Missouri Public Service Commission v. Union Electric Company, D/B/A Ameren Missouri

MISSOURI COURT OF APPEALS WESTERN DISTRICT

MISSOURI PUBLIC SERVICE ) COMMISSION, ) WD79406 ) Respondent, ) OPINION FILED: v. ) ) December 6, 2016 UNION ELECTRIC COMPANY, ) D/B/A AMEREN MISSOURI, ) ) Appellant. )

APPEAL FROM THE PUBLIC SERVICE COMMISSION

Before Division One: Anthony Rex Gabbert, P.J., Thomas H. Newton, and Alok Ahuja, JJ.

Union Electric Co. d/b/a Ameren Missouri (Ameren or Ameren Missouri) appeals a Missouri Public Service Commission order granting Staff’s motion for summary determination and denying Ameren’s cross-motion for summary determination. The ruling arose from a Staff complaint alleging that Ameren had violated the Commission’s Missouri Energy Efficiency Investment Act (MEEIA) 1 rules by using stale “avoided costs” data to calculate Ameren’s performance-incentive award under an approved demand-side program investment mechanism (DSIM). At issue is whether the Commission’s interpretation of the term “methodology” as used in 4 C.S.R.

1 Section 393.1075. All statutory references are to RSMo Cum. Supp. 2010, unless otherwise indicated.

§ 240-20.093(1) (F) 2 was reasonable. Ameren contends that “methodology” simply encompasses the formula used to calculate its avoided costs. The Commission concluded that “methodology” encompasses both the formula and a prescribed data set used in the formula to make that calculation. Also at issue is whether it was reasonable for the Commission to subject Ameren’s DSIM performance-incentive award to that regulation. We affirm.

Factual and Procedural Background

1. Missouri’s Energy-Conservation Statute

The Missouri Legislature enacted MEEIA in 2009 to encourage utilities to conduct programs that “modify the net consumption of electricity on the retail consumer’s side of the electric meter.” § 393.1075.2(3). Recogni zing that lowered energy consumption would affect a utility’s revenue and profits, the Legislature stated, “It shall be the policy of the state to value demand-side investments equal to traditional investments in supply and delivery infrastructure and allow recovery of all reasonable and prudent costs of delivering cost-effective demand-side programs.” § 393.1075.3. As part of its plan for MEEIA’s implementation and to give utilities incentives to adopt demand-side programs, the Legislature allowed the Commission to “develop cost recovery mechanisms to further encourage investments in demand-side programs including, in combination and without limitation: . . . allowing the utility to retain a portion of the net benefits of a demand-side program for its shareholders.” § 393.1075.5. Following notice-and-comment rulemaking, the Commission finalized its MEEIA implementing regulations in February 2011, and this Court upheld the

2 M O . C ODE R EGS . A NN . tit. 4, §240-20.093(1)(F) (2011).

Commission’s four orders of rulemaking in State ex rel. Pub. Counsel v. Pub. Serv. Comm'n, 397 S.W.3d 441 (Mo. App. W.D. 2013), including the rule at issue here. 3 Under the Commission’s rules, a utility’s DSIM may include the “(4) [r]ecovery of lost revenues; and (5) [u]tility incentive[s] based on the achieved performance level of approved demand-side programs.” 4 C.S.R. § 240-20.093(1)(M). How the latter were to be calculated under Ameren’s approved demand-side program is at issue in this litigation.

2. Dispute Overview

As discussed in more detail below, while the Commission’s rulemaking was on appeal before this Court, Ameren’s plan to adopt demand-side programs for its customers was approved by the Commission as modified by the parties. Two documents from the approval process are key to understanding this dispute: Ameren’s DSIM plan, which would, among other matters, allow the utility to recover both lost revenues and a performance incentive, and the modification of that plan negotiated by the parties. The modification adopted the DSIM plan with certain changes, and the

3 Section 240-20.093 of 4 C.S.R., which involves the establishment and operation of DSIMs that “allow periodic rate adjustments related to recovery of costs and utility incentives for investments in demand - side programs,” states in relevant part,

Avoided cost or avoided utility cost means the cost savings obtained by substituting demand-side programs for existing and new supply-side resources. Avoided costs include avoided utility costs resulting from demand -side programs’ energy savings and demand savings associated with generation, transmission, and distribution facilities including avoided probable environmental compliance costs. The utility shall use the same methodology used in its most recently-adopted [sic] preferred resource plan to calculate its avoided costs.

4 C.S.R. § 240-20.093(1)(F). Thus, while customers pay less because they use less energy, a separate line item on their utility bills adds an amount —a DSIM rate—representing, among other matters, “‘[u]tility incentive[s] based on the achieved performance level of approved demand-side programs.’” 4 C.S.R. § 240-20.093(1)(M). See State ex rel. Pub. Counsel v. Pub. Serv. Comm’n, 397 S.W.3d 441, 446 (Mo. App. W.D. 2013). Utilities receive a portion of net -shared benefits, which consist of estimated avoided costs less the costs of running a demand-side program. 4 C.S.R. § 240-20.093(1)(C).

nature and extent of those changes are at issue here, along with how the regulations we upheld intersect with those changes.

Ameren submitted the same underlying data to its independent evaluators to calculate two different plan components, the TD-NSB (or throughput-disincentive net shared benefits, i.e., lost revenues) and the performance incentive, at the end of the three-year program cycle to “true up” the additional DSIM charges that ratepayers had been making under the programs. Commission Staff objected to the data used for the performance-incentive calculation, and the Commission agreed that, under the plan, the modification, and the regulations, Ameren was required to use different, more recent data to calculate the avoided costs underlying the performance incentive. While the actual dollar amount of the difference between a performance incentive calculated using “stale” and “updated” avoided-cost data has not been specified on appeal, Staff counsel noted during argument before the Commission that the performance -incentive component of Ameren’s DSIM represented about 7 percent of the total that the utility could recover under MEEIA as an inducement to adopt demand-side programs, assuming that the plan achieved 100 percent of its energy-saving goals.

“Avoided costs” include those investments that utilities would have made under a traditional supply-side program, such as new energy generation and transmission facilities and the costs of complying with environmental regulations to build these facilities. Shareholders receive a return on those investments in the rates charged to customers, which rates also include the costs of the energy used. Increased capacity and increased demand result in both higher revenues and better investment returns. When customers adopt, under a demand-side program, conservation measures that *5 lower the revenues which utilities receive due to decreased demand, investment in additional supply-side facilities from which utility shareholders can earn a return are no longer required. Under MEEIA, as an incentive to encourage the adoption of conservation measures, utilities implementing approved demand -side programs can charge their customers an additional fee representing a share of both the revenues that are lost and the foregone investments in new generating capacity. Under the regulations, these foregone investments are calculated as part of the p erformanceincentive component of a DSIM and are reduced by the utilities’ costs of implementing the programs (these are recovered in other ways under a DSIM); they are also pegged to the numbers of conservation measures actually adopted and the amount of energy saved. At issue here is whether a measure of the energy saved for purposes of the performance-incentive award includes an updated estimate of actual energy costs over the life of a demand-side program. If so, when such costs rise, shareholders have the potential to recover more under a demand-side program, because customers have saved more money by decreasing energy usage; when they fall, the shareholder recovery will be decreased because customers have saved less money despite decreasing energy usage.

3. Ameren’s Plan and Modifications

Ameren filed a three-year demand-side program with the Commission in January 2012 (MEEIA 1 Plan), while the case challenging MEEIA’s implementing regulations *6 was pending on appeal. 4 The program was based on the utility’s “2013-2015 Energy Efficiency Plan” report (report), which was submitted with Ameren’s application and included a DSIM. Ameren’s MEEIA 1 Plan was not adopted as filed, but was modified under a July 2012 “Unanimous Stipulation and Agreement Resolving Ameren Missouri’s MEEIA Filing” (stipulation). The Commission approved the modified MEEIA 1 Plan (modified plan) in an August 2012 order.

The stipulation that the Commission approved states that the DSIM described in Ameren’s report was to be adopted “[s]ubject to the terms and conditions contained herein” and as “modified to reflect the terms and conditions herein.” Three lengthy stipulation paragraphs specifically address and modify Ameren’s recovery of TD -NSB, including how this component was to be calculated, and the utility’s performance incentive. Nothing in the plan as modified by the stipulation expressly indicates how the performance-incentive award is to be calculated other than that it was to be a percentage of certain net shared benefits set forth in an appendix. The regulations define a utility incentive in terms of annual net shared benefits. 4 C.S.R. § 240 -20.093(1)(EE). They further define “annual net shared benefits” as “the utility’s avoided costs measured and documented through evaluation, measurement, and verification (EM&V) reports for approved demand-side programs less the sum of the programs’ costs. . . .” 4 C.S.R. § 240-20.093(1)(C) (emphasis added). Avoided costs

4 Among the points brought by the utilities (Ameren and Kansas City Power & Light Co.) and the Office of Public Counsel were challenges to the Commission’s definition of “lost revenue” and whether it had the statutory authority to allow single -issue ratemaking and rate adjustments outside of general rate case proceedings. State ex rel. Pub. Counsel, 397 S.W.3d at 450, 454. We decided the case after the Commission approved the DSIM subject to the present appeal. Because we agreed that rate adjustments could be made outside of general rate case proceedings, Ameren was able to include adjustments for lost revenues and its pe rformance incentive by means of “riders” under its plan and the modification.

are defined as “the cost savings obtained by substituting dema nd-side programs for existing and new supply-side resources.” 4 C.S.R. § 240-20.093(1)(F). These cost savings include “avoided utility costs resulting from demand -side programs’ energy savings and demand savings associated with generation, transmission, and distribution facilities including avoidable environmental compliance costs.” Id.

