Office of the People's Counsel v. Public Service Commission of the District of Columbia

482 A.2d 404 (1984)

OFFICE OF the PEOPLE'S COUNSEL, Petitioner,
v.
PUBLIC SERVICE COMMISSION OF the DISTRICT OF COLUMBIA, Respondent,
Washington Gas Light Company, Intervenor.

No. 83-650.

District of Columbia Court of Appeals.

Argued December 14, 1983. Decided October 10, 1984.

Brian J.H. Lederer, Washington, D.C., for petitioner.

Lloyd N. Moore, Jr., Washington, D.C., with whom Roberta Willis Sims, Washington, D.C., was on brief, for respondent.

Lewis Carroll, Washington, D.C., for intervenor.

Before FERREN and TERRY, Associate Judges, and GALLAGHER, Associate Judge, Retired.

FERREN, Associate Judge:

Petitioner, Office of People's Counsel (OPC), challenges a decision of the Public Service Commission (PSC or Commission) in Formal Case No. 787, which grants a rate increase to Washington Gas Light Co. (WGL) for retail service in the District of Columbia. OPC takes issue with the following six aspects of the Commission's decision: (1) the Commission granted WGL an attrition allowance in the form of a forward-looking rate base; (2) it adopted for the first time a flexible rate structure for sales to interruptible customers; (3) it deviated from the Commission's "preferred policy" in calculating WGL's allowance for uncollectible accounts; (4) PSC did not include the tax effects of deductible interest on WGL's short-term debt in calculating WGL's income tax allowance; (5) it approved WGL's estimated construction budget for the years 1982 to 1986; and (6) it adopted a methodology for calculating WGL's marginal commodity cost to be used in future ratemaking proceedings. We conclude that, with respect to all issues, the Commission has adequately explained the reasons for its decision and that its decision is not unreasonable, arbitrary, or capricious. Accordingly, we affirm.

PROCEDURAL HISTORY

WGL filed with PSC on April 14, 1982, an application for an increase in retail service rates for District of Columbia customers. WGL proposed an annual rate increase in the aggregate amount of $26.3 million. The Commission issued a notice of proposed rulemaking on April 30, 1982, and scheduled consideration of WGL's application as Formal Case No. 787.

Several organizations filed comments responding to WGL's application, including the General Services Administration (GSA), the Apartment and Office Building Association of Metropolitan Washington, Inc. (AOBA), and the Consumer Utility Board (CUB). PSC granted each of these organizations intervenor status. OPC also filed comments concerning WGL's application and represented the interests of utility consumers pursuant to D.C.Code § 43-406(d)(1) (1981). The parties conducted discovery, narrowed the issues, and filed testimony and prehearing briefs. The Commission then held fifteen days of evidentiary hearings on WGL's application between October 12 and November 4, 1982, during which each party had an opportunity to present evidence.

On February 25, 1983, the Commission issued Order No. 7749 granting WGL an annual rate increase in the aggregate amount of $14.3 million. In addition to setting an overall rate increase, this 127-page decision addressed numerous issues raised by the parties concerning the rate structure methodologies and accounting practices that govern the manner in which WGL may recover its authorized revenue. WGL, OPC, and AOBA each filed an application for reconsideration of Order No. 7749. By Order No. 7817, issued April 27, 1983, PSC denied these applications for reconsideration and provided additional reasoning and support for its decision. OPC now petitions for review of both Orders.

SCOPE OF REVIEW

The scope of this court's review of PSC decisions "is the narrowest judicial review in the field of administrative law." Potomac Electric Power Co. v. Public Service Commission, 402 A.2d 14, 17 (D.C.) (en banc), cert. denied, 444 U.S. 926, 100 S. Ct. 265, 62 L. Ed. 2d 182 (1979); accord Washington Gas Light Co. v. Public Service Commission, 450 A.2d 1187, 1193 (D.C. 1982). Our review of Commission orders is restricted by statute "to questions of law, including constitutional questions; and the findings of fact by the Commission shall be conclusive unless it shall appear that such findings of the Commission are unreasonable, arbitrary, or capricious." D.C.Code § 43-906 (1981). "We are not to substitute our judgment for that of the Commission. Even though we might arrive at a some-what different decision than did the PSC, if there is substantial evidence to support the Commission's findings and conclusions, we must affirm." Potomac Electric Power Co., 402 A.2d at 17.

