Vote Solar v. PSC

2020 MT 213

Opinion

08/24/2020

DA 19-0223

Case Number: DA 19-0223

IN THE SUPREME COURT OF THE STATE OF MONTANA 2020 MT 213

VOTE SOLAR, MONTANA ENVIRONMENTAL INFORMATION CENTER, and CYPRESS CREEK RENEWABLES, LLC,

Plaintiffs and Appellees, and WINDATA, LLC, Plaintiff-Intervenor and Appellee, v.

MONTANA DEPARTMENT OF PUBLIC SERVICE REGULATION, MONTANA PUBLIC SERVICE COMMISSION,

Defendant and Cross-Appellant,

NORTHWESTERN CORPORATION, d/b/a NORTHWESTERN ENERGY,

Defendant and Appellant,

and MONTANA CONSUMER COUNSEL, Defendant-Intervenor.

APPEAL FROM: District Court of the Eighth Judicial District, In and For the County of Cascade, Cause No. BDV-2017-0776 Honorable James A. Manley, Presiding Judge

COUNSEL OF RECORD:

For Appellant NorthWestern Energy:

Ann Hill, Al Brogan (argued), NorthWestern Energy, Helena, Montana

For Cross-Appellant Montana Public Service Commission:

Zachary Taylor Rogala (argued), Justin Wade Kraske, Montana Public Service Commission, Helena, Montana

For Appellees Vote Solar and Montana Environmental Information Center:

Jenny K. Harbine (argued), Earthjustice, Bozeman, Montana For Appellee Cypress Creek Renewables, LLC:

Maria Phillips Barlow, Attorney at Law, Portland, Oregon For Intervenor and Appellee WINData, LLC:

Monica J. Tranel (argued), Tranel Law Firm, P.C., Missoula, Montana For Intervenor Montana Consumer Counsel:

Jason T. Brown (argued), Montana Consumer Counsel, Helena, Montana

Argued: February 26, 2020 Submitted: March 3, 2020

Decided: August 24, 2020

Filed:

r--6ta•--df

__________________________________________

Clerk

3

Chief Justice Mike McGrath delivered the Opinion of the Court:

¶1 This case stems from challenges by Vote Solar,1 the Montana Environmental Information Center,2 and Cypress Creek Renewables, LLC,3 (collectively “VS-MEIC”) to the Montana Public Service Commission (“PSC”) Order Nos. 7500c and 7500d, in which the PSC reduced standard-offer contract rates and maximum contract lengths for small solar qualifying facilities (“QFs”). On April 2, 2019, Montana’s Eighth Judicial District Court issued an order vacating and modifying PSC Order Nos. 7500c and 7500d. The PSC and NorthWestern Energy (“NorthWestern”)4 appeal. We affirm and remand to the PSC with instructions.

¶2 We restate the issues on appeal as follows:

Issue One: Whether the District Court erred when it determined that the PSC arbitrarily and unlawfully reduced solar QF standard-offer rates by excluding carbon dioxide emissions costs and NorthWestern’s avoided costs of operating its internal combustion engine resource units from the avoided-cost rate.

Issue Two: Whether the District Court erred when it concluded that the PSC arbitrarily and unreasonably calculated solar QFs’ capacity contribution in determining avoided costs.

1 Vote Solar is a nonprofit, grassroots organization working to foster economic opportunity, promote energy independence, and fight climate change by making solar energy a mainstream energy resource across the United States. Vote Solar has more than 70,000 members throughout the United States, with offices in Oakland, California.

2 Montana Environmental Information Center, located in Helena, Montana, is a nonprofit environmental advocate founded in 1973 by Montanans focused on protecting and restoring Montana’s natural environment, including through renewable energy development.

3 Cypress Creek is a renewable energy developer with 11 advanced-stage solar projects in Montana, two of which are at issue in the present action.

4 NorthWestern is a Delaware corporation doing business as NorthWestern Energy, a public energy utility, in Montana, South Dakota, and Nebraska.

4

Issue Three: Whether the District Court erred when it determined that the PSC arbitrarily and unreasonably reduced maximum-length contracts to 15 years for solar QFs.

FACTUAL AND PROCEDURAL BACKGROUND

¶3 Before addressing the merits of the case, we contextualize the issues presented by providing necessary background of applicable federal and state law, historical practices of the PSC in setting contract lengths and standard-offer rates, and the relevant factual and procedural history of the present action. PURPA

¶4 In 1978, Congress enacted the Public Utility Regulatory Policies Act (“PURPA”) to reduce American dependence on fossil fuels, encourage renewable energy development, and promote increased energy efficiency. 16 U.S.C. § 824a-3; FERC v. Mississippi, 456 U.S. 742, 745-46, 102 S. Ct. 2126, 2130 (1982); Small Power Prod. and Congregation Facilities; Regulations Implementing Sec. 210 of the Pub. Util. Reg. Pol. Act of 1978, Order No. 69, 45 Fed. Reg. 12,214, 12,215 (Feb. 25, 1980). PURPA aims to eliminate significant barriers to the development of alternative energy sources, including the reluctance of traditional electric utilities to purchase power from and sell power to non-traditional facilities and the financial burdens imposed upon alternative energy sources by state and federal utilities. Californians for Renewable Energy v. Cal. PUC, 922 F.3d 929, 932 (9th Cir. 2019) (citing Indep. Energy Producers Ass’n, Inc. v. Cal. Pub. Utils. Comm’n, 36 F.3d 848, 850 (9th Cir. 1994)). PURPA charges the Federal Energy Regulatory

5

Commission (“FERC”) with enacting PURPA’s implementing regulations. Californians for Renewable Energy, 922 F.3d at 931.

¶5 Section 210 of PURPA requires large electric utilities to purchase energy from small power production QFs at standard-offer rates. 18 C.F.R. §§ 292.201, 292.203, 292.204. Small power QFs have a nameplate capacity of 80 megawatts (“MW”) or less and produce electric power from biomass, waste, or renewable resources such as wind, water, or solar energy. 18 C.F.R. § 292.204(a), (b); 16 U.S.C. § 796(17)(A). Rates must be “just and reasonable” to consumers, “in the public interest,” and nondiscriminatory to the QF to “encourage” renewable energy development and allow small QFs to “become and remain viable suppliers of electricity.” 18 C.F.R. § 292.304(a); 16 U.S.C. § 824a-3(a), (b); Whitehall Wind, LLC v. Mont. Pub. Serv. Comm’n, 2010 MT 2, ¶ 7, 355 Mont. 15, 223 P.3d 907 (Whitehall Wind I).

¶6 When setting the purchase price, QFs must be compensated at a rate equal to the utility’s full avoided cost. 18 C.F.R. § 292.304(b)(2); Am. Paper Inst. v. Am. Elec. Power Serv. Corp., 461 U.S. 402, 406, 103 S. Ct. 1921, 1924 (1983). Avoided costs are “the incremental costs to an electric utility of electric energy or capacity or both which, but for the purchase from the qualifying facility or qualifying facilities, such utility would generate itself or purchase from another source.” 18 C.F.R. § 292.101(b)(6) (emphasis added). Capacity costs are those costs incurred from providing the capability to deliver energy, consisting primarily of the capital costs of energy storing facilities. FERC Order No. 69 at 12,216. Energy costs are costs associated with energy production, including operating and maintenance expenses. FERC Order No. 69 at 12,216. By limiting the purchase price to

6

a utility’s avoided cost, Congress sought to achieve a balance between the interests of ratepayers and generators. Energy purchased at the utility’s avoided cost is reasonable for consumers because it is equivalent in price as if the utility generated the power itself or purchased it from another source. S. Cal. Edison Co., San Diego Gas & Elec. Co., 71 FERC ¶ 61,269, 62,280 (June 2, 1995).

¶7 The Dissent adopts the incorrect argument promoted by the Montana Consumer Counsel (“Consumer Counsel”)5 that the most critical factor of avoided-cost analysis is protecting the ratepayer. Were that the case, there would be no purpose to PURPA, which is to preclude discrimination in the market place for sources of energy that provide an alternative to fossil fuel development. Of course, that purpose must be undertaken along with the endeavor to hold the ratepayer neutral or indifferent to the source of energy they consume. Moreover, NorthWestern’s frequently-uttered trope that the requirements of PURPA and thus approval of solar sources of energy will wildly increase the rates charged to consumers finds little basis of support in this record.

¶8 FERC does not provide a specific method for establishing the avoided-cost rate but requires that any state-adopted method be consistent with its implementing regulations. See 18 C.F.R. § 292.304(c). FERC has thus provided a list of guidelines for states to consider when developing avoided-cost rates, including (1) the usefulness of the QF’s energy during system emergencies; (2) the individual and aggregate value of energy and

5 The Consumer Counsel represents Montana public utility and transportation consumers before the Public Service Commission, state and federal courts, and administrative agencies in matters concerning public utility regulation.

7

capacity from QFs to the utility’s system; (3) the QF’s smaller capacity increments and shorter lead times—the time it takes to make, produce, or deliver energy; (4) the QF’s ability to enable the utility to defer capacity additions and decrease reliance on fossil fuels; and (5) the utility’s cost savings resulting from decreased line losses of energy during transmission from the QF. 18 C.F.R. § 292.304(e)(2)(v)-(vii), (3), (4). FERC has clarified that energy and capacity avoided costs should be considered together, directing utilities to submit associated energy costs of each planned unit along with the estimated capacity cost of planned capacity additions, such that the calculation of avoided costs “includes the lower energy costs that might be associated with the new capacity.” FERC Order No. 69 at 12,218.

¶9 When determining avoided-cost rates, PURPA mandates states to consider “the terms of any contract . . . including the duration of the obligation . . . .” 18 C.F.R. § 292.304(e)(2)(iii). Neither PURPA nor FERC regulations set a specific contract length requirement for QFs. However, under PURPA, contract duration is closely intertwined with contract rates. Public utilities must encourage QF development in establishing avoided costs, in part by encouraging long-term contracts to “enhance the economic feasibility” of QFs. 16 U.S.C. § 824a-3(a); FERC Order No. 69 at 12,226; § 69-3-604(2), MCA. QFs are to be able to enter contractual commitments based on estimates of future avoided costs to provide certainty regarding potential return on investments. FERC Order No. 69 at 12,224.

8

Montana’s Mini-PURPA

¶10 State utility regulatory agencies implement PURPA through rulemaking. Whitehall Wind I, ¶ 9; Californians for Renewable Energy, 922 F.3d at 931. The relevant state agency here is the PSC. In 1981, Montana adopted PURPA, i.e., “mini-PURPA,” §§ 69-3-601 through 604, MCA. In Montana, QFs with a nameplate capacity of 3 MW or less are eligible for standard-offer rates—NorthWestern’s “QF-1 Tariff Rate”—at issue in this case.6 Admin. R. M. 38.5.1902(5). Like PURPA, Montana’s “mini-PURPA” requires that avoided-cost rates and contract lengths be sufficient to “enhance the economic feasibility of [QFs],” at rates that allow the QF to become and remain viable suppliers of electricity. Section 69-3-604(2), MCA; Whitehall Wind I, ¶ 7. The PSC’s Historical QF Rate Setting Practices Avoided-Cost Rate

¶11 In Montana, the PSC requires NorthWestern to submit biennial updates to its avoided cost information within 30 days of filing a resource procurement plan based on the costs of any planned new resources identified in its most recent plan. Sections 69-8-419, 420, MCA; Admin. R. M. 38.5.1905(1)(b). At the time the present case commenced, NorthWestern’s most recent resource procurement plan was its 2015 plan. NorthWestern did not file a 2017 procurement plan but has since filed one for 2019.

6 QFs between 3MW and 80 MW of nameplate capacity may negotiate avoided-cost rates directly with NorthWestern or opt for a standard avoided-cost rate set by the PSC. Admin. R. M. 38.5.1903(2); 18 C.F.R. § 292.304(c).