Ameren’s MEEIA 1 Plan report contains an extensive discussion of the “Throughput Disincentive” (TD) that was approved as modified by the stipulation. According to Ameren’s report, “the throughput disincentive is about how the reduction in sales volumes impacts the revenues collected by the utility.” During argument before the Commission, Staff counsel stated that the TD-NSB was used in Ameren’s DSIM as a substitute for the “lost revenue” that may be recovered under MEEIA and that Staff did not challenge Ameren’s use of deemed (or “stale”) avoided-cost values in calculating TD-NSB. 5 TD-NSB is not a term used or defined anywhere in MEEIA or in its implementing regulations. And under the regulations, its apparent equivalent— lost revenue—is not defined in terms of net shared benefits or avoided costs.

The stipulation specifically provides that for purposes of determining Ameren’s final recovery for its TD-NSB, the software used in calculating NSB for the MEEIA 1 Plan would be re-run using only:

(i) the actual number of energy efficiency measures (by type) installed in each month up to that point, (ii) the actual program costs in each month incurred up to that point; and (iii) for Commercial and Industrial Custom measures for which the TRM [Technical Resource Manual] does not provide a deemed value, savings determined according to the protocol provided for at pages 85 to 98 of the TRM.

5 In fact, Ameren’s TD-NSB calculation used the same stale avoided costs that Staff challenged in the context of the utility’s performance incentive.

Further, the stipulation provided that “EM&V [Evaluation, Measurement, and Verification] shall not be utilized to calculate the actual NSB for the purposes of determining Ameren Missouri’s TD-NSB Share.” In contrast, when discussing the performance-incentive component, the stipulation does not state how it will be calculated but requires that “[a]ctual net energy savings for each program year will be determined through the EM&V.”

The TD-NSB part of the stipulation, including how it will be calculated, accords in many respects with a report table—Table 2.12, “Description of Update Process”— listing “the items associated with estimating net benefits and whether those items will be updated for purposes of assessing performance and benefits as part of the implementation process.” This table addresses “the mechanics of sharing net benefits” as part of implementing the DSIM’s “program expense tracker,” which under the MEEIA regulations is linked to the “cost recovery of demand-side program costs” and not to a utility’s incentive award. 4 C.S.R. § 240-20.093(1)(M)(2). Items on the list represent data that will be used to calculate whether the company makes progress on its goal of saving a specified number of kilowatt hours of energy “at the meter.” On this list is the item “Avoided Costs,” which has a red “X” instead of a green “√”in the “Update?” box beside the phrase, as well as the following description: “The avoided energy, capacity, and T&D [transmission & distribution] values are deemed.” Text above the table states, “Notice that several items will not be updated, so the focus remains on the cost of the programs and the number of measures implemented.” Neither the table nor the text refers to the performance -incentive award.

While the stipulation is otherwise silent as to the chart appeari ng in Ameren’s report, it further includes a paragraph describing the variances granted to Ameren to the extent that the stipulation’s terms and conditions are inconsistent with Commission rules. The rule at issue here, with the contested term “methodolog y” and the use of the most recently filed IRP methodology to calculate avoided costs for the performance incentive, is not among any of the granted rule variances. As modified by the stipulation, Ameren’s performance incentive was to be recovered “[a]fter the conclusion of the three-year Plan period,” which explains why the Commission granted a number of variances that address the performance incentive but only within the context of “timing of recovery” and of “calculation.”

4. Final Recovery Under Ameren’s DSIM

Ameren filed an integrated resource plan (IRP), as required by Chapter 22 rules, in 2014, or the first time that an IRP had been filed since the utility calculated avoided costs for purposes of submitting its MEEIA 1 Plan to the Commission. The IRP included updated avoided-cost data, reflecting changes in the cost of energy since the prior IRP had been filed. Ameren then provided data to its EM&V contractors to calculate 2014 net shared benefits for use in determining its performance -incentive award over the life of its 2013-2015 demand-side program. Unlike the IRP, however, the data Ameren submitted to its contractors to calculate the avoided costs for the performance incentive had not been updated from its MEEIA 1 Plan submission. The EM&V evaluators filed Ameren’s final DSIM reports in May 2015. In June 2015, Commission Staff filed a complaint, alleging that Ameren had failed to comply with 4 C.S.R. § 240-20.093(1)(F) by not providing “its independent evaluation, measurement *10 and verification contractors [EM&V contractors] with the most recent avoided cost information needed for the calculation of the portion of annual net shared benefits that are to be awarded to [Ameren] as a performance incentive” under its DSIM for the 2014 program year. 6 Ameren admitted that it had not given its EM&V contractors the avoided-cost data used in its most recently adopted IRP. Ameren contended, however, that it had performed its obligations under the stipulation approved by the Commission because its approved modified plan did not require avoided costs to be updated for the performance-incentive award. Ameren also argued that the rule requiring a utility to use “the same methodology used in its most recently-adopted [sic] preferred resource plan [IRP] to calculate its avoided costs” under its DSIM does not require the use of the updated avoided-cost data included in the avoided-cost calculations for its most recently adopted IRP. 4 C.S.R. § 240-20.093(1)(F) (emphasis added). 7

5. The Commission’s Ruling

The Commission heard oral argument on Staff’s complaint following briefing and issued an order in November 2015 granting Staff’s motion for summary determination and denying Ameren’s motion for summary determination. As part of its findings of fact, the Commission explained what avoided costs are and summarized the dispute this way:

Avoided costs are an estimate of future costs over at least a 20-year period. At the time Ameren Missouri’s DSIM was created, that estimate of avoided costs was based on the methodology used in the preferred resource plan set forth in Ameren Missouri’s MEEIA 1 Plan. Ameren

6 The Missouri Department of Economic Development – Division of Energy sought leave to late file an application to intervene, and the Commission granted the application. While the intervenor participated in the summary determination hearing, it has not appealed t he Commission’s order. 7 Under Chapter 22 of our Code of State Regulations, a “preferred resource plan” is part of a utility’s “integrated resource plan” or IRP and refers to the utility’s long -range supply and demand forecasts conducted under the Commission’s resource pla nning rules.

Missouri made its next Chapter 22 Electric Utility Resource Planning Rules triennial IRP filing in 2014. For the 2014 IRP filing, the formula used in the methodology did not change, but the numbers plugged into the formula used to estimate avoided costs did change. As a result, the estimate of avoided costs also changed [due to significant drops in market prices for energy].

EM&V, as performed by Ameren Missouri’s contractors, does not calculate or otherwise determine the avoided costs used to calculate net shared benefits. Instead, the avoided cost estimates are provided to the EM&V contractors by Ameren Missouri. When Ameren Missouri provided the estimate of avoided costs to its independent EM&V contractors for program year 2014, it gave them the estimated avoided costs as calculated using the inputs from the 2012 MEEIA 1 Plan methodology, not the estimated avoided costs calculated using the inputs from the 2014 IRP methodology. Staff asked Ameren Missouri to provide the avoided cost estimates using the inputs from the 2014 IRP methodology to the EM&V contractors, but Ameren Missouri refused to do so, contending that the DSIM established in the 2012 stipulation and agreement does not require the use of updated costs estimates.

While the Commission acknowledged that Ameren’s proposed plan did not allow the use of updated avoided costs estimates, it noted that the stipulation “provides for variances from several rules that would otherwise be inconsistent with the provisions of the stipulation and agreement, [and that] subsection 4 CSR 240 -20.093(1)(F) is not one of the rules from which a variance is provided.” Thus, the Commission ruled that the utility’s approved DSIM “remains subject to that regulation, and Ameren Missouri is required to ‘use the same methodology used in its most recently adopted preferred resource plan to calculate its avoided costs’” for purposes of calculating its performance incentive.

The Commission determined that, in the context of rule 4 C.S.R. § 240 -20.093(1)(F), “methodology includes both the formula by which avoided costs are to be calculated and the inputs used in that formula.” According to the Commission, this “interpretation is consistent with the goal of the MEEIA statute, which is to encourage *12 the electric utility to implement energy-saving measures by protecting the utility’s financial interests while also protecting consumers.” This goal, according to the Commission, is accomplished by connecting the company’s performance incentive “to how much money ratepayers actually saved as a result of the company’s MEEIA program.”

Describing the function of a performance incentive under MEEIA, the Commission determined that the disputed rule must be interpreted as requiring the use of the most recently used data in calculating that incentive. In this regard, the Commission stated,

The sole purpose of a performance incentive under MEEIA is to give the utility an earnings opportunity that will place shareholders in a financial position comparable to the earnings opportunity they would have had if those shareholders had instead made a future supply-side investment. Future earnings opportunities from supply-side investments are dependent on the dynamic character of the energy marketplace. If energy and capacity market prices increase, the utility may be able to earn greater profits. Conversely, if those market prices drop, the utility may be able to earn less profit on its investment. Thus, it is appropriate that the calculation of the utility’s performance incentive should reflect the most current market price information available when avoided costs are calculated. That is the result obtained when the requirements of Commission Rule 4 CSR 240-20093(1)(F) are interpreted correctly, as described in Staff’s complaint.

