Older Archives

Texas Avoids Capacity Market While France Succumbs

November 29th, 2012 by Fereidoon Sioshansi, EEnergy Informer

Few would argue with the claim that Texas has the best wholesale and the most competitive retail electricity market in the US, if not the world. But it suffers from an ailment called inadequate resource adequacy. Private investors, who are the key stakeholders in building generation capacity, appear unable and/or unwilling to invest sufficiently in what the regulators and the market operator would like to see in terms of a comfortable reserve margin.

The last 2 years, the Electric Reliability Council of Texas (ERCOT) has had a few close calls, made more pronounced by a prolonged drought, which increased summer air conditioning load to dangerous levels. ERCOT had to repeatedly call for voluntary conservation to make it through hot summer afternoons in 2011.

Going forward, ERCOT has warned that it expects its reserve margin to plunge to 9.8% as early as 2014, 6.9% in 2015, and to a negative margin by 2022. Making matters worse, ERCOT has set a 13.75% target for how much excess generation capacity it would like to have above the projected demand – don’t ask where this number comes from or why it is as high as it is. One explanation often heard is that ERCOT, being an electric island – i.e., not being interconnected to other grids in the US – is unusually vulnerable to supply shortages, and hence the need for the large reserve margin.

Understandably, the Public Utility Commission of Texas (PUCT), has been sympathetic to ERCOT’s plight, nor do any of its commissioners wish the lights to go out under their watch. In late October 2012, they decided to raise ERCOT’s offer cap from the current $4,500/MWh – the highest of any organized market in the US – to $5,000 in June 2013, $7,000 in June 2014, and $9,000 in June 2015. That would still be lower than the AUS$12,500 offer cap in Australia’s National Electricity Market (NEM), but not by a huge margin.

Offer cap, of course, is the maximum amount that generators can bid in the wholesale auction. The higher the offer cap, the more money generators, especially those with flexible peaking units, typically natural gas-fired plants, can make during those few hours when they are dispatched. This has been identified among the factors prohibiting more investment in new peaking generation, resulting in the dreaded resource adequacy and dangerously low reserve margins predicted by ERCOT.

Another alternative examined but rejected by the PUCT was the introduction of a capacity payment scheme, similar to the one in PJM market in the mid-Atlantic region.

France, on the other hand, has decided to introduce such a scheme starting in 2013:

Concerned about having sufficient capacity during periods of high demand, the French government is introducing a decree that would require electricity suppliers to buy and pay for peak load capacity from the end of 2013. The obligatory certificates would have to match their forecast needs 3 years in advance. The scheme is intended to discourage the closures of uneconomical plants and encourage generators to construct new flexible plants that can be easily turned on and off.

Once introduced, generators will be paid a premium based on their available peak load production capacity, not just electricity sales. According to sources familiar with the scheme, suppliers will have to secure sufficient capacity certificates to meet their demand projections using a competitive auction.

France’s power grid operator, RTE, claims that the country’s supply and demand balance has been weakening and may be in jeopardy by 2016 due to the planned closure of older thermal plants in response to European Union’s environmental directives. E.ON has announced it will close 4 of its 7 coal-fired power plants in France by end-2015 (see related article on E.ON in this issue).

The main challenge in France is not lack of capacity per se, but rather lack of flexible capacity. The French system is dominated by 58 nuclear reactors, providing 75-78% of the generation in a typical year. What it needs is more flexible capacity, the type provided by peaking gas-fired units to meet peak demand during periods of high demand, such as evening hours, especially during cold winter months when a significant electric heating load exacerbates spikes in the residential load.

The French move to establish a capacity payment scheme is closely watched elsewhere in the continent, including in Germany. BDEW, the lobbying group for the German power sector, has indicated that a similar scheme will also be needed for Germany by 2015, if not sooner, to manage the rapid growth of renewable resources.

The debate about paying or not paying for capacity in competitive wholesale markets, of course, is nothing new. For competitive markets to work, existing generators must make sufficient money to remain viable. Attracting investments for new capacity to replace retiring generation or meet growing demand is even more of a challenge since investors want some degree of assurance that they can recover their initial investments plus a reasonable margin over time allowing them to make a profit. Otherwise, why bother.