9

¶12 Three primary methodologies are used to calculate avoided-cost rates, one of which is the proxy methodology.7 The proxy method, regarded by NorthWestern’s and VS-MEIC’s experts as the simplest avoided cost methodology, estimates avoided costs based on the projected capacity and energy costs of the utility’s next planned resource additions. This method assumes that by purchasing energy directly from the QF the utility can delay the construction of its next planned generating unit as identified in its most recent integrated resource plan.8 The fixed costs of this hypothetical proxy unit, such as the capital costs associated with constructing the unit, determines avoided-capacity costs, and the variable costs, including the projected costs of operating the new capacity unit, become the avoidable energy costs.9 In the years before future generating units become active, the cost of short-term market purchases is used to calculate the avoided-cost rate.

¶13 In 2010, the PSC approved NorthWestern’s proposal to classify the nominal, levelized avoided-cost rate into separate energy and capacity elements. In re NorthWestern

7 Two additional methods routinely used to calculate avoided-cost rates are the “peaker”

method and the “hybrid-proxy” method. The peaker method assumes a QF allows the utility to reduce the marginal generation on its system and to avoid building a peaking unit, rather than displacing or delaying the need for a new generation unit. The hybrid-proxy method, considered to be the most equitable and accurate method, essentially distinguishes between whether the utility is in a greater energy or capacity deficit.

8 The PSC has approved NorthWestern’s use of two standard QF-1 rate options: one that reflects avoided costs based on NorthWestern’s identification of the least-cost, least-risk facility in its most recent resource procurement plan, and a second option that offers indexed rates that reflect hourly market prices and does not provide assurance of a long-term fixed rate. In re NorthWestern Energy’s Application for Qualifying Facility Tariff Adjustment, Order No. 7338b, ¶¶ 18, 32, Dkt. D2014.1.5 (May 4, 2015). The second option is not at issue here.

9 See Kavulla Test., Dkt. No. AD 16-16-000, Tech. Conf. on Implementation Issues Under PURPA (FERC, June 29, 2016), available at https://perma.cc/8KYC-U3VC.

10

Energy’s Application for Approval of Avoided Cost Tariff for New Qualifying Facilities, Order No. 6973d, ¶ 134, Dkt. D2008.12.146 (May 6, 2010). In 2011, the PSC adopted the proxy method for calculating avoided costs, also known as the “blended market + combined cycle” approach. In re NorthWestern Energy’s Application for Approval of Avoided Cost Tariff for New Qualifying Facilities, Order No. 7108e, ¶ 68, Dkt. D2010.7.77 (Oct. 19, 2011). The PSC substituted a combined cycle combustion turbine unit,10 scheduled to begin operation in 2015, in place of a Colstrip Unit 4 resource, a coal-fired generating unit, based on a preferred portfolio approach detailed in NorthWestern’s 2009 resource procurement plan. Order No. 7108e, ¶¶ 66, 67. In 2012 and 2014, the PSC again applied the proxy method based on a combined cycle combustion turbine avoided-cost resource. In re NorthWestern Energy’s Application for Approval of Avoided Cost Tariff for New Qualifying Facilities, Order No. 7199d, ¶ 18, Dkt. D2012.1.3 (Dec. 7, 2012); Order No. 7338b, ¶¶ 18-20.11

¶14 A separate consideration when determining avoided costs concerns a QF’s “capacity contribution”—the electricity supplied by a QF to the utility’s system during periods of high demand. 18 C.F.R. § 292.304(e)(3). The PSC has routinely determined a QF’s capacity contribution during “on-peak” hours—the hours that the avoided-capacity

10 A combined cycle combustion turbine uses both gas and steam turbines to produce electricity from the same fuel source. Waste heat from the gas turbine is routed to a nearby steam turbine, which in turn, sends energy to the generator drive shaft to be converted into additional electricity.

11 The proxy method also relies on natural gas forecasts in setting the QF-1 rate. In 2012 and 2014, NorthWestern set QF-1 rates by relying on forward market natural gas forecasts at regional trading hubs. See Order No. 7199d, ¶¶ 27-28; Order No. 7338b, ¶ 23.

11

resource would be expected to perform. See Order No. 6973d, ¶ 134; Order No. 7199d, ¶ 55; In re Crazy Mountain Wind, Order No. 7505b, ¶ 87, Dkt. D2016.7.56 (Jan. 5, 2017).

Carbon Adder

¶15 Since 2012, the PSC has included carbon pricing when determining avoided-cost rates. See In re NorthWestern Energy’s Application for Approval to Purchase and Operate the Spion Kop Wind Project, Order No. 71591, ¶¶ 113-132, Dkt. D2011.5.41 (Feb. 16, 2012). In 2014, the PSC approved NorthWestern’s proposal to collect through customer rates the avoided-carbon costs associated with its purchase and operation of hydroelectric QFs, estimating avoided-carbon costs to be $21.11 per ton beginning in 2021 and escalating at a 5% annual interest rate for the remaining years of the contract. In re NorthWestern Energy’s Application for Hydro Assets Purchase, Order No. 7323k, ¶¶ 83, 89, Dkt. D2013.12.85 (Sept. 26, 2014). Then, in 2017, just seven months prior to issuing Order No. 7500c in this docket, the PSC included carbon emissions in its avoided-cost rate for small wind QFs.12 Order No. 7505b, ¶ 61. Recognizing the speculative nature of forecasts of carbon dioxide emission prices, the PSC nonetheless acknowledged that carbon dioxide emission price is necessarily “a measure of avoided electricity supply costs.” Order No. 7505b, ¶¶ 54, 58. The PSC, however, delayed the onset date for compensating QFs for carbon dioxide emissions from 2022 to 2025 to account for a new presidential

12 Crazy Mountain Wind involved a contested case regarding contract terms for NorthWestern’s purchase of wind energy from small QFs with a nameplate capacity greater than 3 MW but less than 80 MW. In July 2016, at almost the same time as when the present case commenced, Crazy Mountain Wind, LLC, filed a petition with the PSC requesting an order setting contract terms and conditions with NorthWestern.

12

administration hostile to carbon dioxide emission regulation. Order No. 7505b, ¶ 57. The PSC included a fixed $20.00 per ton for avoided carbon dioxide emissions cost adder, escalating at a 4.15% annual interest rate for the remaining years in the contract. Order No. 7505b, ¶¶ 56-57.

¶16 In discussing carbon adder, the Dissent presents a false equivalent that the majority endorses what would be the “first-ever carbon adder in a QF-1 contract.” Dissent, ¶ 84. But of course, as noted, consideration of carbon adders in QF contracts has been the case for other alternative sources, including hydro and wind power projects, since 2012.

Contract Length

¶17 NorthWestern has historically executed maximum-length contracts with QFs in Montana for at least 25-year terms. In fact, prior to issuing its 2015 resource procurement plan, NorthWestern signed two QF contracts for future wind and hydro resources and five QF contracts for future solar resources, all for 25 years. NorthWestern Energy, 2015 Electricity Supply Resource Procurement Plan, Existing Resources, at 8-18 (Mar. 31, 2016). The longest contracts NorthWestern maintains of its own electric resources are between 30 and 40 years. Current Dispute

¶18 On May 3, 2016, NorthWestern filed its biennial avoided-cost application with the PSC, proposing drastic reductions in standard-offer rates for NorthWestern’s purchases from small solar and wind QFs. For solar QFs, NorthWestern’s application proposed reducing the standard QF-1 rate from $66 per megawatt hour (“MWh”) to between $34 and $44 per MWh. NorthWestern requested the PSC adopt the proposed tariff

13

immediately on an interim and final basis. NorthWestern argued it was necessary to revise current QF-1 avoided-cost rates for solar QFs because the current rate exceeded the utility’s avoided costs. NorthWestern’s request was based on an assertion that the current QF-1 rate had caused a significant number of large out-of-state QF developers to seek new QF-1 projects in Montana.13

¶19 NorthWestern’s application sought to abandon the proxy methodology, used to calculate avoided costs, and instead urged the PSC to adopt the “peaker” method, delineating its avoided-cost rate into energy and capacity values. NorthWestern further proposed that avoided-capacity costs be based on the levelized cost of planned internal combustion engines,14 scheduled to come online in 2019, consistent with its 2015 resource procurement plan.15 In addition, NorthWestern proposed that QF capacity contribution be measured using an 85/10 exceedance methodology in which a QF’s capacity contribution

13 NorthWestern continues to assert that the current QF-1 rate would effectively cause a “gold rush” in renewable energy development in Montana, leaving customers on the hook for an oversaturation of renewable energy. At the time NorthWestern filed its emergency application, it had executed five power purchase agreements with small solar QFs, had an additional 43 active interconnection requests for 3 MW facilities under various stages of study, and another 75 interconnection requests for 3 MW solar facilities in pre-application phases.

14 An internal combustion engine is analogous to a diesel engine, using combustion of natural gas to drive pistons to generate electricity. Usually, internal combustion engines produce intermittent combustion, allowing these engines to swiftly provide incremental electricity, even when the electric grid has no power. Because these units can start and stop quickly and operate at partial loads, they have become increasingly important in areas with high shares of renewable electricity generation from wind and solar resources.

15 To determine avoided-energy costs, NorthWestern advocated for the use of a proprietary PowerSimm computer model to calculate resource-specific avoided-cost rates. The PowerSimm model assigns values to a QF’s energy based on a forecast of that QF’s hours of generation, load, market prices, and fuel prices over a 25-year period.

14

is measured based on its performance at 85% or higher of its nameplate energy output during the top 10% of peak hours in NorthWestern’s territory. Applying this new methodology, NorthWestern concluded that solar QFs only contributed approximately 9.6% of nameplate capacity, far less than the 38% capacity contribution value under then-current QF-1 rates. NorthWestern did not propose changes to 25-year maximum- length contracts.

¶20 On June 6, 2016, the Consumer Counsel and VS-MEIC filed petitions for general intervention. On June 9, 2016, the PSC held a hearing concerning NorthWestern’s request to suspend the QF-1 tariff rate. On July 25, 2016, the PSC issued Order No. 7500, granting NorthWestern’s request for a temporary suspension of the scheduled QF-1 tariff for small solar QFs. On October 26, 2016, over 4 months after the deadline for interested parties to intervene, the PSC requested the submission of supplemental testimony addressing additional issues not raised by the parties, including whether implementing 25-year maximum-length contracts imposed undue forecast risk on ratepayers and whether adopting maximum contract lengths shorter than 25 years was reasonable. On June 22, 2017, at a properly noticed work session, the PSC voted to reduce the maximum contract length available to QFs from 25 years to 10 years, with a mandatory rate adjustment after 5 years. The PSC also voted to adopt standard-offer rates for QFs even lower than those proposed by NorthWestern, reducing pre-existing rates by more than half. Order No. 7500c

¶21 On July 21, 2017, the PSC issued Final Order No. 7500c, adopting its work session decision.

15

Avoided Costs

¶22 The PSC decided to retain the proxy methodology to estimate avoided costs. However, the PSC used separate planned generating resources for determining avoided energy and avoided-capacity costs. The PSC declined to use the internal combustion engine unit as the avoided-energy cost proxy resource, even though this unit was the next identified resource in NorthWestern’s 2015 resource procurement plan and NorthWestern was operating at a 28% capacity deficit during peak energy hours. The PSC reasoned that the internal combustion engine unit was a baseload capacity unit, and therefore, “would not be reasonable to estimate avoidable incremental energy costs for all time periods based on a resource that is projected to dispatch economically less than a quarter of the time.”16 Instead, the PSC estimated avoided-energy costs by relying on the blended market + combined cycle proxy method applied in previous orders, again using the combined cycle combustion turbine as its avoided-energy resource. Although the combined cycle combustion turbine was not scheduled to become operational until 2025, the PSC did not

16 Curiously, NorthWestern did not advocate in its Application for a combined cycle combustion turbine resource as the appropriate avoided-cost resource. Instead, as evidenced by NorthWestern’s Prefiled Direct Expert Testimony from expert John Bushnell, proposed updated avoided costs of capacity rates were “calculated using the levelized capital cost of an Internal Combustion Engine [] generation unit to be built in 2019.” Only in NorthWestern’s rebuttal testimony did it advance the argument that the internal combustion engine generation unit was not the appropriate resource to measure avoided capacity. Bushnell explained that he originally included an internal combustion energy unit because “it is the capacity resource selected in the 2015 Electricity Supply Resource Procurement Plan[].” The PSC rejected NorthWestern’s filing in Order No. 7338b, also a QF-1 proceeding, because NorthWestern included a resource not included in its most recent resource procurement plan. Accordingly, Bushnell stated that he used an internal combustion engine in this case “to avoid the possibility that the [PSC] would reject this filing on similar grounds.”