Ameren filed an application for rehearing and request for clarification. Intervenor Missouri Division of Energy also filed an application for rehearing. The Commission denied the requests for rehearing, but agreed with Ameren that, in making its calculation, “the 2014 IRP actual costs begin to apply to the calculation of net benefits only after the 2014 IRP was filed.” Ameren filed this appeal.

Legal Analysis

Ameren raises three points on appeal, arguing that (1) the Commission erred and its order was unreasonable because it incorrectly interpreted and effectively rewrote the applicable rule; and its order is arbitrary and capricious and constitutes an abuse of discretion because (2) “the Commission relies on financial metrics calculated using avoided cost estimates existing when the plan was approved, but then disregards those avoided cost estimates in evaluating the operation of the plan that it approved”; and (3) its rationales “do not support the Commission’s made-up definition of the term ‘methodology’ as used in the MEEIA rules.”

Under section 386.510, 8 we review a Commission order to determine whether the order is lawful and reasonable. Office of Pub. Counsel v. Mo. Pub. Serv. Comm'n, 409 S.W.3d 371, 375 (Mo. banc 2013). The Commission’s order “has a presumption of validity, and the burden of proof is on the appellant to prove that the order is unlawful or unreasonable.” Id. Ameren does not challenge the lawfulness of the order. 9 Accordingly, our sole focus in this appeal is on whether it was reasonable.

A Commission decision “is reasonable where the order is supported by substantial, competent evidence on the whole record; the decision is not arbitrary or capricious or where the [Commission] has not abused its discretion.” Id. “We consider the evidence, along with all reasonable supporting inferences, in the light most

8 RSMo Cum. Supp. 2012. 9 Ameren argues that, because rule 4 C.S.R. § 240-2.117 on summary disposition is similar to Missouri Rule of Civil Procedure 74.04, we “must review the record in the light most favorable to the party against whom judgment is entered.” We disagree. Ameren has cited no case or rule, nor have we located any authority, applying the standard of review for a circuit -court ruling on a motion for summary judgment to a Commission ruling on a motion for summary determination.

favorable to the Commission’s order.” State ex rel. Pub. Counsel v. Mo. Pub. Serv. Comm'n, 289 S.W.3d 240, 246-47 (Mo. App. W.D. 2009). And, “where a decision involves the exercise of [Commission] regulatory discretion, Missouri courts have long recognized that the Public Service Commission Law delegates a large area of discretion to the [Commission] and ‘many of its decisions necessarily rest largely in the exercise of sound judgment.’” State ex rel. Mobile Home Estates, Inc. v. Pub. Serv. Comm'n of Mo., 921 S.W.2d 5, 9-10 (Mo. App. W.D. 1996) (quoting State ex rel. Dyer v. Pub. Serv. Comm'n, 341 S.W.2d 795, 802 (Mo. 1960), cert. denied, 366 U.S. 924 (1961)). Thus, “the reviewing court will not substitute its judgment for that of the [Commission] on issues within the realm of the agency’s expertise.” Id.

1. Updated Avoided Costs

Ameren first argues that the Commission disregarded the terms of the MEEIA 1 Plan it approved, because that plan specifically indicated in Table 2.12 that avoided costs would not be updated for purpose of calculating the DSIM’s utility -incentive component. Citing 4 C.S.R. § 240-20.093(2)(J), Ameren contends that the Commission was bound to the DSIM it approved. 10 We agree that the Commission was bound to the DSIM it approved, but we disagree that the performance-incentive calculation was subject to Table 2.12 in Ameren’s proposed MEEIA 1 Plan, because it appears to us that the table addressed the TD-NSB calculation and not the performance incentive.

10 This regulation states:

If the commission approves utility incentive component of a DSIM, such utility incentive component shall be binding on the commission for the entire term of the DSIM, and such DSIM shall be binding on the electric utility for the entire term of the DSIM, unless otherwise ordered or conditioned by the commission when approved.

4 C.S.R. § 240-20.093(2)(J). *15 Even if Table 2.12 applied to the performance incentive, the modified plan that the Commission approved was based on a stipulation that did not grant the utility a variance from section 240-20.093(1)(F), which governs the calculation of the avoided costs that are used in determining a utility’s performance incentive, but not lost revenue (or “TD -NSB”). Accordingly, the Commission’s approval of the utility-incentive component was conditioned on Ameren’s compliance with section 240-20.093(1)(F). The record supports the Commission’s reasonable conclusion that the utility-incentive component of Ameren’s “approved demand-side program remains subject to the requirements of that regulation.”

Neither MEEIA nor its implementing regulations discuss, mention, or define “throughput disincentive net shared benefit” (TD-NSB). MEEIA and the implementing regulations do, however, address “lost revenue,” narrowly defined as “only those net revenues lost due to energy and demand savings from utility demand -side programs approved by the commission. . . .” 4 C.S.R. § 240-3.163(1)(Q). This Court upheld the lost-revenue definition in State ex rel. Public Counsel, 397 S.W.3d at 454, and rejected Ameren’s argument that the definition of “lost revenue” in Chapter 20 should be consistent with its definition in Chapter 22 of the regulations, finding that “the Commission could have reasonably concluded that the definition of lost revenue used in Chapter 22 to address integrated resource planning is not consistent with the purposes of the MEEIA.” 11 Id.

11 Lost revenue is defined in Chapter 22 as “the reduction between rate cases in billed demand (kW) and energy (KWh) due to installed end-use measures, multiplied by the fixed-cost margin of the appropriate rate component.” 4 C.S.R. § 240-22.020(36). ). Note that the Commission approved a lost-revenue component in Ameren’s DSIM based on “net shared benefits” with deemed avoi ded costs. In essence, the Commission allowed Ameren to recover lost revenue “plus,” while the lost -revenue definition was still under challenge. It did not, however, grant a variance as to the avoided costs calculation applicable under the regulations t o a utility incentive.

In the report and the modified plan to which Ameren and the Commission agreed, TD-NSB, the equivalent of lost revenue, is addressed separately from the performance -incentive component and is subject to different requirements, including how TD -NSB would be calculated, using deemed avoided-cost data, both annually and for purposes of a “true-up” at the end of the program. The report and the modified plan do not detail how the performance incentive is to be calculated for purposes of the EM&V annual reports or for trueing up. The regulations define a DSIM utility lost-revenue requirement as “the revenue requirement explicitly approved (if any) by the commission to provide the utility with recovery of lost revenue based on the approved utility lost revenue component of a DSIM.” 4 C.S.R. § 240-20.093(1)(R). Note that this definition does not base lost revenue on “net shared benefit.” So while Ameren and the Commission agreed that the utility could recover a component dubbed TD -NSB, they had to define how it would be calculated because, as a substitute for “lost revenue,” no regulation prescribes how a TD-NSB is to be determined for purposes of recovery under MEEIA. 12

As to the performance incentive to which Ameren would be entitled under its DSIM, the stipulation states that, “using final Evaluation, Measurement and Verification (“EM&V”) results (with EM&V to be performed after each of the program years 1, 2 and 3), Ameren Missouri will be allowed to recover the performance incentive, which is a percentage of NSB [net shared benefit].” The stipulation does not further define how net shared benefits are calculated for purposes of the performance

12 The dissenting opinion fails to recognize that while Ameren’s plan and the modification refer to the DSIM’s lost-revenue component as TD-NSB, lost revenue is not based under the regulations on net shared benefits and thus would not have been subject to the avoided-costs definition or regulatory requirement at issue here.

incentive. But the implementing regulations do address how the performance incentive will be calculated, from expressly defining the incentive as “a portion of annual net shared benefits” and specifying the methodology that will be used to calculate NSB avoided costs (the same used in the utility’s most recent IRP) to defining EM&V as “the performance of studies and activities intended to evaluate the process of the utility’s program delivery and oversight and to estimate and/or verify the estimated actual energy and demand savings, utility lost revenues, cost effectiveness, and other effects from demand-side programs.” 4 C.S.R. § 240.093(1)(Q) & (V) (emphasis added). Further, the stipulation does not waive the requirements of 4 C.S.R. § 240-20.093(1)(F), which thus subjects the performance incentive, based in the regulations on net shared benefits, to “the same methodology used in its most recently -adopted preferred resource plan to calculate its avoided costs.” Lost revenue and a utility incentive are two discrete DSIM recovery components under the regulations, and the Commission’s conclusion that the utility incentive was subject to a regulation that was not explicitly waived by stipulation was reasonable.

According to Ameren, by including data inputs within the meaning of “methodology,” the Commission has rewritten section 240-20.093(1)(F) and effectively substituted the word “inputs” for “methodology.” It asserts that the dictionary defines “methodology” as “a particular procedure or set of procedures,” or, in other words, “the ‘how one goes about’ achieving something or arriving at an outcome or a result.” In its view, the rule includes the formula used to conduct a calculation, but does not include the inputs used in that formula. Because the methodology, but not the avoided-cost data, which Ameren used in its 2014 IRP filing, *18 was the same as the methodology used in its annual DSIM report for 2014, it contends that under its interpretation, it complied with the rule.

