Renewable energy mandates often accompany ambitious decarbonization policies such as Alberta’s recently announced Climate Leadership Plan. European experience shows that such mandates which generally include subsidized renewables (with near zero short-run marginal costs) can reduce conventional thermal generation facilities’ utilization rates. Importantly, when utilization rates fall, this reduces the economic viability and the incentives to invest in conventional thermal capacity. These diminished investment incentives sit uneasily beside the fact that some—perhaps substantial—thermal generation capacity will always be required to meet demand:  the wind does not always blow and the sun does not always shine. Lower prices and profits in the wake of introducing renewables might drive out some thermal capacity from the market. If this results in demand running ahead of supply, economic theory suggests that prices should rise again to induce thermal generation entry to meet demand. However, the economic literature has identified many factors—especially regulatory intervention and technological limitations—that mean that wholesale prices in electricity markets might not always provide adequate signals of scarcity. Available evidence suggests that subsidized renewables exacerbate this signaling problem. Capacity might not always be built just when it is needed.

In some electricity markets in Europe and the United States, regulators have instituted capacity markets or other mechanisms (including “command and control” mechanisms) that explicitly pay generators for making capacity available. This contrasts with Alberta’s “energy-only” market where generators are only paid for the sale of electricity. The impact of renewables has contributed to Europe’s growing concerns about long-run capacity investment. Interest in capacity mechanisms has correspondingly grown. The evidence that capacity mechanisms actually achieve their intended results, however, is unclear.

The need to provide thermal generation investors with higher and more certain prices to offset (renewables-induced) lower and less certain utilization further motivates interest in capacity mechanisms. But evidence from the country that has the most clear-cut capacity problem—the United Kingdom—shows that regulators continue to find high prices difficult to accept and to commit to. Regulatory recalcitrance means that capacity mechanisms may not achieve their intended results. Investors faced with current low electricity prices and a history of regulatory intervention to protect consumers from price spikes—even when such spikes stem not from the exercise of market power but from a genuine scarcity of generation resources—might well steer clear of investing in additional thermal generation.

Will renewables create similar challenges in Alberta’s electricity markets?  Ensuring that the lights stay on even when wind (the predominant renewable source in Alberta) generation is not available, means that meeting system contingencies requires as much thermal capacity in the future as is currently required.³  Retirement of coal plants means that a large amount of gas-fired generation (CCGTs and peaking facilities) will be needed to replace the coal. But these facilities will likely need to recover their costs—including substantial fixed costs—over fewer and more uncertain hours of operation in order to ensure that they are built. Under current Alberta plans, there will be a substantial period of time in which growing quantities of renewables coexist with substantial remaining coal capacity. While this means no immediate adequacy problem, it also presents a risk that depressed short-term pricing complicates investment in capacity required for the medium-term and long-term. Thermal generation investors will need to be confident that prices will rise when the new capacity they are constructing comes on the market, and by enough to accommodate lower and less certain hours of operation. This confidence is particularly important given that forward contracting in electricity markets does not yet provide a sufficient hedge against medium to long-term price risks.

This confidence could prove elusive if there is uncertainty about the retirement schedule of coal capacity; or if there is uncertainty about the quantity of renewables being procured in the future. Renewables mandates have sometimes been revised and expanded in Europe; anticipation of similar actions in Alberta could send precisely the wrong investment signal to investors. Greater certainty about coal and renewable quantities in the future likely assists investment in thermal generation. Abandoning the current price cap so that prices can rise to levels consistent with economic estimates of scarcity value at super-peak hours could also help. Of course, Alberta could institute capacity auctions or other mechanisms that effectively constitute contracts between the system administrator and generators specifically for capacity. But these mechanisms probably require significant institutional investment in their design; require considerable institutional discipline and commitment to function effectively; and may represent a substantial expansion in the extent to which market outcomes are no longer determined by the market, but by a patchwork of administrative interventions.

The initial wave of electricity restructuring in the late 1990s and 2000s in Alberta and elsewhere emphasized markets or market-like incentives to facilitate consumer choice, innovation and cost efficiency. Market prices were intended to provide the right signals about what capacity got built, and when. Climate change policies need not fundamentally alter this: many economists would prefer to let carbon pricing and emissions trading alone achieve the desired level of carbon abatement. But renewables policies, consistent with carbon abatement but wider-ranging in their socio-economic goals⁴, are here to stay. Given these policies’ focus on achieving an arbitrary penetration target for renewables and the subsidies involved, such policies are inherently not market-compatible. The market no longer chooses what gets built and when. It may be possible to engineer interventions that bring forth adequate investment, but it is doubtful that the energy-only market will continue to be central to achieving this investment. Indeed, if the process is mismanaged, renewables might fundamentally break this market and like Humpty-Dumpty, it may be impossible to put the market back together again.