16

discuss the timing issue with using the combined cycle resource as the avoided-energy cost resource when the internal combustion engine units would be available 6 years earlier.

¶23 The PSC adopted NorthWestern’s proposal to base avoided-capacity costs on a hypothetical aeroderivative combustion turbine,17 scheduled to come online in 2018. The PSC concluded that the aeroderivative unit, while not included in the preferred portfolio of NorthWestern’s 2015 resource procurement plan, provided a reasonable proxy unit to compensate QFs for capacity contributions because of its similar timeframe to that of when the internal combustion engine units would become operational.

¶24 The PSC declined to adopt either NorthWestern’s or VS-MEIC’s proposals for calculating QFs’ capacity contributions, opting instead for the novel Southwestern Power Pool18 (“SPP”) methodology. The SPP methodology requires that the PSC:

(a) Assemble all available hourly net power output (MWH) data measured at the system interconnection point.

(b) Select the hourly net power output value occurring during the top 3% of load hours for the SPP Load Serving Entity for each month of each year for the evaluation period.

(c) Select the hourly net power output value that can be expected from the facility 60% of the time or greater. For example, for a 5 year period with the 110 hourly net power output values ranked from highest to lowest the capacity of the facility will be the MW value in the 65th data point.

17 Aeroderivative combustion turbines are designed from jet engines so that fuel and air are mixed and then ignited to achieve the desired output. Aeroderivative engines are typically designed to make use of a continuous, rather than an intermittent, combustion process. However, aeroderivative turbines recently have increased in popularity for use in electrical power applications for primarily peak and intermittent purposes rather than base power generation.

18 The SPP is a collective of regional power companies throughout the central United States.

NorthWestern is not a member. However, NorthWestern has used the SPP methodology in its South Dakota service area.

17

(d) A seasonal or annual net capacity may be determined by selecting the appropriate monthly MW values corresponding to the Load Serving Entity’s peak load month of the season of interest (e.g., 22 hours for a typical 30 day month and 110 hours for a 5 year period).

In Order No. 7500c, the PSC explained that the SPP methodology involves these steps:

(1) Collecting data for hourly net output (from actual production or, if historical data is not available, applicable proxy sources);

(2) Selecting the net output data during the top 3% of load hours for the load-serving entity for each month of each year of an evaluation period;

(3) Selecting the net hourly output value that can be expected from the facility 60% of the time or greater;

(4) Calculating the annual capacity value (which is the specific measure sought for the QF-1 tariff) by selecting the peak-load month of each year, then selecting and aggregating the top 3% of load hours for each of the peak-load months in the evaluation period;

(5) Sorting of the resulting top 3% hourly values from highest to lowest; and

(6) Determining the facility’s net capacity value—the value which is the facility output that corresponds with the 60th percentile of the sorted list.

The PSC accepted NorthWestern’s interpretation and application of the SPP method. Applying this interpretation to the 10 years of solar and load data available, the PSC and NorthWestern calculated that solar QFs had a solar capacity contribution value of 6.1%.19

19 NorthWestern provided a detailed description of how it interpreted and applied the SPP methodology to the 10 years on record, 2006-2015. NorthWestern first isolated its “on peak” period for these 10 years, defined in NorthWestern’s QF-1 Tariff as the “heavy load hours”— Monday through Saturday between 7:00 a.m. to 10:00 p.m. for the months of January, February, July, August, and December—resulting in approximately 2,038 hours per year. Next, it determined the peak load month for each year. Then, the top 3% of load hours for each of those 10 peak load months were isolated and sorted from largest to smallest and the generation amount exceeded at least 60% of the time was identified. In sum, a nameplate capacity value of 6.1% was determined based on only 22 hours per year. A solar capacity contribution value of zero was assigned to the remaining four of the five on-peak months.

18

The PSC explained that it “interpret[ed] SPP’s method to rely upon the selection of top load hours from only the peak-load month of each year, i.e., 22 hours per year, in a multi-year measurement period . . . .” The PSC concluded that VS-MEIC misapplied the SPP because it used a 60% exceedance value for 10% of all hours in the record—876 hours per year—resulting in a capacity contribution value of 39%.

Carbon Adder

¶25 The PSC declined to use a carbon emission adjustment for calculating avoided-cost estimates when establishing QF-1 standard rates. The PSC acknowledged that it was departing from its own recent precedent in doing so. The PSC explained that excluding carbon costs was reasonable given that “the political forces that once indicated environmental regulatory action at the federal level was likely in the reasonably foreseeable future has diminished and, accordingly, the likelihood of carbon emissions regulation has decreased.” The PSC further stated that “The estimation of avoided costs necessarily entails an assessment of the probabilities, magnitudes, and associated risk of future events that may impact a utility’s avoidable incremental costs of service, and the [PSC] exercises considerable discretion in performance of this task.” By declining to include carbon altogether, the PSC rationalized its departure from its past practice as a “justifiable adjustment to a changed regulatory environment and [] a reasonable recalibration of [its] expectations of the risk associated with unknown, and unknowable, potential regulatory actions at the federal level.”

19

Contract Length

¶26 The PSC found that 25-year maximum contract lengths exposed customers to undue forecast risk, increasing the possibility that customers will pay above-market prices for the output of QFs. The PSC reduced maximum standard-rate contracts for solar QFs from 25 years to 10 years, with a mandatory rate adjustment after 5 years. The PSC concluded that 10 years would provide sufficient encouragement for QF development while adequately mitigating forecast risk for customers. The PSC justified its decision largely on decisions from Idaho and North Carolina, in which each state greatly reduced QF contract lengths to between 2 and 15 years. The PSC also applied a symmetrical treatment to contract lengths for NorthWestern’s own non-QF resources given that NorthWestern’s own resources “contribut[e] to the very risk that they purportedly seek to offset here.” Motions for Reconsideration and Staff Memo

¶27 On July 31, 2017, VS-MEIC filed motions for reconsideration. On October 2, 2017, the PSC’s staff released a memorandum, recommending the PSC reconsider its decision regarding contract lengths. Staff suggested that the PSC instead establish maximum contract lengths of “at least 15 years, but preferably 20 years[.]”20 Staff did not endorse forecast risk as a justifiable consideration for shortening contract lengths. Staff further

20 On June 16, 2017, prior to Order No. 7500c, the PSC staff also released a memorandum.

Like its October memorandum, staff recommended the PSC incorporate a carbon emissions adjustment in setting the avoided-cost rate consistent with the PSC’s decision in Crazy Mountain Wind, which also considered the uncertainty of carbon regulation in light of the 2016 presidential election. Staff further recommended reducing QF-1 maximum-length contracts to 15-20 years as a compromise between encouraging QF development and minimizing forecast risk in determining avoided-energy costs.

20

explained its reluctance to rely on QF contract length policies in other states to justify decisions in this docket “as the facts and the records developed in other states are different than those now before the [PSC].” Staff found that Montana’s statute requiring long-term contracts to enhance the economic feasibility of QFs, § 69-3-604, MCA, is “central to the determination of contract length in Montana, and without knowledge of contextual policy for QFs in other states, staff recommends against relying on state comparisons for a determination of appropriate QF contract length.”

¶28 Staff also advised the PSC to reconsider its decision to exclude a carbon dioxide emission adjustment from its avoided-cost estimate because the PSC did not explain how the facts of this case differed from those in Crazy Mountain Wind. Absent such an explanation, staff believed the PSC’s decision to exclude a carbon dioxide emission adjustment was arbitrary. Staff further recommended that the PSC maintain the adjusted carbon cost adopted in Crazy Mountain Wind, delaying the onset of the carbon adder to 2025. However, staff concluded that the PSC’s calculations of QF capacity contribution and avoided-energy-cost estimates were reasonable and supported by the record evidence. Order No. 7500d

¶29 On November 24, 2017, the PSC issued Order No. 7500d, order on reconsideration. The PSC retained its decision to exclude carbon dioxide emissions from the QF-1 avoided-cost rate because of continued uncertainty surrounding future emissions pricing and heightened federal efforts to deregulate carbon emissions following Crazy Mountain Wind. The PSC affirmed its calculation of avoided costs, finding that the avoided-capacity cost calculated for solar QFs reflects the acquisition of an aeroderivative turbine in 2018

21

such that NorthWestern does not defer capacity payments to 2025. The PSC also declined to vacate its application of the SPP methodology to calculate avoided costs for solar QFs, reasoning that NorthWestern is a winter and not a summer peaking utility, and VS-MEIC failed to aggregate the annual data before measuring the exceedance level.

¶30 The PSC modestly increased the maximum contract length from 10 to 15 years and eliminated the 5-year rate adjustment, finding a 15-year maximum contract length was reasonable, in the public interest, and appropriately balanced QFs’ needs for certainty with the risk to consumers. The PSC acknowledged that NorthWestern and the Consumer Counsel did not provide evidence as to how shortened contract lengths would provide sufficient certainty regarding potential investment returns or enhance the economic feasibility of QFs. The PSC also found that VS-MEIC generally supported a QF contract length of at least 15 years. The PSC declined to apply the definition of “long-term” in Admin. R. M. 38.5.8202(7) to QF contracts, finding the definition fits more appropriately in the context of electricity supply resource planning and procurement and is therefore of limited applicability to QF contracts. The PSC also concluded that North Carolina’s decision to limit QF contracts to 15 years provided support for the PSC to adopt a similar decision because North Carolina operates under a similar statutory regime as Montana, is responsive to the same risk to consumers the PSC is attempting to mitigate, and has a similar biennial ratemaking docket. District Court

¶31 On December 23, 2017, VS-MEIC filed a Petition for Judicial Review in Montana’s Eighth Judicial District Court. On February 21, 2018, the District Court granted WINData,

22

LLC’s21 and the Consumer Counsel’s motions to intervene. This case was ultimately consolidated for judicial efficiency with three other cases. On September 7, 2018, the court heard oral argument.

¶32 On April 2, 2019, the District Court issued an order vacating and modifying PSC Order Nos. 7500c and 7500d. The court found that the PSC’s decision to reduce QF maximum-length contracts to 15 years was arbitrary and unreasonable because the PSC lacked substantial evidence necessary to determine that 15-year contracts were sufficient to enhance small QFs’ economic feasibility. In addition, the court held that the PSC failed to consider the combined negative impact on QF development resulting from reduced contract lengths alongside drastically reduced standard-offer rates, and the PSC failed to provide a reasoned analysis in departing from its past precedent.

¶33 The court also found that the PSC arbitrarily and unreasonably failed to compensate QFs for avoided-energy costs because it did not include a carbon adder in the avoided-cost rate, did not provide a reasoned analysis in departing from its past precedent, failed to follow its own technical staff, and failed to consider NorthWestern’s cost of operating new generating resources that the utility planned to acquire in 2019. Finally, the court held that the PSC arbitrarily failed to compensate solar QFs for their capacity contribution because it discounted Northwestern’s substantial summertime capacity needs and misapprehended the effect of evidence of regional peak demand data reflecting a summertime peak demand.

21 WINData, LLC is a Montana engineering company providing project and site development services to the wind industry. WINData originally filed a motion to intervene before the PSC, but its request was denied as untimely.

23

¶34 NorthWestern and the PSC, joined by the Consumer Counsel, appealed. NorthWestern also filed a motion in District Court to stay judgment pending this appeal. The District Court denied the motion to stay. NorthWestern then filed a motion for relief in this Court. On August 6, 2019, this Court issued an order staying the matter pending resolution of the appeal. Vote Solar v. Mont. Dep’t of Pub. Serv. Regulation, No. DA 19-0223, Or. (Mont., Aug. 6, 2019). On February 26, 2020, this Court heard oral argument as to whether the District Court erred in vacating and modifying Order Nos. 7500c and 7500d.

