Power market seams and the role of arbitragers in market design

Michael Giberson

For class tomorrow I’m reading up on things Enron and California power market melt-down related.

I’m a fan, for example, of Jonathan Falk’s 2002 article in the Electricity Journal on the infamous “Smoking Gun” memo which detailed Enron’s colorfully-named trading strategies like “get shorty” and “death star.” Among other things, Falk points out the several of the strategies provided arbitrage services between the many power markets in and around California, and at least some of the strategies likely helped the California market work more efficiently.

Richard J. Pierce, Jr. has what might be seen as a follow-up article in 2003, also in the Electricity Journal.  Pierce agrees with Falk that many of Enron’s strategies could be fairly described as  arbitraging the California market, but he also asserts that many of the strategies also could be fairly described a manipulative.  As a kind of aside, Pierce said, “If the California debacle has taught us nothing else, it should persuade us that arbitragers should never be given a role in structuring a market. They have a powerful incentive to maximize the flaws in the market design in order to maximize potential arbitrage profits.” (p. 40, emphasis added)

Pierce exaggerates a bit; it is unlikely that maximizing flaws will maximize potential profits.  Rather, some optimum amount of relatively minor flaws probably promises the most overall profit for arbitragers. But Pierce reminded me of the role that arbitragers have played on seams issues between the New York ISO and ISO New England markets.

For years, market monitors for the regions and some market participants have encouraged the New York and New England markets to exchange real-time market information and coordinate power flows as necessary to eliminate price distortions along the border between the markets. For years, other market participants (primarily traders participating in both markets) have continued to support alternative market changes that have the effect of continuing the special role played by traders in determining power flows between the markets. The Federal Energy Regulatory Commission has directed the markets to fix the seams issue in cooperation with market participants, but for years the ISOs have decided that other market changes were higher priority. For years, FERC has accepted that answer from the ISOs.

An estimate by Potomac Economics, external market monitor for the New York ISO, suggested that the net cost of power to New York consumers would have been $177 million lower in 2007 had the two markets better coordinated the power flows between the regions.* Still, the issue is not a priority at the ISOs or the FERC. Maybe when someone notices that efficient market-to-market coordination of power flows between regions would make better use of renewable power and demand-side resources participating in New York and New England markets, then we will see FERC make resolving this seams issue a higher priority. Until then, FERC and the ISOs continue implicitly to support the arbitragers’ favored approach to managing the seams.

(*See Table 1 on page 28 of Potomac’s “2007 State of the Market Report: New York ISO.”

DISCLOSURE: I contributed to the drafting of the 2007 market report as an employee of Potomac Economics last year before taking my current position at Texas Tech University. Of course, nothing I post here should be taken as expressing the views of either my former or current employer.)

Wind power is dispatchable, down

Michael Giberson

Last week the New York ISO filed proposed tariff changes with the Federal Energy Regulatory Commission to revise how it treats wind power generation in its markets.  Under the NYISO’s current rules, wind power generators are not treated as flexible resources.  When the transmission system is overloaded, other generators can be asked to back down but wind power would not be reduced under most circumstances.  (Under current rules wind power can be backed down in extraordinary cases, via a cumbersome process, and sometimes generators back down voluntarily due to low prices.)

Under the proposed rules, wind power generators would be treated as flexible resources by the real time market system for output levels from zero up to the forecasted wind power output level.  The practical effect would be to allow the system to back down wind and non-wind generation in comparable fashion as needed to resolve congestion on the transmission grid.  Wind power generators would submit energy bids into the market like most generators and the ISO market would use the bids to work out the least cost method for resolving congestion.

While the change could result in wind power generators being directed to reduce output more frequently than under current rules, the NYISO said the change would likely increase the overall amount of wind power taken by the system. Under the current, manual procedures for directing reductions in wind power output, it is hard to get just the right amount of power off the system. The NYISO indicated that sometimes more power has been taken off the system for a longer period than was strictly necessary. In addition, even under voluntary curtailment by wind power generators due to low (and sometimes negative) prices, sometimes more power has been taken off the system longer than was needed to resolve the problem.

Both symptoms of the cumbersome approach now used have resulted in more work for the system operator and a less efficient supply of power.  The proposed changes should serve to more finely target any needed reductions in output, allowing a more efficient use of wind power when it is available.

Even with the proposed tariff changes, the NYISO expects that most of the time it will take all of the wind power available to the system. Of course, whether this remains true will depend on the pace of further wind power additions relative to the transmission improvements, if any, needed to support those additions.

The changes represent another step in the direction of the normalization of wind power resources in integrated regional power markets. The rules for wind power will never exactly be the same as the rules for, say, a combined-cycle combustion gas turbine, but then the treatment of that very flexible natural gas unit isn’t exactly the same as the treatment of a less flexible coal-fired steam turbine.

The market design goal should be to maximize the gains from trade produced through the regional power market. The process of adapting rules to the diversity inherent in the generation and demand-side resources available to the system should be undertaken with that goal in mind.

(ASIDE: Elsewhere I have been critical of the way that production tax credits available to wind power generators can distort the efficient operation of power markets. There was some talk about this issue in the recent FERC technical conference on integration of renewable resources, and I suspect I’ll have more to say on the issue in a few days when the transcript is online.

Some critics of current wind power subsidies may object to any normalization of the treatment of wind power so long as the subsidies continue, but I don’t think it desirable to try to fix market rules to compensate for the harmful effects of federal tax policy. Rather, market design should aim to maximize gains from trade given the larger legal and policy framework which the markets must operate in, and opposition to wasteful subsidies in the tax code should be directed to the legislative bodies responsible.)

AWOL today with a headache

Lynne Kiesling

I’m afraid I’m hitting a delay in posting the final part of my smart grid commentary because I’m battling a nasty headache. Thanks for your interest, and I’ll get it up here ASAP!