We believe that the Commission’s rejection of this interpretation of the rule was reasonable for the simple reason that without knowing what data sets are to be used in a formula, it would be impossible to use the methodology to make a calcu lation. If the inputs are not defined as part of the procedure, any random data could be “plugged into” a formula. In other words, a methodology does not include the input of any specific data, but of necessity it does include and must specify which part icular set of data will be used in the formula. Because the rule states that “[t]he utility shall use the same methodology used in its most recently-adopted preferred resource plan to calculate its avoided costs,” a utility must use the formula and data sets in its most recently adopted IRP as part of using “the same methodology” to calculate its avoided costs. 4 C.S.R. § 240-20.093(1)(F). We do not find that the Commission’s understanding of section 240-20.093(1)(F) is unreasonable given the authorizing legislation’s goal, i.e., that a utility adopting demand-side programs be allowed to recover “all reasonable and prudent costs of delivering cost -effective demand-side programs.” § 393.1075.3. The record supports the Commission’s understanding that the avoided-cost calculation is a long-range projection based on probabilities and assumptions about a volatile energy market that can trend up or down at any point in time. This was, in fact, explained in affidavits filed in support of Ameren’s motion for summary determination. To allow a utility to use stale avoided-cost data and projections in calculating its performance incentive when markets are actually down, however, would allow the shareholders to recover a windfall rather than the reasonable *19 and prudent costs of delivering demand-side programs. The record supports the Commission’s conclusion that “to the greatest extent possible, the Commission encourages the use of actual numbers to calculate cost savings. In this case, that requires the use of updated estimates.”

It also makes little sense to require a utility to use the same methodology underlying its most recently adopted preferred resource plan to calculate its avoided costs without also using the most recent data, as well as the formula, in cal culating those costs to recover a performance incentive. 4 C.S.R. § 240-20.093(1)(F). The dissent does not explain what benefit can be derived from making just the DSIM formula used for calculating avoided costs consistent with that used in the most recently adopted IRP. Nor can we think of any, thus leading one to question whether this rule would have any particular purpose, if “methodology” is not understood as the Commission interpreted it. The value of updated data to calculate achieved results, however, cannot be overstated. A methodology with stale data inputs will not fairly calculate the avoided costs on the basis of which periodic rate adjustments for the performance-incentive component of a DSIM are allowed. 13

Finally, Ameren argues that the Commission’s interpretation of “methodology” is inconsistent with the use of the same term in another MEEIA rule. That rule defines a DSIM’s utility-incentive component as “the methodology approved by the commission in a utility’s filing for demand-side program approval to allow the utility to receive a portion of annual net shared benefits achieved and documented through

13 See the stated purpose for this regulatory section: “This rule allows the establishment and operation of Demand-Side Programs Investment Mechanisms (DSIM), which allow periodic rate adjustments related to recovery of costs and utility incentives for investments in demand -side programs.” 4 C.S.R. § 240-20.093.

EM&V reports.” 4 C.S.R. § 240-20.093(1)(EE). According to Ameren, “methodology” as used here could not include 2014 updated data because this data did not exist when the program was approved. Because the term “methodology” as used in this section does not refer to the calculation of avoided costs, as it does in t he contested rule, we do not believe that the Commission’s “methodology” interpretation that includes data sets for purposes of making a performance-incentive calculation presents any inconsistency or is unreasonable. Point one is denied.

2. Consideration of Certain Factors

Ameren next contends that the Commission’s order was unreasonable because it arbitrarily and capriciously failed to carefully consider certain important factors. It argues that the rules require a utility to submit a significant amount o f information when seeking demand-side program approval. Part of that information is an estimate of “the impact of the DSIM on customer rates over the next five years,” which estimate is derived from comparing the utility’s revenue requirements with and w ithout the proposed DSIM. A “significant part” of the “with the proposed DSIM” revenue -requirement analysis, according to Ameren, depends on “the net benefits to be realized from the plan, which in turn depend heavily on the avoided costs estimates used i n the plan filing.” Because the Commission approves a MEEIA plan “at a given point in time,” its decision is based on information then available, i.e., the avoided cost estimates calculated in 2012. In its view, the Commission cannot “ ‘de-approve’ a plan two or three years later if the avoided cost estimates developed for a later IRP go down as compared to the estimates that underlie the MEEIA filing, any more than does the *21 Commission ‘re-affirm’ the MEEIA plan as being even better if avoided cost estimates go up.” Thus, Ameren argues,

[I]t simply makes no sense to evaluate the level of the incentive the utility is to receive arising from the energy savings that its approved demand - side programs are determined to have achieved, by using a totally different set of avoided cost estimates than was used when the utility and the Commission effectively decided, collectively, that the utility should pursue the energy efficiency programs.

Arguing from the premise that the Commission approved a plan that did not require it to update avoided costs for purposes of calculating its performance incentive, Ameren contends that the order reached an unjust result in ignoring these considerations.

Even if we had not already determined that the modified plan required upda ted avoided costs to be used in calculating the utility’s performance incentive, we do not believe that requiring it to use updated avoided costs constitutes a failure to consider important factors. A performance-incentive award, by its nature, would require that the original projections and estimates in an approved plan be updated to make rate adjustments that are based on more accurate estimates of costs expended or avoided in light of actual conditions over the course of the program. The Commission has not “de-approved” or “re-affirmed” Ameren’s modified plan. It has applied its rule consistently with the statute which requires “that utility financial incentives are aligned with helping customers use energy more efficiently and in a manner that sustain s or enhances utility customers’ incentives to use energy more efficiently.” § 393.1075.3(2). Where customers’ savings are less because energy costs have declined, their incentives to use energy more efficiently evaporate if their bills are higher becaus e *22 the net-shared benefit is based on higher energy prices estimated at the program’s inception, which price estimates have not proved accurate. This point is denied.

3. Commission Rationale

In the third point, Ameren argues that the Commission’s rationales do not support its definition of the term “methodology.” First, it contends that “whatever avoided cost estimates are used at whatever point in time they are used will not produce a determination of ‘actual savings,’” thus no one will know what was “actua lly saved.” This is so, because “no one knew (or knows) what those costs will actually be over the upcoming 20-year period, during which energy efficiency measures installed in 2014 will continue to ‘live.’” Ameren claims that it makes no sense for the C ommission to have connected the utility incentive to actual savings. We disagree.

MEEIA mandates, as a matter of policy, that demand-side programs “[p]rovide timely earnings opportunities associated with cost-effective measurable and verifiable efficiency savings.” § 393.1075.3(3). By requiring that a utility with an approved DSIM use updated avoided-cost data as part of its utility-incentive award calculation, the Commission ensures that shareholders receive “timely” earnings linked to “measurable and verifiable efficiency savings.” According to Ameren’s senior manager of corporate analysis, avoided costs “are based upon national and sometimes international market information for items such as gas, coal, electric energy and capacity, capital markets, and economic drivers.” They include estimates of avoided energy, capacity, and transmission and distribution costs. Ameren determines avoided energy costs by establishing ranges of values for variables and addressing best guesses as to uncertain factors as follows:

[B]y modeling the electric grid (the Eastern Interconnection in the United States), including all demand and available generation, using ranges of values for key driver variables, or “critical uncertain factors” that are likely to affect the market price of electric energy. Ranges of values and subjective probabilities for critical uncertain factors are defined through extensive discussion, review and analysis with subject matter experts. Probabilities for different values are also determined as part of this process. The value ranges and probabilities for the various critical uncertain factors are then combined to create scenarios represented by various combinations of values of critical uncertain factors with a commensurate probability. These scenarios are then simulated with a dispatch model of the Eastern Interconnection, yielding hourly estimates for energy prices for each year of the planning horizon, typically 20 years.

The utility similarly determines avoided capacity and transmission and distribution costs by making certain best estimates, comparisons, interpolations, adjustments, and evaluations, and concludes that “[a]voided costs are a prediction; an estimate, made over a very long period of time.” The Commission understood the tentative nature of these estimates and concluded that requiring that they be updated so that “the calculation of the utility’s performance incentive [w]ould reflect the most current market price information available when avoided costs are calcu lated,” best met Ameren’s obligations under MEEIA and its implementing regulations. We defer to its expertise and find that its rationale was not arbitrary and capricious and did not constitute an abuse of discretion.

Ameren also contends that the Commission’s stated rationale of “protecting the utility’s financial interests while also protecting customers” is a made -up goal that appears nowhere in the MEEIA statute. The utility further complains that the Commission has changed “the rules of the game in the middle of it, by changing avoided cost estimates as the term of the programs proceed.” Ameren characterizes the *24 Commission’s replacing different avoided-costs estimates for those in place when the utility’s DSIM is approved as “an ‘energy and capacity cost lottery.’”

