House of Lords - Economic Affairs

The Economics of Renewable Energy - Economic Affairs Committee - Contents

CHAPTER 4: RENEWABLES IN THE ELECTRICITY SYSTEM

97. This chapter considers issues affecting the electricity system as a whole if, as estimated in Paragraph 24, 30-40% of electricity is to come from renewable sources. The first is the problem of intermittency. Second, we consider the impact of different levels of renewable electricity on the overall cost of generation across the industry. Third, connecting renewable generators will require significant investment in the transmission and distribution systems. Finally, we consider the way in which these systems are operated, and the implications of power flows which may be more variable in future than at present.

Intermittency—a constant problem?

98. Matching electricity supply to demand is challenging as it is not presently economic, or technically feasible, to store electricity on a large scale. Electricity can be stored in batteries for portable applications but their costs are too high for use in the national electricity grid. Electricity generation must be matched to demand on a minute-by-minute basis, or power cuts result. Some power plants are therefore kept running at less than full load, to respond rapidly to a sudden increase in demand or to make up for a power plant failure elsewhere in the system.[32]

99. But not all power stations can be "dispatched" to change their output level quickly. Coal, gas- and oil-fired stations are generally straightforward though their response speeds vary. Nuclear stations are relatively inflexible, and are best operated at a constant (full) load. Renewable generators burning biomass, and hydro generators, can generally be dispatched.[33]

100. Wind, wave and tidal stations are inherently not dispatchable. They can only generate when conditions are right—if there is no wind, or too much wind, no electricity can be produced. Tidal generators can produce much more at the spring tides (with a high variation in the water level) than at neap tides (low variation). The tides are predictable far in advance, but the wind is almost impossible to forecast more than a few days in advance, and even day-ahead forecasts can be inaccurate.

SHORT TERM FLUCTUATIONS

101. The first cost imposed by intermittency is that more plant has to be held in reserve to cope with short-term fluctuations in output. At present, National Grid, which operates the electricity system,[34] keeps a number of power stations running at less than their full capacity, providing about 1 GW of spinning reserve—that is capacity which can automatically respond to any shortfall in generation within seconds (Q 293).[35] The company also contracts with other stations to start generation quickly and has arrangements with industrial consumers to reduce their demand at short notice, in order to restore the level of spinning reserves as soon as possible after they are used. The company holds about 2.5 GW of this standing reserve (Q 293); 70% of this comes from generation, and 30% from industrial consumers (p 144).

102. As the amount of wind generation rises, the potential short-term change in wind output will also increase, and National Grid will have to hold more reserve to cope with this increase. The company told us that if renewables provided 40% of electricity generation—the share the company believes would be needed to meet the EU's 2020 energy target—its total short-term reserve requirements would jump to between 7 and 10 GW. Most of this would be standing rather than spinning reserves. This would add £500 million to £1 billion to the annual cost of these reserves—known as balancing costs—which are now around £300 million a year (Q 293). This is equivalent to around 0.3 to 0.7 pence per kWh of renewable output.

103. Estimates of balancing costs vary widely. The government has commissioned research from the consultancy SKM,[36] which estimated that if renewables provided 34% of electricity by 2020, with 27.1% from wind power, the extra cost of short-term balancing would be about 1.4 p/kWh of wind output[37] (Q 481). This equates to a total cost of £1.4 billion, well above that assumed by National Grid. Several pieces of evidence cited a 2006 report by the UK Energy Research Centre (UKERC),[38] which had estimated the balancing costs with up to 20% of intermittent renewable output in Great Britain at 0.2-0.3 pence per kWh. Although the share of renewables in the SKM study was less than double that of UKERC, the balancing costs per unit were more than five times higher. In part, this may reflect higher fuel costs since the studies surveyed by UKERC were performed; but it will also reflect the greater challenges of dealing with larger shares of intermittent renewable generation.

