Unrevised

Members present

Earl of Selborne (Chairman)

Lord Broers (co-opted)

Lord Dixon-Smith

Baroness Hilton of Eggardon

Baroness Manningham-Buller

Lord O’Neill of Clackmannan

Lord Patel

Lord Peston

Viscount Ridley

Lord Rees of Ludlow

Lord Willis of Knaresborough

Lord Winston

________________

Examination of Witnesses

Professor Keith Bell, Scottish Power Professor of Smart Grids, University of Strathclyde, representing the UK Energy Research Centre (UKERC), Professor David Newbery, Director of the Energy Policy Research Group (EPRG), Cambridge University, Research Fellow at Imperial College London, and a Member of the Panel of Technical Experts for DECC on National Grid’s Electricity Capacity Report, and Professor Michael Grubb, Professor of International Energy and Climate Policy, University College London

Q69 The Chairman: Welcome to our three witnesses. Thank you for joining us. If you would like, first of all, to introduce yourself for the record. We are being broadcast and we are on web camera, so it would be helpful for the record to have your names. If any of you would like to make an introductory statement do feel free to do so.

Professor Bell: I am trying to work out which end we start from. I am Professor Keith Bell. I am from the University of Strathclyde. My title there is the Scottish Power Professor of Smart Grids. Although my chair is sponsored by Scottish Power I have the privilege of working with many different companies in the electricity sector. Also I have done work with the Scottish Government, the Government of the Republic of Ireland, with companies across Europe. I describe myself as an independent researcher and I am here today representing the UK Energy Research Centre, which is, we could say, the Research Councils’ flagship energy research initiative. Its particular attraction, as far as I am concerned, is its multi-disciplinary nature and the way it succeeds in bringing together social scientists, economists and engineers.

Professor Newbery: I am Professor David Newbery. I am the Research Director of the Cambridge Energy Policy Research Group. I am also a part-time Research Fellow at Imperial College. I am a Member of the Panel of Technical Experts on National Grid’s Electricity Capacity Market Report, and I also sit on the Ofgem Low Carbon Networks Fund and Network Innovation Competition panels.

The Chairman: Professor Grubb?

Professor Grubb: Michael Grubb. I am Professor of International Energy and Climate Change Policy at UCL. I am also senior adviser to Ofgem but obviously here speaking in my academic capacity and cannot speak on behalf of Ofgem. I have an opening comment if you wish.

The Chairman: Yes, please.

Professor Grubb: It is perhaps a broad and general observation that if I look at the press coverage over recent weeks of National Grid, it does feel slightly like a dilemma for this Committee. It feels a bit like holding a trial when the press has already decided the defendant is guilty. One of the things that may complicate the task of this Committee in sorting out the evidence and realities is the terminological confusion in certain areas and the baggage, which I think reinforces the idea that we have an electricity system like a cliff that somehow has enough generation to meet demand or not, and then the entire thing falls over a cliff. In that sense, the terminologies of supply margin are not helpful and the terminology of loss of load expectation is fundamentally misleading because it has nothing to do with the expectation that anybody actually is involuntarily disconnected.

Neither of those helps us understand the reality, which is trying to deal with a system in which we are evolving many more options and much more flexibility. The real challenge around resilience is understanding that and how robust are those systems.

Q70 Lord Broers: Your opening in respect of this first question: we have heard that the capacity margin will be squeezed over the next two winters. How is this expected to affect the resilience of the electricity system? Will the system be able to cope if there are any more unexpected events, like the recent ones? Are the New Balancing Services that are being implemented by National Grid likely to be required to balance supply and demand on the short term? Have these been tested and are they reliable? Finally, what role do you think industrial backup generation could play in balancing the system if required?

Professor Newbery: I am happy to start off, and I am sure Professor Bell will have other things to add. The first point to make is, of course, unexpected events could cause blackouts, because if you had a system that was so resilient that it never under any circumstances had a blackout you would clearly be spending far too much money. The question really is: is the risk acceptable? I would suggest that it is in the short run. We probably face a tighter situation than we will in 2018/19 when we secure capacity. Perhaps later we can come back to our view on the Panel of Technical Experts that we have arguably prematurely secured too much capacity for that later period, but in the short run the New Balancing Services that have been procured will be tested.

