Volume 2 wind energy 2 15 – wind energy economics

Volume 2 wind energy 2 15 – wind energy economics Volume 2 wind energy 2 15 – wind energy economics Volume 2 wind energy 2 15 – wind energy economics Volume 2 wind energy 2 15 – wind energy economics Volume 2 wind energy 2 15 – wind energy economics Volume 2 wind energy 2 15 – wind energy economics

2.15 Wind Energy Economics D Milborrow, Lewes, East Sussex, UK © 2012 Elsevier Ltd All rights reserved 2.15.1 2.15.2 2.15.2.1 2.15.2.1.1 2.15.2.1.2 2.15.2.1.3 2.15.2.1.4 2.15.2.1.5 2.15.2.2 2.15.2.3 2.15.2.4 2.15.2.5 2.15.2.5.1 2.15.3 2.15.3.1 2.15.3.2 2.15.3.3 2.15.3.4 2.15.4 2.15.4.1 2.15.4.2 2.15.4.3 2.15.4.4 2.15.4.4.1 2.15.4.5 2.15.4.6 2.15.5 2.15.5.1 2.15.5.2 2.15.6 2.15.6.1 2.15.6.1.1 2.15.6.2 2.15.6.3 2.15.7 2.15.7.1 2.15.7.2 2.15.7.3 2.15.7.4 2.15.7.4.1 2.15.7.5 2.15.7.6 2.15.8 2.15.8.1 2.15.8.1.1 2.15.8.2 2.15.8.3 2.15.8.4 2.15.8.5 2.15.8.6 2.15.8.7 2.15.8.8 2.15.8.9 Introduction Basic Financial Issues Definitions Cost inputs Energy prices Net present value Payback period Internal rate of return Price Calculation Methods Recommended Practices Interest Rates Amortization Periods Influence of interest rates and repayment periods Cost and Performance Issues Balance of Plant Costs Operational Costs Size of Wind Farm Installed Costs and Wind Speeds Onshore Wind Historical Cost and Performance Trends Current Plant Costs Current Electricity Generation Costs Small Wind Turbines Offshore wind Historical Price Trends Current Installed Costs Analysis of Offshore Costs Operation and Maintenance Costs Water Depth and Distance from Shore Electricity-Generating Costs Generation Cost Comparisons Key issues Cost Comparisons – Wind and Other Plant Cost Comparisons on a Level Playing Field External Costs Types of External Cost Costing Pollution Market Solutions The UK Climate Change Levy External costs of renewable energies Renewable Energy Support Mechanisms Embedded Generation Benefits Variability Costs Electricity Networks Economies of scale Characteristics of Wind Energy Assimilating Wind Extra Short-Term Reserve Needs and Costs Carbon Dioxide Savings Extra Backup and Its Costs Capacity Credit The Cost of Backup Transmission Constraints Comprehensive Renewable Energy, Volume doi:10.1016/B978-0-08-087872-0.00218-3 470 471 471 471 471 471 471 472 472 472 472 472 473 474 474 474 475 475 475 475 476 477 477 478 478 479 480 480 481 481 482 482 483 483 484 484 485 485 486 486 486 486 487 488 488 489 489 490 491 491 491 492 492 469 470 Wind Energy Economics 2.15.8.10 2.15.8.11 2.15.8.12 2.15.8.12.1 2.15.8.12.2 2.15.8.12.3 2.15.8.12.4 2.15.8.12.5 2.15.8.12.6 2.15.8.12.7 2.15.8.12.8 2.15.8.12.9 2.15.9 2.15.10 2.15.10.1 2.15.10.2 2.15.11 References Wind Surpluses at High Penetration Levels Total Costs of Variability Mitigating the Effects and Costs of Variability Wind forecasting Demand management Energy storage System storage European supergrids Electric cars ‘Smart grids’ and the growth of decentralized generation Electric spaces and water heating Overall effects Total Cost Estimates Future Price Trends Future Fuel Prices Price Comparisons in 2020 Conclusions 492 493 494 494 494 495 495 496 496 496 497 497 497 497 498 499 499 500 2.15.1 Introduction The aims of this chapter are: • To review the generation costs for onshore and offshore wind energy; determine the relative importance of turbine prices, slight costs, and other factors; and indicate how generation costs are influenced by test discount rates and depreciation periods • To discuss the key factors that need to be taken into account when comparing renewable energy generation costs with those from the thermal sources, principally: • External costs • The ‘embedded benefits’ of renewable sources (which can be positive or negative) • The extra balancing costs incurred by the variable sources such as wind, wave, and solar There are no absolutes in energy prices No single number can be assigned to the price of wind energy The same is true of energy prices for thermal plant Unless the relevant assumptions are clearly set out, single numbers are virtually meaningless It is important to distinguish between the cost of plant (such as wind turbines and wind farms) and the price of the electrical energy they produce Capital costs are primarily a function of the size of the installation (due to economies of scale) This chapter focuses principally on large-scale wind farms as these deliver the cheapest electricity costs, but there is a brief review of the costs of smaller installations Wind energy prices depend on wind speeds and institutional factors, and have two components: • Capital charges, including depreciation and interest charges • Operating costs The calculation of wind energy generation prices follows procedures that are reasonably standardized across the power industry ‘Real’, that is, net of inflation, interest rates (test discount rates) are used to determine the capital charges The term weighted average cost of capital is also used National interest rates and repayment periods vary widely, but comparisons can be made provided the levels are quoted Institutional factors account for most of the apparent variations in quoted wind energy prices In Denmark, for example, the costs of grid reinforcement for wind installations are not always charged to the developer and the utilities tend to use relatively low interest rates (5–6%) In Britain and the United States, however, investor-owned utilities tend to use interest rates around 10% In this chapter, for the sake of uniformity, an interest rate of 8% is used, except where stated Comparisons of generation costs are not necessarily the most equitable way of assessing the competitive position of renewables and so there is a brief description of the ‘external costs’ of electricity-generating sources, drawing on the European Union’s (EU’s) ‘ExternE’ project plus other material A complete and fair comparison between wind energy generation costs and those of the fossil fuel sources demands that the additional costs associated with the variability of wind is also taken into account and that the relevant issues are discussed The chapter also discusses methods of valuing renewable energy One of the distinctive features of renewable energy is that most plants are ‘embedded’ in distribution networks – rather than connected to transmission networks They therefore accrue additional value, beyond simple ‘fuel saving’ levels The relevant issues are discussed, but it is also noted that the ‘embedded value’ may be negative if exploitation of renewables demands major transmission reinforcement Finally, examples of ‘total cost estimates’ are Wind Energy Economics 471 discussed, which take into account all the debits and credits associated with renewable energy costs The chapter concludes with a discussion of published forecasts of future price trends Note on currency conversions: Present-day (2010) prices have been converted on the basis that £1 = $1.60 and €1 = $1.36 Historical data have been converted at the exchange rates prevailing at the time, but where the data are intended to show trends, rather than absolute values, some have been left in the original units 2.15.2 Basic Financial Issues 2.15.2.1 Definitions To the layman, costs are paid to acquire items and prices are the amount for which items are sold or offered Value is the worth of an item to the recipient The context must therefore be clear Manufacturers incur costs when building wind turbines and sell them at a price that includes their profit A wind energy developer, however, will incur costs throughout the lifetime of the project, which include the turbines, the ‘balance of plant’ costs plus operation and maintenance costs Developers sell the energy at a price that includes profit Inputs are therefore costs and the output is a price However, conventional terminology sometimes means that rigorous definitions cannot be used without causing confusion The term ‘generation cost’, for example, is very widely used, even though it usually includes an element of profit It has been retained here, even though, strictly speaking, the term should be ‘generation price’ 2.15.2.1.1 Cost inputs There are two cost inputs to a wind developer or owner The capital, or installed, costs of the plant are most frequently quoted in terms of investment cost per installed kilowatt, or $ kW−1, and, broadly speaking, are primarily a function of the size of the installation (due to economies of scale) Operating costs are the second input, and, similarly, depend on size, but may be dependent, also, on energy yield 2.15.2.1.2 Energy prices These depend on installed costs, wind speeds, test discount rates, and capital repayment periods, and are simply the sum of the capital cost element and the operating costs, and are expressed in various ways although $ MWh−1 is used in this chapter In general, there are three principal components of electricity generation costs: Those associated with repayments of capital (plus interest) Quantifying the total capital costs that need to be included when putting a price on electricity-generating plant is relatively straightforward The capital cost includes the cost of the plant, land acquisition (unless a rent is paid, in which case this is a running cost), grid connection (although in some European states, the utility has borne the cost), and initial financing costs (not repayment costs) Operating costs include insurance, rent, and local authority rates, as well as the costs of labor and materials used for operations and maintenance Fuel costs Fuel costs for wind, wave, geothermal, and solar installations are zero The term ‘reference price’ is used in this chapter to denote energy prices calculated using standard procedures, with defined interest rates These must be distinguished from prices relevant to particular states, which use appropriate levels of interest rate and repayment period; these are given the generic name ‘national prices’ 2.15.2.1.3 Net present value A number of financial parameters are used when evaluating the viability of wind energy projects The net present value (NPV) is [1] the net present worth of a time series of cash flows, both incoming and outgoing NPV is a central tool in discounted cash flow (DCF) analysis, and is a standard method for using the time value of money to appraise long-term projects Used for capital budgeting, and widely throughout economics, finance, and accounting, it measures the excess or shortfall of cash flows, in present value terms, once financing charges are met The NPV is calculated by summing the yearly cash flows that have been discounted by an appropriate discount rate That rate may be set by government – if a wind project is a public sector initiative – or by the individual developer Public sector projects tend to use lower rates – 5% or 6%, say – and private sector developments usually use rates in the range 8–10% 2.15.2.1.4 Payback period When calculating NPVs, the first term in the time series – in the case of a wind project – is the outlay for the capital cost of the plant That has a negative value and the remuneration from the electricity sales minus the provisions for operation and maintenance charges and other outgoings generally has a positive value When the sum of these positive monies equals the initial capital outlay, 472 Wind Energy Economics this defines the ‘payback period’ ‘Simple payback’ is calculated in a similar way, but no discount rate is applied to the summation of the yearly cash flows and so it is invariably less 2.15.2.1.5 Internal rate of return The internal rate of return (IRR) of an investment is the interest rate at which the NPV of costs (negative cash flows) of the investment equals the NPV of the benefits (positive cash flows) of the investment IRRs are commonly used to evaluate the desirability of investments or projects The higher a project’s IRR, the more desirable it is to undertake the project Assuming all other factors are equal among the various projects, the project with the highest IRR would probably be considered the best and undertaken first [2] 2.15.2.2 Price Calculation Methods The basic procedures for setting selling prices for electricity are set out in standard texts [3] The first step is to calculate the annual capital charge, which depends on the project interest rate, or ‘test discount rate’ and repayment period An ‘annual charge rate’ is used, which expresses the fraction of capital cost that needs to be charged each year in order to yield the required rate of return over the specified period The annual capital charge includes capital depreciation and interest charges and is divided by the annual energy output to yield the capital element of energy price The energy output of wind plant is primarily dependent on the wind regime, that is, on its geographical location, and on the performance of the wind turbines Institutional factors arguably account for most of the apparent variations in quoted wind energy generation costs These factors may also influence the exact makeup of capital and operating costs In Denmark, for example, the costs of grid reinforcement for wind installations have not always been charged to developers Similarly, utilities in Denmark and elsewhere may not always charge their overheads to ‘operating costs’, whereas in Britain, by