STANDARDS OF REVIEW

¶35 The Montana Administrative Procedures Act (“MAPA”) standards of review govern an administrative appeal of an agency decision in a contested case. Section 2-4-704, MCA; Whitehall Wind, LLC v. Mont. Pub. Serv. Comm’n, 2015 MT 119, ¶ 7, 379 Mont. 119, 347 P.3d 1273 (Whitehall Wind II). This 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 I, ¶ 15; McGree Corp. v. Mont. Pub. Serv. Comm’n, 2019 MT 75, ¶ 6, 395 Mont. 229, 438 P.3d 326 (citing Nw. Corp. v. Mont. Dep’t of Pub. Serv. Regulation, 2016 MT 239, ¶ 25, 385 Mont. 33, 380 P.3d 787). The agency decision must be judged on the grounds and reasons set forth in the challenged order(s); no other grounds should be considered. Montana-Dakota Utilities Co. v. Mont. Dep’t of Pub. Serv. Regulation, 223 Mont. 191, 196, 725 P.2d 548, 551 (1986) (citation omitted).

24

¶36 The court may not substitute its judgment for that of the agency in weighing the factual evidence. Section 2-4-704(2), MCA. The court may reverse or modify an agency decision if the substantial rights of the appellant have been prejudiced because the agency’s decision exceeds its statutory authority, is affected by legal error, clearly erroneous in light of the whole record, arbitrary or capricious, or characterized by an abuse of discretion. Section 2-4-704(2)(a)(ii), (iv), (v), (vi), MCA. 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 made. Nw. Corp., ¶ 27. While agencies possess specific, technical, and scientific knowledge exceeding that of this Court, an agency must still articulate a satisfactory explanation for its actions and provide a rational connection between the facts found and the choice made. Mont. Envtl. Info. Ctr. v. Mont. Dep’t of Envtl. Quality, 2019 MT 213, ¶ 26, 397 Mont. 161, 451 P.3d 493; Clark Fork Coal. v. Mont. Dep’t of Envtl. Quality, 2008 MT 407, ¶ 47, 347 Mont. 197, 197 P.3d 482 (Clark Fork Coal. I).

¶37 In addition, pursuant to § 69-3-402, MCA, a party in interest dissatisfied with an order of the PSC setting or fixing rates may proceed in the district court to vacate the order on the grounds that it is unlawful or unreasonable. An agency action is arbitrary if it fails to consider relevant factors, including the standards and purposes of the statutes the agency administers. Clark Fork Coal. I, ¶ 21. This Court will not defer to an agency’s incorrect or unlawful decisions but will only defer to an agency action within permissible statutory bounds. North Fork Pres. Ass’n v. Dep’t of State Lands, 238 Mont. 451, 459, 778 P.2d 862, 867 (1989). We afford an agency’s interpretation of its own rule great weight and

25

will defer to that interpretation unless it is plainly inconsistent with the spirit of the rule, considering a range of reasonableness permitted by the regulation’s wording. Mont. Envtl. Info. Ctr., ¶ 23.22 DISCUSSION

¶38 Issue One: Whether the District Court erred when it determined that the PSC arbitrarily and unlawfully reduced solar QF standard-offer rates by excluding carbon dioxide emissions costs and NorthWestern’s avoided costs of operating its internal combustion engine resource units from the avoided-cost rate.

¶39 The District Court held that the PSC’s decision to reduce the standard-offer QF-1 rates was arbitrary and unreasonable because the PSC failed to consider future carbon costs and failed to provide a reasoned decision in departing from its recent precedent. In addition, the court found that the PSC unreasonably failed to consider NorthWestern’s cost of operating its new internal combustion engine resources, scheduled to come online in 2019, when setting the avoided-cost rate. We address these considerations separately.

Carbon Adder

¶40 The PSC argues that the District Court erred because it has never included carbon costs in determining the QF-1 tariff rate as its past decisions all assumed a delayed onset date of carbon dioxide pricing. The PSC also argues that carbon forecasting is inherently speculative in nature, particularly following the election of a presidential administration hostile to carbon regulation, such that it was not required to include a carbon adder in the

22 The Dissent posits that there is sufficient record-based evidence to support deferral to the agency decision. However, as demonstrated, the PSC decisions here were not predicated on the underlying record; rather, as the District Court noted, they were clearly arbitrary and capricious and failed to articulate reasoned decisions between the facts, the law, and the choices made.

26

avoided-cost rate.23 NorthWestern asserts that including future carbon costs within the avoided-cost rate is a “discretionary factor” that the PSC need not consider when setting rates.

¶41 PURPA requires that utilities purchase electricity generated by QFs at rates that are “just and reasonable” to the consumer, “in the public interest,” and nondiscriminatory to the QF. 16 U.S.C. § 824a-3(b). Thus, in setting contract rates, the agency must fairly balance the interests of its ratepayers with that of the QF such that it complies with PURPA and “encourages” renewable energy development while making the ratepayer indifferent as to the energy source. S. Cal. Edison Co., 71 FERC at 62,080. In doing so, a utility purchasing electricity must compensate the QF at a rate equal to the utility’s full avoided cost—that is, “the incremental costs to an electric utility of electric energy or capacity or both which, but for the purchase from the qualifying facility or qualifying facilities, such utility would generate itself or purchase from another source.” 18 C.F.R. § 292.304(b)(2); 16 U.S.C. § 824a-3(d); Am. Paper Inst., 461 U.S. at 406, 103 S. Ct. at 1924; 18 C.F.R.

23 Additionally, the Consumer Counsel argues that small QFs may still monetize avoided future carbon dioxide costs without including such costs in the avoided-cost rate because QFs can retain and sell renewable energy credits (“RECs”). RECs are certificates representing environmental attributes of energy generated by renewable energy sources, created as a means for states to measure progress toward compliance with Renewable Energy Portfolio Standards. Buying RECs is not equivalent to buying electricity, however, because RECs do not provide a guarantee that a particular amount of carbon emission is avoided; thus, buying and selling RECs has no effect on the avoided-cost rate. See Californians for Renewable Energy, 922 F.3d at 940 (noting that RECs are not covered under PURPA and therefore do not factor into the avoided-cost determination). In fact, the PSC has previously rejected such arguments in establishing the avoided-cost rate, acknowledging that avoided-carbon costs are a consideration separate and independent of the value of RECs. Order No. 7505b, ¶¶ 58-59. And the PSC did not base its decision in this docket to exclude carbon costs from the avoided-cost rate on the ability of QFs to retain and sell RECs. Accordingly, we decline to review the PSC decision based on a reason not set forth in Order Nos. 7500c and 7500d. See Montana-Dakota Utilities Co., 223 Mont. at 196, 725 P.2d at 551.

27

§ 292.101(b)(6); Admin. R. M. 38.5.1901(2)(a). FERC has held that carbon costs are a necessary aspect of a utility’s avoided costs if those costs are incurred by the utility. Cal. Pub. Utilities Comm’n, 133 FERC ¶ 61,059, 61,268 (Oct. 21, 2010).

¶42 While the PSC has certainly delayed the onset date of a carbon adder in recent QF dockets, the PSC acknowledged in Order No. 7500d that, since its 2012 order adopting present QF-1 rates, it has regularly included estimated avoided carbon emission costs as part of the avoided-cost projections for QF and non-QF projects. In issuing Order No. 7323k, addressing the avoided-cost rate for hydro QFs, the PSC described the essential nature of including carbon emissions pricing in the avoided-cost rate, stating that NorthWestern must “consider risk costs related to potential carbon emissions regulation when planning for and acquiring new resources.” Order No. 7323k, ¶ 88 (citing Admin. R. M. 38.5.8213). The PSC further provided that “One of the key inputs in NorthWestern’s comparative cost modeling and [discounted cash flow] analysis was the future cost of carbon emissions. Although highly uncertain, all parties agreed that future carbon costs should not be considered zero.” Order No. 7323k, ¶ 81 (emphasis added).

¶43 Similarly, in Crazy Mountain Wind, the PSC acknowledged that carbon-free qualifying facilities avoid future carbon costs such that including a reasonable estimate of those avoided costs is an essential component of nondiscriminatory implementation of PURPA. Order No. 7505b, ¶ 59. In Crazy Mountain Wind, the Consumer Counsel argued that carbon dioxide should be excluded from the avoided-cost rate due to the “speculative nature” of the calculation, particularly so given the results of the 2016 presidential election. Order No. 7505b, ¶ 50. While recognizing the fundamentally abstract nature of carbon

28

price forecasting, the PSC declined to adopt the Consumer Counsel’s request. The PSC suggested that if the Consumer Counsel continued to argue to limit carbon dioxide emission price forecast risk in future proceedings, it “should propose language that could be included in a contract that would allow an emission price to be clearly ascertained when it appears in the market and to correspond with a clear and forecastable adjustment.” Order No. 7505b, ¶ 60. Otherwise, the PSC explained that the Consumer Counsel’s proposals would violate PURPA’s requirement that rates be forecast at the time when the obligation is incurred. Order No. 7505b, ¶ 60 (citing 18 C.F.R. § 292.304(d)(2)(ii)).

¶44 Here, the PSC declined to include a carbon adder when setting the avoided-cost rate, explaining “The reason for changing [its] practice relates to an assessment that the political forces that once indicated environmental regulatory action at the federal level was likely in the reasonably foreseeable future has diminished and, accordingly, the likelihood of carbon emissions regulation has decreased.” We conclude that the record evidence does not support the PSC decision, such that its exclusion of consideration of carbon dioxide emissions cost from the avoided-cost rate violates PURPA.

¶45 Setting QF rates despite market volatility and uncertainty inherently underlies the entire purpose of PURPA: to encourage and incentivize renewable energy development over a long-term contract in order to escape fossil fuel dependency. Am. Paper Inst., 461 U.S. at 405, 103 S. Ct. at 1924; FERC, 456 U.S. at 750, 102 S. Ct. at 2132. While carbon price forecasting may be innately difficult, to assign carbon pricing a value of “zero” because of its speculative nature simply does not compensate QFs for the full avoided-cost rate. See Order No. 7505b, ¶¶ 59, 60. Nor does conjecture about an increasingly hostile

29

political climate establish a “clear and forecastable adjustment,” necessary to comply with its own precedent. Order No. 7505b, ¶ 60.24 And, given that hydro and wind QFs enjoy a carbon adder in its avoided-cost rates, to exclude carbon costs from the avoided-cost rate exclusively for solar QFs, particularly given greatly reduced maximum-length contracts,25 is discriminatory to solar QFs and in violation of PURPA. 16 U.S.C. § 824a-3(b). The PSC’s decision to exclude a carbon adder from the avoided-cost rate was unlawful.

¶46 Likewise, the PSC’s justification for its actions was arbitrary. The PSC noted that it was reasonable to depart from past precedent given the “unknowable, potential regulatory actions at the federal level[,]” citing its “authority and technical fact finding expertise to appropriately balance the future risk of carbon costs to be borne by customers.” However, the PSC found such reasoning proffered by the Consumer Counsel in Crazy Mountain Wind, decided only months earlier, to be inadequate. In declining to reconsider including a carbon adder in the avoided-cost rate, the PSC made no effort to distinguish the facts in this case from those in Crazy Mountain Wind or explain why its decision was acceptable

24 As noted in ¶ 15 of the Opinion, prior contracts that have been approved postponed the imposition of carbon adders to future years when the obligation may actually be incurred.

25 In fact, a hot mic recording at a properly noticed work session captured PSC commissioner Bob Lake, in discussion with a PSC staff member, stating that reductions of contract lengths, together with modified avoided-cost rates, should effectively kill QF development entirely.

