October 10, 2007

Mr. Ken Kozlik

Independent Electricity System Operator

655 Bay St., Suite 410

Toronto, Ontario  M5G 2K4

Dear Ken,

I am writing to provide comments on the IESO’s proposals regarding development of day-ahead mechanisms.

The IESO has identified a number of current issues which it hopes a day-ahead market will address:

• Gas fired generation—asymmetry between gas and power markets create risks for gas-fired generators which increase contract costs and electricity prices;
• Day-ahead Commitment Process—although it has been effective in addressing reliability, the DACP could be more efficient; pre-dispatch prices are poor indicators of real time prices; the Market Surveillance Panel has recommended the IESO review the DACP to reduce the costs and improve the effectiveness of the Generator Cost Guarantee; unit commitment efficiency could be improved through three-part bidding and 24 hour optimization.
• Prices—day-ahead prices could provide better signals for conservation, demand response and embedded generation

 The IESO has considered 5 options. Implementation of option 5, based on the model currently used in the New York Power Pool, would require fundamental policy changes in Ontario, including implementation of location-based pricing. Option 4 has already been considered in detail and at length. The conclusion then was that it was unworkable and undesirable. The IESO has decided already that option 2 should be abandoned. That leaves options 1 and 3.

The IESO has suggested that option 1 “improved price forecasting” will not achieve the benefits of a day-ahead market because it would not produce an “accurate HOEP forecast” so would not provide certainty to consumers and suppliers, would not improve the situation for gas fired generators and would raise concerns about the IESO’s role in forecasting price. With respect, the IESO already produces price forecasts, both in the DACP process and more generally via published pre-dispatch prices. IESO staff noted during the consultations that there are clear reasons why pre-dispatch prices are inaccurate: they exclude exports, allow imports to set price and are based on forecast peak demand in the hour. This suggests a relatively simple remedy to the issue of “improved price forecasting” which could provide immediate benefits to the market.

We recommend that the IESO include improvements to pre-dispatch price forecasting among the ‘base case’ items it proposes to address regardless of which option if any is selected for further study.

Since option 1 should be pursued in any case, and options 2, 4 and 5 are non-starters, the IESO’s proposal is to pursue option 3 exclusively. The IESO is asking stakeholders generally if Option 3 is likely to generate efficiency improvements that would exceed costs, maintain or improve future reliable operations or introduce any unintended market outcomes (e.g. creation of gaming opportunities or anti-competitive drivers).

Our review of option 3 suggests that it has not been defined to a level of detail sufficient to allow us to answer the questions the IESO has posed or to evaluate whether we should support it as is.

1.       Problem definition

A primary motivation to pursue a day-ahead market is to reduce risks created by asymmetrical markets for natural gas and electricity. The IESO has retained a consultant to assess the value of these risks and the potential benefits of a day-ahead market in mitigating the risks, but this report is not concluded and its results are not available to us at present. Even if this value is material, it is not clear to us how existing contracts might be changed or new ones amended to ensure that the benefit is shared with consumers.

Evidence from other markets suggests that a day-ahead market can be beneficial to consumers who have access to demand response programs by providing greater certainty, an ability to contract at a price a day-ahead, and as an aid in operational planning. Since the IESO is not proposing any kind of demand response program be developed in concert with a day-ahead market, referring that issue to the OPA, one of the principal advantages of a day-ahead market to consumers is not available in the current process. This absence reduces the appeal of a day-ahead market to consumers.

Anecdotal evidence from other markets suggests that a day-ahead market can be effective, but we are reluctant to draw parallels given the numerous and significant differences between Ontario’s electricity sector and the situation in other jurisdictions. Also, it is important to acknowledge that Ontario’s economy and geography is dissimilar from those of jurisdictions with day-ahead markets, suggesting that what works elsewhere cannot simply be transplanted here.

2.       High-level design

The current definition of Option 3 makes it very difficult to predict outcomes or to evaluate risks and opportunities for consumers or suppliers. The current design suggests that participation will be voluntary. This issue alone allows for a large number of permutations of energy price and quantity offers, between the day-ahead and real-time markets, creating almost complete uncertainty about outcomes in both markets. We cannot understand how option 3 will deliver claimed benefits (improved optimization, more efficient unit commitment, and more accurate prices) in this scenario. We can envision multiple opportunities for gaming and arbitrage between the day-ahead and real-time, all of which we fear would be prejudicial to consumers.

3.       Risks to consumers

(a)     Operating reserve

The IESO suggests that operating reserve would be settled in the day-ahead market. So long as the IESO is the sole buyer of OR in real time, this would create a situation where a consumer selling OR a day-ahead would be at a significant and unmanageable risk if operational or other considerations prevented him from supplying the contracted amount of OR in real-time. Our assessment is that this would cause customers to leave the OR market.

(b)    Scenarios for gas-fired generation

As we understand it, contract arrangements with new gas-fired generators make payments based on a deemed dispatch (heat rate) and the spot gas price at Dawn. We assume, were a day-ahead market to be implemented, that future contracts would be modified to refer to a day-ahead gas price. It is not clear how current contracts might be amended or future contracts modified to ensure that the benefits of these lower costs and risks are shared with consumers.

We can understand how a mid-merit or intermediate gas generator could benefit from nominating gas in the day ahead. It is not so clear that a day-ahead market would offer much benefit to gas generators called in real time to provide peaking energy or load-following capacity. These generators will still be challenged to buy and balance gas consumption in real time.

(c)     Implications for real-time market

Low liquidity has been identified as an issue in the current real-time market. Since the day-ahead option put forward by the IESO does not address fundamental causes of low liquidity, and is put forward without the context of a road-map for electricity sector development,  it raises questions about the impact on liquidity in real-time. Given that the current market is significantly inhibited by the high concentration of ownership, barriers to entry, and price intervention through contracts and regulation, it is not clear that the introduction of a day-ahead market will deliver any incremental benefits to customers through increased forward price transparency.

(d)    Implementation costs

The IESO has produced some rough estimates of its costs in developing and implementing a day-ahead market. No estimates have been provided of the potential administrative, systems and training costs for market participants.

4.       Benefit-cost methodologies

We are prepared to participate as much as possible in the IESO technical working group looking at how to analyze the benefits and costs of the IESO’s day-ahead proposals. We do not, however, support the IESO’s reliance on a narrow Kaldor-Hicks approach to estimating total efficiencies. The Kaldor-Hicks criterion implicitly assumes that marginal social utility of income is equal across all market participants. Such an assumption would recommend moving forward where total benefits exceed total costs,  even if that required significant transfers from consumers to suppliers. Efficiencies generally arise when costs and prices go down. We are not prepared to support initiatives that claim efficiencies but cause prices or consumer costs to rise.

Given these concerns, we have serious reservations about supporting further investigation of option 3 as it is currently defined. A way forward might be to investigate and develop additional detail around variations of the option, including, for example, a mandatory version and a voluntary version. An advantage of this approach is that it would allow us to evaluate and select the best of a range of feasible options rather than simply identifying a single option that compares favourably to the status quo.

Sincerely yours,

Adam White

President