contrast – where private developers undertake all wind developments – all costs are included 2.15.2.3 Recommended Practices The calculation of wind energy generation costs follows procedures, which are reasonably standardized across the power industry The International Energy Agency (IEA) has published guidelines in the form of a ‘Recommend Practice’ [4] for wind energy, and these are similar to those used for other renewables and for thermal plants The IEA document advocates the use of ‘real’, that is, net of inflation, interest rates – more accurately, test discount rates – for the calculations, which is also common practice The following items are included in energy price calculations: • • • • • • • • Planning costs Capital cost of plant Construction costs Interest during construction Land costs (either as part of the capital or as annual leasing payments) Fuel costs – zero for renewable energy plants Operating costs (O&M), including labor, materials, rents, taxes, and insurance Decommissioning The IEA document recommends that capital costs are amortized over the technical life of the plant and that standard test discount rates are used While this may produce useful data for comparative purposes, actual interest rates and regulatory or institutional frameworks, as noted above, control amortization periods Each of these parameters needs, therefore, to be considered in more detail 2.15.2.4 Interest Rates Public sector companies use test discount rates set by government These vary, generally between 5% and 8% Private companies set their own rates In practice, many projects are financed using a mixture of loan and equity funding A typical ratio is 80/20 If the loan interest rate is 6% and the equity return is 25%, the equivalent test discount rate for the project as a whole is 10% This is a fairly typical value for the private sector Irrespective of the sources of finance, a ‘cost of capital’ can be derived that expresses the interest rate that can be applied to the total cost of the project to determine generation costs Alternative nomenclature for this parameter is the ‘test discount rate’, ‘project interest rate’, or ‘weighted average cost of capital’ 2.15.2.5 Amortization Periods Amortization and depreciation periods also vary, and are not necessarily as long as the life of the plant This is rarely used outside the public sector The IEA Recommended Practice notes that “20 years is commonly used for proven grid connected wind turbines” and is used in this chapter as a default value Wind Energy Economics 2.15.2.5.1 473 Influence of interest rates and repayment periods National institutional factors have a major influence on the financing terms for borrowing money and on investors’ expectations for the return on the equity they put into a project In some places, all the investment is provided by the state, or by nationalized industries, and so the equity contribution is zero Before privatization of the electricity industry in the United Kingdom, the test discount rate was set by the Treasury and was 8% in 1990 In Denmark today, utilities almost invariably use ‘public sector’ test discount rates, typically 5% or 6% In Britain and the United States today there are no fixed criteria and project developers use criteria that tend to be strongly influenced by the financiers The overall interest rate is dependent on the relative proportions of debt and equity and the appropriate interest rates; most tend to be in the range 8–11% The other parameter that strongly influences generation costs is the repayment period The longer this is, the lower the annual payment to cover depreciation and interest and hence the lower the generation cost In Denmark, this period often corresponds to the expected life of the project – 20–25 years is common In the private sector, depreciation periods also vary, they were generally in the region of 12–15 years in the early years of wind energy development, but increasing investor confidence means that finance can now be secured for up to 20 years in some instances The ‘annual charge rate’ is the fraction of capital payable each year to cover repayments of the original investment plus interest, and depends on both interest rate and repayment period Typical values are shown in Table Once the annual cost has been determined, the corresponding generating cost component is simply the annual cost divided by the annual energy generation Financing terms can have a significant impact on the ‘capital cost’ element of electricity-generating costs The most stringent criteria (bottom line of Table 1) mean that annual charges will be double those associated with more relaxed criteria (top line of table) If the capital cost element of the total generation cost is small (as in the case of gas-fired generation), changes in the financing terms will have very little impact on generating costs In the case of renewable energy technologies such as wind, wave, photovoltaics (PVs), and tidal, where capital costs form the largest element of generation cost, the impact will be much greater The data shown in Figure illustrate just how important interest rates are to capital-intensive electricity-generating technologies such as wind The estimates of electricity generation costs have been derived using the cost data listed in Table Note that these Table Typical annual charge rates Interest rate (%) Repayment period (yr) Annual charge rate Comments 8 10 12 20 20 15 20 12 0.0802 0.1019 0.1168 0.1175 0.1614 Used in Denmark and in some American States Old ‘nationalized industry’ UK figure Typical UK Non-Fossil Fuel Obligation criteria Widely used now that wind is well established For investments seen as more risky, for example, offshore wind Energy price ($/MWh) 90 80 70 WIND 10% 5% GAS 10% 5% 60 50 40 30 10 15 20 25 Amortization period (years) Figure Gas and wind energy prices – influence of interest rates and repayment periods Table Data used to illustrate effects of interest rates and repayment periods Technology Wind Gas Capital cost, $ kW−1 Build time, yr Load factor, % Fuel cost, $ MWh−1 O&M cost, ECU MWh−1 1000 0.5 0.26 10 724 0.8 33 474 Wind Energy Economics Table UK Renewable energy prices Technology NFFO5 price (£ MWh−1) August 2006 price (£ MWh−1) Landfill gas Energy from waste Hydro Wind projects > 2.3 MW 27.3 24.3 40.8 28.8 108.3 54.8 98.0 102.3 data are purely for illustrative purposes Whereas the spread of prices from gas-fired generation is only about $10 MWh−1 (from a 10-year amortization period and 10% interest rate to 25 years amortization and 5% interest rate), the corresponding spread for wind energy is about $42 MWh−1 With a 20-year amortization period and 10% interest rate, wind is dearer than gas, but a change in the interest rate to 5% brings the prices into line Extending the amortization period to 25 years brings wind prices below gas prices It is assumed that gas prices remain constant in real terms over the period Changes in national institutional frameworks illustrate the maxim “Prices are what you want them to be,” and price variations in recent years for renewable energy in the United Kingdom bear this out In the last round of contracts under the Non-Fossil Fuel Obligation in 1998, the average price for ‘large’ wind energy projects was £28.8 MWh−1, whereas the prices realized in the auction by the Non-Fossil Purchasing Agency in August 2006 was £102.3 MWh−1 Table illustrates how dramatically headline prices for four technologies have changed; the 1998 prices have not been corrected for inflation but this would make little difference to the comparison Under the previous arrangements – the Non-Fossil Fuel Obligation – the prices bid by renewable generators were (assuming they were successful) the prices they were paid Under the new arrangements, prices are subject to negotiation between generators and suppliers and are a function of ‘supply and demand’ While the plant costs of renewable generation have not changed dramatically, the price paid, under the new arrangements, is now a complex amalgam of numerous different components The generators are paid less than the price in Table 3, but the supplier takes the risk that prices into the future may be less than the prices quoted here Further details of wind energy support mechanisms are discussed later It may be noted that national institutional factors are responsible for most of the apparent variations in renewable energy generation prices These influence the ‘cost of capital’ and depreciation periods – the latter are not necessarily the same as project lifetimes 2.15.3 Cost and Performance Issues Wind turbine and wind farm costs are often quoted on the basis of a price per unit of installed capacity ($ kW−1), but such data can be misleading as manufacturers have differing design philosophies – there is no fixed relationship between rotor size and rating A 40 m machine, for example, may have a rating anywhere between 350 and 500 kW The price of a machine with a high rating, relative to its size, will therefore tend to appear low when compared with one that has a low rating for its size More reliable comparisons may be made by comparing prices on the basis of price per unit swept area ($ m−2) However, although this may be more rigorous, it is probably less readily understood and is used less often 2.15.3.1 Balance of Plant Costs Balance of plant costs – the extra costs, additional to those of the wind turbines – add between 15% (onshore) and 100% (offshore) to wind turbine costs, depending on the number and size of machines in the wind farm, and the location The windiest onshore sites – on hilltop sites, often remote from a grid connection, or coastal locations where deep piling into silt is needed, tend to incur costs above average Similar considerations apply offshore, where sites in deep water and far from shore are the most expensive Although wind turbine prices (in $ kW−1) may increase at the large size ranges, there are nevertheless sound reasons for pursuing the development of such machines A number of items in the ‘balance of plant’ costs decrease with size (in $ kW−1 terms) of machine and/or machine numbers such as: • Foundation costs • Electrical interconnection costs • Access tracks The way in which foundation costs decrease with increasing size has been examined in a study carried out by the author [5] It was found that foundation costs for 800 kW machines were around one-third less than those of 200 kW machines 2.15.3.2 Operational Costs Operational costs also fall with increase of turbine size Analysis of data from German wind installations has shown that the price of insurance and guarantees both halved, approximately, as ratings increase from 200 to 600 kW Total costs, which are similar across Wind Energy Economics Table 475 Operational costs Item Cost basis Service contract Administration Insurance Land rent Local rates Electricity usage Reactive power $ kW−1 and $ MWh−1 both used Cost basis varies, $ kW−1 common $ kW−1 more usual 1–2% of revenue, if land not bought $ kW−1 more usual Standard tariffs Up to 0.4 kVArh most of Europe, fell from around 38 electronic control unit (ECU) kW−1 yr−1 at the 200 kW size to around 15 ECU kW−1 yr−1 at 600 kW [5] (the ECU was slightly above parity with the US$ during this period) Operational costs vary between countries and wind farms Some elements are fixed annual sums, so wind farms on high wind speed sites may have lower costs per unit of electricity generated Table shows the main components and the usual basis of charging is denoted Several costs are size-dependent Total costs are in the range $32–60 kW−1 yr−1 onshore and $90–150 kW−1 offshore Not all utilities charge for reactive power – and not all machines need it Land rent may be an explicit cost if a developer builds on land he does not own, but may not appear if the wind farm operator owns the land In the latter case, the cost of the land may be included in the overall capital cost, or the owner may simply receive remuneration from the overall profit 2.15.3.3 Size of Wind Farm The size of a wind farm influences its cost, as large developments • οften attract discounts from wind turbine manufacturers • enable site infrastructure costs to be spread over a number of machines, reducing the unit cost • enable more effective use of maintenance staff 2.15.3.4 Installed Costs and Wind Speeds The attraction of hilly sites onshore and remote sites offshore is the higher wind speeds Developers can afford higher costs, if they get a higher yield Increasing the wind speed, for example, from to m s−1 at hub height will typically increase output from a wind turbine by 10% However, remote hilltop sites generally cost more to develop than flat, low sites The link between onshore installed costs and winds has been examined [5, 6] and it was found that installed costs in Britain increased by between 16% and 25% as the site wind speed increased from 7.5 to m s−1 Another factor that influences this correlation is that tall towers are often used in places such as Germany to obtain higher wind speeds, but at increased turbine cost 2.15.4 Onshore Wind 2.15.4.1 Historical Cost and Performance Trends The cost of wind energy plant fell substantially during the period from 1980 to 2004 Figure shows the average worldwide selling price of wind turbines from the early days (1980) in