FERC's Market Design: The End of a 'Noble Dream'

Deck: 
How state opposition cowed the feds and turned a powerful rule into just a set of talking points.
Fortnightly Magazine - February 15 2003

How state opposition cowed the feds and turned a powerful rule into just a set of talking points.

A funny thing happened on the way to a standard market design (SMD). What began as a full-fledged rulemaking-with the Federal Energy Regulatory Commission (FERC) giving instructions and imposing deadlines on the electric utility industry-now has degenerated into little more than a set of talking points.

Talk about cold feet.

After first asking for industry reaction by Nov. 15 and weathering a storm of protest, a chastened FERC invited a second round of comments by Jan. 10, and then extended that deadline to Feb. 28. With each round of criticism, the commission seemed to lose just a little more nerve.

One state, Iowa, suggested that the SMD commits such an illegal trespass on state jurisdiction that FERC would get more mileage out of the effort if it would simply cancel the compulsory aspects of the plan. The feds, said Iowa, should convert the docket (RM01-12-000) into a so-called "statement of policy"-offering only guidance to those students who might be interested.

FERC appeared to do exactly that, when, in mid-January, it issued a press release promising yet another white paper and-more than likely-yet another re-evaluation of the plan, followed by additional rounds of constituent feedback.

In short, regulators from more than a dozen states see the plan as dead in terms of regulatory discipline.

"Profound and unjustified," say regulators from Washington state. Solves "problems that do not exist," adds Louisiana. Regulators in Arizona call SMD "a noble dream," yet one that is "unwarranted" in its expansion of federal authority.

The industry appeared to have good reason to question FERC's legal authority to promulgate a new market architecture for wholesale power transactions that would bring some traditionally state-regulated features under federal purview.

As many parties have noted in their filed comments, the Federal Power Act bars federal intrusion into questions relating to retail electricity distribution service. They have cited that point in opposing FERC's bid to seize jurisdiction over electric transmission employed in retail service and to create a tariff with a single transmission product ("network service"). That move would de-list "point-to-point" service and kill the so-called "native load preference," whereby utilities can reserve grid capacity for the future potential use of retail customers who take electricity as a fully regulated product.

Indeed, the Tennessee Valley Authority argues that Congress ratified the native load preference as late as 1992, in the Energy Policy Act (sec. 212a), when it ruled that transmission (wheeling) customers must pay all costs, including costs of new facilities required for the service. TVA argues that this provision "essentially codifies the native load preference," since any existing grid capacity that has been used or relied upon to serve native load cannot be appropriated or subordinated to the wheeling request. By contrast, FERC's SMD would bar the traditional allowance for "capacity benefit margin" (CBM) to allow for future growth of native load. Instead, FERC would "monetize" CBM and put native load at risk for future load growth, by requiring retail customers who take bundled service to acquire congestion revenue rights (CRRs) to guarantee the same degree of grid access.

And indeed, a state public utility commission (PUC) as progressive as Pennsylvania has gone on record opposing any erosion of CBM rights, as proposed in the SMD.

Particularly galling to state regulators and some others is that FERC itself seems so willing to waffle on details of SMD for some beneficiaries, even while coming down hard on others.

For example, the commission has announced that, in the interest of moving things along, it will accept regional market models that do not mesh perfectly with the SMD. Thus, in recent decisions that approved initial plans to get regional transmission organizations (RTOs) started in the Southeast (SeTrans) and out West (WestConnect and RTO West), FERC declared:

"We do not intend, in the final SMD rule, to revisit prior approvals or acceptances of RTO provisions because of possible inconsistencies with the details of the final rule. This Commission intends to take all appropriate steps at the final rule stage of the SMD to ensure that, to the extent we have already approved or conditionally approved RTO elements, these approvals would remain intact." ()

No wonder a state like Arizona asks why, "if the SMD rule will be trumped by RTO orders," it should not be allowed to devise a unique regional solution for its own constituents.

These and other concerns spell a heap of trouble for FERC's SMD. The issues are numerous-far too many to be listed here-but it is possible to outline several of the most significant questions that have been raised over the past couple of months by the nation's state regulators.

The Western Problem

Can the Western Interconnection live with locational marginal pricing (LMP) to reconcile transmission congestion and govern a spot energy market? That question continues to divide utilities and regulators out West, with no clear resolution in sight.

PG&E Corp., coming fresh from the California experience, where the state's independent system operator (ISO) eschewed a fully developed and security-constrained LMP model in its first disastrous fling with markets, now clearly prefers LMP, both in theory and in practice, across California and the West.