"`It is especially important to accord great respect to the Commission in a complex, esoteric area such as rate making in which the Commission has been entrusted with the difficult task of deciding among many competing arguments and policies.'" People's Counsel v. Public Service Commission, 455 A.2d 391, 393 (D.C.1982) (quoting Goodman v. Public Service Commission, 162 U.S.App.D.C. 74, 78-79, 497 F.2d 661, 665-66 (1974) (footnote and citations omitted)). "It is the Commission, not this court, that must balance the competing interests of utility consumers and investors in the ratemaking process." Id. Indeed, because "[t]heories of ratemaking in particular fall within the special province of the Commission," they are "`not subject to the same substantiation principle as the substantial evidence test applicable to fact-finding.'" Washington Gas Light Co., 450 A.2d at 1193 (citation omitted). Although "`[a] theory of ratemaking must be reasonable, explained, and supported,'" id. (quoting Continental Air Lines, Inc. v. Civil Aeronautics Board, 179 U.S.App. D.C. 334, 342, 551 F.2d 1293, 1301 (1977)), we may not "`reassess the weights given by a rate-making agency to different factors, absent a legislative direction as to precisely what gravity each factor bears.'" Id. (quoting Association of American Publishers, Inc. v. Governors of the United States Postal Service, 157 U.S.App.D.C. 397, 403-04, 485 F.2d 768, 774-75 (1973)).

Accordingly, in addressing the issues of ratemaking theory and methodology raised by OPC here, our task is limited to determining whether the Commission took into account "`all the relevant factors and no others,'" id. (quoting Association of American Publishers) as well as ensuring that the methodology is sufficiently explained to reveal its effect on the reasonableness of the overall decision. Id. at 1193-94; accord Washington Public Interest Organization v. Public Service Commission, 393 A.2d 71, 76-77 (D.C.1978), cert. denied, 444 U.S. 926, 100 S. Ct. 265, 62 L. Ed. 2d 182 (1979). OPC thus bears a "heavy burden of demonstrating clearly and convincingly a fatal flaw in the action taken," Washington Gas Light Co., 450 A.2d at 1194, "a burden not met by merely setting forth an acceptable alternative to Commission action." People's Counsel, 455 A.2d at 394. As long as the Commission "has fully and clearly explained" what it has done and why it did it, and "the agency decision is supported by substantial evidence," we will, "upon a finding that the Commission order is reasonable in its overall effect," affirm the Commission's rate-making decision. Washington Gas Light Co., 450 A.2d at 1194.

THE MERITS

I.

OPC first claims that the Commission erred in adjusting WGL's rate base to compensate for the effects of "attrition." Attrition "is the tendency of a utility's rate of return to diminish over time as a result of high construction costs, an expanding rate base, and increasing operating costs." Washington Gas Light Co., 450 A.2d at 1219 n. 42. The high cost of bringing a new plant into service "`tends to increase the applicable rate base [investment] at a more rapid pace than the resultant earnings [net operating revenues], and the rate of return decreases accordingly.'" Potomac Electric Power Co., 402 A.2d at 23 (emphasis in original) (quoting Telephone Users Association v. Public Service Commission, 304 A.2d 293, 298 (D.C.1973), cert. denied, 415 U.S. 933, 934, 94 S. Ct. 1448, 1449, 39 L. Ed. 2d 492 (1974)).

When the Commission finds that attrition is undermining a utility company's ability to earn its authorized rate of return, PSC may consider at least two possible types of adjustments. First, the Commission may simply add an attrition allowance onto the utility company's authorized rate of return, thereby enabling the company to make a reasonable return on investment even if it falls short of earning the full, attrition-adjusted authorized rate. This court recently has held that this type of attrition allowance should be granted only under "extraordinary conditions." Washington Gas Light Co., 450 A.2d at 1221. Second, the Commission may use a forward-looking rate base—i.e., a rate base which is not matched to the test period used for balancing revenues and expenses and which reflects the growing level of utility investment—instead of following the normal procedure of calculating rates based on the average test period rate base. This type of adjustment has been consistently approved by this court as a means of offsetting earnings erosion resulting from attrition. People's Counsel, 455 A.2d at 394-95; Washington Gas Light Co., 450 A.2d at 1219-22; Potomac Electric Power Co., 402 A.2d at 23; Telephone Users Association, 304 A.2d at 298.

In the present case, WGL asked the Commission to grant two attrition allowances: (1) a specific adjustment in the rate of return to reflect rising gas storage and labor costs in 1983; and (2) a forward-looking rate base (calculated for year-end 1982 with adjustments for nonrevenue-producing distribution plant replacements through mid-1983) totalling $158.2 million. OPC, on the other hand, claimed that WGL had failed to demonstrate the likelihood of future attrition and, accordingly, that it would be inappropriate to grant any attrition allowance. OPC urged the Commission to adopt an average rate base for the calendar year 1981, the test period adopted by the Commission for the purpose of balancing revenues and expenses. OPC claimed that this average test period rate base totalled $136.1 million.