30

under the facts in this docket.26 Indeed, beyond its brief references to the speculative nature of carbon pricing and a presidential administration hostile to carbon regulation, the PSC did not attempt to explain why carbon emissions should not be considered in the avoided-cost rate. Mere speculation based on political forecasting hardly constitutes technical or scientific knowledge worthy of deference.27 See Mont. Envtl. Info. Ctr., ¶ 26. Nor does an agency’s reference to its own technical expertise constitute a reasoned decision. See Clark Fork Coal. I, ¶¶ 21, 47. The District Court did not err in concluding that the PSC’s decision to exclude carbon emissions from the avoided-cost rate was arbitrary and unlawful.

Avoided-Energy Costs

¶47 In applying the proxy methodology to calculate the avoided-cost rate, the PSC utilized different avoided-cost resources to separately calculate avoided-capacity and avoided-energy costs, calculating avoided-capacity costs based on the 2018 aeroderivative unit and avoided-energy costs based on the 2025 combined cycle combustion turbine unit.

26 The PSC’s staff also noted in its October memorandum that (1) the record evidence in this case did not reflect additional changes in the political landscape since the PSC’s decision in Crazy Mountain Wind, and (2) the PSC failed to distinguish the facts in this docket from Crazy Mountain Wind, relegating its decision to exclude a carbon adder arbitrary. Although an agency has no duty to adhere to the recommendations of its technical staff, contrary to the finding of the District Court, staff’s repeated warnings further illustrate that the record evidence did not support excluding a carbon adder from the avoided-cost rate and the PSC did not provide a reasoned decision for its actions.

27 Furthermore, the PSC’s speculation is without merit. While the PSC based its decision to exclude carbon costs from the avoided-cost rate on the Environmental Protection Agency’s (“EPA”) proposed repeal of the Clean Power Plan, the EPA notes in its proposed repeal that the Clean Air Act still requires it to regulate power plant carbon emissions. Proposed Rule, 82 Fed. Reg. 48,035, 48036-37 (Oct. 16, 2017).

31

The District Court held that the PSC unreasonably failed to consider the cost of operating new generating resources that the utility planned to acquire in 2019 even though the record evidence demonstrated that solar QFs would allow NorthWestern to avoid such costs in high-demand hours.

¶48 The PSC first argues that VS-MEIC failed to exhaust its administrative remedies because of its evolving theory of the case before the PSC such that judicial review of this issue is not available. Section 2-4-702, MCA, requires exhaustion of administrative remedies and states that judicial review is only available to a person aggrieved by a final agency decision. See also Wilson v. Dep’t of Pub. Serv. Regulation, 260 Mont. 167, 172, 858 P.2d 368, 371 (1993). The exhaustion doctrine serves to provide administrative agencies an opportunity to utilize their expertise, correct any mistakes, and avoid unnecessary judicial intervention. Buckingham v. Sec’y of U.S. Dep’t of Agric., 603 F.3d 1073, 1080 (9th Cir. 2010). A party forfeits argument as to an issue not raised during the administrative process; however, so long as a claimant provides enough clarity such that the decision maker understands the issues raised for the agency to use its expertise to resolve the claim, the claimant will have met this burden. Lands Council v. McNair, 629 F.3d 1070, 1076 (9th Cir. 2010).

¶49 In Order No. 7500c, the PSC estimated avoided costs by relying on the blended market + combined cycle combustion proxy methodology. The PSC, however, did not discuss the timing issue with using the combined combustion turbine unit as a proxy resource to calculate avoided-energy costs. VS-MEIC challenged this decision in its motion for reconsideration, arguing that “Even if it was reasonable for the [PSC] to

32

establish avoided costs based on a [combined cycle combustion turbine], rather than an [internal combustion engine], it was unreasonable and unlawful for the [PSC] to assume that NorthWestern can defer capacity additions to 2025 when its [resource procurement plan] calls for new capacity resources in 2019.” VS-MEIC further argued that the PSC’s order “failed to address the timing of NorthWestern’s capacity needs.”

¶50 In Order No. 7500d, the PSC attempted to retroactively address the timing issue of the delayed energy resource addition, explaining that “the avoided capacity cost calculated for solar QFs reflects the acquisition of an aeroderivative combustion turbine in 2018 . . . reasonably reflect[ing] the timing of NorthWestern’s next planned resource addition and does not assume NorthWestern defers capacity additions to 2025.” At the District Court and now before this Court, VS-MEIC advances fundamentally the same argument that it did before the PSC—using the combined combustion turbine as the proxy resource still fails to compensate QFs for the full avoided costs because of the timing issue created by using this resource in place of the internal combustion engine. VS-MEIC exhausted its administrative remedies as to this issue. It is ripe for review.

¶51 The PSC and NorthWestern also argue that energy and capacity costs are different considerations. The PSC therefore asserts that its decision to adopt the combined cycle combustion turbine in place of the internal combustion engine units was appropriate for its calculation of avoided-energy costs because the combined combustion turbine, unlike the internal combustion engine, is primarily an energy and not a capacity resource. The PSC argues that QFs do not have to defer capacity payments until 2025, when the combined combustion turbine comes online, because it used the 2018 aeroderivative unit as a proxy

33

for the internal combustion engine for calculating capacity payments. VS-MEIC argues that PURPA requires capacity and energy costs to be considered together when setting the avoided-cost rate such that the PSC’s decision to use proxy resources on a different timeline than its next scheduled resource in its 2015 resource procurement was arbitrary and unlawful.

¶52 Again, PURPA requires that QFs must be compensated for the utility’s full avoided costs. 18 C.F.R. § 292.304(b)(2). Avoided costs include the “incremental costs to an electric utility of electric energy or capacity or both which, but for the purchase from the [QF], such utility would generate itself or purchase from another source.” 18 C.F.R. § 292.101(b)(6) (emphasis added). Avoided-capacity costs occur when a utility is reaching maximum demand for its current energy supply throughout its service area and the utility needs additional power to meet demand but building a new centralized power plant would be excessive. FERC Order No. 69 at 12,216 (explaining that the capacity component serves to ensure a reliable production of electricity during high-demand periods). Capacity costs are therefore those costs associated with providing the capability to deliver energy, composed primarily of the capital cost of the facility itself. FERC Order No. 69 at 12,216. Conversely, energy costs are the variable costs associated with the production of electric energy, entailing the costs of fuel and operating expenses. FERC Order No. 69 at 12,216.

¶53 PURPA provides a number of factors to be considered when setting the avoided-cost rate, including “[t]he availability of capacity or energy from a [QF] during the system daily and seasonal peak periods,” and “[t]he relationship of the availability of energy or capacity from the [QF] . . . to the ability of the electric utility to avoid costs, including the deferral

34

of capacity additions and the reduction of fossil fuel use[.]” 18 C.F.R. § 292.304(e)(2), (3). Moreover, in submitting avoided-cost rates, FERC directs utilities to submit energy costs of each planned unit with the estimated capacity cost of planned capacity additions in order “to ensure that the calculation of avoided costs includes the lower energy costs that might be associated with the new capacity.” FERC Order No. 69 at 12,218 (emphasis added). While energy and capacity costs are technically different costs, they are fundamentally two parts of the same equation that must be considered together when calculating avoided costs.

¶54 The PSC creates an arbitrary distinction between avoided capacity and energy costs when setting the avoided-cost rate. To determine the rates for purchase from the QF, the PSC explained that it takes avoided-capacity costs starting in 2019, divides the amount by the expected energy production during on-peak hours, and then adds this amount to avoided-energy costs during those on-peak hours. This explanation implicitly acknowledges that energy and capacity costs are inextricably related. Calculating avoided costs is necessary to compensate a QF for all avoided costs from the time the resource is used as a proxy; if the PSC uses a proxy unit for determining avoided-capacity costs beginning in 2019, then its avoided-energy costs must necessarily be sourced from the same period. See 18 C.F.R. § 292.304(e)(3); FERC Order No. 69 at 12,218. Further, rates for purchase from QFs “must be reasonable and based on current avoided least cost resource data.” Whitehall Wind I, ¶ 21. The District Court did not err in determining that the PSC’s decision to source avoided-energy costs from a resource that will not be online

35

until 6 years after its scheduled 2019 internal combustion engine resource was arbitrary and in violation of PURPA and Montana law.

¶55 Likewise, the PSC’s assertion that the internal combustion engine is primarily a “capacity” resource and not an “energy” resource is also arbitrary. The PSC argues in its opening brief that the internal combustion engine “provides relatively on-demand power which can quickly ramp up and down to meet periods of high demand.” But again, this explanation tacitly concedes that building a new capacity facility allows NorthWestern to produce more energy. And as a capacity-strapped utility, it is unreasonable to assume that NorthWestern is adding capacity resources simply for capacity purposes. In fact, as NorthWestern’s own witness John Bushnell admitted, even if the primary purpose of the internal combustion engine unit is to provide capacity, the resource will also serve to supplement NorthWestern’s energy purchases during high-demand hours. Because the internal combustion engine resources will provide energy, regardless of its primary purpose, this energy must be considered in the avoided-cost rate in order for the PSC to comply with PURPA’s mandate that QFs be compensated for the utility’s full avoided cost. 18 C.F.R. § 292.304(b)(2); Am. Paper Inst., 461 U.S. at 406, 103 S. Ct. at 1924.

¶56 Further, the PSC arbitrarily applied the proxy methodology in calculating avoid costs, limiting its application of this method exclusively to calculate capacity costs. In Order No. 7500d, the PSC approved NorthWestern’s proposal to use the aeroderivative unit as a proxy resource for avoided-capacity costs, describing a calculation to compensate the QF for its “capacity contribution.” The PSC explained that its avoided-cost calculation “results in rates that compensate QFs for their capacity contributions consistent with the

36

timing of the [internal combustion engine] units identified in NorthWestern’s 2015 [resource procurement plan].” However, “capacity contribution” involves a calculation concerning the SPP method—an entirely different consideration—reflecting a facility’s assumed capacity contribution during peak demand hours. Conversely, when determining avoided costs under the proxy method, the PSC must consider the relationship of availability of capacity (i.e., capacity contribution) to the ability of the electric utility to avoid costs. 18 C.F.R. § 292.304(e)(3). Put differently, a resource’s capacity contribution informs the avoided-cost rate; the avoided-cost rate does not inform capacity contributions. The PSC’s erroneous conflation of these separate considerations does not escape PURPA requirements which mandate that energy and capacity costs be considered together when calculating the avoided-cost rate. 18 C.F.R. § 292.304(e).

¶57 The District Court did not err in concluding that the PSC’s calculation of the avoided-cost rate was arbitrary and unlawful, in violation of PURPA and Montana law.

¶58 Issue Two: Whether the District Court erred when it concluded that the PSC arbitrarily and unreasonably calculated solar QFs’ capacity contribution in determining avoided costs.

¶59 The District Court held that the PSC arbitrarily determined that solar QFs contribute 6.1% of their overall generating capacity to NorthWestern’s capacity needs because the PSC discounted NorthWestern’s substantial summertime capacity needs and disregarded regional peak demand data. The court found that the PSC focused only on a handful of peak demand hours—220 hours over a ten-year period—that reflect primarily infrequent

37

wintertime spikes while overlooking evidence that NorthWestern lacks sufficient capacity to meet peak customer demand in both summer and winter.

¶60 The PSC and NorthWestern argue that the District Court erred because its capacity contribution was reasonable when applying the SPP methodology. VS-MEIC asserts that the PSC’s aggregation of the top 3% of load hours caused NorthWestern to exclude summertime peak hours. In fact, NorthWestern’s 2015 resource procurement plan acknowledges that its loads reflect dual-peaking periods, with maximum annual peak demand occurring in both winter and summer seasons. 2015 Electricity Supply Resource Procurement Plan at 2-12. The core of the issue, however, stems from the PSC’s interpretation and application of the SPP methodology.

¶61 The PSC and NorthWestern provided some of the SPP planning criteria detailing the steps used to calculate capacity contribution. These criteria require the PSC to:

(a) Assemble all available hourly net power output data;

(b) Select the hourly net power output values occurring during the top 3% of load hours for the SPP Load Serving Entity for each month of each year for the evaluation period; and

(c) Select the hourly net power output value that can be expected from the facility 60% of the time or greater.

