JUSTICE RICE

delivered the Opinion of the Court.

¶1 Appellants Northwestern Corporation, doing business as Northwestern Energy (Northwestern), the Natural Resources Defense Council (NRDC), and Human Resources Council, District XI (HRC), appeal the decision of the Second Judicial District Court affirming the Final Order of the Montana-Public Service Commission (Commission), which disallowed $1,419,427 in claimed excess electric regulation costs and adjusted energy efficiency savings calculations. We affirm, considering the following issues:

1. Did the Commission apply the correct legal standard in reviewing NorthWestern’s claim for excess outage costs?

2. Were the “free ridership” and “spillover”calculations adopted by the Commission supported by substantial evidence?

FACTUAL AND PROCEDURAL BACKGROUND

¶2 This matter involves a challenge to the Commission’s Final Order in NorthWestern’s 2011-2012 annual tracker filing.1 Therein, Northwestern requested, inter alia, a $1,419,427 increase in rates for *35unexpected electricity supply costs due to an outage at its Dave Gates Generating Station (DGGS), located near Anaconda.2 As part of the proceeding, the Commission also ordered Northwestern to present evidence for purposes of conducting a “true-up” to actual costs for lost revenues that had been previously estimated in NorthWestern’s demand-side management (DSM) programs. Ultimately, the Commission (1) denied NorthWestern’s request to include the DGGS outage costs in customer rates, and (2) rejected NorthWestern’s expert’s conclusion that the “free ridership” and “spillover” values of its DSM programs were perfectly offsetting, adopting instead the same expert’s actual calculations used in a draft report.

DGGS Outage Costs

¶3 In 2008, Northwestern sought Commission approval to build the DGGS. The DGGS was intended to provide regulation and frequency response service in NorthWestern’s service area. The Commission approved the project in 2009, and the DGGS commenced commercial operation on January 1, 2011.

¶4 The DGGS was a first-of-its-kind facility that Northwestern presented as having “the potential to be a model facility for the supply of regulation service.” It consisted of three generation units made by Pratt & Whitney Power Systems, Inc. (PWPS) and was an application of a simple cycle natural gas turbine generator designed to increase or decrease generation (ramp) in response to variations in NorthWestern’s load, “on a moment-by-moment basis.” NorthWestem’s General Manager of Generation testified that the plant had a “very unique” control mechanism and “early on we knew that the plant was going to have a very unique control application.”

¶5 Northwestern was aware that the ramp capabilities of the DGGS were critical to its operation and that the DGGS was a first-of-its-kind application, stating:

[The DGGS] is one of the first power plant installations to be built specifically for electrical transmission grid regulation duty. The design requirements for grid regulation are stringent since they require the plant to continually change load in a short time frame (seconds to minutes).

This load requirement was necessary because Northwestern “anticipated variable operating conditions,” largely due to wind generation variations, and the DGGS needed to be able to ramp up or *36down by at least 15 mega-watts (MW) per minute per unit to “offset the continuous variation between system generation and system load.”

¶6 The contract between Northwestern and PWPS included a waiver of consequential damages, but Northwestern purchased, with customer revenue, an extended warranty to cover the innovative technology. Northwestern did not purchase or evaluate the feasibility of outage insurance in case the DGGS had an operational failure.

¶7 On January 31, 2012, thirteen months after Northwestern brought the DGGS online, it suffered a complete outage. Unit cycling had caused “thermal stresses” by going from a cold state to a very high temperature, damaging the rotating equipment. PWPS concluded the outage resulted from ramp rates “much greater” than anticipated, excessive temperatures, and cycle-related hardware failures. The Commission was unable to precisely examine the ramp data because Northwestern failed to maintain minute-by-minute records.

¶8 Pursuant to the extended warranty, PWPS repaired the damaged turbines at its cost, including removal, installation, and shipping costs. However, due to the waiver of consequential damages in the contract, PWPS was not obligated to cover the costs associated with purchasing replacement regulation service during the outage. On February 3, 2012, Northwestern began purchasing replacement service from Powerex Corp. (Powerex) and Avista Corp. (Avista). PWPS took “extraordinary measures” to repair the DGGS as soon as possible. Individual generators were put back online as PWPS restored them and Northwestern proportionally decreased its regulation service purchases accordingly. The DGGS was fully back online on May 1, 2012.

