Every Last Penny


Transmission cost allocation, the worth of the grid, and the limits of ratemaking.

Fortnightly Magazine - August 2010

A year ago this month, when Judge Richard Posner and his 7th Circuit court of appeals upheld a challenge from state utility regulators in Illinois and Ohio, and tabled FERC’s plan for PJM to spread the $6.6-billion cost of some 18 new extra-high-voltage (EHV) transmission projects across the entire RTO footprint, from the Jersey shore to Chicago’s suburbs, the judge exposed what now has become the single most important policy rift in the electric utility sector: How to decide who benefits and who should pay for the cost of new transmission lines designed to serve customers situated across a broad swath of multiple utility service territories.

But while Judge Posner sent this most important question back to the commission for a rethink, he took pains to explain that FERC need not recalculate costs “to the last penny,” as traditional ratemaking would imply.

In fact, the judge left room for FERC to reinstate the original scheme, provided that after collecting more evidence, the commission could articulate a plausible case to show that with region-wide cost sharing, known as “socialization” or “postage-stamp” allocation, Illinois and Ohio ratepayers would gain benefits from grid expansion roughly commensurate with the higher costs they’d be paying, and so would not fare appreciably worse than ratepayers residing east of the Alleghenies where all of the projects would be built.

But with the evidence now in, FERC’s policy is looking a bit tenuous. Opponents would tie cost responsibility to grid power flows. They suggest that socialization would shift over $2 billion in costs unfairly from eastern to western reaches of the PJM region. By contrast, FERC supporters believe that grid expansion creates certain region-wide intangible benefits as a counterbalance, but courts don’t much like taking arguments on faith.

FERC’s original ruling had come not from evidentiary findings, but simply by ratifying a settlement agreement. (See Opinion 494, April 19, 2007, 130 FERC ¶61,233.) Yet, it reflected the idea that EHV transmission lines (e.g., 500 kilovolts and above) yield benefits across wide regions, such as greater carrying capacities, reduced line losses, and better access to distant and cheaper generating plants, making power markets more vibrant. Also, a joint United States-Canada task force found that 500- and 765-kV lines had stopped the spread of cascading faults and outages during the 2003 August blackout, showing that EHV lines protect consumers from economic losses stemming from region-wide outages.

Nevertheless, the case on remand raises certain red flags.

First, of the 18 new EHV projects in question in the 7th Circuit appeal, it now appears that the very largest were approved by PJM during the years 2007-09 through its RTEP regional transmission planning process primarily to offset future gen plant retirements in New Jersey, and lessen the risk of reliability violations in the next few years in the Baltimore-Washington corridor. (Fig. 1 shows the six biggest grid projects at issue in the remand case, representing a combined cost of some $6.2 billion.)

And second, as for those regional benefits, they seem tangible enough, but appear aimed in the wrong direction. It so happens that PJM’s operative transmission line planning documents showed that certain key projects (e.g., the PATH and TrAIL lines, in particular) could facilitate power exports from west to east, and so predicted significant drops in wholesale power prices in areas served by PEPCO, BG&E, and Dominion, on the order of $5-$10 per MWh in terms of locational marginal prices, with complementary price hikes for western PJM companies utilities such as AEP, Duquesne Light, Commonwealth Edison, and Dayton Power & Light. The intangibles, if anything, seem to militate against an equal regional allocation of grid-expansion costs.

ELCON, the Electricity Consumers Resource Council, argued that FERC must base its cost allocation on precise ratemaking formulas to ensure efficient planning:

“In [our] experience,” ‘unquantifiable’ costs or benefits … are all too often a fig leaf justification for a project that does not otherwise make economic sense.”

ELCON and others say that socialization masks price signals; if gen developers in remote areas need not factor in the added cost of transmission, infrastructure gets built in the wrong place.

And the court itself had laid down the law against regional favoritism:

“FERC is not authorized,” wrote Judge Posner, “to approve a pricing scheme that requires a group of utilities to pay for facilities from which its members derive no benefits, or benefits that are trivial.” (Ill. C.C. v. FERC, 7th Cir., Aug. 6, 2009, 576 F.3d 470.)

But now the story grows more complicated.