Additionally, subsection (d) provides that “A seasonal or annual net capacity may be determined by selecting the appropriate monthly MW values corresponding to the Load Serving Entity’s peak load month of the season of interest[].” The PSC accepted NorthWestern’s application of this criteria, which explained that it applied these criteria by “using the top 3% of load hours for each month, and a 60% generation exceedance value

38

for these hours in all of the annual peak months of the calculation period.” After determining the peak load month for each year, the PSC further isolated the top 3% of load hours for each of those 10 peak loads months, sorting all hours by generation from largest to smallest to determine the generation amount exceeded at least 60 percent of the time.

¶62 The plain language of the SPP Planning Criteria demonstrates that subsections (b) and (d) are different methods to aggregate data. However, the PSC conflates these two factors, first using subsection (d) to limit its evaluation to only the highest peak load months per year, and then, based on subsection (b), further limits its review to the top 3% of those load hours of those 10 months. The PSC failed to clarify why its 3% isolation of the highest hourly loads for each of those peak load months is only applied to the 10-peak load annual months, when the 3% section clearly applies to “each month of each year for the evaluation period.” Further, the PSC did not justify how the annual calculation alternative in subsection (d) can first limit the months considered to only those peak load months of the season of interest before applying the 3% aggregation of the top load hours in subsection (b). The PSC also failed to explain why it utilizes part of the subsection (b) VS-MEIC asserts is proper for determining capacity contribution, but disputes applying the section fully by instead limiting its application by first isolating the peak load month for each year, thus resulting in low capacity contributions.

¶63 The SPP methodology is a novel methodology not yet applied in Montana. While it is certainly within the authority of the PSC to adopt an entirely new methodology to calculate capacity contribution, it must still consider all relevant factors and adequately explain why it interpreted the methodology in such a way. See Mont. Envtl. Info. Ctr., ¶ 26.

39

Moreover, the PSC cannot adopt a new methodology simply to circumvent PURPA’s objective to encourage alternative energy development of small power production facilities. 16 U.S.C. § 824a-3(a); Am. Paper Inst., 461 U.S. at 405, 103 S. Ct. at 1924. Because the PSC’s interpretation contravenes the plain language of the SPP methodology and the PSC did not explain why it interpreted the SPP as such, its interpretation and application of the SPP methodology in calculating avoided-capacity costs was arbitrary and unreasonable.

¶64 The District Court held that the PSC misapplied the SPP methodology because it excluded regional peak demand in the data and NorthWestern is a dual peaking utility. We conclude, however, that the PSC instead misapplied the methodology by acting contrary to the plain language of the SPP criteria and did not articulate a satisfactory explanation for its actions. Clark Fork Coal. I, ¶ 47. We will affirm the district court if it reaches the right result, even for the wrong reason. State v. Ellison, 2012 MT 50, ¶ 8, 364 Mont. 276, 272 P.3d 646. The District Court did not err in determining that the PSC’s application of the SPP methodology was arbitrary and unreasonable.

¶65 Issue Three: Whether the District Court erred when it determined that the PSC arbitrarily and unreasonably reduced maximum-length contracts to 15 years for small solar QFs.

¶66 The District Court held that the PSC lacked substantial evidence necessary to determine that 15-year contracts are sufficiently long term to enhance the economic feasibility of qualifying small power production facilities in violation of PURPA and Montana law. The PSC and NorthWestern argue the District Court erred because Congress and FERC did not set specific contract-length requirements for QFs; the PSC implemented

40

this policy consistent with testimony offered by VS-MEIC; and the District Court improperly relied on PSC staff recommendations in concluding that the PSC’s decision was arbitrary and unlawful.

¶67 VS-MEIC argues that reducing contract lengths to 15 years violates PURPA and Montana law because it does not support, let alone enhance, the economic feasibility of QFs. VS-MEIC further asserts that there was an insufficient evidentiary record to support its contract-length decision; that the PSC impermissibly relied on a North Carolina Public Utilities Commission decision; that the PSC did not provide adequate notice prior to shortening contract lengths; and that the decision to cut contract lengths must be analyzed in light of the reduced QF rate.

¶68 PURPA and Montana law require that long-term contracts between the utility and the QF must be encouraged in order to enhance the economic feasibility of the QF. 16 U.S.C. § 824a-3(a); FERC Order No. 69 at 12,226; § 69-3-604(2), MCA. Contract lengths are important to consider in light of standard-offer QF rates; FERC provides that long-term contracts allow for any overestimations and underestimations of avoided costs to balance out so that neither the utility nor the QF is negatively affected by market fluctuations, creating certainty with regard to return on investment in new technologies for qualifying facility investors. FERC Order No. 69 at 12,224. Neither PURPA, FERC, nor Montana statute offers a definition of “long term.”

¶69 Here, the PSC concluded that 15-year maximum-length contracts was supported by the record evidence. We disagree. In fact, driven largely by its unilateral decision to request additional testimony on contract lengths far along in the docket and its altered

41

reasoning set forth in Order No. 7500d following its staff recommendations, the PSC’s decision to adopt 15-year maximum QF contract lengths was based almost entirely on a 2014 North Carolina Utilities Commission decision. However, the PSC lacks any intimate knowledge regarding QF development policies in North Carolina or other states. Indeed, we find the PSC’s justification especially dubious given its wholesale rejection of out-of-state decisions as a consideration when setting the avoided-cost rate.

¶70 Further, the PSC did not explain why 15 years was feasible in balancing a need with certainty regarding a return on investment in Montana, nor could it, given that the PSC also found that NorthWestern and the Consumer Counsel failed to submit any evidence supporting such a decision. And, it is undisputed that NorthWestern’s own resources enjoy contracts for at least 25 years. The PSC’s decision was clearly erroneous and not supported by substantial evidence in the record. Section 2-4-704(2)(v), MCA.

¶71 We are similarly unpersuaded by the PSC’s and NorthWestern’s assertions that VS-MEIC “generally supported” 15-year contract terms. While VS-MEIC submitted expert testimony acknowledging that in some instances 15-year contracts may be the minimum length necessary to secure financing, the PSC opportunely glosses over the fact that VS-MEIC’s argument rests on the contingency that shorter contracts may only be sufficient assuming there are appropriate standard-offer rates. Moreover, in misstating VS-MEIC’s position, the PSC attempts to lessen PURPA’s directive, arguing that contract lengths need only be long enough to “obtain” financing. But again, contracts must be long enough to “encourage” and “enhance” the economic feasibility of QFs, not merely long enough such that financing may be “possible.” 16 U.S.C. § 824a-3(a); § 69-3-604(2),

42

MCA. The PSC’s misinterpretation of VS-MEIC’s position does not constitute a reasoned decision. Clark Fork Coal. I, ¶ 21.

¶72 The PSC’s decision further violated PURPA’s requirement that contracts be sufficiently long term to “encourage” and “enhance” QF development. PURPA, FERC, and Montana law are replete with the requirement that service commissions consider both the length of contracts alongside contract prices, recognizing the synergistic effect of these dual considerations. See 18 C.F.R. § 292.304(e)(2)(iii); FERC Order No. 69 at 12,224; § 69-3-604(2), MCA. Obtaining financing for QF projects fundamentally rests on whether investors may ensure a return on investment. By establishing longer contract lengths, investors remain insulated from forecast risk, as overestimations and underestimations of avoided costs may balance out. FERC Order No. 69 at 12,224. However, amidst abridged QF-1 rates, there remains little incentive for financers to fund prospective projects over shortened contract terms.

¶73 To be sure, 15-year contracts, standing alone, are not per se unreasonable. But because the PSC failed to consider shortened contract lengths in conjunction with greatly reduced standard-offer QF-1 rates, 15-year contracts cannot be considered sufficient to encourage and enhance QF development. 16 U.S.C. § 824a-3(a); § 69-3-604(2), MCA. The District Court did not err in determining that the PSC’s decision to reduce maximum solar QF contract lengths was arbitrary and unreasonable, not supported by the record evidence, and did not comply with PURPA and Montana law.

43

CONCLUSION

¶74 For the reasons stated, the District Court did not err in concluding that the PSC’s calculation of the avoided-cost rate was arbitrary and unlawful. The District Court also did not err in concluding that the PSC arbitrarily and unreasonably calculated QF “capacity contribution” values and arbitrarily and unreasonably reduced maximum-length QF-1 contracts to 15 years. The court may reverse or modify an agency decision if the substantial rights of the appellant have been prejudiced because the agency decision violates constitutional or statutory provisions, is made upon unlawful procedure, is affected by other error of law, clearly erroneous in view of the substantial evidence on the whole record, arbitrary or capricious, or characterized by an abuse of discretion. Section 2-4-704(2)(a), MCA. However, NorthWestern has since issued its 2019 resource procurement plan detailing its most recent avoided-cost data. We leave the stay in place and remand to the PSC for consideration of the factors set forth in this Opinion when setting QF-1 standard-offer rates and contract lengths in all future regulatory proceedings.

/S/ MIKE McGRATH

We Concur:

/S/ JAMES JEREMIAH SHEA /S/ INGRID GUSTAFSON /S/ DIRK M. SANDEFUR

44

Justice Beth Baker, concurring in part, dissenting in part.

¶75 I agree generally with the Court that the PSC did not comply with PURPA and Montana’s “mini-PURPA” when it set the standard-offer contract rates and maximum contract lengths for qualifying small (QF-1) solar power producers in July 2017, but I do not join its Opinion in full.

¶76 First, it bears emphasis that the twin goals of non-discrimination against independent renewable electric-energy producers and financial indifference to ratepayers are subsumed in the statutory standards by which the PSC determines rates and conditions of QF-1 contracts. Under § 69-3-604(4), MCA, the PSC is to “us[e] the avoided cost over the term of the contract” to set these rates. As the Opinion discusses at some length, “avoided costs” are the costs the utility would incur “but for the purchase from the qualifying facility[.]” Opinion, ¶ 6 (quoting 18 C.F.R. § 292.101(b)(6)). Energy purchased at avoided-cost rates thus leaves the consumer neutral. See Opinion, ¶¶ 6, 41. I thus disagree with the Dissent’s view that the Court ignores consumer financial indifference. I share the Court’s view instead that the PSC’s determination of avoided costs in this case was faulty on several levels and must be reconsidered.

¶77 Second, however, it bears further emphasis that the statutes confer on the “commission” the authority and responsibility to determine the rates and conditions of QF- 1 contracts. Section 69-3-604, MCA. The dispute between the parties over consumer indifference and the impact on rates—reflected in the Court’s Opinion and the Dissent— highlights why the courts should avoid any specific pronouncement of terms or conditions that will govern such contracts. If, as the Court concludes, the PSC failed to balance a

45

fifteen-year contract length with appropriate consideration of other rate factors or either misapplied or failed to explain adequately how it applied the SPP methodology, the PSC must have latitude to consider each of these failures and decide for itself how to rectify them.

¶78 I cannot agree to affirm the District Court when it not only vacated the PSC’s orders but modified them by imposing specific requirements for the rates and conditions of the QF-1 contracts. For example, the trial court expressly adopted the commission staff’s recommendations to “us[e] the values for contract terms of twenty-five years”; mandated that “the QFs must be compensated for [the] avoided energy costs beginning in 2019”; made a finding that “the average capacity contribution of solar resources is 36%”; and ordered the PSC to “direct NorthWestern Energy to make a compliance filing consistent with the District Court’s findings” within twenty days of its order.

¶79 On appeal, although the Court leaves in place its stay of the District Court’s directives and does not adopt its findings seriatum, it nonetheless affirms the decision. Opinion, ¶ 1. We afford “deference to an agency’s evaluation of evidence insofar as the agency utilized its experience, technical competence, and specialized knowledge in making that evaluation.” Northwestern Corp. v. Mont. Dep't of Pub. Serv. Regulation, 2016 MT 239, ¶ 27, 385 Mont. 33, 380 P.3d 787 (citations omitted). In my view, the Court does not leave adequate room for the PSC, applying its technical competence and specialized knowledge, to consider and balance the “closely intertwined,” Opinion, ¶ 9, factors of contract length, avoided costs, and availability and reliability of the electricity produced in determining new rates and conditions. See § 69-4-604(1)-(4), MCA. I accordingly do not

46

join its Opinion but instead would reverse the District Court’s June 17, 2019 Order to Alter or Amend and direct the court to remand the matter to the PSC for further consideration in light of the flaws the Court identifies in the Commission’s analysis.1 The passage of time means that such consideration would be applied to Northwestern’s next biennial application to update its QF-1 avoided cost tariff. Reversing the District Court, instead of affirming and staying its decision, would provide more clarity that the discretion to consider the technical evidence and to determine the appropriate rates and conditions of QF-1 contracts rests with the PSC.