Early and detailed consideration of renewables’ impact on adequacy and on investment incentives ought to be an important focus for Alberta’s efforts to adapt its electricity sector to the Climate Leadership Plan.  We hope that highlighting European experience will facilitate the process.

Missing Money and Scarcity Pricing

In energy-only markets, such as Alberta’s,⁵ suppliers must earn profits that are sufficient to cover the fixed costs associated with providing generation capacity, through the prices that they face in the power pool.⁶ Typical real-time markets rely on a relatively heterogeneous mixture of generating resources, with significant variation in the marginal costs of dispatch as between, say, a coal plant and a simple-cycle gas-fired peaking facility. This heterogeneity of resources generates an upward-sloping supply curve for electricity, with the market clearing price set by the marginal cost of the last unit dispatched when capacity equals or exceeds demand. There are infra-marginal producers who earn “quasi-rents”⁷ based on the fact that they receive a market clearing price that exceeds their own marginal cost. In peak demand hours, when demand bumps up against capacity constraints, prices should rise steeply. Since the demand side of electricity markets, to date, is generally inflexible (as most end-use customers do not or cannot react to real-time prices), prices ought to rise towards the “value of lost load” (what customers would be willing to pay to avoid service curtailment). Although generator profits rise sharply in these circumstances, profits earned from these “scarcity hours” might be critical to generators’ ability to recover fixed costs and earn a return on capital.⁸ This is likely particularly true for high and intermediate-cost generation facilities which operate for only a few hours each year. Economists and others believe that robust scarcity pricing signals are essential to ensuring long-term investment and thus long-term generation adequacy.

However, many electricity markets in North America have not let market forces entirely determine the price, at peak hours. For example, Hogan states:

“The missing money problem arises when occasional market price increases are limited by administrative actions such as offer caps, out-of-market calls, and other unpriced actions. By preventing prices from reaching high levels during times of relative scarcity, these administrative actions reduce the payments that could be applied towards the fixed operating costs of existing generation plants and the investment costs of new plants.”⁹

Alberta imposes a price cap on offers of $999.99 per MWh¹⁰, similar to many other North American jurisdictions. Joskow describes the typical $1000 per MWh price cap used in many U.S. jurisdictions as being “clearly below what the competitive market clearing price would be under most scarcity conditions.”¹¹  He also describes other aspects of electricity markets that create the “missing money” problem—which he defines as wholesale markets producing total revenues that are too low to support investment in an efficient (least-cost) portfolio of generating capacity: for example,  demand that does not respond to real-time prices, resulting in actions taken by regulators such as “must offer” obligations to control price spikes;¹² “out of market” calls; and voltage reductions to avoid rolling blackouts in times of scarcity.¹³

Additionally, at times of peak system demand, with little real-time price response, prices may not increase—as economic efficiency suggests that they should—to reflect the willingness of consumers to pay to avoid curtailment rather than the marginal cost of the last generator dispatched.¹⁴  Hogan describes this facet of conventional energy-only markets as a de facto price cap.¹⁵

Renewables: Merit Order and Market Power Effects

The economic literature identifies two theoretical and mutually offsetting effects of introducing large quantities of renewables into the generation mix:

A price-depressing Merit Order Effect. This effect arises because a large quantity of zero-marginal cost renewables are added alongside low-marginal-cost base-load and higher-marginal-cost energy sources such as CCGTs and simple cycle peaking facilities. The effect of this is to shift out the supply curve to the right, and by doing so, depress the market-clearing price for any given level of demand. In economic terms, renewables (absent withdrawal of other capacity from the market) can substantially reduce the quasi-rents available to existing conventional facilities. This might exacerbate the perhaps inherent scarcity pricing problems associated with current electricity markets.¹⁶

A price-enhancing Market Power Effect. Many generation markets are at least somewhat concentrated. In these concentrated markets, if conventional thermal generation owners are also diversified into renewables, they may have enhanced incentives to withhold supply from the market if those renewables receive market prices.¹⁷ Each generator is a monopolist on its own “residual” demand curve and will trade off elevated profits on infra-marginal units against lost sales of marginal units. Withholding (economically or physically) supply of otherwise “in-merit” facilities induces a higher market-clearing price and thus higher rents on infra-marginal facilities, most especially renewables. Even though renewables have well-known intermittency issues and even though forward markets theoretically mitigate incentives to exercise market power, the economic literature shows that diversified firms will have increased incentives to exercise market power when renewables are introduced into the market.¹⁸