After reviewing the evidence presented by each party, the Commission found that WGL's evidence of attrition was persuasive. It refused, however, to grant either of the two attrition allowances that WGL requested. Instead, the Commission agreed to use a forward-looking rate base in the form of a year-end 1981 rate base, adjusted for known and measurable changes. This rate base equaled $145.5 million. OPC raises two objections to this ruling.

A.

OPC first challenges the Commission's finding that WGL has suffered and is continuing to suffer from attrition. OPC argues that WGL failed to meet its burden of proving the existence of attrition and that the Commission misapplied its own established criteria for evaluating evidence of attrition.

The Commission began its analysis of WGL's request for an attrition allowance by recognizing that its past orders had established a framework for evaluating such requests:

In the past, this Commission has allowed an end-of-year rate base to be used only where evidence of attrition, as compared to evidence of mere inflation, was demonstrated.... We have stated our opinion that the factors to be considered in determining whether attrition exists include "trends in [1] rate of return on investment, [2] operating ratio, and [3] operating revenue per $100 of intrastate net investment." Chesapeake & Potomac Telephone Co., 4 PUR 4th 1, 42-43 (D.C. 1974).

The Commission also noted that it has never considered these three types of trend evidence to be the exclusive measures of attrition, and that additional evidence of attrition will be considered when it is helpful in providing "as complete a financial picture of a utility as possible."

The Commission's opinion next evaluated WGL's evidence on each of the three established factors for showing attrition. As to the first factor, WGL presented exhibits showing that it had failed to earn its authorized rate of return in any year between 1972 and 1981; that for three of these years (1975, 1978 and 1979), its rate of return on equity was negative; and that during this ten-year period its overall equity earnings were deficient by more than $30 million. The Commission noted that "WGL's A-Rated Bond has been downgraded twice in 1982 by bond rating agencies who cite `consistently low interest coverage' as the reason for this downgrading." WGL also presented exhibits which addressed the second and third trend factors for the years 1972-1981. Although WGL conceded that the evidence concerning these two factors did not demonstrate the existence of attrition as clearly as did the evidence on the first factor, it argued that the second and third factors were nonetheless consistent with the existence of attrition.

Finally, WGL presented evidence of a fourth trend which it claimed was helpful in assessing the effects of attrition: the trend in "average net investment versus growth in net operating income." WGL argued to the Commission that evidence on this fourth trend revealed that, over the 1972-1981 period, the District of Columbia rate base grew faster than the revenues available to WGL after it had paid its operating expenses. The amount of investment required to produce a dollar of net utility operating income in 1981 was nearly double the amount required to produce a similar income in 1972.

OPC presented the Commission with evidence which, it claimed, showed that attrition was no longer a serious problem for WGL. In particular, OPC pointed out that in 1981 WGL earned 98.75 percent of its authorized rate of return and that WGL's operating ratio in that year was lower than in any of the previous eight years. Although OPC recognized that WGL had not earned its authorized return during the previous eight years, it attributed this short-fall to problems other than attrition^[1] and argued that data from the most recent years showed a trend away from such deficiencies in earnings.

After reviewing this evidence, the Commission ruled: "Based on the record, we find that WGL, utilizing the three measures of attrition set forth in [earlier PSC decisions], has shown some evidence of earnings attrition for which an attrition adjustment should be made." In its brief to this court, OPC attacks the Commission's attrition analysis on several grounds.

First, OPC claims that the Commission erroneously "ignor[ed] the trend indicated by the 1981 data." This contention is plainly wrong. The Commission's decision expressly addressed OPC's argument that the 1981 data reflect a positive trend:

We have carefully reviewed the arguments and the record evidence on attrition of all of the parties. While we recognize that 1981 data showed WGL earning 98.75 percent of its authorized overall rate of return, we also recognize that WGL's 1981 data was influenced by a number of factors (e.g., interest rates, income taxes and the Crab Run gas strike) and can not be relied upon in isolation. Indeed, we have always required the use of trend data for the purpose of determining an attrition allowance precisely to avoid the situation of being influenced by a single event or a limited set of circumstances....

When WGL's performance is reviewed over several years, we believe that the picture is different. We note that WGL has consistently earned a low percentage of its authorized rate of return (footnotes omitted).

Thus, the Commission did not ignore OPC's position that the 1981 data reflected a trend away from attrition; it expressly rejected it. Viewing the evidence in the record as a whole, we cannot say that this finding was "unreasonable, arbitrary, or capricious." D.C.Code § 43-906 (1981). Although OPC is correct in pointing out that a "trend" cannot be established merely by aggregating all of the data for a preceding ten-year period, without some attention being given to patterns evidenced by data in the most recent years, this is not to say that data for the most recent year always reflect an accurate projection of a future trend. PSC adequately explained its reasons for viewing the 1981 data as a deviation from an established trend of attrition, rather than as a signal that attrition was no longer a problem.

