The Too-Perfect Hedge

Deck: 

Congress gives FERC an impossible task: Craft long-term transmission rights to save native load from paying grid congestion costs.

Fortnightly Magazine - May 2006

If “perfect” be the enemy of the “good,” then look no further for proof than in Federal Power Act section 217(b)(4), enacted by Congress in last year’s landmark energy bill, EPACT 2005.

That section, as many see it, would grant a virtual long-term immunity against liability for transmission congestion costs—a “perfect hedge,” if you will—to any electric utility that retains the traditional obligation to serve native load (a load-serving entity, or LSE), even if that utility operates within the footprint of a regional transmission organization (RTO), and even if the RTO essentially has adopted a day-ahead spot market for wholesale power, complete with locational marginal pricing (LMP) and financial transmission rights (FTRs).

Before RTOs, the utility seeking base-load generating capacity simply would build its own power plant and transmission lines, or sign a long-term supply contract with a plant owner and buy or reserve enough transmission service to ensure physical delivery. Congestion? Not a factor, except for the occasional line outage or blackout. The power flows, or it doesn’t.

Fast-forward to today’s financial world. The LSE as RTO customer gains access to all manner of cheap power supplies—a cornucopia of supply options—but to purchase any actual quantity for delivery to consumers it must “buy through” any congestion it encounters, as measured in the difference between the price of power at the source and sink.

FTRs offer a financial hedge against congestion risk, of course, but no more, since the typical RTO rarely will guarantee the sufficiency of FTR payments through what is called “full funding.”

Even if the LSE owns its power plant and chooses to “self-schedule” the unit through the RTO’s dispatch process, the LSE in effect “sells” the plant output to itself at the wholesale price that prevails at the point of injection into the grid (the “nodal” price), and then “buys” back the output at point of withdrawal from the grid and delivery into its local distribution network, again at the prevailing local price. In each case, the nodal price reflects the opportunity cost of alternative power available at the given location.

Under this RTO market regime, a utility might pay $1 billion to construct a 600-MW coal-fired generating plant, with running costs of about $20 per megawatt-hour (or pay dearly for that capacity over a long-term supply contract), but then end up “buying” the output on behalf of its native load at $80/MWh—the approximate running cost of a gas-fired peaking turbine that would cost a third as much to construct. The extra payment represents congestion—the premium required for moving power in the same direction as everyone else (sparsely populated producing area to the densely populated consuming area).

Attorneys Robert McDiarmid and Cynthia Bogorad (Spiegel & McDiarmid), representing the Transmission Access Policy Study Group (TAPS), an informal association of transmission-dependent utilities (TDUs) in more than 30 states that comprises mostly municipal utilities and other public power entities, argue (as many others have) that this required payment of congestion costs thwarts best efforts at power-plant development:

“The high installed cost of base-load coal generation,” they say, “cannot be justified if, as a result of congestion, the LSE is likely to pay gas-based LMPs for the energy produced.”

According to testimony from Thomas W. Ingwers, director of energy trading and contracts for the Sacramento Municipal Utility District (SMUD), RTO market regimes built on nodal LMPs and FTR congestion hedges threaten to create a new version of “stranded investment” in utility-owned generation located far from load because of the steep energy price differentials seen between producing and consuming areas.

In SMUD’s case, notes Ingwers, the absence of any long-term firm rights in transmission service may prove “a serious obstacle” to the utility’s planned Solano Wind Project, a 205-MW development to be located within the control area of the California Independent System Operator (Cal-ISO).

Ingwers adds that last September, during California Independent System Operator (Cal-ISO) stakeholder discussions regarding the development of a new market design with full nodal locational pricing and financial transmission rights known as CRRs (“conges- tion revenue rights”), Bonneville Power had warned the ISO that congestion-cost exposure in the style of the northeastern RTOs could lead BPA to reconsider decisions to buy or sell in the new market. Ingwers cites written stakeholder comments from BPA, available on Cal-ISO’s Web site, stating that California’s expected configuration for CRRs was “not suited to the sporadic nature of our marketing of surplus energy to the ISO.”

(Cal-ISO has since submitted its new market design to the commission, including financial CRRs. See FERC Docket No. ER06-615, filed Feb. 2, 2006.)

This new problem with “stranded” generation becomes even more vexing, however, because RTOs generally offer FTRs only for terms of one year or less. By contrast, a load-serving utility in a state that retains full, traditional cost-of-service regulation, without retail choice, might well want a 30-year long-term FTR to match the life of its coal-fired power plant or power supply contract.

None of this, however, will come as any surprise to the Federal Energy Regulatory Commission (FERC), which has recognized the problem for some time now, and which last year (in its Docket No. AD05-7) had asked the utility industry for ideas on how to solve it. (See, “Coal’s Raw Deal: The bias in RTO markets, and how FERC might fix it.” Public Utilities Fortnightly, September 2005, p. 20.)

