Connecticut Light & Power Co. v. Public Utilities Regulatory Authority

************************************************ The “officially released” date that appears near the beginning of an opinion is the date the opinion will be published in the Connecticut Law Journal or the date it is released as a slip opinion. The operative date for the beginning of all time periods for the filing of postopinion motions and petitions for certification is the “officially released” date appearing in the opinion. All opinions are subject to modification and technical correction prior to official publication in the Connecti- cut Law Journal and subsequently in the Connecticut Reports or Connecticut Appellate Reports. In the event of discrepancies between the advance release version of an opinion and the version appearing in the Connecticut Law Journal and subsequently in the Connecticut Reports or Connecticut Appellate Reports, the latest version is to be considered authoritative. The syllabus and procedural history accompanying an opinion that appear in the Connecticut Law Journal and subsequently in the Connecticut Reports or Connecticut Appellate Reports are copyrighted by the Secretary of the State, State of Connecticut, and may not be reproduced or distributed without the express written permission of the Commission on Official Legal Publications, Judicial Branch, State of Connecticut. ************************************************  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority THE CONNECTICUT LIGHT AND POWER COMPANY v. PUBLIC UTILITIES REGULATORY AUTHORITY (SC 21123) McDonald, D’Auria, Ecker, Alexander, Dannehy and Bright, Js. Syllabus In 2017, the plaintiff electric supplier filed an application with the defen- dant, the Public Utilities Regulatory Authority (PURA), seeking approval of amendments to the base rates it charges customers. In January, 2018, the plaintiff entered into a settlement agreement with various parties, includ- ing PURA’s prosecutorial division, that resolved the 2017 rate case and set new base rates for 2018 through 2020. The agreement also created a new mechanism, referred to as the new capital tracker, that allowed the plain- tiff to recover the costs of certain investments in infrastructure, including “core capital” spending, in its base rates. Under the new capital tracker, if the plaintiff spent more than a specified amount per year on core capital, it could recover its excess expenses during the term of the settlement agree- ment without having to wait for its next general rate case; procedurally, the plaintiff’s annual costs and collections under the tracker would be reconciled and added to its base rates each year at a rate adjustment mechanism (RAM) proceeding. PURA ultimately approved the settlement agreement in April, 2018. While the agreement was being negotiated and approved, five cata- strophic storms impacted Connecticut, damaging certain of the plaintiff’s equipment. The settlement agreement was then amended to provide that the plaintiff could seek review and recovery of all storm related costs incurred after December 31, 2016, either during its next general rate case or by initi- ating a separate, contested case. In November, 2018, the plaintiff initiated a contested case to recover the operation and maintenance costs incurred in connection with the five storms, but it did not seek to have PURA review or approve the nonincremental capital costs that it had incurred as a result of those storms. Instead, the plaintiff sought to recover the capital costs in its base distribution rates when it filed its annual RAM application in 2021. In its final decision in the 2021 RAM proceeding, PURA rejected the plaintiff’s argument that, under the terms of the settlement agreement, the plaintiff could recoup in its base rates the capital costs that it had incurred in connection with the five storms, subject only to a subsequent prudence review by PURA. Thus, PURA ordered the plaintiff to deduct more than $17 million in such spending from its new capital tracker request but indicated that, if the $17 million subsequently was deemed to be prudently spent, it would be recoverable in the plaintiff’s next general rate case. The plaintiff appealed from PURA’s final decision to the Superior Court. In upholding PURA’s decision, the trial court did not construe the disputed portions of the settlement agreement but, instead, concluded that PURA had discre- tion, pursuant to the statute (§ 16-19e) specifying how PURA is to assess the reasonableness of rates, to resolve the matter as it had done, and that  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority there was substantial evidence to sustain PURA’s final decision. On appeal from the judgment of the trial court, the plaintiff claimed that the court had improperly deferred to PURA’s decision to delay recovery of the capital spending when the court should have construed the settlement agreement de novo and concluded, as a matter of law, that the agreement unambiguously permitted the plaintiff to recoup capital costs incurred in connection with the five storms in its base rates. Held: The trial court first should have interpreted the settlement agreement and determined whether it was clear and unambiguous, rather than deferring to PURA’s rate-making discretion prior to doing so. When reviewing an agency’s interpretation of a contract or agreement that establishes utility rates and rate related policies and procedures, the trial court must first examine the contract or agreement itself and determine initially whether its language is clear and unambiguous in light of the context of the parties’ situation, insofar as PURA’s rate-making discretion under § 16-19e is constrained by binding settlement agreements that it has approved. Although the trial court did not initially determine whether the settlement agreement was clear and unambiguous, this court elected to address that issue in the interest of judicial economy and because the record was adequate for this court to make that determination in the first instance. This court concluded that the settlement agreement was facially ambiguous with respect to how the plaintiff was to recoup its capital costs incurred in connection with the five storms, and neither relevant, contemporaneous documents associated with the approval of the agreement, nor the available extrinsic evidence of the parties’ course of dealing, definitively resolved those ambiguities. Although the language of the settlement agreement, as executed by the parties before it was approved, was ambiguous with respect to how the plaintiff was to recover its capital spending on storm recovery, insofar as the agreement did not expressly address that issue, the plaintiff could not prevail on its claim that the approach established by the agreement, namely, the new capital tracker, unequivocally entitled the plaintiff to recover, in its 2018–2020 base rates, capital spending arising from the five storms. The plaintiff’s claim that the settlement agreement unambiguously permitted it to recoup its capital spending through the new capital tracker was premised largely on its argument that capital equipment damaged during a catastrophic storm qualifies as “emergent equipment failures,” as evidenced by a head- ing in a chart that was included in certain prefiled testimony submitted on behalf of the plaintiff that was incorporated by reference into the agreement. In the present case, the inclusion of “emergent equipment failures” as a row heading in a chart that was incorporated into the settlement agreement did not clearly and unambiguously indicate the intent of the parties, in early 2018, with respect to the treatment of catastrophic storm damage. Given the various dictionary definitions of “emergent,” the settlement agree- ment was facially ambiguous with respect to how capital spending on storm  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority recovery was to be treated, and the administrative record failed to clearly resolve that ambiguity, particularly in light of the plaintiff’s inconsistent references as to whether storm damage and emergent equipment failures represent distinct categories of capital spending for purposes of the plain- tiff’s internal accounting. A second ambiguity related to how the settlement agreement treated core capital spending that significantly exceeded the budget projections the plaintiff had included in its 2017 rate case application and that PURA had preapproved, as the language of the agreement, on its face, was susceptible to two plausible interpretations, and a review of the administrative record in the 2017 rate case did not clearly resolve that ambiguity. A third source of ambiguity related to whether damage from both catastrophic storms and noncatastrophic storms fell within the ambit of the core capital spending that the settlement agreement preapproved for recovery in base rates. The plaintiff’s position that the settlement agreement drew no distinction between catastrophic storms and other major storms and PURA’s position that only routine, noncatastrophic storm recovery costs were built into the base rates were both reasonable, and the available extrinsic evidence and the parties’ subsequent course of dealing did not resolve that ambiguity. In light of this court’s conclusions that the settlement agreement was ambigu- ous in various respects and that the relevant portions of the administrative record did not clearly and definitively resolve those ambiguities, this court remanded the case to the trial court so that it could resolve the ambiguities in light of the complete administrative record or so it could remand the case to PURA for consideration of all of the evidence relevant to resolving the ambiguities. Moreover, the parties did not specifically brief the issue of what deference was to be afforded an agency’s factual findings regarding the meaning of an ambiguous settlement agreement when PURA’s prosecutorial division was a party to that agreement, and, to the extent that the outcome of the case hinges on the degree of deference to be afforded to PURA’s factual findings with respect to the intentions of the parties in entering into the agreement, the trial court will have an opportunity to consider that issue in the first instance on remand. Argued December 11, 2025—officially released May 12, 2026 Procedural History Appeal from the decision of the defendant preclud- ing the plaintiff from recovering the costs of certain capital investments related to storm repairs, brought to the Superior Court in the judicial district of New Brit- ain, where the court, Cordani, J., granted the motion  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority to intervene filed