Sithe petitions for review of two orders by the Federal Energy Regulatory Commission, one conditionally accepting a proposal to restructure wholesale electricity sales and transmission services in New York State, the second denying Sithe's application for rehearing. Because FERC failed to explain adequately its decision to depart from its longstanding cost-causation principle in approving a component of the proposal, we grant Sithe's petition for review in part.

Sithe describes itself as a non-utility generator that owns and operates an electric generation facility in New York. Its facility interconnects with the utility Niagara Mohawk Power Corporation's transmission system. Petitioner sells energy, on a wholesale basis, to Niagara and Consolidated Edison Company of New York, Inc. under privately negotiated power purchase agreements. Sithe challenges the Commission's approval of new tariff provisions governing charges for socalled transmission losses. In Order 888,1 FERC generally required each public utility to file tariffs for open access transmission services to remedy undue discrimination in access to their monopoly-owned transmission wires and required certain power pools, including the New York Power Pool (comprising eight utilities, including Niagara and Con Ed, referred to as the "Member Systems"), to file reformed pooling agreements by December 31, 1996. Under the Order, open access transmission tariffs were to contain at least equivalent terms and conditions for non-discriminatory service to those set out in a Commission-prescribed pro forma tariff. It also encouraged the formation of independent system operators (ISOs) to administer transmission services and new markets for wholesale electricity transactions. These ISOs were to adopt transmission (and ancillary services) pricing policies to promote the efficient use of, and investment in, generation, transmission, and consumption. In response, the Member Systems filed a proposal to replace the existing New York power pool structure with a newly created ISO: the New York Independent System Operator. According to the proposal, the NYISO would be an independent, non-profit administrator of transmission services and of the new markets for wholesale electricity transactions in New York — the accompanying tariff was intended to provide a single open access tariff over the entire New York State transmission system.

This case involves the propriety of the Member Systems' proposed transmission services pricing, specifically the pricing for transmission losses. Transmission losses refer to the amount of electric energy lost when electricity flows across a transmission system: it is a function of the square of the amount of the current flowing on the wire and of the resistance it encounters. In general, the current on a given transmission line remains a constant, and the loss associated with a single transmission of electricity is primarily a function of the distance the electricity is transmitted. Northern States Power Co. (Minn) v. FERC, 30 F.3d 177, 179-80 (D.C.Cir.1994). Utilities (who control the transmission lines) must deliver to the electricity customer the entire amount contracted for, regardless of the inevitable loss, so a transmission customer (an entity such as Sithe who pays a utility to transmit electricity across its lines) generally compensates a utility for lost energy either by providing more energy at the injection point than the electricity customer receives at the withdrawal point, or by providing energy in-kind to the transmitting utility.

The Member Systems propose tariffs that would charge transmission customers for transmission losses predicted on a "locational based marginal pricing" (LBMP) method, which would take into account costs imposed by congestion on the system. As its name implies, the methodology has both a marginal and a locational element. We gather from the parties' submissions that it predicts costs taking into account location, because the efficiency of a given electricity withdrawal or injection location may depend on the relative crowedness of a transmission route at that time. It also projects costs on the margin, meaning that it takes into account how much electricity is already on the system when an additional unit is added, because the more electricity on the system at a given time the higher the costs. For example, in the event of ten consecutive transactions each adding one unit of electricity to the system, the tenth transaction imposes a greater cost than the first additional unit. The parties do not make clear whether the LBMP method also includes traditional cost factors such as the distance load traveled. Previously, losses had been calculated using either a "rolled-in" rate or through modeling techniques that allocated charges to each transmission customer based on average (or approximate) system-wide costs imposed by the addition of load. When utilities allocated losses through a "rolled-in" average system transmission loss factor, system-wide losses were divided pro rata, and each customer paid a standard, per-unit amount. Similarly, utilities have previously allocated losses based on their transmission customers' incremental usage, using load flow models and prioritization systems to approximate the marginal losses caused by each transmission customer. The Member Systems asserted that their proposed LBMP-based tariff would be an improvement. It would send efficient price signals to market participants because it could calculate the actual marginal losses transmission customers impose on the system.

But in its tariff, the Member Systems did not propose to rely on a straight LBMP methodology; instead, they added a "simplifying assumption" that is the crux of this case. Every MWH of energy injected into the system is treated as the "last" MWH of energy on the system, and therefore this assumption would lead to the systematic overcollection of the amount of revenue needed to offset the transmission system's actual losses. The record indicates that the overcollection would be as high as 31%. The Member Systems proposed using the excess amounts to offset the NYISO's Scheduling, System Control and Dispatch Service charge (the "Scheduling Charge"), before allocating such costs among transmission customers. Generally, the total Scheduling Charge paid by a particular transmission customer is the product of the Scheduling Charge rate and the amount of power withdrawn by the customer from the system. That rate equals the NYISO's monthly overhead and other costs and expenses, minus certain credits such as overcollections for transmission losses, divided by the total number of billing units in the system. The Member Systems proposed to provide a sort of indirect refund of overcollections to those transmission customers subject to the Scheduling Charge. Unfortunately, as discussed below, FERC has never established that each entity that would be overcharged by the LBMP methodology is subject to and would benefit from a reduced Scheduling Charge.