OPC next argues that WGL's evidence on the second and third trend factors did not persuasively demonstrate the existence of attrition. We note that the Commission expressly stated in its decision that it weighed the evidence presented by all parties on both the second and third factors in making its finding on the question of attrition. More importantly, we note that neither this court nor the Commission has ever required a utility to demonstrate that all three attrition factors, or even two of the three factors, support a finding of attrition. We conclude that a finding of attrition is appropriate as long as it is supported by the record evidence presented on all three factors, taken as a whole, even if the evidence presented on any one or two of the factors is inconclusive. A review of the evidence in the record here convinces us that the Commission's attrition finding meets this requirement.

Finally, OPC challenges the attrition finding on the ground that "the Commission nowhere articulates any basis for adopting [a] new fourth attrition standard." We do not, however, read the Commission's decision to establish a new attrition factor—"average net investment versus growth in net operating income"— which will hereafter be considered or relied on in every case involving a request for an attrition allowance. Instead, we understand the Commission to have accepted evidence of a fourth trend because it was persuaded that the particular evidence presented by WGL would be a useful supplement to the three established trend factors in the context of this proceeding. The Commission's earlier decisions had indicated that evidence of attrition other than the three established attrition factors would be considered on an ad hoc basis and relied on when it provided the Commission with a more complete picture of the attrition problem.

OPC does not dispute that the fourth type of trend evidence relied on by the Commission was probative of the attrition issue here. Indeed, the relationship between "average net investment" and "net operating income" closely parallels the relationship between a utility's rate base and its earned rate of return, which lies at the heart of the attrition phenomenon. Given these facts, we do not find the Commission's reliance on evidence of a fourth trend to render the attrition finding unreasonable, arbitrary, or capricious.

B.

OPC also argues that, even if the Commission's attrition finding was appropriate, the Commission erred in granting WGL an attrition allowance. OPC claims that an adjustment compensating WGL for attrition was built into the overall rate of return authorized by the Commission as a result of the Commission's use of Discount Cash Flow (DCF) analysis in calculating the cost of equity capital. It is OPC's position that, because attrition was accounted for in the Commission's DCF analysis, the use of a forward-looking rate base as an adjustment for attrition results in a "double recovery" to WGL for anticipated losses because of attrition. We find no basis for concluding that the DCF analysis used here made any allowance for attrition and further conclude that OPC's reliance on an earlier opinion of this court, Washington Gas Light Co., 450 A.2d at 1219-22, is misplaced.

DCF analysis is used to determine the rate of return which a utility must provide to its equity shareholders in order to attract new capital investment. The DCF method of calculating the cost of equity can be stated in a "simple formula": market capitalization rate (costs of equity) equals the current dividend yield (risk component) plus anticipated growth (growth component). Washington Gas Light Co., 450 A.2d at 1210. Thus, DCF analysis "takes into account an investor's anticipated income from dividends and capital gain upon eventual sale of the stock in order to arrive at an estimated rate of return which the investor must receive in order to induce him to invest in the utility stock." Washington Gas Light Co. v. Public Service Commission (Washington Gas Light Co. II), 452 A.2d 375, 382 (D.C.1982), cert. denied, ___ U.S. ___, 103 S. Ct. 2454, 77 L. Ed. 2d 1334 (1983).

In practice, the risk and growth components for a DCF analysis are arrived at by undertaking a comparative study of bonds and stocks issued by a number of utility companies. See, e.g., Washington Gas Light Co., 450 A.2d at 1210-11. The companies chosen for comparison are selected because they have certain characteristics in common with the company seeking a rate adjustment. In the proceedings at issue here, WGL, OPC, GSA, and PSC staff each submitted a cost of equity estimate based on independent DCF analyses. Each of the parties selected different utility companies for comparative analysis, with the proposals (combined risk and growth) ranging from OPC's recommended 14.50 percent to WGL's recommended 18.00 percent. After reviewing these analyses, the Commission determined that WGL's cost of equity was 15.7 percent.

OPC contends that the type of market-oriented DCF analysis relied on by the Commission has a "built in" allowance for attrition, if attrition indeed exists. In its brief, OPC maintains that "if such a risk of earnings erosion is real, then it will necessarily be perceived by investors, who will thereby require that they be compensated for such risk through their required equity return." OPC contends that the Commission's use of a forward-looking rate base to offset attrition was unreasonable because the Commission failed to take this built-in attrition allowance into account. In its Order denying reconsideration of its rate decision, the Commission found this argument "unconvincing." We agree.