by the Office of Consumer Counsel; thereafter, the court, Hon. Henry S. Cohn, judge trial referee, granted the defendant’s motion to dismiss as to counts one and two of the complaint, and the plaintiff appealed to the Appellate Court, Suarez, Clark and See- ley, Js., which dismissed the appeal; subsequently, the plaintiff withdrew the remaining count of the complaint; judgment for the defendant, from which the plaintiff appealed. Reversed; further proceedings. James J. Healy, with whom were Allison D. White and Vincent P. Pace, for the appellant (plaintiff). Seth Hollander, assistant attorney general, with whom, on the brief, was William Tong, attorney gen- eral, for the appellee (defendant). Andrew W. Minikowski, staff attorney, for the appel- lee (intervenor Office of Consumer Counsel). Opinion BRIGHT, J. The plaintiff, The Connecticut Light and Power Company, doing business as Eversource Energy, appeals from the judgment of the trial court dismissing its administrative appeal from a final decision of the defendant, the Public Utilities Regulatory Authority (PURA), which delayed until the plaintiff’s next rate case recovery of certain capital costs1 that the plaintiff believes are immediately recoverable in its current base rates pursuant to a 2018 settlement agreement. The plaintiff claims that the trial court (1) rather than inter- preting the settlement agreement, improperly deferred to PURA’s decision to delay recovery of the expenses, and (2) should have construed the settlement agree- ment de novo and concluded, as a matter of law, that 1 We are aware that the terms “capital expenditures,” “capital plant additions,” and “capital projects” or “capital programs” have distinct, technical meanings in the context of the settlement agreement. Because those distinctions are not directly relevant to the present dispute, however, and for convenience, we generally will use the terms “capital costs” or “capital spending” as shorthand for the capital investment funds that the plaintiff claims it was authorized to recoup from rate- payers under the agreement.  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority the agreement unambiguously permitted the plaintiff to recoup in its current base rates capital costs associated with recovery from five catastrophic storm systems that struck Connecticut between October, 2017, and May, 2018 (five storms). We agree with the plaintiff’s first claim but disagree that the settlement agreement is clear and unambiguous as to when the plaintiff is entitled to recoup its capital costs from the five storms. Because we conclude that the settlement agreement is facially ambiguous, we reverse the trial court’s judgment and remand the case for further proceedings. I A Before we discuss the history of this dispute, it is help- ful to briefly review the regulatory regime that governs the two private electric distribution companies (EDCs) that serve the Connecticut market: the plaintiff and The United Illuminating Company. PURA regulates the level and structure of rates set by the EDCs; General Statutes § 16-19e (a); and no EDC is permitted to charge rates in excess of those previously approved by PURA. General Statutes § 16-19 (a). Section 16-19e (a) provides the standards by which PURA is to assess the reason- ableness of such rates. Those standards include, among other things, “that the level and structure of rates be sufficient, but no more than sufficient, to allow public service companies to cover their operating costs includ- ing, but not limited to, appropriate staffing levels, and capital costs, to attract needed capital and to maintain their financial integrity, and yet provide appropriate protection to the relevant public interests, both existing and foreseeable”; General Statutes § 16-19e (a) (4); and “that the level and structure of rates charged custom- ers . . . reflect prudent and efficient management of the franchise operation . . . .” General Statutes § 16-19e (a) (5). PURA periodically approves new rates, either upon application by an EDC in a quasi-judicial general rate case; see General Statutes §§ 16-19 (a) and 16-19a (a) (2); or through a settlement agreement. See General Statutes  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority § 16-19jj. However the rates are set, “at intervals of not more than four years from the last previous general rate hearing of each . . . [EDC, PURA must] conduct a complete review and investigation of the financial and operating records of each such company and hold a pub- lic hearing to determine whether the rates of each such company are unreasonably discriminatory or more or less than just, reasonable and adequate . . . .” General Statutes § 16-19a (a) (1). In addition, PURA conducts an annual rate adjustment mechanism (RAM) proceed- ing for each EDC, during which it reconciles authorized additions to and subtractions from the EDC’s rates dur- ing the prior year. B The administrative record in this case is voluminous, and we therefore will confine our description of the rel- evant facts and procedural history to the following sum- mary of considerations pertinent to the issues on appeal.2 In 2017, the plaintiff filed an application with PURA, seeking approval of amended rate schedules (2017 rate case). Following public hearings and extensive negotia- tions between the plaintiff, consumer advocates from the intervenor, the Office of Consumer Counsel (OCC), and PURA’s prosecutorial division (PRO), those par- ties entered into a settlement agreement on January 11, 2018. In addition to establishing new electric base rates for at least the 2018–2020 period, the agreement embraced a novel approach to the ongoing need for the plaintiff to invest substantial capital in maintaining, expanding, and upgrading the state’s electric distribu- tion network. Section 3 of the settlement agreement “approved” annual “core capital” spending of $270 million for the first three years of the agreement and allowed the plain- tiff to recover those funds in its current base distribution 2 Additional relevant procedural history is set forth in a prior decision of the Appellate Court in this matter. See Connecticut Light & Power Co. v. Public Utilities Regulatory Authority, 223 Conn. App. 136, 139–42, 307 A.3d 967 (2023).  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority rates. Core capital spending includes, among other things, spending intended to improve the reliability of the electric distribution system. The agreement also envisioned that the plaintiff could spend more than $270 million per year on core capital and recover part or all of that excess in base rates during the term of the agreement, without having to wait for the next rate proceeding. The mechanism for recovering that excess core capital spending was referred to as the “new capital tracker” or the “electrical system improvement” (ESI) tracker. Although the settlement agreement set no absolute limit on the total core capital spending that the plain- tiff could recover through the ESI tracker each year, it specified that only the subset of core capital spend- ing identified as reliability spending could be recovered through the ESI tracker. This provision effectively set an upper limit on the total core capital spending that could be automatically recovered under the agreement.3 Under the agreement, the plaintiff’s annual costs and collections under the ESI tracker would be reconciled and added to customers’ base rates each year through a RAM proceeding. The primary purpose of the ESI tracker, and the related provisions of the settlement agreement, was to balance two competing aims. On the one hand, the agreement sought to ensure that any base rate funds allocated to capital improvements would in fact be spent thereon, with associated benefits to electricity customers, as well as secondary benefits for Connecticut businesses and the state’s tax base, and would not simply become addi- tional profits for the plaintiff. On the other hand, the ESI tracker and the annual RAM reviews afforded the plaintiff some flexibility to reprioritize and increase its capital spending to respond to changing needs, without 3 If, for example, the plaintiff had spent $30 million on reliability improvements in 2018, then the maximum total core capital that it could recover in base rates for that year would be $300 million (the $270 million preapproved in base rates, plus the full $30 million in reliability spending that could be recovered through the ESI tracker).  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority having to reduce ongoing investment in the electrical system’s reliability and resiliency. The first of the five storms—a major windstorm— already had occurred by the time the plaintiff, the OCC, and the PRO entered into the settlement agreement in early January, 2018. Over the next several months, during which PURA reviewed the agreement for its approval, three damaging nor’easter storms hit the state. Accordingly, one focus of the parties’ discussions dur- ing this period centered around how the additional costs associated with catastrophic storm recovery would be handled. PURA ultimately approved the agreement in April, 2018, conditioned on several changes, which were memorialized in amendment one. One set of changes addressed how PURA would review for prudence the more than $60 million in incremental storm recovery costs the plaintiff had incurred since the end of 2016. Amendment one provided, for example, that “the [p]ost- December 31, 2016 [s]torm [c]osts are subject to a full prudence review by PURA, the OCC, [the] PRO and any other entities granted intervenor or party status in a future contested case convened by PURA,” and that the plaintiff could “elect [either] to have PURA evaluate the prudence of all [p]ost-December 31, 2016 [s]torm [c]osts in . . . its next rate case or . . . to initiate such review sooner in a separate contested case involving such storm costs any time after the conclusion of this pending rate case . . . .” It is undisputed that the plaintiff chose the second option and, in November, 2018, initiated a proceeding to recover the incremental, or operation and maintenance, costs associated with the five storms (five storms pro- ceeding). The parties also agree that the plaintiff did not seek in that proceeding to have PURA review or approve the nonincremental capital costs it incurred as a result of the five storms. As we will discuss, the parties do not agree as to the actual scope of PURA’s review in the five storms proceeding. See footnote 22 of this opinion and accompanying text.  