture of highly intermittent capacity and combined-cycle gas turbines (CCGT). CCGTs are inflexibly dependent on dwindling energy supplies half way around the world, in countries of doubtful political stability. By design, they are also unsuited to operating flexibly in a system also having a stochastically intermittent output. This latter situation can only be addressed by a massive capacity of preferably distributed storage. Storage, other than pumped hydro, is a feasible option for balancing future stochastic output. Compressed air energy storage and various electro-chemical systems are becoming steadily more economic as fossil fuels rise in price. As to what might constitute the ideal generation mix for the UK to aim for in 15 years time, I am neutral. It seems likely to me that both ‗clean‘ coal and nuclear should contribute a major part of the mix. Author’s reply to Peter Hinson I fully support the development an appropriate amount of wind power. The UK's wind capacity is concentrating in two areas, Wales and Scotland, where, unlike the Midlands or most of the home counties where demand is concentrated, excellent wind conditions prevail. The clustering of wind capacity in this way makes commercial logic as the investment in wind is paid off by generating as many MWh per unit of investment as possible. West Denmark is not ‗a small part of Denmark‘. It is most of Denmark by land area. It is significantly sized in relation to both Wales (it is larger than Wales) and those areas of Scotland where most applications for wind connections are clustered. This is why, on the whole, I deem it can be valid to extrapolate West Danish experience when dealing with relatively small parts of the UK where most wind is accumulated. I am glad Mr Hinson mentions the cost of wind. He is quite right to draw attention to the convergence of fossil fuel price with energy systems that are based on paying off a fixed amount of capital, like wind. It follows that providing the wind turbine owner receives an equal amount of reward for each MWh of production, whether this happens at 2 am on a Sunday morning in summer or 9 am on a Wednesday morning in winter, its investment will be paid off more quickly in a windy area than a calm area. This distinguishes the reward for wind from other generating systems that are forced to be more sensitive to market conditions. As a renewable energy consultant, Mr Hinson must be aware that the system for rewarding wind is highly profitable for those developers fortunate enough to receive permission and grid connections to build at good sites in UK. The wind market in Ireland is developing well, with excellent returns for developers, at a price of €57–59 /MWh, or about £40. This should be enough for UK developments in similar areas of high wind output. Instead, at present, thanks to the ludicrous system of ROCs, UK developers are receiving an average of £90 /MWh, twice the rate necessary to obtain a 20% return on the investment. In Faro islands, the wind generator receives £27 /MWh and ‗would like to build more‘ except, at 4 MWh, that is all the small local grid can accept. The undifferentiated ROC subsidy, although not enough to make offshore wind power and more reliable tidal stream and wave power feasible, is grossly overrewarding much onshore wind power, thus more than doubling its cost to consumers, as Mr Hinson must be aware. Equally, given the very low load factor of the national wind carpet (Table 1 in the paper), it seems to have motivated the construction of many wind farms where the wind resource was never enough to justify the landscape destruction caused. Author’s reply to Bob Johnson I thank Mr Johnson for identifying a less than fully satisfactory argument within my overall case. I was hampered to some extent by the rich menu of possibilities that were available to present my case but the strict limit of only 6000 words understandably imposed by the editor. I chose to present three arbitrarily selected storms that can occur at any time of the year, graphically. The main point at issue will be the nature and amount of the base load plants that will be operating in tandem with wind. My paper points out that most of today‘s firm, base-load capacity, in the form of 12 GW nuclear, will be closed by 2020. Any amount of wind cannot replace the need to replace this ‗firm‘ capacity. The replacement can be with CCGTs, ‗clean‘ coal or new nuclear. For the reasons explained, increased dependence on gas would be unwise. That leaves the need for building not less than 12–16 GW of nuclear and/or clean coal, the latter probably in the form of integrated gasification combined cycles (IGCCs). Neither nuclear nor IGCC are amenable to cycling, let alone turning on and off. So, in summer, much of the 12 MW (or so) of wind capacity that will not present huge problems for the rest of the year, must often be curtailed during windy weather. This situation could be altered if storage is built on a very large scale. I cannot pretend