I am reassured by the statements that have been made to this Committee from generators that more plant will be returned to service. As Professor Grubb has already said, the idea that when the system is tight you instantly lose load is mistaken and there are a whole sequence of actions that can be taken. The one that seems to have been seriously neglected is the ability to import over the interconnectors. The panel strongly criticised National Grid for assuming that they would make no net contribution, which was odd given that both DECC and Ofgem had commissioned reports sometime before, trying to estimate the contribution that the interconnectors would make to security of supply, and suggested that they could make at least half of their nominal capacity available. It is also particularly odd—now that DECC has consulted on the ability to secure capacity over the interconnectors—that no space was made for that in the upcoming auction.

We think that understandably perhaps politicians, and particularly Ministers, are so nervous about the concept of the lights going out—and in particular the Daily Mail-type views that that might happen—that they are overcautious, and that has high costs that we can elaborate on.

The Chairman: Did Professor Bell want to add anything at that stage?

Professor Bell: Yes, I would add that I have some sympathy with the notion that we might be over-procuring through the capacity market and I would agree with what Professor Grubb said at the beginning, that we should be clear about what the metrics are and on what basis those metrics are being developed. As Professor Newbery has said, the metric that has been proposed and is implemented in the Capacity Mechanism makes a certain assumption about the availability of power over the interconnector.

You can make a different assumption and then you could adjust the metrics, so that whatever you assume you should have your choice of the standard that understands what the assumptions are that are being made. It certainly has been the case in the past that you can procure system operator to system operator services—as they are sometimes known—over the interconnector to support the meeting of demand. We could get into the technicalities of the way the assessment is done but I would agree, certainly, that we need clarity on what the reference standard means.

In terms of whether we face some risk of loss of supply, as has been said already, there are circumstances under which the lights can go out across a very large area very quickly. Fortunately, they are very rare. We have seen them in other countries. So far we have not seen them to a large extent in Britain, but the risks are there. That is just one example of the fact that the causes of loss of supply are many and various. I think in a previous session there was some discussion about individual Members of the Committee who have had their electricity disconnected. Most people’s experience of disconnection is to do with an event that originates on the distribution network and is nothing to do with the availability of generation on the system as a whole.

When you get to the layer of simple availability of generation to meet demand you are quite down the tail of the distribution of individual disconnection events that people experience. Again, as has already been said—and this comes back to the point about the metric—it is important to be clear that there are three hours per winter that represent, as I understand it, not the physical disconnection or the probability or some kind of indication of the probability of the disconnection of demand but rather the need for the system operator to start taking actions. The first one is this maxgen thing that was talked about in the earlier session; the next one would be, yes, emergency actions on interconnectors; and a further one would be voltage reduction. It would be rather useful to see analysis that articulated precisely—or as far as you can, given the uncertainties in the available data—the disconnection chance. Given all that, I would share Professor Newbery’s views that we are not facing a crisis.

Q71 Lord Broers: Professor Dieter Helm told us that the costs would be lower if we had a larger margin. As an economist, Professor Newbery, do you agree with that?

Professor Newbery: Yes. This is one of the points we tried to stress in our Panel of Technical Experts’ reports, that if you over-procure capacity then the nominal cost is about £2.5 billion but the net cost to consumers is less than that because the prices will be lower. A larger capacity margin means lower prices in the wholesale market. But that has consequential effects. Two of them are that the cost of supporting renewables goes up because the difference between the wholesale price and the strike price increases.

If that goes up then the Levy Control Framework restricts the amount of renewables you can put on the system, so there are adverse consequences for one of the main targets of the electricity market reform. If we lower the price in this country relative to other countries the economics of building interconnectors is undermined somewhat and, since renewable generation is imperfectly correlated the wider the area over which you trade, that disadvantages the penetration of renewable generation. It is true that the prices may come down but it would be unwise to ignore the adverse consequences of that.

Viscount Ridley: To follow up on Lord Broers’ point, I think we probably accept that if it looks dangerous that the lights are going to go—it is a very cold winter and there is a lot of capacity offline and so on—there are all sorts of things that can be done to help it out, but is it not a price spike at that point that is the thing that consumers will most feel? That will be just as big a political crisis for politicians if there is a huge spike in energy prices because of a crisis of that kind.