Now however, comes Congress, attempting to fix the problem in EPACT sec. 1233. That section gives a complex and ambiguous set of instructions to FERC, and then commands the commission to carry them out within the year. Those instructions appear to many to guarantee to LSEs the same degree of protection from congestion charges that they would have enjoyed before RTO markets, through a physical, long-term reservation of transmission capacity rights (an “LTTR”) and to do it within the operating and market protocols of the RTO itself.

There’s a problem, however. It very likely cannot be done.

An Ambiguous Assignment

FERC took the first step toward compliance three months ago, when it issued a notice of proposed rulemaking (NOPR) governing long-term transmission rights (LTTRs) in “organized electricity markets” (defined as the day-ahead markets run by RTOs that feature LMPs and FTRs). Yet it remains unclear how FERC should attempt to carry out this near-impossible assignment.In fact, the task appears so daunting that, rather than craft a clear set of rules, the commission has chosen instead to lay out a bare outline of eight so-called “guidelines” to govern the process, and then punt the question back to the RTOs. This tactic only has heightened the confusion, as the guidelines appear so vague and internally inconsistent as to permit virtually any interpretation. (See Docket No. RM06-8, Feb. 2, 2006, 114 FERC ¶61,097, plus industry comments filed through April 3, 2006.)

Above all, the new law appears vague, lacking adequate definition on a number of points, as does FERC’s proposal:

1. Duration. How long is long-term?

2. Physical vs. Financial. Must LTTRs be physical, or will it suffice simply to lengthen the one-year financial rights already offered by RTOs?

3. Subsequent Modification. If the hedge, once allocated, “should not be modified,” as FERC insists (except upon extraordinary circumstances), then isn’t the LTTR in reality a physical carve-out that could distort RTO markets?

4. Mitigating Impacts. Again, if physical, then how can FERC mitigate (“balance”) the adverse impact on RTO market participants that will occur because of fewer available short-term FTRs?

5. Revenue Shortfalls. If financial, should RTOs make LTTRs fully funded, and if so, then who backs up any revenue shortfall?

6. Priority Issuance. Should LSEs with native-load service obligations enjoy priority in allocations or auctions of LTTRs? Even if the obligation derives from contract rather than operation of law? How long a term is required for such contracts?

7. Sham Transactions. If priority is warranted, then do LSEs lose priority if they lose load or contracts? Can they assign excess LTTRs to lower-priority retailers, or resell LTTRs at a profit? If so, wouldn’t that encourage sham transactions, requiring RTOs to scrutinize or “police” contract claims?

8. Conflicting State Laws. How should FERC deal with retail access states that may conduct auctions to reassign the native-load service obligation (Ohio, perhaps?), or impose term limits (3 years or so) on contracts purchased through statewide gen supply auctions, such as Illinois or New Jersey?

9. Simultaneous Feasibility. If initial LTTR allocations take place, as FERC proposes, without requiring LSEs to participate in auctions, how can RTOs assure simultaneous feasibility of all issued FTRs?

10. Participant Funding. If RTOs should allocate extra LTTRs to those who fund grid upgrades that increase availability of FTRs, as FERC suggests, then how do RTOs reconcile two conflicting ideas: (1) That LTTRs must identify a particular source and sink, versus (2) that RTOs issue FTRs and verify simultaneous feasibility without regard to actual grid usage or schedules, because economic theory teaches that Simultaneous Feasibility is assured if the grid topology assumed in the FTR allocation matches the topology employed in the security constrained dispatch and setting and clearing of nodal LMPs.

Congress offers little guidance. It simply instructs FERC to act in a way that facilitates grid planning and expansion, and to meet the “reasonable needs” of LSEs to satisfy their service obligations and secure “firm transmission rights or equivalent tradable or financial rights on a long-term basis for long-term power supply arrangements.”

The Bonneville Power Administration proposes a five-year term length for LTTRs. TAPS suggests 10 years. One popular idea would link LTTR term lengths to the RTO’s planning horizon for transmission, on the theory that Congress designed EPACT sec. 1233 to promote grid expansion.

The American Public Power Association (APPA) worries that RTOs could “go short,” and in fact, both Cinergy and the New York ISO suggest that a one-year duration ought to qualify as “long-term,” since current FERC rules in many areas (such as Order 888, the original open-access mandate) say that power contracts as short as one year in length will qualify as “long term.” If that’s the correct definition, then the RTOs already would satisfy EPACT by offering one-year FTRs, and the law would serve no purpose.

Financial Dynamics

A coalition of municipal, government- or consumer-owned utility systems from New England contends that simply extending the length of RTO-issued FTRs, while retaining the same RTO-designed attributes for allocation, trading and risk management would not do the trick because of an overall lack of simultaneously feasible FTRs or auction revenue rights. (Some RTOs award revenues from FTR auctions to utilities in place of direct allocations of FTRs. Utilities then can use these “ARRs” to purchase FTRs, or can choose simply to keep the money.)

