Fitchburg Gas & Electric Light Co. v. Department of Public Utilities

Cordy, J. Fitchburg Gas and Electric Light Company (Fitch-burg or company) appeals from a ruling of the Department of [801] Public Utilities (department)2 requiring it to reimburse its customers over $4.6 million in gas supply costs incurred during the 2007-2008 and 2008-2009 purchasing seasons. In its ruling, the department determined that Fitchburg had failed to seek preap-proval of its gas purchasing plans for those years as it had been required to do, because the plans were intended to stabilize gas supply prices through the use of forward contracts.3 The department separately found that Fitchburg acted imprudently when executing its gas purchasing plans. An appeal was filed with a single justice of this court, pursuant to G. L. c. 25, § 5, and was reserved and reported, without decision, to the full court. We conclude that the department’s determination that Fitch-burg’s purchasing plans required preapproval was erroneous, as the plans incorporated only traditional risk management techniques that had previously never been subject to the department’s preapproval. Penalizing Fitchburg for failing to seek preap-proval, when such preapproval was never required, exceeds the department’s authority and amounts to an error of law. With respect to the allegedly imprudent purchases, we agree with the department that one of the purchases at issue, made in November, 2008, was unreasonable and imprudent, but hold that the department’s findings of imprudence with regard to the balance of the purchases in 2007 and 2008 were not supported by substantial evidence. 1. Regulatory background. The department supervises rate setting by natural gas local distribution companies (LDCs). G. L. c. 164, § 94. The department reviews gas procurement [802] contracts and rate setting consistent with the legislative policy “to provide a reliable energy supply for the commonwealth with a minimum impact on the environment at the lowest possible cost.” G. L. c. 164, § 69H. A natural gas bill consists of two components: a supply cost and a distribution cost. Bay State Gas Co., D.T.E. 01-09 at 3 (2001). The supply cost refers to the expense involved in purchasing and transporting gas on the interstate pipeline to Massachusetts. Id. Distribution costs refer to an LDC’s operating costs for bringing gas from the interstate pipeline to a customer’s meter. Id. The present case concerns only supply costs. Supply costs are recovered from customers on a dollar-for-dollar basis, through a reconciliation mechanism known as the cost of gas adjustment clause (CGAC). Id. That is, LDCs do not profit from the supply portion of a gas bill, and the cost of purchasing gas is a straight pass-through to the customer. Id. See 220 Code Mass. Regs. §§ 6.00 (2008). The rate that LDCs bill customers for gas supply is called the gas adjustment factor (GAF). 220 Code Mass. Regs. § 6.06. LDCs recalculate GAFs on a semiannual basis, once for the peak season of November 1 to April 30 and once for the off-peak period of May 1 to October 31. See 220 Code Mass. Regs. § 6.11. In addition, LDCs are required to submit revised GAFs any time their deferred gas-cost balances at the end of a period are projected to be greater or less than five per cent of the total seasonal gas costs stated in their effective GAF. Investigation Concerning the Cost of Gas Adjustment Clause, D.T.E. 01-49-A at 8 (2001). For supply costs to be recovered, the expenditures must have been reasonably and prudently incurred. G. L. c. 164 § 94G. “It is well-settled that while public utilities should be permitted to charge rates which are compensatory of the full cost incurred by efficient management, they may not recover costs which are excessive, unwarranted, or incurred in bad faith.” Boston Gas Co. v. Department of Pub. Utils., 387 Mass. 531, 539 (1982), and cases cited. When conducting a prudence review, the department determines whether a utility’s actions, based on all that it knew or should have known at the time, were reasonable and prudent in light of the circumstances which then existed. Hing-ham v. Department of Telecomm. & Energy, 433 Mass. 198, [803] 202 (2001). Such a determination may not be made on the basis of hindsight judgments, nor is it appropriate for the department merely to substitute its own judgment for the judgments made by the management of the utility. Attorney Gen. v. Department of Pub. Utils., 390 Mass. 208, 229 (1983). The present case concerns risk management techniques used in the procurement of natural gas. During the 1980’s and 1990’s, the price of natural gas remained essentially stable, at between two dollars and $2.50 per Dekatherm (DTh).4 Bay State Gas Co., D.T.E. 01-09 at Executive Summary (2001). LDCs would purchase approximately one-third of their winter needs during the off-peak season, and would store the gas in underground reservoirs. The