/S/ BETH BAKER

Justice Jim Rice, dissenting.

¶80 Mandated by law, the most critical factor in the process of setting QF avoided cost tariffs is protecting the customer—the ratepayer—from bearing any increase in utility rates as a result of the tariff. See 16 U.S.C. § 824a-3(b) (“no such rule prescribed under [this section governing tariffs] shall provide for a rate which exceeds the incremental cost to the electric utility of alternative electric energy.”); Cal. Pub. Utilities Comm’n, 134 F.E.R.C. ¶ 61,044, 61160, see also n.67 (Jan. 20, 2011) (“Congress’s intent [is] that utility customers be financially indifferent” concerning whether the utility purchases power from a QF, purchases it from another source, or generates the power itself.). The language in 16 U.S.C.

1 I would affirm the trial court’s decision to vacate the symmetry finding, which no party challenges on appeal.

47

§ 824a-3 about seeking to “encourage” small facility power production—specifically, to “encourage . . . small power production” by “requir[ing] electric utilities to . . . purchase electric energy from [small power] facilities”—is the Congressional statement of purpose for creating the program of electrical energy purchases by utilities. Contrary to Appellees’ arguments herein, this policy is not the standard for setting the rates for such purchases. Rather, ratepayer financial indifference is. While encouraging small facility development is without question a general objective of both federal and state law, that objective is not a mandate that supersedes ratepayer indifference, which is manifested in the calculation of avoided costs. Opinion, ¶¶ 5, 6, 39.

¶81 In more common terms, “financial indifference” for purposes of setting QF tariffs means that the consumers/ratepayers are to be held harmless. Tariffs for QF contracts are to be “revenue neutral” to the consumer—imposing no greater cost for the provision of power than utilities would otherwise pay for the same power. As clearly articulated by Jason Brown, attorney for the Consumer Counsel, to the District Court:

I want to speak to the idea of consumer indifference, because it is the captive consumers of Northwestern Energy who I represent that ultimately pay these costs. It’s consumers that ultimately pay for QF projects. And [Petitioners]

here want to ignore this bedrock principle of PURPA, of consumer indifference, which is embodied in the definition, your Honor, the very definition of avoided costs.

By definition, rates must not exceed the avoided cost of new[,] that is[,]

marginal or incremental energy and capacity. And what that means in plain English is we should pay no more for these projects than we otherwise would if we didn’t have these projects. . . . [A] ratepayer should be indifferent to whether the utility buys its energy capacity from a QF, from a utility-owned project, or from a third party that’s not a QF. [Emphasis added.]

48

¶82 Unfortunately, though accurately described as the “bedrock principle of PURPA,” and “the very definition of avoided costs,” ratepayer financial indifference has been almost entirely ignored in the judicial review of the PSC’s decision in this case. Despite receiving the above argument about this controlling principle, the District Court’s decision did not even mention it, let alone enforce it. This Court’s Opinion, though extensive and well- written, mentions ratepayer financial indifference only once in passing, Opinion, ¶ 39, but neither conducts any analysis of its import nor gives any express consideration to this mandate. In contrast, the Court emphasizes PURPA’s language of “encouraging” the development of QF projects 11 times, repeatedly employing this purpose in its reasoning on individual factors—despite the fact this language states the general Congressional purpose of creating the utility buying program in the first place, not the statutory standard for setting rates. As the Consumer Counsel well explains in its briefing, citing federal authority:

No party disputes that “Congress enacted Section 210 of [PURPA] to encourage cogeneration and small power production.” Whitehall Wind, LLC v. Mont. Pub. Serv. Comm’n, 2015 MT 119, ¶ 2, 379 Mont. 119. This Court has recognized the general policy of “allow[ing] the small facilities to become and remain viable suppliers of electricity.” Whitehall Wind I, ¶ 7.

However, these goals do not override PURPA’s express requirement that “these rates may not exceed the incremental cost to the utility of purchasing alternative electric energy.” [Indep. Energy Producing Assoc., Inc. v. Cal. Pub. Utils. Comm’n, 36 F.3d 848, 856 (9th Cir. 1994)]. Nor do PURPA’s policy goals override its black-letter definition of “avoided cost.”

[Emphasis and underlining added.]

¶83 If, as the Court’s opinion implies by its repeated emphasis, “encouraging” QF development was the sine qua non for setting tariff rates, then every factor would presumptively be resolved in favor of higher rates. And, indeed, the Court’s analysis leads

49

to that result: every factor is decided in favor of higher rates, with no consideration of the “very definition of avoided costs”—ratepayer financial indifference, and no consideration of the financial impact of its decision upon ratepayers. Consequently, the unfortunate effect of judicial review in this case is the imposition of new “judicial rates” that drastically increase costs to the ratepayers, in violation of PURPA.

¶84 One may wonder, if the QF tariff process simply makes mathematical calculations of those costs avoided by Northwestern Energy because of energy alternatively provided by QFs, then how can the effect on ratepayers be anything but financially neutral? The Court’s Opinion assumes so, but the real answer to the question is that tariff-setting is not a purely mathematical calculation, despite being framed within language that gives that impression. To the contrary, the process is replete with subjective and speculative assumptions about the future, and the weighing of policy alternatives. While we may prefer to think of QF tariffs as premised upon mathematical calculations, in reality the process is as much “crystal ball” as it is “calculator.” For example, the carbon cost factor has been ruled “an uncertain cost stream” in past PSC decisions, rejected for all prior QF-1 tariffs, and subjected to an ongoing debate over its viability (currently nonexistent) within the federal regulatory environment—all of which rendered the threshold determination of whether to here impose a carbon adder to the QF-1 tariff a decidedly nonmathematical determination. The Court premises its decision to impose the first-ever carbon adder in a QF-1 contract, not on science, but on the perceived insufficiency of the PSC’s rationale in rejecting the adder, and the adder’s use in different regulatory contexts. Opinion, ¶ 44. The carbon adder determination alone, with its theoretical value of $6.77 per MWh,

50

increases the cost Northwestern will be obligated to pay for QF-1 contracts, and, correspondingly, adds millions of dollars to the costs consumers will pay in increased utility rates. See 18 C.F.R. § 292.303; 16 U.S.C. § 824 a-3(M)(7) (utilities authorized to recover costs of QF purchases directly from their customers).1

¶85 About that there is no need for speculation here, because the record demonstrates the recent Montana history of overpriced QF tariffs in falling markets, as well as the drastic impact judicial review will have upon the ratepayers as a result of this case. In 2012, the PSC established QF-1 rates of $46.97 per MWh (off-peak) and $86.56 per MWh (on-peak). After a scheduled review of the rates, the PSC adjusted them upwards to $53.14/$92.73 for 2013. When the PSC ruled that NorthWestern’s 2014 petition failed to provide sufficient information to adjust tariff rates, the 2013 rates remained in place as market energy prices dropped significantly. The resulting inflated QF rates led to a Montana “gold rush” of QF applicants seeking to participate in the bonanza. In 2012, six eligible QFs had sought use of the tariff; in 2015, 45 did.2 NorthWestern entered five new QF contracts and then sought emergency suspension of the tariff, which was ordered by the PSC. Because the inflated tariffs exceeded NorthWestern’s avoided costs, the PSC determined that failure to suspend

1 Notably, such decisions over the years have resulted in a trend of over-priced QF contracts. “It is fairly obvious today that ‘many state [public utility commissions] and legislatures greatly overestimated long-run avoided costs, thus forcing utilities to buy huge amounts of overpriced power.’ Some state commissions overestimated the amount (as well as price) of the power needed. SoCal, one of the countries’ largest utilities, claims that QF contracts will be its largest source of stranded costs.” John Burritt McArthur, Cost Responsibility Or Regulatory Indulgence For Electricity’s Stranded Costs?, 47 Am. U. L. Rev. 775, 919 (1998) (internal citation omitted) (emphasis added). 2 At that time, Solar was combined with wind. Beginning with the current cycle, solar was tariffed separately.

51

the rates for solar QFs would impose $60 million in increased costs upon Montana ratepayers under the contracts sought by the remaining QF applicants—and, to boot, a carbon adder had not been included in the tariffs.

¶86 That led to the tariff determination now before us, and upon its rulings on the rate factors, discussed below, the PSC set the new rates at $28.14/$37.26. Upon the scheduled periodic review of the rates for updated market forecasts, the PSC increased the rates to $32.78/$41.90. However, upon review, the District Court, after concluding the PSC’s determination on every rate factor was “arbitrary” (see Order Vacating and Modifying, April 2, 2019, ¶¶ 15, 17, 18, and 21), decided every factor in favor of higher rates and, instead of remanding, set the new rates itself—without any consideration of ratepayer financial indifference—at $42.71/$96.75. While PSC Order 7500d, as updated upon scheduled review, established the value of a 3 MW solar QF-1 contract at $2.3 million, the District Court established the value at $4.9 million, a 113% increase over the PSC’s determination. Readily obvious is that the District Court’s on-peak rates exceed even the inflated rates the PSC suspended during the 2015 bonanza. Based upon the 2015 calculation made by the PSC, that the inflated rates would impose $60 million in new costs upon ratepayers, the District Court’s new rates would impose rate increases upon ratepayers in the amount of $104.5 million under the same assumptions ($2.613 million increased contract value x 40 QF-1 applicants).

¶87 Locking NorthWestern into long-term, inflated QF contracts will also lock ratepayers into long-term, inflated rates, in direct violation of what is the clear mandate of PURPA, ratepayer financial indifference. As NorthWestern states, “[t]he new rates would

52

create an unlawful customer subsidy for solar development for 25 years.” As the PSC puts it, “[t]he resulting decision is consequentially harmful to ratepayers.” Or, as the Consumer Counsel puts it, “captive customers would be harmed by the district court’s order. . . . [T]he district court upset the careful balance struck by the Commission between PURPA’s general goal of encouraging QF development and its specific mandate that consumers remain indifferent between QF and non-QF power.” The most revealing point about these statements concerning the large increase in ratepayer costs is that Appellees do not contradict them—Appellees simply argue that errors made by the PSC require the District Court’s rate factor corrections, and steadfastly avoid any consideration of ratepayer indifference. The Court agrees, affirming the District Court’s decision on every factor, and in its entirety.

¶88 The above discussion underscores the critical importance of properly applying the standards of review in cases involving technically complex, yet also significantly subjective, determinations. The standards bear careful repeating. “The scope of review under the ‘arbitrary and capricious’ standard is narrow and a court is not to substitute its judgment for that of the agency.” Motor Vehicle Mfrs. Ass’n v. State Farm Mut. Auto. Ins. Co., 463 U.S. 29, 43-44, 103 S. Ct. 2856, 2866-67 (1983); see also Clark Fork Coal. I,

¶ 27. The judiciary’s “role is not to reweigh the evidence.” Nw. Corp., ¶ 47. “[T]he 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

53

P.3d 595 (citing § 2-4-612(7), MCA). If there is reasonable record support for the PSC’s determination, we are to affirm it.