In the European setting, the empirical literature unambiguously supports a dominant merit order effect (we discuss this in the following section). This may be linked to many countries’ choices (in particular that of Germany) to use feed-in tariffs, wherein payments to renewables are not linked to the market price for electricity. European experience of premature asset retirements and mothballing also supports the idea of straightforward stranded assets problems induced by renewables policy. The introduction of a large quantity of renewables has rendered some existing and even some brand new thermal generation facilities uneconomic.

Even if the market power effect fully offsets the merit order effect, the increased incentive to exercise market power is not a good market outcome. Scarcity pricing in a competitive market allows generators to earn quasi-rents but in a fashion consistent with allocative and dynamic efficiency. The exercise of market power, on the other hand, preserves generators’ profits, but only at the expense of allocative efficiency—too little generation is supplied to the market. Further, it is likely to produce dynamic inefficiencies. The “long-run equivalent” of withholding is simply not to invest in the most frequently withheld types of generation capacities (assuming that there are non-trivial barriers to new entry in generation). These could be mid-merit generation sources, such as CCGTs.

Europe’s Experience

The EU’s 2008 renewable energy directive bound member states to national renewables targets in the context of an EU-wide objective of achieving 20 per cent of final energy consumption from renewable sources by 2020.¹⁹ Generous subsidies helped countries make substantial progress towards these targets. Between 2005 and 2014, renewables’ share of electricity generation grew from about 15 per cent to almost 30 per cent (at a compound annual growth rate of 7 per cent). In the five largest electricity markets in Europe, renewable share of electricity grew at 10 per cent per annum over the same period.²⁰  Although country-level progress varies, the EU is expected to meet the “20 per cent by 2020” target. A new target of 27 per cent of final energy consumption by 2030 has thus been set.²¹

A growing body of literature highlights two major effects from increased renewable generation on electricity producers in some countries:

1. Wholesale electricity prices are reduced (and can be more volatile); and
2. The incentives for investment in new thermal generation have been reduced, with implications for future system reliability.

In 2014, the European Commission wrote:

“Increasing amounts of electricity generated from wind and solar have also exerted downward pressure on wholesale prices particularly in regions with high shares of these renewable energy sources …”²²

Various authors have analysed ex-post data on electricity prices and renewable capacity in several countries with significant amounts of renewable generation with similar conclusions on the direction of effect.²³  They have identified an increased correlation between the availability of wind generation and electricity prices and so confirm the primary merit order effect and that this has (other things being equal) reduced electricity prices. The studies (because of differences in methodology) are difficult to compare but the estimated effects, as measured, in some markets (where renewable penetration is high such as Spain and Germany) have been very significant.²⁴

Renewable generation’s rise has displaced fossil-fuelled generation. Given relatively (compared to natural gas) cheap coal prices in Europe, natural gas generation has suffered the most displacement, with coal-fired generators able to protect their position in the merit order in the short-run.  Traber and Kemfert and Van den Bergh et al find that financial support for renewable generation may have also dampened EU emissions prices by reducing demand for fossil fuel generation and consequently lowered demand for emissions credits.²⁵ As noted, an unintended consequence is that lower emissions prices have also disproportionately benefitted coal generation over gas-fired generation in the merit order. After a long period of growth, electricity generation from natural gas peaked in 2008 and has declined since.²⁶

Falling prices, resulting from renewable generation, has in some countries contributed to a decline in quasi-rents available to thermal generation. Industry association, Eurelectric, concluded that incorporating renewable energy supplies reduced the operating hours and profitability of thermal generation. It also found that scarcity pricing in the fewer remaining operating hours has, “generally not been enough to cover the costs of ‘peaking’ plants (such as CCGT).”²⁷ This results from the increased margin of supply over demand that has resulted from increased renewable energy supply and a contemporaneous decline in demand from the financial crisis and ensuing recession. Renewable energy supplies have effectively increased the share of load being supplied by notionally baseload plant which has made peak hours more competitively traded.

Fewer profitable generating hours, to which renewables (in some markets) have in part contributed, for thermal – and in particular gas generation—have predictably led to plant mothballing and closures. Caldecott and McDaniels²⁸ reported write-downs for natural gas power assets to six major utilities of €6 billion. IHS estimated that 21GW of natural gas fired power plants was closed between 2008 and 2014.