balance of the purchasing would take place during the peak winter season, using either “first of the month” or “spot” (i.e., daily) pricing. The winter of 2000-2001 was the coldest nationwide in the 105 years of United States Weather Bureau records. After fifteen years of essentially stable prices, the price of gas nearly quintupled. Bay State Gas Co., D.T.E. 01-09, supra. This sharp increase was passed directly to consumers through the CGAC. Both LDCs and the department recognized the desirability of shielding consumers from such dramatic price fluctuations. Id. at 10-11. In December, 2001, the department opened an investigation concerning the use of risk management techniques to mitigate gas price volatility. Investigation of Risk-Management Techniques, D.T.E. 01-100-A (2002) (Order). The investigation considered, in pertinent part, whether LDCs should be allowed or required to implement a risk management program, and what standard of review the department should use when reviewing a proposed program. Id. at 1-2. The department issued its findings in the October, 2002, Order. Order, supra. The Order permitted risk management programs that incorporated financial derivatives,5 subject to certain conditions. See Order, supra at 3-8. [804] The Order evaluated numerous concerns relating to the use of financial derivatives to manage risk and stabilize the price of gas. At the outset, the Order highlighted the distinction between derivatives trading and traditional forms of risk management, including forward contracting: “A distinction should be made regarding traditional forms of risk-management, which include the use of storage to offset winter demand, through the use of physical gas purchases and forward contracts, and modern risk-management functions, which use financial futures and option contracts to effectuate various forward pricing strategies. For purposes of this Order, both ‘risk-management’ and ‘hedging’ are used interchangeably to mean the use of financial derivative products in combination with physical gas purchases to mitigate commodity price volatility” (emphasis added). (Footnotes omitted.) Order, supra at 2. The Order then explained why derivatives could be risky. Purchasing derivatives can be costly; and, while the use of derivatives may dampen volatility, it is unlikely to produce prices below index levels averaged over time. See Order, supra at 5-6. Additionally, LDCs that attempt to “beat the market” may engage in speculative purchases that result in over-all higher risk. Id. at 6. Further, overhead expenditures, such as professional staffing, information technology, and consulting, may render the use of financial derivatives prohibitive, especially for small companies. Id. Nevertheless, the department refused to bar categorically risk management plans that use financial derivatives, finding that the risks of such plans “can appropriately be reviewed on a case-specific basis, and in unison with an LDC’s supply design, size, and position in the market, before any determination could be made as to whether the risks are too high.” Order, supra at 7. Therefore, the department concluded that it would “allow, but not require, Massachusetts LDCs to use financial risk-management instruments to mitigate commodity price volatility,” and would “review and approve LDC risk-management proposals on a case805 specific basis.” Id. at 8. At the end of the Order, the department set out the standard of review for such risk management plans.6 Between 2002 and 2009, three LDCs sought preapproval of risk management plans. Only one of the plans specifically incorporated financial risk management instruments. In each case, the department applied the standard elucidated in the Order but failed to clarify whether the Order had mandated the companies to seek preapproval. On August 25, 2003, KeySpan Energy Delivery New England (KeySpan)7 sought approval of a risk management plan. Key-Span, D.T.E. 03-85 (2003). KeySpan had already been purchasing approximately thirty-two per cent of its winter gas requirements during the preceding spring and summer months, storing the gas in underground storage facilities or in liquefied natural gas tanks. Id. KeySpan proposed to “lock in” the price of an additional quantity of gas through the use of forward contracts, so that the price would be locked in for approximately two-thirds of KeySpan’s winter requirements. KeySpan would lock in prices for equal quantities of gas, once a month, over the twelve months preceding a given winter.8 Id. at 1-2. KeySpan did not propose to use financial derivatives.9 Id. at 4 n.1. The department applied the standard of review articulated in [806] the Order, even though KeySpan’s proposal did not incorporate financial derivatives. It explained: “For the sake of accuracy, we note that the standard of review established in D.T.E. 01-100-A, at 28-29 was for ‘risk-management plans.’ [KeySpan’s] proposal here addresses only timing of gas purchases and not derivatives. Despite some terminological blurring when discussing ‘risk-management’ and ‘hedging,’ (id. at 2, n.5 and n.6) the Department did distinguish ‘financial derivative products’ and ‘physical gas purchases,’ even where both might be used to manage or hedge risks. The conceptual or analytic difference is worth preserving. [KeySpan’s] proposal amounts to only part of a risk-management plan such as that contemplated in [the Order]. Nonetheless, the six criteria in [the Order] provide a useful framework for analyzing this gas purchasing plan.” (Emphasis added.) KeySpan, D.T.E. 03-85, supra at 4 n.1. The department approved KeySpan’s plan on November 13, 2003. It did not address whether KeySpan had been required to seek preapproval for this plan, but did specify that, “[i]n the event that [it] intends to use financial derivatives, KeySpan must seek department approval prior to entering into this type of transaction.” KeySpan, D.T.E. 03-85, supra at 5 n.2. In June, 2004, NStar Gas Company (NStar) sought approval of a similar purchasing plan. NStar, D.T.E. 04-63 (2005). Like KeySpan, NStar had already been purchasing and storing approximately one-third of its winter gas needs in underground reservoirs and liquefied gas facilities. Id. at 1. NStar proposed to lock in the prices of an additional one-third of its anticipated winter demand. It would make these purchases in equal increments over the course of twelve months. Id. at 2. However, NStar also proposed to resolve discrepancies between the locked-in prices and the date of delivery prices by entering into swap arrangements with large financial institutions. The department approved NStar’s plan on February 28, 2005, including its contracts with financial institutions. Id. at 5. The purpose of the arrangement was not to beat the market, explained the department, and the contracts would not cause NStar to incur transaction costs. Id. [807] Finally, in January, 2006, New England Gas Company (New England Gas) sought approval to purchase up to fifty per cent of its winter gas supply, at even increments over a twenty-month period. New England Gas Co., D.T.E. 06-3 (2006). New England Gas had previously been purchasing and storing twenty-eight per cent of its winter gas needs. The department approved the plan. Id. at 1, 5. As in the previous two cases, the department did not comment on whether New England Gas had been required to seek preapproval for its plan. 2. Fitchburg’s gas purchase plans and the present enforcement proceedings, a. Fitchburg’s gas purchase plans. Following the winter of 2000-2001, Fitchburg developed its own risk management program. Fitchburg had already been storing approximately twenty-two per cent of its winter peak period gas requirements in underground tanks. Fitchburg supplemented this strategy by executing forward contracts. Fitchburg did not lock in prices using derivative products, such as futures, options on futures, or swap contracts. Fitchburg initially operated this risk management strategy without specific goals as to the quantity of gas or the number of times in which to lock in prices. The company made its price-locked purchases based on a number of market indicators, such as weather, forward-looking price trends, and anticipated market volatility. In order to secure a stable price of gas for the 2007-2008 peak period, Fitchburg began locking in prices after it executed a supply contract on June 29, 2007. The company locked in the 2007-2008 peak winter period supply requirements by making nine purchases, evenly spaced over a nine-week period from June 29, 2007, through August 27, 2007. These purchases amounted to 69.3 per cent of the company’s total peak winter period requirements, of which 17.9 per cent was planned storage volumes and 51.4 per cent was price-lock purchases. The amounts of these purchases varied from week to week, from a low of 68,647 DTh to a high of 98,662 DTh. In January, 2008, Fitchburg promulgated internal purchasing guidelines for risk management. The purchasing guidelines called for Fitchburg to lock in one-tenth of its peak period baseload10 [808] requirements each month over a ten-month period, from the previous November through August.11 Having missed making purchases in the first three months of the program — November, December, and January — Fitchburg consolidated its purchases into seven months, locking in prices at eight points, approximately evenly spaced, between February and August, 2008.12 The monthly amounts of these purchases varied from 90,000 DTh to 127,855 DTh. In addition to these preplanned price locks, Fitchburg executed an unplanned price-lock on November 25, 2008, purchasing 107,846 DTh at $6.52 per DTh for delivery during January through April, 2009. This price-lock followed a rapid and unprecedented decline in the price of gas between July and November, 2008.13 Fitchburg based