that 10 GW of wind power is a scientifically calculated ‗upper‘ limit to how much should be built. However, at the point where large scale curtailment or expensive balancing by fossil plants operating sub-optimally is reached, that is the likely, rational, upper limit for this type of capacity. Author’s reply to Ben Tatham It is useful to read the SDC report, to which Mr Porritt refers, which is downloadable from www.sd-commission.org.uk. First, apart from Mr Porritt, no author is mentioned. I wonder what qualifies Mr Porritt to come to his conclusions? It would have been helpful to know who was responsible for what reads to some of us in the engineering profession as propaganda masquerading as science. The references (page 143) exclude the two weighty wind reports by EON Netz, 2004 and 2005, which, unlike the SDC‘s report are fully attributed and recount real experience. It astonishes me that the UK appears to remain in wilful denial of what is happening elsewhere. In Germany just 4 .5 % wind penetration by MWh, or 14% by capacity (MW) is causing enormous strain on the grid system, despite the fact that Germany is inter-connected with all its neighbours. It is therefore valid to ask how the UK can painlessly achieve even 10% penetration by MWh given its relative lack of inter-connection and its very limited storage, (in the form of pumped hydro in North Wales). Some astonishing statements are made in the report. For example, in the foreword, Sir Tom Blundell states: ‗Another frequent misunderstanding related to wind is the implication of its variability. In fact, with modern meteorology, wind is very predictable over the time scales relevant for balancing the electricity system‘. It is true that wind forecasting is much better than it was and is improving. Enormous sums of money and huge talent is devoted to wind forecasting by the Danes and Germans. But both make it clear that small discrepancies in forecasted as against actual wind output make it a real challenge to keep their systems in balance and that the art is far from developed. As mentioned in my paper (see Fig. 6), +/- 1 m/s wind speed makes an enormous difference to how much wind power is being generated. And wind forecasting‘s ‗Holy Grail‘ is an RMS of 1 m/s. Author’s reply to John Richardson I general I agree, although the UK‘s engineering professionals have badly mishandled nuclear power in the past. These remain in denial of the cost overruns and the still unresolved disposal issues of nuclear waste. ‗Nuclear Power UK Ltd‘ must do convincingly better in the future if nuclear power is to be a part of the future energy mix. In the mean time, it is has already fallen far behind its counterparts in the USA and Europe and is even behind China and India. It lacks two generations of engineers having an education and career in exploiting this vital resource and there are almost no teaching and training resources. May I mischievously suggest this task be out-sourced to the French who have always been more pragmatic and apparently more successful about energy policy and developments than the UK? I have little more to add except to note that the unfortunate ROCs process is over rewarding onshore wind power to the detriment of developing more reliable, long-term, equally renewable energy resources such as wave power, tidal power, tidal stream and less-landscape-damaging offshore wind power. Electricity storage remains the last unconquered challenge that might make some sense of intermittent energy sources. Even here, the UK Department for Trade and Industry (DTI) allowed the development of the promising Regenesys electricity storage process to be terminated and this is now in the competent hands of a start-up Canadian company. Author’s reply to Mark Petterson In calculating the load factors in Table 1, I used the average of the national capacity at the beginning and the end of each Ofgem period. This takes into account the growth of the capacity and gives a reasonable if not perfect approximation, for the load factor achieved during the period. Of course, the UK wind carpet at 1 GW produces a tiny fraction of all generation, and this is easy to balance in real time. Our main point of contention arises from the question of how easy this is to balance when the wind penetration becomes more significant. At what point does the growth of the wind carpet start to deliver more problems than solutions? Here we need to draw on the experience of systems where the wind penetration is much deeper. When the west Danish system is producing more than 500 MW of wind power, this almost always coincides with net electricity outflow. This is because it is the least costly way for the transmission system operator to keep the grid in balance. Thus 500 MW represents the roughly ‗optimum‘ level of wind capacity in the west Danish system, where all the Danish-consumer-subsidised wind output would be consumed by the Danish rate payer. 