104. Fluctuations in wind speed lead to short term changes in electricity output from wind farms. Greater use of wind power and other intermittent renewable sources therefore requires more backup generation capacity to respond very quickly to, for example, reductions in the output of wind turbines when the wind drops. But the technical challenges and costs of backup generation on a scale large enough to balance an electricity system with a high proportion of intermittent renewable generation are still uncertain. There is currently no experience elsewhere in Europe of the scale of dependency on intermittent renewables expected in the UK. Whereas the highest share of intermittent renewable electricity now being generated is 15% in Denmark, the UK is expected to reach a share of some 30%-40%. We recommend that the Government should ensure that further work is carried out to clarify the costs and encourage development of technical solutions to deal with intermittency.

105. Running a conventional plant at part load to provide spinning reserve reduces efficiency which leads to higher emissions per unit of electricity actually generated at that plant. Some commentators, such as Campbell Dunford of the Renewable Energy Foundation, argue that this might have offset the CO₂ savings from renewable generation in Denmark. Denmark's carbon emissions per kWh generated have fluctuated from year to year, although the trend is steeply downwards, as set out in Appendix 7. Calculations based on the loss of efficiency from running a power station at part load, and the amount of extra reserve required, also suggest that the extra carbon emissions in the UK from additional spinning reserve would be very small in comparison to the savings from renewable generation. The Government has estimated the net saving from raising the share of renewable electricity to 32% to be about 45-50 million tonnes of carbon dioxide—about 8-9% of total CO₂ emissions—after taking account of the cost of part-loading plant.[39] The need to part-load conventional plant to balance the fluctuations in wind output does not have a significant impact on the net carbon savings from wind generation.

PEAK DEMAND AND CAPACITY CREDIT

106. The second cost due to intermittency comes from the need to have enough capacity available to meet peak demand. No power station is guaranteed to be available at peak demand. So the industry holds extra capacity over and above the expected peak demand to cope with stations that turn out not be available when most needed, or higher than expected demand. As a rule of thumb, a 20% margin of extra capacity has been sufficient to keep the risk of a power cut due to insufficient generation at a very low level, given the characteristics of the current system.

107. A fossil-fuelled station has around a 5% chance of not being available to generate at the time of the system peak because of breakdowns or essential maintenance (p 119). One plant's breakdown is rarely correlated with another. Nuclear plants have a similar risk, although they sometimes suffer from generic issues that require maintenance at all of the stations of a similar design.

108. But for renewables it is very different. At peak demand not only are the chances of a wind farm not being fully available much higher but it is very likely that, if so, nearby wind farms will also be at least partially unavailable because it is not windy in the area. This correlation will fall for distant wind farms—for example, the wind could well blow in Scotland when conditions in Cornwall are calm. But within the UK, the correlation does not fall to zero.

109. As a result, the proportion of renewable generation which can be relied on at peak demand is much lower than for fossil fuel plants and more complicated to calculate. We received several estimates of how far wind capacity could be counted on to contribute to meet peak demand—its "capacity credit". BERR uses a range of between 10% and 20% of wind stations' capacity, so that 25 GW of wind plant could displace between 2.5 and 5 GW of conventional plant (Q 483). E.ON suggest that the capacity credit of wind power in the UK should be only 8% (p 119).

110. As wind generation increases, its capacity credit will tend to fall because low winds over part of the country can affect many wind turbines simultaneously. Extra, offsetting conventional plant is needed. The Renewable Energy Foundation's rule of thumb is to treat the square root of the wind capacity in GW as if it were conventional capacity (Q 112). On that basis, for example, 25 GW of installed wind generation capacity could be counted on for the same contribution to peak demand as 5 GW of conventional capacity; and it would take 36 GW of wind plant to match 6 GW of conventional plant.

111. Under any of these assumptions it is clear that much conventional capacity will be required to support renewable generators coming on stream in the period up to 2020, during which many of Britain's coal and nuclear power plants are scheduled to close. To replace them, the Government has calculated that 20-25 GW of new power stations will be needed by 2020—the equivalent of more than a quarter of today's 76 GW of electricity capacity. But that calculation assumes replacement on a like-for-like basis and does not take account of the target for renewables. If some 30 GW of additional (Q 487) renewable capacity were required to meet the EU's 2020 target for the UK (and its capacity credit did not exceed 6 GW), a further 14-19 GW of new fossil fuel and nuclear capacity will still be needed to replace plants due to close and meet new demand. The total new installed electricity generating capacity required by 2020 would thus be roughly double the level needed if renewable generation were not expanded.