"There is no fundamental reason," says the company, "that makes LMP incompatible with hydro systems."

As PG&E explains, "PJM, New York, California, and New England all have hydro resources and pumped storage facilities that operate (or will operate) under LMP electric resources."

The company recognizes that early debates "raised concerns about centralized unit-commitment-based designs," including "overly complex optimization algorithms, limited transparency, large uplifts and no-load costs, a consequent departure from uniform price auctions." But since then, says PG&E, "the reliably successful experience in the New York and PJM day-ahead markets has lessened the scope of such fears."

Regulators from the state of Wyoming would appear to agree, but not so the Washington Utilities and Transportation Commission (UTC), which continues to insist, as do the utility members of RTO West, that LMP and hydro don't mix.

The reasons lie with the unique topology of the Western grid. As explained by the utility participants in RTO West, the regional grid system sports a Jeckyll-and-Hyde personality. The system is not homogeneous. On one hand, the companies say, the West enjoys a "relatively robust grid system" strung out along the length of certain major river systems. Beyond that, however, the remaining grid system is best characterized as "relatively lean."

The robust half of the Western grid does a good job of integrating the high-capacity (but energy-constrained) hydro resources located in those valleys, allowing highly efficient regional planning. The other half, however, delivers energy from remote thermal baseload plants, such as the coal-fired plants located in the eastern part of the RTO West region. It is marked by relatively expensive long-haul lines designed "to just fit" the local load or generating plant, spanning a region with a sparse population. These lines could just as well be classified as generation assets for rate-making or allocation. In no way do they resemble the spider-web sort of pattern that you would see in an integrated grid system back East. In many instances, the companies say, the grid owners and operators must pay close attention to specific local generator characteristics (voltage, VARs, output levels, remedial action schemes, etc.), just to support the transmission capability into, out of, or through the area.

These two bifurcated grids, the companies say, have encouraged a highly complex structure of bilateral transmission contracts that carry physical rights and maximize system performance. The structure has led RTO West to propose a variant at odds with SMD, whereby the region "catalogues" physical rights and allows participants to choose if and when to convert those physical rights into financial congestion rights ("financial transmission options"-FTOs-in Northwest parlance).

If these contracts were abrogated, the companies say, and if all physical rights were redistributed as financial congestion rights, whether as full FTO "strips" or chopped into hourly pieces, then contract rights holders would be certain to see a significant reduction of their pre-existing rights. The whole, then, is greater than the sum of the parts. Full conversion of physical contract rights into financial hedging rights would cause short-term transmission revenues to fall, making it difficult for transmission owners to recover fixed costs and creating "a substantial cost shift" among grid owners in the RTO.

Against this backdrop, the Washington UTC insists that the region cannot flourish with LMP:

"The hydropower system offers great dispatch flexibility. … This is already our most cost-efficient dispatch, since the hydropower system has no direct marginal fuel costs. LMP, transmission congestion pricing, day-ahead markets, single-system dispatch, and mandatory real-time balancing markets could, in theory, be implemented in the Pacific Northwest, but to what purpose?

"Worse yet, the [SMD] provides no assurance that the flexibility benefits of the hydropower system in the Pacific Northwest will not be lost under the proposed congestion management and CRR mechanisms."

In particular, the UTC chafes at claims by FERC staff that New Zealand makes LMP work with its largely hydro system. As Washington state points out, New Zealand can dispatch its hydro plants more or less independently, while most hydro plants in the Pacific Northwest are placed along a single river system and "cannot be operated independently or pitted in competition with each other."

Yet they see things differently in Wyoming.

"We generally favor the Commission's LMP approach," says the Wyoming PSC.

"Although some in the West argue that an LMP approach … is incompatible with the West's historic operating protocols (particularly in the Northwest) we believe LMP is central to effectively managing congestion.

"Although it will not be a simple task, we believe that the West will ultimately reach agreement on an LMP method that substantially preserves the principles articulated by the FERC in its SMD."

Mitigating Unlawful Prices

In its SMD rule, FERC proposes four remedies to mitigate anticompetitive activity and keep power prices reasonable:

  • A bid cap in centralized spot markets to offer an overall safety net, such as the cap of $1,000 per megawatt-hour (MWh) now in place in ERCOT and Eastern grid regions;
  • A resource adequacy requirement for retail utilities, similar to the traditional reserve margin;
  • A voluntary price mitigation scheme, such as the automatic mitigation procedure (AMP) in place in the New York ISO, which compares bidding conduct to historical bid reference levels, and evaluates the impact of outlier bids on prices; and
  • Some form of must-run regime for generating plants that might exercise market power within a local area ("load pocket") plagued by transmission constraints.