its decision to lock in the additional supply on the following factors: (1) the New York Mercantile Exchange (NYMEX) prices had declined by approximately fifty per cent since July, 2008; (2) Fitchburg estimated that it could reduce its total peak period costs by an additional ten to fifteen per cent when combined with the price-locks already in effect; (3) temperatures for the next thirty to sixty days were forecasted to be below normal; and (4) market indices indicated an upward price trend for December, 2008, through April, 2009. However, gas prices continued to plummet throughout the winter of 2008-2009. On February 1, 2009, the price went down to $4.49 per DTh; by April 1, it was $3.65 per DTh. In response to this dramatic decrease in prices, a number of LDCs sought department approval in January, 2009, to decrease their GAFs from those currently in effect. Fitchburg was not among them. On January 21, 2009, the department requested [809] Fitchburg to demonstrate that its projected gas cost balance at the end of the peak winter period would be within five per cent of its projections as of October 31, 2008. As a result of Fitch-burg’s response to that inquiry, the department commenced an investigation into Fitchburg’s gas procurement practices. b. The department’s investigation and findings. At the conclusion of its investigation, the department made the following findings: (1) Fitchburg had been required to seek preapproval of its risk management plan; (2) Fitchburg engaged in gas purchasing methods that were unreasonable and imprudent; (3) the proper measure of damages was the difference between what Fitchburg actually paid and what it would have paid had it purchased all its gas without any risk management plan whatsoever, buying each month’s supply on the first day of the month.14 The department ordered Fitchburg to repay its customers $4,648,075, plus interest, in gas supply costs. i. Preapproval. According to the department, both the Order and prior regulatory history15 required Fitchburg to seek pre-approval for its gas procurement strategy. The department rejected Fitchburg’s argument that it was exempt from preapproval because it did not use derivatives, concluding that the Order never restricted the requirement of preapproval to companies using financial derivatives. The department concluded that Fitch-burg should also have taken notice of the subsequent regulatory history, during which two of the three companies that had submitted requests for preapproval had not used financial instruments of any kind. Finally, Fitchburg’s plan raised concerns similar to derivatives, because Fitchburg incurred transaction costs in its price-lock contracts.16 [810] ii. Prudence. The department found that Fitchburg’s gas purchasing costs were not reasonably and prudently incurred. In support, the department adduced the following evidence: (1) In the November 25, 2008, purchase, Fitchburg engaged in speculation in an attempt to “beat the market,” which the company should have known not to have done, based on the policy directives in the Order. (2) The purchases made during 2007 and 2008 were spread out over a shorter time period than previously approved risk management programs. The KeySpan, NStar, and New England Gas plans had spread out purchases over twelve to twenty months. By contrast, Fitchburg locked in its prices for the 2007-2008 period over just nine weeks. The seven-month purchasing period of the 2008-2009 period was longer, but it was “nonetheless inconsistent with Department findings on the period of time needed to effectively mitigate price volatility.”17 (3) Fitchburg had varied its increments of time and volume for the price-locked purchases. According to the department, a reasonable and prudent risk management plan would have locked in precisely equivalent amounts, in uniform intervals. Moreover, the department found that Fitchburg failed to lock in sufficient price points for each individual winter month. For example, for the month of November, 2008, Fitchburg had locked in only one price point, taken the previous February.18 iii. Remedy. The department barred Fitchburg from recovering [811] from customers supply costs that, due to Fitchburg’s price-lock plan, were in excess of reasonable and prudent prices. The parties submitted alternative baseline prices against which to compare Fitchburg’s gas purchases. Fitchburg suggested comparing its actual purchases to the prices it would have paid for the same amount of gas purchased under its own proposed ten-month guidelines. Under that analysis, Fitchburg actually spent $282,844 less than a “prudent plan.” The Attorney General suggested comparing Fitchburg’s average prices to the average costs of the three companies that had utilized preapproved risk management plans. Under this analysis, Fitchburg overpaid by $863,368. The department rejected both of these approaches. The department concluded that Fitchburg could not enjoy the benefit of comparison to preapproved plans when it had failed to seek preapproval. Moreover, Fitchburg could not assume that the department would have found its ten-month purchasing plan prudent. The only types of purchasing that Fitchburg was permitted to use without prior approval were first of the month and spot market prices. Applying this benchmark, Fitchburg overcharged its customers by $4,648,075. The department ordered Fitchburg to refund this amount to its customers, with interest, and Fitchburg timely appealed pursuant to G. L. c. 25, § 5. 