112. The intermittent nature of wind turbines and some other renewable generators means they can replace only a little of the capacity of fossil fuel and nuclear power plants, if security of supply is to be maintained. Investment in renewable generation capacity will therefore largely be in addition to, rather than a replacement for, the massive investment in fossil-fuel and nuclear plant required to replace the many power stations scheduled for closure by 2020. The scale and urgency of the investment required is formidable, as is the cost.

113. The UKERC study calculated the cost of building additional conventional capacity to maintain reliability in Great Britain, with up to 20% of intermittent generation would be between 0.3-0.5 pence per kWh of that intermittent generation. These costs are dominated by the fixed costs of building plant and payments to generators to keep the capacity available even though they may rarely need to generate the power.

114. The SKM report for BERR estimated the cost of additional generating capacity to maintain security of supply with a renewable share of 34% would be £316 million a year, or 0.31 pence per kWh of wind output.[40] This estimate is at the bottom of the UKERC range, because SKM assumed that the reserve capacity would have lower costs per kW than did UKERC.[41] If the studies had used the same cost of capacity, the SKM cost per kWh of wind generation would have been slightly higher than the UKERC figure, at 0.6 pence per kWh. This is because, as we have noted at paragraph 110, the capacity credit of each additional wind power station—i.e. the amount it can reliably contribute to peak electricity demand given the problems of intermittency—declines as more of them are added.

STORAGE—A PERMANENT SOLUTION TO INTERMITTENCY

115. A sufficiently great advance in electricity storage technology would help solve many of the problems of intermittency (Q 98). If the storage could be charged and emptied quickly, this would be an attractive way of balancing the system. If the cost of storage capacity is sufficiently low, it would be an effective alternative to building additional generation capacity to deal with the peak levels of demand. The Royal Society of Edinburgh reported that a range of alternative storage technologies are being considered alongside the existing use of pumped storage hydroelectricity (p 453). Ofgem told us that fuel cells could become economically viable if their costs continued to fall, or electricity prices rose (p 171).

116. Dr Clarke of the Energy Technologies Institute told us more resources had recently been applied to developing energy storage, with major industrial corporations becoming involved. He pointed out that large-scale schemes might be located close to generators, and would then smooth out the load on the transmission system, reducing its costs. Small-scale storage could help to manage local demand. High-temperature batteries, mainly used by the military at present, are more efficient than conventional batteries, and could provide a significant opportunity where waste heat from combined heat and power schemes could keep them hot enough to work properly (QQ 318-20).

117. A breakthrough in cost-effective electricity storage technology would help solve the problem of intermittency and remove a major stumbling block to wider use of renewable energy in the longer term. However, no evidence we received persuaded us that advances in storage technology would become available in time materially to affect the UK's generating requirements up to 2020. We recommend that the Government should as a matter of urgency encourage more research, development and demonstration in energy storage technologies.

The impact on the system average cost of electricity generation of an increased share from renewables

118. Chapter 3 discussed the evidence we received on the cost of individual technologies for renewable generation. We now consider the total cost of generation across the system. This requires us to take account of the mix of capacity, the load factors that different plants achieve, and the costs of intermittency discussed in the previous section.

119. In Table 3, we use the figures in Table 2 to estimate the impact of increased renewable generation on base generation costs.[42] We assume that the amount of onshore wind generation rises from 2% to 8% of the total, and that offshore wind output rises to 19%. We assume that 75% of this extra renewable output would replace gas-fired generation, and 25% would replace coal. We find that if the share of renewable output rose to 34%, the base cost of generation would rise by £4.3 billion, or 1.1 pence per kWh of total output. We also consider a case with higher coal and gas prices—50% above 2007-8 prices. The increase in generation costs is somewhat smaller, at 0.8 pence per kWh or £3.0 billion.