Professor Newbery: Let us be quite clear, domestic consumers do not face half-hourly wholesale price spikes. They get, effectively, smoothed average prices.

Viscount Ridley: They get fed on eventually, do they not?

Professor Newbery: Then the question is: what is the average price over the course of the year and in how many half hours does the price reach that level? Let me give you an example. In 2012 a cold winter in France had the wholesale price hitting the price cap of €3,000 a megawatt hour for many hours. But the domestic consumers in France, by and large, enjoy rather low prices because the rest of the time the prices are pretty low. In the industrial and commercial market you have the choice, you can choose to buy spot—most people contract—and the contract will hedge you against that sharp spike. But the sharp spikes are extremely important for motivating short-term responses to improve the resilience of the system. One of the adverse things you could do is to suppress those short-term price signals.

Q72 Lord Peston: Yes, can we go on to an explanation of the capacity market? I freely confess the harder I try the less I understand how, first, it actually works, and secondly, how it is meant to work. But am I right that this is a payment system that is meant to get the optimum amount of capacity available and that is its ultimate objective? Am I right on that?

Professor Newbery: Yes. You start with the security standard, which, if you properly interpret it, would lead you—given the kinds of generation and interconnection on the system and the responsiveness of the demand side—to determine how much you need to meet that security standard and then you—

Lord Peston: Yes. Just to interrupt you at that point, that standard would be a standard of the probability of running out, is that right?

Professor Newbery: Taking a very large run of years—

Lord Peston: Yes, but it is a probable—

Professor Newbery: Yes, and, again—as has been repeatedly stressed—you have to be quite clear what “running out” means. It is not this knife edge; it is the manner to which you move into a number of other—

Lord Peston: Dangerous situations essentially or difficult situations.

Professor Newbery: When you have run out of all of the ability of the system operator to manage the system, including controlled disconnections of people who are willing for a lower normal price to bear the risk that they will be occasionally disconnected but that is going very far down the list of actions. But perhaps Professor Grubb will know more about these actions that can be taken.

Professor Grubb: Perhaps I could pick up a couple of those points but the first is to stress that the measure the Government uses in determining security standard to then inform the Capacity Mechanism is a value of loss load, which is about 100 times the typical price of electricity. In that sense, it is a pretty high measure and the question that I think all of us are interested in is about the whole range of things that can kick in between the normal long or average wholesale price and something that is more than 100 times as high, and there are a lot of things that can come in.

To stress in relation to the first and last of Lord Broers opening questions, maybe I will give a brief illustration around “Will the system be able to cope and how worrying is the supply margin?” In what other sectors do we have this language?” For example, what is the supply margin in food? I had a look and last year the UK produced 60% of its food consumption. Obviously we can produce more but let us say we have a supply margin of minus 30% in food. None of us starved last year. Of course, that is not a perfect analogy. We are more interconnected—more trade routes, more storage capacity—but, nevertheless, we can say, “Be careful about this idea there is a supply margin and once it is too small all hell breaks loose”. No, there are lots of other things that then happen, like prices rise a bit and attract more imports and so forth.

In terms of the economics of this in relation to your specific question regarding Dieter’s remarks that we should rely much more on the capacity market, I was a little surprised if that is a recommendation that we rely far more on Government procurement than on market-based signals of any potential scarcity, given that—as David said—the impacts on consumers are anyway smoothed out at the retail end. It is about the relative economics of those occasional price spikes driving the responses of generators and others on the system, vis-à-vis the economics of a wholesale subsidy or a wholesale payment to enduring capacity.

To finalise that point in a sense, the last question in this opening list was: what role do you think industrial backup capacity could play?” I suspect the answer is: very considerable. I did have a look. There appear to be no firm published statistics on this, but the evidence I have is the amount of industrial backup capacity is somewhere probably between 10% and 40% of total peak demand on the system. That is not generally included in this indicator of the supply margin. In other words, the amount of existing capacity of plant, which is either industrial backup or in some cases retired—mothballed, potentially awaiting use—is much greater than any discussion about the supply margin. The combination of Capacity Mechanism and New Balancing Services should—but may not sufficiently in current designs—bring some of that reservoir of resilience into the electricity system, which is not adequately done at present.