California to around 1998 (when worldwide capacity was around 10 GW) [5] Average machine price ($1996/kW) 5000 3000 2000 1000 500 10 100 1000 10 000 Cumulative capacity (MW) Figure Average price of wind turbines, as a function of worldwide capacity 100 000 476 Wind Energy Economics Energy productivity (kWh m−2) 2000 1500 1000 500 0 20 40 60 80 100 120 Rotor diameter (m) Figure Productivity of wind turbines in Denmark, 2005 Source of data: Danish Energy Agency (1999) Presentation of the data in this way enables the ‘learning curve’ reduction to be derived (this is the reduction in cost that is achieved for every doubling of capacity) The data that are plotted suggest this ratio is about 15%, which is consistent with estimates derived by other analyses Prices in the 1980s were around $3000 kW−1, or more, and by 1998 they had come down by a factor of During that period, the size of machines increased significantly – from around 55 kW to MW or more – and manufacturers increased productivity substantially In 1992, for example, one of the major manufacturers employed over seven people per megawatt of capacity sold, but by 2001, only two people per megawatt were needed The energy productivity of wind turbines also increased during this period This was partly due to improved efficiency and availability, but also due to the fact that the larger machines were taller and so intercepted higher wind speeds This is illustrated in Figure 3, using data from the Danish Environment Ministry The small machines in 2005 delivered around 500 kWh m−2 of rotor area, or less, whereas the largest machines delivered over twice that amount A further factor that led to a rapid decline in electricity production costs was the lower operation and maintenance costs With capital costs halving between 1985 and the end of the century, and productivity doubling, it may be expected that electricity production costs might fall by a factor of This general trend has been confirmed by the Danish Energy Agency; they suggest that generation costs fell from 1.2 DKK kWh−1 in 1982 to around 0.3 DKK kWh−1 in 1998 (1 DKK1998 = $0.144) [7] Shortly after the turn of the century, the downward trend in wind turbine and wind farm prices halted and prices moved upward, as shown in Figure This was partly due to significant increases in commodity prices and partly due to shortages of wind turbines Prices appear to have peaked in 2008, with wind farms averaging just under $2200 kW−1 and wind turbines at just under $1600 kW−1 Although the prices may now be falling, based on data available to the autumn of 2009, it should be noted that a complete data set was not available for 2009 at the time the graph was compiled 2.15.4.2 Current Plant Costs The price of modern turbines around 55–65 m in diameter, for onshore installations, is around $1500 kW−1 The ‘most economic size’ of machine has changed over the years and is still moving upward The larger the machines, the fewer are required for a given capacity This brings savings in site costs and in operation and maintenance costs, as noted earlier Overall, site costs add between 15% and 40% to wind turbine costs, depending on the number and size of machines in the wind farm, and the location Figure shows a typical cost breakdown for a complete wind farm [8], which now (late 2009) costs around $2150 kW−1 An analysis of published data from about 30 wind farms around the world, which were completed between January and October 2009 with a total capacity of over 3000 MW, suggests that the average installed cost was $2150 kW−1, with a standard deviation of $466 kW−1 [9] An analysis by the Lawrence Berkeley Laboratory in the United States reached a very similar result – $2120 kW−1 [10] Price ($/kW) 2500 2000 1500 1000 Farms 500 2004 2005 2006 2007 Year Figure Wind farm and turbine prices, 2004–09 Turbines 2008 2009 Wind Energy Economics 477 Financial costs 1.2% Road construction 0.9% Control systems 0.3% Consultancy 1.2% Electric installation 1.5% Land rent 3.9% Foundation 6.5% Grid connection 8.9% Turbine (ex works) 75.6% Figure Cost breakdown for an onshore wind farm 2.15.4.3 Current Electricity Generation Costs Using estimates Ỉ1 standard deviation from the average installed cost, quoted in the previous paragraph, and rounding slightly, suggests that generation costs based on installed costs between $1700 and $2600 kW−1 should have a wide relevance These estimates, which are shown in Figure 6, were derived using an 8% discount rate and a 20-year amortization period; operating costs have been set at $32 kW−1 yr−1 for the lower capital cost and $60 kW−1 yr−1 for the higher capital cost The link between wind speed and energy productivity has been established by examining the performance characteristics of a number of large wind turbines that are currently available Although there is not a unique link between wind speed and capacity factor, the spread is quite small All wind speeds refer to hub height The estimates suggest that generation costs at $2600 kW−1 range from just under $200 MWh−1 at m s−1, falling to $87 MWh−1 at 9.5 m s−1 At $1700 kW−1, the corresponding range is $125 to $55 MWh−1, respectively 2.15.4.4 Small Wind Turbines The discussion and analysis so far has related to wind turbines of MW size and above, and to wind farms of greater capacity As with most technologies, economies of scale yield significant savings of installed costs, but these effects become progressively smaller at the larger scales Conversely, below MW (roughly), installed costs rise Although these higher costs (and lower energy yields) translate to higher electricity-generating costs, that is not necessarily a drawback as the price comparator is likely to be different Large wind farms compete with gas and coal-fired power stations, but small- and medium-sized wind turbines may be competing with expensive electricity from diesel generators on remote islands or perhaps on remote farms in the developed world There is such a wide variety of applications for small wind systems that it is difficult to lay down guidelines as to what are the target electricity-generating costs Figure suggests that fully installed costs for small wind turbines range from around $8000 kW−1, for machines with a rating of around kW, around $5500 kW−1 for a 10 kW machine, down to around $4000 kW−1 for a 100 kW machine [11] These data come from a recent generalized analysis, but the trends are similar to those from an earlier analysis for turbines, only, rather than installed costs shown in Figure A very similar trend is evident in the earlier data, with prices for very small machines straddling around $8000 kW−1, falling to around $2050 kW−1 at 10 kW size and $1560 kW−1 at 100 kW rating [12] At that time (1999), much more information on selling prices was available than at the present time; there are not enough data at the present time to enable the Generation cost ($/MWh) $2600/kW 150 $1700/kW 100 50 Site mean wind at hub height (m s−1) Figure Estimates of electricity generation costs from onshore wind turbines 10 478 Wind Energy Economics Installed cost ($/kW) 10 000 8000 6000 4000 2000 10 100 1000 Turbine size (kW) Figure Small wind turbines: fully installed costs Price ($/kW) 8000 6000 4000 2000 0.1 10 100 1000 10 000 Turbine rating (kW) Figure Wind turbine prices, 1999 information to be updated The original turbine prices were in ECU, and in 1999, ECU was roughly equal to $1.04 Figure – for installed costs – and Figure – for turbine prices only – both clearly illustrate that economies of scale are significant 2.15.4.4.1 Offshore wind Offshore wind has the potential to deliver substantial quantities of energy, but it is more expensive than onshore wind Cost reductions may be expected as the technology is further developed A number of factors combine to increase the cost of offshore wind farms above onshore costs: • The need to ‘marinize’ the wind turbines, to protect them from the corrosive influence of salt spray These measures may add up to 20% to turbine costs • The cost of the cable connection from the wind farm to the shore; this increases with the distance from the shore, and accounts for between 17% and 34% of the total cost • More expensive foundations The cost increases with water depth and can account for up to 30% of the total cost • Increased operation and maintenance costs, with a risk of lower availability due to reduced access to the wind turbines during bad weather Foundation and grid connection costs are substantially more expensive when compared with onshore wind energy In the budgets for the first batch of large Danish offshore wind farms, from which Figure is drawn, these items together account for around a third of the total cost (Onshore foundations are typically less than half this amount, whilst grid connection costs are frequently even lower.) This observation underscores the reasons for the interest in larger wind turbines and also for the enthusiasm for larger numbers of machines 2.15.4.5 Historical Price Trends Table summarizes the principal operational data for the early Danish wind farms at Vindeby and Tuno Knob [13], together with the pilot Dutch farm in the Ijsselmeer [14], and a later Swedish wind farm Although the cost of the Vindeby wind farm was 85% higher than the cost of an onshore installation, the anticipated energy yield was 20% higher Concerns about low availability offshore – due to problems of access – have not generally been realized The costs of the early wind farms were significantly higher than those of the later installation, at Bockstigen in Sweden Just as with onshore wind farms, prices vary depending on the exact location, with distance from shore, seabed conditions, and water depth being key factors Another key determinant in offshore wind farm costs is likely to be the number of machines There is a trend toward larger wind farms than onshore, to spread the cost of offshore transport, cable connection, and operation and maintenance costs Wind Energy Economics Table 11 487 Summary of support mechanisms for renewable energy Location Initial tariff ($ MWh−1) Periods, initial/total (yrs) Onshore China Denmark France Germany Ireland Netherlands 80–95 +54 120 92 94 +$100 kW−1 10/20 5/20 15 15 Ontario South Africa United Kingdom United States 125 158 +90 (variable) +21 20 20 20 10 Offshore Denmark France Germany 100–125 195 220 10/20 12/20 Ireland Ontario 208 175 15 20 Second stage tariff ($ MWh−1) 40–120 73 Market price 45–195 51 Notes, source Depends on wind speed For 22 000 full-load hours Second stage tariff depends on wind speed [32] Initial period depends on site wind IEA Rate MWh−1 depends on wind speed Target is ∼$145 MWh−1 Ontario Power Authority Reference [33] Average total price ∼ $155 MWh−1 in 2008/09 For 50 000 full-load hours Second stage tariff depends on wind speed [34] Initial period depends on water depth/distance to shore IEA Ontario Power Authority his/her house and export the surplus electricity to the local electricity company, they would resell it at around $100 MWh−1 (depending on the tariff) to neighbors There would be no backflow through the local 33/11 kV transformer and, as a bonus, the locality might suffer fewer power cuts This reasoning may be simplistic, but it is, nevertheless, the principle of ‘net metering’, which is used by several American utilities to remunerate renewable energy generators A modified version of the approach has been used in Denmark and Germany, for example, used to pay renewable generators 80–90% of the domestic tariff Although studies in America and elsewhere point to the fact that renewable energy may have a high value in certain locations, this point is not widely accepted by utilities, many of whom continue to assign renewable energy a value based on its fuel saving capability In many instances, renewable electricity-generating technologies deliver energy closer to consumer demand than centralized generation It substitutes for electricity that has accrued a higher selling price than the generation costs of large thermal plant, due to its passage through the network This rising level of ‘value’ with reducing voltage is reflected in higher charges for customers connected at lower voltage levels In view of complexity of the issues involved and the fact that ‘embedded generation benefits’ are dependent on location, it is difficult to quote definitive values Broadly speaking, however, the embedded benefits rarely exceed about $24 MWh−1 in the United Kingdom and are often much lower The converse problem, where substantial quantities of renewable generation may trigger reinforcements, result in a negative benefit In the United Kingdom, for example, the need for significant (and costly) transmission reinforcements has been addressed by a Transmission Issues Working Group Their report examined the implications of installing up to 6000 MW of renewable generation in Scotland and concluded that reinforcement costs would be roughly $800 million per 2000 MW of generation 2.15.8 Variability Costs Variability costs acknowledge the fact that wind energy has a variable output that is not totally predictable System Operators, as a result, incur extra costs as they need to deal with the extra uncertainty when balancing supply and demand Those extra costs are, however, modest They need to be considered carefully, as