An Inconvenient Fact


Why the standard market design refuses to die.

Fortnightly Magazine - December 2006

Hold on to your hats. The vaunted and vilified “standard market design” (SMD), once thought dead and buried, has been resuscitated, with all attendant chaos and rhetoric, but this time in the guise of a new proposal that debuted in early August under the code name “open dispatch.”

This new construct, as remarkable in its way as Einstein’s theory of indeterminate space and time, declares that electric transmission, long seen as one of a triumvirate of unique and essential utility industry sectors (along with generation and distribution), is little more than a mirage.

Rather, it is the sequencing and dispatching of generating units, accomplished in real time along principles of marginal cost, and with full respect for the physics and security constraints of the grid, that constitutes the real “service”—the single substantive activity and one that warrants full federal regulation. In this view, the three true industry sectors are: (1) generation; (2) dispatch; and (3) “load assets” (or distribution). Transmission, along with his sidekick “open access,” becomes odd man out—not so much a function as a playing field, in which transactions take place.

Open dispatch frowns on the Newtonian concept of a precise contract path, or that transactions and service should depend on static measurements of grid capacity. The only useful guidepost becomes the dynamic and security-constrained market equilibrium, as dictated by the unit dispatch—the only facet of grid operation that is truly amenable to ironclad mathematical modeling, and efficient execution by computer software.

A simple notion, perhaps. Yet it mocks the Federal Power Act, plus a full century of legal and economic precedent. Critics see open dispatch as a poorly veiled attempt to resurrect the SMD, leading inexorably to mandatory participation in regional transmission organizations (RTOs), or at least formation of independent grid coordinators, in the style of Duke or Mid-American Energy. They decry this “second bite at the apple,” noting that Congress had wanted to ban the SMD outright in the 2005 Energy Policy Act (EPACT), but removed such language only after FERC itself had agreed to drop SMD as a policy alternative. These opponents warn that advocacy of open dispatch could “re-open old wounds” in the relationship among Congress, state regulators, and the Federal Energy Regulatory Commission (FERC)—wounds, they say, that “only recently have begun to heal.”

The testing ground for this new theory is the notice of proposed rulemaking (NOPR) issued last spring, in which FERC took a surgical approach and recommended a series of micro refinements to the pro forma Open Access Transmission Tariff (OATT) it first approved in 1996 in Order 888. On its face, the NOPR proposes dozens of modest modifications to the tariff to make the point-to-point (PTP) transmission service that is offered to merchants more comparable to network service more often provided to native load:

Capacity benefit margin (CBM)—FERC wants uniform standards, or else offer CBM as a separate service;

Energy and generator imbalances—FERC touts ideas from Bonneville Power, which employs a partial netting to avoid penalties, plus a multi-tiered series of deadbands and different penalty categories;

Hourly firm PTP service—FERC says it might be needed;

Regional transmission planning—FERC lays out eight principles to govern coordination and meetings, but doesn’t say what a “region” is, or tell planners how to recognize “significant and recurring congestion,” which is to be the focus of such efforts;

Service denials—FERC asks grid providers to give narrative explanations when they deny transmission to merchant generators;

Joint ownership—FERC asks whether it should create an “open season” for large regional backbone grid projects to facilitate joint ownership (on the wish list for public power);

Secondary resales of grid capacity—FERC would relax price caps, as was done with gas pipeline capacity;

Rollover rights—a guaranteed renewal only for those transmission contracts of five years or more, says FERC; and

Public power participation—FERC avoids any blanket mandate for non-jurisdictional utilities to file an OATT, but will review the idea case- by-case.

Space limitations prevent much further analysis of these or the many other OATT proposals, though each has provoked controversy, as seen by the sheer length of the FERC’s NOPR (536 pages), plus the thousands of pages of comments filed by some 150 industry stakeholders. (See, Docket Nos. RM05-25, RM05-17, May 19, 2006, 115 FERC ¶61,211. All comments cited in this column come from these dockets. See also, Commission Watch, “Tariff Tinkering,” Jan. 2006. That prior column reported on FERC’s initial inquiry on how it might revamp the OATT.)

By contrast, two other ideas from FERC warrant a closer look:

1. Measuring Grid Capacity. Standardizing the calculation of ATC (available transfer capability), to determine how much transmission capacity that transmission service providers have available, after satisfying all prior service commitments (including, in particular, network-transmission service promised in support of native load); and

2. Redispatch to Boost Capacity. Requiring transmission providers either: (a) to offer “conditional firm” service (instead of simply denying a service request) if firm PTP is available in all but a very few defined periods during the term; or (b) to consider redispatching resources to reconfigure grid flows, without first conducting a facilities study.