There are two schools of thought on this. Some experts believe that energy-only markets, where generators are only paid when they are dispatched, are adequate, provided the offer cap is sufficiently large to allow them to make a killing during times of scarcity. These people say high offer caps are necessary to attract new investment, especially for new peaking plant. ERCOT and Australia’s National Electricity Market (NEM) are often mentioned as examples of successful energy-only markets.

Another group of experts believe that there are inherent limits and/or flaws to energy only markets, usually referring to “missing money problem,” the fact that not all generators make enough to remain viable – it is a rather long and complex story. These experts have proposed markets with both energy and capacity payments. The main objective is to pay sufficient amounts to keep some of the marginal thermal plants at least marginally viable.

This can be done through a capacity payment for having capacity available just in case they are needed. In this case, some generators may get paid even if they are rarely, if ever, dispatched. This, of course, introduces perverse incentives to keep old and inefficient plants around just to get paid for having capacity available when/if needed.

The Texas commission’s recent decision did not surprise many given a report prepared by The Brattle Group in June 2012. The Brattle report concluded that new capacity investments in ERCOT were impeded by low wholesale prices due to low natural gas prices and an insufficient number of peaking units were being built partly due to limits on the offer caps.

Increasing the offer caps to $9,000/MWh would help somewhat but would only raise the reserve margin to around 10%, not 13.75% as recommended by the Federal Energy Regulatory Commission (FERC) and sought by ERCOT. The Brattle Group recommended instead that either the market design be adjusted or reliability objectives be revised downward.

The Brattle report also suggested that ERCOT’s higher reliability target could be achieved with a significant increase in price-responsive demand, also referred to as Demand Response but cautioned that “it would likely take several years before a sufficient level of demand response could be achieved.”

The PUCT’s decision to stick to its energy-only market was hailed by pure energy-only market enthusiasts. The move by France to introduce a capacity scheme, likely to be adopted by Germany and possibly others, suggests that the debate about the virtues of energy-only vs. energy and capacity payment schemes is far from over.

F.P. Shioshansi

This post is extracted from EEnergy Informer, December 2012 issue.

Send this to a friend

4 Responses to “Texas Avoids Capacity Market While France Succumbs”

  1. Steven Stoft Says:

    Interesting story, but the theory part is nonsense — although it is believed by many.
    “Some experts believe energy-only markets are adequate provided the cap is sufficiently high.”
    “Other experts believe there is a flaw in energy-only markets referred to as ‘missing money.”

    Of course “missing money” is defined to mean “the cap is not sufficiently high.”
    The two sides agree on missing money. Here’s a rough summary of where they disagree.

    Energy-only advocates believe:
    1. The cap can be set correctly to VOLL so it is not a real regulatory intervention.
    2. The increase costs of risk from setting the cap to VoLL should be ignored.
    3. The increased market power problem should be handled by regulators

    Capacity-market advocates believe:
    1. Regulators have no idea what VoLL equals, so setting that is just as strong a market intervention as choosing a reserve margin. Setting either will determine the reserve margin.
    2. For political reasons the cap may be set too low, in which case it’s better to have a CM than to have blackouts and risk more political backlash — remember California.
    3. By adopting a CM with a reliability hedge –
    a) The cost of market risk can be reduced, saving more than the cost imagined for excess capacity.
    b) Energy market power can be tamed without regulatory intervention
    c) The public will tolerate the high energy prices that are needed to eventually bring about enough demand elasticity to eliminate the adequacy problem. (because the high prices are hedged)

    Energy-only advocates are wrong on #1, naive on #2, and confused on #3.
    However, because there is so much confusion over basic market theory, the chance of a good solution via either path is slim.

  2. Stephen Woodhouse Says:

    I like the comment of the previous commentator on the importance of political influence. A precondition for an energy-only market to deliver timely investment is that political influence must be exercised in a predictable way. In our energy-dependent societies and with the climate change agenda upon us, it is probably too much to expect that the degree of political influence will diminish. Perhaps the challenge of market design is to provide frameworks that channel the future exercise of political influence.

    Regarding point 2 from capacity market advocates, there are two conflicting political issues; firstly pressure to cap prices below the economic level, and secondly pressure to maintain reliability margins above the economic level. The reality – that average prices are more significant to consumers than rarely-occurring spot prices – does not seem to diminish political concern over price spikes.