We agree with the Commission that the Legislature intended to protect both shareholder and customer interests by (1) allowing the recovery of “all reasonable and prudent costs of delivering cost-effective demand-side programs,” (2) requiring that the Commission “[e]nsure that utility financial incentives are aligned with helping customers use energy more efficiently and in a manner that sustains or enhances utility customers’ incentives to use energy more efficiently,” and (3) associating “timely earnings opportunities” with “verifiable efficiency savings.” § 393.1075.3. Accordingly, the Commission did not make up a goal and did not err in relying on this rationale to support its interpretation of the word “methodology.” 14

Regarding Ameren’s allegation that the Commission changed the rules of the game thereby effecting a lottery, even if we assume that the Commission’s interpretation of “methodology” represented a change, because imposing an incentive award is similar in some respects to setting a rate, “the Commission is not bound to any set methodology in ensuring a just and reasonable return in setting rates.” State ex rel. Praxair, Inc. v. Pub. Serv. Comm'n, 328 S.W.3d 329, 339 (Mo. App. W.D. 2010). “The Commission has considerable discretion in rate setting due to the inherent complexities involved in the rate setting process. It is not the theory or methodology, but the impact of the rate order which counts.” Id. (citations omitted). Indeed, “[i]f the total effect of the rate order cannot be said to be unjust and unreasonable, judicial

14 Note that the Commission’s interpretation will also best protect the utility’s shareholders who will reap the benefit by collecting a higher performance incentive from ratepayers during those periods when energy prices spike. Commission counsel agreed during oral argument that this would be the case under its interpretation.

inquiry under the Act is at an end. Where ratemaking is at issue, determinations by the Commission are favored by a presumption of validity.” Id. (citations omitted). We cannot say that a Commission order requiring that Ameren use updated avoided costs in calculating net-shared benefits to determine the utility’s performance incentive is unjust or unreasonable. Likening future earnings opportunities from supply -side investments, which “are dependent on the dynamic character of the energy marketplace,” to earnings opportunities from demand-side investments is wholly consistent with the Legislature’s policy of valuing “demand-side investments equal to traditional investments in supply and delivery infrastructure.” § 393.1075.3. Because the performance-incentive award here represents 7 percent of the total that Ameren could recover for adopting a demand-side program, assuming 100 percent of its energy savings goals were realized, we cannot say that the total effect of the Commission’s ruling would be unjust and unreasonable. This point is denied.

Conclusion

The Commission’s order granting Staff’s motion for summary determination and requiring that Ameren use the avoided-cost data from its 2014 IRP in calculating the incentive award under its modified MEEIA plan in 2014 was supported by substantial, competent evidence on the whole record and was accordingly reasonable. 15 We affirm.

/s/ THOMAS H. NEWTON Thomas H. Newton, Judge

Thomas H. Newton, Judge, writes for the majority. Anthony R. Gabbert, Presiding Judge, concurs with the majority. Alok Ahuja, Judge, writes a dissent.

15 The dissenting opinion makes assumptions about the meaning of the stipulation that adopted a modified DSIM plan without granting a variance from § 240-20.093(1)(F). We do not believe that such assumptions are justified in light of the deference we are required to accord the Commissio n.

IN THE MISSOURI COURT OF APPEALS WESTERN DISTRICT MISSOURI PUBLIC SERVICE ) COMMISSION, ) Respondent, ) ) v. ) WD79406 ) UNION ELECTRIC COMPANY ) F ILED : December 6, 2016 d/b/a AMEREN MISSOURI, ) Appellant. )

DISSENTING OPINION

Under its Commission-approved energy-efficiency program, Ameren is entitled to receive, as a performance incentive, a percentage of the “net shared benefits” generated by the energy-efficiency measures which are implemented. The dispute in this case involves the manner in which “net shared benefits” are calculated. The Commission held that, in calculating net shared benefits, Ameren is required to employ updated estimates of the costs which are avoided due to decreased electri city consumption. Ameren argues that the Commission erred, because its Commission -approved Energy Efficiency Plan specifies that the avoided cost estimates would not be updated during the Plan’s three-year life.

I respectfully dissent from the majority’s affirmance of the Commission decision. The Energy Efficiency Plan which Ameren submitted to the Commission for approval in 2012 makes absolutely clear that, for purposes of calculating its *2 performance incentive, Ameren would not be required to update its avoided cost estimates during the Plan’s three-year life. Although the majority disputes that Ameren’s Plan addressed the calculation of the performance incentive, the Commission found that it did, and the parties conceded the p oint. The Unanimous Stipulation which led to the Commission’s approval of the Plan did not address – and therefore did not modify – this aspect of Ameren’s Plan. And the regulation cited by the Commission and the majority opinion does not alter this outc ome, because that rule does not require a utility to employ the same avoided cost estimates contained in a later preferred resource plan.

Once the Commission approved the performance incentive mechanism contained in the Plan and Stipulation, it was bound by the terms of that Plan. The Commission cannot “change the rules of the game” midway through Ameren’s implementation of the Plan.

The record does not reflect the specific financial consequences of the parties’ dispute. In the complaint which commenced this proceeding, however, the Commission’s staff alleged that the issue “is worth millions of dollars to [Ameren’s] ratepayers.” Presumably Ameren and its shareholders face the identical financial impact.

I.

Understanding the specific issue Ameren raises requires a review of the Missouri statutes and rules which govern an electric utility’s implementation of an energy-efficiency program.

The Missouri Energy Efficiency Investment Act is codified at § 393.1075. 1 The Act declares that “[i]t shall be the policy of the state to value demand-side investments equal to traditional investments in supply and delivery infrastructure and allow recovery of all reasonable and prudent costs of delivering cost -effective demand-side programs.” § 393.1075.3. A “demand-side program” is defined as “any program conducted by the utility to modify the net consumption of electricity on the retail customer’s side of the electric meter, including but not limited to energy efficiency measures, load management, demand response, and interruptible or curtailable load.” § 393.1075.2(3). While “demand-side programs” are intended to reduce electricity consumption, “supply-side programs” increase the supply of electricity (such as by increasing a utility’s generating capacity).

Demand-side programs do not fit neatly within the traditional electric utility ratemaking model. Traditionally, an electric utility’s rates are designed to permit the utility to recover its prudently-incurred fixed and marginal costs, as well as a return on investment, through usage-based charges. As a general proposition, a utility’s revenues, and its profits, increase if its customers use more electricity.

Demand-side investments operate in a fundamentally different way. The goal of demand-side programs is to reduce customer demand for the utility’s product: electricity. To the extent a utility’s rates are set using the traditional model, its revenues – and profits – will go down if a demand-side program is successful, since the utility will collect its per-unit rates on a smaller quantity of electricity consumed. Thus, under the traditional ratemaking model, a utility may have a disincentive to

1 Statutory citations refer to the 2000 edition of the Revised Statutes of Missouri, as updated through the most recent cumulative and non-cumulative supplements.

aggressively implement demand-side energy-efficiency programs. A fundamental purpose of the Act is to make demand-side investments as attractive to utilities as supply-side investments, with the goal of reducing overall electricity consumption.

In order to “level the playing field” between demand-side and supply-side investments, the Act authorizes the Commission to “develop cost recovery mechanisms to further encourage investments in demand-side programs including . . . allowing the utility to retain a portion of the net benefits of a demand -side program for its shareholders.” § 393.1075.5.

The Commission has adopted detailed regulations to implement the Act. See § 393.1075.11 (grant of rulemaking authority); State ex rel. Pub. Counsel v. Pub. Serv. Comm’n, 397 S.W.3d 441 (Mo. App. W.D. 2013) (rejecting challenges to the Commission’s implementing regulations). As relevant here, the regulations authorize utilities to seek Commission approval of a “[d]emand-side programs investment mechanism, or DSIM.” Rules 3.163(1)(F), 20.093(1)(M). 2 The regulations provide that a DSIM is “a mechanism . . . to encourage investments in demand-side programs,” and may include “[r]ecovery of lost revenues” resulting from decreased electricity consumption, and a “[u]tility incentive based on the achieved performance level of approved demand-side programs.” Id. The utility incentive component is implemented through a “DSIM utility incentive revenue requirement,” which is designed “to provide the utility with a portion of annual net shared benefits based on

2 Rule citations refer to the Commission’s regulations implementing the Act, which are found in Title 4 of the Code of State Regulations, Division 240 ( i.e., 4 C.S.R. 240). Several of the relevant definitions appear in multiple rules.

the approved utility incentive component of a DSIM.” Rules 3.163(1)(J), 20.093(1)(Q).

This case concerns the manner in which “annual net shared benefits” are to be calculated. That calculation will materially affect the performance incentive Ameren receives for successful implementation of its Energy Efficiency Plan. The regulations define “[a]nnual net shared benefits” to mean:

the utility’s avoided costs measured and documented through evaluation, measurement, and verification (EM&V) reports for approved demand- side programs less the sum of the programs’ costs including design, administration, delivery, end-use measures, incentives, EM&V, utility market potential studies, and technical resource manual on an annual basis.

Rules 3.163(1)(A), 20.093(1)(C). “Avoided cost” is in turn defined to mean: the cost savings obtained by substituting demand-side programs for existing and new supply-side resources. Avoided costs include avoided utility costs resulting from demand-side programs’ energy savings and demand savings associated with generation, transmission, and distribution facilities including avoided probable environmental compliance costs. The utility shall use the same methodology used in its most recently adopted preferred resource plan to calculate its avoided costs.