These twin ideas—ATC reform and redispatch—offer key clues in understanding the theory of open dispatch.

In Defiance of Law

This monster NOPR was pretty much the pet project of FERC boss Joseph Kelliher, and he has talked it up ever since he took the chairman’s seat. Yet he has cautioned also that the NOPR should only improve on Order 888 and the OATT framework—not create any “new market structures.” Industry observers took Kelliher at his word: that FERC would not rock the boat. Certainly, there was no inkling that FERC would revisit the SMD or RTO markets, be they “day 1” (real-time balancing) or “day 2” (day-ahead, bid-based energy price auctions).

But then came trouble. On Aug. 2, consultant John Chandley (LECG) and the more famous professor and consultant William Hogan (Harvard Univ.) filed comments advising FERC that in seeking to standardize ATC calculations, it was bent on a fruitless quest:

“The inconvenient fact is that even if [we] were somehow able to define … [a] detailed set of rules for calculating ATC, the ATC calculations would still be largely irrelevant.” (See initial comments of John Chandley & William Hogan, filed August 2, 2006, p. 16.)

Moreover, however (and herein lies the real message), Chandley/Hogan implied that FERC’s intended focus on ATC also was dangerous, as it would only worsen the problem.

That’s because FERC’s ATC construct dares do nothing more than to ask, in essence, “how much capacity is left over” for merchants and third parties, after incumbent utilities, native load, and others holding prior, grandfathered rights are served first, regardless of the commercial value of their transactions.

Chandley/Hogan instead would put everyone in the pool as equals—before assigning any priority or privilege to assets linked to native loads or incumbent utilities (save, perhaps, for resources with unique physical and geographic properties essential for grid security). Starting from that point, the transmission provider then must make everyone happy. In other words, the job of transmission provider is not to ration the grid, but to find a way to honor every request. Open dispatch would achieve this end, in effect, by ensuring that transactions are properly valued, and then are properly balanced against the “price” of congestion and other real-time physical constraints. The provider deploys assets and re-arranges unit dispatch to respond to these prices and physical dictates—all for the greatest good for the greatest volume of transactions, as measured in terms of commercial value.

Thus, according to Chandley and Hogan, FERC had misconstrued the very nature of transmission service. The authors warned that FERC was leading the industry “inexorably down false paths towards dead-end solutions that cannot solve the problem.” (Chandley/Hogan, initial comments, p. 8.)

It then took only five days for another chip holder to come in and up the ante. On Aug. 7, the PJM Interconnection (a certified RTO) went one better and proposed to FERC an entirely new doctrine, grounded on the Chandley/Hogan ideas. PJM endorsed many of FERC’s incremental improvements to make the OATT more “transparent,” but advised that the real need was for a different sort of transparent tariff. This new tariff would have little to do with the wires, per se. Instead, PJM proposed a new tariff that would open the generation “bid stack” to regulatory scrutiny.

Of course, vertically integrated load-serving utilities that own both transmission and generation generally do not run actual auctions or take bids. Their stacks are more likely reflective of internal costs, and shielded from view. Thus, to be more precise, PJM’s proposal would simply require such utilities to make visible the full set of cost inputs, operational characteristics, and reliability constraints that come into play when they operate their transmission networks. These cost inputs might include gen plant parameters such as fuel costs, start-up and minimum-run costs, unit ramp times (up and down), emissions restrictions and such, plus how they interact with model assumptions about the workings of the grid.

PJM coined the term “open dispatch,” to distinguish its proposal from what it called the “closed dispatch” practiced by vertically integrated utilities and other transmission providers operating in non-RTO areas. It also assured FERC that its proposal would not require RTOs, or mandatory formation of independent grid operators. It would simply make the bid stack visible, just as FERC’s OATT had forced utilities to reveal their open-access practices.

As PJM wrote in its comments, “control over the dispatch sequence is … in today’s information-based society, as much as ‘essential facility’ as the transmission grid itself.” (PJM, initial comments, Aug. 7, 2006, p. 25.) PJM’s inference was clear: An unregulated dispatch actually may violate antitrust law. PJM added that comments filed in prior SMD dockets by the Department of Justice and the Federal Trade Commission had lent implicit support to the idea.

