Beware even the best of attempts at apportioning grid rights and costs.
Bruce W. Radford is editor-in-chief for Public Utilities Fortnightly.
Consider the fate of the borough of Chambersburg, Pa., population about 17,800, and its not-for-profit municipal electric utility, which serves about 10,000 retail customers, give or take.
Earlier this year, the town heard from the PJM regional grid operator that for the 2006-2007 planning year it would receive only about 53 percent of its nominated and requested auction revenue rights (ARRs). That news came despite the fact that Chambersburg had not experienced significant load growth, nor changed its pattern of power-supply sources or ARR nominations in any material way from recent prior years, when typically it had received 100 percent of its requested hedging rights.
(ARRs, allocated to load-serving entities and keyed to identified source/sink pairs, serve within the PJM scheme of things as a financial derivative exchangeable for an equivalent share of FTRs—financial transmission rights—as the more commonly known term. PJM chooses to allocate ARRs, rather than FTRs, as the practice affords more flexibility to the recipients.)
As a result, the Chambersburg municipal utility faces the potential risk of having to pay some $5.7 million in added congestion costs. And this risk, according to a complaint filed recently at the Federal Energy Regulatory Commission (FERC), represents no less than 31 percent—nearly a third of its entire retail-revenue requirement—collected from native-load customers for the 12 months ended in April. To hear it from Chambersburg, PJM’s apportionment has simply failed. Its tariff, therefore, must be unlawful.
The town of Front Royal, Va., of roughly similar size, has joined the complaint. It claims it faces the risk of $1.4 million in unhedged congestion costs, plus $1.9 million in lost FTR revenues, meaning that each customer of the town’s municipal utility could expect to pay an additional $470 per year for electricity, or around $40 per month.
These examples, while perhaps unique, mark only the tip of the iceberg. In fact, this case, one of several involving PJM and now at FERC, poses fundamental questions on how regulators and grid operators should attempt to price and allocate grid rights and costs, and whether such allocations should continue to respect historical artifacts, such as the boundaries between utility service territories:
• Congestion and FTRs. Should regional FTR allocations serve to maximize system-wide commerce and revenues, on a transactional basis, or rather, protect market participants from the sort of injury that resembles traditional rate discrimination? (Borough of Chambersburg, FERC Docket Nos. EL06-94, filed Aug. 1, 2006);
• Regional Grid Expansions. Should grid operators allocate costs of required grid upgrades across the entire system, or only across affected zones? How does one determine which zones are “affected?” Moreover, should they focus primarily on reliability or market commerce in choosing which upgrades to fund? (PJM Regional Transmission Expansion Plan (RTEP), FERC Docket Nos. ER06-1271, filed July 21, 2006; ER06-1474, filed Sept. 8, 2006);
• Transmission Rate Design. Should FERC preserve license-plate pricing for the regional grid-access charge, to respect differences in the traditional revenue requirements of transmission-owning utilities (TOs), or should regulators equalize rates across a broad multi-state area, using a postage-stamp pricing method, or one of a number of other methods, commonly known as “Highway-Byway.” (FERC Docket No. EL05-121, initial decision issued, July 13, 2006, 116 FERC ¶63,007, briefs on exceptions filed through Sept. 15, 2006.)
Each of these cases, in all their complexity, boils down to a simple inquiry. Is the transmission network a public asset, with costs that must be apportioned on principles of equity? Or, rather, is transmission an instrument of commerce, to be priced so as to maximize trade?
FTRs: Just DFAX, Ma’am
Civil War history buffs will know Chambersburg as the “Day-1” entry point for the rebel army as it marched toward its ill-fated rendezvous with federal troops at the Battle of Gettysburg. For our purposes, however, the problem has to more to do with “Day 2”—the so-called “Day-2” wholesale power markets operated by regional independent transmission system operators (ISOs) and regional transmission organization system operators (RTOs), of which PJM is one. More to the point, however, Chambersburg finds that to gain transmission access for its power imports, it must rely largely on the Beddington-Black Oak transmission line (hereinafter “Black Oak”), notorious as one of the most heavily congested flowgates in the PJM system, if not the worst. PJM also has included Black Oak as one of its scarcity pricing regions, meaning that locational marginal prices (LMPs) can exceed the bid cap under certain prescribed situations. This fact, according to the complaint, helped create LMPs in excess of $2,000 on May 30 at the related Grand Point node, no doubt leading other market participants to review their hedge protection against congestion on the line.