In June, FERC OK’d a highway/byway plan for Southwest Power Pool, with postage-stamp cost allocation for grid facilities operating at 300 kV or above, showing that it still likes socialized cost recovery for new grid lines. (See, Docket ER10-1069, June 17, 2010, 131 FERC ¶61,252.)

And more important, on July 15, as this column went to press, the Midwest ISO proposed region-wide cost sharing for a new category of grid upgrades, to be known as MVPs, or multi-value projects. These new projects qualifying for regional socialization would be those that address reliability and economic issues affecting multiple transmission zones, along the lines of a high-voltage regional grid overlay, or regional grid superhighway. As MISO explains, MVPs would help support “documented energy policy mandates,” such as renewable portfolio standards, or other state or federal mandates governing resource planning and overall energy policy. (See, FERC Docket ER10-1791, proposed tariff filed July 15, 2010.)

This new initiative from MISO bears watching.

Second, by proposing this policy, MISO stakeholders certainly must feel they have some sort of chance to win FERC approval. But by proposing to link grid-cost allocation to the furtherance of national energy policy, MISO presumes to act both as judge and jury on what projects, exactly, will serve the national interest.

But Dayton Power argues that if socialized costs are valid as a policy exception to promote wind energy development, then “such exception should be a policy determination made by Congress and not through a regulatory process.”

And so while FERC still appears to favor transmission cost allocation with the widest possible geographic reach, the 7th Circuit case, involving 18 specific high-voltage grid projects planned by PJM, likely offers the worst possible test case for seeking court approval. That means FERC must somehow appease Judge Posner to save its skin, yet preserve enough wiggle room to OK MISO’s proposal at a later date.

FERC could choose to give ground in the PJM case, adopting perhaps partial use of a more prosaic bean-counting method, as the Pennsylvania Consumer Advocate in fact recommends in the remand case. Yet even a strict flow-based allocation is not free of baggage, as this column will attest.

Dayton Power sums it all up: “FERC,” it states, “has an exceptionally difficult task to perform in this proceeding.”

Taking a Snapshot

The notion that grid-cost socialization in PJM could shift over $2 billion in costs from eastern to western RTO members stems from studies that PJM conducted earlier this year, on instructions from FERC, to illustrate hypothetically how a postage-stamp allocation for EHV transmission projects would differ if PJM instead would employ its “beneficiary pays” allocation method (e.g., the DFAX model), which otherwise applies only to grid projects below 500 kV.

The PJM study showed that while the choice of method would make little cost difference for some utilities (e.g., Dominion and Allegheny, for example), others would come out either winners or losers under socialization, in some cases by ratios greater than 100 to 1. (See, Response of PJM, p. 17, Docket EL05-121, filed Apr. 13, 2010.)

Eyeing such numbers, Exelon’s v.p for federal regulatory affairs, Karen Hill, noted that if FERC recanted on socialization, switched instead to a strict “beneficiary pays” principle based on the DFAX model, then “western zones of PJM would be allocated approximately $2.4 billion less and eastern zones approximately $2.4 billion more for RTEP projects 500 kV and above at issue in this proceeding.” (Exelon Comments, Doket EL05-121, p. 3.)

Nevertheless, the DFAX model as applied to high-voltage lines remains highly contentious.

Simply put, DFAX identifies what loads in which local utility transmission zones are flowing over a grid facility that is constrained beyond what reliability allows. Or, as described more fully by PJM, DFAX is a computer model used in conjunction with the RTO’s annual regional transmission plan (RTEP), to calculate individual relative percentage distribution factors that “express the portions of a transfer of energy from a defined source to a defined sink that will flow across a particular transmission facility or group of facilities.”

As explained by Steven Naumann, Exelon’s v.p. of wholesale market development, DFAX “identifies the loads that cause the costs of the reliability upgrades to be incurred. Since PJM must operate in a manner that is not in violation of NERC [standards], these loads would have to be curtailed if the reliability upgrades were not installed. These loads are by definition the beneficiaries of the reliability upgrades. (Exelon comments, Affidavit at p. 8.)