¶9 During the outage, Northwestern customers continued to pay the fixed costs for the operation of the DGGS ($6,742,625), including NorthWestern’s usual rate of return, as well as the variable costs ($1,527,714) Northwestern did not actually incur, but would have incurred had the plant been operational. However, the outage caused Northwestern to incur an additional $1,419,427 in charges to Powerex and Avista for regulation service. Northwestern requested reimbursement of these costs, arguing they were reasonably incurred because it obtained an extended warranty that covered all repairs, it purchased regulation service on the competitive market at 2011 rates, it structured its regulation market purchases to enable it to incrementally reduce the purchases as generators were repaired, and it had worked quickly to get the DGGS back online.

¶10 The Montana Consumer Counsel (MCC) opposed reimbursement *37of the replacement service costs, contending that Northwestern failed to undertake risk mitigation by failing to investigate whether outage insurance was available. The MCC offered the testimony of Dr. John Wilson:

No. I don’t fault the company for not procuring it [outage insurance]. What I think was imprudent was not looking into it, not evaluating it, not finding out whether it was available and what the cost would be for a plan like this. I think you have to do that before you make a determination as to whether you acquire it or not.

The MCC argued that evaluation of insurance was fundamental to risk management where the contract contained an exclusion for consequential damages:

[T]he most imprudent thing that occurred here, is the failure of the company to take steps to protect itself against the outage, given the fact that they had this exclusion under the warranty, given the fact that they knew... that there were unknowns about this plant and where it was going to go and how it was going to operate.

¶11 Northwestern responded by providing evidence that in its experience it had never purchased replacement power insurance and, instead, always relied on the market for replacement power. NorthWestern’s General Manager of Generation testified that after receiving inquiry from the Commission and the MCC regarding insurance, he “went and solicited input from other utilities ... [a]nd they indicated that they simply do not get outage insurance because it is not economical to do so.” Northwestern put on evidence that outage insurance could be $1 million per year, thus potentially costing more than the replacement power itself, but acknowledged it did not “investigate or purchase insurance that might have covered the additional electricity supply costs.”

¶12 The Commission inquired into NorthWestern’s operation of the DGGS through data requests and found that Northwestern was aware the units needed to change load quickly, that quick response was critical, and that the units could experience unique thermal stresses due to ramping up and down. The outage was directly tied to “ramp rates ‘much greater’ than anticipated, excessive temperatures and cycle-related hardware failures,” yet Northwestern used software allowing excessive ramping and did not retain precise ramp rate data.

¶13 The Commission determined that NorthWestern’s management of the DGGS was not reasonable and that the excess regulation costs *38were not prudently incurred because Northwestern (1) failed to prudently manage risks; and (2) did not “exhibit the level of situational awareness that the Commission would expect from a utility managing a one-of-its-kind power plant.” The Commission reasoned:

Given the warranty’s exclusion of consequential damages and the uniqueness of DGGS, Northwestern should have identified the risk of incurring replacement costs in the event of an outage. ... [NorthWestern’s] failure to identify risk ensured that incremental costs of replacement service would be incurred in the event of an outage.

The Commission found that outage insurance was available and, even though it may not have been cost-effective, because Northwestern failed to “evaluate the availability, price and terms of outage insurance,” it “guaranteed that any incremental replacement costs would be unavoidable in the event of an outage.” Citing both NorthWestern’s failure to manage risk and reasonably operate the DGGS, the Commission denied NorthWestern’s request to include the outage costs in customer rates.

DSM Program

¶14 Fixed costs are those the utility will incur regardless of how much energy it actually sells to consumers. Utilities typically recover fixed costs through volume based charges built into customer rates. Consequently, there is no financial incentive for a utility to encourage energy efficiency because decreases in consumption would hamper the utility’s recovery of its fixed costs. A lost revenue adjustment mechanism (LRAM) is designed to compensate a utility for the revenue lost due to the utility’s energy efficiency efforts. In essence, it allows the utility to estimate and recover the revenue it lost due to energy efficiency efforts directly attributable to the utility, such as by DSM programs.

