On Tuesday 17 September, the California Public Utilities Commission released a study indicating that the five largest private generators in the California wholesale market did not produce up to their capacity during the winter 2000-2001. Looking at 38 blackout days between November 2000 and May 2001, the PUC study concludes that because of this lack of production, blackouts occurred, and that if Duke Energy, Dynegy, Mirant, Reliant, and AES/Williams had sold all of their available capacity, that all of the Southern California and most of the Northern California blackouts could have been avoided.
This study raises several practical and theoretical economic questions. First, the practical. The study did not take into account the fact that many facilities in California were operating as “reliability must run” (RMR) facilities for the California Independent System Operator (ISO), and that therefore much of who put power into the grid in what hour and in what amounts was under the automated control of the ISO. In fact, the authors claim that any and all hours in which plants did not comply with ISO orders to offer power were the deliberate actions of the generators, when many other studies and news reports have indicated the extent to which the ISO engineers were completely overwhelmed and could not keep up with the volume of system balancing work that was required of them. Because of the interplay of the utility bid underscheduling and the generator movement of transactions into the ISO real-time market, the ISO ended up processing many more real-time transactions than had ever been intended, or for which it had been designed. Thus the ISO was scrambling to keep the grid in balance, and probably did not perform optimal dispatch, but the PUC study does not acknowledge that reality.
The report does point out the important fact that the ISO could not compel generators to comply with ISO requests in Stage 2 or Stage 3 alert periods. The ISO did not receive this authority until June 2001.
Amazingly, the report trots out the usual arguments against the generators for their bid withholding from the day-ahead market, but fails to discuss the role of utility bid underscheduling in working in conjunction with that withholding to shift so many transactions to the ISO real-time market. This combination of bid underscheduling and supplier bid withholding tends to happen in bifurcated wholesale market designs, such as the PX/ISO system that California mandated in its restructuring legislation.
Now, the theoretical. On page 14 the report’s authors assert that “the generators should have offered all of their available power supplies to the ISO at all times. Indeed, after FERC imposed comprehensive market controls in June 2001, including a price cap, trading barriers to prevent some types of market manipulation and a ‘must-offer’ obligation, blackouts and service interruptions nearly ceased even though California’s power demand was at its highest in the summer.” Suggesting that generators should have offered all of their available capacity at all times is economically absurd.
Contrary to what some politicians would like to believe, generating electricity costs money, those costs fluctuate even in functioning markets, and those costs were unusually high in the period in question. Two of the cost components that increased the most were natural gas and emissions permits. The report mentions high natural gas prices (alluding also to the ongoing FERC investigation of natural gas price manipulation at the California border), but does not incorporate the important, and scarce, emissions permits. Other studies have suggested that some plants could not operate in some hours because the price they could get for their electricity did not make buying the permits worthwhile.
One revealing piece of economic ignorance in the study is on page 18, where the authors state that “in these calculations, bids are not considered valid if the ISO ordered a generator to produce power pursuant to a bid by that generator, and the generator failed to respond to or rejected the dispatch for ‘economics.'” In other words, the PUC would have preferred to force the generators to produce power even if they would lose money by doing so. I would be interested in knowing how many such bids were excluded from their calculations, and what proportion those bids were of the total capacity that the study attributed to the generators. Unfortunately, though, the appendices that apparently report the data are not available on the CPUC website, so I am unable to replicate their results.
Furthermore, statements like “Generators Failed to Bid in All Supplies During Blackout and Service Interruption Hours Even Though the Power Was Needed” on page 56 seem to indicate that the PUC expects the electricity industry to operate on the basis of charity, not on the economic exchange of value for value between buyers and sellers.
Another economic feature that the study fails to understand or interpret correctly is payment and credit risk. In almost any transaction there will be some credit risk. In the dysfunctional universe that was the California wholesale market in 2000-2001, credit risk was huge – one reason why prices went as high as they did was that generators factored in the probability that they were not going to get paid. Given that PG&E subsequently declared bankruptcy to avoid its creditors, that risk premium in the wholesale electricity price looks like it was pretty important. On page 53 the authors of this study show the extent to which they fail to understand credit risk and the risk premium portion of the wholesale price, when they relate an ISO request to a generator to power up in a Stage 2 hour, but the generator refused because they did not think they would get paid. Apparently the ISO operator took umbrage at being told by the generator that, given how close demand was to supply, the ISO should consider reducing demand. In the archaic, one-sided world of supply dispatch, the only way to reduce demand is involuntary interruption (California did have some voluntary interruption contracts, but only rudimentary ones). In fact, the report characterizes the generators as having disputed prices and terms “improperly” with the ISO. The attitude put forth in this report is not one that understands and appreciates mutually beneficial exchange as a foundation of civil society.
There is not one single mention in the report of the role of exports, or the fact that California gets much of its power from plants that are located out of state and were not required at the time to sell into California, although it counts that capacity as part of what the five generators should have sold into the ISO upon demand. The report obliquely refers to the exports and Western market issue in its final chapter, with the shrill demand that FERC extend the “must serve” conditions that are currently in place for the whole Western interconnection.
The report also invokes the studies that showed that generators had incentives to withhold power to raise prices, and that they did in some hours. Again, though, they add no economic insight or nuance to what has been shown in those other studies. Most importantly, they fail to acknowledge the role that poor, politicized policymaking processes and bad rules played in creating an environment in which generators did have market power. As many have said before me, and many will after, the dysfunctional California restructuring labyrinth gave the generators market power on a silver platter.
The extent to which the authors demand in the final chapter that FERC “do something” indicates precisely the extent to which the PUC intends this report to further its claim on retaining control over the fates of electricity producers and consumers in California. I agree with Jan Smutny-Jones, who was quoted in this LA Times article saying “Clearly, the target here is to affect standard market design in Washington, which is an actual attempt to fix the problem.” FERC has issued a Notice of Proposed Rulemaking in their standard market design process, and the California PUC continues to perceive FERC’s moves as a threat to its power over and control of the electricity industry in California, as a recent San Diego Union Tribune article suggests.
Other news stories on this report are:
Lynne Kiesling is director of economic policy at Reason Foundation and senior lecturer in economics at Northwestern University.