some critics suggest that the addition of these costs destroys the economic viability of wind energy This analysis draws on material that examines the issues in some detail [36] The variations from distributed wind are generally less than the demand fluctuations regularly encountered on electricity networks To cope with these, every network has reserves scheduled at all times and the key issue is the additional reserve that is needed to cope with the variability of wind That is only a few percent of the rated power of wind plant There are also concerns that a system with a high proportion of variable renewables would risk power cuts at times of peak demand The ability of wind energy to contribute to these peak demands needs to be examined This introduces the concept of ‘capacity credit’ – how much thermal plant can be retired with the introduction of wind energy? The aim of this section is to clarify the issues in more detail, drawing upon the analysis that has been carried out during the past 30 years (one recent review identified over 150 references [37]) 488 Wind Energy Economics This section is structured as follows: • A brief description of electricity network issues comes first as this is essential background to discussions of issues surrounding the integration of variable renewable energy sources • Next comes an examination of exactly what is meant by ‘wind variability’ (‘Variable’ is a better description of the power fluctuations from wind, wave, etc., than ‘intermittent’.) • The next section deals with the integration of the variable sources into power systems, the costs, and other issues 2.15.8.1 2.15.8.1.1 Electricity Networks Economies of scale Although large integrated electricity systems are efficient, they still require ‘operational reserve’ to deal with short-term mismatches in supply and demand and a ‘plant margin’ (additional plant, over and above that required to meet the maximum demand) to deal with plant breakdowns and other outages The effect of wind energy on the short-term reserves and margins is the subject of much discussion and each issue is discussed in the context of electricity networks and the additional costs that may be incurred 2.15.8.1.1(i) Demand fluctuations Although aggregation smooth variations in consumer demand, there are still substantial fluctuations when numbers of consumers together increase power needs – at the morning and evening rush hours, for example Figure 14 illustrates the demand variations on the network in Great Britain on January 2009 From a nighttime minimum of just over 40 000 MW at 05.00 h, demand rose rapidly to just under 54 000 MW at 10.00 h It then fell off slightly until there was another surge in demand at the evening peak of just under 60 000 MW, reached at 17.30 h Demand then fell off steadily The intra-half hourly changes in demand are shown in Figure 15 During the morning peak, the maximum change between two successive half hours was 2300 MW and during the evening peak was 3100 MW Negative changes in demand were recorded from 18.00 h onwards, reaching a maximum of 2600 MW at 22.30 h 2.15.8.1.1(ii) Operational reserve All power systems, with or without wind energy, need short-term operational reserve, often called ‘spinning reserve’, to deal with demand fluctuations The terms cover various types of plant with different characteristics, outlined briefly below Nomenclature varies between utilities, but the exact details are not central to this discussion The principal types of plant are: • Frequency response: Such plant automatically adjusts its output, increasing it when system frequency is low and vice versa • Fast reserve: This plant is able to increase or decrease its output, under instruction from the System Operator on a short timescale (typically 30 min) • Standing reserve: Similar to fast reserve, but on a longer timescale (typically 1–4 h) System demand (MW) 60 000 55 000 50 000 45 000 40 000 10 20 30 40 50 40 50 Half-hour period Figure 14 Demand variations on the British electricity network System demand changes (MW) 3000 2000 1000 −1000 −2000 −3000 10 20 30 Half-hour period Figure 15 Half-hourly demand changes on the British electricity network Wind Energy Economics 489 In most utilities, the operational reserves are generally part-loaded thermal or hydro units They operate at below maximum capacity, so that the output can be increased or decreased to cater for mismatches between generation and demand Pumped-storage plant is also used, as it can respond very rapidly to a need for more generation 2.15.8.1.1(iii) Costs of reserve The costs of reserve reflect the fact that they need to operate at part-load (and lower efficiency) Annual costs of frequency response plant in Great Britain are around $230 million and each of the other types was expected to cost around $100 million in 2008/09 [38] The costs of reserve depend on the precise type and price structure International values are in the range $6–12 MWh−1, although some fast-response plant, such as pumped storage, secures higher values These costs compensate the plant owners for the lower efficiency of plant whose output is below its maximum, extra wear and tear, and possibly extra controls; they are additional to payments for the energy actually generated 2.15.8.1.1(iv) Plant margin or backup Reserves ensure minute-by-minute system security, but longer-term security is managed by making sure that there is always enough plant available to meet the highest demands on the electricity network The ‘Plant margin’ is a measure of the difference between the total capacity of the plant on the network and the expected maximum demand The desirable plant margin (plant capacity minus maximum demand) for a large system is around 24% [39] This does not guarantee that the lights will never go out, but ensures this will happen very rarely The high level of security and low plant margin stems from the fact that a large system has a number of generating units each with a quantifiable probability of failure, but the combined probability of, say, units failing at the same time, is much less It may be noted that the plant margin in many regions is higher than the theoretical figure It is very difficult to design a market that delivers the theoretically desirable optimum, simply because power plant equipment takes a long time to build, and plant closures are not always easy to predict in advance 2.15.8.2 Characteristics of Wind Energy An analysis of the wind power fluctuations in Western Denmark in 2007 [40] suggests that for 42% of the year (37.00 h), the intra-hourly fluctuations were within the range Ỉ25 MW (1% of the wind capacity) Extending the range to Ỉ50 MW captures another 18.00 h of fluctuations At the extremes, fluctuations in excess of Ỉ375 MW (16% of capacity) only occurred 10 times in the year The complete histogram of power swings is shown in Figure 16 The standard deviation of the fluctuations is around 3% Information of this kind provides a basis for estimating the effects of integrating wind energy into an electricity network 2.15.8.3 Assimilating Wind When considering the introduction of the variable renewable energy sources, it is important to preserve the advantages of an integrated electricity network as that minimizes the extra costs to electricity consumers The UK’s National Grid has made this point [41]: However, based on recent analysis of the incidence and variation of wind speed we have found that the expected intermittency of wind does not pose such a major problem for stability and we are confident that this can be adequately managed … It is a property of the interconnected transmission system that individual and local independent fluctuations in output are diversified and averaged out across the system The effects of adding wind to an electricity network may be illustrated by the case of Western Denmark, and examining the changes in demand that need to be managed If there had been no wind installed there in 2007, the maximum h increase in system demand would have been 675 MW With 26% wind (the amount on the system that year), sometimes the wind fluctuations Frequency of swings (h) 10 000 1000 100 10 –400 –200 200 400 1-hour power swing (MW) Figure 16 Intra-hourly power swings observed in Western Denmark in the year 2007 The maximum swing never exceeded around 18% of the installed capacity of the wind plant (up or down) and the standard deviation was about 3% 490 Wind Energy Economics augmented the demand fluctuations, and sometimes they reduced them The maximum increase in demand that the System Operator had to deal with went up from 675 to 900 MW In that hour, there was an increase in demand at the same time as the output from the wind plant fell However, the number of times that the net demand increased by more than 600 MW in an hour only went up from 55 (consumer demands only) to 63 (consumer demand net of wind production) Even with 39% wind (scaling up 2007 wind outputs by 50%), there would only be about 75 occasions when the net increase in demand exceeded 600 MW The impacts of variability and the corresponding possibilities for mitigation can be quantified by examining the three principal ‘cost centers’: • the costs of extra operational reserves (balancing costs), which can be reduced by ○ reducing the ‘unpredictability’ of wind (by better forecasting), or ○ reducing the cost of balancing services • ‘backup’, which can be reduced if the capacity credit can be increased, and, • ‘constraint costs’, due to surplus wind, which can be reduced if the surpluses can be reduced 2.15.8.4 Extra Short-Term Reserve Needs and Costs Electricity networks with wind energy need extra reserves to deal with the extra uncertainty associated with the presence of wind on the network It is important to emphasize that this extra uncertainty is not equal to the uncertainty of the wind generation, but to the combined uncertainty of wind, demand, and thermal generation The combined uncertainty is determined from a ‘sum of squares’ calculation that provides the basis for estimates of additional reserve needs: totalị ẳ demandị ỵ windị where is the standard deviation of the uncertainties Although the quantity of operational reserve rises with increasing amounts of wind energy, the UK’s National Grid (NGT) is confident that it will be able to procure the necessary amounts, and, moreover, that there is no ‘ceiling’ on the amount of wind-generated electricity that can be accommodated [42]: Based on recent analysis of the incidence and variation of wind speed the expected intermittency of the national wind portfolio would not appear to pose a technical ceiling on the amount of wind generation that may be accommodated and adequately managed It may be noted that system operators such as NGT says nothing about the type of plant that may be needed for reserve – that is left for the market to decide, provided it can meet the technical specifications set by operator In practice, it may be coal-fired plant, combined cycle gas turbines (CCGTs), hydro, or storage The former tends to be the most economic option, while the latter tends to be the most expensive However, pumped-storage plant can respond extremely rapidly and so is well suited to a particular type of ‘fast reserve’ The characteristics of most electricity systems are similar, so estimates of the extra reserve needed to cope with wind energy are also similar With wind supplying 10% of the electricity, estimates of the additional reserve capacity are in the range 3–6% of the rated capacity of wind plant With 20% wind, the range is 4–8%, approximately The UK’s National Grid has recently quoted estimates of the extra balancing costs for wind in the United Kingdom for 40% wind [43] These would increase balancing costs in 2020 by £500–1000 million per annum (£3.5–7 MWh−1 of wind, or $5.6–11.2 MWh−1) The uncertainty arises partly because the future trajectory of balancing services costs is uncertain (they are dependent on fuel prices), partly because increased use of the demand-side management (DSM) could reduce the overall costs The way in which balancing costs are likely to increase with wind penetration level is illustrated in Figure 17 This makes use of the National Grid data as ‘anchor points’ and uses information on demand and wind uncertainty (discussed earlier) to synthesize the rest of the curve So 10% wind energy is likely to incur extra costs in the range $ 4–8 MWh−1 and 20% wind energy in the range $5–10 MWh−1, approximately A recent American review [44] quotes a study that looked at 30% penetration on a peak load basis (probably about 15% on an energy basis) and that suggested an extra balancing cost of $8.84 MWh−1 – within the range of the British data Cost of extra balancing ($/MWh of wind) 10 High 0 10 20 30 Wind Energy penetration (%) Figure 17 Estimates of additional balancing costs for Great Britain Low 40 Wind Energy Economics 2.15.8.5 491 Carbon Dioxide Savings As the extra reserve operates at part-load, its lower thermal efficiency means that its emissions increase The reserve still contributes useful energy to the system, so the extra emissions are those associated with the reduced efficiency of part-loaded plant Taking a pessimistic estimate of 10% for the reduced efficiency, and taking into account the fact that the load factor of wind plant is just under half