¶15 In 2005, the Commission approved the use of a LRAM to account for revenue losses incurred as a result of NorthWestern’s energy efficiency efforts, finding that “the lost revenue disincentive is real and puts at risk a full and complete ramp-up of cost-effective energy efficiency resource acquisition programs in the near-term.” It authorized Northwestern to include in rates an estimate of the income lost due to DSM programs with a requirement that, after the programs had been implemented, the “estimated lost... revenue amount must be trued-up based on actual program activity in [the given years] and again following a comprehensive program evaluation and independent verification of actual savings.” This “true-up” ensures that *39Northwestern is only including in rates the revenue lost from its DSM programs, and not from independent causes.

¶16 Analysis of a DSM program includes examination of “free ridership” and “spillover.” Free ridership occurs when a consumer takes advantage of a program incentive to install an energy efficient device, but would have installed the device with or without the incentive. As such, the utility did not effectuate the customer’s usage reduction and is not entitled to recover the associated lost revenue. On the other side of the ledger, spillover occurs when a consumer does not respond to a DSM program incentive, but later chooses energy efficient products or practices as a result of the utility’s general advocacy. As such, the utility is credited with the energy reduction it only indirectly induced, and can include those lost revenues in its LRAM.

¶17 Northwestern selected Nexant Energy Management Group (Nexant) to evaluate its DSM programs for its first true-up process in 2006-2007. Nexant measured free ridership and spillover and included them in its final assessment. The Commission adopted the Nexant assessment, concluding that it “satisfies the DSM program evaluation and savings verification requirements” the Commission had established.

¶ 18 The next true-up of NorthWestern’s DSM programs was presented in the subject proceeding. Northwestern hired SBW Consulting (SBW), who partnered with Research into Action (RIA), to conduct the required independent, comprehensive true-up for the periods 2006-2007 to 2010-2011. In its draft report to Northwestern, SBW included the values for free ridership and spillover it had calculated. The draft report concluded that Northwestern was responsible for 79% of the energy efficiency savings it had estimated and included in customer rates through the LRAM.

¶19 However, in its final report, SBW came to the conclusion that the values calculated for free ridership and spillover should not be used in the assessment of NorthWestern’s DSM programs. The final report assumed that the two values, since they work in contradiction to each other, offset each other equally. In statistical terms, this offset was considered a 1.0 net-to-gross (NTG), meaning the net is no different than the gross savings. By completely offsetting spillover and free ridership values, SBW’s final report concluded that Northwestern was responsible for 87% of the energy efficiency savings it had previously *40estimated in the LRAM.3 Northwestern agreed that this difference in over-collected revenues ought to be refunded to Northwestern ratepayers.

¶20 During her testimony before the Commission, Dr. Marjorie McRae (Dr. McRae), the RIA researcher responsible for free ridership and spillover calculations, explained that when she met with Northwestern to discuss the draft report, she had informed Northwestern that she believed “we are not able, as a profession, to measure these accurately, and that the effects are offsetting.” Dr. McRae testified that Northwestern had advised her to revise the draft “according to [her] professional opinion.” Thus, the final report utilized a 1.0 NTG value for comparison between the two values instead of using the actual values derived from the research. Dr. McRae affirmed that she had conducted the free ridership and spillover research using “national common practices, and best practices,” and the actual data was “comparable to those found for similar programs conducted by other respected program evaluators.” However, the SBW final report stated that there were problems with using the calculated free ridership and spillover calculations:

[T]he economic analysis [should] use the value 1.0 for the net-to-gross ratio ... [due to] known limitations to standard practices for the estimation of free ridership and spillover estimation — limitations that confound their effects and result in the overestimation of free ridership and the underestimation of spillover — and on current net-to-gross practices in 31 jurisdictions with active energy efficiency programs, many of which recognize that free ridership and spillover are offsetting phenomena.4

¶21 Dr. McRae concluded that researchers cannot truly ascertain free ridership and spillover values, and opined the Commission should use *41a 1.0 NTG ratio that treats the numbers as if they perfectly offset each other. To support her conclusion, Dr. McRae cited various studies, one notably finding that thirteen regulatory jurisdictions used a 1.0 NTG, while two jurisdictions, Michigan and New York, used a 0.9 NTG.