TABLE 3

Prediction of the impact of increasing amounts of renewable power on the system average base cost of generation

	Current fuel prices		Fuel prices rise by 50%
All figures use 2008 prices	Base cost in pence per kWh	Base cost in £ billion	Base cost in pence per kWh	Base cost in £ billion
Base cost of generation with current share of renewable output (6%)	4.3 p/kWh	£16.2 billion	5.2 p/kWh	£19.6 billion
Base cost of generation with an additional 5% onshore and 13% offshore wind (25% renewable in all)	5.0 p/kWh	£18.6 billion	5.6 p/kWh	£21.2 billion
Base cost of generation with an additional 6% onshore and 19% offshore wind (34% renewable in all)	5.4 p/kWh	£20.5 billion	6.0 p/kWh	£22.6 billion

120. The SKM study[43] cited by government witnesses also predicted a sizeable, but lower, increase in generation costs.

Investment in the Electrical Grid

121. National Grid told us that capital investment to reinforce the onshore transmission networks—the wires and pylons that carry electricity over long distances—to accommodate 40% of renewable generation would cost around £3.5 billion. This included reinforcements to the transmission network to accommodate an additional 10 GW of renewable generation in Scotland, developments in Eastern England to accommodate up to 19 GW of offshore wind generation in the North Sea and an overhead line in mid-Wales to accommodate an extra 1 GW of onshore wind generation (p 127 and appendix 2; Q 267).

122. These figures cover only the cost of upgrading the onshore transmission system. Laying cables along the sea bed to connect offshore wind farms will be expensive. National Grid expected the cost for the 19 GW of offshore wind—which it views as necessary if the EU 2020 targets are to be reached—to be in the region of £6-10 billion (Q 271). To connect 33 GW of offshore capacity to the Grid, Ofgem expected a cost of around £10 billion which is at the more optimistic end of National Grid's range of costs. Any of these figures would be well above the amounts for local connection costs included in the estimates of power station costs presented in Chapter 3.

123. National Grid's figures did not include the costs of improving local distribution networks that may be necessary in some areas to connect up the new generators. In areas where renewable resources are plentiful, the distribution system is often sparse, and new generation will trigger significant infrastructure investment, in many cases including the construction of new overhead lines. The Energy Networks Association described the provision of infrastructure to accommodate 2020 targets as challenging (p 285).

124. The SKM study cited by the Government estimated £10.2 billion would need to be spent in total on the transmission and distribution networks. That is at the low end of National Grid's range of estimates, despite the fact that SKM have included distribution costs. But SKM's figures apply to 34% of electricity coming from renewables while National Grid's are estimated for a 40% share.

125. SKM calculated these transmission and distribution costs would add a further 0.34 pence per kWh to the cost of the renewable scenario, or 1.25 pence per kWh of wind generation. Added to the other costs outlined earlier for balancing and security of supply the renewable scenario would be 27% more expensive than the conventional scenario, at 6.17 pence per kWh as opposed to 4.86 pence per kWh.

126. Table 4 gives our own estimates of the total cost of moving to a high level of renewable electricity generation in 2020. The top line of the table gives the predicted base cost of generation in 2020, on the assumption that there is no further increase in the share of renewable generation, taken from Table 3. The second line includes the cost of system balancing and consumers' payments for the existing transmission network. The third line gives our prediction for the total costs of generation and transmission in 2020, with 6% of renewable power.

127. The rest of the table considers the additional costs imposed by increasing the share of renewable generation. First, there is the higher base cost of renewable generation, from Table 3, which would add £4.3 billion. Second, there are the costs of system integration—additional balancing and reserve costs, and extra investment in the onshore and offshore transmission networks. We use the middle of the ranges given to us by National Grid for balancing and transmission costs, and the upper end of the UKERC range for reserve costs. These add £2.5 billion to the predicted cost. In total, increasing the share of renewable generation to 34% would raise the annual cost of generation and transmission by £7.5 billion. In other words, the cost of generation and transmission would rise from 4.7 pence per kWh of total output to 6.7 pence per kWh.