Though this rule might appear simple, it has raised questions among state regulators. Should price mitigation schemes recognize the right of power producers to recover not just variable and fixed costs, but opportunity costs that reflect scarcity rents collected during periods of regional supply shortages, as FERC has proposed? Connecticut regulators say yes (with scarcity value based on the highest-cost losing bidder), but others disagree.

Some states, such as Wisconsin, call for market mitigation for bilateral trading, as well as for centralized spot markets, as FERC has proposed. In similar fashion, regulators at the Connecticut and the New England PUC association (NECPUC) urge FERC to extend mitigation beyond cases involving transmission constraints or local market power to cover situations such as extreme demand peaks, in which virtually all regional units are dispatched, and suppliers can take advantage, knowing that grid operators must dispatch nearly all bidders.

The New England regulators fear market power "even in areas free of transmission constraints," and they describe the current safety net bid cap of $1,000/MWh as "wholly inadequate" to protect consumers. Georgia regulators share that view:

"The idea of FERC imposing a system of electricity price regulation on Georgia and the Southeast where caps of $1,000 are deemed a necessary part of the FERC systems sets off alarm bells."

Nevertheless, NECPUC admits that FERC gave fair warning last fall in a case that reviewed market design in ISO New England, when it asked the ISO to explain why a safety net bid cap of $1,000/MWh would not provide sufficient consumer protection in areas without transmission constraints. .

Policy disagreements can emerge even in regions with a lot of spot market experience.

In PJM, for example, market rules trigger mitigation in either the day-ahead (DAM) or real-time markets whenever and wherever an interface constraint creates a local load pocket that requires generation to be dispatched out of merit order. Once triggered, the plan will mitigate bids for must-run plants operating within the load pocket, usually by reducing the nodal clearing price to a level equal to the generator's estimated incremental cost, plus 10 percent. The Maryland People's Counsel praises the PJM method and urges FERC to incorporate it in the SMD for load-pocket relief. (PJM also allows other alternative mitigation options for load pockets: either a simple price negotiation, or a reduction in the clearing price to the level of the average of LMPs at the same generator bus for hours that featured a merit-order dispatch.)

One might think PJM would serve as an exemplary model for FERC's SMD. Especially here, where the SMD lays down a heavy hand to keep prices low. But it seems that FERC can't win for losing, because even in this case, state regulators in PJM's home state of Pennsylvania actually would prefer a different rule-one that would adopt less of a regulatory approach and more of a competitive or market solution.

Of course, the Pennsylvania PUC acknowledges that some have proposed to improve the PJM rules by substituting some sort of guaranteed generator payment reflecting variable and fixed costs plus a rate of return. But again, the PUC sees such solutions as so much "formula tweaking, headed in the wrong direction towards more RTO rate setting, rather than less."

For its part, the Pennsylvania PUC would prefer some sort of "proxy competitive price," derived either from historic unconstrained hours in the region on different days, or proxy LMP prices for the hour in unconstrained regions outside the load pocket (or perhaps even the historical average successful LMP bids by the generator itself, during unconstrained hours, which would have the effect of making the unit its own price proxy).

The point, says the PUC, is to avoid an "irrationally generous" mitigation scheme that destroys any incentive for efficient plant operation by setting an administratively mitigated price that still exceeds the plant's incremental cost by a substantial margin.

Otherwise, the PUC says, such plants could still operate, despite mitigation, as "the only big fish in the local pond, sure of a risk-free market for power, and fiercely contesting any new entrants."

Funding Transmission Expansion

Politics makes for strange bedfellows. As an example, consider that many of the states that oppose FERC's SMD plan most ardently-Georgia, Arkansas, Kentucky, North Carolina-are themselves some of the biggest boosters of perhaps the most progressive element of the plan: participant-funded transmission (PFT), which calls on merchant generators to pay for grid expansion in place of rolled-in pricing. Indeed, these states typically will defend PFT as an essential adjunct to an LMP congestion pricing system, so as not to distort price signals. Kentucky has even passed a state legislative resolution endorsing PFT.

"Participant funding leaves decisions over 'economic' grid upgrades to free market forces," says the Arkansas commission. "This allows for full and fair competition between generation, transmission, and demand-side resource solutions to congestion."