¶22 Under cross-examination, Dr. McRae admitted she cannot know what the actual values are due to the state of the science. “I would say that’s [(measuring spillover and free ridership)] not possible with any methods that I know to know what they are.” In response to questions from Commissioner Kavulla about whether there was data to support her conclusion that free ridership and spillover perfectly offset in a 1.0 NTG relationship, Dr. McRae admitted:

If you take 1.0 as the null hypothesis that these effects are offsetting, then, I think the burden is — especially if you’re going to be in a lost revenue calculation or something like that, I think the burden of proof is to say, no, these aren’t offsetting. These savings would have happened anyway.... I don’t think we have a way of saying that the null hypothesis is rejected, that it’s anything other than what 1.0. And if you want to say for argument’s sake it’s [0].9, well, then for argument’s sake why don’t we say it’s 1.1.

Commissioner Kavulla’s asked: “why is the 1.0 rather than a [0].09 or a 1.1 the null hypothesis?” Dr. McRae concluded: “I think in the absence of any other information, you just assume that one is positive and one is a negative; they’re offsetting. That’s how I think of it.”

¶23 The Commission rejected Dr. McRae’s conclusion that free ridership and spillover perfectly offset each other in a 1.0 NTG ratio and, instead, adopted the values she provided in her draft report. The Commission held that “[a]lthough free ridership and spillover may be difficult to estimate, the remedy is not to discard the only empirical data that attempts to ascertain those values.” The Commission disagreed with Dr. McRae’s conclusion that offsetting meant equal offsetting:

Offsetting does not imply perfectly offsetting, and Northwestern has not demonstrated that an NTG of 1.0 is more reasonable as a null hypothesis than an NTG of 0.9 or any other fixed relation of the effects of free ridership and spillover. Because SBW did not test the null hypothesis proposed by [Dr.] McRae, it cannot be supported.

Noting the Commission’s duty to “approve an accurate level of savings and associated lost revenues,” the Commission reasoned that Dr. McRae’s conclusions were problematic because they forced the *42Commission to assume both that: (1) a fixed ratio (1.0 NTG) between free ridership and spillover was more accurate than actual measured numbers; and (2) 1.0 NTG was a better assumption than any other fixed value, for example, 0.9 NTG. Using the data from the draft report indicating a 0.908 NTG correlation between free ridership and spillover, the Commission lowered NorthWestern’s true-up realization rate from 87% to 79%.

Procedural History

¶24 Northwestern appealed the Commission’s order on both issues to the Montana Second Judicial District Court, Silver Bow County. The District Court affirmed the Commission’s Final Order. Northwestern, NRDC, and HRC appeal.

STANDARD OF REVIEW

¶25 In an administrative appeal, we apply the same standards of review that the district court applies. Whitehall Wind, LLC v. Mont. Pub. Serv. Comm’n, 2015 MT 119, ¶ 8, 379 Mont. 119, 347 P.3d 1273 (Whitehall Wind II); Molnar v. Fox, 2013 MT 132, ¶ 17, 370 Mont. 238, 301 P.3d 824. Administrative appeals are governed by § 2-4-704, MCA. “A district court reviews an administrative decision in a contested case to determine whether the agency’s findings of fact are clearly erroneous and whether its interpretation of the law is correct.” Whitehall Wind, LLC v. Mont. PSC, 2010 MT 2, ¶ 15, 355 Mont. 15, 223 P.3d 907 (Whitehall Wind I); accord Molnar, ¶ 17 (conclusions of law are reviewed de novo). Judicial review of a final agency decision “must be confined to the record.” Section 2-4-704(1), MCA; Molnar, ¶ 17.

¶26 “The court may not substitute its judgment for that of the agency as to the weight of the evidence on questions of fact.” Section 2-4-704(2), MCA; accord Whitehall Wind II, ¶ 7. “A finding of fact is clearly erroneous if it is not supported by substantial evidence in the record, if the fact-finder misapprehended the effect of the evidence, or if a review of the record leaves the court with a definite and firm conviction that a mistake has been made.” Williamson v. Mont. PSC, 2012 MT 32, ¶ 25, 364 Mont. 128, 272 P.3d 71. “In reviewing findings of fact, the question is not whether there is evidence to support different findings, but whether competent substantial evidence supports the findings actually made.” Mayer v. Bd. of Psychologists, 2014 MT 85, ¶ 27, 374 Mont. 364, 321 P.3d 819. The court may reverse or modify the agency decision if the “substantial rights” of the appellant were prejudiced because the administrative findings are “in excess of the statutory authority of the agency,” “affected by error of other law,” “clearly *43erroneous in view of the reliable, probative, and substantial evidence on the whole record,” or “arbitrary or capricious or characterized by abuse of discretion or clearly unwarranted exercise of discretion.” Section 2-4-704(2)(ii), (iv), (v), (vi), MCA.