In relevant part, FERC conditionally accepted the Member Systems' proposal; summarily rejected Sithe's objections to the proposed treatment of transmission losses; and set remaining issues for hearing. Sithe petitioned for rehearing, which the Commission denied.

Petitioner challenges as arbitrary and capricious both FERC's approval of the LBMP methodology — with its simplifying assumption — as well as the agency's endorsement of the proposed refund mechanism. FERC disputes our jurisdiction over the first issue; section 313(b) of the Federal Power Act, 16 U.S.C. § 825l(b), provides that "[n]o objection to the order of the Commission shall be considered by the court unless such an objection shall have been urged before the Commission in the application for rehearing unless there is a reasonable ground for failure to do so."

And in its rehearing request, Sithe made the following statement:

Sithe does not seek rehearing of the Commission's determination that the ISO may charge for losses using a marginal methodology. However, Sithe requests that the Commission reconsider its decision permitting the ISO to re-allocate to all loads (through the Scheduling charge) the amount of revenue that the ISO admittedly overcharged customers for transmission losses. (Emphasis added.)

Sithe, rather unpersuasively, contends that its statement below should be construed as not challenging the concept of "a" LBMP charge but as raising a challenge to this particular application of the methodology. Nice try — but no cigar. Although the record provides some indication that Sithe was unhappy with the simplifying assumption (for example, it submitted expert testimony as to different methodologies for assessing transmission loss costs), it is also clear that Sithe limited its challenge to the refund mechanism.

On the other hand, FERC reads Sithe's waiver too broadly. The Commission would have us accept for purposes of this case the LBMP methodology with the simplifying assumption as if it were graven in stone, and therefore in judging its response to Sithe's objection to the refund mechanism that methodology must be regarded as sacrosanct. But although its challenge was directed to the refund mechanism, petitioner did clearly say that "[u]ntil [the member systems] can deliver a means of assessing losses on a marginal basis that permits refunds of overcharges to the customer that overpays, the Commission should reject [the] proposal." Although Sithe ostensibly has grabbed hold only of the tariff's tail, the tariff cannot go forward so long as petitioner's legal hold on its tail is sufficient. And FERC cannot justify a perfunctory response to petitioner's refund complaint on grounds that the basic tariff methodology justifies it.

It will be recalled that the proposed refund mechanism would reduce the Scheduling Charge for transmission customers. Sithe objects on two grounds. It contends that it does not even pay a Scheduling Charge because that charge is not imposed on transmission customers, such as itself, that "schedule" wholesale bilateral transactions within New York State and do not serve end-use consumers. The Commission's counsel thought otherwise but could point to nothing in the record to support her belief.

Assuming arguendo that Sithe is correct, it would be part of a class of transmission customers who have been overcharged for transmission losses. The Federal Power Act provides that a utility may not charge rates that "make or grant any undue preference or advantage to any person or subject any person to any undue prejudice or disadvantage." 16 U.S.C. § 824d(b); see also Electricity Consumers Resource Council v. FERC ("ECRC"), 747 F.2d 1511 (D.C.Cir.1984) (remanding a marginal rate scheme that resulted in cross-subsidization of certain customers by other customers). Similarly, under section 205(a) of the Federal Power Act, a utility may charge only rates that are "just and reasonable." 16 U.S.C. § 824d(a).2 Interpreting that mandate, we have explained that such rates "should be based on the costs of providing service to the utility's customers, plus a just and fair return on equity." Alabama Elec. Coop. v. FERC, 684 F.2d 20, 27 (D.C.Cir.1982). We have consistently upheld rates based on such a cost-causation principle, see, e.g., id.; K N Energy, Inc. v. FERC, 968 F.2d 1295, 1300 (D.C.Cir.1992).

Even if Sithe gains some benefit from the refund mechanism, it also claims that it would be entitled to a refund that is equivalent to the amount it has been overcharged. FERC's response in its order on rehearing was merely that the tariffs and refund mechanism produced "efficient price signals," and that petitioner's requested refunds would somehow disrupt that price signaling, would be "infeasible," and a matter of "unending controversy." To be sure, we have acknowledged that feasibility concerns play a role in approving rates, indicating that FERC is not bound to reject any rate mechanism that tracks the cost-causation principle less than perfectly, see, e.g., Tejas Power Corp. v. FERC, 908 F.2d 998, 1005 (D.C.Cir.1990). But the Commission's cursory response simply will not do. At no point did the Commission explain how these considerations applied. Why, we wonder, would a different method of refunds, based more closely on cost-causation principles, jeopardize desirable price signaling or be infeasible?

It may well be that the Commission's refusal to consider petitioner's refund proposal is really attributed to a desire to protect the simplifying assumption. It could be thought that a more precise refund mechanism — matching the refunds to overpayments — would render the simplifying assumption useless. But we do not see how the Commission can justify its refusal to insist on equitable refunds, based on its approval of a presumably discriminatory tariff, just because petitioner challenged squarely only the refund mechanism. If FERC's position were to be that the refund mechanism is inextricably intertwined with the simplifying assumption, then petitioner's challenge to the refund mechanism would perforce question the simplifying assumption.

* * * *

Accordingly, the petition for review is granted in part and we remand the case to the Commission to more adequately respond to petitioner's contentions.

So ordered.