Q73 Lord Peston: To follow the lead you are giving us, you said we should compare it with other examples. The area that I used to make a decent living from was doing operational research involving inventory control. Following David Newbery’s point again, it was incredibly difficult to persuade firms that the marginal cost of never running out of spare parts was so high that there would be very much more profit to be made if they would run out. That is why, David, I used the expression “running out” rather than the relevant one here. What I am trying to get clarified from you is that the optimum position, and the optimum outcome in the area we are discussing, is not to have available supply at some appropriate rate. Is that the correct view: that the lights should go out occasionally?

Professor Grubb: No, not personally speaking.

Lord Peston: Is that right or wrong? I mean on average, if it is an average proposition.

Professor Newbery: The lights do go out—at least in my part of the world—quite often because the distribution network is overstressed or because the wind has been blowing or a tree has fallen over the local loop. These are not unexpected events. The question is: how much security and confidence do we want at the system level?

Lord Peston: Yes, that is what I am trying to press you on.

Professor Newbery: Obviously, we want to have enough and that is what exactly a security standard aims to set. One can discuss whether or not we have set the right one. All I can observe is the Loss of Load Expectation is similar to most other European countries but it is lower than quite a few. Whether we have interpreted that Loss of Load Expectation correctly is very doubtful and I think we have erred on the side of caution. We have a very secure standard that we have overegged. So in terms of security of supply, at the level of generation adequacy and access to power, I think we are doing well. Whether we have the right resilience elsewhere in the system depends on where you are.

Lord Rees of Ludlow: To follow up that, what are the figures on our relative resilience and number of breakdowns, compared to, say, France or Germany?

Professor Newbery: We have the same loss of load expectation standard as France and Germany. Belgium has two and a half times as high. As far as I can tell, the only event that we have had since 1990 that was exciting was the loss of Long Gannet and—30 seconds later—the loss of Sizewell, which was a massive and very unexpected uncorrelated event happening within a very short period of time and that was managed without much happening.

The Chairman: To be fair, if we are looking at generation failure or outage rather than distribution, probably it was a three-day week, was it not—which is now 40 years ago—that we lost?

Professor Newbery: Exactly.

Q74 Baroness Hilton of Eggardon: Basically my question has just been asked. On the international comparisons with our generating capacity, do you think we have the same level of security as elsewhere and how would you offset costs against the benefits—we have already been talking about this—of having a secure system?

Professor Grubb: As you say, it has been partly answered. I will just point to a paradox that if, as may be the case, the measured margins are tighter than some of our continental countries that implies there is a high chance that interconnectors will flow in our direction when needed. It is not a simple thing to say, “Oh, look, we are worse off than those”. I am inclined to agree with David; I think that we have standards that err on the conservative side—perhaps quite rightly—and have not taken enough account of the range of options.

But perhaps underlying this is my reaction to the question phrased as “the lights going out”; there are some critical uses of electricity for which the standards should be extremely high. It should be pretty much always the case that we try to make sure enough generation is available to meet those, compared with the risk of transmission and distribution failures, which do dominate forced disconnections. That is a reasonable benchmark to say the generation system should not be the dominant cause of blackouts, compared with distribution. It is not a numerically optimal view; it is a very cautious security-orientated view. But, again, I come to this problem where you get the phrase about lights going out as if all electricity use is of the same value and essential nature.

I may have one example in relation to the industry side of this. I was rather struck a week or two ago to see a CBI spokesman say something along the lines of, “This is terrible. Industry is being offered contracts to sometimes disconnect from the grid or reduce its demand, and this is a terrible thing for British business”. It seemed to me very strange to hear the CBI propositioning that giving a company a new option to make more money than it otherwise would do was going to damage its competitiveness. It seems to me it is a perfectly rational thing to do because over-engineering the system in unnecessary ways ultimately requires all of consumers, including industry, to pay higher prices.

Q75 Lord Rees of Ludlow: I would like to ask about the consequences long-term about a growing dependence on decarbonisation types of energy generation. This is more intermittent, of course, and may need new technologies and also will make a more complicated grid. I wonder if you would like to comment generally on some of these issues.