¶27 “Except as otherwise provided by statute relating directly to an agency, agencies shall be bound by common law and statutory rules of evidence.” Section 2-4-612(2), MCA. “The agency’s experience, technical competence, and specialized knowledge may be utilized in the evaluation of evidence.” Section 2-4-612(7), MCA. “Substantial evidence is evidence that a reasonable mind could accept as adequate to support a conclusion; evidence beyond a scintilla.” Mayer, ¶ 27 (internal quotations omitted). “Moreover, the court should give deference to an agency’s evaluation of evidence insofar as the agency utilized its experience, technical competence, and specialized knowledge in making that evaluation.” Knowles v. State ex rel. Lindeen, 2009 MT 415, ¶ 21, 353 Mont. 507, 222 P.3d 595 (citing § 2-4-612(7), MCA; Johansen v. Dept. of Natural Res. and Conservation, 1998 MT 51, ¶ 29, 288 Mont. 39, 955 P.2d 653).

DISCUSSION

¶28 1. Did the Commission use the correct legal standard in reviewing NorthWestern’s claim for excess outage costs'?

¶29 Northwestern argues that the Commission used the incorrect legal standard when reviewing the outage costs associated with purchasing replacement regulation service during the DGGS outage. Northwestern contends that “prudently incurred electricity supply costs,” § 69-8-210(1), MCA, is an objective, reasonable person standard, which in the context of utilities, is a “reasonable utility standard.” Northwestern notes that other jurisdictions consider such costs under a reasonable utility standard. Under this standard, Northwestern argues that “prudently incurred” costs are those that a reasonable utility in NorthWestern’s similar situation would have incurred, and argues that it acted as any other reasonable utility would have in the same situation.

¶30 The Commission argues that “prudent” must be interpreted in light of the statutes and Commission rules referenced by the statute. The Commission does not dispute that the reasonable utility test is one factor to be considered, but argues that it is not the complete definition of “prudent.” The Commission offers that it reviewed NorthWestern’s actions to determine whether the electricity supply costs were prudent pursuant to § 69-8-210(1), MCA, whether the assets purchased and *44owned by Northwestern were managed reasonably under §§ 69-8-419 and -421, MCA, and whether rates that included the outage charges would be excessive or confiscatory pursuant to § 69-3-201, MCA. The Commission argues it applied the appropriate review and, under the facts in this case, made an appropriate determination that the costs were not prudently incurred because the plant was not reasonably managed, and that any rates that included those costs would not be reasonable.

¶31 At issue in this case is the meaning of the word “prudent” in § 69-8-210(1), MCA, which, as the parties note, is not defined by the Legislature. Section 69-8-210(1), MCA, reads in full:

The commission shall establish an electricity cost recovery mechanism that allows a public utility to fully recover prudently incurred electricity supply costs, subject to the provisions of 69-8-419,69-8-420, and commission rules. The commission may include other utility costs and expenses in the cost recovery mechanism if it determines that including additional costs and expenses is reasonable and in the public interest. The cost recovery mechanism must provide for prospective rate adjustments for cost differences resulting from cost changes, load changes, and the time value of money on the differences.

¶32 Section 69-8-210(1), MCA reflects the full authority the Legislature granted to the Commission to review electricity supply costs. The Commission is an administrative agency created by statute. Section 69-1-102, MCA; Schuster v. Northwestern Energy Co., 2013 MT 364, ¶ 9, 373 Mont. 54, 314 P.3d 650. The Commission does not have judicial powers, Schuster, ¶ 9, Williamson, ¶ 31, and its jurisdiction is “limited to the regulation of rates and service as provided by the Montana statutes.” Billings v. Pub. Serv. Comm’n, 193 Mont. 358, 370, 631 P.2d 1295, 1303 (1981); accord Great N. Utils. Co. v. Pub. Serv. Comm’n, 88 Mont. 180, 203, 293 P. 294, 298 (1930) (“[T]he Commission is a creature of, owes its being to, and is clothed with such powers as are clearly conferred upon it by the statute.”); Mont. Power Co. v. Pub. Serv. Comm’n, 206 Mont. 359, 371, 671 P.2d 604, 611 (1983). As we noted in the cases following the deregulation of the Montana electrical industry, see, e.g., Mont. Power Co. v Mont. PSC, 2001 MT 102, ¶ 46, 305 Mont. 260, 26 P.3d 91 (“[W]e observe that the Commission is statutorily charged with applying and enforcing the [deregulation] Act.”), the Commission was specifically charged with carrying out the statutes in question: “[t]he commission shall establish an electricity cost recovery mechanism.” Section 69-8-210(1), MCA (emphasis added). *45As such, the statute in question clearly confers authority on the Commission for this purpose.

