1

The Cost Effectiveness of Renewable Energy Support Schemes in the European Union

Arjun Mahalingam1, David Reiner and David Newbery, Energy Policy Research Group (EPRG), University of Cambridge

Introduction

The EU Renewables Directive (2009/28/EC) seeks to deliver by 2020 a 20% share of renewable energy in gross final energy production, as part of the 2020 climate and energy package, which also includes a binding target of reducing greenhouse gas (GHG) emissions by 20% relative to 1990 and an aspirational energy efficiency target of reducing energy consumption by 20% relative to projected 2020 levels.

The overarching goals are shared out across member states (MSs) to reflect differences in national circumstance such as endowment and starting point. Each MS is required to meet its specific binding national target for the 2020 share of renewable energy, which ranges from 10% for Malta to 49% for Sweden. The existing National Allocation Plans under the EU Emissions Trading System (ETS) specify CO2 targets for each MS to be met by 2020, while the energy efficiency target is not binding, but as of April 2013 all MSs are meant to have set indicative national targets. Overall, the EU is well on track to meeting its 2020 renewable targets (having exceeded its indicative target for 2011/12) and its GHG targets (which stood at -18% in 2012, thus almost reaching its target ahead of schedule), but is lagging in terms of energy efficiency where only four MS have been deemed to be making ‘good progress’ by the European Environment Agency (EEA, 2013).

In January 2014, following on from the 2020 climate and energy package, the European Commission published A policy framework for climate and energy in the period from 2020 to 2030, which proposed an EU-wide target of a 40% reduction in GHG emissions by 2030 (again relative to 1990) which would be translated into binding national-level GHG targets, complemented by an EU-level target for renewable energy of 27%. EU leaders will finalize the framework no later than October 2014.

The 2030 GHG target is equivalent to a 43% decrease for the sectors covered by the ETS and likely to require even deeper reductions from the electricity sector. The Policy framework argued that as the 40% GHG target would likely deliver the proposed 27% EU Renewable Energy Supply (RES) target, there was no need to require country-specific RES targets. Moving away from national targets would afford MSs the ‘flexibility to transform the energy system in a way that is adapted to their national preferences and circumstances’. Nevertheless, the 2020 targets remain binding, and in fact, the day after the Commission’s 2030 proposal was announced, Ireland was referred to the European Court of Justice for failing to fully transpose the 2009 Renewables Directive (Lynch, 2014).

The drive towards greater renewables must be implemented by each MS and interacts with other key EU objectives including its policies on competition and market integration. MSs are free to choose their own policy instruments such as green taxes, investment subsidies and

1 Energy Policy Research Group, Judge Business School, University of Cambridge, Trumpington Street, Cambridge CB2 1AG, UK. Ph +44 1233 748814, E-mail: am2257@cam.ac.uk

2

feed-in tariffs in order to meet their national targets, provided they accord with the State Aids Guidelines (European Commission, 2014). The electricity sector offers the greatest potential for switching to RES, as technically it only requires changes to generation, leaving the final product unchanged and hence requiring the least adaptation by consumers (although demand side response will become increasingly valuable to handle the intermittent nature of most RES in the electricity sector, or RES-E). In terms of market integration, the electricity sector (together with gas) are required to make further reforms as part of the Third Energy Package to complete the internal market in electricity and specifically to meet the requirements of the Target Electricity Model. That, and the need to support a higher share of renewables in electricity, has prompted many MSs to reform their electricity markets. Thus, for example, the UK Government published its plans for Electricity Market Reform (EMR) in the Energy Act 2013 in December 2013 setting out the policies to meet its climate change obligations (Great Britain 2013).

Our aim here is to examine the various ways in which RES-E is supported in different Member States, and then contrast the UK’s proposals under the EMR with Germany’s support schemes, which are part of its wider Energiegewende (or energy transition). We concentrate on the cost-effectiveness of these policies and their success in minimizing unnecessary rent transfer to developers and reducing the cost of finance.

Climate change mitigation is predicated on taking the future seriously, which requires discounting future damage at a rather low discount rate (Stern, 2007). At low discount rates, low-carbon electricity and particularly RES-E, which are highly capital intensive but have low, often zero, running costs, are increasingly attractive, so the central message for delivering cost-effective decarbonization is to find effective ways of lowering the cost of capital. By that test many support mechanisms do a poor job.

We first review the different support mechanisms for RES-E across the EU and then carry out a deeper investigation of the UK and German cases. The UK EMR proposes interventions to address various market failures, but the specific form of contract for RES-E compares somewhat unfavorably with the fixed feed-in-tariff that has been both cheaper and more effective in Germany. Some of the arguments against that scheme are refuted and the potential benefits of lowering the cost of capital quantified for UK on-shore wind. The challenge of deploying renewable electricity in liberalized markets

Electricity market liberalization in developed countries took place under benign conditions of excess capacity, concentrated markets and the advent of cheap, efficient Combined Cycle Gas Turbines (CCGTs). While electricity prices were set by fossil fuel prices and supported by market power, new investment in gas generation (either by dominant incumbents or backed by long-term power purchase agreements) was easy to finance. Decarbonization was aided by initially quite high CO2 prices in the ETS, while low gas prices further encouraged fuel switching – and CCGTs halved the carbon intensity of the coal-fired plant they displaced. Following the failure of the international climate negotiations at Copenhagen and the economic crisis in 2008, the ETS carbon price collapsed, which was accompanied by a trebling of EU gas prices, and the US shale gas revolution driving world coal prices down, which together put decarbonization at serious risk.