In fact, PJM acknowledges that region-wide load growth, together with increased loop flows, plus an increase in demand to hedge congested paths such as Bedington-Black Oak, led to a substantial increase (6,000 MW) in ARR nominations earlier in 2006.
Thus, when ARRs become over-subscribed, forcing PJM to cut back (prorate) ARR nominations to achieve that single combination of simultaneously feasible source/sink pairs yielding the maximum total of auction revenues, it looks first to curtail those ARRs nominated across the key choke points, such as Black Oak. In that way, PJM ensures that the congestion revenues it earns from the differentials in LMPs on actual wholesale transactions in its spot energy market will prove sufficient to pay off the hedging rights claimed by holders of FTRs.
On the face of it, PJM’s ARR allocation appears discriminatory as applied to Chambersburg, though it apparently was conducted according to the letter of the tariff.
The PJM tariff prorates ARRs “in proportion to the MW-level requested and in inverse proportion to the effect on the binding constraints. In practice, PJM prorates excess ARR nominations by reference to the power supply distribution factor (DFAX), which denotes what proportion of a given transaction must flow over a given constraint or grid feature. Each ARR is treated equally. Thus, given a line with a 30-MW capacity, and two separate requests for ARRs worth 100 MW (the total exceeding the line’s capacity), with one request scoring a 40 percent DFAX and the other only 20 percent, PJM will prorate the first request down to 37.5 (30 x 100/200 x 1/0.40), and the second to 75 (30 x 100/200 x 1/0.20).
This method also means that with a line capacity of 180 MW, and two competing ARR requests—one for 10 MW, with an 80 percent DFAX, and the second ARR 50 times bigger, at 500 MW with a 40 percent DFAX—PJM will cut the first request by half (from 10 MW to 4.41 MW), while cutting the second by less than one-twelfth (from 500 MW to 441.18 MW).
Chambersburg cries foul: Why cut its ARR request by a half, when its 20-MW flow marks only a small share of the 2,400-MW capacity of the Black Oak line? Surely it argues, it must be that other, much larger LSEs—perhaps in the higher-priced and populous eastern zones of PJM — are exerting a greater impact on the scarce resource. Chambersburg argues, in essence, that just as PJM assigns responsibility for mandated grid upgrades through its RTEP process on the basis of the grid impacts exerted by entire pricing zones (based roughly on utility service territories), so also it should prorate ARR nominations in the same way. By contrast, PJM’s method, says Chambersburg, will always discriminate against LSEs with small flows and load requirements in geographic locations that force them to rely disproportionately on congested lines.
In fact, here is what is happening: PJM’s ARR allocation method does not compare the treatment of differing geographic zones or corporate entities, as might occur in a traditional investigation of rate discrimination by a state public utility commission. Rather, its method seeks fairness among transactions. It assures the fewest curtailments of ARR nominations across the entire system, and thus the maximum amount of commerce and the maximum amount of hedged congestion revenue. In this way, it represents a complete reversal of traditional notions of rate making.
Upgrades: Planet Neptune?
In an ironic twist, PJM’s recommended allocations of cost responsibility to individual LSEs for transmission upgrades approved by the PJM board of managers as part of the RTO’s most recent RTEP iteration, as per its report of July 21, have raised objections opposite to those of the Chambersburg case. In essence, opponents say that PJM’s allocation method is too primitive, that it focuses too much on company and zonal impacts, while ignoring broader impacts on region-wide commerce that can be perceived only by taking a much broader view.
For example, several parties object the fact that PJM nets the DFAX calculation, combining and netting all counterbalancing needs or benefits within TO-specific territories and zones, thus masking their effects. In particular, the Old Dominion Electric Co-op notes that by modeling the system only at peak, when west-to-east flows are greatest (ignoring the other 8,759 hours of the year), PJM adopts a “snapshot” view that ignores broader trends, and instead searches only for the “straw that broke the camel’s back.” This use of the DFAX method, says ODEC, tends “to skew upgrade allocations to the east, as the prevailing direction of flows in PJM during the peak hour used in the DFAX approach is west to east.” By contrast, says ODEC, “There are many other hours when flows are from east to west.”