These developments helped inform concerns – also raised by Eurelectric—about growing long-term capacity challenges: the EU worries that the current diminished economics of thermal generation provides little indication of future capacity needs (EU 2014). Long lead times in power sector investments mean that today’s depressed and uncertain market economics for thermal generation provide a significant disincentive to invest. The EU also found that price caps and other measures such as operating reserves, emergency demand response, and voltage reductions are suppressing price signals in hours of scarcity which otherwise might have signalled the need for capacity investment. Most countries face no imminent capacity crunch (Britain excluded). But the impending retirement of ageing coal and nuclear facilities could change that picture.²⁹

Policy Responses: Capacity Mechanisms and Capacity Markets

The growing adequacy-related concerns discussed above have motivated increased policy-making interest in mechanisms designed to ensure adequate generation capacity. This is particularly true of the interest around “capacity mechanisms” that specifically reward generators for capacity rather than energy. Such mechanisms were originally developed in several U.S. and European electricity markets as a response to the “missing money” problem.³⁰ The advent of a large amount of subsidized and highly intermittent generation capacity has been one important contributory factor to the sustained interest and innovation in capacity mechanisms in Europe in recent years.³¹ In effect, capacity mechanisms reflect the belief that electricity market prices cannot provide thermal generation investors with the assurance of higher prices for future capacity to offset renewables-induced lower utilization and higher uncertainty.

The European Commission’s recent study of capacity mechanisms—impelled by the increased salience of these mechanisms³²—identifies two broad types of capacity mechanisms: (a) targeted and (b) market-wide. In the former case, system administrators determine how much capacity is required over and above what the market would provide. System operators then either provide payments (at an administratively determined price) for specific types of capacity, or hold tenders to elicit the required capacity. Alternatively, system operators can procure capacity through a centralized auction, or they can require electricity suppliers or retailers to contract for top-up capacity with generators. The system operator can also construct estimates of how much capacity is required on a going-forward basis and pay would-be capacity providers on the basis of its estimates of the cost of providing new capacity. These various mechanisms differ substantially in terms of the degree to which they represent a genuine “market” for capacity—indeed some are simply “command and control” processes.

The experience of the United Kingdom is perhaps the most interesting development in European capacity mechanisms and there are some analogies to the situation in Alberta. The UK, perhaps unlike mainland Europe, needs new capacity and it needs it in the short to medium term. It committed to completely phase out coal generation by 2025. Until recently coal was responsible for about 30 per cent of UK electricity production. Some plant are retiring over the next few years. For the remaining plant, this commitment was made contingent on new gas-fired power plants being developed. However, the electricity energy-only market and other short term measures used by the system operator for system reliability are not providing sufficient incentives to invest in new large-scale capacity.

The UK has conducted two capacity auctions to date (in 2014 and 2015) for capacity in 2018 and 2019 respectively with longer term contracts being available for new generation. Both auctions, however, cleared at prices well below what is considered to be necessary to build a new CCGT.  In fact only two CCGTs, one of which was already in development, received long-term capacity contracts.³³ The only other new plants have been diesel-fired peaking plants. This was not necessarily the outcome intended when the auctions were conceived.

This outcome reflects a balance the Government struck between the amount of capacity to procure and prevailing concerns about the cost of electricity to consumers. It also highlights the fact that a capacity market brings another set of non-market interventions which may not, without significant design consideration, deliver new capacity as required at least cost to consumers. If the introduction of a large quantity of renewables to Alberta’s market has the same effect on thermal generators’ investment incentives, instituting capacity auctions or capacity markets does not represent an easy-to-design “fix”.

Does European experience hold lessons for Alberta?

If there are lessons to be drawn from Europe’s experience with renewables, what are those lessons and how might they direct policy-makers in the development of Alberta’s electricity system?

Economists generally seem to accept that the positive supply shock from renewables worsens the “missing money problem”.  European experience indeed shows that there are lower prices and growing concerns about how to effectively ensure long-term adequacy and reliability. These concerns are focused on the incentives of thermal generators, perhaps particularly natural gas fired plants, to add capacity over a multi-year period. Generators in Europe have responded to a renewables-assisted supply shock (creating temporary over-supply) with rationalization of gas facilities and gas investments. Leaving aside the policy preference for gas over coal, this short-run rationalization is not creating an adequacy problem in the short run. Europe’s concern is that investors might, however, continue to rationalize for as long as they think prices will remain low.