that of thermal plant, Dale et al [45] suggested that this reduces the emission savings from the wind, at the 20% penetration level, by a little over 1% In other words, if the displaced fuel is coal, for the sake of argument, with CO2 emissions of 900 kg MWh−1, then the effective CO2 saving would be around 890 kg MWh−1 If the displaced fuel were gas, with CO2 emissions of 400 kg MWh−1, the effective saving would be 395 kg MWh−1 At higher penetration levels (40%, say), the nonlinear increase in the necessary reserves brings these figures down to around 875 and 388 kg MWh−1, respectively 2.15.8.6 Extra Backup and Its Costs A distinction must be drawn between the extra reserves needed for short-term balancing in an electricity system with wind and the extra backup (if any) needed to guarantee the security of the system at all times That means making sure that there is always enough power available to meet the peak demands of the system Although it is suggested that there is a need to provide ‘backup for windless days’ to ensure that demands are always met [46], this is misleading on two counts: • When a new thermal power station is built, there is no discussion as to how the electricity system will manage when the station is unexpectedly out of action for emergency repairs during the winter The ‘plant margin’ is a common pool of ‘extra’ plant that ensures peak demands are met No power stations are 100% reliable, as discussed earlier • Not even the most zealous of renewable energy enthusiasts would suggest that System Operators should rely on the full rated capacity of wind power plant When wind is added to an electricity network, the situation is not fundamentally different from an addition of thermal plant If the wind plant has some ‘capacity credit’ (discussed next), then it will be possible to retire some of the older plant, without compromising system security If the new plant has zero capacity credit, then no plant can be retired, but, either way, no new plant needs to be built for ‘backup’ – it already exists Estimates of capacity credit that are based on wind electricity production during a single winter are unlikely to provide accurate estimates It is a statistical quantity that requires adequate data – as for nuclear plant During the winter of 2008/09, for example, at the time of peak demand, the metered wind electricity production in Great Britain was about 18% of its rated output However, about 5000 MW of nuclear output was not available, for various reasons – nearly 50% of the total [47] It would be misleading to assign a capacity credit of 18% to wind on the basis of this one instance, and equally misleading to assign a ‘firm power’ contribution from nuclear as 50% of its rated output It is important to emphasize that the capacity credit of wind will never be greater than the plant margin, and even if Britain had 26 000 MW of wind and had been completely becalmed at the time of the peak demand on January 2009, the plant margin would not have been used up, despite the missing 5000 MW of nuclear It is also worth reiterating that the plant margin is generally greater than the theoretically desirable minimum 2.15.8.7 Capacity Credit The term ‘capacity credit’ for wind, introduced above, tends to be controversial It may be defined [48] as follows: The reduction, due to the introduction of wind energy conversion systems, in the capacity of conventional plant needed to provide reliable supplies of electricity Despite the controversy, numerous studies have confirmed that wind can substitute for thermal plant and enable power systems to operate with the same level of reliability Figure 18 shows how the capacity credit varies as a function of the installed wind energy penetration level Capacity credit/rated power 0.4 DK F 0.3 GR IR IT NL P UK 0.2 0.1 0 10 Wind energy contribution (%) Figure 18 Capacity credits as a function of wind energy contribution 15 492 Wind Energy Economics Comparisons of capacity credits across national boundaries must be made with care In northern Europe, where peak demands on most electricity systems occur around 18.00 h during the winter months [49], capacity credits for wind energy at low penetrations are mostly close to ‘winter quarter’ capacity factors There are differences in the numerical values, however, due to the differing wind speeds All the results show a similar trend, with capacity credits declining by around 40% between 3% wind energy and 15% Data from the Irish system operator [50] have yielded similar results to those from British studies, when normalized to take account of differing assumptions about capacity factor [51] There are a number of references to capacity credits in the American literature Winds in many inland sites are, however, driven by local thermal effects – particularly in the Californian passes As a result, wind speeds are more predictable but often have a pronounced diurnal trend Capacity credits therefore often depend on the coincidence (or lack of it) between the peak winds and peak demands Some installations claim high capacity credits, whereas others can only claim low values There are therefore only limited lessons to be drawn from American experience, although the procedures used are similar to those discussed in this chapter 2.15.8.8 The Cost of Backup Although the ‘extra costs of backup’ are not derived by assuming the whole of the wind plant capacity needs to be duplicated by standby thermal plant, there are extra costs associated with the low capacity credit of wind at penetration levels above about 8% With thermal plant, the average power and the ‘capacity credit’ are the same, but wind energy is different In Great Britain, above the 8% penetration level (approximately), the capacity credit of wind is less than its capacity factor This means that 26 000 MW of wind (roughly 20% energy penetration) delivers electricity that corresponds to around 10 700 MW of thermal plant (assuming a wind capacity factor of 35% and a thermal plant load factor of 85%), but only displaces around 5000 MW of thermal plant This has the effect of reducing the load factor on the remaining thermal plant Their generation costs increase, as capital cost repayments are spread over fewer kilowatt-hours This provides a basis for estimating the ‘additional costs of backup’, using the methodology used by Dale et al [45] Using an up-to-date price for CCGT plant of $960 kW−1, these amount to around $4 MWh−1 of wind (at 20% penetration), rising to around $10 MWh−1 of wind at 40% penetration Even where capacity credits are much lower than has been assumed, the effect on the variability costs would be modest At 20% wind energy penetration level, for example, the additional variability cost would be about $2.8 MWh−1 [47] 2.15.8.9 Transmission Constraints The foregoing discussion has implicitly assumed that the electricity network can be operated as a single unit, with unrestricted flows of energy In practice, this is not always the case and there are sometimes occasions when the power production from renewable plant exceeds the transmission capacity that is required to deliver it to the demand centers This means that there may be occasions when renewable plant may be required to cease generation, or be ‘constrained off’ The effect of such constraints will be to increase the costs of the renewable plant, as the capital costs will be spread over fewer units of electricity than was anticipated Whether or not these costs are borne by the renewable generator depends on the structure of the market as designed by the regulator and government In Britain, such constraints are likely to occur due to the large quantities of wind energy – installed or planned – in the north of England and Scotland For many years, there have been large North to South power flows, as generation capacity exceeds demand in the north, and vice versa There is increasing concern over the cost of these constraints [52], although these can be alleviated by additional transmission connections at an estimated cost of £4700 million [53] 2.15.8.10 Wind Surpluses at High Penetration Levels The discussion so far has focused on wind energy penetration levels up to around 40% In practice, as noted earlier, higher levels are achievable, albeit at increased cost A detailed analysis by the Danish system operator, Energinet, has examined the implications of operating with 100% wind and quantified the additional costs [54] The analysis was carried out for Western Denmark, but ignored the existence of the connections to Germany, Sweden, and Norway and did not assume that any storage was available Of course, 100% wind is not feasible, but the System Operator assumed that sufficient wind power capacity was installed in order to meet 100% of the electricity requirements With that level of capacity, around 30% of the wind energy had to be rejected when wind power production exceeded system demand A similar amount of wind had to be supplied from thermal sources of generation when the system demand exceeded the wind power production The possibility that wind power production may occasionally exceed system demand first occurs at penetration levels around 25% Figure 19, which uses actual data from Western Denmark, suggests that it occurred twice during 13/14 January 2007 However, the amounts of ‘surplus’ wind energy are initially modest and similar estimates come from the Energinet study and from a British study by consultants Sinclair Knight Merz (SKM) [55] With 30% wind energy, the Danish study suggested that around 1.7% would need to be constrained off or rejected and the SKM study a slightly lower level – although the precise value depended on assumptions about interconnectors and pumped storage With 40% wind, both studies projected about 4% would need to be rejected and at 50% wind about 7.5% Data from the two studies are compared in Figure 20 Wind Energy Economics 493 Power demand and output from wind plant (GW) Demand Wind 0 10 20 30 40 Time (hours), from 00:00, 13 January 2007 Figure 19 Demand and wind production in Western Denmark, 13/14 January 2007 Wind energy surplus (%) 15 Energinet SKM 10 20 30 40 50 60 Nominal wind energy contribution (%) Figure 20 Surplus wind: estimates for Denmark and Britain If no market can be found for this ‘surplus’ wind energy, a small penalty is attached to this ‘lost’ electricity, as the fixed costs of the wind plant are spread over fewer units of electricity With 30% wind, this amounts to around $1 MWh−1 of wind, rising to around $2.4 MWh−1 with 40% wind, based on current installed prices of around $2100 kW−1 Ways in which this ‘surplus’ may be utilized are discussed later 2.15.8.11 Total Costs of Variability The total costs of variability to the electricity consumer – defined as additional balancing costs plus backup costs plus constraint costs – are shown in Figure 21 The ‘high’ estimate uses National Grid’s upper balancing cost estimate and an installed cost for CCGT plant of $1000 kW−1 in the calculation of backup costs The ‘low’ estimate uses National Grid’s lower balancing cost and an installed cost of $800 kW−1 for a mixture of CCGT and open cycle gas turbine plant To derive the constraint costs at penetration levels above 30%, it has been assumed that 12 GW of onshore wind costing $2150 kW−1 has been installed and 45 GW of offshore wind costing $3200 kW−1 With 10% wind energy, the extra costs are below $2 MWh−1 in each case; at the 20% level they rise to a little over $3 MWh−1 in the ‘high’ case ($2.4 MWh−1 in the ‘low’ case) and with 40% wind, the estimates are $12 MWh−1 and $8.6 MWh−1, respectively A ‘central’ figure would add about 5.5% to domestic electricity bills Extra cost to consumer ($/MWh) 12 10 High Low 0 10 20 30 Wind energy contribution (%) Figure 21 Variability costs estimates 40 494 Wind Energy Economics 2.15.8.12 Mitigating the Effects and Costs of Variability Progress toward high levels of wind penetration will inevitably be gradual and so it is likely that technologies and strategies will develop that will mitigate the impacts of variability Some of these are already in use, such as improved methods of wind forecasting and this is likely to have a significant impact Numerous other ideas are under development or discussion and it is likely that increased use of DSM (possibly aided by the installation of ‘smart meters’) will also play a key role in reducing variability costs The use of storage is often advocated, but its use for ‘levelling the output’ of wind power may be difficult to realize However, it may have a role to play – for the benefit of the electricity network as a whole – in systems with a high penetration of variable renewables The construction of more international transmission links should aid the assimilation of variable renewables – and will also work to the benefit of the system as a whole Further ahead, the widespread introduction of electric cars or a switch to electric heating would be beneficial to wind energy, as it would facilitate the absorption of ‘surplus’ wind at times of high wind and low demand 2.15.8.12.1 Wind forecasting There is considerable work in progress on improvements in wind forecasting and the emergence of forecasting services, in both Europe and America, testifies to the fact that it is worthwhile improving the accuracy of projected power outputs A large