At the same time, however, these states fervently oppose LMP and market pricing for energy as unworkable for their own constituents. Meanwhile, the states that prefer traditional rolled-in pricing over participant funding, such as North Dakota, tend also to favor postage-stamp pricing over license-plate pricing.

States in the West and the Great Plains generally prefer socialization of grid expansion costs over the broadest possible region, as they must seek to open up their resources to boost power exports to urban consuming areas. In the words of the Crescent Moon grid group, they fear that FERC may be "backsliding" in favor of license-plate pricing and ignoring its original vision of wide-area, pancake-free pricing.

The Louisiana PSC has panned SMD outright. It sees disaster in any mandatory plan to force utilities to divest themselves of transmission assets to form independent transmission providers (ITPs), as FERC would require through its market design. Yet two Entergy utility subsidiaries, in a pending case in that state (La. PSC Docket No. U-25965), have offered studies showing that any plan for them to join either SeTrans (planned as an RTO) or the Midwest ISO (already RTO-certified) will likely increase costs unless the deal is coupled with a PFT policy that compels new merchant generators to pay for grid expansions they need for new power plants. The plan offers an alternative to traditional rolled-in pricing, which would impose the costs of such expansion on retail ratepayers of transmission-owning utilities.

In a similar example, the Southeast Association of State Utility Regulators (SEARUC) commissioned a study from Charles River Associates that found virtually no consumer benefits from a SeTrans RTO without participant-funded transmission.

FERC indicates a willingness to allow participant funding, but only for new transmission facilities that are included in a regional planning process conducted by an independent grid operator, be it an RTO, ISO, or ITP. And therein lies a key issue: Can participant funding exist on its own, outside the rest of the SMD framework?

In Arkansas, the commission says there are times when participant funding needs to be supplemented by transmission planning studies, and when rolled-in pricing might be warranted "on a limited basis." For example, a grid expansion plan funded by one merchant generator might end up benefiting several such new plants in the same region. In that case, Arkansas says, the ITP would need to develop a cost-sharing protocol applicable to all merchant plant beneficiaries.

Consider another example offered by Arkansas regulators:

A new merchant power plant might require a network upgrade, achievable with a new line with capacity of either 138 kV or 230 kV. This merchant plant might well choose to select the smaller upgrade in order to leave some congestion intact and optimize the value of CRRs, even if the 230-kV option would eliminate 100 percent of the congestion (driving CRR values to 0) at only a slightly higher construction expense.

"In such a case," says Arkansas, "the ITP may wish to require the 230-kV expansion option but only charge the merchant plant for the suboptimal 138-kV cost. The difference in cost … would be rolled-in or reallocated. The final [SMD] rule should allow for such modifications to the pure participant-funded structure (as part of an RTO-ITP-approved transmission plan)."

Nevertheless, regulators in North Carolina (another pro-PFT state) oppose FERC's RTO/ITP overlay as a requirement for participant funding. North Carolina sees an immediate need for a new transmission expansion policy that does not involve socialization of costs. Participant funding, says North Carolina, "is unrelated to the existence of independent operation of the transmission system … so that implementation should not depend on the existence of an ITP."

Of course, there is more going on here than meets the eye. The planned SeTrans RTO has invested a lot of planning effort in reliance on participant funding as a viable new policy to avoid forcing transmission owners to foot the bill for grid expansion designed to accommodate new local generators, only to see the local power exported out of the region.

In fact, the Pennsylvania PUC has observed that participant funding has become a two-edged sword. As the PUC explains, it sounds like a free market idea, but it can easily be turned into a weapon (a "cudgel," Pennsylvania says) against the introduction of competitive markets. Perhaps there is a little bit of that going on in the states that have embraced participant funding, yet have rejected LMP and SMD, as the Pennsylvania commission hints in its SMD comments filed Jan. 10:

"Participant funding, raised as a high-level issue primarily by some states and stakeholders in the Southern and Northwestern U.S., has been a flashpoint for criticism of the proposed standard market design. … Typically, the issue has been framed as [one] of fairness. … But the concept … may enable the remaining incumbent vertically integrated monopoly electric utility companies to disadvantage new market entrants and potential competitors.

"Incumbent utilities, having built the existing transmission grid, are likely to be entirely satisfied with its topology and transfer limits.

"Participant funding as a general principal of SMD … would enable the incumbent monopoly utility to protect its own generation business."



Learning from comments filed by state utility regulators in FERC's SMD case.

ARKANSAS-CONGESTION UP?