However, others point out, such as FERC supporters BG&E, PEPCO, Atlantic Electric, Delmarva Power, and the ad hoc Fair Pricing Group (PP&L, PSE&G, Rockland Electric), PJM designed the DFAX method initially to allocate costs only for small, local, low-voltage grid upgrades designed to fix a single, immediate, and easily identifiable reliability or deliverability violation, and so to provide a “snapshot” of line flows at the precise time the RTO includes the local upgrade in its RTEP. That pinpoints cost causation.

By contrast, a major 500-kV line might well address 20, 30 or more violations—some not expected for as long as five to 15 years. But meanwhile, each year that PJM revisits and updates its RTEP, it gets a different picture of likely future conditions. A reliability violation and upgrade “need” cited in this year’s RTEP could well vanish the next year, when the RTEP is retooled, reflecting the latest figures for economic activity, load growth, line loadings, and gen plant development.

The $1.8 billion PATH line (Potomac-Appalachian Transmission Highline), planned jointly by Allegheny Power and AEP, and first approved by PJM in RTEP 2007, as needed in 2012 to ensure deliverability to growing load in Northern Virginia, was deferred year-to-year by successive RTEP studies as the economy deteriorated. On January 27, the PATH sponsors withdrew their certificate application before the Virginia State Corporation Commission, as the line no longer appeared to be needed by 2014, but AEP and Allegheny filed updated testimony before the West Virginia commission just last month, reflecting PJM findings in RTEP 2010 that the line must be built by June 15, 2015. (See, www.pathtransmission.com.)

And in its April report to FERC comparing the DFAX and postage-stamp methods, PJM showed how a DFAX allocation would vary from year-to-year for the 500-kV Susquehanna-Roseland line, planned for northern Pennsylvania and New Jersey, as assumptions changed.

For example, PSE&G, JCP&L and Atlantic Electric, taken together, would pay for 78 percent of the $1.16 billion Susquehanna-Roseland line under RTEP 2007, but 90 percent under RTEP 2009. PECO’s share would fall from $119 million (2007 retool) to $33 million (2008) to nothing at all (2009). On the other hand, the PSE&G zonal allocation goes from $473 million (2007) to $710 million (2009).

Timing is everything.

What if wind projects are developed offshore in the Mid-Atlantic, altering or even reversing line flows in the future?

At the risk of being labeled disingenuous, BG&E argues that it could become a net loser in the future under cost socialization if wind project development out West spawns a raft of new 500- and 765-kV lines built to deliver wind power to Chicago or other Midwest load centers, producing a west-to-east cost shift—opposite to what Exelon’s Naumann complains of today. A short-sighted reliance on DFAX, notes BG&E, could eventually, “shield the BG&E rate zone from being allocated those costs.”

Yet even Judge Posner had discounted this possibility, noting in his opinion that, “so far as it appears, few if any of such facilities will be built in … the Midwest within the foreseeable future.”

Exelon v.p Naumann also points out that PJM’s 2009 RTEP report, while finding some overloads on the ComEd system in 2024, did not see any need at that time for 765-kV lines on the ComEd system.

Instead, as Naumann testifies, any new wind-driven grid connections into the ComEd zone likely will come in at the 345-kV level:

Several large wind farms, specifically the 300-MW Top Crop Wind Farm, the 400-MW Twin Groves Wind Farm and the 300-MW Cayuga Ridge South Wind Farm are already in service, or will be in service by the end of 2010 and connected to the ComEd 345-kV system, and any reinforcements needed to allow continued operations are likely to be at 345 kV.” (See, Comments, Naumann affidavit at pp. 54-55.)

Naumann also cites the 345-kV Central Transmission project, proposed recently as the first-ever merchant transmission line to be approved as part of PJM’s regional transmission plan. (See FERC Docket EL10-52, filed March 25, 2010.)

Nevertheless, the record appears clear that DFAX won’t work very well for EHV lines without continuous revising and updating to reflect the changing grid topology—a task that PJM claims will prove impossible in practice for lines 500 kV and above:

“Performing recurring DFAX allocations over a period of years would be virtually impossible as this would require unwinding the transmission grid, line by line, to determine whether the impacts driving the need for a previously approved project had changed. (Response of PJM, p. 27.)

ELCON appears untroubled about calculating precise line flows and allocating costs accordingly: “It is not only possible to determine who benefits from transmission projects; it is done all the time.”