Professor Bell: Yes, I think it does make operation of the power system more complicated. I tend to start from the perspective that decarbonisation is not just an option, it is an essential thing that we need to do. For me, the challenge is: how do we facilitate it in such a way that we achieve whatever we regard as being a sufficient level of reliability of supply in the cheapest way? There are all sorts of options that are out there. In respect of many of them, exactly how you do it is not yet determined. It is too early to say. There is still more research that is needed. Lots of people have jumped on the smart grid bandwagon and say that smart grids are the answer. I am not sure they always know what the question is. As we go on we will articulate that question much more precisely.

In the previous session, Mike Calviou from National Grid articulated some of the issues to do with the variability of renewables—the uncertainty of renewables, although the forecasting of a few hours ahead is not too bad. To get the net balance correct you need to fill in the gap between renewables and net demand. You have to have some kind of ability to schedule and ramp up and ramp down the other generation resources and the ability to withstand faults or failures. Professor Newbery alluded to an event back in 2008 where two large generating units were tripped within a short space of time.

The system is operated now in such a way that you can keep the system frequency within statutory limits for the loss of one very large source of power, the reference for which happens to be the reactor at Sizewell. The amount of reserve and response that you need to carry to do that depends on the inertia of the system. One of the sources of the replacement energy for that lost energy is, of course, the store of kinetic energy. If you have a lot of wind farms or any generation or source of power sitting behind a power electronic interface you do not get the same natural coupling, electromagnetically, with that source of energy. One of the ideas is that, when we talk about reduced inertia of the system, that is what we mean. There is still rotating mass. There is research going on on how you can articulate a controlled response to synthesise the rapid utilisation of any store of energy even if it is not electromagnetically coupled. These things are not fully sorted out, to my understanding, but they offer some promising avenues and the challenge for us engineers is to work out which ones are going to work best and most cost effectively.

Professor Newbery: Perhaps I can add something, I sit on the Single Electricity Market Committee of the island of Ireland, which is planning by 2020 to have 75% non-synchronous generation connected and it does not have very strong links to GB. They are actively investigating what services they need now to define in terms of fast frequency response and various other new services and how best to procure them. So we can admire our cousins to the west and see how they deal with this system and learn from it. Of course, we have the advantage of a much larger inertial mass—to use that phrase—and lower penetration, so we face the problem later than some other people do.

Professor Grubb: Your question, Martin, is a very important one, it is a long-term evolution one. In effect, the more one relies on renewable resources, certainly wind and solar, what we are doing is emphasising the move to a system in which the volume of low-marginal cost generation available will fluctuate much more than previously and the cost will be more geographically varied than used to be the case.

Those together imply a greater value, you would expect, from interconnection under stronger transmission and a broader transmission system and associated governance because of the geographical variation. The temporal variation obviously increases the value of storage or access to other systems or technologies and end-uses that can help to provide storage to the system, of which there are quite a range of varying characteristics and costs. But all of those things become potentially more important the further you try to decarbonise the system.

Lord Rees of Ludlow: I have two more questions. One is: what about the grand role of small-scale off-grid generation?

Professor Grubb: As I alluded to earlier, we still do not have quite enough information even around the present situation, let alone where it may go. But it is clear that there is a growing capacity of what is generally called embedded generation, broadly below the level of direct connection to the transmission system, and including industrial backup of various sorts that is usually considered a subset of potential embedded generation capacity.[1] Ofgem itself issued a licence modification in the summer to try to improve and accelerate the reporting of that kind of data, which has been rather below the radar screen. The implication, again, is a more intelligently managed and intelligently informed system to get the best balance of all those inputs.

Lord Rees of Ludlow: My last question is about the fact that this distribution is getting more complicated, more vulnerable to cyber attacks and all that. Do you have any comments on whether that should lead to a greater need for a margin?