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority In any event, by mid-2020, it had become clear that the parties interpreted the settlement agreement differ- ently with respect to the question of recovery of capital costs related to the five storms. The issue appears to have first arisen during the 2020 RAM proceeding in which PURA reviewed the plaintiff’s 2019 finances and performed its annual ESI reconciliation. In May, 2020, Michael L. Shelnitz, the plaintiff’s manager of revenue requirements for Connecticut, responding to an interrogatory from the OCC, stated that, under the plaintiff’s interpretation of the agreement, “[c]apitalized storm costs are considered part of the [plaintiff’s] ‘basic business’ program and therefore [are] included in base distribution rates.” In November of that year, PURA issued a draft decision in the 2020 RAM case, rejecting the plaintiff’s interpretation of the agreement. Specifi- cally, PURA took the position that, although the ESI tracker allows for some deviation from the plaintiff’s original proposed capital programs as priorities change from year to year, the plaintiff nevertheless “is required to abide by the authorized budget” that the plaintiff “pre- sented in the original testimony of the 2017 rate case.” Noting that the plaintiff’s claimed capital additions for 2018 and 2019 “greatly exceed[ed] the levels authorized by the 2017 [r]ate [c]ase [d]ecision” given the inclusion of capital spending on catastrophic storm remediation, PURA stated that those costs should be subtracted from the plaintiff’s claimed capital costs for those years. In both its written exceptions to the draft decision and during oral argument before PURA, the plaintiff requested that PURA remove any language from its deci- sion purporting to definitively construe the settlement agreement in this respect. The plaintiff asked PURA to defer determination of the issue to a future proceeding in which PURA would examine the prudence of the capi- tal costs at issue. In its final decision in the 2020 RAM case, issued in February, 2021, PURA reiterated that it remained unconvinced by the plaintiff’s arguments, but it removed the draft language construing the agreement. At the same time, PURA observed that the adoption  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority of the ESI tracker had rendered the plaintiff’s capital spending less transparent and indicated that it intended to conduct “a full prudency review of all core and capital programs undertaken since the establishment of the ESI tracker,” which would be “conducted in the context of [PURA’s] interpretation of the associated provisions in the [s]ettlement [a]greement . . . [and which would] take place during the [plaintiff’s] next RAM proceeding or . . . next general rate case . . . .” The following month, in March, 2021, the plaintiff submitted its RAM filing for the 2020 calendar year (2021 RAM proceeding). As indicated in its final decision in the 2020 RAM case, PURA used this opportunity not only to commence a prudence review of the plaintiff’s capital costs but also to formally construe the disputed provisions of the settlement agreement. In September, 2021, PURA issued a final decision in the 2021 RAM pro- ceeding. This decision is the subject of the present appeal. PURA formally rejected the plaintiff’s argument that, under § 3 of the agreement, the plaintiff could recoup its capital costs related to the five storms in its current base rates, up to the limits discussed in footnote 3 of this opinion, subject only to a subsequent prudence review by PURA. PURA rejected this argument because PURA (1) construed the agreement as preapproving for inclusion in current base rates only those capital expenditures that had been projected in the plaintiff’s 2017 rate case application, (2) concluded that capital spending on storm recovery costs had not been included in those projections, and (3) read amendment one to the agreement to mean that all storm recovery costs, including capital costs, must be reviewed and approved either in a future rate case or in a separate proceeding. Accordingly, PURA concluded that the plaintiff had inappropriately included its capital spending on recovery from the five storms in its base distribution rates. PURA ordered the plain- tiff to subtract more than $17 million in such spending from its ESI core capital recovery request and indicated  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority that, if found to be prudently spent, that amount would be recoverable in the plaintiff’s next rate case. The plaintiff appealed from that decision to the Supe- rior Court pursuant to General Statutes §§ 4-183 (a) and 16-35 (a), contending, among other things,4 that PURA erroneously had disallowed the $17 million in storm recovery capital spending. The trial court dismissed the appeal,5 thereby upholding PURA’s decision, without construing the disputed sections of the settlement agree- ment. Instead, the trial court concluded that administra- tive agencies have “significant latitude in setting rate policies,” that “PURA has the discretion under § 16-19e to resolve this matter as it has done,” and, therefore, that “there is substantial evidence to sustain PURA’s final decision and . . . its decision was reasonable.” (Internal quotation marks omitted.) The plaintiff appealed to the Appellate Court, and, pursuant to General Statutes § 51-199 (c) and Practice Book § 65-1, we transferred the appeal to this court. Additional relevant facts will be set forth in the legal analysis that follows. II The trial court determined that the standard of review was dispositive of the plaintiff’s claim, and it is there that our analysis begins. It is well established that, under the Uniform Administrative Procedure Act, the trial court “shall not substitute its judgment for that of the agency as to the weight of the evidence on questions of fact. The court shall affirm the decision of the agency unless the court finds that substantial rights of the person appealing 4 The plaintiff’s other principal claim in its administrative appeal, that PURA set an incorrect interest rate for carrying charges, has been resolved and is not at issue in the present appeal. 5 The trial court initially dismissed the first two counts of the plaintiff’s complaint but remanded the case to PURA with direction to render a supplemental decision regarding the issue raised in count three. See Connecticut Light & Power Co. v. Public Utilities Regulatory Author- ity, 223 Conn. App. 136, 140–42, 307 A.3d 967 (2023). The Appellate Court dismissed the plaintiff’s appeal from that decision for lack of a final judgment. See id., 139, 149. Thereafter, the plaintiff withdrew the remaining count of its complaint.  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority have been prejudiced because the administrative find- ings, inferences, conclusions, or decisions are: (1) In violation of constitutional or statutory provisions; (2) in excess of the statutory authority of the agency; (3) made upon unlawful procedure; (4) affected by other error of law; (5) clearly erroneous in view of the reliable, probative, and substantial evidence on the whole record; or (6) arbitrary or capricious or characterized by abuse of discretion or clearly unwarranted exercise of discre- tion.” General Statutes § 4-183 (j). Moreover, as the trial court correctly noted, we frequently have recognized that the legislature has afforded administrative agen- cies “significant latitude” when it comes to rate setting matters, both in terms of the specific formulas that are utilized and broader policy choices. Greenwich v. Dept. of Public Utility Control, 219 Conn. 121, 126 n.3, 592 A.2d 372 (1991); see id., 126–27 and nn.3 and 4. Less well established is the manner in which these standards apply to the judicial review of an agency’s interpretation of a contract that establishes utility rates and rate related policies and procedures. In the ordinary case, “[t]he interpretation of a contract, being a deter- mination of the parties’ intent, is . . . a question of fact that is subject to reversal on appeal only if it is clearly erroneous. . . . When there is definitive contract lan- guage, however, the determination of what the parties intended by their commitments is a question of law over which [judicial] review is plenary.” (Citation omitted; internal quotation marks omitted.) Clinton v. Aspinwall, 352 Conn. 597, 608, 338 A.3d 1103 (2025). Moreover, because the ambiguity of a text is a question of law, the trial court typically must perform the threshold analysis of whether a contractual provision is plain and unambigu- ous. See, e.g., id., 619–20. The plaintiff contends that these principles of contract construction apply with equal force in the administrative context. It argues that the trial court was incorrect when it deferred to PURA’s rate-making discretion without first interpreting the settlement agreement and deter- mining whether the relevant language is ambiguous.  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority PURA, relying primarily on Connecticut Light & Power Co. v. Dept. of Public Utility Control, 219 Conn. 51, 591 A.2d 1231 (1991), counters that the intention of the parties to a settlement agreement presents a question of fact and that the trial court properly applied a deferential standard of review, inquiring only as to whether PURA’s interpretation was supported by substantial evidence. At the outset, we agree with the plaintiff that the trial court, through its analysis, did not interpret the settlement agreement but instead appeared to defer to PURA’s rate-making discretion. Although the court set forth the parties’ competing positions regarding the meaning of the capital cost recovery provisions of the agreement, it did not resolve their dispute over the contract language. Instead, it concluded that “PURA has the discretion under § 16-19e to resolve this matter as it has done.” We disagree that this is the correct analysis of contractual disputes of this nature. PURA’s rate-making discretion under § 16-19e is constrained by binding settle- ment agreements it has approved with EDCs such as the plaintiff. For example, PURA may not set rates that are contrary to a clear and unambiguous contract. See Southeastern Connecticut Regional Resources Recovery Authority v. Dept. of Public Utility Control, 244 Conn. 280, 290–96, 709 A.2d 549 (1998) (Southeastern Con- necticut) (predecessor agency to PURA improperly set rate that was contrary to clear and unambiguous terms of controlling electricity purchase agreement). Thus, the proper analysis must begin with an examination of the settlement agreement’s terms. In Southeastern Connecticut, we discussed the judicial standard of review of an agency’s interpretation of a con- tractual agreement setting rates. We held that the first step, as in any contract