But shortcomings over the DFAX model leads PEPCO to state that Judge Posner and the 7th Circuit should not waste effort comparing DFAX results to a socialized allocation:

“The issue,” write PEPCO, “has to an extent been framed in a manner that appears to treat DVAX-based cost allocation as a kind of ‘default’ approach from which socialization is purportedly a departure. However, the relevant issue in this proceeding is not how socialization compares …

“The relevant issue is whether the broad regional benefits of 500 kV-and-above facilities support socialization.

“The commission’s order accepting the settlement, moreover, made clear that it made no finding on the merits that … DFAX … was just and reasonable as to any voltage level.”

Having It Both Ways

In one sense, this fight never had to happen.

RTO membership remains voluntary, so, as PEPCO has pointed out, “had Commonwealth Edison elected to remain in the Midwest ISO instead of joining PJM, there would be no issue today about the allocation of PJM’s 500 kV-and-above facilities in Illinois—and thus no Seventh Circuit remand and no paper hearing.”

And as FERC’s postage-stamp method would allocate the expense of grid upgrades across all utility service areas in the RTO by load-ratio shares, ratepayers would be treated substantially the same across the region. Only the utilities, as corporate entities, would show significant disparity in rate treatments.

BG&E writes that as RTOs mature, “the boundaries between the individual franchise service territory of each transmission owner has less meaning with respect to high-voltage 765-kV and 500-kV facilities.

“The actual source of energy delivered to one zone within PJM from another … cannot even be traced, and is even independent of contractual arrangements.”

The record in the case contains studies that purport to show the likely effect on monthly retail utility bills for typical residential ratepayers within the PJM region.

Testifying for the Fair Pricing Group, the private consultant Richard Wodyka estimates that a socialized allocation for all grid project costs at issue in the remand case would boost retail bills for all residential customers across the entire PJM footprint by an average of 15 cents per month—as low as 12 cents for Penelec; 13 cents for Allegheny, 14 cents for AEP, BG&E, Duquesne and others; 15 cents for Dayton, Dominion Virginia Power, PECO, and PEPCO; 16 cents for Delmarva and ComEd; and up to 18 cents for Atlantic Electric, Jersey Central, and Rockland Electric.

Wodyka’s study assumes an annual carrying charge of 19.1 percent for all allocated grid costs, plus typical monthly residential consumption of 1,000 kWhs. (See PEPCO comments, Wodyka affidavit, pp. 30-36.)

Another witness however, calculates rate impacts higher by a full order of magnitude. Perhaps someone has misplaced a decimal point.

Thus, David J. Scarpignato, director of RTO regulatory affairs for Old Dominion Electric Cooperative, but testifying for PEPCO, shows residential rates climbing 1.7 mills per kWh for ComEd, or about $1.30/month on average for Illinois monthly residential consumption of 765 kWhs. For Dayton, Scarpignato sees socialized costs adding 1.6 mills per kWh, or about $1.46/month on average (Ohio) consumption of 910 kWhs. (See PEPCO comments, Scarpignato affidavit, attach. A.)

Nevertheless, ELCON’s argument about masked price signals still appears troubling for adherents of socialization.

As ELCON points out, “attempts to socialize costs across FERC-created organized markets are ironic because the locational (nodal) pricing regime was intended to facilitate locational—not regional—solutions to reliability, congestion, and resource adequacy. The commission cannot have it both ways.”

Even PJM recognizes this potential inconsistency in its recent landmark (and copyrighted) study on the issue:

“Transmission cost allocation methods,” writes PJM, “should at least be neutral and not run counter to incentives provided in the energy and capacity market designs.” (See, A Survey of Transmission Cost Allocation Issues, Methods and Practices, PJM, March 10, 2010, p. 20.)

It is notable that First Energy applied to move from MISO to PJM only eleven days after the 7th Circuit announced its ruling. And just five weeks ago, Duke Energy sought the same move for its Ohio and Kentucky utility subsidiaries. (See FERC Docket EL10-1562, filed June 25, 2010.)

As pointed out by consultant Michael Schnitzer, director of the NorthBridge Group, and testifying for Dayton Power, these recent decisions point out the importance of grid-cost allocation.

“Prospective new entrants,” he states, “are obviously watching this proceeding carefully to make sure that FERC does the right thing.”