¶33 The meaning of “prudent” is largely self-evident. “Absent statutory definitions, the plain meaning of the words used controls.” City of Great Falls v. Mont. Dept. of Pub. Serv. Regulation, 2011 MT 144, ¶ 18, 361 Mont. 69, 254 P.3d 595; accord Williamson, ¶ 36. The word has been applied in prior Commission decisions, which have used such terms for “prudent” as “marked by wisdom or judiciousness” or “circumspect or judicious in one’s dealings” and its synonyms are “ ‘careful,’ cautious,’ ‘sensible,’ ‘practical,’ ‘discreet,’ ‘wise,’ and ‘farsighted.’ ” In re Mont. Power Co., Mont. Pub. Serv. Comm’n, Docket D2001.10.144, Order No. 6382d 12 (June 21, 2002) (internal citations omitted). The Montana Legislature gave the Commission express latitude to determine if the given costs were prudent — careful, sensible, practical, discreet, wise, or farsighted or, more apt in the regulatory environment, avoiding unnecessary risks — through its own fact finding and administrative authority. Further, this analysis is undertaken in light of the statutory requirement that “prudently incurred electricity supply costs” must be determined “subject to the provisions of 69-8-419, 69-8-420, and commission rules.” Section 69-8-210(1), MCA.5

¶34 Section 69-8-419, MCA, governs the utility’s duties for building and maintaining its “electricity supply resource” portfolio, including contracts for power generation or capacity, electricity plants owned or leased by the utility, customer load management, or any other means of providing reliable and adequate electricity service to customers. Section 69-8-103(9), MCA (defining “electricity supply resource”). The provision requires utilities to “plan for future electricity supply resource needs; manage a portfolio of electricity supply resources; and procure new electricity supply resources when needed.” Section 69-8-419, MCA. The utility is required to conduct this planning in accordance with, inter alia, the following objectives: (1) “provide adequate and reliable electricity supply service at the lowest long-term total cost”; and (2) “identify and cost-effectively manage and mitigate risks related to its obligation to provide electricity supply service.” Section 69-8-419(2)(a), (c), MCA. Thus, the utility must plan for future needs, manage its portfolio, and procure resources when necessary at the lowest long-term cost and, when doing so, identify and mitigate *46risks related to those obligations.

¶35 Commission administrative rules also address prudent utility resource procurement. “Prudent electricity supply resource planning and procurement includes evaluating, managing, and mitigating risks associated with the inherent uncertainty of wholesale electricity markets and customer load.” Admin. R. M. 38.5.8219(1) (2016) (emphasis added). The Commission has specifically identified sources of risk that, among others, may be evaluated: fuel prices and price volatility, environmental regulations and taxes, retail supply rates, supplier capabilities, construction costs, and contract terms and conditions. Admin. R. M. 38.5.8219(1) (emphasis added). The Commission’s rules require that the “utility’s strategy for managing and mitigating risks associated with the identified risk factors should be developed in the context of the goals and objectives of these guidelines and include an evaluation of relevant opportunity costs.” Admin. R. M. 38.5.8219(2). Finally, prudence involves documenting and carrying out the resource procurement plans:

The commission must allow a utility to recover all costs it prudently incurs to perform this function. Whether the costs a utility incurs are prudent is, in part, directly related to whether its resource procurement process was conducted prudently. It is vital that a utility document its portfolio planning, management and electricity supply resource procurement activities to justify the prudence of its resource procurement decisions.

Admin. R. M. 38.5.8220(2).