The New England public systems complain also that ISO New England allocates ARRs among congestion-paying LSEs on the basis of zonal load-ratio shares that reflect only the average zonal market-clearing price of FTRs in the auctions, offering no assurance, then, that ARR recipients can use the revenues to buy specific FTRs to hedge congestion on a particular grid path between a specific source and sink.

The requirement of simultaneous feasibility, alluded to above, tends to make sequential allocations or issuances of financial transmission particularly problematic, and instead favors an all-in-one calculation performed by RTO software. The underlying reason, according to testimony from Scott Harvey (director) and Susan Pope (principal), from LECG LLC consulting, submitted to FERC in support of Cal-ISO’s recently filed new market design, stems from a mathematical proof developed in 1992 by Harvard economist William Hogan, concerning the adequacy of congestion revenue collected by RTOs to pay back the congestion costs incurred by FTR holders. (See, testimony of Scott Harvey and Susan Pope, pp. 16-17, FERC Docket No. ER06-615, filed Feb. 8, 2006, citing William W. Hogan, “Contract Networks for Electric Power Transmission,” Journal of Regulatory Economics, Vol 4, No. 3, Sept. 1992, pp. 211-42.)

As Harvey and Pope explain, Hogan’s work proved that if an RTO relies upon the same grid topology both to (A) conduct a security-constrained dispatch of generating resources and calculate nodal locational clearing prices, and (B) determine the simultaneous feasibility of FTRs (i.e., whether the identified pairs of sources and sinks can support the nominated quantities of power at the same time without violating reliability rules), then, by definition, the congestion revenues that the RTO collects (LMP differentials) will exactly equal the amount needed to pay off the FTR holders (assuming FTRs are “obligation” instruments, and not options), regardless of whether the actual energy transaction schedules on the grid match the FTRs, or even whether the FTR holders choose to schedule any power at all.

This theorem thus allows RTOs to maximize savings in generation and in grid efficiency, and at the same time determine the winning bids in FTR auctions by clearing that single set of bids that produces the largest auction value (the sum of winning bids), while satisfying the simultaneous feasibility requirement.

In other words, if FERC should attempt to mix this system of financial rights with a limited set of carved-out long-term physical rights, the entire construct will come crashing down.

At TAPS, however, they believe they have a solution. TAPS points to a concept called a “dispatch-contingent FTR,” which the group likens to a transmission right developed in MISO to deal with carved-out physical rights, known as “MISO Option B,” and which FERC defines in its proposal at paragraph 81.

According to FERC and TAPS, the RTO software would “model” this type of right as a fully financial right, with the RTO instead of the LSE holding the rights to the revenue payout. Thus, the RTO would collect congestion revenue and then pay it back to itself. The LSE, meanwhile, could treat the right as a fully physical right, and would be required to schedule transmission in the day-ahead market to match the quantity and path of the right. This concept would also avoid another criticism of physical transmission rights: namely, that they give rise to the phenomenon known as “phantom congestion.” (See, Initial Comments of Transmission Access Policy Study Group, pp. 3-4, 30-33, FERC Docket No. RM06-8, filed March 13, 2006.)

A Stale Business Model?

A common refrain heard from many public power interests insists that EPACT sec. 1233 represents legislative validation of the traditional business model of meeting the utility obligation to serve native load through physical asset rights and long-term contracting. This idea purportedly justifies a preference for LSEs in allocation of LTTRs, even as it imposes adverse impacts on RTO market participants that hold short-term FTRs. Such impacts would be nonexistent, the argument goes, if enough transmission was built; the fact that such impacts are worrisome proves only that transmission infrastructure is too thin in the first place to support RTO market structures.

“Clearly, this bias is necessary,” writes engineer Dennis Delaney, representing an ad hoc group of small, consumer-owned electric systems from Arizona.

Nevertheless, many utilities argue also that business models have changed. As Exelon attorney and vice president Karen Hill puts it, “Serving native load using long-term static resources versus shorter-term dynamic market forces and generation resources is not a preferred business model.”

Representing Constellation Energy, attorney Deborah Carpentier (Dickstein, Shapiro, Morin & Oshinsky) reminds FERC that fared no better in the past:

“Customers that had a fixed-price, long-term physical transmission right in the past were not sheltered from congestion costs. They simply incurred those costs differently … through higher power rates.”

At San Diego Gas & Electric, attorney Paul Szymanski sees the quest for the perfect hedge as clearly unattainable:

“While long-term physical transmission rights might have theoretical advantages,” he notes, “it has yet to be demonstrated how physical rights can be effectively identified and deployed in a looped network configuration where the actions of each market participant instantaneously affect the fights of all other market participants.

“To date, the only practical way to address these interactions is through the exchange of money.”