500 MW corresponds to 15 % by capacity of the Danish peak demand and delivers roughly 4 % of Danish demand. It is clear from the EON Netz wind reports of 2004 and 2005 that optimum levels of wind power in the German system have been exceeded. EON Netz, RWE and Vattenfall are struggling to keep the German system in balance with a wind capacity of 17 000 MW, 24 % of peak demand but delivering just under 5 % of German electricity demand. The Danes and Germans have the same access to wind forecasting and hightech system balancing as enjoyed by the UK. Engineers in both countries admit that, lacking systems for electricity storage, from their national perspective, it would have been sensible to limit the growth of wind capacity to levels that would not require high levels of balancing. Unless it develops a large scale capacity of distributed storage, I cannot agree that the UK should develop so much capacity that it needs to employ wind power curtailment or an even heavier use of inefficient fossil units to balance the wind output of a wind carpet much larger than 10–12 GW. Author’s reply to Robert Freer I completely agree that publicly subsidised financial transactions of any sort, in any democratic country, ought to be completely transparent. In researching this paper, I found that the availability of technical and trading data about this sector is generally impossible to obtain in fine enough detail to deliver truly satisfactory conclusions. In this respect, the consumer-subsidy-driven expansion of wind energy in the UK seems to be driven forward in a way that is very often contrary to normal democratic principles and good business practice. Author’s reply to David Milborrow I agree with Mr Milborrow that much theoretical work has been done by fine, clever and dedicated engineers in the UK to predict the amount of wind capacity that be built without upsetting the security of the grid. I have read some but not all of it. I may never have got involved in this controversy had I not read Quantifying the system costs of additional renewables in 2020, ILEX Consulting, October 2002, commissioned by the DTI. This document, widely referred to as the Scar report, laid the foundations of the Government‘s policy toward wind energy. Presumably the authors of this were more familiar than I with the whole body of UK work to 2002. As part of its terms of reference, the authors were instructed to study the experiences of Denmark, where I live. Naturally and mostly out of curiosity at the time, I read this section. I was sufficiently shocked by the superficial treatment and the simple but fundamental errors of fact written up in this section to become more interested in the methodology and conclusions of the whole report. Incidentally, even by 2002, much experience had also been gained in Germany that the report ignored completely. In analysing the report further, I identified an obvious calculation error in annex B, where the report‘s assumptions about the load factor of onshore and offshore wind had been reversed. Offshore wind in this section is assigned a load actor of 30 % and onshore wind a load factor of 35 %. Thus, if these figures were carried through to the conclusions, the results of the whole study are quite simply wrong. In fact, according to the UK DTI statistics, during the ten years from 1994– 2004, the UK‘s national, onshore wind carpet has just once, in 1998, achieved 30 %. The table, giving load factor (in %) below is downloadable from www.dti.gov.uk/energy/inform/index.shtml Year 1997 1998 1999 2000 2001 2002 2003 2004 Onshore wind 27.2 30.7 28.2 28.2 26.4 29.9 24.1 26.6 My analysis of the much more transparent data downloaded from Ofgem brings even the DTI‘s data into question (see Table 1). In fact, I regret to state that much of the data used in the national, DTI statistics are based on guesses and assumptions; how much, is not stated. Undoubtedly, there are high-performing wind farms, probably in Scotland and Wales. But their high performance (and enormous profitability, given ROCs) must obviously be offset by the majority of wind farms that drag the average, national statistic so low. It is deeply regrettable that so little faith can be placed in the authority of the Scar report, almost all of it highly theoretical. My paper is at least based on the real experience of just two countries that between them, at this moment, contain over 20 GW of wind capacity. This is over 20 times the installed wind carpet of the UK. Considering their close proximity, considerable experience and many similarities as well as differences, I find it extraordinary that the UK authorities simply refuse to accept that they have anything of substance to learn from Germany and Denmark. In closing this reply, it is worth considering the very different and much more realistic approach of the Irish, whose national wind resource is even better, on average, than the UK‘s. Ireland has at least engaged in an open, national debate and has placed the stability of the grid and its reinforcement at the heart of the discussion.