As the European Commission’s recent Staff Working Document points out, it is difficult to calibrate the timing of capacity investments with actual surpluses or shortages in the energy markets. Generators might not respond to shortages until the shortage becomes apparent, and critically, until it is reflected in actual energy market prices.³⁴ Low prices, today, resulting from renewables might, therefore, embed expectations that prices will be lower for longer (than rationally might be expected). This may dampen thermal generation investment intentions, particularly if investors are already scarred by their experience of the regime changes that renewables have wrought. Renewables also add incremental uncertainty to the investment decision-making process, particularly if renewables targets and procurement mechanisms are continually revised. These effects are all the more pronounced because electricity forward markets are insufficiently liquid to handle contracts for a significant volume of delivery over the long-term.

There are straightforward differences between Alberta and Europe: a preference for coal rather than gas investments is a problem that will not arise in Alberta. But there is no denying that renewables will make investment in thermal resources less attractive simply because of lower and less certain operating hours for thermal generators. Timely and adequate investment in these facilities will depend on investors’ confidence that prices will ultimately rise, and crucially that they will be allowed to rise as necessary. Europe’s renewables push was substantially crafted without regard to incentives in the restructured electricity market. Consequently, European countries have had to adapt and develop or enhance institutions such as capacity markets in an effort to cope with the aftermath. The results of this continuing institutional improvisation are unclear. Alberta has an opportunity, however, to consider the role of policy factors that might facilitate investment within the context of the energy-only market. We offer three considerations below, based on Europe’s experience and the economic literature, which might be relevant to Alberta’s transition. These considerations are crafted with the energy-only market in mind. There are, of course, other policy responses such as capacity markets and their design, but these are farther-reaching than the considerations we detail below.

Consideration 1: Coordination between Quantity Commitments for Renewables and Coal

Europe’s lack of extensive coordination in matching renewables introduction with the retirement of some conventional base-load partly accounts for the reduction in available quasi-rents and reduced investment incentives.

Alberta’s proposed retirement of coal plant and its replacement (in part) through renewable generation might benefit from careful coordination. Alberta might specify a schedule that lays out what quantity of renewables will be procured and when. The more stable are investors’ expectations about the quantity and timing of renewables introduction, the easier it is to predict future prices and adjust investment decisions accordingly. Obviously such commitments would need to be credible in order to be effective. The temptation to frequently revise renewables targets would need to be resisted, for instance.³⁵

Providing certainty and commitment as to the schedule of base-load retirements improves, at least marginally, market participants’ ability to anticipate future prices and capacity availability. The commitment to retire coal capacity expeditiously, but in an orderly fashion, will offset some of the possible supply shock effects of introducing renewables to Alberta’s market. In practical terms, given what has already been proposed, this means sticking to the 2030 sunset date for coal. An unpredictable and chaotic process would harm investor confidence—this is perhaps the danger that Alberta faces if investors perceive the ongoing unwinding of PPAs as chaotic and subject to political uncertainty.³⁶ A structured transition ought to be feasible through the periodic renewables auction process as envisaged by the AESO.

Consideration 2: Rethink Price Caps

If investment incentives do emerge as an issue, Alberta might consider revising the current price cap towards a level more consistent with estimates of the Value of Lost Load (VOLL). With potentially fewer hours of scarcity resulting from increased renewables generation, it will be more important to allow for the market price to effectively price that scarcity. While prices will only ever hit such levels in a few hours every few years, the profits from such scarcity hours could be very important in sustaining peaking capacity that is required in scarcity hours, while increasing quasi-rents for low and intermediate-cost capacity.

Alberta could also allow demand response to participate in any future capacity mechanisms as the U.K. has, but demand response’s ability to be an effective participant depends on technological progress substantially determined outside Alberta’s control. Also there is the risk that demand response that does not have stringent performance requirements attached to it will do little for reliability, while damping price signals for thermal generation.

Consideration 3: Role of Market Prices in the Renewables Auction

The design of any renewables auction may also need to consider whether payments to renewables providers should or should not be linked to the market price. There is increased pressure in Europe to expose renewable generation to market forces in order to reduce the cost of subsidies and transition those technologies ultimately into the market. Recent literature suggests the possibility, however, that during hours when renewables are generating, diversified generators’ incentives to exercise market power increase. This incentive to withhold arises because of diversified generators’ ability to earn higher margins on their infra-marginal renewables capacity when they engage in withholding.