This modern notion of simply leaving congestion in place and then managing it with prices to improve grid efficiency very much resembles the cap-and-trade plans that have become so well accepted for controlling emissions, such as sulfur dioxide and carbon dioxide. It stands opposite to comments from those like the city of Santa Clara and Silicon Valley Power, who say that congestion pricing only diverts funding from the more important job of expanding the grid to eliminate congestion (Santa Clara, initial comments, p. 9.), and also from Progress Energy, which argues that “the use of redispatch as a virtual substitute for additional transmission” runs counter to EPACT “because it does not encourage the expansion of the transmission system.” (Progress Energy, initial comments, p. 44.)

Nevertheless, by late September, PJM had attracted enough allies to form an ad hoc coalition known as the Transparent Dispatch Advocates (TDA). Coalition members included ELCON (the Electric Consumers Resource Council), EPSA (Electric Power Supply Association), Exelon, AWEA (American Wind Energy Association), and several environmental groups, including, most notably, the Natural Resources Defense Council.

Also, while not officially a coalition member, San Diego Gas & Electric appears quite sympathetic to the TDA platform. SDG&E attorney Don Garber, who has been quoted previously in this column on various occasions, has offered the most substantive and comprehensive explanation of the open dispatch concept, plus a full rundown of the TDA agenda. (SDG&E, reply comments, filed Nov. 3, 2006.)

As one might imagine, the reaction was swift and deadly. Many saw it as crazy to invoke the wrath of Congress after it had so clearly forced FERC to kill SMD and to stick to the straight and narrow. For example, the American Public Power Association said its members were not eager to repeat the SMD experiment, since Einstein had defined insanity as “doing the same thing over and over again and expecting different results.”

Many stakeholders cited Kelliher’s intent to narrow the focus and said PJM had picked the wrong docket. One was the Sacramento Municipal Utility District (SMUD): “These proposals so dramatically depart from the NOPR that they could not lawfully be adopted in this proceeding.”

Many opponents faulted the open dispatch theory for involving FERC with power plants, thus violating the long-established federal-state jurisdictional divide over the regulation of transmission (a federal matter) and generation (left to the states).

For example, as a counterweight to PJM’s TDA coalition, Entergy, Southern Co., Progress Energy, and the Salt River Project formed the so-called Community Power Alliance to defend the status quo. It saw open dispatch as “insensitive to the concerns of state regulatory commissions.”

Progress Energy, filing its own comment, then blasted open dispatch as violating native-load protections and due-process guarantees: “Open dispatch is a shorthand term for … confiscating the power supply resources of vertically integrated electric utilities in order to serve loads that they have no obligation to serve.”

Entergy advised that FERC and its staff had long recognized and understood “all of these concepts,” but had simply “knowingly chosen to adopt a different approach than that previously taken with the SMD.” Continuing in the same vein, the large Public Power Council warned that “Chandley/Hogan may feel free to quarrel with Congress, but the commission does not have that luxury.”

The Evidence

The analysis starts with short-term grid imbalances, as when generators run short or long of scheduled output, or load exceeds forecasts or fails to show up. That’s when mischief begins—but only if imbalances are priced incorrectly.

Entergy’s service area offers the prime example, where some 17,000 MW of new merchant generation has put down roots, though Entergy’s grid system claims only 4,000 MW of simultaneous transport capability. For years, Entergy complained how merchant plants leaned on its system over short-term intervals. They allegedly would ramp up and down ahead of schedules, shaving minutes so as not to leave money on the table (not to mention start-up and minimum-run costs), while Entergy’s native units (even coal plants) would scramble to maintain frequency control. That led Entergy to negotiate a special generator imbalance agreement (GIA) with merchants. (See Docket ER04-901, June 1, 2005, 111 ¶FERC 61,314.)

In fact, while it opposes open dispatch, Entergy defends its GIA as offering real-time features for managing imbalances that go far beyond the crude monthly accounting requirements of FERC’s OATT revised plan. Thus, Entergy urges FERC not to jeopardize its own “uniquely tailored” plan. (Entergy, initial comments, pp. 26-35.)

Another open dispatch opponent, the Edison Electric Institute (EEI), recognizes nevertheless how important it is for FERC to distinguish between imbalances in peak- and off-peak periods: “Otherwise,” says EEI, “energy imbalance customers will have an incentive to underschedule in high-load periods when costs and prices are high, and to repay … when costs and prices are low.” (EEI, initial comments, p. 71.)

Meanwhile the Western Electricity Coordinating Council (the western reliability region) has noticed that FERC-mandated imbalance penalties have proven problematic. It seems that some generators use “set-point controllers” to “override governor action” to avoid imbalance charges, but erode frequency control in the process. Meanwhile “those with properly operating governors” will help control frequency, but “may be punished,” says WECC, “for deviating from scheduled output to respond to system reliability needs.” (WECC, initial comments, pp. 19-20.)