¶36 Considering these sources, we disagree with Northwestern that the “reasonable utility standard” — i.e., what would a reasonable utility do in similar circumstances — is the appropriate interpretation of “prudent” or the appropriate inquiry under Montana law. The Montana Legislature used the term “prudent,” not “reasonable utility,” to describe how the Commission was to review electricity supply costs. Adopting NorthWestern’s proposed standard would read a contradictory idea into the statute. If “prudent” was restricted to what a reasonable utility would do in similar circumstances, the Commission would be deprived of its own discretion to evaluate and determine whether the utility’s actions were prudent. Tying the outcome to evidence of what other utilities did or would do would remove or reduce the discretion of the Commission to rely on its own expertise.

¶37 In sum, § 69-8-210(1), MCA, grants authority to the Commission to determine whether energy supply costs were prudently incurred — i.e., the utility’s incurred costs were wise, judicious, or *47sought to avoid unnecessary risk — in light of the planning requirements set forth in § 69-8-419, MCA, § 69-8-421, MCA, and Commission rules, which specifically require risk analysis and mitigation, including an examination of the relevant contract terms. The Commission was correct to apply these standards.

¶38 The remainder of NorthWestern’s arguments challenging the Commission’s decision assumes that the reasonable utility standard governs the outcome. Having rejected that view, we need not address all of NorthWestern’s further arguments based thereon. In brief, and to the extent that the reasonable utility standard is an appropriate factor to consider, as the Commission did, the Commission’s determination was supported by the record. The DGGS was a “one-of-a-kind” plant and the purchase and installation contract contained a provision that excluded consequential damages. Waiver of consequential damages on a first-of-its-kind regulation plant without extensive industry use supported the Commission’s determination that NorthWestern’s failure “to identify risk ensured that incremental costs of replacement service would be incurred in the event of an outage,” and was imprudent. To defend its actions, Northwestern asked other utilities — after the MCC and the Commission inquired into its risk mitigation efforts — about their insurance practices and presented evidence that those utilities did not purchase it. However, this is risk justification, not risk management.

¶39 Even if it is accepted that insurance was cost-prohibitive and would not have been a viable alternative, the Commission also determined that Northwestern did not reasonably manage the DGGS and that the outage costs were also imprudent for that reason.6 Northwestern was aware that the DGGS had “very unique” controls and was different from other plants. Northwestern was also aware, as the Commission found:

(1) “[T]he units need[ed] to change load rapidly’ as measured in “MW change per minute,” and that a single engine in operation could “ramp up or down at a rate of at least 15 MW per minute”;

(2) “the ability to respond to demand within seconds” was critical to the operational mission of DGGS; and (3) the units could *48experience unique “thermal stresses,” and that going “from a cold start to a very high temperature” can cause “a lot of distress within rotating equipment.”

(Internal quotations in original.) The outage specifically resulted from these known factors. PWPS’s investigation concluded “[o]ver temperatures resulted in reduction of material properties,” “[hjigher motion resulted in higher stress on the affected parts,” and “hardware failures are cycle related.” Northwestern admitted the ramp rate was “much greater” than Northwestern had requested due to software configuration and Northwestern had not installed anything to monitor the actual ramp data on a per-minute basis. In addition, Northwestern cycled each unit frequently, which PWPS concluded was the cause of the hardware failures.

¶40 The Commission did not commit clear error in finding that Northwestern had failed to appropriately plan for and operate the DGGS. The Commission’s decision to disallow the outage costs incurred by Northwestern when the DGGS went offline was well within its authority to determine whether those costs were “prudently incurred.” Section 69-8-210(1), MCA. Accordingly, the Commission’s order regarding the outage costs is affirmed.

¶41 2. Were the “free ridership” and “spillover” calculations adopted by the Commission supported by substantial evidence?

¶42 NRDC and HRC argue that the Commission erred when it adopted the free ridership and spillover values presented in Dr. McRae’s draft report when she, as the only witness to testify on the subject, repudiated those very numbers in her testimony. This, they argue, was clearly erroneous because there is no evidence in the record supporting the use of those numbers.

¶43 Citing problems with the methodology, the SBW final report concluded that the actual calculations for free ridership and spillover should not be used. SBW concluded that the best approach was to assume the numbers perfectly offset each other. Dr. McRae echoed this conclusion in her testimony before the Commission.

¶44 However, NRDC and HRC are incorrect to argue that there was no testimony regarding actual free ridership and spillover calculations. When pressed on her conclusions, Dr. McRae hedged her testimony in several ways. First, Dr. McRae stated affirmatively that actual free ridership and spillover calculations were conducted using “national common practices, and best practices,” and that the actual data derived was “comparable to those found for similar programs conducted by other respected program evaluators.”