Rules 3.163(1)(C), 20.093(1)(F).

The performance incentive which a utility is entitled to receive for implementing energy-efficiency programs is analogous to the return on investment which it would receive for implementing supply-side programs. The terms of the performance incentive undoubtedly affect a utility’s decision making in numerous ways, including decisions concerning the resources it devotes to demand -side versus supply-side programs. The Commission’s rules recognize this reliance interest by providing that, upon Commission approval, the utility incentive component becomes binding on both the Commission and the utility:

If the commission approves [the] utility incentive component of a DSIM, such utility incentive component shall be binding on the commission for the entire term of the DSIM, and such DSIM shall be binding on the electric utility for the entire term of the DSIM, unless otherwise ordered or conditioned by the commission when approved.

Rule 20.093(2)(J). II. A.

Ameren submitted its 2013-2015 Energy Efficiency Plan to the Commission for approval on January 20, 2012. The Plan unambiguously provides that Ameren would not be required to update its avoided cost estimates in calculating its performance incentive.

The Plan contained Ameren’s proposal for a demand-side programs investment mechanism or DSIM, having “two main components: direct program cost recovery and a sharing of net benefits to remove economic disincentives and provide timely earnings opportunities.” The Plan explained that the sharing of net benefits was intended to address

two main issues: removal of the throughput disincentive and providing an earnings opportunity equivalent to a supply-side alternative. Removing the throughput disincentive simply makes the utility whole for the revenues it would have collected absent the implementation of its energy efficiency programs whereas the earnings opportunity compensates for the foregone earning opportunities associated with supply-side investments.

The Plan explained the “throughput disincentive” in this way: The implementation of energy efficiency programs causes a decrease in electricity sales, which causes the utility to lose revenue that it would have otherwise collected. But even more importantly, it prevents the utility from recovering a portion of its fixed costs. . . . To fully align utility incentives such that the utility can partner with third party energy efficiency or conservation efforts, the throughput disincentive must be adequately addressed.

In addition to eliminating the throughput disincentive, the Plan also ex plained the need to compensate Ameren for the lost earnings opportunities associated with foregone supply-side investments:

Sharing a portion of net benefits to cover the aforementioned decline in net income only removes the disincentive associated with energy efficiency. But without some way to match the earnings potential of supply-side projects, the utility will continue to favor investments in energy infrastructure projects. . . . Ameren Missouri’s 2011 [Integrated Resource Plan] . . . called for the construction of a combined cycle plant to be completed in 2029. Therefore, if Ameren Missouri engaged in energy efficiency it would forfeit the potential equity earnings associated with that construction investment. In order for energy efficiency investments to be on an equivalent economic footing, the earnings opportunities must be equivalent.

Thus, Ameren proposed that it be entitled to a share of net shared benefits to compensate it for two separate disincentives associated with demand-side programs: (1) the throughput disincentive (representing lost revenues from foregone electricity sales); and (2) the loss of earnings opportunities which are associated with supply -side investments (which would be compensated by a “performance incentive”). The Plan proposed that each of these disincentives be addressed by permitting Ameren to retain a separate specified percentage of net shared benefits.

Ameren estimated that the present value of the total net benefits which would be achieved by implementing its Energy Efficiency Plan would be $364.3 million. It estimated that the present value of three years’ of lost net income associated with decreased electricity consumption (i.e., the throughput disincentive) was $56 million, and that the present value of the performance incentive necessary to compensate for lost earnings opportunities was $17 million. Ameren therefore proposed a total net benefit sharing percentage of 20.2% (assuming that it achieved 100% of the Plan’s *8 performance targets), which would entitle it to retain a portion of the program’s net benefits having a total present value of $73 million.

Determining net shared benefits, and Ameren’s share of those benefits, requires consideration of a number of factors. First, total energy savings must be determined based on the number and kind of energy-efficiency measures actually implemented by Ameren’s customers, and applying a per-measure energy saving amount associated with each different measure. Once total energy savings are determined in megawatt hours, the energy savings must be converted into a dollar figure using an estimate of the avoided costs associated with that reduced electricity usage (including avoided costs of energy production, energy transmission and distribution, and environmental compliance). The total avoided costs must then be compared with the costs of implementing Ameren’s energy-efficiency program, to determine the net monetary benefit of the energy-efficiency program. That net monetary benefit, which includes avoided costs over a multiple-year period, must be reduced to present value using a discount rate. Finally, under the Plan Ameren’s sharing percentage varies depending on its performance (in terms of total energy savings) compared to a goal stated in the Plan. The Plan specified that Ameren’s performance goal would be adjusted based on the number of customers who opted out of the energy-efficiency program. Therefore, it would be necessary to determine the actual number of opt -outs before the amount of Ameren’s share of net shared benefits could be determined.

Ameren’s Energy Efficiency Plan clearly provides – contrary to the Commission’s decision – that the estimates of avoided costs due to decreased energy *9 consumption would not be updated during the life of the Plan as part of the calculation of Ameren’s performance incentive. The Plan explains:

[T]he mechanics of sharing net benefits need to be precisely defined. Table 2.12 shows the items associated with estimating net benefits and whether those items will be updated for purposes of assessing performance and benefits as part of the implementation process. Notice that several items will not be updated, so the focus remains on the cost of the programs and the number of measures implemented. The TRM [Technical Resource Manual] provides significant value in simplifying this process as several important inputs are deemed. Table 2.12 Description of Update Process

(Emphasis added.)

As if Table 2.12 and the discussion above it were not clear enough, immediately below the table the Plan reiterates that only three items will be updated before determining Ameren’s incentive payment: (1) the number of energy -efficiency measures actually implemented; (2) the costs of the energy-efficiency program; and (3) the percentage of customers who opted out of the program. The Plan explains that “when the final performance is judged, the MWh target shall be increased or *10 decreased according to how the opt-out magnitude actually compared to the planning estimate.” The Plan continues:

Once the three year plan implementation is complete, Ameren Missouri will update its DSMore model with the evaluated number of measures implemented and the final program costs. With that updated analysis the final value for net benefits will be calculated an d the sharing percentage applied.

Thus, under Ameren’s Plan, only three of the values needed to calculate its performance incentive would be updated based on actual experience: the number of energy-efficiency measures actually implemented; the costs of implementing the energy-efficiency program; and the number of customers who opted out of the program. The Plan states in no uncertain terms that other variables – including the avoided cost estimates – would remain unchanged throughout the Plan’s three-year life.

B.

The majority opinion concludes that Table 2.12 and the related discussion in Ameren’s Energy Efficiency Plan are irrelevant here. The majority denies “that the performance-incentive calculation was subject to Table 2.12 in Ameren’s proposed MEEIA 1 Plan,” and instead concludes “that the table addressed the [‘throughput disincentive’ or] TD-NSB calculation and not the performance incentive.” Maj. Op. at 14.

The majority’s claim that Ameren’s Plan did not address the manner in which the performance incentive would be calculated is plainly incorrect, and is remarkable from a number of perspectives.

1. First, the majority’s reading of Ameren’s Energy Efficiency Plan is directly contrary to the Commission’s own reading of the Plan. The Commissio n Report and Order expressly recognized that, under the Plan, Ameren was correct that it need not update its avoided cost estimates for purposes of calculating net shared benefits. The Commission expressly found that “the DSIM as proposed by Ameren Missouri in its 2012 MEEIA filing, specifically, subsection 2.6 and Table 2.12 of that filing, does not allow for the use of updated avoided cost estimates.” (Emphasis added.) Although the Commission went on to hold that the terms of the Plan were subject to the Commission regulations defining “avoided cost,” it conceded that Ameren would prevail under the Plan itself. The majority offers no justification for rejecting the Commission’s reading of Ameren’s Plan.

2. Equally significant, the Commission has not argued in this appeal that Ameren’s Plan did not address the calculation of the performance incentive. Instead, as in its Report and Order, the Commission has contended in this Court that the Plan does not constitute the “last word” concerning calculation of the performance incentive, but is instead subject to the definition of “avoided cost” found in Rules 3.163(1)(C) and 20.093(1)(F).

In its appellate briefing, Ameren relied heavily on the terms of its Energy Efficiency Plan, and on Table 2.12 in particular, to argue that it was not bound to update its avoided cost estimates when calculating the performance incentive. If the Commission believed that the Plan did not address the performance incentive calculation, presumably it would have said so. It is telling that the Commission failed to make the argument on which the majority opinion now relies so heavily.

3. To my knowledge, no party to this proceeding has advocated the

majority’s reading of Ameren’s Plan. Before the Commission, both the Commission Staff and the Office of Public Counsel conceded that, under Table 2.12, Ameren would not be required to update its avoided cost estimates when calculating its performance incentive. During oral argument before the Commission, Ameren’s counsel described Table 2.12. The following exchange then occurred:

MR. LOWERY [Ameren counsel]: Okay. So if we stop here, I think that – if this was the only source of information and we didn't have a stipulation that might or might not modify this table, then I think that Staff and OPC would both agree we wouldn't be here today, there would be no complaint, they would agree that’s what you approved, you don't update avoided costs, period. But I agree – CHAIR HALL: Let me – let me cut to the chase. Do you guys agree with that statement? MR. THOMPSON [Staff counsel]: Could you repeat the statement, so I – MR. LOWERY: If we – MR. THOMPSON: – can make sure I understand it? MR. LOWERY: – imagine that we filed the MEEIA plan and the Commission just approved it, said – MR. THOMPSON: Approved it – MR. LOWERY: – approved. MR. THOMPSON: – as filed? MR. LOWERY: Approved it as filed. MR. THOMPSON: Right. MR. LOWERY: In that case I think you would agree that we wouldn't be here today. MR. THOMPSON: I agree. MR. OPITZ [Office of Public Counsel]: I agree.