dispute, is to determine whether the contract is clear and unambiguous. “Although ordi- narily the question of contract interpretation, being a question of the parties’ intent, is a question of fact . . . [w]here there is definitive contract language, the determi- nation of what the parties intended by their contractual commitments is a question of law.” (Internal quotation marks omitted.) Id., 290. We further noted that, “[i]n  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority construing the meaning of the parties’ agreement, we are guided by well established rules. A contract must be construed to effectuate the intent of the parties, which is determined from the language used interpreted in the light of the situation of the parties and the circum- stances connected with the transaction. . . . [T]he intent of the parties is to be ascertained by a fair and reasonable construction of the written words and . . . the language used must be accorded its common, natural, and ordinary meaning and usage where it can be sensibly applied to the subject matter of the contract. . . . Where the language of the contract is clear and unambiguous, the contract is to be given effect according to its terms. A court will not torture words to import ambiguity where the ordinary meaning leaves no room for ambiguity . . . . Similarly, any ambiguity in a contract must emanate from the lan- guage used in the contract rather than from one party’s subjective perception of the terms.” (Emphasis omitted; internal quotation marks omitted.) Id., 291.6 Although the trial court did not make the initial deter- mination of whether the settlement agreement is clear 6 In support of the trial court’s decision, PURA cites to Connecticut Light & Power Co. v. Dept. of Public Utility Control, supra, 219 Conn. 66–67, for the proposition that “[t]he interpretation of the intention of the parties to a settlement agreement is a question of fact, and such determinations by an administrative agency are reviewed to determine if [they are] supported by substantial evidence.” We do not understand that language in Connecticut Light & Power Co. to be in conflict with our later decision in Southeastern Connecticut or with the general rule that clear and unambiguous contractual language is to be construed de novo by the court, as a matter of law, and given effect according to its terms. In Connecticut Light & Power Co., both parties had acknowledged that no provision of the agreement spoke directly to the issue at bar. See, e.g., Connecticut Light & Power Co. v. Dept. of Public Utility Control, Conn. Supreme Court Records & Briefs, January Term, 1991, Pt. 4, Plaintiff’s Brief p. 10. More important, Connecticut Light & Power Co. was not a contract case. The settlement agreement at issue was simply one of several factors that each party thought supported its position in a rate dispute; the plaintiff did not allege breach of the agreement or seek to enforce its terms. See Connecticut Light & Power Co. v. Dept. of Public Utility Control, supra, 64–66. Accordingly, we read that decision to stand for nothing more than the general principle that deference is afforded to administrative agencies in rate setting matters.  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority and unambiguous, which is the foundation of the plain- tiff’s claim on appeal, because the record is adequate for us to make that assessment in the first instance, and for purposes of judicial economy, we turn next to that question.7 III The following additional principles of law guide our analysis of the settlement agreement. “A contract is unambiguous when its language is clear and conveys a definite and precise intent. . . . In contrast, a contract is ambiguous if the intent of the parties is not clear and certain from the language of the contract itself. . . . The contract must be viewed in its entirety, with each provi- sion read in light of the other provisions . . . and every provision must be given effect if it is possible to do so. . . . If the language of the contract is susceptible to more than one reasonable interpretation, the contract is ambigu- ous.” (Internal quotation marks omitted.) Simpson v. Simpson, 352 Conn. 81, 95, 335 A.3d 472 (2025). “When only one interpretation of a contract is possible, the court need not look outside the four corners of the contract. . . . Extrinsic evidence is always admissible, however, to explain an ambiguity appearing in the instrument.” (Internal quotation marks omitted.) Id. As we discussed in part II of this opinion, the court’s “threshold determination” of whether contractual lan- guage is clear and unambiguous must be made “in light of the context of the situation” of the parties. Isham v. Isham, 292 Conn. 170, 181, 972 A.2d 228 (2009). That context may shed light, for example, on whether 7 The fact that each party contends that the language of the settlement agreement unambiguously favors its preferred interpretation does not preclude us from concluding that the language is nevertheless ambigu- ous, which may necessitate supplementing the record with extrinsic evidence. See, e.g., Simpson v. Simpson, 352 Conn. 81, 100–101, 335 A.3d 472 (2025). Indeed, the existence of more than one plausible interpretation is the very hallmark of ambiguity. See id., 95. But see id. (“the mere fact that the parties advance different interpretations of the language in question does not necessitate a conclusion that the language is ambiguous” (internal quotation marks omitted)).  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority contractual language that the parties have not defined has a specialized, technical or trade meaning, or is oth- erwise a term of art. See, e.g., id., 182–83; Tallmadge Bros., Inc. v. Iroquois Gas Transmission System, L.P., 252 Conn. 479, 499–500, 746 A.2d 1277 (2000); Spen- cer v. Higgins, 22 Conn. 521, 526–27 (1853); see also Southbury Land Trust, Inc. v. Andricovich, 59 Conn. App. 785, 789, 757 A.2d 1263 (2000) (“[a]lthough the words in a restrictive covenant are to be interpreted in their ordinary and popular sense, if any of the words have acquired a particular or special meaning in the particular relationship in which they appear, such particular or special meaning will control”). “In addition, documents that are part of the same transaction or subject matter will be construed together in determining the intent underlying the contracts. . . . Moreover, [as] long as two or more instruments are part of the same transaction, different execution times will not prohibit [the] instruments from being construed together.” (Citations omitted; internal quotation marks omitted.) Gold v. Rowland, 325 Conn. 146, 160–61, 156 A.3d 477 (2017); see also Schubert v. Ivey, 158 Conn. 583, 587, 264 A.2d 562 (1969) (“[i]n considering the expressed intent of a contract evidenced . . . by multiple writings, all of the writings should be considered and an endeavor made to ascertain the expressed intent of the contract as a whole”); 2 Restatement (Second), Contracts § 202 (2), p. 86 (1981) (“all writings that are part of the same transaction are interpreted together”); 5 M. Knif- fin, Corbin on Contracts (Rev. Ed. 1998) § 24.21, p. 216 (when terms of agreement are expressed in separate docu- ments, “[the] documents should be interpreted together, each one assisting in determining the meaning intended to be expressed by the others”); 11 R. Lord, Williston on Contracts (4th Ed. 1999) § 30:26, p. 239 (“the principle that all writings [that] are part of the same transaction are interpreted together also [applies when] incorpora- tion by reference [to] another document may be inferred  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority from the context in which the documents in question were executed”). Applying these principles to the present case, we look not just at the terms of the settlement agreement at the time it was executed, but also at the parties’ commu- nications to PURA before it approved the agreement, regarding the meaning of its terms. We do so because the agreement did not become final and binding on the par- ties until PURA approved it. Thus, any communications from the parties to PURA during the approval process had the potential to resolve an otherwise ambiguous term of the agreement by clarifying the intentions of the parties. If PURA then approved the agreement based on this clarified intent, the terms of the agreement would be unambiguous despite the original ambiguity. By way of example, if the undefined phrase “emergent equipment failures,” as it appears in the agreement, has two reason- able interpretations and is therefore ambiguous, but the parties’ communications to PURA prior to approval of the agreement resolved the ambiguity, then the mean- ing of that phrase, at the time it became binding on the parties, would be rendered unambiguous. Thus, we start, but do not end, with the language of the settlement agreement as executed by the parties before it was approved by PURA. Each party contends that its own respective construction of the agreement is based on the plain and unambiguous language therein. We disagree and conclude that the agreement is ambigu- ous as it relates to the recovery of capital spending on storm recovery. The settlement agreement itself does not expressly address the question of how the plaintiff was to recoup its capital spending on storm recovery. The agreement references the plaintiff’s “core capital program” but makes no mention of storm recovery in that context. PURA contends that the ordinary course of dealing had been for the EDCs to recoup their capital spending on unforeseeable events, such as catastrophic storms,  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority in the company’s next rate case, and that nothing in the settlement agreement purports to change that long- standing practice. The plaintiff challenges PURA’s char- acterization of the historical practice and argues that, in any event, the agreement established a new approach and unequivocally entitles it to recover capital spending arising from the five storms in its 2018–2020 base rates. The plaintiff’s argument that the settlement agree- ment is unambiguous in this respect proceeds in four parts. First, § 3 of the agreement permits the plain- tiff to recover up to $270 million in unspecified plant additions for “core capital projects” (plus additional reliability spending through the ESI tracker) in its cur- rent base rates. Second, a footnote to the agreement indicates that core capital programs8 are “identifie[d]” by the prefiled testimony of Kenneth B. Bowes, the plaintiff’s vice president of transmission performance. Third, Bowes’ testimony includes a chart (chart KBB-3) with row headings identifying “[e]mergent [e]quipment [f]ailures” as one subset of “[b]asic [b]usiness,” which in turn is one subset of “[c]ore [c]apital [p]rogram.” Fourth, the plaintiff contends that it is clear (for reasons to be discussed) that capital equipment damaged during a catastrophic storm qualifies as an emergent equipment failure. PURA disputes only the fourth proposition, as well as the plaintiff’s conclusion that the incorporation of this chart into the agreement necessarily means that the plaintiff could recoup its capital spending on recov- ery from the five storms in its 2018–2020 base rates. For the following reasons, we are not persuaded that, as a matter of law, this syllogism unambiguously describes how capital spending on storm recovery was to be recouped under the settlement agreement. First, storms and storm recovery are not mentioned in the relevant section of the agreement or in chart KBB-3, and it is not clear that storm damage necessarily falls under the heading of emergent equipment failures. Sec- ond, regardless of how storm damage is classified, it is 8 PURA does not dispute the plaintiff’s contention that the terms “core capital program” and “core capital project” are used interchangeably in the settlement agreement.  