TABLE 4

Predicted total costs in 2020 of electricity generation and transmission with 34% of generation from renewables, including allowance for back-up and grid integration

	Pence per kWh of total output	£ billion per year
Predicted base generation cost with 6% renewables	4.31 p/kWh	£ 16.2 bn
Cost of balancing and existing transmission system	0.41 p/kWh	£ 1.5 bn
Predicted total cost with 6% renewables	4.82 p/kWh	£ 17.7 bn
Extra costs of moving from 6% to 34% renewables
Generation base cost	1.14 p/kWh	£ 4.3 bn
Predicted additional costs of system integration
Intermittency (See para 102 and 113)	0.33 p/kWh	£ 1.3 bn
Transmission (See para 121-3)	0.32 p/kWh	£ 1.2 bn
Predicted total integration costs	0.65 p/kWh	£ 2.5 bn
All extra predicted costs for moving from 6% to 34% renewables	1.79 p/kWh	£ 6.8 bn

Predicted overall cost of generation and transmission with 34% renewables	6.61 p/kWh	£ 24.5 bn

128. Our calculations suggest that the total extra annual cost of increasing the share of renewables in electricity generation from 6% to 34% in 2020 would be £6.8 billion or an extra 38%—the equivalent of an extra £80 a year for the average household. Emissions of carbon dioxide would be reduced by 52 million tonnes a year—in 2007, the UK's emissions were 544 million tonnes. This implies that the additional cost is about £130 per tonne of carbon dioxide emissions avoided.

GRID CONNECTION POLICY

129. The current policy is for each project developer to arrange a separate connection between an offshore generating plant and the electricity network on land. Ofgem is required to seek the best route forward through competition and markets wherever appropriate, and to secure value for money for consumers. They and BERR have decided to base future arrangements around competitive tenders for the major offshore transmission projects. (QQ 442-443, p 171).

130. National Grid and EDF agreed that this "radial connection" approach was fit for purpose when the decision was made. But the need to develop renewables offshore had changed significantly since then. They questioned whether this approach could be sustained for an offshore wind programme of three times the size originally envisaged. The proposed regime appeared overly complex to National Grid, with many areas still uncertain and undecided. The company believed simple, co-ordinated, regulated transmission build would be more effective to help ensure the infrastructure was in place when new renewables were ready to connect. Ofgem said if it transpired that they needed to develop an offshore grid, rather than taking a radial approach, they would not rule it out.

131. We note that the regulator's statutory duties require the use of competition wherever appropriate, and therefore give it some discretion about the use of markets. Although competition is usually preferable, we are concerned that the use of competitive tenders implies a piecemeal approach to building the networks of wires and cables required to connect offshore wind farms to the electricity grid, and that as a result the programme could become overly complex and costly. We recommend that the regulator implements the new system in a way that allows a coordinated approach for organising grid connections to offshore wind farms.

GRID CHARGES AND ACCESS

132. A second set of transmission-related issues concerns the terms on which renewable generators are able to access the grid. First, renewable generators in some parts of the country (and particularly in Scotland) face significant delays before they can be connected. In the face of insufficient transmission capacity, National Grid's response has been to delay connecting generators to the system, rather than to connect them and ration capacity (e.g. through market prices) when required. Second, there will be times when the grid cannot accept all the power generated within an area, and generators will be unable to sell their power. Third, we received some evidence querying whether the charges levied for using the transmission grid are appropriate.

DELAYS IN GRID CONNECTIONS

133. National Grid has signed agreements with 49 GW of new generation since 2005, equivalent to nearly two-thirds of the 77 GW of capacity currently connected to the system. This has created a backlog for getting projects connected to the Grid. National Grid has established the "GB queue", which promises to give generators access to the grid in the order in which they signed connection agreements with the company. But new entrants may receive a connection date in ten years' time, according to Scottish and Southern Energy (p 85).

134. Ofgem and BERR have recently concluded a Transmission Access Review, which asked National Grid to take a more proactive approach to managing this queue, giving priority to projects that had received planning permission for their plant. Ofgem told us that National Grid is now taking a more robust approach to removing unviable or purely speculative projects (p 171). We welcome these measures to organise better the queue of renewable generation projects awaiting connection to the electricity grid. They should reduce delays in connecting viable generation projects and push back schemes unlikely to get off the ground.

SURPLUS POWER

135. When National Grid is unable to accept power from a generator, through a lack of transmission capacity, it has to compensate the generator accordingly. But as wind output grows, flows on the transmission system will become more variable. British Energy pointed out that at very high levels of wind-generated electricity, there may be periods when it will be necessary for National Grid to instruct wind generators to spill power because renewable generation exceeds demand, net of other plants that are required to run. National Grid's policy at present is to invest in a way that keeps the cost of transmission constraints at about £100 million a year. They noted that without additional investment on the lines between Scotland and England, the cost of constraints might rise to between £500 million and £1 billion a year. This would cost a household consuming the average amount of electricity between £7 and £14 a year.