While this has the effect of restoring prices and offsetting the merit order effect, it does so inefficiently, through the exercise of market power. If the magnitude of this inefficiency is large relative to the benefits from tying renewables prices to market prices, then the auction design might be modified such that bidders bid on their costs and not the difference between their costs and their expectation of future market prices. In any case, any given market participant’s expectation of future market prices will depend on the aggregate quantity of renewables that it expects to be supplied in future years, and on the strategic response of owners of thermal generation to the introduction of renewables. Needless to say this is a tough calculation to make, even if commitment to providing quantity certainty with respect to renewables may make it slightly easier.³⁷ Additionally, persistently low price expectations would mean high subsidy bids and thus defeat any benefit arising from the linkage between subsidies claimed and market prices.

Alberta may consider designs such as the UK in which an auction is held to determine the price support that will be guaranteed to winning producers over a fixed time period. It is structured, however, as a fixed-for-float swap. As the market price rises, the level of subsidy reduces leaving aggregate support to the producer unchanged. This appears to be the option preferred by the AESO.

Finally, recent political developments in the United States cast some doubt on the ability of Canadian Federal and Provincial governments to stick to currently announced climate change mitigation initiatives. This makes the structured approach, outlined above, more relevant to the design process.

1.   Berkeley Research Group LLC.
2.   DeGroote School of Business, McMaster University. The opinions expressed in this paper reflect the views of the authors, not the corporate views of Berkeley Research Group or McMaster University, or the views of other individuals associated with these institutions. Responsibility for errors and omissions rests solely with the authors. We are indebted to Matthew Barmack for providing comments on this paper.
3.   The Alberta Electricity System Operator (AESO) currently assigns a zero rating to wind capacity in calculating the availability of a supply cushion to meet contingencies. See Alberta Electric System Operator, “Long-Term Adequacy Metrics” (August 2016) at p 10.
4.   These policies are frequently described as “complementary” to carbon pricing and emissions trading, but are best understood as having goals such as attracting investment, fostering innovation, and boosting economic development that go beyond mere carbon abatement. This paper does not comment on the overall desirability of renewables.
5.   Alberta operates both a real-time market and a day-ahead market (the latter for ancillary services). There is some use of financial instruments—e.g., forward contracts—by market participants, but our understanding is that forward market volumes are relatively small. Alberta’s expectation is that market participants can exchange electricity on non-discriminatory terms and manage spot market volatility through appropriate use of financial instruments. Although Alberta has some ex-ante limits on generators’ market power (e.g., a $999.99 per MWh bidding cap and a limit on any one firm controlling more than 30 per cent of generation capacity), the mere exercise of market power (“extraction”) is not censured.
6.   The term “power pool” is used to reflect Alberta’s specific circumstances.
7.   In this context, quasi-rents are defined as the margins that firms need to earn to pay back the fixed costs incurred in providing generation capacity.
8.   If demand is elastic and participates in wholesale markets, then the marginal flexible load might require a payment close to VOLL to curtail voluntarily.  This marginal load will set the market price.
9.   William W. Hogan, “Electricity Scarcity Pricing through Operating Reserve”, (2013) 2:2 Economics of Energy and Environmental Policy at 1.
10.   Since Alberta expressly does not censure the mere exercise of generator market power, both scarcity rents and monopoly rents are potentially available to generators. Monopoly rents are earnings in excess of long-run average cost (i.e., more than what is required to generate a normal return on capital) by firms that are able to materially influence the market price by their choice of output (or in the context of the power pool, their bidding strategy). By contrast, scarcity rents are consistent with competitive markets, with scarcity pricing signaling the opportunity cost to society of not providing generation capacity in hours of scarcity. It is erroneous to conclude that economic withholding—the exercise of market power usually causing too little output to be supplied to the market—is an offset to scarcity rents in peak hours. In any case, incentives to withhold are highest at times of scarcity.
11.   As an example: London Economics on behalf of the UK Department of Energy and Climate Change investigated VOLL across different customer segments and found a load weighted average of £16,940 (roughly $29,000 at current exchange rates) for domestic and small/medium sized commercial customers for peak winter workdays in Great Britain. See London Economics, The Value of Lost Load (VOLL) for Electricity in Great Britain (London: London Economics, 2013) at p 54.
12.   These obligations reflect regulators’ concerns that scarcity conditions, in the presence of a vertical demand curve, present inviting opportunities to exercise market power by withholding supply from the market.
13.   Paul L. Joskow, “Capacity Payments in Imperfect Electricity Markets: Need and Design” (2008) 16:3 Utilities Policy at 16-18.
14.   “Conventional” markets have lacked one of the desired features of an efficient, idealised energy-only market: demand-side response. See Joskow, supra note 13 at 161, for a description of the four conditions that characterize an energy-only market that does not suffer from the “missing money” problem. One of these conditions is that there are both price-sensitive and price-insensitive consumers.  Another is that retailers can offer consumers contracts that specify the conditions under which they can be rationed. Historically, at least, real-time metering technology has not existed to satisfy these conditions.