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority not clear how capital expenditures that substantially exceeded the plaintiff’s 2017 budget projections were to be handled under the agreement. Third, it is unclear whether the relevant documents drew a distinction between catastrophic and other storms for purposes of capital spending recovery. We discuss each point in turn. A The plaintiff’s argument hinges in significant part on the theory that the heading “[e]mergent [e]quipment [f]ailures” in chart KBB-3 in Bowes’ prefiled testimony, which was incorporated by reference into the settlement agreement, necessarily encompasses catastrophic storm damage. In support of that theory, the plaintiff offers another chart, one that Jeffrey D. Turano, the plaintiff’s manager of budgets and investment planning, presented in June, 2021, in response to a PURA interrogatory. The Turano chart identifies various subcategories that the plaintiff includes under the heading of “[e]mergent [e]quipment [f]ailures,” presumably for purposes of its internal accounting, one of which is labeled “major storms.” Setting aside for the moment the question of whether the five storms did in fact qualify as “major storms” for purposes of the plaintiff’s accounting system; see part III C of this opinion; we note that the Turano chart does little to advance the plaintiff’s argument. The ques- tion is what the parties to the settlement agreement mutually intended in early 2018 and, potentially, what PURA understood when it approved that agreement.9 The Turano chart is evidence of the plaintiff’s own understanding more than three years later, after the present dispute already had arisen. Absent evidence that the chart is reflective of the parties’ practices when the agreement was executed, it reveals little about the mutual understanding of the various parties regarding 9 The parties have not briefed the question of whether, and in what way, the meaning of a contract is altered by the understanding of the regulatory body that must approve the contract. We leave that question to be addressed in the first instance by the Superior Court on remand.  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority the construction of the contractual language in 2018. Consequently, the Turano chart, standing alone, does not inform our analysis of whether the meaning of “emergent equipment failures,” as used in the agreement, is clear and unambiguous. We turn, therefore, to the ordinary meaning of the con- tractual language—particularly the word “emergent.”10 Contemporary dictionaries indicate that the word “emer- gent” has at least two potentially relevant and quite dis- tinct meanings. Consistent with the plaintiff’s theory, “emergent” can refer to something that emerges suddenly and requires immediate attention, such as storm or other unplanned for emergency damage. See, e.g., Merriam- Webster’s Collegiate Dictionary (12th Ed. 2026) p. 534 (definitions of “emergent” include “arising unexpectedly” and “calling for prompt action”). But “emergent” need not connote suddenness or emergency, and, indeed, the word can refer to something that comes into being gradu- ally. See, e.g., The American Heritage Dictionary of the English Language (5th Ed. 2011) p. 583 (definitions of “emergent” include “[c]oming into view, existence, or notice . . . [such as] an emergent political leader,” and “[o]ccurring as a consequence . . . [such as] economic prob- lems emergent from the restriction of credit” (emphasis omitted)); Random House Webster’s College Dictionary (2001) p. 402 (definitions of “emergent” include “coming into existence . . . [such as] an emergent nation,” and “characterized by evolutionary emergence” (emphasis omitted)). This second meaning is used in legal, political, scientific, and industrial settings, among others.11 We 10 There is no dispute that the capital spending at issue was to repair or replace failed equipment. The question is simply whether storm related failures qualify as “emergent.” 11 See, e.g., E. Mertz, “Jane Larson’s Sociological Jurisprudence,” 28 Wis. J.L. Gender & Society 111, 118 (2013) (“[t]he remedies themselves were conceptualized as gradual and emergent in an evolving situa- tion”); M. Perlin & H. Cucolo, “ ‘See This Empty Cage Now Corrode’: The International Human Rights and Comparative Law Implications of Sexually Violent Predator Laws,” 23 New Crim. L. Rev. 388, 430 n.272 (2020) (“ ‘[d]esistance’ is a gradual or emergent process . . . through which people cease and refrain from persistent offending”);  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority thus conclude that the settlement agreement is facially ambiguous in this respect. We next turn to whether the parties’ communications to PURA before it approved the settlement agreement resolved the ambiguity. The manner in which the par- ties used the term “emergent” when the agreement was approved, including in the plaintiff’s own submissions to and testimony before PURA, offers support for both parties’ positions and, therefore, fails to resolve this ambiguity. On the one hand, there is evidence that the plaintiff was using the same classifications as in the Turano chart—defining major storms as a subset of emer- gent equipment failures—during the 2017 rate case. On January 26, 2018, just two weeks after the agreement was signed and well before PURA approved it, Shelnitz submitted a spreadsheet in response to a PURA inter- rogatory. The spreadsheet labeled “major storms” as one subcategory of “total emergent equipment failures.” On the other hand, other submissions the plaintiff made to PURA in support of approval of the settlement agreement suggest the opposite. For example, in the weeks leading up to the signing of the agreement, the plaintiff appeared to distinguish storm damage from emergent equipment failures, implying that the latter category is limited to equipment that wears out gradually over time. In late 2017, in response to an interrogatory from the OCC, Shelnitz submitted a different spreadsheet that provided detailed breakouts of the plaintiff’s capi- tal spending forecasts by project (2017 spreadsheet). In that spreadsheet, one general category is labeled simply as “equipment failures.” The spreadsheet then lists sev- eral subcategories of equipment failures, one of which is “emergent equipment failures.” Other subcategories include “inspection failed equipment,” “insurance claim,” see also U.S. Nuclear Regulatory Commission, Docket No. 05000247, Report No. 2001-002 (January 16 to February 9, 2001) p. iii (citing “feedwater pump oscillations” and “[a] heater drain pump flow element leak” as examples of “emergent equipment failures,” despite finding of overall acceptable system performance), available at https://www.nrc. gov/reading-rm/doc-collections/congress-docs/correspondence/2006/ clinton-05-31-2006.pdf (last visited May 10, 2026).  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority and, notably, “major storms.”12 Although the conclusion is not compelled, this additional consideration could be read to reflect the understanding that equipment can fail (or failed equipment can be identified) in various contexts, and that storm damage and emergent failure are two different modalities of failure. Although the 2017 spreadsheet was prepared before the settlement agreement was signed, at a public hearing that PURA held in February, 2018, to review the agreement, Bowes testified that “[w]hat is in the settlement is actually the plant additions in [the 2017 spreadsheet].” We also question whether the Bowes prefiled testimony that is incorporated in the settlement agreement undis- putably supports the plaintiff’s position. At one point, Bowes walked PURA through the major elements of the plaintiff’s core capital program. During that discus- sion, he made no reference to catastrophic storm dam- age, which was addressed in a different chapter of his testimony, and each of the elements of core capital that he discussed fits into a narrower definition of emergent failures as limited to aging and deteriorating equip- ment. And, as PURA points out, at several points, Bowes stated that core capital programs are those that address “day-to-day” system requirements. That portion of his testimony is consistent with PURA’s theory that rare and unpredictable catastrophic storm damage is distinct from run-of-the-mill emergent equipment failures. In the course of his prefiled testimony, Bowes also appeared to distinguish age related equipment failures from weather damage. He explained that the “majority” of the plaintiff’s “overhead [distribution] equipment is [more than fifty] years old. Due to its age, this equip- ment requires increasing amounts of maintenance, and it is becoming increasingly susceptible to outages caused by equipment failures and weather events.”13 (Emphasis added.) 12 Capital spending (specifically, “plant additions”) for major storms was forecast to be approximately $1.1 million per year between 2018 and 2020. 13 We note that footnote 5 of the settlement agreement, which addresses core capital spending, provides in relevant part that “[§] IV of the . . .  