Professor Newbery: Perhaps I can comment and pick up the point that was made in the previous session that Ofgem, through the Low Carbon Networks Fund and the Network Innovation Competitions, is actively looking at how low-carbon generation can be better connected to the distribution networks, especially embedded generation. A lot of that has to be active network management, increased visibility of what is happening on the distribution network, better co-operation and co-ordination between the transmission and the distribution networks. The research results from the first four years of the Low Carbon Networks Fund have been extraordinarily promising. From being probably one of the laggards in Europe in the smartening of the distribution network, we are probably now very much at the cutting edge. At a recent conference some 800 people were present listening to the progress that has been made and what we now know that we did not know four years ago.

The Chairman: Lord Ridley, do you want to follow up that point?

Q76 Viscount Ridley: Yes. I want to press you a little bit further on the intermittency issue. The system is designed to cope with variability of demand. If you add variability of supply to that as well, of an unpredictable nature, it must increase the degree to which you have to call on supplemental balancing, or whatever it might be, which is inevitably going to be in price spikes, which inevitably is going to make the things more expensive. For example, if the wind drops just at the moment when everybody arrives home and turns on the television, that puts more of a strain on the system than if you were not relying on wind power at all.

Professor Newbery: Let me challenge one of your comments that the price spikes make it more expensive. The complaint on the continent of the generating companies is the massive solar and wind installations in Germany have so crashed the prices that they are retiring almost brand new Combined-Cycle Gas Turbines, so that does not follow.

Viscount Ridley: Yes, but that was hugely subsidised, so it did cost the customer somewhere else.

Professor Newbery: It did, big time. But to come back to the question, it depends on what kind of intermittency you have. Wind: as you get closer and closer to real time, the reliability of that improves, compared to a nuclear power station, which could fall over on a blink at any moment—very improbably. But your confidence in the wind supply, as you get to within half an hour and five minutes of dispatch, is very high. Solar, on the other hand, is very tricky because the sun can go behind a cloud. If you have what I would think an unreasonable amount of PV in a rather cloudy part of the world connected to your system, you can stress it. We have discovered that there are rather important design features that you need to incorporate into that. One of the things I think that the Germans are discovering is that having all PV panels disconnect to the certain frequency, the same frequency, is a bad idea. We are learning quite a lot about how to improve the resilience of including these new technologies.

Professor Bell: If I might just concur there, the nature of the risk is different. We have to distinguish between variability and uncertainty, so something can be highly variable but we can have a good forecast of it some time ahead. The probability of losing a certain amount of power within a certain period of time is different, depending on what the resources are. The wind turbines are distributed across the whole country. It does happen that when a storm front comes through you can go from 100% output to near enough to 0% output but it takes a few hours for that to happen. The critical thing then is identifying when it starts, so you can start ramping up the replacement generation. Yes, the 50.2 Hz problem, as it is known in Germany, when the PV panels were being specified and some standards articulated they would look across the system in Europe and say, “How high does the system frequency go? It almost never gets above 50.05, never gets above 50.1. Let us put a safety margin in, 50.2. That will do”. Then in the last few years they found it is creeping up to 50.1 and a bit above and, “We had better go back and retrofit new software in the panels to make sure it does not all disconnect at 50.2”. It is slightly ironic that if you went from a surplus of generation and you did disconnect them all—the lights, a lot of the lights, talking about the lights going out—there would be a serious disturbance in Germany. I heard that it was maybe costing €1 billion to fix that but it is one of these things you learn. As you get experience with this it does not mean that PV in the right places in the right parts of the world is a bad idea; you just have to manage the engineering in the right way.

Viscount Ridley: No, but it does mean that it is added to a resilience problem.

Professor Bell: If you managed it correctly, there is always a resilience problem. As we were talking about it earlier, the system is not 100% reliable and if you were making it 100% reliable you are probably overvaluing your electricity supply. Different people at different times will value it in a different way. It is about making the right judgments about what you do at the right time. On occasions I have given talks on how come the lights go out or, as engineers, we can get very interested in big disturbances. It is very interesting to see the complexities and how they work out. We can learn a lot from them. Another time I changed the title and I said, “How come the lights do not go out more often?” It is a great engineering achievement that we have managed to have the reliable supply that we do at the price that we have electricity.

Professor Newbery: Just to give a little more comfort, the laws of physics work everywhere in the world. One of the things we can do is look to other countries and learn from what problems they have run into before we have to face them. Germany is a very good example.