At the resulting wholesale prices, low-carbon generation was commercially unattractive, and if the 20/20/20 targets were to be met, some form of politically determined market

3

intervention was inevitable. As soon as prices are set through a political process, their future trajectory ceases to be predictable on the basis of commercial fundamentals. Even with an adequate carbon price (i.e. one high enough to make mature low-carbon generation profitable), less mature renewables would not be commercially viable. Additional support can be justified by the public good they create in the form of induced innovation and cost-reduction that attends mass deployment. However, neither future carbon prices nor public commitments to renewables support have much credibility without firm, legally binding contractual backing. Absent that, the cost of financing what are highly capital-intensive investments from the private sector becomes excessive, and further reduces support for the climate change agenda.

Once renewables are adequately supported to meet the 2020 targets, the additional RES-E capacity and low demand growth (caused by higher prices, the financial crisis and the quest for energy efficiency) depresses prices and makes conventional generation investment unattractive. Peaking capacity to meet shortfalls from intermittent generation is unlikely to be investible, as the prices needed to justify its cost in the few hours when it is needed would not be credibly allowed to reach adequate levels without some form of contractual guarantee of the kind that a Capacity Payment Mechanism might provide. The UK EMR debate offers a clear diagnosis of the problems facing liberalized electricity markets in meeting climate change targets that are commercially unattractive at prevailing fuel and carbon prices and will be one of the case studies examined below. Forms of renewables support

RES support mechanisms are defined by the European Commission as those schemes ‘‘originating from a market intervention by a Member State that help energy from renewable sources to find a market by reducing the cost of production of this energy, increasing the price at which it can be sold, or increasing, by means of a renewable energy obligation or otherwise, the volume of such energy purchased’’. The main forms of support schemes in the European Union are Feed-in Tariffs (FiTs), Feed-in Premiums (FIPs), quota obligations, investment grants, fiscal incentives and tenders (Batlle, 2011; Kanellakis, Martinopoulos, & Zachariadis, 2013). Table 1 shows the various support schemes in operation in each MS. In rough order (for the first three) of prevalence and risk they are as follows.

A Feed-in Tariff (FiT) pays a guaranteed price to eligible RES-E producers and so provides them with a fixed payment per MWh produced. The counterparty, often the System Operator, takes on the balancing and marketing responsibilities, which is automatic in pool systems in which all generators offer supply, the cheapest of which is dispatched to meet consumer demand. FiTs pass all the marketing and balancing risks and costs through to final consumers, leaving only volume risk, so producers benefit from the certainty of receiving a fixed-level of support over a defined period, thereby eliminating most investment risks. At present, they are the most common policy support instruments in the EU, used in France, Germany, Spain, the Baltic countries and a few others in Central Europe. The important lesson to draw is that the overall costs for FiT-supported renewables are considerably lower than for other support mechanism with higher risks.

A Feed-in Premium (FIP) is a guaranteed payment per MWh in addition to the revenue from selling to the wholesale market. Its claimed advantage is that RES-E generators bear the risks and rewards for their exposure to the electricity price and balancing risks, so they are

4

motivated to minimize these costs, which are also made explicit and become part of the required subsidy. They are commonly used as stand-alone support mechanisms for electricity in the Netherlands and Denmark. In some countries (such as Italy, Germany and the Czech Republic), both FIP and FiTs co-exist. A variant of this type of support is the floating premium, where the sum of the wholesale price and the premium are stabilized to some extent. The new UK Contract for Difference (CfD) is such an example, as the CfDs are standard two-sided contracts which pay or charge (s – p) per MWh, where s is the strike price for each technology published in the UK Energy Act 2013, and p is the reference price (typically the hourly wholesale price). This reduces revenue volatility, leaving only basis risk (the difference between the reference price and the price actually secured when selling) and the underlying volume risk.

Quota obligations are obligations placed on agents to supply (or produce or consume) a minimum share of RES-E. The obligations normally increase in line with the annually rising RES targets. The agents may be subject to a penalty for any shortfall, which is recycled to suppliers in proportion to their RES-E supplied. In certain cases, these quota obligations are combined with Renewable Obligation Certificates (ROCs) or Tradable Green Certificates (TGCs) that can be sold. The advantage is that the certificates serve as a proof of compliance while generating extra revenue for the RES-E generators. The downside is that uncertainty over the future price of the ROCs leads to additional financial risk on top of the wholesale price risk. Quotas with TGCs are used as policy instruments in the U.K., Italy, Poland, Sweden, Belgium, and Romania.

Both quota obligations and FIPs are compatible with energy-only markets trading through power exchanges and all three instruments provide output rather than capacity support, although it is investment in capacity that leads most directly to the leaning benefits from deployment. They also risk distorting location signals as they over-reward deployment in good resource locations (e.g. the windier places) which may not be economically attractive if the power produced is valued at its market price when compared to the additional transmission costs. On the other hand the Renewables Directive measures success by output not deployment and to that extent creates this distortive incentive.

Investment grants or subsidies are available in many MSs for RES-E and renewable heating and cooling (RES-H) systems. Such incentives reduce the cost of investment and thereby stimulate more market diffusion of less mature technologies. In general, they are provided at the beginning of the project and are calculated as a percentage of the output of renewable energy expected or the total cost of investment. These support mechanisms are the only policy instruments for RES available in Finland. However, they are also available in combination with different policy instruments in the other MSs. They have the great advantage of directly reducing the up-front cost of the investment. Where based on investment cost rather than predicted (as opposed to benchmarked) output, they support the activity most likely to deliver learning benefits without distorting location decisions.