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority Thus, the PURA approval process does not resolve the ambiguity in the phrase “emergent equipment failures.” We therefore consider whether the extrinsic evidence available to us so clearly points to one interpretation of the phrase that we can resolve the ambiguity as a matter of law. See, e.g., Serby v. First Alert, Inc., 664 Fed. Appx. 105, 108 (2d Cir. 2016) (“[w]hen faced with ambiguous contractual language, an appellate court has two options: it may resolve the ambiguity in the contract language as a matter of law if there is no extrinsic evidence to support one party’s interpretation of the ambiguous language or if the extrinsic evidence is so one-sided that no reason- able [fact finder] could decide contrary to one party’s interpretation, or, it may remand for the [trial] court to consider and weigh extrinsic evidence to determine what the parties intended” (internal quotation marks omitted)).14 We conclude that it does not. First, the parties have not referred us to anything in the drafting history of § 3 of the settlement agreement that provides any clarity on this issue. Nor have they offered any evidence or authority as to whether “emer- gent equipment failures” is an established term of art in the electric distribution business. Second, the parties’ course of conduct prior to the execution of the settlement agreement is not illuminat- ing. Although PURA argues that the parties could not have intended to alter their long-standing practice of the plaintiff’s waiting until the next rate case to recoup capital expenditures incurred due to storm damage, it acknowledges that the agreement provided the plaintiff with an unprecedented opportunity to recover capital [prefiled] [t]estimony of . . . Bowes identifies ‘core’ capital programs.” By contrast, footnote 6, which addresses reliability spending, provides in relevant part that “[Bowes’ prefiled] [t]estimony at [chart] KBB-3 . . . describes the core capital projects in the system ‘[r]eliability’ cat- egory of [the plaintiff’s] core capital program.” The fact that footnote 5 references Bowes’ testimony in general, whereas footnote 6 expressly references chart KBB-3, raises further questions as to whether the row headings in that chart were intended to resolve the storm damage issue. 14 In the present case, we engage in this further review only because it is an administrative appeal with an extensive record on which both parties relied in support of their suggested constructions of the settle- ment agreement.  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority costs more quickly. The agreement is not clear on the limits, if any, on this new recovery process or whether storm cost recovery was meant to be part of it. Third, a review of the parties’ subsequent course of dealing; see, e.g., 2 Restatement (Second), supra, § 202 (5), pp. 86–87; also fails to definitively resolve these ambiguities. Although the plaintiff generally has taken the same position that it does in the present litigation, in the proceedings giving rise to the present appeal, the plaintiff submitted testimony that distinguished emergent equipment failures from catastrophic storm damage: “In summary, the variances between the 2016 [c]apital [p]lan [f]orecast of [c]ore [c]apital additions and [the] 2018 actual cost of [c]ore [c]apital additions was due to unanticipated catastrophic storms ($30 mil- lion); a substantial unanticipated increase in the volume of pole attachment applications ($7.6 million); and an unanticipated increase in emergent equipment failures ($10.3 million).” At the least, then, there is evidence that the plaintiff was inconsistent in its own use of the term “emergent equipment failures” and whether that is an umbrella term that encompasses storm damage or, alternatively, a different subset of equipment failures. To summarize, PURA’s position in this litigation is that a mere row heading in a chart, obliquely referenced in a footnote to a contract, simply cannot suffice to dem- onstrate a clear, mutual intent to alter the parties’ well established course of dealing. Regardless of whether that is true as a general matter, in the present case, we are not persuaded that the “emergent equipment failures” row heading in chart KBB-3 clearly and unambiguously evidences the intent of the parties in early 2018 with respect to catastrophic storm damage. Given the various possible dictionary meanings of the word “emergent,” we conclude that the settlement agreement is facially ambiguous as to how capital spending on storm recovery was to be treated. The available administrative record fails to clearly resolve that ambiguity, particularly in light of the plaintiff’s own inconsistent references as to whether storm damage and emergent equipment failures  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority represent distinct categories of capital spending for purposes of the plaintiff’s internal accounting. B Second, even if storm damage did qualify as a type of emergent equipment failure, the settlement agreement is ambiguous as to how cost overruns and unbudgeted expenses are to be handled. This is important because, in its original 2017 rate application, the plaintiff projected that it would invest just over $1 million annually in capital spending on major storm recovery. See footnote 12 of this opinion. In fact, the plaintiff spent approxi- mately $45 million during the 2018–2020 period, more than thirteen times the amount that it had budgeted and PURA had preapproved. Once again, the settlement agreement, on its face, is susceptible to two plausible interpretations. Section 3 (a) (ii) (1) of the agreement (approval provision) pro- vides: “Core Capital. The [plaintiff’s] proposed capital expenditures for ‘core’ capital projects for calendar years 2018, 2019 and 2020, as filed in the [a]pplication, are approved.” (Emphasis added; footnote omitted.) Section 3 (a) (ii) (1) (a) of the agreement (recovery provision) then further specifies that “[p]lant [a]dditions for core capital projects of $270 million for calendar years 2018–2020 shall be recovered in base distribution rates. If [p]lant [a]dditions for core capital projects are greater than $270 million per year for calendar years 2018–2020, then the [plaintiff] shall recover ‘reliability’ [p]lant [a]dditions (which are a subset of core capital projects) for that time period through the [n]ew [c]apital [t]racker up to the amount in excess of $270 million.” (Footnote omitted.) PURA, relying on the emphasized language in the approval provision, contends that the capital expendi- tures that it preapproved for inclusion in the 2018–2020 base rates were only those that the plaintiff specifically included in the proposed budget, or spending forecasts, that it filed with its 2017 rate case application. Other spending—even within the categories covered by the  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority settlement agreement—would have to be approved ret- roactively, in a future rate case or proceeding, as had been the historical practice. The logic behind this posi- tion is that PURA could not have departed so far from historical practice as to preapprove spending of which it was not yet aware. The plaintiff, by contrast, relies on the recovery pro- vision, which both permits recovery of unspecified core capital spending up to $270 million per year and provides a mechanism for the plaintiff to recover spending above that cap. There is a logic to the plaintiff’s position as well: what would be the point of a flexible spending cap and the ESI tracker if the plaintiff could immediately recover only the amount of capital spending that it had specifically forecasted? The settlement agreement seems to necessarily envision that there would be unexpected costs and cost overruns.15 A conflict between contractual provisions creates an ambiguity. See, e.g., Thoma v. Oxford Performance Materials, Inc., 153 Conn. App. 50, 60, 100 A.3d 917 (2014). As was the case with the emergent equipment failures language, consideration of the circumstances surrounding approval of the settlement agreement fails to resolve and, indeed, only serves to heighten this ambi- guity. The following additional procedural history is relevant. In February, 2018, PURA conducted a series of hear- ings to help it evaluate the proposed settlement agree- ment. One set of issues that PURA explored during those hearings was how § 3 of the agreement was to function with respect to unanticipated costs in general and storm recovery costs in particular. Several aspects of the 15 This is consistent with the position that the plaintiff took in a late filed exhibit, filed between the signing of the settlement agreement and PURA’s approval thereof, stating that “[t]he proposed new capi- tal tracker was intended to recover costs associated with . . . excess core capital project costs that exceed the annual caps described in the [s]ettlement [a]greement.”  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority plaintiff’s (and the OCC’s) responses to PURA’s ques- tions during those hearings are especially noteworthy. It was undisputed during the hearings that PURA’s preapproval of capital spending was limited to the amounts that the plaintiff presented in the 2017 spreadsheet, and that spending substantially in excess of those budgeted amounts would require PURA’s fur- ther approval. Bowes testified that, “in my prefil[ed] testimony and in some of the interrogatories . . . we had quite a bit of detail around the specific programs. [It] is our intention to execute those programs as we filed. . . . I believe that we filed by program and intend to execute by program as we entered into the case.” Bowes later clarified that the numbers incorporated into the base rates were those presented in the 2017 spreadsheet. For his part, Shelnitz expressed agreement with the OCC’s opinion that the annual reconciliation process was not intended to relitigate the entire capital budget but that, “if there are [spending] categories where it’s . . . 100 percent above or . . . significantly above what was forecasted . . . that might raise some concern from PURA or from [the] OCC . . . .” PURA then sought clarification as to how substantial cost overruns would be handled under § 3 of the settlement agreement, given the apparently conflicting language.16 Representatives of the plaintiff sought to assure PURA that (1) such overruns would be handled consistent with historical practice, and (2) PURA would have adequate opportunity to review such overruns. But the plaintiff’s representatives did not provide details as to what sort of review they envisioned: would significant, unbudgeted 16 For example, Stephen Capozzi, a member of the PURA staff, expressed his confusion in a series of questions: “So, for both new system resiliency and core capital programs, we touched on this a little bit, but . . . how will we handle them in the [new] capital tracker proceedings for any repri- oritization that comes up over the years, especially if it seems that . . . those proceedings may just be . . . more of a cost true-up than anything else . . . ?” “I’m just trying to get a sense of . . . the language that says . . . that the capital programs are approved as filed; I’m just trying to understand how that’s going to work in the new capital tracker program and how much flexibility we’ll have to address those. The language seems somewhat opposing each other . . . .”  