*49¶45 Second, Dr. McRae testified her opinion of the state of the science is that she simply cannot know what the actual values are, including the 1.0 NTG she suggested the Commission adopt. “I would say that’s not possible with any methods that I know to know what they [free ridership and spillover] are.” Regarding whether there was actual, hard data to support her conclusion for a 1.0 NTG, Dr. McRae testified there was no way to prove or disprove her conclusion:

If you take 1.0 as the null hypothesis that these effects are offsetting, then, I think the burden is — especially if you’re going to be in a lost revenue calculation or something like that, I think the burden of proof is to say, no, these aren’t offsetting. These savings would have happened anyway.... I don’t think we have a way of saying that the null hypothesis is rejected, that it’s anything other than what 1.0. And if you want to say for argument’s sake it’s [0J.9, well, then for argument’s sake why don’t we say it’s 1.1.

(Emphasis added.) When asked why 1.0 would be used instead of 0.9 or 1.1, Dr. McRae responded: “in the absence of any other information, you just assume one is positive and one is negative; they’re offsetting. That’s how I think of it.”

¶46 The Commission was faced with: (1) an expert’s conclusion that one cannot know the precise spillover and free ridership numbers; and (2) testimony stating they could neither prove nor disprove that given hypothesis. The same expert provided a range of hypothetical values from 0.9 to 1.1 and provided anecdotal evidence of other states using a 0.9, while some used 1.0. Finally, the expert admitted the only hard research available in the proceeding was done according to best practices and was comparable with that done by other respected researchers.

¶47 Our role is not to re-weigh the evidence, but rather, to determine if substantial evidence existed “and not whether, on the same evidence, [we] would have arrived at the same conclusion.” Johnson v. W. Transp., LLC, 2011 MT 13, ¶ 18, 359 Mont. 145, 247 P.3d 1094 (citing Ward v. Johnson, 242 Mont. 225, 228, 790 P.2d 483, 485 (1990)). We hold the Commission’s facts were supported by substantial evidence. The actual data collected by Dr. McRae and SBW provided a 0.908 NTG, which falls in the range of hypothetical values provided by the expert. It is also in the range of values used by other commissions, as testified to by Dr. McRae. Dr. McRae admitted there was no actual, hard data to support her conclusion that the values perfectly offset each other. And, finally, the only hard data available was collected per best practices and was consistent with the research done by other *50respected firms.

¶48 As an administrative agency, the Commission’s “experience, technical competence, and specialized knowledge may be utilized in the evaluation of evidence.” Section 2-4-612(7), MCA. The Commission had substantial evidence to rely upon and it appropriately used is expertise to evaluate that evidence. As such, the Commission’s determination to adopt the calculated values for free ridership and spillover is affirmed.

¶49 For the foregoing reasons, the Commission’s Order No. 7219h is affirmed.

CHIEF JUSTICE McGRATH, JUSTICES SHEA, COTTER and WHEAT concur.

In re NorthWestern Energy's 2011-2012 Electricity Supply Tracker, Mont. Pub. Serv. Comm’n, Dkt. D2012.5.49, Order No. 7219h (Oct. 28, 2013).

The DGGS was formerly called the Mill Creek Generating Station.

Northwestern had projected 309,336 megawatt-hours (MWh) of total energy savings. In its final report, SBW was able to verify 270,564 MWh in savings.

Specifically, Dr. McRae opined that while the free ridership and spillover numbers were reliable (they consistently returned similar results from similar data sets), the numbers were not valid because researchers are unsure what the research was actually measuring. For free ridership, Dr. McRae stated various biases were the core of the problem, notably asymmetric perceptions of gains versus losses, attribution errors, cognitive dissonance, and the inability to accurately report events and predict participants’ behavior. For spillover, McRae noted difficulty identifying non-incentivized efficiency actions, estimating baseline energy usage, and showing a causal relation to an efficiency program.

Section 69-8-420, MCA, covers a utility’s utility procurement plan, which are not directly at issue in this proceeding.

Section 69-8-421(9), MCA, allows the Commission to “disallow rate recovery for the costs that result from the failure of a public utility to reasonably manage, dispatch, operate, maintain, or administer electricity supply resources in a manner consistent with 69-3-201, 69-8-419, and commission rules.”