3. Analysis, a. Standard of review. Because Fitchburg’s petition raises no constitutional questions, we review the department’s decision to determine whether it is “based on an error of law, unsupported by substantial evidence, unwarranted by facts found on the record as submitted, arbitrary and capricious, an abuse of discretion, or otherwise not in accordance with law.” Massachusetts Inst. of Tech. v. Department of Pub. Utils., 425 Mass. 856, 868 (1997), citing G. L. c. 30A, § 14 (7). “ ‘[W]e give deference to the department’s expertise and experience in areas where the Legislature has delegated to it decision-making authority, pursuant to G. L. c. 30A, § 14,’ and ‘we have acknowledged that the department has broad authority to determine rate-making matters in the public interest.’ ” Southern Union Co. v. Department of Pub. Utils., 458 Mass. 812, 819 (2011), quoting Massachusetts Inst. of Tech. v. Department of Pub. Utils., supra at 867, 868. The question whether Fitchburg was required to seek pre812 approval turns on the interpretation of prior department rulings, and in particular, the Order. Ordinarily courts grant great weight to an agency’s interpretation of its own rulings. However, as we have noted in the context of interpreting statutes and regulations, this principle is “ ‘one of deference, not abdication,’ and this court will not hesitate to overrule agency interpretations of statutes or rules when those interpretations are arbitrary or unreasonable.” Moot v. Department of Envtl. Protection, 448 Mass. 340, 346 (2007), S.C., 456 Mass. 309 (2010), quoting Boston Preservation Alliance, Inc. v. Secretary of Envtl. Affairs, 396 Mass. 489, 498 (1986). We have recently declined to defer to the department’s interpretation of a rate settlement contract approved under its jurisdiction, when the plain language of the contract unambiguously led to the opposite conclusion. Southern Union Co. v. Department of Pub. Utils., supra at 820. With respect to the prudence review, the department’s factual findings are governed by the substantial evidence test. G. L. c. 30A, § 14 (7) (e). See Hingham v. Department of Telecomm. & Energy, 433 Mass. 198, 209-211 (2001). We will uphold the decision so long as “the findings by the authority are supported by substantial evidence in the record considered as a whole.” Id. at 210, quoting 1001 Plays, Inc. v. Mayor of Boston, 387 Mass. 879, 885 (1983). See G. L. c. 30A, § 14 (7). Substantial evidence is “such evidence as a reasonable mind might accept as adequate to support a conclusion.” G. L. c. 30A, § 1 (6). Agencies are required to make adequate subsidiary findings, including determination of each issue or fact necessary to the decision, so that this court may properly exercise its function of appellate review. Costello v. Department of Pub. Utils., 391 Mass. 527, 533 (1984), citing G. L. c. 30A, § 11 (8). When an agency bases its decision on specialized or expert knowledge, it must delineate the basis for that knowledge. “While the board is free to evaluate evidence in light of its expertise, it cannot use its expertise as a substitute for evidence in the record.” Arthurs v. Board of Registration in Med., 383 Mass. 299, 305 (1981). b. Requirement of preapproval. The company argues that neither the Order nor any other department regulation required it to seek preapproval for risk management plans that utilized only forward contracting. We agree. [813] The Order expressly limits its application to “the use of financial derivative products in combination with physical gas purchases to mitigate commodity price volatility.” Order, supra at 2. All the arguments submitted to the department during the proceedings that led to the issuance of the Order, in opposition to risk management techniques, focused on the dangers of derivative transactions. See id. at 3-7. The concluding portion of that section of the Order stated that the “Department will allow, but not require, Massachusetts LDCs to use financial risk-management instruments to mitigate commodity price volatility,” subject to case-by-case department review (emphasis added). Id. at 8. Based on the plain language of the Order, there was no obligation to seek preapproval for risk management plans such as Fitchburg’s plan. The regulatory history subsequent to the Order also failed to establish an obligation