A review of the DCF analysis presented by the parties here reveals that, in selecting utility companies for comparative study, none of the parties specifically considered whether the companies chosen were experiencing attrition in a manner similar to WGL. Nor did the Commission, in arriving at its cost of equity figure, indicate that it was considering WGL's attrition or the likelihood that the companies considered in the DCF analyses were experiencing earnings erosion. Thus, we are not persuaded by OPC's argument that the DCF analysis relied on by the Commission automatically compensated WGL for attrition. Given the DCF presentations made by all of the parties, there is simply no way of knowing the extent to which investors' perceptions of the risk of attrition affected the cost of equity calculations.

Moreover, we reject OPC's contention that the Commission acted unreasonably in failing at least to make an express finding as to whether its DCF analysis accounted for attrition before ruling on the appropriateness of a forward-looking rate base. In this context, OPC's reliance on this court's recent opinion in Washington Gas Light Co., 450 A.2d at 1222 n. 45, is misplaced. A careful reading of that case reveals that this court approved the grant of a forward-looking rate base as an adjustment for attrition, even though the Commission determined the utility's cost of equity by using DCF analysis and never made any finding as to whether its DCF analysis embodied a built-in adjustment for attrition. Id. at 1215, 1221-22. Thus, far from supporting OPC's position, Washington Gas Light Co. actually approves of the very practice challenged here.

II.

Most of WGL's customers are "firm" customers, i.e., their facilities are designed to burn gas and they lack the capacity to switch to an alternative energy source on a short-term basis. A small portion of WGL's customers, however, are "interruptible" customers who can use either natural gas or some other fuel—generally oil— to fulfill their energy needs. These customers ordinarily choose between purchasing gas or oil based on the price of each fuel. They also have no right to service from WGL; service to interruptible customers can be discontinued if that would be advantageous to WGL and its firm customers.

The Commission's decision abandoned the traditional practice of fixing a rate that the WGL must charge interruptible customers; it adopted instead a flexible rate design which permits WGL to vary the price of gas sold to such customers on a month-to-month basis. Having reviewed the record, we conclude that the parties were given an adequate opportunity to present their positions on this issue and that the Commission's findings are supported by substantial evidence. We address only OPC's claim that the newly-adopted rate design unreasonably discriminates against firm customers.

The Commission has always recognized that it is both necessary and fair to maintain separate rate structures for firm and interruptible customers. In the past, WGL charged interruptible customers for their gas at a rate fixed several cents below that charged to firm customers. Moreover, while firm customers traditionally have been forced to pay a monthly customer charge allocated to WGL's fixed costs of service and unrelated to the amount of gas used, interruptible customers have paid only for commodity purchases.

This is not to say, however, that interruptible customers did not bear some portion of WGL's fixed costs. At each rate proceeding, the Commission would allocate to interruptible customers some portion of WGL's fixed costs to be recovered from revenues derived from gas sales. Thus, the Commission relieved firm customers of a portion of the responsibility for covering WGL's fixed costs "up front"; i.e., firm customers were not required to pay for the designated portion of the fixed costs even if WGL was unable to make sufficient interruptible sales to cover such costs. If WGL's interruptible sales declined so that net revenues from commodity sales to interruptible customers fell short of the level of fixed cost allocated to such sales, WGL absorbed the loss in lower earnings. In the short run, WGL, not its firm customers, bore the risk of declining interruptible sales. To the extent that interruptible sales declined over time, however, the amount of fixed costs properly allocated to such sales would decline and firm customers would be hurt.

In Order No. 7749, the Commission first recognized that the number of WGL's interruptible customers had increased dramatically since 1980. At the same time, the Commission took official notice of "the collapse of the OPEC oil price negotiations and newspaper reports of sharply falling oil prices." It expressed concern that a fixed commodity rate for interruptible sales would not permit WGL to "maintain the competitiveness of interruptible gas service," finding evidence for this concern in the presentations made by both WGL and GSA. The Commission recognized that it was in the interest of WGL's firm customers to have revenues from interruptible sales remain as high as possible.

The Commission also concluded that the traditional practice of allocating a predetermined percentage of fixed costs to interruptible sales failed to "ensure that WGL's firm customers' rates remain just and reasonable." Because interruptible customers did not pay a monthly customer charge, and because their contribution to fixed costs was limited to a predetermined percentage rather than being tied to the actual level of interruptible sales, the Commission felt interruptible customers did not always bear a fair share of WGL's fixed costs. This would be the case whenever interruptible sales for a year turned out to be higher than anticipated.