136. It is important that the transmission companies strike the right balance between investing in transmission and paying stations to be constrained off (p 311). It would not be economic to build a network in which transmission constraints were very rare, as the capacity needed to achieve this would cost too much. For an analogy, a motorway with six lanes in each direction might never see a traffic jam, but it would be a better use of resources to build a three-lane motorway and accept a few hours of congestion each month. A two-lane road that was congested for several hours a day would be inadequate, however. If the electricity transmission companies do not invest enough in the grid, congestion costs will be excessive; the cost of inadequate capacity is not a traffic jam, but that the system operators may have to use expensive stations near the loads rather than cheaper stations further away. Transmission access arrangements should address this issue. A series of proposals are outlined in Appendix 8. The key feature is that if new generators in areas likely to have a surplus of power must pay more to use the grid, they will tend to avoid these areas, reducing the amount that needs to be spent on transmission investments.

TRANSMISSION CHARGES AND LOSSES

137. The Transmission Use of System charges are set by National Grid to recover the costs of all three transmission system owners in Great Britain. Generally, the more remote a generator, the higher the charges it has to pay because of the greater investment required in the transmission network to carry the electricity to centres of high demand. So a wind farm in the Highlands faces higher charges than an identical one near London. However, Professor Bain argued that the transmission charging system did not take account of the electricity lost in transmission (p 227). While these losses are only equal to 2% of electricity generated on average (p 144), they rise with the distance that power has to travel (Q 291). Furthermore, when power flows increase, the marginal losses are twice the average level. National Grid's Seven Year Statement[44] shows that an extra 100 MW of generation in the Highlands would only replace 90 MW in the Midlands, as a result of the additional transmission losses incurred. In 2002, Ofgem proposed that transmission losses should be taken into account in the industry's trading mechanisms, but was over-ruled by the Government.

138. The Government was concerned about the impact on the costs of renewable generators located in Scotland. The higher the cost, the greater the financial support required (cf. chapter 6). If the system of support gives similar payments to every generator (of a given type), then the costs of the most expensive generator deemed to deserve support (the marginal generator) determine the payments to each of them. If the marginal generator is located in Scotland, then charging for transmission losses would increase the amount of support it required, and raise the amount of profit made by those generators in areas with lower transmission losses.

139. In the broader context, E.ON told us that charging for use of the transmission system should continue to reflect the costs to the system associated with generating from renewables and other generation at that location on the system. This would help ensure that these costs are taken into account in the decision where to site the project in the first place (p 108). But Scottish and Southern Energy, which has wind farms in remote parts of Scotland, argued against "the current perverse mechanism of regional charges". Ofgem has successfully defended a judicial review on the basis that "it was absolutely right that people who were at the extremities of the system should pay very high charges that reflected the economic costs of transmitting electricity a long way from where it is produced to where it is used" (Q 422). We agree with this position. We consider that the current system of Transmission Use of System charges sends broadly appropriate signals of the costs of locating generators at different points on the system.

MITIGATING INTERMITTENCY BY MORE CONNECTIONS TO THE CONTINENTAL GRIDS

140. Over the long term the costs of intermittency could be mitigated by greater interconnection between the electricity grids of Britain and the rest of Europe (p 56, Q 20). Unlike Denmark, for example, Britain has very little capacity to import or export electricity to other countries. There is a link to France with 2 GW of capacity. National Grid and the Dutch transmission operator are building a cable to the Netherlands with 1 GW of capacity, expected to cost £480 million. Three other lines—to Belgium, Norway and the Republic of Ireland—have been studied, but no contracts have been signed. Even if all three were built, the total import capacity would be roughly 6 GW, compared to our peak demand of over 60 GW. Britain is, in effect, an 'island generator'. This complicates the task of managing intermittent generation.[45]