on Fitchburg to seek preapproval. Nowhere in the KeySpan, NStar, or New England Gas orders did the department expand the scope of the preapproval requirement. To the contrary: in the KeySpan order, the department reaffirmed the “conceptual or analytic difference” between forward contracting and financial derivative products. KeySpan, D.T.E. 03-85, supra at 4 n.1. While two of the three requests for preapproval did not involve financial transactions, the decisions by its competitors to seek preapproval did not bind Fitchburg to do the same.19 On appeal to this court, the department argues that attempting to mitigate price volatility without department approval is inconsistent with Fitchburg’s statutory obligation to purchase gas at the least possible cost. See G. L. c. 164, § 69H. This premise is flawed for two reasons. First, the Order itself recognizes that “the primary objective of a risk-management program is to mitigate price volatility, not to reduce the overall cost of gas.” Order, supra at 2 n.6. When articulating the standard of review for risk management plans, the department ordered companies to “maintain the objective of volatility mitigation and price stability rather than the objective of procuring prices below indices.” Id. at 28. Plainly, the department views the objectives [814] of lowest price and mitigation of volatility to be compatible, because the department may not take action or render decisions that are inconsistent with statutory provisions. See, e.g., Providence & Worcester R.R. v. Energy Facilities Siting Bd., 453 Mass. 135,144 (2009). Second, the traditional way for a company to satisfy its least cost obligation is to use competitive solicitations for market supply. See Hingham v. Department of Tele-comm. & Energy, supra at 211; Boston Gas Co. v. Somerville, 420 Mass. 702, 704-705 (1995). An LDC can thus satisfy its least cost obligation by engaging in competitive bidding, while simultaneously mitigating price volatility through forward contracting. The record contains ample evidence that Fitchburg did so. In sum, no statute or department precedent required Fitchburg to seek preapproval of its gas purchasing plan. The department’s conclusion to the contrary was an error of law. c. Prudence. The department concluded that all of Fitch-burg’s price-locked purchases — from summer, 2007; February-August, 2008; and November 25, 2008 — were unreasonable and imprudent. Fitchburg argues that this conclusion was not based on substantial evidence. We agree in part and disagree in part. With respect to the summer, 2007, and February-August, 2008, price-lock programs, we reverse the department’s finding as not based on substantial evidence. Concerning the November 25, 2008, purchase, however, we find that the department has amply supported its findings. The department found that the summer, 2007, and February-August, 2008, purchases were imprudent because they were spaced too closely together. As stated by the department, “A nine-week purchasing period could not allow for a sufficient number of pricing points to smooth prices and, thus, mitigate price volatility.” While the seven-month purchasing period of 2008 “allowed the Company to obtain more pricing points . . . , it is nonetheless inconsistent with Department findings on the period of time needed to effectively mitigate price volatility.” In support of this contention, the department pointed to risk management plans that other companies had submitted, which were for twelve months. See KeySpan, D.T.E. 03-85, supra at 4-5; NStar, D.T.E. 04-63, supra at 3-4. See also New England Gas Co., D.T.E. 06-3, supra at 2-3 (twenty months). In [815] each of those cases, the department had found that the plans would limit the exposure of ratepayers to price volatility. The flaw in the department’s analysis is that while the department found that twelve months of price-locking was sufficient to achieve price stabilization, it never determined that twelve months of price-locking was the minimum necessary to achieve rate stabilization. Nor did the department ever produce expert opinion in support of such a contention. Ultimately, the department seems to assert that any price-locking plan not strictly following its own precedents cannot be exercised in proper business judgment, and therefore must be deemed unreasonable and imprudent. Such a sweeping contention must be supported by expert testimony in the record; otherwise, it is unsupported by substantial evidence and must be rejected. Arthurs v. Board of Registration in Med., 383 Mass. 299, 305 (1981).20 By contrast, the department’s finding of imprudence with respect to the November 25, 2008, purchase, was amply supported by substantial evidence. The November 25,2008, purchase was not done pursuant to any price-lock program. Fitchburg conceded that it made that purchase in an effort to take advantage of one-time market conditions. While well