Getting Reliability

The most obvious lesson learned from the blackout report is that the electric power industry and its regulatory organizations are better at diagnosing system failure ex post than at divining ways to foster growth of a self-correcting, self-reinforcing, and dynamically reliable system.

The blackout report does an excellent job of diagnosing the recent failure, and provides a helpful tutorial on power system operations for the non-specialist at the same time. But the 46 recommendations it offers won’t get us to success we want.

In some sense the report is trapped by the “one big interconnected machine” picture of the grid, and an accompanying view that reliability can only be attainted by a mixture of planning, regulation, spending and hope. Reliability on the transmission system is not the mystery that it once was.

The relevant factors that add to or subtract from system reliability are well understood. Most of these factors are attributed to, or could be measured and attributed to, the responsible party. The responsible party could then be either charged or paid an appropriate amount. The key is to bring reliability into the commercial realm, where choices can be made in the presence of relevant tradeoffs.

As the blackout report says “reliability is not free” and the way to ensure that we get what we pay for, and not too much more or less than we want, is to move reliability into the market. The report agrees that “market mechanisms should be used where possible,” but worries that there may be conflicts between reliability and commercial objectives that cannot be reconciled. (Quotes from p. 139) In such cases, says the report, high reliability must trump commercial concerns.

This reliability as trump card could be a bad rule. Let’s consider provision of reactive power as a concrete case. The blackout report does a fair job of introducing reactive power and its role on the transmission grid. It suffices for our purposes to note that (1) generators can get market rates for real power, (2) generators can be paid cost-based rates for providing reactive power, (3) the transmission grid needs reactive power to operate efficiently and to remain stable, and (4) generators face a trade-off between the production of real and reactive power, at least at higher levels of producing real power.

Lack of reactive power support was cited as a contributing factor to the blackout. What happened? In part, as real power prices increased, generators became more reluctant to produce reactive power at cost-based rates. Reportedly, system operators asked an independent generator to produce more reactive power, and the generator declined because it would make more profit selling real power.

This example looks like a direct conflict between reliability and commercial objectives, and it is, but only because current rules treat reactive power as a public-service type duty to the network, for which a regulated rate is just compensation. If, on the other hand, the generator was paid a price for reactive power which reflected the opportunity costs incurred in terms of lost profits for real power, the conflict goes away.

The ultimate objective is healthy, thriving wholesale power markets, which rely on transmission networks. Reliability is a crucial element in enabling those power markets to continue developing. But that doesn’t mean that reliability is a “one size fits all” characteristic of the network. Treating reliability as a public good leads to conflicts, treating it as a private good could avoid those conflicts. The metering, monitoring and switching technology exists to treat reliability as a private good. Now we just need the institutional and legal structure.

Healthy, thriving markets are a pipe dream without a demand side. Active retail demand transmits end-customer preferences into the wholesale market, smoothing out peaks and optimizing load factors (and curbing the exercise of supplier market power along the way). Furthermore, allowing demand reduction to bid into capacity markets can reduce the construction of new generation and transmission capacity, and is therefore a good long-run strategy for conservation of resources and for making investment more efficient. Demand response is the Swiss Army Knife of the electricity policy world – it is one compact tool that does a lot of things in a very parsimonious way.

In the end, we think that the most important changes to make in the industry are really just a continuation of industry restructuring. Let’s commercialize reliability – reform the reliability rules to properly line up incentives and information flows. Reliability is valuable to consumers. What has been lacking is a way for consumers to express that value, and for suppliers to be appropriately paid for providing it.

The regulator’s report provides a regulatory solution. That fact in itself is not too surprising. But the regulatory solution won’t give consumers what they could really use, which is a more efficient, more resilient and more dynamic power grid.


2 thoughts on “Getting Reliability

  1. When “managed” markets fail, the failure is always attributed to the markets, rather than to the “managers” (regulators). The “solution” is always more “management”, rather than the operation of more market forces, especially when the solution is prescribed by the “managers”.

    The most dangerous aspect of the restructuring of energy markets is the transition period, during which the industry must rely on temporary structures and rules. The longer the transition lasts, the more likely that it will result in failures caused by these temporary structures and rules.

    Regulators have characteristically accused utilities of “gold plating” infrastructure to maximize their returns. Regulators have also characteristically favored delaying investments in new technology because of their impacts on rates. (Baling wire and bubble gum are fine until either fails.)

    The estimates of $50-100 billion to “fix” the grid after the Northeast blackout would suggest that regulatory and utility decisions resulted in underinvestment of this magnitude in the grid during the period when a multitude of independently designed, constructed and operated transmission systems were being “baling wired” together into an interconnected and interdependent network.

    The questions which apparently were not objectively and adequately addressed in the assessment of the causes of the blackout include: “What were the incentives which resulted in the decisions on interconnection design and operation which most directly contributed to the massive failure of the interconnected system?”; and, “What are the incentives which will most likely result in revised interconnection structures and protocols which will assure optimal reliability at minimal cost?” This is not surprising, since the assessment was performed by the “managers”. If the “managed” grid is to persist, the managers must eventually realize that it is not enough for them to mandate reliability; they must also permit it. If a transition to a competitive market is to occur, the transition should be thoroughly studied and then rapidly implemented. (The basic rule followed by carpenters applies here: measure twice, cut once.)

  2. The distinction between market and regulators (managers) leaves out the roles of transmission owner/provider and grid-operator (MISO). The problem as I understand it, from working with a number of grid systems and ISOs, is that (1) there was inadequate metering/SCADA information about the loadings at particular points on the Ohio (First Energy) part of the grid and (2) there were faulty circuit breakers that would have tripped (and then ideally signaled) the overloading of a single set of circuits. Either proper circuit metering/SCADA — assuming this information was communicated to MISO and detected — or circuit breaker upgrades would have forclosed this problem. This is the joint responsibility of the TO (transmission owner) and the ISO (MISO) to ensure that adequate systems are in place. Regulation cannot be detailed enough to ensure these systems are in place, though policies and monitary penalties by regulators or MISO may work to reduce the incidence of such problems. Thus, I suggest that the specific technical problems be defined in more detail and that remedies be defined from this vantage point.

Comments are closed.