Fiscal incentives may take the form of tax incentives or exemptions, tax rebates, tax refunds, lower VAT rates or attractive depreciation schemes, soft or low-interest loans, etc., with the first two the most common in the EU. They are usually used to complement other renewable support schemes. As such they can target a specific RES technology to stimulate its growth and impact that market segment. Spain, the Netherlands, Greece and Finland provide tax incentives for investments in RES, while Poland, Sweden, UK, Slovakia and Latvia provide deductions against income tax or production incentives. Soft loans at sub-market interest rates lower the

5

financial costs of investment. These loans may also allow longer repayment periods and interest holidays. They are available in Germany, the Netherlands, Poland and a few other EU countries.

Finally, and likely to be increasingly important, tenders and auctions create competition for either a predetermined capacity or a financial budget. Generators bid at the price that they consider makes the investment profitable. The cheapest bids are awarded contracts, usually at the auction-clearing price (per MW or MWh over a time period, that itself could form part of the bid). The main advantage of auctions is that they avoid less-well informed bureaucrats setting the subsidy rate, which instead is revealed by the auction. As such, it is therefore both demonstrably least cost and fair, meeting the main criteria for acceptable State Aid. They have been used in the UK, Spain, the Netherlands and Denmark.

Thus auctions have been used to good effect in financing connections to UK offshore wind farms, where the expensive sub-sea link from the wind farm to the on-shore connection point is built by the developer, who will receive the auction price from the winner who takes on the responsibility of maintaining the link in return for a flow of revenue per MWh. This model has substantially lowered the weighted average cost of capital needed to finance the investment, but the winders (usually banks) know exactly what they are buying. The closest analogy would be if an agency secured sites with planning permission for on-shore wind farms and grid connections, and then auctioned off the right to construct and operate a wind farm. This would clarify the product to be auctioned and remove much of the risk and should lower the cost of finance. Past auctions in the UK have instead just auctioned the payment per MWh for defined volumes of specific technologies such as on-shore wind, leaving the winner to find a site, secure planning permission and a grid connection and then build the wind farm or other RES-E plant. As such it is cannot extract site-specific rent, and, as evidence suggests, suffers from under-delivery when winners find the costs of further development to great and walk away from their option. The lesson to draw is that designing the auction is critical to successful delivery of RES-E.

In response to the combination of the more ambitious 20/20/20 targets and the requirements of the Target Electricity Model, a growing number of Member States are reforming their electricity markets, of which the most explicit to date is the UK Electricity Market Reform (EMR). It provides an excellent case study to identify the costs and benefits of different forms of renewables support, particularly when contrasted with Germany. The UK Electricity Market Reform (EMR)

The UK led Europe in creating a liberalized, privatized and unbundled electricity supply industry. Since the British2 market reform of 2001 replaced the electricity pool, it has an energy-only market with no capacity payments, a two-price balancing mechanism, and until recently, a thinly traded day-ahead OTC market (but a more liquid forward market), six incumbent generators (the Big Six) who are integrated into supply (retailing), with distribution and transmission under separate ownership and subject to incentive regulation by the National Regulatory Agency Ofgem (Newbery, 2012).

Before EMR, RES-E was supported by tradable Renewable Obligation Certificates (ROCs). The Renewable Obligation (RO) requires all RES-E generators to market and balance their power (unless they hold a contract, usually with one of the Big Six), on top of which is 2 Northern Ireland is part of the UK but electricity is traded in the all-island Single Electricity Market, which remains a mandatory pool with capacity payments and therefore structurally quite different from GB.

6

added the value of the number of ROCs granted per MWh produced. ROCs traded in March 2014 at £41.40 ($70)/ROC when the wholesale electricity price was around £40/MWh. On-shore wind farms receive 0.9 ROCs per MWh, so the RO scheme nearly doubles the value of on-shore wind. As offshore wind receives 2 ROCs/MWh, the RO scheme trebles their value. Despite the apparently generous support and the fact that she has the best wind resource in Europe, the UK lags far behind other countries like Germany and Spain, and by 2010 looked most unlikely to reach the EU RES 2020 target. Ofgem reported its disquiet with the market arrangements3 and the Government also started consulting on market reform (DECC, 2010). Following tradition, it published an Energy White Paper (DECC, 2011, the fourth since 2003) that led to the publication of the Energy Act 2013 in December.

The Energy Act 2013 has four main elements: the Carbon Price Support (CPS), which was enacted by the Treasury in the 2011 Budget to start in April 2013; Contracts for Differences (CfDs) for low-carbon electricity (DECC, 2012a), an Emission Performance Standard of 3,350 tonnes CO2/MW capacity/year (described as 450gm/kWh operating at a capacity factor of 85%) to rule out any unabated new coal stations, and a Capacity Payment Mechanism. As laid out in the 2011 Budget, the CPS would levy a tax on the carbon content of fuel used for electricity generation (but no other fuel) to bring its level up to £16/tonne CO2, rising to £30/tonne in 2020, and projected to rise to £70/tonne by 2030 (all at 2009 prices).4 The CfDs are standard two-sided contracts as described above (a form of floating FIP) and are intended to de-risk RES-E while still exposing them to market and balancing requirements (in accordance with the implied requirements of the State Aid Guidelines).