EU-funded R&D project involved a large number of contributors [56], and a utility-funded project has been completed in the United States, managed by the Electric Power Research Institute [57] Commercial forecasting services are also available, with software that improves forecasts up to 24 h ahead One company claims, for example, that the error in h ahead forecasts is typically 15–25% lower than that of persistence forecasts [58] Much of this forecasting work is focused on providing services to operators of wind farms, rather than system operators, but estimates are available of countrywide improvements that might be expected [59] The analysis, for Germany, suggests that the standard deviation of the uncertainty h ahead might be reduced from 3% to below 2% and, similarly, the h uncertainty can be reduced from 6% to around 4% This would reduce the estimates of balancing costs that were quoted earlier by about 30%, provided the system operator felt that the forecasting accuracy was sufficiently robust 2.15.8.12.2 Demand management DSM has been an integral part of the load management strategies of system operators for many years It has the potential both to reduce peak loads and lower the costs of reserve The variable sources of renewable energy, such as wind, are likely to benefit from the wider adoption of load management in the future, although there are potential benefits for the electricity system as a whole There are various types of DSM, although the boundaries are not sharp: • Provision of reserve and frequency response under contract to the system operator In the United Kingdom, in 2005/06, users, rather than generators provided about one-third of the ‘standing reserve’ and ‘frequency response’ balancing service requirements of National Grid [60] • Teleswitching This technology has been slowly developing for around 40 years It enables demands to be modulated in response to radio signals sent by the System Operator The signals are sent to a special meter and the technology has the potential to control or modulate interruptible demands • Dynamic demand While teleswitching is actively (and remotely) controlled, dynamic demand is a passive system that relies on sensors in equipment used by consumers to modulate demands If all domestic refrigerators in the United Kingdom, for example, included a frequency-sensitive device that inhibited its operation when the frequency fell below (say) 49.8 Hz and switched the fridge on (provided it was not already too cold) at, say 50.2 Hz, then this could substitute for between 728 and 1174 MW of frequency-response plant [61] • Smart meters These give electricity consumers access to information about the price of their electricity on a continuous basis The most common perception of such meters at present is to provide information to consumers, rather than intervene to restrict demand on a selective basis • ‘Time of use’ tariffs have been in existence for many years in the industrial and commercial sectors, and, in simplified form, in the domestic sector The tariffs aim to discourage use at peak demand times – that normally coincide with peak prices – and so iron out, to some extent, demand fluctuations If that enables the quantity of rarely used (and expensive) ‘peaking plant’ to be reduced, that reduces both costs and emissions The most sophisticated development of ‘time of use’ pricing responds continuously to changes in market prices Whether or not consumers can react to high or low prices depends on the type of electrical equipment they are using ‘Time of use’ tariffs are common in France and their effectiveness appears to be reflected in a lower ratio between maximum and minimum demands On 21 May 2009, for example, the French ratio was 1.39, whereas in Britain it was about 1.6 Wind Energy Economics 495 All these concepts act to improve the efficiency of the electricity system as a whole Any benefits to wind would come through reductions in the costs of balancing services Some benefits are already being realized (first bullet point) but it is difficult to estimate the magnitude of any additional benefits One possible downside (from the point of view of wind) is that a reduction in the uncertainties in the supply/demand balance might mean that the uncertainties in wind power production would become more significant, thus increasing additional balancing costs 2.15.8.12.3 Energy storage 2.15.8.12.3(i) Dedicated storage Energy storage is often seen as a means of ‘levelling the output’ variable renewables, possibly increasing its capacity credit and so increasing its value Such ‘dedicated storage’ faces a number of challenges as it adds to the generation cost of the variable renewable That additional cost needs to be less than the additional value of ‘firm power’ over variable power There are additional challenges with dedicated storage, as the store needs to be large to ensure that the ‘leveling’ continues during long periods of low wind An early integration study concluded [62]: There is no operational necessity in associating storage plant with wind-power generation, up to a wind output capacity of at least 20% of system peak demand This does not imply that 20% is a ceiling, or threshold It was simply the upper limit that was investigated in the study A later American study made the same point [63]: Storage may increase the value of intermittent generation However, studies generally show that dedicated storage systems for renewables are not viable options for utilities because of added capital costs of current storage technologies Storage can add flexibility and value to utility operations, but it should generally be a system-wide consideration, based on the merit of the storage system More recently, the American Electric Power Institute suggested [64]: Installing energy storage … practically eliminates wind integration issues … Unfortunately the high cost of storage systems limits the situations in which they are useful In Britain, the cost threshold that storage would need to meet for viability in the market can be gauged from the difference between ‘continuous’ and ‘variable’ power sources The difference in the prices realized for landfill gas (firm power) and for wind energy (variable power) in the auctions conducted by the Non-Fossil Purchasing Agency is a guide Between the summer of 2006 and the end of 2008, the minimum price difference was $1.8 MWh−1, the maximum $18.5 MWh−1, and the average $8 MWh−1 The average roughly corresponds to the theoretical ‘capacity value’, based on the replacement cost of CCGT plant 2.15.8.12.4 System storage There is a difference between ‘dedicated storage’ for variable renewables and storage for an electricity system – with or without variable renewables Storage has the potential to enable power systems to operate more efficiently – absorbing power at periods of low demand and releasing it at periods of high demand that otherwise needs to be met by expensive generating plant This is generally a less challenging role than ‘dedicated storage’ However, storage has a generation cost, just like any other generation technology, and will only be economic if the differential between the energy prices paid to the storage operator at times of high demand, and by the operator for electricity at times of low demand, is sufficient to cover its costs A recent UK Select Committee observed [65]: 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 A recent analysis that examined the prospects for Western Denmark concluded [66]: The conclusion is that energy storage systems are for most cases uneconomical for day-to-day trading in Western Denmark If the introduction of large quantities of wind energy into electricity networks widens the difference between ‘high’ and ‘low’ spot prices in the electricity market, that may enable the construction of cost-effective storage Most of the technologies are able to provide system services (short-term operating reserve, reactive power, and ‘black start’ capability) and these can provide additional revenue There are a number of technologies in existence and under development, with a wide range of applications, apart from those discussed here [67] 496 Wind Energy Economics 2.15.8.12.4(i) Additional international connections Additional international connections give system operators access to more sources of power, effectively increasing the size of the system The advantages of large systems were discussed earlier and perhaps the simplest way of looking at the effects of additional connections is to view them as additional plant The ‘effective’ renewable energy penetration level then drops The next stage in the argument for additional interconnections is that, with the two interconnected systems operated as one, the wind variability comes down However, the evidence on this point is not clear On the one hand, the standard deviation of hourly wind fluctuations in Britain is similar to that in (smaller) Western Denmark On the other hand, Foley et al [68] combined British and Irish wind records and showed that the joint occurrence of wind speeds below m s−1 was reduced to 16% of the time, whereas the individual occurrences were 22% (Britain) and 28% (Ireland) 2.15.8.12.5 European supergrids A number of proposals for more extensive international grid connections – mostly using high voltage direct current (HVDC) – have been put forward in recent years, mostly with some or all of the following objectives: • Facilitating the connection of offshore wind farms • Smoothing wind fluctuations on a continental (rather than national) scale • Facilitating progress toward very high proportions of renewable energy in the European network including, for example, concentrated solar power plant in Africa Hurley et al [69] provided data that illustrated the smoothing effects They took data from 60 well-distributed sites over a 34-year period and showed, for example, that there were very few occasions when wind power production fell below 12% of rated output in the winter This suggests that the capacity credit of this widely distributed wind might be higher than the values calculated from individual national studies Similarly, their analysis of power swings also suggests that there would be a lower uncertainty, probably leading to lower additional balancing costs Decker et al [70] have recently summarized most of the proposals that are currently being discussed Some cater, in the longer term, for up to 100 GW of offshore wind They note that increased interconnections would be beneficial to European electricity networks as a whole, quite apart from their role in facilitating the connection and smoothing of renewable energy They also provide data on the smoothing effects, suggesting that hourly variations in excess of plus or minus 10% appear to be negligible (This may be compared with the maximum hourly swings around 14–18% observed in Western Denmark and estimated for Britain.) There are, nevertheless, difficulties on the way to the construction of a European Supergrid Several witnesses who appeared before the UK’s Energy and Climate Change Select Committee in April/May 2009 drew attention to technical and regulatory difficulties, although there was support for further analysis of the concept Another possible difficulty is that national plans for the connection of offshore wind farms are already well advanced and so the possibility of looking at ‘the big picture’ may already have passed Nevertheless, the broad concept has support at the European level, through the ‘Ten E’ programme of support for improved interconnections [71] 2.15.8.12.6 Electric cars The prospects for electric cars are being studied with a view to reducing greenhouse gas emissions There is a double benefit as air quality in cities will be improved and national emissions will be reduced, even though the electricity used to charge them at present comes from a mixture that includes lots of coal and gas-fired plant The attraction from the standpoint of the electricity industry is that it may enable the more efficient use of generating plant, provided most of the charging takes place during the night In the United Kingdom, between 22.30 and 06.30 h in the winter, the demand drops below 45 000 MW, whereas during the rest of the day, it is above this level, peaking at just under 60 000 MW Inspection of data for later in the year (21 May 2009) suggests that there was a similar difference between the nighttime demand and the peak A margin of around 15 000 MW would be sufficient to charge the entire fleet of British cars if all were to be powered electrically [36] The additional attraction, from the standpoint of integrating wind, is that this would enable surplus wind power to be utilized at times when wind power generation exceeded demand There would be no guarantee, of course, that surplus wind would always be generated during the night, but that is a realistic scenario, given the lower demand Daytime charging would be possible, provided it could be controlled by some form of ‘smart meter’, or by teleswitching Although much of the discussion surrounding electric cars has focused on their use in high-wind scenarios, it is likely that they might be to provide reserve services at modest cost As noted in the previous paragraph, control of the magnitude of the charging load is a strong possibility, and