141. Greater inter-connectedness would allow Britain to tap renewable sources over a wider area which would reduce the problems with intermittency. For example, when the wind is blowing in Denmark but not Britain, electricity from Danish wind farms could then be imported. The wider the area of interconnectedness, the more likely it is that variations in wind patterns will cancel out, although the weather may sometimes be similar over even a wider area. For example, we received some evidence that low wind speeds in the UK could coincide with similar conditions in Germany, Ireland and even as far away as Spain.[46] Furthermore, it would not be economic to build enough interconnector capacity to solve our problems with intermittency, for wholesale electricity prices in the UK and on the Continent would then converge, whereas the interconnectors need different prices in the two markets to profit from trading between them. Greater interconnector capacity with the Continent would reduce, but not solve, the problems of intermittent renewable generation.

32 The limited number of pumped storage stations in Scotland and Wales also increase output when necessary. They use electricity to pump water uphill, and release it when necessary to turn a turbine to generate power. But pumped storage carries an efficiency penalty, in that less electricity is generated at the end of the cycle than is required at the start. Nonetheless, it can be economic if the electricity used in pumping is cheap (which is typically the case overnight), and the water is released at peak times when power can be sold for a high price. Back

33 However, a hydro station may have to "spill" water if its storage is full and its generation is not required. Back

34 In Great Britain, the system operator is National Grid, controlling its own transmission system and those owned by Scottish Power and Scottish and Southern Energy. In Northern Ireland, SONI (System Operator for Northern Ireland Ltd) is a subsidiary of Northern Ireland Electricity. Back

35 For the purpose of comparison current total generation capacity is 76 GW. The amount of spinning reserve that National Grid holds is currently based on the size of the largest single generator on the system. This allows the company to cope with any single failure, on the basis that the near-simultaneous failure of two large generators is sufficiently unlikely. Back

36 Sinclair Knight Merz (2008) Growth Scenarios for UK Renewables Generation and Implications for Future Developments and Operation of Electricity Networks BERR Publication URN 08/1021 Back

37 SKM presented these costs as 0.07 pence per kWh of total generation in a scenario with low levels of wind power (3.1%), and as 0.45 pence per kWh of total generation in a scenario with wind power making up 27.1% of total generation (table 7.12). We divided the difference of 0.38 pence per kWh by 27.1% to give the figure for the increased cost per kWh of wind generation. Back

38 Gross, R., P. Heptonstall, D. Anderson, T.C. Green, M. Leach and J. Skea (2006) The Costs and Impacts of Intermittency: An assessment of the evidence on the costs and impacts of intermittent generation on the British electricity network, London, UK Energy Research Centre. Back

39 Department for Business, Enterprise and Regulatory Reform (2008), UK Renewable Energy Strategy; Consultation Document, June 2008. The UK emitted 557 million tonnes of CO₂ in 2006. Back

40 SKM assumed that in a scenario with 27% of wind generation (and 34% renewable generation in total) 19.5 GW of conventional capacity would be needed but would not generate throughout the year, compared with 10.4 GW in a scenario with 3% of wind generation. The cost of the extra 9.1 GW of conventional capacity would be £316 million a year, which gives 0.31 pence for each of the 102 billion kWh of wind generation. (table 7.13) Back

41 UKERC based their estimates on a new Combined Cycle Gas Turbine plant costing £67 per kW per year, whereas SKM assumed that already existing plant could be kept open at a cost of just £35 per kW per year. Back

42 We have followed the Government's consultants in assuming that the capital costs of generators will fall slightly over the years to 2020-if we were to assume constant costs over the period, the additional cost of renewable generation would be higher by about £500 million a year. Back

43 Sinclair Knight Merz (2008) Growth Scenarios for UK Renewables Generation and Implications for Future Developments and Operation of Electricity Networks BERR Publication URN 08/1021 SKM estimated the average base cost of generation in 2020 with 6% of renewable output would be 4.68 pence per kWh. If the share of renewable output rose to 34% the average base cost of generation would be 5.19 pence per kWh. This gives an increase of 0.51 pence per kWh of total output, or £1.9 billion, which in turn implies that the base cost for each kWh of renewable output is about 2 pence higher than the base cost of the kWh of conventional output it replaced. Our figures are based on Table 2, which predicts offshore wind to cost roughly 4 pence per kWh more than coal- or gas-fired electricity. Back