The ROC bandings (i.e. the number of ROCs per MWh for each technology) were reformed in 2011 (after the start of the EMR process, see DECC, 2012b) and slightly modified from their original levels (on-shore wind was reduced from 1 ROC/MWh to 0.9 ROCs/MWh) but remained at a uniform rate across GB regardless of wind conditions, and had to be set at a level to make the marginal wind farm needed to meet the RES-E targets profitable. The new CfD strike prices were set to ensure comparability with the existing generous ROCs as ROCs continue to be available until 2017 and it was felt important to wean developers away from the RO Scheme. They were based on forecast wholesale prices that were expected to rise because of the CPS, and which were thus expected to make ROCs even more attractive than before the CPS. However, in the 2014 Budget the CPS rates were capped at a maximum of £18/tonne from 2016-17 until 2019-20, effectively freezing the CPS rates at around 2015-16 levels and casting doubt on the future value of the RO, but not leading to any change in the now possibly over-generous CfD strike prices.

The CfD strike prices were again set uniformly across GB (except for the islands of Okney, Shetland and the Hebrides) to make the marginal supply profitable at rather generous hurdle rates (effectively the Weighted Average Cost of Capital, WACC) for financing the renewable projects. Thus for on-shore wind the hurdle rate is 7.1% real and for offshore wind is 9.7% (Round 2) rising to 10.1% (R3), which are generous compared to the WACC for the offshore grid connections that were auctioned (DECC, 2013a). NERA (2013) suggests that on-shore wind might expect to be financed with 75-85% debt under a CfD. The standard formula for the WACC is f + gd + (1-g)*β(r – f). If the risk-free rate f is as high as 2% and if g is the gearing

3 http://www.ofgem.gov.uk/markets/whlmkts/discovery/Pages/ProjectDiscovery.aspx 4 HM Treasury, Budget 2011, HC 836, March 2011

7

or debt share is 80%, d is the debt risk premium (UK regulated water companies take f + d as 2%, so d = 1% would be generous), r is the market equity return, so r - f is therefore the equity risk premium (perhaps 6%), and β measures the correlation of the returns of the asset with the equity market (all rates are real) then if β = 1 (i.e. the CfDs had not reduced risk at all) the WACC might be 2% + 0.8*1% + 0.2*1*6% = 4%. In practice if β were as high as 1, the gearing would be lower, or, putting it the other way round, as there is no reason why wind output should be correlated with the equity market, β should be low, further lowering the real WACC.

The higher rate might be defended as investors were thought to be unsure about the new CfD instrument, were wary of reforms underway to the balancing mechanism, were worried about the then pending (now published in April 2014) new State Aid Guidelines that might rule some supports illegal, and were quite understandably skeptical about the stability of any government legislation (and have been subsequently vindicated in the capping of the CPS). As such, there is an additional political/regulatory risk premium that makes support more costly. Just how costly can be estimated given estimates of the WACC under different support regimes.

DECC (2013b) published the strike prices for various low-carbon electricity technologies, which, for on-shore wind will be £95/MWh until 2017/18, at which point they will drop to £90/MWh, while off-shore wind will be £155/MWh until 2016/17 when they drop first to £150 and then to £140/MWh in 2017/18. Large Solar PV will fall from £120/MWh to £115 in 2016/17 and £110 in 2017/18 (all in 2012 prices and indexed to the CPI and for their contract life). Green and Staffell (2013, in turn derived from the rebanding review, DECC 2012b) provide the cost and performance data for on-shore wind needed to estimate the additional costs of the higher WACC. Thus if the WACC for on-shore wind had been reduced from 7.1% to 5% and if the average capital cost of an on-shore wind turbine is £1,486 per kW of capacity operating at an average capacity factor of 28% (decreasing from 31.3% initially at 1.6% p.a.), with fixed operating costs (including insurance and transmission charges) of £46 per kW per year, variable costs of £3/MWh and zero residual value after 15 years and with a reference price of £45/MWh, then the strike price could be reduced from £95/MWh to £80/MWh and the NPV of the additional support cost to consumers (discounting at the Treasury discount rate of 3.5%) would be reduced by 53%. If the turbine continues to operate for 20 years but selling at a wholesale price of £45/MWh after 15 years the strike price could have been reduced to £75/MWh, saving consumers 71% of the support cost. Extracting rent from windier locations on the German model discussed below would reduce consumer costs even more. Feed-in-Tariffs in Germany

Germany has been very successful in delivering high rates of on-shore wind investment despite having a considerably inferior wind resource to the UK. Most of the UK has wind energy of above 300 W/m2 on plains (6.5-7.5 m/s, considerably higher on hills and ridges), and above 500 W/m2 in Scotland, while most of Germany is below that figure.5 Germany has a comparable population density to the UK (225/km2 compared to 260/km2 for the UK). Spain is similarly challenged in terms of wind power with only a few areas exceeding 300 W/m2. Fig 1 compares installed capacity in Germany and Spain with the UK, and shows that in 2011 the UK had the same installed capacity as Germany in 2000. Over the decade 2000-2010 Germany increased 5 Data available at: http://www.windatlas.dk/europe/europeanwindresource.html although the period over which the resource was measured (in W/m2) is not given in the source.