at modest cost 2.15.8.12.7 ‘Smart grids’ and the growth of decentralized generation The term ‘Smart Grids’ is used to describe various technologies that may need to be developed in the future to enable electricity networks to function more efficiently – especially with large amounts of wind energy A paper [72] that was commissioned by the UK Government Office for Science has discussed the issues surrounding the concept The paper suggests that “Fully decentralised energy supply is not currently possible or even truly desirable,” but that Wind Energy Economics 497 current evidence points towards the deployment of an increasingly decentralised energy supply infrastructure, which will still rely on and benefit from common centralised infrastructures The rate at which decentralized generation will grow is somewhat uncertain [73], although both PB Power [74] and National Grid UK [75] appear to be in broad agreement that the additional plant capacity in the United Kingdom will be in the range 3–5 GW by 2020 The present capacity of embedded generation is around 7500 MW 2.15.8.12.8 Electric spaces and water heating At present, electric space and water heating is significantly more expensive than gas or oil-fired heating If, however, there were a move toward more electric heating, this potentially would provide system operators with a large source of inexpensive DSM This would enable wind surpluses to be utilized and avoid the necessity for wind turbine outputs to be constrained Such a shift in emphasis might be the result of high fossil fuel prices, government incentives, or both Electric water heating, and, to a lesser extent, space heating can be modulated without significantly affecting the comfort of the consumer, although wider dead bands may be necessary for central heating systems The signals that would enable system operators to influence demand levels could be transmitted through ‘smart controls’ or one of the other technologies discussed earlier The concept is actively being considered in Denmark [76] 2.15.8.12.9 Overall effects Precise estimates for impacts of the mitigation measures cannot easily be made in every instance The prospects for improved methods of wind forecasting are good and the most significant impacts would be at the higher levels of wind energy penetration, where it may be possible to reduce the extra costs of variability by about 20–25% 2.15.9 Total Cost Estimates A number of governments have set targets for renewables, or for wind energy of achieving 20% renewable energy (or more) in electricity supply by 2020 This provides a reference point for estimating the total additional cost to the electricity consumer, taking into account all the factors discussed in this chapter An analysis completed in 2003 [45] suggested that the extra cost would increase domestic electricity bills in the United Kingdom by about 5% As wind energy is expected to contribute the majority of the renewable energy target, the analysis assumed that all the renewable energy would come from wind and made projections about future trends in wind energy costs Similar analyses were later made elsewhere, such as the Republic of Ireland, with broadly similar conclusions [77] The Irish study suggested that electricity consumers might realize savings from about 2010 onward when the additional costs associated with extra balancing were outweighed by the lower costs of wind compared to gas In all such studies, however, it is essential to examine the assumptions carefully As the assumptions about gas prices made in the 2003 UK study were too low, a later analysis quantified ‘break-even’ criteria, when the addition of wind would bring about no increase in electricity prices [78] If the price of gas is €1 per therm, wind can cost up to €1350 kW−1, but if gas falls back to €0.5 per therm, wind needs to cost less than €820 kW−1 The extra costs of switching from fossil fuels to renewables have also been quantified elsewhere – in Europe and in America – and they are also extremely modest In the United States, the effect of a 15% ‘Renewables Portfolio Standard’ was projected to increase the national average electricity price by 2%, compared with the reference case ($82 MWh−1 with the RPS compared to $81 MWh−1 in the reference case) [79] By 2030, prices for natural gas and coal, two key fuels for the electric power sector, are lower with the RPS than in the reference case These reductions come about because of the reduced demand for the fossil fuels In a more ambitious American scenario, it was suggested that 300 GW of wind could generate 1200 TWh of electricity at a cost of $61 MWh−1, or less [80], and made allowances for transmission constraints The study concluded that the extra cost on consumer electricity bills would be around $0.6 MWh−1, or around 1% The extra costs that are coming out of these American studies are very similar to those emerging from Europe One recent European study [81] suggested that the costs of meeting the 20% renewable energy target (this covers all energy, not just electricity) are estimated to be €18 800 million, which is an additional 1.25% on the EU’s total energy bill 2.15.10 Future Price Trends The prices of oil and gas have fallen back from the peak levels reached in 2008, and so the competitive position of wind weakened When the oil price was $140 per barrel, and gas and coal prices were also rising, wind needed little in the way of support With oil at $70 per barrel however, that picture changes Although the downward trend in turbine prices halted around 2004, this was largely due to substantial increases in commodity prices – including fuels However, there are a number of factors that are still likely to lead to lower costs for both wind turbines and wind farms in the future, including: • The continuing trend toward larger wind turbines • More efficient methods of manufacture 498 Wind Energy Economics Installed cost, onshore ($/kW) 2500 2000 1500 1000 EIA, ref EIA, rising EIA, falling EEA 500 2005 2010 2015 2020 2025 2030 Date Figure 22 Estimates of future onshore costs EIA, Energy Information Administration, US DoE; EEA, European Environment Agency • A better understanding of wind loads and materials properties • Larger wind farms As the demand for wind turbines in 2009 is still high, there is little incentive for manufacturers to reduce profit margins for sake of competition On the other hand, high production levels generally foster more productive methods of working – the so-called ‘learning curve effect’ Manufactured items that are produced in quantity benefit from increased production, due to improved manufacturing and assembly techniques The way in which costs fall depends on the product, and is a function of the relative inputs of material and labor For wind turbines, it has been estimated that the price is likely to fall by 8–15% for each doubling of production The major manufacturers mostly require between 2.4 and employees to produce MW of wind turbine and there may be some modest scope for this figure to be reduced Weight reductions may also be a route to cheaper wind turbines Although wind turbine weights, per megawatt of rated power, fell steadily as machine sizes increased, they now appear to have stabilized at around 100 kg kW−1 However, weight is not the only criterion that governs the price of a wind turbine and the recent move toward concrete towers for very large machines may enable cost savings to be made The future trajectory of wind turbine costs, both onshore and offshore, is uncertain Onshore, the US Department of Energy suggests three scenarios, reference, rising costs, and falling costs [21] and the progression of installed costs up to 2030 is shown in Figure 22 With a baseline cost in 2008 of $1923 kW−1, the range in 2030 lies between $1143 kW−1 and $2389 kW−1 The European Environment Agency suggests lower costs From a baseline in 2005 of $1500 kW−1, they suggest costs will fall to $835 kW−1 by 2030 Future offshore wind costs are equally uncertain One recent study has identified a range of factors that are likely to influence costs and suggested that installed costs could rise or fall by 20% by 2015 from the current British value of €3570 kW−1 [82] The corresponding cost range by that date lies between $3760 kW−1 and $6000 kW−1 However, some of the cost drivers are particularly related to the British market – particularly the link between the British pound and the euro Other factors would, however, put an upward pressure on prices internationally, particularly high commodity prices, supply chain difficulties in the offshore market, and competition for turbines with the onshore market Given that the offshore market is still relatively new and that manufacturers and installers are still innovating, the prospects for ‘learning by doing’ in the medium to long term are good and so, given stable commodity prices, the prospects for cost reductions are also good The US Department of Energy suggests a wide range of cost estimates for 2030, ranging from $4230 kW−1 to $2023 kW−1, while the European Environment Agency again suggests the lowest cost – at $1475 kW−1 A selection of these estimates is shown in Figure 23 2.15.10.1 Future Fuel Prices There is a wide range of projections for future fuel prices There is, however, a good measure of agreement between the reference scenario from the US Department of Energy and one of the scenarios suggested by the UK Department of Energy and Climate Installed cost, offshore ($/kW) 6000 5000 4000 3000 2000 EIA, ref EIA, rising EIA, falling EEA GH, best GH, worst 1000 2005 2010 2015 2020 2025 2030 Date Figure 23 Projections of future offshore wind energy costs EIA, Energy Information Administration, US DoE; EEA, European Environment Agency; GH, Garrad Hassan Wind Energy Economics Oil price ($/bbl) 140 130 120 110 100 90 80 70 60 50 2010 2015 499 UK DECC3 US EIA 2020 Year 2025 2030 Figure 24 Future oil price estimates Change, as shown in Figure 24 The latter’s High Demand, Producers’ Market Power scenario matches the most recent price changes well and suggests $84 per barrel in 2010, $102 per barrel in 2015, and $120 per barrel in 2020 – very similar to the American figures The various other scenarios suggested by the UK government put the oil price between $60 per barrel and $150 per barrel in 2020 The American projections have an even wider price range – up to $180 per barrel by 2020 2.15.10.2 Price Comparisons in 2020 Despite the uncertainty in future wind energy costs, there is an equal uncertainty in future fuel costs However, once a wind farm has been built, its generation costs are virtually fixed The same is not true for the fossil sources of generation Their generation costs are subject to fuel price uncertainty, and this gives wind energy a competitive advantage The implications of this uncertainty have been examined in some detail by the European Wind Energy Association and it is argued that this gives wind energy a competitive advantage [8] As there are so many uncertainties in price projections, it is hazardous to make generation cost comparisons into the future However, if oil, by 2020 is at $120 bbl−1 and if, for the sake of argument, gas prices match oil prices (as they have done in the past), that would suggest that the fuel price would be about $74 MWh−1 The corresponding electricity generation price – with no allowance for any carbon price – would be about $168 MWh−1 With no change at all in the installed cost of onshore wind up to that time, that generation cost would ensure onshore wind was competitive at a wide range of sites For offshore wind to be competitive, the installed costs would need to be about $4100 kW−1 Even a modest adder for the price of carbon would ensure that both onshore and offshore wind were comfortably competitive The American Electric Power Institute has examined the issue of competitiveness in 2015 and 2025 In 2015, they found that wind could compete with gas (costing $34 MWh−1 approximately) with the carbon price of around $12 tonne−1 of CO2 In 2025, they suggested wind (with a higher productivity) would compete with gas at a similar carbon price 2.15.11 Conclusions After falling steadily for about 20 years wind turbine and wind farm costs eased upward from around 2004 – largely due to increased commodity costs – but it seems likely that wind turbine prices have now (late 2009) stabilized at around $1500 kW−1 As more manufacturers enter the market, competing for nearly 30 GW of installations annually, prices are likely to resume their downward trend, provided commodity prices remain stable Currently, an average installed cost of $2150 kW−1 may be taken as a typical onshore value There is a very large variation of wind speeds, plant costs, test discount rates, and other factors that influence wind energy prices around the world With this caveat, the corresponding generation costs (8% interest rate, 20-year amortization) range from about $140 MWh−1 on a low wind speed site, falling to about $90 MWh−1 on a high wind speed site Offshore installed costs on average are around $ 4200 kW−1, but generation costs are less than double the onshore values on account of the higher wind speeds A direct comparison between wind energy prices and those of thermal plant is misleading, as wind has lower external costs and often has a higher value At