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority expenditures be recoverable annually through the ESI tracker, subject only to the routine prudence review, or would they have to be specially approved in the next rate case, as per historical practice?17 The same is true of the OCC and the PRO; their representatives sought to reas- sure PURA that it would be able to review future cost overruns, but they offered little in the way of details.18 These various discussions were summarized in a joint brief filed prior to PURA’s final approval of the amended settlement agreement.19 In that brief, the plaintiff, the 17 Shelnitz appeared to take the former position. He testified that “the settlement recognized that there would be a change in the plant addition numbers and provided a mechanism for truing those up. So, from the perspective of the settlement parties, that was cared for through the settlement language.” Bowes’ position is less clear: “I would say it would be the same process used in historical proceedings prior to this where we indicate in a capital plan what we plan to do. We then go ahead and execute and report . . . the results of that. So [in] the tracker program, or the base capital program, we’re going to use a similar process, and it’s really just how those costs are recovered, whether from [the] dis- tribution rate or the tracker rate.” 18 For example, Richard Sobolewski, an OCC representative, testified: “[T]he track[er] will give the participants, PURA [and the] OCC, at least two opportunities annually to review the [capital spending] details . . . before and . . . after they’re incurred. And . . . I can guarantee that [the] OCC will be keeping an eye on that in those filings. As things come up . . . if there’s a large variance or a large change in the totals, I can guarantee that [the] OCC will be asking questions and asking PURA for hearings. . . . We believe the oversight and review opportunities protect ratepayers and regulators [from] any [run-ups] in [capital spending] in future years . . . and it clearly does not create a blank check for runaway spending.” Steven Cadwallader of the PRO likewise expressed his opinion that, “even though it is a soft amount in the sense that if the [plaintiff] spends more than [$270 million], [it has] an opportunity to recover that. [The set- tlement agreement] gives [PURA] plenty of oversight opportunities. . . . So there’s lots of opportunity for oversight, and PURA can determine what’s appropriate and what’s not. And, if it thinks that something is not appropriate in the [plaintiff’s] expected expenditures . . . [PURA] can say [that it does not] think it’s appropriate and give the [plaintiff] direction at that point.” Cadwallader further expressed his doubts that PURA’s review of storm costs could be handled through the new capital tracker mechanism and suggested that future storm costs would be handled in the next rate case. 19 The section of PURA’s April 18, 2018 decision that incorporates the settlement agreement by reference also expressly refers to the joint brief and the various hearings and related exhibits discussed herein.  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority OCC, and the PRO agreed that the base rate would include only preapproved capital spending that the plaintiff had forecast and filed with PURA, and that any spending in excess of those forecasts would be subject to “further review . . . .” The parties wrote: “During the hearing, PURA staff inquired whether the utilization of a [n]ew [c]apital [t]racker provided the [plaintiff] with authoriza- tion to spend unlimited future amounts on core capital and system resiliency programs. The [parties] explained that the [s]ettlement [a]greement . . . approved [only] the forecasted core capital and system resiliency capital expenditures that were filed in the record in this case. Additionally, [representatives of the] OCC and [the] PRO both testified that their position is that this approach imposes appropriate limits and oversight on the [plain- tiff’s] future spending. Any amounts spent in excess of the forecasted expenditures in the record in this case are subject to further review by PURA.” (Footnotes omit- ted.) The parties to the agreement further agreed that, consistent with historical practice, each November, the plaintiff would provide PURA with its budgeted core capital spending for the upcoming calendar year and, each March, it would report to PURA and identify any variances of more than 10 percent from the budgeted amount in the prior year’s capital spending, in any ini- tiative or category, or in total, and explain the reason for that variance. But, again, the joint brief failed to clearly answer the questions that PURA raised dur- ing the hearings, namely, what was the legal status of any spending that varied by more than 10 percent from the forecast levels, and did “further review” by PURA encompass (1) a traditional, formal review in a future base rate proceeding, (2) a prudence review of the sort that PURA can conduct with respect to any preapproved spending under the settlement agreement, or (3) some other unspecified, discretionary review. We thus conclude that the settlement agreement is ambiguous as to how capital spending that significantly exceeded the projected amounts filed with PURA in the 2017 spreadsheet, or in subsequent annual projections, was to be handled, and that a review of the administrative  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority record in the 2017 rate case does not clearly resolve that ambiguity. C A third source of ambiguity relates to whether dam- age from catastrophic storms in particular falls within the ambit of the core capital spending that the settle- ment agreement preapproved for recovery in base rates. PURA contends that, even if storm damage qualifies as a subset of emergent equipment failures for purposes of the agreement, only routine, noncatastrophic storm recovery costs are built into the plaintiff’s base rates. The plaintiff counters that the agreement and the supporting spending projections encompass all capital spending on “major storm” recovery, drawing no distinction between catastrophic storms and other major storms, and, in any event, that PURA’s argument is unpreserved because it was not the basis for the agency’s decision. The plaintiff’s interpretation of the settlement agree- ment is certainly a reasonable one. The agreement expressly distinguishes between catastrophic and non- catastrophic storms with respect to its treatment of incremental, operation and maintenance costs,20 but it draws no such distinction for capital spending purposes. Nevertheless, we cannot conclude that the plaintiff’s construction of the settlement agreement is the only reasonably plausible one. Specifically, we agree with PURA that, at times, the parties have used the term “major storms” as limited to noncatastrophic storms. For example, in the Bowes prefiled testimony that is incorporated by reference in the agreement, the KBB- 11 chart identifies and classifies the damaging storms that struck Connecticut in 2015 and 2016. The title of that chart, “Major Storms in 2015 and 2016,” supports the plaintiff’s interpretation by classifying all of the storms, including catastrophic ones, as major. Within 20 Section 15 (d) of the settlement agreement, for instance, reduced the benchmark for accessing the catastrophic storms reserve to $4 million of incremental expenses per event, effectively redefining noncatastrophic storms as those requiring less than $4 million in incremental recovery costs. Amendment one also addresses catastrophic storm costs.  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority the chart, however, there is a column labeled “[s]torm [c]ategory.” In that column, each storm is identified as either “[p]re-[s]taging,” “[m]ajor,” or “[c]atastrophic,” consistent with PURA’s theory that the agreement treated major and catastrophic storms as distinct. Thus, the agreement as approved by PURA is ambiguous as to whether catastrophic storms fall within the category of core capital spending. And, once again, the extrinsic evidence available to us, including the parties’ subsequent course of conduct, does not definitively resolve the ambiguity. There is a tension, for example, in the positions that the plaintiff has taken in this litigation. As discussed, beginning with its rate application in 2017, and continuing through its annual true-up in 2021, the plaintiff has forecast that it would invest approximately $1.1 million annually in capital spending on “major storms” recovery. At the same time, however, the plaintiff’s representatives have stated that the plaintiff does not forecast capital spending on catastrophic storm recovery. In their prepared joint tes- timony in this matter, four of the plaintiff’s executives testified, with respect to the plaintiff’s core capital cost overruns, that the plaintiff “cannot reasonably budget for catastrophic storm events. Neither the original 2018 [c]ore [c]apital plan, nor the [plaintiff’s] capital plan for any other period, includes a forecasted budget for cata- strophic storms.” If the plaintiff budgets capital spend- ing for major storms, but not for catastrophic storms, then that distinction may imply (at least for this purpose) that the major storm costs allegedly included in the core capital program are limited to noncatastrophic storms.21 In addition, throughout the 2018 five storms proceed- ing, the plaintiff represented that it was not seeking PURA’s approval of capital spending associated with the five storms in that case because it intended to recoup those costs in a future rate case. In its initial brief in that case, the plaintiff stated that “storm-related capital costs 21 We agree with PURA that it is not foreclosed from raising the cata- strophic/noncatastrophic distinction on appeal, insofar as it is simply attempting to reconcile an apparent contradiction in the plaintiff’s representations in this litigation.  