8

installed capacity by about 20 GW – a task the UK has set itself for the decade to 2020, although not all on-shore. Fig 1 shows that Spain (with a lower population density of 93/km2) as also catching up rapidly and matching Germany’s pace with a slight delay, although the Spanish economic crisis (and in the power sector in particular) has recently undermined Spanish support for wind. Denmark was of course the early leader and has excellent wind conditions (and a lower population density than the UK at 130/km2) but in absolute magnitude is quite modest (although with a larger share of generation). In light of the remarkable success of Germany compared to the UK in stimulating wind investment, it is worth comparing their different support mechanisms, as Butler and Neuhoff (2004) suggests that the German system is more cost-effective.

Fig 1 Installed wind capacity in various EU countries, 1995-2011 The German FiT for on-shore wind in 2015 is €86.6 (£74.1)/MWh for the first 5 years (in

windy locations) and 20 years in less windy conditions, counting from commissioning, and thereafter is €47.3 (£40.5)/MWh. Moreover, unlike the GB CfD, it is not indexed to the price level. The lack of indexation is partly offset by the front-end loading that DECC seems to have ignored in its consultations. It is somewhat surprising that €47.3/MWh is likely to be below the future electricity wholesale price in GB, so effectively wind farms in Germany in windy locations are only really subsidized for the first 5 years.

If we compute the real indexed levelized value of the German FiT for windy locations over 15 years (assuming a 2% inflation rate) it would be €56.8 (£48.6)/MWh at 5% discount rate. For less windy locations (which would be inferior to most GB locations) the real indexed value would be €76.3 (65.3)/MWh. GB wind-farms would have to pay grid charges (TNUoS) for which the median estimated cost is £9,912/MW/yr or, for a rather low capacity factor of 2,000 hrs per year (23%), just under £5/MWh. The CfD for GB on-shore wind net of transmission charges

Installed wind capacity in MW

0

5,000

10,000

15,000

20,000

25,000

30,000

1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011

MW

Denmark Germany Spain UK

9

would then be £90/MWh compared to an indexed equivalent in Germany of £49-65/MWh (depending on discount rate and wind conditions). It should be noted that the rapid growth shown in Fig 1 was based on higher earlier FiTs, as the tariff has been degressing at 1% p.a. Thus in 2009 the FiT was €97/MWh, which, coincidentally, is £100/MWh at 2013 prices (but not indexed). It fell to €50/MWh after 5 years in windy locations, so again in £2013 at 5% real discount rate this was worth £63/MWh now.

The other salient difference between the German and UK system is that Germany extracts infra-marginal rent from windier sites by making the contract length depend on the first three years’ output, as a predictor of how long a contract would be needed to cover the investment costs. The UK gives all wind farms the same length of contract at the same strike price or ROC and thus allows developers to keep such rents, at additional costs to the consumer. Newbery (2013, p21) estimates that this transfer of rent for on-shore wind might have cost consumers £70 million per year after the rebanding review.

Spain has also been very successful in stimulating wind development and the Spanish FiTs appear more generous at €81.247 (£69.5)/MWh for 20 years, then dropping to €67.902 (£58.1)/MWh but it is not clear if this is indexed, nor whether it is subject to retrospective adjustment (which has created considerable uncertainty in the past and may be expected to raise perceived risk and hence cost).

Risk allocation, the cost of capital and the new State Aid Guidelines

The major cost of supporting RES-E derives from its high capital cost per MWh (a combination of a high capital cost per kW capacity and a low capacity factor, which for solar PV may be as low as 10% and even for on-shore wind in Britain averages only 27%). If, as is surely defensible, the public sector discounts future social costs and benefits at a lower discount rate than private RES-E developers (in the UK the Treasury discount rate is 3.5% real) then the cheapest form of support is a sufficient up-front capital grant that lowers the cost of the generation investment to the point that it is commercially viable selling into the wholesale market. The cost can be further reduced be reducing the risk of the electricity sales through a FiT at the expected wholesale price (less imbalance costs), transferring the risks of marketing and balancing to their cost minimizing locus with the System Operator.

It might be argued that this is unwarranted and discriminatory state aid, and merely shifts the very real risk and selling costs through to final consumers, but these arguments are mistaken. First, the total cost decreases directly in proportion to the number of participants who bear it. That is, if the cost of risk borne by one agent is C, and the risk is divided equally between n agents, then each bears a cost of C/n2 (assuming they all have the same risk aversion) and the total cost of risk is n times this, or nC/n2 = C/n. Transferring a large risk from a single modest-sized wind farm owner to 27 million electricity consumers dramatically lowers the total cost of the risk.

Second, the cost of risk depends on its correlation with other risks the participant bears. If it is exactly negatively correlated then it cancels out, which is why generators like vertically integrating with retailers - each perceives the risk of wholesale price movements in opposing ways. If there is no correlation, then the risk does not materially increase costs provided the holder has an adequately diversified portfolio. Given that wind is not correlated with anything economically very relevant, transferring that wind output risk to final consumers is essentially costless. Finally, transferring weather forecasting, marketing and balancing actions from large

10

numbers of small wind farm developers to a single large System Operator reduces transaction costs dramatically. The only reason to leave these costs upstream is if the generators are better placed to manage them. With controllable (fossil) generators, this is desirable as it gives them incentives to schedule adequate maintenance suitably and ensure reliable fuel supplies. Neither holds for uncontrollable wind (or sunshine).