Order No. 7749 addressed these concerns by adopting a flexible interruptible rate structure under which WGL could change the price of gas for interruptible customers on a month-to-month basis. These monthly price changes are to be filed with the Commission for review. WGL will be permitted to keep 20 percent of the net revenue from interruptible sales as earnings and the remaining 80 percent of the net revenue will be allocated to fixed costs. In addition, interruptible customers will pay $60 per month as a customer charge and this will be allocated to fixed costs.

OPC argues, however, that this flexible rate structure also has negative effects on firm customers and that these effects render the system discriminatory and unreasonable. First, OPC points out that the flexible rate structure shifts the risk of an unanticipated decline in interruptible sales from WGL to the firm ratepayers. Whereas in the past WGL would absorb the loss if interruptible sales fell below the level used to determine the interruptible customers' contribution to fixed costs, under the new flexible rate structure a loss of interruptible sales would result in higher rates for firm customers (presumably because WGL set gas prices for interruptible customers too high or because the price of oil fell so low that gas simply could not compete). Second, OPC contends that, if WGL sets the price charged interruptible customers too low, WGL will fail to maximize the revenues derived from interruptible sales. This would decrease the size of the fixed cost contribution made by interruptible customers and increase the burden on firm customers.

We recognize the validity of OPC's concerns. We also note, however, that the flexible rate structure adopted in Order No. 7749 contains two features—(1) Commission review of monthly rate changes and (2) WGL earnings linked to net interruptible revenues (20 percent)—which mitigate the fear that WGL may set rates at an undesirable level. The Commission considered the possible problems with a flexible rate structure raised by OPC and balanced them against the considerable benefits of such a system. This court may not engage in a second such balancing. Our review of the record convinces us that the Commission has given reasoned consideration to all of the relevant factors and concerns discussed here and that the Commission's decision to adopt a flexible rate structure for interruptible customers was not unreasonable. Accordingly, we affirm this aspect of the Commission's decision.

III.

OPC next challenges the Commission's decision to deviate from its recently established policy pertaining to the method for calculating WGL's allowance for uncollectible accounts. In Formal Case No. 768, completed shortly before WGL filed its application for increased rate in the present proceedings, the Commission established a policy whereby WGL's allowance for uncollectible accounts would be determined by looking at the utility's average level of uncollectible accounts over the previous five-year period. In Order No. 7749, challenged here, the Commission set WGL's uncollectible account allowance at "the current book uncollectible accrual rate of 2.30 percent," instead of using WGL's five-year average accrual rate of 1.36 percent. Thus, the Commission abandoned its five-year average method in this proceeding.

When an agency departs from a settled practice or rule, it has a special duty to explain the reasons for that departure. See, e.g., Atchison, Topeka & Sante Fe Railway Co. v. Wichita Board of Trade, 412 U.S. 800, 807-09, 93 S. Ct. 2367, 2374-75, 37 L. Ed. 2d 350 (1973). However, so long as it supplies "a reasoned analysis indicating that prior policies and standards are being deliberately changed, not casually ignored," People's Counsel, 455 A.2d at 396 (quoting Greater Boston Television Corp. v. FCC, 143 U.S.App.D.C. 383, 394, 444 F.2d 841, 852 (1970), cert. denied, 403 U.S. 923, 91 S. Ct. 2233, 29 L. Ed. 2d 701 (1971)), an agency "may find that, although the rule in general serves useful purposes, peculiarities of the case before it suggest that the rule not be applied in that case." Atchison, Topeka & Santa Fe Railway Co., 412 U.S. at 808, 93 S. Ct. at 2375.

We find that the Commission has met its burden of giving a reasoned explanation for its departure from the policy of using a five-year average as the measure for uncollectible accounts. Order No. 7749 states:

The evidence presented in this case indicates a dramatic increase in WGL's uncollectibles since our decision in Formal Case No. 768. Therefore, based on this evidence of changed circumstances regarding WGL's collection problems, particularly with master-metered apartments, and in consideration of new developments in District of Columbia laws, we are persuaded to deviate for purposes of this proceeding only from our preferred practice of using a five-year average as the measure for determining uncollectible accounts.

The District of Columbia law referred to here is D.C.Code § 43-542 (1981), which became effective in 1980 and prohibits WGL from terminating gas service to master-metered apartment buildings without first giving the tenants an opportunity to receive service in their own names.

OPC argues that the above-quoted passage cannot serve as a reasonable explanation for departing from the policy of using a five-year average method. It contends that, because the Commission's stated purpose in adopting the five-year average method was to eliminate distortions in rate calculations resulting from the "erratic up-again-down-again pattern" in WGL's uncollectibles, the Commission cannot reasonably explain its departure from that method by pointing to WGL's unusually high level of uncollectible accounts. We disagree.