Q77 Lord Broers: Professor Grubb, you have said a couple of times that your estimate of the potential industrial backup is between 10% and 40%. That is a huge range. Should initiatives be put in place to better determine what that is?

Professor Grubb: Yes.

Lord Broers: Should this be one of our recommendations?

Professor Grubb: First, they are not my estimates. Simply, the only two points I was able to get anybody to offer any quantified indication were, respectively, 5 gigawatts and 20 gigawatts, but they are big numbers. As I mentioned, Ofgem did issue a licence modification this summer to try to improve the data on that. But, yes, recommending—there appears to be an important resource here that we should understand better—absolutely would make sense.

Maybe just to build on the theme: first, the discussion we just had emphasises that innovation is not just about widgets. It is about understanding the system and how best to utilise all the different kinds of widgets and associated options that may be available, contingent often in part on policy and regulatory design. Some of it may involve creating new kinds of markets to tap into some of these. In some countries you have, more or less, developed markets around reserve requirements and balancing and other stuff. But the essential point is, it is changing the notion of: what do we mean when we talk about the system? It used to be desperately simple. In the ‘good old CEGB’ days, we knew the power system was what the CEGB ran and electricity demand was what we demanded of it. That is not the world that we are in. As an example, in terms of the variability versus uncertainty, I agree. Uncertainty, for a number of the intermittent sources, is not the problem; variability is, depending on the time horizon. One can address that better through a whole mix of issues, some of which we have touched on.

Broadly, anything that either uses fuel, which can be gas or biomass or various other things, or supplies heat to something reasonably well insulated, has capacity to help manage variability on various sorts of timescales. For example, if we do end up in a world of substantial electric vehicles, along with pricing as time-varying signals and other means to help determine when people plug in and when they charge up, that could add significant storage. And if the battery has reached the end of its useful life in the vehicle, instead of throwing it away, should they put it in their garage and connect it to the system and should they then receive a capacity payment and so forth? The world is remarkably wide in terms of options when one starts thinking through these possibilities. As I say, technological innovation is often the relatively minor part of what is required. It is also expanding our own understanding and management of the systems.

Professor Bell: If I could just add something briefly on industrial generation or on-site plant: that is likely to be one of the things that is most prominent in these New Balancing Services that National Grid are procuring. That is where the offers are coming from. If National Grid puts out an invitation to tender for these services that is one of the ways in which you reveal how much of this industrial on-site generation there is. Okay, you want to be sure you can rely on it and, as Mike Calviou said earlier, they will be conducting some tests.

Q78 Viscount Ridley: I want to come to the question of the impact of climate change on resilience and reliability of the electricity system. It is very clear that weather is a big source of non-resilience or potential non-resilience in the electricity system because it blows trees down and knocks out a distribution network or something. The question is: how much percentage does potential climate change add to that weather risk and in some of the written evidence we have seen, I think, you, Professor Bell, talk about there being wetter summers and colder winters, but is it just as plausible we are going to see drier summers and milder winters as a result of climate change? I have seen projections that say that, and where you talk about more snowfall, more ice storms and so on but ice storms are a very rare event. What percentage probability are you adding to the weather problem from climate change?

Professor Bell: I was very careful in my written evidence not to make any claims about the climate science, since I am not a climate scientist, so I would not dare to say whether it is more likely to be drier summers or wetter summers. What we can do—and I can try to help with, as an engineer, and my colleagues I work with—is to understand if the climate means these kinds of weather effects, what that would mean for the power system. We can at least be forewarned at such time as the climate scientists are able to give us some better projections on the weather impacts—and the IPCC’s Syntheses Report that came out on the weekend was starting to make some stronger statements about likely weather impacts and beyond just changes of average temperature. Then we can have an idea of what that might mean.

You are absolutely right that weather is one of the major influences on loss of supply now.

Coming back to something we talked about earlier, there are different mechanisms by which one’s supply might be lost and I tend to think there is a bit of a psychological thing that goes on here as well to do with the impact—how we perceive the impact. I am looking forward to working with some of my social science colleagues in UKERC on this who can help me understand it a bit better.