    Broad-based capacity mechanisms (in which all providers of capacity receive common payments) can be designed in isolated areas to deliver effective dispatch and investment signals: essentially, the capacity prices must be formed in a way which mimics the price formation in an energy-only market. Such a scheme would lead to incentives for availability at the most critical times and would also provide sharp prices to consumers (or at least retail companies) at the same critical times.

    Matters get more confused at the boundaries between interconnected systems. Reliability crosses boundaries, whereas political concerns tend to be more local (typically nation state level in Europe; countries are entitled to choose their generation mix and levels of security supply under the Lisbon Treaty). The real challenge for capacity payments is at boundaries between systems; the market coupling between France (which in future will have a capacity payments) and Germany (which is less likely to do so) will deliver some perverse outcomes, both in terms of dispatch and investment.

    Perhaps the idealised scheme is one in which capacity is treated as an option to deliver energy. This is similar to the reliability options arrangement, except that these would be traded voluntarily by market participants as a hedge against peak prices (and exposure to imbalance). Generators would choose whether to sell their capacity as firm energy or options, and the sale of options would deliver a more stable revenue stream for peaking generators and demand response. The trading emphasis would be on energy prices (with high peak prices), and there would be no capacity obligations. This system would still require political and regulatory stability.

    The longer term solution must be more to encourage proactive demand response and demand management, so that both supply and demand are responsive to price. This would reduce the cost of reliability, reduce the extent to which peaking generators are required (meaning that the residual fleet has better load factors and more stable revenue). It would have the added benefit that there would be increased competition at the peak hours, mitigating the market power problem.

    With the spend of billions of € on smart metering across Europe by 2020, perhaps this new world is already on its way. However, there is a big step from a smart meter to a smart customer and I won’t hold my breath for this to happen.

  3. David Devon Says:

    Just to comment on the lack of flexibility capacity in France, I think that with all these new green energy issues and cost cutting problems building new power plants can be a lengthy process. There are many EU countries that have this same problem. Some times when you try to please everyone nothing gets done or not enough. There needs to be strong decisions made

  4. Lorraine Gregoire Says:

    I wanted to give additional explaintanions to this article on some questions I deemed interesting:

    First: Why France, as well as many other countries, needs to change its market design?

    France’s TSO, RTE, forecasted that by 2016 resource adequacy would not be ensured anymore. Several factors can explain this need of higher flexibility:
    •The promotion of intermittent renewables such as wind and solar
    •78% of French electricity production comes from nuclear power plants which are not renowned for their high flexibility.
    Even if they have improved a lot on this topic, they are far from being able to follow solar load’s brisk changes (or even wind).

    In addition to the need for more flexibility comes a need for higher reserve margin:
    •High peak load due to a high electrification rate of the domestic heating sector
    •Moreover peak load has been growing faster than average demand over the last 10 years. The ratio was of 1,4 in 2000 and is nearly 1,8 in 2012 (observatoire de l’électricité website)
    •According to RTE there is an increasing sensitivity to temperature gradient
    •Several flexible plants such as gas or coal turbine are being mothballed

    Second: Why France did not choose to simply raise its price cap on its power exchange?

    •Texas is known as one of the most competitive market which might not be the case of France
    •France, unlike Texas has a few more year ahead of her to ensure its resource adequacy which gives her time to design its capacity mechanism (with the loi NOME)
    •As said in a previous comment : setting VOLL is just as strong a market intervention as choosing a reserve margin since no one know exactly what VOLL equals
    •Choosing a capacity mechanism (like the one in PJM, which is one of the inspiration sources for France’s design) allows participation of Demand Response but also Energy Efficiency Resources and potentially Storage Resources

    Third: Some major issues with the centralized capacity market design

    •Adds high complexity to the current market design through the calculation of capacity revenue (peak capacity, availability, wind and solar contribution…)
    •Might be very intrusive for generation owners compelled to give a huge number of sensible data for the calculation of their capacity rights
    •Complex interaction with bordering regions that have separate legislation. The level of external capacity needs to be limited
    •Complex market design carries risks of initial design flaws and inefficiency

Leave a Reply