That leaves the main cost to impose on RES-E as the efficient location cost – the cost of strengthening the transmission grid to deliver from that location, as well as incremental transmission losses. The aim should be that all generation locates to deliver power to final consumers at least total system cost (generation plus transmission). On that score the UK with its locational transmission charges fares quite well compared to almost all EU MS’s transmission charging systems.

The State Aid Guidelines (EC, 2014, hereafter the Guidelines) sets out the guiding principles and their interpretation (section numbers, §, and paragraph numbers are to the Guidelines). First, interventions or support must be “compatible with the internal market within the meaning of Article 107(3)(c) of the Treaty” so that they do not “affect trading conditions to an extent contrary to the common interest.” (§3 (23)). For the support to be “considered compatible with the internal market” the design of the aid measure must ensure “that the positive impact of the aid towards an objective of common interest exceeds its potential negative effects on trade and competition.” (§3.1 (26)). In addition, it must remedy a well-defined market failure that cannot be directly addressed (§3.2.2), it must be least distortionary (§3.3.3 (39)), provide appropriate incentives (§3.2.4) to change the behavior of the investor, without leading to profits in excess of comparable investments (§3.2.4.2) and be proportionate, i.e. least cost (§3.2.5). The irremovable (or at least hard to remove) set of market failures start with the inadequate carbon price (§3.3 (115)), followed by the production of otherwise unremunerated public good benefits from learning how to lower RES-E costs, and finally, the lack of credible future markets for the support schemes (including the future price of carbon), other than contracts.

On that score FiTs trump other forms of support, and should not require each RES-E producer to bear efficiently transferable marketing transaction costs. Much the same defense can be made in favor of power pools and central dispatch rather than energy-only power exchanges. However, the Guidelines (at §3.3.1, (125)) require “that that beneficiaries sell their electricity directly in the market and are subject to market obligations. The following conditions apply from 1 January 2016 to all new aid schemes and measures:

(a) Aid is granted as a premium in addition to the market price (premium) whereby the generators sell its electricity directly in the market.

(b) Beneficiaries6 are subject to standard balancing responsibilities, unless no liquid intra-day markets exist.

(c) Measures are put in place to ensure that generators have no incentive to generate electricity under negative prices.”

(127) “From 1 January 2017, the following requirements shall apply. Aid is granted in a competitive bidding process on the basis of clear, transparent and non-discriminatory criteria, unless: …” there are too few bidders or sites or the outcome would be higher support costs. “If such competitive bidding processes are open to all generators producing electricity from

6 Beneficiaries can outsource balancing responsibilities to other companies on their behalf, such as aggregators.

11

renewable energy sources on a non-discriminatory basis, the Commission will presume that the aid is proportionate and does not distort competition to an extent contrary to the internal market.”

It is not clear whether the premium can be a floating premium as with CfDs, and the let-out that balancing (and presumably marketing) can be delegated to an aggregator (perhaps the System Operator) might allow FiTs to be translated into an approved form of State Aid, with the advantage that the additional support costs of marketing and balancing would become an explicit part of the subsidy cost.

Finally, the favored system of financing RES-E support by imposing the costs on electricity consumers is fiscally illiterate. Given that the support is justified by the public good of driving down costs to benefit all future users of RES-E and hence the planet’s climate, the funds should be made available from general taxation, not inefficiently loaded on to the production sector, nor by raising the tax on one specific final good, electricity.7 The State Aid Guidelines (at §3.7.1) defines environmental taxes as those “are imposed in order to increase the costs of environmentally harmful behavior.” and that legitimately applies to carbon taxes, as recognized at §3.7.1 (180). §3.7.2 discusses reductions of taxes or charges designed to collect revenue to provide the public good renewable subsidies, but makes the invalid point that “they should be recovered in a way that does not discriminate between consumers of energy.” It (grudgingly?) accepts that “to avoid that undertakings particularly affected by the financing costs of renewable energy support are put at a significant competitive disadvantage, Member States may wish to grant partial compensation for these additional costs. Without such compensation the financing of renewable support may be unsustainable and, public acceptance of setting up ambitious renewable energy support measures may be limited. On the other hand, if such compensation is too high or awarded to too many electricity consumers, the overall funding of support to energy from renewable sources might be threatened as well and the public acceptance for renewable energy support may be equally hampered and distortions of competition and trade may be particularly high.” (183). To repeat, this violates the fundamental principles of good public finance. Auctions

The Guidelines advocate auctions to establish the least cost way of supporting RES-E, and the auctions for off-shore grid connections to Off-shore Transmission Owners (OFTOs) are a good example of the considerable reduction in the implied WACC that can be achieved by the combination of competition and clarity about what is being auctioned. We noted this above, and argued that if sites with planning and environmental permission and grid connections could be secured (after carefully examining the best places for on-shore wind) then each site could be auctioned off for the least cost of support. Even if this is not possible, then a multi-dimensional auction (Che, 1993) in which bidders specify a range of different packages of characteristics, such as the payments contingent on different aspects of the contract, such as the achieved capacity factor over the first three years, whether the subsidy is a capital grant or per unit output, and/or whether it transfers marketing and balancing to another party) might allow some rent transfer to consumers. Given the stress that the Guidelines place on the transparency and desirability of auctions, we suggest that work is needed to scope out suitable auction designs. 7 The counterpart to this is that energy should pay the standard rate of VAT, which automatically exempts the production sector and in the UK would go some way to making up the lost revenue from RES-E and other charges.