Chandley/Hogan were quick to spot this anecdote from WECC as an indictment of FERC’s OATT revise: “In other words, reliability entities in the West … are telling the commission that Order 888’s somewhat arbitrary mechanisms for penalizing imbalances may be posing a threat to grid reliability.” (Chandley/Hogan, reply comments, Sept. 20, 2006, p. 13.)

A second example comes from PJM itself, where the RTO suggests that pricing inconsistencies between its open-market dispatch versus closed dispatch procedures in adjacent areas led to higher unscheduled power flows through its system during the 13 months ending April 30, 2006. PJM explains that this “free-rider” problem has disrupted RTO operations and contributed to the “revenue adequacy issues” for financial transmission rights (FTRs) that PJM has experienced over the last year. (PJM, initial comments, pp. 33-35.)

PJM argues that FERC’s plan for separate-but-equal open and closed dispatch regimes in adjacent regions will prove untenable. It urges FERC to mandate reciprocal and open redispatch protocols for RTO-bordering and vertically integrated utilities, and even to authorize RTOs to include provisions in their tariffs to charge non-RTO members for making use of the RTO’s open dispatch markets.

PJM’s reference to FTR revenue adequacy recalls the situation in the recent Chambersburg complaint (see Commission Watch, “Bad Day at Black Oak,” Oct. 2006), which opponents have cited widely as reason to reject the open dispatch concept.

PJM’s free-rider example shows how the failure to assign proper prices to grid services can spark arbitrage, undermining markets even in areas favorable to trading.

Redispatch and Native Load

Of course, FERC’s reform plan in actuality seeks to have it both ways. It disclaims any interest in RTO market regimes, yet proposes a very slimmed-down “lite” version: a new “conditional firm” PTP service, plus a modest redispatch option that goes beyond the scope of redispatch now required under Order 888.

The conditional firm idea has won the more favorable review. FERC sees constraints occurring during a few hours in the year and asks why that should give cause to transmission providers to deny a service request in its entirety. So it proposes a CF service that would qualify as firm in all but those few delineated hours.

On the other hand, open dispatch opponents also have widely panned FERC’s more limited redispatch proposal, whereby denied grid customers could ask the transmission providers to consider redeploying power plants under the provider’s control as a potentially cheaper alternative to a grid expansion. But here is the irony: The provider must make the decision before conducting the facilities study that reveals the cost of the redispatch.

Some opponents reject redispatch as an unlawful extension of FERC jurisdiction over the generating sector. PJM counters, however, that OATT redispatch should be seen essentially as another form of transmission service, in the same way that FERC justifies its authority over generator interconnections.

Quite naturally, EPSA joins PJM in arguing that FERC has authority to control dispatch. EPSA predicts that this question must be decided eventually, if not immediately in the OATT NOPR case. For example, it references the complaint filed by the Arkansas Public Service Commission, asking FERC to compel Entergy to deploy merchant gen units to displace some of Entergy’s fleet of rate-based assets. (See FERC Docket EL06-76, filed June 7, 2006.)

Then there remains the enigma of “native load.” Many NOPR comments claim FERC’s redispatch would violate native-load rights, as redefined in EPACT sec. 1233, now codified at Federal Power Act sec. 217. These objections, however, seem far off base. Taken together, in fact, the hundreds of comments reveal considerable industry confusion about what is meant by native load.

One view sees two different types of load—merchant load and “utilities native load.” (N.M. Atty. Gen., initial comments p. 10) This view defines native load as that “distinct set of customers that are regulated at the state level.” (South Carolina Elec. & Gas, initial comments p. 6.) Yet FPA sec. 217 declares that a load obligation secured only by contract still will qualify as “native” load.

Another view sees the native-load protection extending to generating assets, arguing that open dispatch would undermine “the current dedication of an integrated utility’s generation to first meet its native load service obligation.” (Large Public Power Council, reply comments, p. 18) Yet again, FPA sec. 217 appears to protect only the transmission rights of native load. No mention is made of rights tied to power plants.

EPSA counters that native load as defined in EPACT 1233 and FPA 217 should apply to all obligations, “whether … served by a competitive supplier, an affiliate of the transmission provider, or by the transmission provider itself.”

This notion comports more fully with the Chandley/Hogan world, where “all … are treated as native loads; thus all loads receive all of the benefits and protections.”