Notes that PJM has reported dramatic increases in congestion charges even as its bid-based market has developed, with costs rising from $53 million in 1999 to $271 million in 2001, a more than five-fold increase. Compare that situation to the recent study conducted by Charles River Associates, which estimated intrazonal congestion charges at $200 million a year for the planned SeTrans RTO, and $475 million annually for the entire Southeast region, plus an additional $260 million annually for interzonal and export congestion fees.

CALIFORNIA-BETTER BOARDS.

Warns that FERC's selection process for RTO board members could produce a "moral hazard." Notes that under FERC policy, the boards of the four existing "stakeholder" RTOs (MISO, PJM, New York, New England) have come to be made up 50 percent of former utility executives, a significant number of whom are retired. An added 8 percent had substantial utility ties (consulting, etc.), while 21 percent came from finance and 6 percent from an information technology background. A cursory review showed less diversity of experience on RTO boards than for traditional utilities.

ILLINOIS-LET FERC BE BOSS.

Wants market monitoring (MM) staff at RTO or ITP to function as contract agent of FERC, and be funded through a mechanism that remains separate from the RTO/ITP funding protocol. "It is simply not reasonable to expect the MM to be accountable to the FERC and independent of the RTO, under a framework in which the MM is selected by the RTO, contracts with the RTO, has its budget set by the RTO, has terms of payment controlled by the RTO, and has its invoices paid by the RTO."

KENTUCKY-WATCH PAYROLLS!

State regulators warn that FERC policy already is taking a "toll" in Kentucky, as retail rates there are likely to rise in tandem with the tens of million of dollars in extra administrative costs incurred annually at the Midwest ISO. "With an annual operating budget approaching $70 million, a capital budget of more than $30 million, and a staff of 210 people, MISO is clearly not an inexpensive undertaking."

Kentucky also is a party to a legal appeal in the federal courts (Midwest ISO Trans. Owners v. FERC, Case Nos. 02-1121, 02-1122, D.C.Cir.) that seeks to overturn a FERC ruling that rejected a settlement between MISO and the Kentucky PSC that would have saved the state's native load customers from paying MISO administrative costs.

MAINE-RESERVE MARGINS?

State opposes the new resource adequacy requirement (RAR) contained in the SMD. It argues that FERC's plan for a 12 percent margin, as measured against retail "load," cannot work, because in states that have moved to retail access, the retail supplier no longer has any captive load. Maine PUC says that retail suppliers could avoid FERC-imposed penalties for failure to maintain the required margin simply by exiting the market prior to "real time," in effect ignoring the RAR rule entirely.

The PUC proposes instead to adopt a structure that it calls a "central buyer model," whereby the RTO/ISO/ITP determines the capacity required for the market as a whole through an auction, and then buys the commitments needed to provide that product for the relevant period. There would be no direct link between any particular load and any particular capacity; that would be "impractical and illogical," the PUC says.

OHIO-FULL DISCLOSURE, NOW.

Says states lack access to data needed to make market-monitoring work on all levels. Notes that PJM has a 6-month moratorium on release of MM information to state regulators, and then can release only the fact of market abuse-not the identity of the culprit. Says New York has a three-month moratorium, while New England keeps information confidential for seven days. By contrast, the Midwest ISO prohibits all state regulatory access to MM information, forever. "This situation is unconscionable," says Ohio.

TEXAS-CALIFORNIA REDUX?

As the PUC explains, ERCOT lacks a day-ahead energy market (DAM) and in the past has tried to manage congestion on a zonal basis (as California tried to do and failed).

Now, however, the PUC claims that its Market Oversight Division has developed and implemented a system for congestion management that is based not on nodal locational marginal pricing (LMP) but on an administrative assignment of hypothetical congestion charges needed to achieve a pre-determined, formulaic optimization of the transmission grid.

Texas fears that, like California, zonal congestion management method in ERCOT has led to a "dec game" (manipulation of decremental bids). Also, a huge amount (nearly 900 MW) of wind-powered generators has located unexpectedly in West Texas, on the wrong side of a major local transmission constraint, in reliance on faulty locational market signals supplied by the old zonal congestion system.

The PUC now says it has visions of creating an energy DAM with ex- ante pricing that would be operated privately, because with its newfangled congestion management system, the DAM could receive bids and execute dispatch without a security-constrained protocol.

Articles found on this page are available to Internet subscribers only. For more information about obtaining a username and password, please call our Customer Service Department at 1-800-368-5001.