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority were appropriately identified by an after-the-fact analysis of the total costs charged to the storm work orders. . . . The plant assets identified in this manner were removed from the storm work orders to be added to the [plaintiff’s] rate base when new rates go into effect . . . .” (Citation omitted.) The plaintiff made similar representations in its reply brief: “Capital storm costs are excluded from this docket. . . . [T]his issue is not ripe for adjudication in this docket. Parties can elect to reraise this issue in the appropriate future docket when [the plaintiff] seeks to recover in rates its storm-related capital costs.” (Cita- tions omitted; emphasis in original.) Those representa- tions appear to contradict the plaintiff’s position in the current litigation, namely, that it did not need to seek approval or recovery of its five storms related capital spending in the storm case because, under the settlement agreement, it already could recover those costs in its 2018–2020 base rates.22 Given these inconsistencies, we conclude that the administrative record fails to resolve this third source of ambiguity in the agreement.23 IV Our conclusions that (1) the trial court first should have interpreted the settlement agreement and determined whether it is clear and unambiguous rather than defer- ring to PURA’s rate-making discretion, (2) the agree- ment is, in fact, facially ambiguous in various respects, and (3) neither related contemporaneous documents 22 We note that PURA’s stance regarding the 2018 five storms proceed- ing also appears to have shifted over time. In its decision in that case, PURA—relying on the plaintiff’s representations—indicated that it intended to address capital costs for the five storms in the plaintiff’s next rate case. Subsequently, in its decision in the present matter, PURA took the position that it already had approved some, but not all, of the plaintiff’s storm related capital spending in that case. 23 Because these three aspects of the settlement agreement and the circumstances surrounding its adoption are more than sufficient to establish ambiguity as to the parties’ intent, we need not address a fourth potential source of ambiguity on which PURA relies, namely, the language in amendment one providing that the plaintiff could seek to have the prudence of all post-2016 storm costs evaluated in a future storm docket.  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority associated with PURA’s approval of the agreement nor the available extrinsic evidence of the parties’ course of dealing definitively resolves those ambiguities, bring us full circle to the question of what deference, if any, is to be afforded to PURA’s determinations in this matter. At the very least, our precedents suggest that we should afford due deference to a neutral administrative agency in the interpretation of an ambiguous contract with respect to any relevant factual determinations made by the agency. See, e.g., Tomlinson v. Board of Educa- tion, 226 Conn. 704, 723–28, 629 A.2d 333 (1993). The plaintiff, by contrast, argues that we should not defer to PURA’s factual findings because PURA is not a neutral arbiter. Rather, its prosecutorial arm, the PRO, was a party to the settlement agreement. The plaintiff argues that, as a general matter, it would be unfair to allow an agency to make self-serving factual findings to sup- port a post hoc reinterpretation of a contract to which it is a party. It also argues that, in this particular case, deferring to PURA’s findings as to the parties’ intent in entering into the agreement would strongly dissuade regulated entities from entering into such agreements, contrary to the public policy of Connecticut. Finally, the plaintiff notes that, although PURA has alluded to historical rate setting practices, it has neither introduced nor evaluated any extrinsic evidence of the parties’ intent with respect to this particular contract. The parties have not specifically briefed the question of what deference is due to an agency’s factual find- ings regarding the meaning of an ambiguous contract under these circumstances, and we are not aware of any controlling Connecticut authority.24 Given the various complexities involved, we conclude that prudence mili- tates against adopting a general rule at this juncture, 24 Neither of the cases on which the parties respectively rely is directly on point. See Southeastern Connecticut Regional Resources Recovery Authority v. Dept. of Public Utility Control, supra, 244 Conn. 290 (affording no deference because contractual language was not ambigu- ous); see also footnote 6 of this opinion (discussing Connecticut Light & Power Co. v. Dept. of Public Utility Control, supra, 219 Conn. 51).  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority particularly when the trial court’s resolution of the mat- ter on remand may render the issue moot. The complexi- ties that motivate this conclusion include the following. First, although the PRO was a party to the settlement agreement, PURA itself was not. It is unclear, therefore, to what extent the relationship between these two enti- ties would warrant any exception to the general rule of deference with respect to an agency’s factual findings. Second, we understand the prevailing federal rule to be as follows: an administrative agency’s interpretation of an ambiguous contract to which it is a party deserves deference when (1) the agency’s expertise will be of some value in construing the contract, and (2) the agency is serving as a neutral arbiter in its regulatory role as a guardian of the public interest such that neither the agency itself nor the government has an interest in the outcome.25 In this context, an interest typically refers to a financial interest, although courts also have been reluctant to defer when an agency’s interpretation of a contract would result in expanding its jurisdiction or otherwise accreting power to itself. See footnote 25 of this opinion. The federal courts are far from uniform in their adoption and application of that rule, however, and, given its origins in the Chevron doctrine; see Chevron, U.S.A., Inc. v. Natural Resources Defense Council, Inc., 467 U.S. 837, 104 S. Ct. 2778, 81 L. Ed. 2d 694 (1984); see also footnote 25 of this opinion; we have some ques- tion as to the ongoing vitality of the rule in light of the United States Supreme Court’s recent repudiation of that doctrine. See Loper Bright Enterprises v. Raimondo, 25 See, e.g., Southern California Edison Co. v. United States, 226 F.3d 1349, 1357–58 (Fed. Cir. 2000); Muratore v. Office of Personnel Man- agement, 222 F.3d 918, 922–23 (11th Cir. 2000); Cemex, Inc. v. Dept. of the Interior, 560 F. Supp. 3d 268, 280–81 (D.D.C. 2021); Bottoms Farm Partnership v. Dept. of Agriculture, Docket No. 4:15-CV-1073- SPM, 2017 WL 1105124, *7–9 (E.D. Mo. March 24, 2017), aff’d sub nom. Bottoms Farm Partnership v. Perdue, 895 F.3d 1070 (8th Cir. 2018); see also, e.g., Guilford Transportation Industries v. Public Utilities Commission, 746 A.2d 910, 912–14 (Me. 2000); T. Armstrong, “Chevron Deference and Agency Self-Interest,” 13 Cornell J.L. & Pub. Policy 203, 213–30 (2004).  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority 603 U.S. 369, 411–13, 144 S. Ct. 2244, 219 L. Ed. 2d 832 (2024) (overruling Chevron, U.S.A., Inc.). Third, to the extent that these questions come down to legislative policy preferences with respect to settle- ment agreements, there is some reason to be skeptical of the plaintiff’s assertion that its position aligns with the current preferences of the legislature.26 But, again, these questions have not had the benefit of full brief- ing. On remand, to the extent that the outcome of the case hinges on the degree of deference to be afforded to PURA’s factual findings as to the intentions of the parties in entering into the settlement agreement, the trial court will have the opportunity to consider them in the first instance. To summarize, the administrative record in this mat- ter and the related PURA dockets is voluminous, and our independent review has only scratched the surface thereof. In addition, the parties have not fully briefed the significance of all of the relevant documents and 26 Following, and partly in response to, the events that gave rise to the present appeal, the legislature qualified its previous endorsement of settlements as an alternative to contested rate cases. In 2023, the legislature removed the language from § 16-19jj encouraging the use of settlements to resolve contested cases. See Public Acts 2023, No. 23-102, § 4 (P.A. 23-102). At that time, legislators also (1) added a requirement that any settlements be consistent with the principles set forth in §§ 16-19 and 16-19e, (2) authorized PURA to retrospectively reject or modify provisions of signed settlement agreements under certain circumstances, and (3) clarified that a settlement agreement does not function as a general rate case for purposes of postponing the periodic rate review required under § 16-19a. See P.A. 23-102, § 4. The legislative history of P.A. 23-102 consistently indicates that, although legislators remained favorably disposed to the settlement of rate disputes, in light of the present matter, they wanted to ensure that formal rate cases be held on a regular basis and that PURA not get locked into settlement agreements that committed the agency and ratepayers to unintended negative consequences, such as was the case with the ESI tracker in the 2018 settlement agreement. See, e.g., 66 S. Proc., Pt. 4, 2023 Sess., pp. 2606–2607, remarks of Senator Norman Needleman; id., pp. 2609–10, remarks of Senator Ryan Fazio; 66 H.R. Proc., Pt. 10, 2023 Sess., pp. 7283–84, 7330–31, 7371–72, 7390, 7430–34, remarks of Representative Jonathan Steinberg; id., pp. 7407–7408, remarks of Representative Charles J. Ferraro.  Connecticut Light & Power Co. v. Public Utilities Regulatory Authority procedural history discussed herein; nor have they had an opportunity to address the questions of (1) whether fact finding is necessary to resolve the various ambigui- ties in the settlement agreement, and (2) what degree of deference (if any) is owed to PURA’s factual findings regarding the construction of an ambiguous settlement agreement that it approved and to which its prosecuto- rial arm is a party. Accordingly, given our conclusions that the settlement agreement is ambiguous in various respects and that the portion of the administrative record that we have reviewed does not clearly and definitively resolve those ambiguities, we conclude that the most prudent course is to remand the case to the trial court to resolve the contractual ambiguities in light of the complete administrative record and/or to remand the matter to PURA for consideration of all of the evidence relevant to resolving those ambiguities. The judgment is reversed and the case is remanded to the trial court for further proceedings according to law. In this opinion the other justices concurred.