the present time (late 2009), wind is slightly more expensive than coal or gas-fired generation if there is no carbon price in lieu of the extra costs Variability costs need to be taken into account once wind energy penetration levels exceed about 10% but, even at the 40% level, the additional balancing costs add less than 10% to the price of wind energy Other factors, such as the cost of backup, also need to be taken into account, but the impact on consumer prices is modest (around 7% at 40% penetration level) Although there is some uncertainty about the competitive position of wind in the future, it is likely that oil prices continue to rise If they reach $120 bbl−1 – as projections in America and Britain suggest – wind becomes competitive with gas, onshore A modest carbon price (about $12 tonne−1 of CO2), or a 10% reduction in offshore costs from the present levels, would enable offshore wind to become competitive as well 500 Wind Energy Economics References [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [45] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] http://en.wikipedia.org/wiki/Payback-period (accessed 15 November 2010) http://en.wikipedia.org/wiki/Internal_rate_of_return (accessed January 2011) Wood A and Woollenberg B (1996) Power Generation, Operation and Control New York, NY: Wiley Tande JO and Hunter R (1994) Estimation of cost of energy from wind energy conversion systems Submitted to Executive Committee of IEA Wind Programme Milborrow DJ (1995) An analysis of UK wind farm statistics British Wind Energy Association, 17th Conference Warwick, UK, 19–21 July London, UK: MEP Brocklehurst F (1997) A review of the UK onshore wind energy resource ETSU Report ETSU-R-99 ETSU, Harwell, UK Danish Energy Agency (1999) Wind power in Denmark: Technology, policies and results Danish Energy Agency, Copenhagen http://193.88.185.141/Graphics/Publikationer/ Forsyning_UK/Wind_Power99.pdf (accessed November 2009) Krohn S (ed.) (2009) The Economics of Wind Energy European Wind Energy Association, Brussels http://www.ewea.org Milborrow DJ (2010) What a difference a year can make Windpower Monthly 26(1): 41–47 These and later data were reported in: Wiser R and Bolinger B (2010) 2009 Wind technologies market report LBNL–3716 E Lawrence Berkeley Laboratory, California Element Energy Limited (2009) Design of feed-in tariffs for sub-5 MW electricity in Great Britain quantitative analysis for DECC URN 09 D/704 Department of Energy and Climate Change, London http://www.decc.gov.uk European Commission (1999) Wind Energy: The Facts Luxembourg: Office for Official Publications of the European Communities Madsen PS (1996) Tuno Knob offshore wind farm EU Wind Energy Conference Göteborg, Sweden, 20–24 May H S Stephens and Associates, Bedford Van Zanten W (1996) Lely wind farm Caddet Newsletter September Ernst and Young for the Department of Energy and Climate Change (UK) (2009) Cost of and financial support for offshore wind URN 09 D/534, London L.E.K Consulting (2003) Perspectives on Renewable Power in the UK London, UK: The Carbon Trust Department for Business Enterprise and Regulatory Reform (2008) Scoby sands offshore wind farm 3rd Annual Report URN 08/P48 Department of Energy and Climate Change, London http://www.decc.gov.uk/ Morgan C (2006) Optimisation of operations for onshore and offshore wind farms EWEC Conference Athens, Greece European Environment Agency (2009) Europe’s onshore and offshore wind energy potential EEA Technical Report No 6/2009 European Environment Agency, Copenhagen http://www.eea.europa.eu/ Carbon Trust (2008) Offshore wind power: Big challenge, big opportunity The Carbon Trust, London http://www.carbontrust.co.uk Energy Information Administration, US Department of Energy (2009) Annual energy outlook http://www.eia.doe.gov/ Ofgem (2009) Project discovery Energy market scenarios http://www.ofgem.gov.uk Commission of the European Communities (2008) Energy sources, production costs and performance of technologies for power generation, heating and transport SEC (2008) 2872 Commission of the European Communities, Brussels http://ec.europa.eu/energy/strategies/2008 California Energy Commission (2010) Comparative costs of California central station electricity generation CEC-200-2009-07 SF California Energy Commission, Sacramento, California Geothermal Energy Association (undated) c.2008 http://www.geo-energy.org/reports.aspx (accessed 16 June 2011) City Analysts Voice New Nuclear Build Scepticism (2009) New Power 10: 54–55 Hohmeyer O (1988) Social Costs of Energy Consumption Berlin, Germany: Springer European Commission (1995–1999) ExternE Externalities of Energy, 10 vols Luxembourg: Office for Official Publications of the European Communities Stern L (2006) Review on the Economics of Climate Change Cambridge, UK: Cambridge University Press Energy Information Administration, US Department of Energy (1995) Electricity generation and environmental externalities: Case studies DOE/EIA-0598 US Department of Energy, Washington Energy and Environmental Economics, Inc (E3) (2009) http://www.ethree.com/home.html Dodd J (2008) Now vying for third slot in European rankings Windpower Monthly News Magazine March: 55 Sun and Wind Energy 9/09: 43 Sun and Wind Energy 2/2006: 74 Fouquet D (2009) Prices for Renewable Energies in Europe European Renewable Energies Federation, Brussels http://www.eref-europe.org/ Milborrow DJ (2009) Managing variability World Wildlife Fund, Godalming, Surrey http://www.wwf.org.uk/ Gross R, Heptonstall P, Anderson D, et al (2006) The Costs and Impacts of Intermittency London, UK: UK Energy Research Centre Ofgem (2008) National Grid Electricity Transmission System Operator Incentives from April 2008 (accessed 14 June 2011) Davies E (ed) (1991) Modern power station practice In: System Operation, vol L Oxford, UK: Pergamon Press Milborrow DJ (2009) Quantifying the impacts of wind variability Proceedings of the Institution of Civil Engineers 162: 105–111 National Grid UK (2004) Seven year statement http://www.nationalgrid.com/uk/ National Grid UK (2008) Seven year statement http://www.nationalgrid.com/uk/ National Grid plc (2008) Evidence to house of Lords European Union Committee 27th Report, Session 2007–08 The EU’s target for renewable energy: 20% by 2020 Report HL 175-II London, UK: The Stationery Office DeMeo E, Jordan G, Kalich C, et al (2007) Accommodating wind’s natural behaviour IEEE Power and Energy Magazine November/December, pp 59–67 Dale L, Milborrow D, Slark R, and Strbac G (2004) Total cost estimates for large-scale wind scenarios in UK Energy Policy 32: 1949–1956 EoN UK (2008) Requirement for thermal generation to back-up wind capacity Evidence submitted to House of Lords Economic Affairs Committee into the Economics of Renewables HL Paper 195 London, UK: The Stationery Office Milborrow D (2009) Is wind reliable? New Power UK 1, pp 59–67 British Wind Energy Association (1982) Wind Energy for the Eighties Stevenage, UK: Peter Peregrinus Ltd Union for the Co-ordination of Production and Transmission of Electricity (UCPTE) (1995) Annual Report Paris ESB National Grid (2004) Impact of wind power generation in Ireland on the operation of conventional plant and the economic implications www.eirgrid.com/…/2004%20wind% 20impact%20report%20 Milborrow D Assimilation of wind energy into the Irish electricity network Report for Sustainable Energy Ireland www.seai.ie/Archive1/Files_Misc/ DavidMilborrowIreland.ppt Ofgem (2009) Managing constraints on the GB transmission system Open letter from Ofgem to National Grid, 17 February http://www.ofgem.gov.uk/Pages/MoreInformation aspx?docid=119&refer=Networks/Trans/ElecTransPolicy/tar Electricity Networks Strategy Group (2009) Our electricity transmission network: A vision for 2020 webarchiv.nationalarchives.gov.uk/…/ensg.gov.uk/…/ ensg_transmission_pwg_full_report_final_issue_1.pdf Pedersen J, Eriksen P, and Orths A (2006) Market impacts of large-scale system integration of wind power European Wind Energy Association Conference Athens, Greece European Wind Energy Association Sinclair Knight Merz (2008) Growth scenarios for UK renewables generation and implications for future developments and operation of electricity networks BERR Publication URN 08/121, London Wind Energy Economics 501 [56] Kariniotakis G, Moussafir, J, Usaola, J, et al (2003) ANEMOS: Development of a next generation wind power forecasting system for the large-scale integration of onshore and offshore wind farms European Wind Energy Conference, Madrid, Spain [57] (2003) Wind generation forecasting for power dispatching EPRI Journal Online May http://www.epri.com [58] Gow G (2003) An adaptable approach to short term wind forecasting American Wind Energy Association (AWEA) Conference, Austin, Texas, AWEA [59] Wessel A, Jiang J, and Dobschinski J (2009) Improving short-term forecasting with online wind measurements European Wind Energy Conference Marseille, France European Wind Energy Association [60] Perkins J (2006) Demand side participation in balancing services UK: National Grid, London http://www.nationalgrid.com/NR/rdonlyres/ [61] Department of Energy and Climate Change (2008) The potential for dynamic demand URN 08/1453 UK Department of Energy and Climate Change, London [62] Farmer ED, Newman VG, and Ashmole PH (1980) Economic and operational implications of a complex of wind-driven power generators on a power system IEE Proceedings A 127: [63] Wan Y and Parsons B (1993) Factors relevant to utility integration of intermittent renewable technologies NREL/TP–463–4953 National Renewable Energy Laboratory, Golden, Colorado [64] Douglas J (2006) Putting wind on the grid EPRI Journal Spring [65] House of Lords, Select Committee on Economic Affairs (2008) The economics of renewable energy 4th Report of Session 2007–08, HL Paper 195–I London, UK: The Stationery Office Limited [66] Greiner C, Korpas M, and Gjengedal T (2009) Optimal operation of energy storage systems combined with wind power in short-term power markets European Wind Energy Conference Marseille, France European Wind Energy Association [67] Electric Power Research Institute (2008) Emerging technologies to increase the penetration and availability of renewables EPRI, Palo Alto, California http://www.epri.com [68] Foley A, Leahy P, and McKeogh E (2009) Wind energy variability, the Wales-Ireland interconnector and storage The Proceedings of the IEEE PES/IAS Conference on Sustainable Alternative Energy at Instituto de Ingeniería Energética Spain, Universidad Politécnica de Valencia, 28–30 September [69] Hurley B, Hughes P, and Giebel G (2007) Reliable power, wind variability and offshore groups in Europe In: Boyle G (ed.) Renewable Electricity and the Grid: The Challenge of Variability London, UK: Earthscan [70] De Decker J, Vu Van T, and Woyte A (2009) The Greenpeace offshore grid report: Development drivers and benefits European Wind Energy Conference Marseille, France European Wind Energy Association [71] 2nd TEN-E Information Day, Brussels http://ec.europa.eu/energy/…/2008_05_20_ten_e_infoday_slide_presentation.pdf (accessed 20 May 2008) [72] Bouffard F and Kirschen D (2008) Centralised and distributed electricity systems Energy Policy 36: 4504–4508 [73] Ofgem (2009) Electricity distribution price control review: Methodology and initial results paper Ofgem 47/09 Office of Gas and Electricity Markets, London [74] PB Power (2008) The PB Power report on future network architectures BERR file 46168 UK Department of Energy and Climate Change, London [75] National Grid Briefing Note ‘Gone Green’ a scenario for 2020 http://www.nationalgrid.com/uk/Electricity/Operating+in+2020/2020+Consultation.htm [76] Elkraft System (2005) Long-term challenges in the electricity system [Elkraft was the system operator in Eastern Denmark and is now merged with Energinet] [77] Milborrow DJ (2004) Wind economics set to beat gas in Ireland Windpower Monthly December: 31 [78] Milborrow DJ (2006) Nuclear suddenly the competitor to beat Windpower Monthly January: 43 [79] Energy Information Administration (2007) Impacts of a 15% renewable portfolio standard SR/OIAF/2007–03 US Department of Energy, Washington [80] DeMeo E (2007) A picture of 20% US electrical energy from wind UWIG Technical Workshop Anchorage, Alaska 24 July [81] Poyry Energy Consulting (2008) Compliance costs for meeting the 20% renewable energy target in 2020 URN 08/757 Department for Business Energy and Regulatory Reform, London http://webarchive.nationalarchives.gov.uk/ [82] Garrad Hassan and Partners, Bristol (2009) UK offshore wind: Charting the right course British Wind Energy Association http://www.bwea.com ...470 Wind Energy Economics 2. 15. 8.10 2. 15. 8.11 2. 15. 8. 12 2 .15. 8. 12. 1 2. 15. 8. 12. 2 2. 15. 8. 12. 3 2. 15. 8. 12. 4 2. 15. 8. 12. 5 2. 15. 8. 12. 6 2. 15. 8. 12. 7 2. 15. 8. 12. 8 2. 15. 8. 12. 9 2. 15. 9 2. 15. 10 2. 15. 10.1 2. 15. 10 .2. .. Coal with FGDa CCGTa Nuclear [26 ] 1900 26 00 380 0–4 600 180 0–5 400 340 0–3 900 410 0–5 000 170 0–3 400 [25 ] 1900 21 00 960 320 0–4 000 32 60 90 150 1 3–8 0 5 7–1 75 12 100 165 28 –8 5 13 90 a FGD, flue gas desulfurization;... 0.5 24 0 1 7–4 00 7–6 0 5–1 3 3 7–1 87 3–7 3–7 8. 3–1 9 0.3 Small Small 1.1 25 2 25 0 .20 3. 6–5 0 Note: During the period these studies were carried out €1 = $1.3, approximately 2 5 0.18 4–1 0 486 Wind Energy