But the experience of having the lights in your street go out for an hour or however long you think it was—when you happened to be out of the house at the time and you come back in you need to reset all the clocks—is an annoyance but it is not a big thing. You could compare that with if you are depending on electricity for your heating and you lose that heating for 12 to 24 hours. That is serious. If the whole country is blacked out, which does happen—it has not happened in Britain but it happens in places—it is a very high-impact/very low-probability event and we maybe respond to that in different ways. I expect your Committee would be desperate to know what on earth was going on if we did black out the whole of Britain. It can happen, unfortunately.

The mechanisms—we can look around the rest of the world, as Professor Newbery was saying and try to learn from these—are complex but there are often some common features about individual items of control plant that did not behave in the way they were supposed to. Maybe there was some natural event that triggered it; it might have been a lightning strike or some high winds and things cascaded in a way that was not intended but was just a phenomenon of a very large complex system.

If we got changed weather patterns then the individual kind of single events that the system is designed to withstand would happen more often, but you would not notice an impact because the system is designed to withstand a single lightning strike or a single double-circuit overhead line coming out of service because of the high winds and there is a short circuit. That is okay. But if the weather events become so severe that you see multiples of these events occurring simultaneously, then you start to impact on reliability of supply and then you run into some of these complex situations of cascades and the system running away—a widespread disconnection, not just a local disconnection. We should not underestimate the impacts of local disconnection. The previous session talked about the ability of the distribution networks, in particular to restore supply.

Clearly, there were lessons to be learned from what happened last winter. If I understand it the regulator has been quite firm with some of DNOs in particular that they judged underperformed, and one looks forward to improved performance. One of the things that will improve that is the information that is available to the DNOs. I know that software providers are working on devices where individual consumers can take a picture on their mobile phone and from there the DNOs can understand where the picture was taken and can see from the picture what the nature of the outage is. They can then line up the right resource with the right equipment to get to the right place to restore supply as quickly as possible.

Q79 Viscount Ridley: I suppose my point is that all of that is already a good idea for dealing with weather. Do we need anything extra for dealing with climate change? What I am concerned about is that we might end up over-procuring resilience to climate change, just as we are over-procuring Capacity Mechanisms. For example, if we say that winds are going to get 10% stronger in 100 years, does that mean we have to replace the whole high-voltage transmission network because it was only designed for a certain wind speed or something like that?

Professor Bell: That is a good example of something that might need to happen but it does not all happen overnight; it cannot happen overnight. You are replacing a system that you need to use, so you have to take bits of it out of service to replace it, so it has to be planned over a period of time. We are talking about something that is probabilistic, so the probability of a certain wind speed exceeding what apparently is the safe limit—how often is that going to occur? Over how big an area are you going to experience it?

For example, in 1999 in December—I do not know whether the French Michael Fish saw it coming or not—but there were severe storms that affected north-west France and a number of transmission towers were physically damaged. It was not just a temporary outage; there was a lot of work to repair the towers and that took some weeks to get the system back up and running. What you can do in the meantime if you think the risk of that is increased is adopt some policies on spares, temporary towers that you can transport and erect relatively quickly. As old towers need to be replaced with new ones because they have reached their end of life—the metalwork needs to be replaced or the foundations—at that point you can use a new specification to be able to withstand higher wind speeds.

Again, a bit like what we were saying earlier, it is not an all or nothing, there are ways of managing it and managing the transition. But, yes, one of the things we need to be looking at is, what is the likelihood of seeing more outages due to high winds? In what way are those outages caused? What proportion of them lead to permanent damage to the towers and then making a judgment about whether new towers, as they replace old ones, need to be designed to a higher standard?

Viscount Ridley: How much we should be prepared to spend.

Professor Bell: Yes, absolutely, how much we should be prepared to spend, yes.

The Chairman: That probably brings us to a conclusion because it is 12.30 pm; we try to end by then. Thank you to Professor Grubb, Professor Newbery and Professor Bell for a very informative session. You have a great deal of expertise and we have benefited enormously from it. We will be sending a transcript for minor corrections, so we have the record correct. Thank you once more for the help you have given us.

[1] Note this is also potentially a source of confusion, because it may then get considered as demand-side response, insofar as it means these entities would run their back-up generation instead of drawing power from the grid.