12

The balance between deployment and research support

EU MSs are providing massive subsidies to meet their 20/20/20 targets. The UK alone has capped (and expects) subsidies at £7.5 billion per year, and this excludes various costs that have been transferred to consumers, particularly the extra transmission investment costs, 73% of which are borne by consumers. In addition the original aim of the Renewables Directive was to minimize costs by encouraging trade in RES-E, but the very specific forms of support for each technology in each MS militates against that. At the same time the EU’s aspirations for the Strategic Energy Technologies (SET) Plan of trebling the level of R&D in these fields lacks adequate financial support other than modest additional EU funds.

Both the inefficiency of the RES allocation across MSs and the lack of finance for the SET Plan might be addressed by reforming RES Research, Development, Demonstration and Deployment (RDD&D) support. The RES targets were set with reference to GDP per head and the potential for additional RES. One possible change would be to replace these RES targets with an equivalent financial target (perhaps set equal to the cost of meeting the original RES target with an uplift for extra R&D described below). That money must be spent on a designated set of low-carbon energy RDD&D actions. The support would be justified in terms of the objectives motivating the RES Directive, which would have to be spelled out and agreed. These designated actions would include R&D for very immature but promising technologies (e.g. wave, tidal stream, far off-shore wind etc.), and hence address the SET Plan underfinancing. It would also include demonstrations where technologies and their costs need to be tested at scale (CCS, possibly biomass, off-shore wind), and perhaps the funds set aside for that could be rolled into the financial targets or provided in counterpart funds from the EU, such as the NER300 funding for CCS and innovative renewables. Finally, it would include the deployment for near-market technologies currently delivered by the RES targets, where deployment leads to learning-by-doing economies of scale (on-shore wind, solar PV).

The objectives and technology readiness would determine the optimal method for finance. For immature technologies EU-wide competitions are probably best, for demonstration plant tender auctions, and for near-market technologies, tender auctions for feed-in tariffs for MW of available capacity as discussed above.

The credit that each MS would receive for its actions would be suitably benchmarked, so that for e.g. solar PV, it would be credited with an annual value per MWp based on the extra annual revenue needed to justify installation in a reference sunny location (such as southern Spain compared to a CCGT there). Each MS would be free to offer (and be credited for) finance to support deployment in any location (possibly including approved developing countries, perhaps with a discount as with CERs that can be traded in the EU ETS). Benchmarking would remove the temptation to gold-plate domestic delivery of RES as covert industrial or employment policy) and would encourage optimal location. Conclusions

The main cost of decarbonizing the electricity sector is the cost of financing the very capital-intensive investments needed in generation and transmission. Transmission investments are regulated and enjoy low weighted average costs of capital (WACC), but RES-E and nuclear power are costly to finance in liberalized markets subject to uncertainty over future carbon prices

13

and policy risks. The solution is to provide contracts, and we have argued that good contract design allocates risks to those best placed to manage and bear them. With intermittent RES-E, that is almost certainly the System Operator who can offer a fixed payment per MW or MWh, ideally involving the lowest transfer of rent to the developers. Carefully designed multi-dimensional auctions are potentially the best way to reveal that and the format most consistent with the new State Aid Guidelines, although the German example of benchmarking contract length on revealed capacity factors may be an adequate and simpler substitute. The evidence suggests that contract credibility is key to reducing the WACC and hence support costs. Finally, there is arguably a growing mismatch between the large subsidies provided to RES-E support and the underinvestment in low-carbon R&D that has been exacerbated by privatization and liberalization. We suggest a way of both redressing that failure and improving the tradability of RES-E across MS borders, further lowering the cost of support. References Batlle, C. (2011). A method for allocating renewable energy source subsidies among final energy

consumers. Energy Policy, 39(5), 2586–2595. doi:10.1016/j.enpol.2011.02.027 Butler, L., Neuhoff, K. (2004) ‘Comparison of feed in tariff, quota and auction mechanisms to

support wind power development’, Cambridge Working Papers in Economics CWPE 0503at https://www.repository.cam.ac.uk/bitstream/handle/1810/131635/ep70.pdf?sequence=1

Che, Y.K. (1993). ‘Design Competition Through Multidimensional Auctions,’ RAND Journal of Economics, 24, 668-680.

DECC (UK Department of Energy and Climate Change) (2010). Electricity Market Reform: A consultation document, London: Department of Energy & Climate Change available at http://www.decc.gov.uk/en/content/cms/consultations/emr/emr.aspx

DECC (2011). Planning our electric future: a White Paper for secure, affordable and low-carbon electricity. London: Department of Energy & Climate Change, 12 July, available at http://www.decc.gov.uk/en/content/cms/legislation/white_papers/emr_wp_2011/emr_wp_2011.aspx DECC (2012a) Electricity Market Reform: Annex A: Feed-in Tariff with Contracts for Difference:Operational Framework London: Department of Energy & Climate Change

DECC (2012b) Government response to the consultation on proposals for the levels of banded support under the Renewables Obligation for the period 2013-17 and the Renewables Obligation Order 2012, London: Department of Energy & Climate Change. July, at https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/42852/5936-renewables-obligation-consultation-the-government.pdf

DECC (2013a) Electricity Market Reform: Annex F: EMR Panel of Technical Experts’ Final Report for DECC, London: Department of Energy & Climate Change, July, at https://www.gov.uk/government/.../emr_consultation_annex_f.pdf

