The notion of “reliability to the nines” — that electric utility service should be so reliable as to suffer an outage event on only one day in ten years — stands as a touchstone of  today’s electric utility industry. Yet it’s no more than myth, a sort of urban legend that likely sprouted wings back in the 1960s, after the first big New York City blackout, when the industry told Congress it could regulate itself by setting up voluntary reliability standards.  In truth, there is not now nor never has been a formal, nationwide electric reliability standard to require utilities to maintain a loss of load expectation (LOLE) no greater than 0.1 per year. It’s a canard, a chimera.  

That fact was made abundantly clear at the close of last year when, at long last, the North American Electric Reliability Corp. (NERC) finally got around to filing a formal application with the Federal Energy Regulatory Commission (FERC) for approval of the proposed Regional Reliability Standard BAL-502-RFC-02 — Planning Resource Adequacy Analysis, Assessment and Documentation, to give the ”one day in ten years” standard the force of law – but only in the particular region known as Reliability First, which covers a wide swath of the mid-Atlantic and near Midwest, blanketing PJM and covering a substantial chunk of the Midwest ISO.

“The petition is the first request by NERC for FERC approval of this proposed Regional Reliability Standard,” NERC explained, in its application. “Continent-wide reliability standards do not presently address the issues covered in the proposed … standard.” (See FERC Docket RM10-10, filed Dec. 14, 2009.)

So what should we make of this belated omission, and NERC’s proposed fix? Well, two things, at least.

First it’s not clear that the one-day-in-ten requirement even qualifies as a true reliability standard. Is it crucial to the workings of the mechanical machine we know of as the bulk power system? Or does it have more to do with resource adequacy, and how much in dollar terms that we are willing for consumers to have to pay to keep the lights on? Second, even if it does qualify, and even if it gains formal adoption (which is far from a slam dunk), it’s not clear whether FERC enjoys the necessary authority to enforce it. That’s because the Federal Power Act, the 2005 Energy Policy Act, and all other U.S. utility laws deny to the Feds the right to regulate directly either electric generation or resource adequacy.

(Of course, careful readers of Public Utilities Fortnightly magazine no doubt will want to point to last year’s federal appeals court ruling in Connecticut Dept. of Pub. Util. Control v. FERC, 569 F.3d 477, which was reported by Fortnightly editor-in-chief Michael T. Burr in his piece “Policy Shift: 2009 Groundbreaking Law and Lawyers Report,” Public Utilities Fortnightly, Nov. 2009, p. 24. The article explains how the D.C. Circuit upheld FERC’s authority to approve the installed capacity requirement in ISO New England’s Forward Capacity Market. Naysayers claimed that FERC had intruded on traditional state regulation of generation and resource adequacy. But the court affirmed FERC’s rulings on the ground that installed capacity requirements in RTO markets serve more as rate-setting algorithms, since the ICAP target, coupled with capacity purchase bids from utilities, ultimately sets the RTO capacity market price. All of which makes my point – that reserve margins and the one-day-in-ten standard are more about setting a price that consumers can live with, rather than safeguarding the mechanical functioning of the grid as a complex machine.) 

How is one-day-in-ten bound up in rates and energy pricing? Under the typical wholesale energy pricing regime in place in the nation’s RTO’s and ISO’s, the forward locational capacity markets (sometimes referred to as “ICAP”) take their clues from one-day-in-ten. Both the calculation of CONE (the cost of new entry for proposed new electric power plants) and the design shape of the administrative, sloping demand curves that define RTO capacity markets depend on a fundamental equation: the Loss of Load Probability = the gross cost of building a new plant divided by the Value of Lost Load. LOLP = CONE / VOLL.

If it costs $72 per kilowatt-year to build a new, aero-derivative simple-cycle gas-fired combustion turbine, then our one-day-in-ten equation dictates that that the opportunity cost of an outage, the Value of Lost Load, will equal $72,000/megawatt-year divided by 2.4 hours, or about $30,000 per megawatt-hour (MWh), which is a pretty good approximation, say the experts, of what users would be willing to pay to avoid an outage, if they had such a choice to make.

 Whether we want this equation to define the wholesale capacity price that is bound up in wholesale power prices set in RTO wholesale energy markets is purely a political decision for regulators. It’s a guess about the ratepayer’s ability to pay – a compromise struck to balance the twin evils of price caps and price spikes.

In this light, the one-day-in-ten idea hardly seems to qualify as a “reliability standard,” as FERC defined such in Order 672, where the commission ruled that a proposed reliability standard “must address a reliability concern that falls within the requirements of section 215 of the Federal Power Act — that is, it must provide for the reliable operation of Bulk Power System facilities. It may not extend beyond reliable operation of such facilities.”

And, in an virtual admission against interest, even NERC appears wonder whether one-day-in-ten pertains more to resource adequacy,  which exceeds FERC authority. In its application filed last December in Docket RM10-10, NERC explained that its proposed one-day-in-ten reliability standard actually imposes two separate duties: (1) to document the analysis of  loads and resources, and (2) to document and publicly post the projected load and resource capability obligations that would demonstrate the sufficiency of planning reserves for meeting the one-day-in-ten standard.

But as NERC writes in its application: “The enforcement mechanism for planning reserve margin obligations … is not expressly permitted by the Energy Policy Act of 2005 and it is also outside the scope of the proposed standard.”

And why would NERC even to run the risk that FERC might deny the application, and actually reject one-day-in-ten as a mandatory reliability standard? After all, the noted economist and utility expert James F. Wilson argues in the current Fortnightly issue that the one-day-in-ten standard may well be overkill — that it is probably a full order of magnitude more strict than needed to assure economically efficient and reliable service. See his piece, Reconsidering Resource Adequacy, Public Utilities Fortnightly, April 2010, p. 33.

So far, pure myth has sufficed to convince regulators alike that one-day-in-ten is just right. It’s the utility industry’s version of “don’t ask, don’t tell. But in converting the legend to a full-blown formal reliability standard, NERC runs the risk that the myth may die — that once open to scrutiny and the legend may not survive the full light of day.