We understand the Commission to have found that enactment of D.C.Code § 43-542 resulted in a previously unforeseen, but now predictable and perhaps permanent, increase in the level of WGL's uncollectibles. Thus, reliance on a five-year average method in this case would have resulted in an underestimation of WGL's uncollectible accounts expense. The Commission's acceptance of the five-year average method as a means of accounting for factors such as unemployment that push WGL's uncollectibles up some years and down others did not preclude it from taking account of a new and different type of factor that rendered the five-year method an inaccurate indicator level of WGL's future uncollectibles. Because we find the Commission's findings to be supported and its explanation for the methods it used reasonable, we affirm its decision to use WGL's current book uncollectible accrual rate.

IV.

In calculating WGL's income tax expense allowance, the Commission did not include any adjustment for the tax deduction generated by the interest paid on WGL's short-term debt. OPC's motion for reconsideration pointed out the omission of a short-term debt adjustment and suggested that the Commission had made a computational error. The Commission rejected this suggestion and indicated that it had decided not to make any adjustment for the tax effects of short-term debt because short-term debt had not been included as a component of WGL's capital structure. By excluding short-term debt from WGL's capital structure, the Commission relieved ratepayers of any responsibility for paying for the financing of such debt. Accordingly, the Commission decided that the principle of interest synchronization dictated that ratepayers not receive the tax benefits generated by short-term debt.

OPC does not argue that the Commission's interest synchronization methodology is unreasonable or contrary to acceptable accounting procedures. Rather, OPC maintains that the failure to make an adjustment for short-term debt in the income tax expense allowance constitutes a departure from the methodology used by the Commission in Formal Case No. 768—a case in which the Commission also excluded short-term debt from the capital structure. OPC contends that the Commission has failed sufficiently to articulate its reasons for this departure from an established practice.

A careful reading of the Commission's orders in Formal Case No. 768, however, reveals that the Commission did not establish a general practice or policy of including the tax effects of short-term debt in its income tax calculations. PSC Order No. 7537, issued by the Commission in Formal Case No. 768, explained that the recognition of a short-term debt adjustment to the tax expense allowance in that case resulted from the particular facts presented there. The Commission used WGL's short-term debt interest figure in Formal Case No. 768 as a means of estimating the tax benefit that WGL would receive from a $30 million postponed debt issue. The Commission had properly excluded the postponed debt issue from WGL's capital structure for the year in question, but that debt would nonetheless generate tax benefits for WGL and the interest on that debt would ultimately be paid by the ratepayers.

Thus, the recognition of a tax deduction attributable to short-term debt in Formal Case No. 768 did not establish a practice of including such a deduction in every case. The Commission included WGL's short-term debt only as a substitute for postponed debt issue interest deduction, not for its own sake. Because there was no postponed debt issue excluded from the rate structure in the instant proceedings, there was no need to include a short-term debt adjustment to the income tax expense allowance.

V.

D.C.Code § 43-516 (1981) states that "[t]he Commission shall keep itself informed of all new construction, extensions, and additions to the property of all public utilities." Furthermore, D.C.Code § 43-1002 (1981) provides that "[n]o gas corporation or electrical corporation shall begin the construction of a gas plant or electric plant without first having obtained the permission and approval of the Commission." Pursuant to these statutory requirements, WGL petitioned the Commission for approval of its 1982-1986 construction budget. Although the budget approval sought by WGL would permit the utility to begin work on its proposed construction projects, it would not foreclose the issue of whether the costs actually incurred as a result of these projects were reasonable or should be passed on to WGL's customers. The Commission could address any challenge to the reasonableness of particular construction costs at future ratemaking proceedings before authorizing the inclusion of those costs in the rate base. Thus, the approval of a proposed construction budget reflects only the preliminary judgment of the Commission, consistent with its statutory duty to supervise WGL's construction plans even before those plans will have an effect on the rates charged customers.

After reviewing the evidence presented by all parties, the Commission noted several reservations concerning certain aspects of the proposed budget but gave the overall budget its approval:

It would appear from this record that WGL can use better internal coordination with respect to its forecasting efforts and planning processes. We also note that WGL's elasticity calculations need some technical improvement. In future rate cases, we recommend that WGL investigate the stability of its elasticity calculation and that such calculation be made on a more comprehensive basis using different time periods in order to determine whether or not changes are occurring. This procedure is warranted in forecasting activities because it determines whether or not a change in elasticity has taken place. It is our hope that WGL will improve the deficiencies we have noted with its internal coordination and planning processes. We invite all parties to address this issue in WGL's next rate case. However, since WGL's estimates are the best available on the record, we accept WGL's 1982-1986 construction budget as not being unreasonable.