(Emphasis added.)

Although Staff counsel later clarified that the Plan would still be subject to the Commission’s rules defining “avoided cost,” the essential point remains: both the Commission’s Staff, and the Office of Public Counsel, agreed with Ameren (and with the Commission) that Table 2.12 specifies that avoided cost estimates would not be updated for purposes of calculating the performance incentive. The majority reaches a different conclusion, with no attempt to explain why it adopts a different reading of the Plan than all of the interested parties.

4. The majority’s claim that Table 2.12 does not address the calculation of the performance incentive disregards the Plan’s clear terms. The Plan mak es clear that the sharing percentages which Ameren proposed for both the performance incentive, and to address the throughput disincentive, would operate on the same underlying number: net shared benefits. Ameren’s Plan advocated a 15.4% share of net shared benefits to address the throughput disincentive, and a 4.8% share as a performance incentive. The Plan then combined the two sharing percentages into a single 20.2% share, explaining that “[t]he overall Performance Mechanism must both offset the financial disincentive and provide equivalent earning opportunities to supply-side alternatives.”

Obviously, the separate percentage shares Ameren advocated to address the throughput disincentive, and as a performance incentive, could not be combined unless both percentages operated on the same underlying amount. But the majority opinion contends that, under Ameren’s Plan, the throughput disincentive sharing percentage and the performance incentive percentage would be applied to “net shared benefits” which were calculated in two different ways. If the majority were correct, *14 the Plan’s addition of the two percentages together would have constituted a basic arithmetical mistake.

The text surrounding Table 2.12 makes clear that the table is intended to address the variables which will be applied in calculating both the throughput disincentive share, and the performance incentive share. The combined sharing percentages which Ameren proposed, at different performance levels, are depicted graphically in Figure 2.6 of the Plan. Immediately below Table 2.12 – which the majority contends is relevant only to calculation of the throughput disincentive share – the Plan states that “Figure 2.6 shows the sharing percentages that are applicable at different performance levels.” Figure 2.6, however, contains sharing percentages that combine the throughput disincentive share and the performance incentive share.

5. Finally, the majority’s contention that Ameren’s Plan “do[es] not detail how the performance incentive is to be calculated,” Maj. Op. at 16, creates a fundamental problem: how then are net shared benefits to be determined for purposes of calculating Ameren’s performance incentive? The majority suggests that the Commission’s regulations, standing alone, provide sufficient detail. But the regulations the majority cites do not address all of the issues which Ameren’s Plan indicates are necessary to determine net shared benefits. Given that no party has argued that the Commission’s regulations alone provide sufficient information to calculate net shared benefits, I fail to see how the majority can so confidently assert that Ameren’s Plan (and the Unanimous Stipulation) are irrelevant to that calculation.

III.

As explained in § II above, the Commission correctly found (contrary to the majority opinion) that Ameren’s Energy Efficiency Plan, and specifically Table 2.12, provided that Ameren would not be required to update its avoided cost estimates when calculating the performance incentive. Nothing in the Unanimous Stipul ation alters the plain and unambiguous statements in Ameren’s Plan that avoided cost estimates would not be updated.

The Unanimous Stipulation begins by making clear that the parties agreed that Ameren’s Energy Efficiency Plan should be approved, subject t o the modifications contained in the Stipulation itself.

4. Approval of Plan. Subject to the terms and conditions contained herein, the Signatories agree that Ameren Missouri’s demand - side program plan should be approved. For purposes of this Stipulation, Ameren Missouri’s three-year demand-side program plan (the “Plan”) consists of the 11 demand-side programs (“MEEIA Programs”) described in Ameren Missouri’s January 20, 2012 MEEIA Report, the demand-side programs investment mechanism (“DSIM”) described in the MEEIA Report, modified to reflect the terms and conditions herein, and the Technical Resource Manual (“TRM”) attached as Appendix A to the surrebuttal testimony of Ameren Missouri witness Richard A. Voytas.

Therefore, except as modified by the Unanimous Stipulation itself, the parties recommended that Ameren’s Plan be approved.

The Unanimous Stipulation provides that “Ameren Missouri will be allowed to recover the performance incentive, which is a percentage of [net shared benefits or] NSB as described on Appendix B.” Notably, the performance incentive discussion in the Unanimous Stipulation describes the use of only two actual, updated figures to calculate net shared benefits: (1) the section refers to a determination of actual net *16 energy savings as determined through the Evaluation, Measurement and Verification (“EM&V”) process; and (2) use of the actual number of customer opt-outs. Similarly, Appendix B to the Unanimous Stipulation, addressing the performance incentive calculation, states that “[a]ctual net benefits are based on actual program costs for the three-year MEEIA plan and the actual net MWh savings as determined by EM&V.” The Unanimous Stipulation’s discussion of the performance incentive makes no reference to any other element of the net shared benefits calculation, and expresses no intent to modify any other aspect of the Plan’s discussion of net shared benefits and the performance incentive. Instead – and consistent with Table 2.12 – the Stipulation refers to the use of actual, updated figures only with respect to (1) Ameren’s energy-efficiency program costs; (2) the determination of actual energy savings (based on the number of energy-efficiency measures actually implemented); and (3) the number of customers who opted out of the energy-efficiency program. 3

The majority opinion itself acknowledges that “[n]othing in . . . the stipulation expressly indicates how the performance-incentive award is to be calculated other than that it was to be a percentage of certain net shared benefits set forth in an appendix.” Maj. Op. at 6. The consequence is that, since the Unanimous Stipulation did not modify the statement in Ameren’s Plan that avoided cost estimates would not

3 The Commission claims in its Brief that “ Ameren’s performance incentive is described exclusively by the 2012 Stipulation.” That is simply inaccurate. The Unanimous Stipulation does not discuss a variety of factors which would be necessary to calculate net shared benefits (such as applicable discount rates, or the performance attributes of various energy -efficiency measures). The Unanimous Stipulation cannot be interpreted as a “stand -alone” discussion of the performance incentive, without reference to Ameren’s Plan. The Unanimous Stipulation itself states that it is not a stand-alone description of Ameren’s DSIM, but instead that the DSIM is “described in the MEEIA Report, modified to reflect the terms and conditions herein.”

be updated in calculating net shared benefits, the terms of the Plan continue to govern.

Although it is not our task to evaluate the wisdom of the Commission’s actions, it is worth noting that there are plausible reasons supporting the Commission’s approval of the net shared benefits calculation described in Ameren’s Plan and in the Unanimous Stipulation. As both the Plan and the Unanimous Stipulation explain, the principal values which will vary based on actual program implementation are (1) the number of energy-efficiency measures actually implemented, and (2) the amount of Ameren’s program costs. These two items are, to a substantial degree, within Ameren’s control, and they constitute a reasonable measure of the quality of Ameren’s performance. It seems sensible that the amount of Ameren’s performance incentive would be based, to a significant degree, on its success in implementing energy-efficiency measures, and its success in controlling the program’s costs. The avoided cost estimates, on the other hand, may vary over time based on a host of extraneous factors over which Ameren has no contr ol, such as the performance of national or international capital and energy markets. Holding avoided cost estimates constant over the life of the energy-efficiency program, for the purpose of calculating Ameren’s performance incentive, is a perfectly rati onal approach.

IV.

The Commission’s order, and the majority opinion, support the use of updated avoided cost estimates by referring to Rule 20.093(1)(F), which states that “[t]he utility shall use the same methodology used in its most recently adopted pre ferred *18 resource plan to calculate its avoided costs.” (The identical language appears in Rule 3.163(1)(C).) The reference to use of “the same methodology” in Rule 20.093(1)(F) cannot justify the Commission’s decision.

The Commission’s order concludes that, “in the context of [Rule 20.093(1)(F)], methodology includes both the formula by which avoided costs are to be calculated and the inputs used in that formula.” (Emphasis added.) As Ameren points out, however, the Commission’s interpretation of the word “methodology” is inconsistent with dictionary definitions of the term. A “methodology” is defined as “the processes, techniques, or approaches employed in the solution of a problem or in doing something: a particular procedure or set of procedures.” W EBSTER ’ S T HIRD N EW I NTERNATIONAL D ICTIONARY at 1423 (1993). Under this definition, “methodology” refers to the formula or process for calculating avoided costs, not the specific numerical values inserted into that formula or process to perform a calculation. It is undisputed that Ameren used the same methodology (in the sense of the same process) in calculating its avoided cost estimates for purposes of its Energy Efficiency Plan, as it used in connection with its 2014 preferred resource plan. Nothing in Rule 20.093(1)(F) requires that the avoided cost estimates be updated when Ameren adopts a new preferred resource plan.