15.   William W. Hogan, On an ‘Energy-Only’ Electricity Market Design for Resource Adequacy (Cambridge: Havard University, 2005), online: <https://www.hks.harvard.edu/fs/whogan/Hogan_Energy_Only_092305.pdf >. In theory, it should be possible to construct an appropriate demand curve for operating reserves and thus offset the missing money problem by effectively “completing the market”.  Hogan writes: “The absence of an appropriate operating reserve demand curve is one of the difficulties in market design that result in de facto price caps and missing money”.  He adds, however, that if the “reserve demand curve does not raise prices towards VOLL (the value of lost load) when operating reserves approach the minimum then the demand curve is not capable of representing… the true opportunity cost at the margin”.
16.   Looked at another way, renewables reduce thermal generation facilities’ average utilization rates. Theoretically, prices could rise by enough in the hours when such facilities actually operate that it could offset the lower utilization rates. This scenario likely requires the retirement of substantial amounts of existing capacity and relaxed regulatory policies towards high or even very high peak or super-peak prices. As discussed below, this has not happened in Europe.
17.   In many markets in the US and Europe (and in Ontario), however, renewables are effectively under long-term contracts or feed-in tariffs that do not face spot market pricing.
18.   Acemoglu et al confirm our intuition in this regard. Assuming Cournot competition between thermal generators, and assuming that these thermal generators’ portfolios include renewables, they find that strategic withholding of output dulls or even fully neutralizes the merit order effect (rightward shift in the supply curve) that the empirical literature on renewables discusses. The existing literature on power pools suggests that the Cournot assumption is a reasonable approximation to competitive behaviour among generators (see, for example, Bert Willems et al, “Cournot versus supply functions: What does the data tell us?” (2009) 31:1 Energy Economics at 38–47). When all thermal generators are vested in renewables—“full diversification” – the merit order effect of renewables is fully neutralized. See Daron Acemoglu et al, Competition in Electricity Markets With Renewable Sources (Cambridge: MIT, 2015), online: MIT <https://asu.mit.edu/sites/default/files/documents/publications/MAIN-submit.pdf>.
19.   They are also each required to have at least 10 per cent of their transport fuels come from renewable sources by 2020; National renewables targets have ranged from 10 per cent in Malta to 49 per cent in Sweden, European Commission, Renewable Energy, online: EC <https://ec.europa.eu/energy/en/topics/renewable-energy>. Countries also developed renewables action plans that included sectoral renewable energy targets and goals for different mixes of renewables deployed.
20.   Eurostat, “Electricity generated from renewable sources”, 2016, European Commission, online: EC <http://ec.europa.eu/eurostat/tgm/table.do?tab=table&plugin=1&language=en&pcode=tsdcc330>: “Electricity produced from renewable energy sources comprises the electricity generation from hydro plants (excluding pumping), wind, solar, geothermal and electricity from biomass/wastes. Gross national electricity consumption comprises the total gross national electricity generation from all fuels (including auto production), plus electricity imports, minus exports.” Top 5 markets selected by total consumption of electricity.
21.   European Commission, 2030 Energy Strategy, online: EC <https://ec.europa.eu/energy/node/163>.
22.   European Commission, A policy framework for climate and energy in the period from 2020 to 2030 (2014) at p 9.
23.   See Klaas Würzberg et al, “Renewable generation and electricity prices: Taking stock and new evidence for Germany and Austria”, (2013) 40 Energy Economics, for a comparative survey of recent literature on merit order effects in various European electricity markets.
24.   Hugo A. Gil et al, “Large-scale wind power integration and wholesale electricity trading benefits: estimation via an ex post approach” (2012) 41 Energy Policy at 849–859; Sensfuss et al,  “Analysen zum merit-order effekt erneuerbarer energien: Update für das jahr 2010” (2011) Frauenhofer ISI, Karlsruhe, estimated an effect in the period 2006-2010 of about €6/MWh in Germany (equivalent to roughly as much as 10 per cent of the prevailing baseload price).
25.   Thure Traber & Claudia Kemfert , “Impacts of the German Support for Renewable Energy on Electricity Prices, Emissions and Firms” (2009) 30:3 The Energy Journal at 155–178;  Kenneth Van den Bergh et al, “Impact of renewables deployment on the CO₂ price and the CO₂  emissions in the European electricity sector” (2013) 63 Energy Policy at 1021-1031.
26.   Eurostat, Gross electricity generation by fuel, GWh, EU-28, 1990-2013, online: EC <http://ec.europa.eu/eurostat/statistics-explained/index.php/File:Gross_electricity_generation_by_fuel,_GWh,_EU-28,_1990-2013.png>.
27.   Eurelectric, “RES Integration and Market Design: Are Capacity Remunerations Mechanisms Needed to Ensure Adequacy” (2011) at p 4.
28.   Ben Caldecott & Jeremy McDaniels, “Stranded generation assets: Implications for European capacity mechanisms, energy markets and climate policy” (2014) Smith School of Enterprise and the Environment Working Paper, online: <http://www.smithschool.ox.ac.uk/research-programmes/stranded-assets/Stranded%20Generation%20Assets%20-%20Working%20Paper%20-%20Final%20Version.pdf>.
29.    Many nuclear plant in Europe will be over 30 years old by 2020, while some countries such as the Netherlands and the UK have issued explicit coal retirement mandates.
30.   The economic literature identifies several other potential solutions to the inherent scarcity pricing conundrum of EOMs. Among these are (a) relaxing or abolishing administratively imposed price caps, (b) allowing prices to rise to the price cap level whenever out-of-market resources are called upon to generate, (c) enhancing demand-side response mechanisms in the market. Another approach is the development of a market for operating reserves.
31.   European Commission, Interim Report on the Sector Inquiry on Capacity Mechanisms: Commission Staff Working Document, (Brussels: 13 April, 2016), at 4, 12, 30, and nn 23, 36. These excerpts highlight the belief that renewables worsen the missing money problem.
32.   Capacity mechanisms raise concerns about so-called “State Aid”, which was the proximate reason for the Commission’s investigation. However, the Commission Staff Working Document, supra note 31, discusses the deep-seated economic causes at 4:

    “The large-scale roll-out of renewables combined with the overall decline in demand and the decreasing cost of fossil fuels have curbed the profitability of conventional generators and reduced incentives to maintain existing power plants or invest in new ones. In many Member States, these developments have been accompanied by increased concerns about security of supply. Member States are concerned that the electricity market will not produce the investment signals needed to ensure an electricity generation mix that is able to meet demand at all times…Some Member States have reacted by taking measures designed to support investment in the additional capacity that they deem necessary to ensure an acceptable level of security of supply. These capacity mechanisms pay providers of existing and/or new capacity for making it available.”

33.   ICIS, “UK CCGT developers keep faith with capacity market” (2016), online: ICIS <http://www.icis.com/resources/news/2016/04/22/9990602/uk-ccgt-developers-keep-faith-with-capacity-market/>.
34.   The European Commission Staff’s Working Document states, supra note 31 at 25, that expectations about future prices are more important to investment decisions than current prices. In many other commodity markets, well-developed futures markets offer an offset against the uncertainty inherent in long-term pricing expectations. This is perhaps less true for electricity.  The Commission Staff also notes (n 37, citing De Vries) that given uncertainty about future prices, investment decisions could be delayed so as to result in significant periods of actual shortages. The EC’s document thus points to two distinct but inter-related possibilities. First, there is the problem that adjusting from one equilibrium to another is not frictionless. Unlike in other long-cycle commodity industries—e.g., oil sands—there is a compelling policy interest in avoiding a disequilibrium that leads to actual shortages of electricity. Thus at a minimum policies that contribute to uncertainty and thus create frictions in the adjustment process should be avoided. Second, there is the problem that investment levels in equilibrium might be inefficient. Even if regulators avoid actual physical scarcity, can they do so in the efficient (least-cost) fashion?
35.   This may suggest that the renewables process be governed by an agency that is not incentivised one way or another by achieving higher renewables penetration.
36. The Alberta AESO’s 2016 Long-Term Outlook does provide details of an assumed retirement schedule for coal and assumptions about wind capacity additions in the future. AESO, AESO Long-Term Outlook (2016), online: AESO <https://www.aeso.ca/grid/forecasting/>.
37. Additionally, persistently low price expectations would mean high subsidy bids and thus defeat any benefit to the government arising from the linkage between subsidies claimed and market prices.