intentioned, Fitch-burg’s executives engaged in speculative purchasing practices in an attempt to obtain below market prices. This practice was expressly forbidden in the Order. See Order, supra at 6. We therefore leave the department’s finding of imprudence undisturbed as to that purchase. d. Remedy. Applying the department’s method for determining a proper remedy, the calculation of the remedy is straightforward. Fitchburg must repay its customers for gas supply costs incurred on the November 25, 2008, purchase, over and above the prices that it would have paid had it purchased the same gas at the first of each month in which the gas was to be delivered. [816] Applying the first of the month standard to Fitchburg’s November 25, 2008, purchase, Fitchburg incurred $198,227 in unreasonable and imprudent gas costs.21 Fitchburg must refund this amount, with interest, to its customers. Finally, we note the dispute between the department and Fitchburg over whether Fitchburg incurred transaction costs when executing its price-locking program. See note 16, supra. This dispute is best addressed by the department in the first instance. See Southern Union Co. v. Department of Pub. Utils., 458 Mass. 812, 826 (2011) (remanding to department to decide new factual issues resulting from this court’s decision). If the department intends to pursue these costs on remand, it must reopen the record to allow Fitchburg to respond to these allegations. If Fitchburg imprudently contracted for higher prices than the market, the department may order Fitchburg to refund the difference between the market price and the lock-in price.22 4. Conclusion. We remand the case to the single justice, who is directed to remand the matter to the department. On remand, the department shall vacate its penalty of $4,648,075, plus interest, in favor of a penalty of $198,227, plus interest. The department may also reopen its investigation into whether Fitchburg incurred unreasonable and imprudent transaction costs. If the department finds that Fitchburg did incur such costs, it may take action as it deems appropriate and not inconsistent with this opinion. So ordered. Footnotes [2] In 1997, the Legislature reorganized the Department of Public Utilities into the Department of Telecommunications and Energy. St. 1997, c. 164, § 28. General Laws c. 25, § 1, as amended through St. 2007, c. 19, § 21, reverted its designation to the Department of Public Utilities. The present case references decisions issued both before and after 2007, and we shall refer to “the department” throughout. [3] Forward contracts, also known as “price-locks,” are agreements to buy or sell a quantity of gas, at the price as of the date of the contract, that is to be delivered on a certain date in the future. “If a forward contract contains a fixed price provision, the seller and the purchaser know exactly what the price will be upon delivery of the commodity and have, therefore, eliminated the risk of adverse price fluctuations on a certain quantity of gas.” S.P. Geurin & DJ. May, The Use and Regulation of Financial Risk Management Products in the Natural Gas Industry, Natural Gas Contracting: Strategies for the New Marketplace 23 (AJ. Wright & C.J. Sharpe eds. 1998). [4] A dekatherm (also spelled decatherm) is a unit of heat energy equivalent to ten therms or one million British Thermal Units (BTUs). [5] “Derivatives are financial instruments that derive their value from the price of another product. Some examples of derivative products are: futures, options [804] on futures, and swap contracts.” Investigation of Risk-Management Techniques, D.T.E. 01-100-A at 2 n.5 (2002) (Order). [6] Proposed plans had to: “(1) allow customers to volunteer to participate in the plan; (2) maintain the objective of volatility mitigation and price stability rather than the objective of procuring prices below indices; (3) ensure fair competition in the gas supply market; (4) allocate all costs to program participants only; (5) demonstrate the effect that the plan would have on the reliability and transparency of the commodity price; and (6) contain no incentive mechanisms.” Order, supra at 28. [7] KeySpan Energy Delivery New England (KeySpan) became a wholly owned subsidiary of National Grid USA (National Grid) on August 24, 2007. See Massachusetts Elec. Co., D.P.U. 09-39 at 263-264 (2009). The parties have represented that, at all times relevant to this proceeding, National Grid continued to operate the plan set up by KeySpan. [8] In a July, 2006, letter to the department, KeySpan subsequently sought to extend the duration of its forward contracts from twelve months to sixteen months. [9] In a subsequent filing, KeySpan claimed to use “financial ‘swap’ arrangements” to lock in prices in advance of the 2005-2006 winter season. KeySpan, D.T.E. 06-65 at 2 (2006). However, no mention was made of such arrangements during the initial request for approval. [10] Baseload