OPC focuses on the last sentence from this portion of the Commission's decision and contends that the Commission has effectively shifted onto OPC the burden of proving that WGL's construction budget is unreasonable. It argues that WGL should bear the burden of proving the reasonableness of its proposed budget.

We do not read the Commission's decision to place a burden of production or persuasion on OPC. Instead, we understand the Commission to have required WGL to present evidence, which WGL did, and to have found that WGL's overall budget approval was not unreasonable. Although the Commission found several aspects of WGL's proposed budget to be problematic and targeted those matters for future consideration, it determined that, on the whole, WGL's estimates were the "best available." We conclude that these findings are sufficient to meet the Commission's statutory obligation to review WGL's future construction plans in advance of considering the extent to which those costs should be included in the rate base.

VI.

In previous proceedings, the Commission has indicated an interest in developing a ratemaking theory which used marginal cost pricing as a means of allocating WGL's revenue requirement between different classes of customers and between customer and commodity charges. For instance, in Formal Case No. 680, the Commission recognized that "marginal cost pricing is a valid means of better tracking costs to the activities which incur them and of improving the price signal communicated to consumers." The Commission designated an issue in the present proceedings asking each party to propose a method for calculating WGL's marginal costs for each component of its service.

The Commission's consideration of a working definition for WGL's marginal cost of gas supply was complicated by the fact that the cost of gas to WGL's suppliers varies greatly depending on the source of the gas. The higher cost of exploiting new sources of gas, combined with the regulatory policies of the Natural Gas Policy Act, 15 U.S.C. §§ 3301-3432 (1982), makes gas from new wells considerably more expensive than the gas from old wells. The price which WGL's suppliers actually charge WGL for gas is a "rolled-in average" of the cost of old gas and new gas, referred to as the "pipeline price." This is to be contrasted with the "wellhead price" of new gas, which may be two or more times as high as the pipeline price.

As supplies of old gas are depleted, the pipeline price at which WGL must purchase gas will steadily rise and approach the wellhead price of new gas. For this reason, OPC argued to the Commission that, in order to send the proper price signal to consumers regarding the rising replacement cost of gas, WGL's marginal commodity cost should be defined as the wellhead price of new gas. OPC argued that this definition of marginal commodity cost would encourage the proper level of conservation.

Both WGL and PSC staff urged the Commission to reject OPC's proposal and to adopt the rolled-in average pipeline as WGL's marginal commodity cost. They contended that OPC's definition of marginal commodity cost would unnecessarily discourage the use of gas in the District of Columbia based on an inflated cost estimate that would not be realized until some unspecified point in the future. They advocated the use of WGL's pipeline price, which would increase gradually over time, as a means of matching the price signal sent to consumers with the changing cost experienced by WGL in the real world. Finally, PSC staff argued that the conservation that would result from an adoption of OPC's proposal would not inure solely to the benefit of District of Columbia ratepayers but would be spread to all gas consumers in the United States.

The Commission reviewed both of the arguments presented by both sides on this issue and concluded that "WGL's marginal commodity cost is the cost which WGL incurs in providing an additional unit of natural gas and is to be determined by the average cost of pipeline supplies." This court has recognized the wisdom of previous efforts by the Commission to anticipate and take account of the rising replacement cost of gas in allocating costs. Washington Gas Light Co., 450 A.2d at 1208. At the same time, we have consistently recognized the need to defer to the "expertise of the commissioners in the complex and esoteric area of utility regulation," especially where "[t]heories of ratemaking" are at issue. Id. at 1193. We conclude that the Commission has considered all of the relevant factors concerning the marginal cost issue and that its decision is "`reasonable, explained, and supported.'" Id. (citation omitted).

Affirmed.

NOTES

[1] OPC argued to the Commission that factors such as poor management, inefficient operations, or prevailing economic conditions could be responsible for WGL's failure to earn its authorized return. The Commission rejected this argument, noting that OPC had not presented any specific evidence to support its theory that problems other than attrition were responsible for WGL's earnings erosion.

Before this court, OPC argues that the Commission's decision is flawed because the Commission failed to exclude the possibility that general economic conditions or similar factors which are not necessarily linked to attrition were responsible for the downgrading of WGL's bond rating and the other signs of earnings erosion. We reject this argument. When a finding of attrition is supported by substantial evidence in the record, we will not set it aside simply because the Commission has not explicitly considered and ruled out every possible alternative explanation for the utility's deficient revenues.