Some question why RTEP should seek to allocate costs to zones, rather than simply to customers, on a system-wide, postage-stamp basis. By contrast, the various operating utilities and affiliates that make up the FirstEnergy, PEPCO, and PSEG corporate families argue that PJM should seek more granularity and assign costs to smaller, so-called “electrically cohesive areas” within zones, such as the Delmarva Peninsula, or the northern sector of the PSE&G zone.
Two additional objections appear particularly telling.
First, PJM in its RTEP plan treats the proposed Neptune high-voltage direct current transmission line as a load, since it creates a fixed, firm 20-year right for the Long Island Power Authority to siphon off 685 MW of power from PJM. Thus, PJM’s RTEP proposal would allocate a significant portion of costs for required grid upgrades to the Neptune Line (to its owners? to LIPA ratepayers?), even though Neptune plainly does not qualify as a load “zone” in the classic sense.
This strikes some as a prejudice against power exports. For instance, for the summer 2005 and winter 2005-06 capability periods, certain PJM generators had reserved 1,300 MW of ICAP import rights in New York State, to allow for firm and non-firm deliveries of real-time energy from PJM to the New York ISO, to help satisfy requirements for installed capacity. Yet, as pointed out by attorneys for Neptune Regional Transmission System LLC, these exports had remained immune from any assignment of upgrade costs under prior RTEP iterations. What’s the difference?
Second, the Maryland cities of Hagerstown, Thurmont, and Williamsport, (which also played key roles in the Gettysburg campaign, by the way), make an interesting and clever observation. They note that PJM’s RTEP plans in practice tend to take large, integrated grid upgrade projects and bifurcate them into smaller, manageable project segments, to be proposed piecemeal over months or years in successive RTEP proposals. The effect of this, they say, is to create a series of small projects that emerge from the modeling process as providing localized benefits only, suggesting a more narrow cost allocation. By contrast, if such projects were treated as a single integrated project, the critics say, the analysis would imply benefits covering a broader swath of zones, resulting in a wider cost allocation.
Rate Design and Beyond: Mixed Messages
In his initial decision issued July 13, 2006, administrative law judge William Cowan endorsed the commission staff’s position that PJM should move to a new design for its system-wide grid access charge, junking the license-plate pricing scheme for existing transmission (with grid costs tallied and allocated by zone), and instead adopt a postage-stamp pricing scheme, with costs and rates averaged across the entire RTO footprint. This new policy would acknowledge that the benefits from grid investment and expansion tend to benefit region-wide commerce, so that costs should be allocated likewise.
Meanwhile, however, PJM’s RTEP allocation scheme for new transmission upgrades would remain in place, leaving it open to the types of objections noted above.
In the last few weeks, however, in an interesting twist, PJM has proposed an entirely new RTEP regime. This new regime, says PJM, would dispense with attempts to distinguish upgrades driven by reliability needs, versus upgrades driven by economic ones—e.g., to relieve congestion or promote access to cheaper power supplies. This change, notes PJM, would allow it to use its RTEP process to solicit upgrades to reduce total production costs (fuel plus O&M), total generator revenue, or total system-wide zonal load payments. In other words, it would free the RTO to build a grid of the future aimed more at promoting commerce over a wide geographic area, rather than simply solving various local problems. (See FERC Docket No. ER06-1471, filed Sept. 8, 2006.)
Also, in a brief filed in the transmission rate-design case, PJM appears to accept the inevitable rejection of license-plate pricing, with a wider regional allocation of costs, such as for grid facilities of 500 kV or higher. (See, PJM Brief on Exceptions, pp. 9-10, FERC Docket No. EL05-121, Aug. 14, 2006).
And if that occurs, says PJM, FERC “should ensure than any implementation of such a change is synchronized with a concurrent change in the methodology for allocating ARRs.”
For that, however, Chambersburg may need to wait until Day 3.