refers to the minimum amount of gas an LDC expects to need to fulfil customer gas requirements over a given period of time regardless of changes in temperature. [11] Fitchburg chose a ten-month duration, rather than a twelve-month or longer period, due to the expected demands of its gas suppliers. Because Fitchburg distributes a relatively small quantity of gas annually, about one per cent of the total gas supply in the Commonwealth, Fitchburg expected that its suppliers would be unwilling to lock-in prices for smaller quantities than one-tenth increments. [12] Specifically, the purchases took place on February 15, March 17, April 17, May 19, June 27, July 18, July 25, and August 14. [13] In July, 2008, gas prices on the New York Mercantile Exchange (NYMEX) were approximately $13.30 per DTh; by November, they had plunged to below seven dollars per DTh. [14] The department also found that Fitchburg’s mention of its gas procurement strategies in other regulatory filings did not suffice to meet its disclosure and preapproval requirement. Because we hold that preapproval was not required, we need not reach the substance of this point. [15] The department expanded on prior regulatory history in the underlying proceeding. Prior to 2002, it reasoned, LDCs had not been permitted to engage in risk management at all, including forward contracting. The Order enabled LDCs to engage in risk management for the first time, but only if the LDCs sought approval. We have examined the cited precedents and we find no examples of express disapproval of forward contracting prior to 2002. In any case, the department has abandoned this portion of its argument on appeal. [16] The department reached this conclusion by comparing the prices paid by [810] Fitchburg with the NYMEX prices. NYMEX publishes three sets of daily prices —■ high, low, and settlement (i.e., average). According to the department, Fitchburg’s locked-in prices exceeded the NYMEX prices on fourteen of the eighteen dates on which it contracted. The department identified these prices by taking judicial notice of the NYMEX futures prices published in Platts Gas Daily. Fitchburg did not object to the taking of judicial notice, but it requested a hard copy of the data, which the department supplied. On appeal, Fitchburg claims that the department misapplied the numbers in Platts Gas Daily, a subscriber-only publication, and that Fitchburg was unable to correct the error because the department conducted this analysis only after the record was closed. The data were not added to the record, in respect of the proprietary rights of Platts Gas Daily. The record therefore does not indicate whether there were transaction costs. [17] The department did not analytically justify, or cite any expert opinion, explaining why a minimum of twelve months is necessary in order to mitigate price volatility. [18] The department did not provide an analytical explanation, or cite any expert opinion, suggesting the minimum number of lock-in points that are needed to mitigate price volatility. [19] We need not decide whether KeySpan and New England Gas sought preapproval out of a genuine belief that preapproval was necessary, or whether preapproval would have been necessary based on other features of their plans. [20] In addition to its failure to support its findings adequately, the department also appears to have ignored data in the record that suggested that Fitchburg’s price-lock program was indeed mitigating price volatility. For example, during the peak winter seasons between 2002 and 2009, the ten Massachusetts LDCs applied for a total of 101 interim GAF changes during the peak season. Fitch-burg applied for just five interim price changes during that period, the lowest of any of the gas suppliers. The department failed to address the import of this data at all. [21] On November 25, 2008, Fitchburg purchased 107,846 DTh at $6.52 per DTh, for $703,156. This gas was delivered in the following increments (first of the month cost in parentheses): January, 2009, 30,000 DTh ($6.16 per DTh); February, 2009, 27,846 DTh ($4.49 per DTh); March, 2009, 30,000 DTh ($4.07 per DTh); April, 2009, 20,000 DTh ($3.65 per DTh). Total costs based on the first of the month standard would have been $504,929. [22] The department presumed that Fitchburg was not entitled to engage in risk management at all. For this reason, the department calculated the $4,648,075 penalty as if Fitchburg had not entered into any forward contracts. The department did not, and would not have needed to, calculate the extra expense arising specifically from transaction costs. We have rejected the department’s fundamental premise and held that Fitchburg was entitled to engage in its risk management program. In light of our ruling, the question thus arises whether Fitchburg acted imprudently when executing its price-locks, and, if so, how much it overcharged its customers.