The Commission’s current reading of Rule 20.093(1)(F) essentially rewrites it. Instead of requiring that Ameren “use the same methodology used in its most recently adopted preferred resource plan to calculate its avoided costs,” the Commission now reads Rule 20.093(1)(F) to require that Ameren “use the same avoided cost estimates used in its most recently adopted preferred resource plan.” But that is not what the *19 rule says, and the Commission should not be permitted to rewrite the rule after -the-fact. 4

In addition, the Commission’s construction of the word “methodology” in Rule 20.093(1)(F) is inconsistent with the use of the very same word in another relevant Commission rule – Rule 20.093(1)(EE). Recall that, under the Commission’s rules, “[i]f the commission approves [the] utility incentive component of a DSIM, such utility incentive component shall be binding on the commission for the entire term of the DSIM.” Rule 20.093(2)(J). Therefore, the Commission is bound by the “utility incentive component” of Ameren’s approved Demand-Side Investment Mechanism or DSIM. What is the “utility incentive component” by which the Commission is bound? Rule 20.093(1)(EE) supplies the answer: the “utility incentive component” “means the methodology approved by the commission in a utility’s filing for demand-side program approval to allow the utility to receive a p ortion of annual net shared benefits achieved and documented through EM&V reports.” (Emphasis added.)

Therefore, under Rules 20.093(1)(EE) and 20.093(2)(J), the Commission is bound by “the methodology” it approved “to allow [Ameren] to receive a portion o f annual net shared benefits” as a performance incentive. Under the definition of

4 Notably, when the Commission’s Chair stated during oral argument before the Commission that Staff’s argument essentially rewrote Rule 20.093(1)(F) to require that “the utility shall use the same avoided costs used in its most recently adopted preferred resource plan,” Staff counsel agreed: “I think that’s the effect of the rule.” Staff was even more explicit in its briefing to the Commission. It argued:

The word ‘methodology’ as used in the rule necessarily encompasses the formula, the inputs, and the results of the calculation. What Rule 4 CSR 240-20.093.1(F) requires is that the avoided costs from AmMo’s most recently adopted preferred resource plan be used in calculating NSB for the purposes of the [performance incentive].

(Emphasis added.) Staff’s arguments – which the Commission adopted – go well beyond the language of a rule which merely requires that Ameren use a particu lar “methodology . . . to calculate its avoided costs.”

“methodology” adopted in the Commission’s order, the Commission approved – and is bound by – “both the formula by which [Ameren’s portion of net shared benefits] are to be calculated and the inputs used in that formula.” Under that reading, the Commission would be bound by the avoided cost estimates contained in Ameren’s Energy Efficiency Plan, because those avoided cost estimates are an essential input in determining Ameren’s share of net shared benefits. This reading would create a conflict with Rule 20.093(1)(F), however, because the latter rule says that the Commission is not bound by “the methodology” approved as part of Ameren’s DSIM, but instead that “the methodology” contained in Ameren’s most recent preferred resource plan must be used to calculate avoided costs.

This conundrum can be avoided by interpreting the word “methodology” consistent with the dictionary definition – as the process or formula employed to calculate avoided costs, not the inputs inserted into that formula as part of a particular calculation. The Commission cannot “have it both ways,” and interpret the word “methodology” differently under Rules 20.093(1)(F) and 20.093(1)(EE).

The majority dismisses the use of the term “methodology” in Rule 20.093(1)(EE) with the statement that it “does not refer to the calculation of avoided costs.” Maj. Op. at 20. But avoided cost estimates are an essential component in the calculation of net shared benefits, which is in turn an essential component in determining Ameren’s performance incentive amount. The “methodology” to which Rule 20.093(1)(EE) refers would plainly include a determination of the relevant avoided cost estimates (as Table 2.12 to Ameren’s Plan makes clear).

Besides the fact that it is inconsistent with the dictionary, and with the use of the term “methodology” in another regulation, the Commission’s – and the majority’s – interpretation of the word “methodology” in Rule 20.093(1)(F) also leads t o an absurd outcome. If the Commission’s current order is correct, and the reference to “methodology” in Rule 20.093(1)(F) “includes both the formula by which avoided costs are to be calculated and the inputs used in that formula,” then the Commission in 2012 approved a DSIM which on its face conflicted with the Commission’s own regulations, in two significant respects. First, Ameren’s Energy Efficiency Plan employed avoided cost estimates which were different than the avoided cost estimates contained in its most recent preferred resource plan (filed in 2011), although the avoided cost estimates incorporated into the Plan were derived using the same process used in connection with the 2011 preferred resource plan. The Plan specifically advised the Commission that “[t]he avoided energy costs [used in the Plan] represent an update to the [Integrated Resource Plan] planning scenarios,” and provided the Commission with “a description of those updates.” 5 Those avoided cost estimates were used, among other things, to determine the cost-effectiveness of Ameren’s proposed energy-efficiency program, and to determine the likely impact of the energy-efficiency program on customer rates. If the Commission’s current interpretation of Rule 20.093(1)(F) were correct, Ameren violated the rule in 2012 when it used new avoided cost estimates developed specifically for use in conjunction

5 Notably, the Commission itself acknowledges in its Brief that “ the methodology Ameren used to calculate avoided costs of energy for its 2012 MEEIA request is not identical to the avoided costs as determined in Ameren’s 2011 IRP filing.” The Commission’s Staff made the same concession during argument before the Commission: “the avoided costs that Ameren used in its MEEIA plan were already a change from the 2011 IRP.” These admissions are curious, since they essentially acknowledge that Ameren’s 2012 Plan – which the Commission approved – violated the Commission’s current interpretation of Rule 20.093(1)(F).

with its Energy Efficiency Plan, rather than the avoided cost estimates from its 2011 preferred resource plan. Yet, no one objected to the avoided cost estimates Ameren used in connection with Plan approval in 2012.

Second, as explained in §§ II above, Ameren’s Energy Efficiency Plan unambiguously provides that Ameren’s avoided cost estimates would be “deemed” for the purpose of calculating net shared benefits, and that those avoided cost estimates would not be updated during the three-year life of the Plan. Yet, according to the Commission’s current interpretation of Rule 20.093(1)(F), use of unchanging avoided cost estimates is prohibited. On that interpretation, the explicit references in the Plan to the use of unchanging, “deemed” avoided cost estimates were patently unlawful at the time the Plan was submitted. But the Commission’s Staff expressly advocated the approval of this performance incentive mechanism, and the Commission approved it. We should hesitate to adopt a reading of Rule 20.093(1)(F) which renders the Commission’s 2012 approval of Ameren’s Plan a nullity in any material respect.

The majority also seizes on the fact that the Unanimous Stipulation did not give Ameren a variance from Rule 20.093(1)(F). But there was no need for a variance, if Rule 20.093(1)(F)’s reference to “the same methodology” is properly interpreted. If properly interpreted, the rule is fully consiste nt with using unchanging avoided cost estimates throughout the three-year term of the DSIM. But even if the Commission’s current interpretation of Rule 20.093(1)(F) were correct, I would hold that the Commission granted Ameren a variance from that rule wh en it approved Ameren’s DSIM. With respect to regulatory variances, the Commission’s rules simply provide that, “[u]pon request and for good cause shown, the commission may *23 grant a variance from any provision of this rule.” Rule 20.093(13). The rule doe s not require that variances take any particular form.

Here, as I have described in detail above, Ameren sought approval of a Demand-Side Programs Investment Mechanism which expressly provided that avoided cost estimates would remain constant throughout the life of the DSIM, and which employed avoided cost estimates different from those appearing in Ameren’s most-recent preferred resource plan. The Commission’s Staff did not object to the avoided costs estimates Ameren used in seeking Plan approval, or to the Plan’s specification that the avoided cost estimates would remain unchanged throughout program implementation; indeed, Staff itself requested that the Commission approve Ameren’s Plan as modified by the Stipulation. The Commission then approved the Plan in the form submitted. In these circumstances, if the express terms of the Plan are deemed to conflict with Rule 20.093(1)(F), I would hold that the Commission granted a variance from the Rule’s requirements when it approved the Plan as modified by the Stipulation. Notably, the majority itself essentially acknowledges that the Commission gave Ameren an unstated regulatory variance, since it acknowledges that the throughput disincentive share which the Commission approved, to compensate Ameren for lost revenues, is inconsistent with the definition of “lost revenue” contained in the Commission’s own Rules 3.163(1)(Q) and 20.093.(1)(Y). 6

6 The Commission also justified its interpretation of Rule 20.093(1)(F) by asserting that i t would more closely align Ameren’s performance incentive to the amount consumers “actually saved” as a result of the deployment of energy-efficiency measures. But the avoided cost estimates are intended as estimates of expected savings over the multi-year life of various energy-efficiency measures; they are not simply an accounting of savings actually achieved during any particular period of time. Ameren will be recovering a performance incentive based on complex estimates of net benefits – not “actual savings” – whether we uphold the Commission’s position or Ameren’s.

Conclusion

For the foregoing reasons, I respectfully dissent from the majority’s affirmance of the Commission’s order.

/s/ ALOK AHUJA Alok Ahuja, Judge