DECC (2013b) Investing in renewable technologies – CfD contract terms and strike prices at https://www.gov.uk/government/uploads/system/uploads/attachment_data/file/263937/Final_Document_-_Investing_in_renewable_technologies_-_CfD_contract_terms_and_strike_prices_UPDATED_6_DEC.pdf

European Commission (2009). Directive 2009/28/EC of the European Parliament and of the Council, Official and COM(2010) 2020 Final of 3.3.2010 Available at http://faolex.fao.org/docs/pdf/eur88009.pdf

European Commission (2014). Guidelines on State aid for environmental protection and energy 2014-2020. Brussels at http://ec.europa.eu/competition/sectors/energy/eeag_en.pdf

14

EEA (European Environment Agency) (2013). Trends and projections in Europe 2013: Tracking progress towards Europe's climate and energy targets until 2020, EEA Report No 10/2013. Luxembourg: Publications Office of the European Union.

Great Britain. Energy Act 2013; Elizabeth II. (2013) [Online] Available at: http://www.legislation.gov.uk/ukpga/2013/32/contents/enacted/data.htm

Green, R and Staffell, I. (2013) ’Gold on them thar hills? Estimating wind farm rents in the UK’s Electricity Market Reform’ mimeo Imperial College Business School

Kanellakis, M., Martinopoulos, G., & Zachariadis, T. (2013). ‘European energy policy — A review,’ Energy Policy, 62, 1020–1030. doi:10.1016/j.enpol.2013.08.008 Lynch, S. (2014). ‘EU Commission refers Ireland to European Court of Justice over failure to meet renewable energy targets’, Irish Times, 24 January.

Mir-Artigues, P., & del Río, P. (in press). ‘Combining tariffs, investment subsidies and soft loans in a renewable electricity deployment policy,’ Energy Policy, doi:10.1016/j.enpol.2014.01.040

NERA (2013). Changes in Hurdle Rates for Low Carbon Generation Technologies at https://www.gov.uk/government/publications/nera-economic-consulting-report-changes-in-hurdle-rates-for-low-carbon-generation-technologies Newbery, D.M. (2012). ‘Reforming Competitive Electricity Markets to Meet Environmental Targets,’ Economics of Energy & Environmental Policy, 1(1), 69-82

Newbery, D.M. (2013) “Evolution of British electricity market and the role of policy for the regulation toward low carbon future”, Ch.1, pp. 3-33 in Evolution of Global Electricity Markets: New paradigms, new challenges, new approaches, edited by F.P. Sioshansi. Waltham, MA: Academic Press.

Stern, N. (2007). The Economics of Climate Change: The Stern Review. Cambridge and New York: Cambridge University Press

15

Table 1. Targets and Support Schemes for Renewables in the European Union by Member State

% share of renewable energy in gross final energy consumption

SUPPORT SCHEMES

COUNTRY 2012 2020 Target

Progress towards 2011/12 target

FiT FiP Investment subsidy/ Soft Loans

Quota system

Tax regulation Tenders Net

metering

EU (28 countries) 14.1 20 n.a. EU (27 countries) : 20 ↑

Belgium 6.8 13 ↓

ü ü ü Bulgaria 16.3 16 ↑

ü

ü

Czech Republic 11.2 13 ↔ ü ü ü

Denmark 26 30 ↔ Germany 12.4 18 ↑ ü ü ü

Estonia 25.2 25 ↑

ü ü

ü Ireland 7.2 16 ↔

ü

ü

Greece 15.1 18 ↑ ü

ü

Spain 14.3 20 ↑ France 13.4 23 ↓ ü

ü

ü ü

Croatia 16.8 20 n.a. ü

ü

Italy 13.5 17 ↑ ü ü

ü ü ü ü

Cyprus 6.8 13 ↔ ü

ü

Latvia 35.8 40 ↓ ü

ü

ü

Lithuania 21.7 23 ↑ ü

ü

ü

ü

Luxembourg 3.1 11 ↑ ü

ü

Hungary 9.6 14.65 ↑ Malta 1.4 10 ↓ ü

ü

ü

16

% share of renewable energy in gross final energy consumption

SUPPORT SCHEMES

COUNTRY 2012 2020 Target

Progress towards 2011/12 target

FiT FiP Investment subsidy/ Soft Loans

Quota system

Tax regulation Tenders Net

metering

Netherlands 4.5 16 ↓

ü ü

ü Austria 32.1 34 ↔

ü

ü

Poland 11 15.48 ↔

ü ü ü Portugal 24.6 31 ↔

ü

ü

Romania 22.9 24 ↑

ü ü Slovenia 20.2 25 ↑

Slovakia 10.4 14 ↑ Finland 34.3 38 ↑

ü ü Sweden 51 49 ↑

ü ü ü

United Kingdom 4.2 15 ↓ ü

ü ü ü

Iceland : 64 n.a. ü

ü ü

Norway 64.5 67.5 ↑

ü Switzerland : : n.a.

ü

Sources: EEA (2013) for columns 1-3 and Mir-Artigues & del Río (in press) for Support Schemes Notes: In the ‘Progress towards 2011/12 target’ column, ↑ indicates both the indicative and the expected trajectory were met, ↔ indicates the indicative trajectory for 2011–2012 was achieved, but the expected trajectory was not met and ↓ indicates neither the indicative and the expected trajectory were met; FiT = Feed in Tariff; FiP = Feed in Premium