¶89 More, I believe the analytical approach taken by the Court in this case has distorted the proper application of the standards of review to the PSC’s tariff decision. The Court analyzes each factor in isolation and renders a decision on each as if they were sub-cases to be decided individually. While perhaps not intended as a “divide and conquer” strategy, this analysis nonetheless limits much of the PSC’s discretion, particularly with regard to its ability to achieve ratepayer financial indifference. If each factor must be judged as its own case, and decided on the strongest basis, the PSC will be unable to weigh and balance between the factors as a tool to ensure that ratepayers receive the protection required by PURPA. For example, the PSC may decide to adopt a supported methodology or assumption that is more favorable to QFs under one factor, while balancing that advantage by adopting a supported methodology or assumption that is less favorable to QFs on a different factor. While the PSC may certainly err on an individual factor determination, Whitehall Wind I, ¶ 27, the focus in these tariff cases must be on PURPA’s express mandate that rates guarantee ratepayer indifference, and thus, the tariff decision is more than just “the sum of its parts.” Here, there is an overlay: PURPA requires the PSC to keep its eye on the bottom line, and courts should allow it the discretion to do so. Under the Court’s approach, in combination with its “encourage QFs” quasi-rate standard, there is a presumptive thrust to individually decide each factor in favor of the highest QF rate. This precedent will likely force the PSC’s hand in future decisions and lead to much higher costs and rates in violation of ratepayer indifference, which was buried by the courts in this case.

54

¶90 The District Court also broadly applied the standard of review to conclude every sub-decision of the PSC was arbitrary and capricious, in spite of the principle that this standard is a narrow one. In so doing, the District Court was quite obviously moved by the PSC’s rejection of staff recommendations, from which the court drew heavily for its own findings of fact and conclusions of law. This line of reasoning is a trend in recent district court decisions. In three cases now pending before this Court, including this one, the District Court has concluded the PSC acted arbitrarily by rejecting staff recommendations. Regarding the carbon adder, as just one example, the District Court reasoned here: “The Commission’s contrary finding, disregarding the recommendation of its own technical staff, was arbitrary”; see also MEIC v. Northwestern Energy, 19-0565 (“the Final Order did not address several issues raised by its own staff. . . . This omission itself renders the Commission’s 2015 waiver decision arbitrary”) (emphasis added); MTSUN v. PSC, et al., 19-0363 (“MTSUN is entitled to a carbon adder of $9.65 per MWh as recommended by the staff”). The idea that a decision-making body’s rejection of staff recommendations should be suspected as arbitrary has no authority in law, is arbitrary itself, and is particularly inappropriate when applied to the PSC. By design, QF tariff decisions are made in Montana by five public officials, elected in partisan elections, who are accountable for these decisions to their constituents, even when acting in a quasi-judicial capacity. Moreover, their constituents are ratepayers who are protected by federal law from inflated QF tariff rates that translate into higher utility rates. It is not the staff who bear the duty to represent and protect the constituent-ratepayers. The Commissioners are entitled to make

55

these decisions with appropriate deference and without fear that their failure to adopt a contrary staff recommendation will suggest arbitrariness.

¶91 Turning to the factors, I would affirm the PSC’s determination on the carbon adder or “adjustment,” reversing the District Court’s imposition of the first-ever adder in a QF-1 rate. To explain, “the Social Cost of Carbon . . . [is a] tool, developed by an interagency working group, [that] attempts to value in dollars the long-term harm done by each ton of carbon emitted.” Sierra Club v. FERC, 867 F.3d 1357, 1375, (D.D.C., 2017) (rejecting use of the adder). The adder is “a federal working group’s calculation of the social cost of carbon emissions. (Coal and gas plants emit carbon dioxide; nuclear, wind, solar, and hydro plants don’t.).” Elec. Power Supply Ass’n v. Star, 904 F.3d 518, 522 (7th Cir. 2018). Its purpose is to “rais[e] the costs that carbon-releasing producers incur to do business.” Elec. Power Supply Ass’n, 904 F.3d at 524.

¶92 First, the adder is technically not a “factor,” as it is not a mandated consideration under either state or federal law. Section 69-3-303, MCA; 18 C.F.R. § 292.304. The Court cites Cal. Pub. Utilities Comm’n, 133 FERC at 61,268, Opinion, ¶ 39, but the language there was discretionary in the context of that case, and the Court offers no authority that, in the subsequent 10 years, the adder has been made a general requirement. Consequently, without legal force, the adder’s assessment lies entirely within the PSC’s discretion. While a carbon assessment is considered in a variety of different contexts, in the QF tariff context it runs headlong into PURPA’s definition of avoided costs, premised upon ratepayer indifference. “[A]voided cost rates, in short, are not intended to compensate the QF for more than capacity and energy.” Covanta Energy Group, 105 F.E.R.C. ¶ 61,004, 61,007

56

(Oct. 1, 2003) (explaining that environmental considerations are not included within the factors for consideration under 18 C.F.R. § 292.304). As a discretionary matter, the PSC should be given wide deference in its use.

¶93 As alluded to above, I believe there were legitimate reasons for the PSC’s determination. The Court faults the PSC for inadequate reasoning, but the PSC came to this issue from a long history in which it had addressed the issue often, and its reasoning was clear and supported. The cost was speculative, had never been imposed, and its future consideration within federal regulatory efforts was doubtful, having first been noticed for repeal, and then repealed. As the Consumer Counsel’s expert testified, “[w]e have no way of knowing if, and when, the [] market in Montana will incorporate a value of carbon and what that value will be.” The District Court reasoned that the PSC had “departed from precedent,” but was referring to the use of the adder in other contexts, not QF-1 tariffs; even so, given the highly speculative and unproven nature of the adder, and the lack of legal mandate, the PSC should not be locked into a position to forever impose it when conditions are subject to change or reassessment. These reasons provide a more than sufficient justification for the PSC’s determination, particularly in light of the PSC’s duty to ensure ratepayer indifference. Courts should not be imposing costs that regulatory bodies have not found a basis to impose.3

3 Indeed, courts should not be setting rates at all. If the agency’s decision is determined to be incorrect and reversible, the matter should be returned to the agency for reconsideration. Whitehall Wind I, ¶ 14 (“The [District] [C]ourt remanded the case with instructions to the PSC to set a new rate that would take into account the avoided cost data submitted by Whitehall and the PSC.”).

57

¶94 The PSC has long been concerned that 25-year contracts expose ratepayers to substantial risk, as rates based upon current market prices and necessarily speculative future projections are set for the duration, even as the PSC also seeks to provide stability for QF investment as required by state law. See § 69-3-604(2), MCA (“Long-term contracts . . . must be encouraged in order to enhance the economic feasibility of qualifying small power production facilities.”). However, “long-term” is not defined by law. The PSC initially set the length at 10 years, but upon reconsideration, raised contract length to 15 years, concluding that a shorter length would not “provide QFs sufficient certainty with regard to the potential return on investment in qualifying generating technologies or enhance the economic feasibility of QFs.” The decision is a difficult one, but the question is not whether another length was better supported in the record, or whether the Commission went far enough in its determination to choose 15 years. The Court rejects the PSC’s reliance on the North Carolina Utilities Commission’s approval of 15-year contracts as “dubious given its wholesale rejection of out-of-state decisions as a consideration when setting the avoided-cost rate.” Opinion, ¶ 69. It also rejects the PSC’s reliance upon Appellees’ expert testimony that 15 years was within the duration range for viable QF contracts because the PSC failed to acknowledge the conditions upon which the expert testimony was offered. Opinion, ¶ 69. In my opinion, there is a sufficient record basis for the PSC’s decision, particularly in light of the requirement of ratepayer indifference, and I believe reversing for the reasons stated is substituting the judgment of the court for that of the agency. As we have held, the factfinder is entitled to make credibility assessments, weigh the evidence, and draw implications from the record to make

58

valuation decisions—even if rejecting the precise testimony offered by experts for both sides. See In re Marriage of Clark, 2015 MT 263, ¶ 24, 381 Mont. 50, 357 P.3d 314 (affirming trial court’s ranch valuation of $2.4 million despite lack of specific testimony for that valuation, where it fell between the $1,172,513 offered by husband’s expert and the $2,667,940 offered by wife’s expert). Here, the PSC’s decision fell between the contract lengths offered by the parties.4

¶95 The avoided energy and avoided capacity factors have been merged in the Court’s analysis, which I believe is based upon an overreading of 18 C.F.R. § 292.101(b)(6),5 at the expense of other provisions. The Court reasons that “[w]hile energy and capacity costs are technically different costs, they are fundamentally two parts of the same equation that must be considered together when calculating avoided costs.” Opinion, ¶ 53. Further, the Court concludes that PURPA “mandate[s] that energy and capacity costs be considered together when calculating the avoided cost rate,” citing 18 C.F.R. § 292.304(e). Opinion,

¶ 56. The Court thus concludes that the PSC’s analysis was “an arbitrary distinction between avoided capacity and energy costs when setting the avoided-cost rate.” Opinion,

¶ 54.

4 Although not emphasized by the parties’ arguments, the District Court also reversed the PSC’s “symmetry” determination on QF contract length, wherein the PSC applied the length of a QF contract symmetrically to a utility’s other, non-QF, resources. In doing so, the PSC departed from many years of its contrary practice, and offered no reasoned analysis for doing so. The PSC does not challenge the District Court’s holding on appeal. In light of the lack of record support and lack of challenge, I would affirm the District Court’s reversal of this holding. 5 18 C.F.R. § 292.101(b)(6) provides that avoided costs include “incremental costs to an electric utility of electric energy or capacity or both.” See Opinion, ¶ 52.

59

¶96 However, the implementing regulations maintain a continuing distinction between avoided energy and avoided capacity, and they expressly indicate that energy and capacity must be considered as individual factors. 18 C.F.R. § 292.304(e) provides, in pertinent part:

Factors affecting rates for purchases. In determining avoided costs, the following factors shall, to the extent practicable, be taken into account:

(2) The availability of capacity or energy from a qualifying facility during the system daily and seasonal peak periods, including:

(vi) The individual and aggregate value of energy and capacity from qualifying facilities on the electric utility’s system;

(3) The relationship of the availability of energy or capacity from the qualifying facility as derived in paragraph (e)(2) of this section, to the ability of the electric utility to avoid costs, including the deferral of capacity additions and the reduction of fossil fuel use;

(Emphasis added.) While it cannot be denied that there are intersections between the analyses, I believe these regulations are inconsistent with the Court’s statement that PURPA “mandate[s] that energy and capacity costs be considered together,” Opinion, ¶ 56, because I read them to require that the energy and capacity factors to be considered individually—as analyzed by the PSC—and then their individual impacts combined and assessed as an aggregate of the utility’s avoided costs. The Court reasons that the two analyses do not fit together well, and that merger provides a superior analytical approach. That may well be true, but I believe the regulations require otherwise.

¶97 This analytical clarification justifies the PSC’s argument that Vote Solar altered its position on energy and capacity before the District Court, and, further, led the District Court to erroneously conflate the factors in its analysis. While I would not dismiss Vote

60

Solar’s arguments as beyond the court’s jurisdiction, I would agree with Appellants’ arguments that the District Court erred. The District Court both approved the PSC’s 6.1% capacity contribution (within its adoption of staff recommendations, which included the 6.1% capacity) and rejected it (in its later determination that the 6.1% capacity contribution was arbitrary). I believe there was substantial support for the PSC’s approach and the methodologies it employed in coming to its individual assessments of avoided capacity and avoided energy, and that it accurately determined those costs.

¶98 The Southwest Power Pool (SPP) studies the nation’s electric grid and power market, and its SPP Planning Criteria and Net Planning Capacity program is an industry- wide assessment tool for determining the capacity of small QF solar facilities. Neither Vole Solar nor Cypress Creek disputed the use of this program. The PSC’s application of the program yielded an avoided capacity calculation of 6.1%, and the PSC concluded that Vote Solar’s 36% calculation was inconsistent with application of the SPP. (“The Commission continues to rely on the SPP method . . . . [It] better represents the challenge of meeting load obligations at all times in all years, rather than adopting the VS-MEIC approach of finding the simple average of exceedance values calculated separately for each year. . . . VS-MEIC’s failure to aggregate the annual data before measuring the exceedance level has a significant impact on the final result and is not consistent with evidence on how the SPP method works.” Order 7500d, ¶¶ 48-49). While it does appear the Commission initially used factors from alternate subsections of the SPP in a way that may have limited the sample size, upon reconsideration, it reassessed and considered a larger sample size of months and loads than were required under either subsection. While in future proceedings

61

the PSC should bring clarity to the application of the subsections of the program, it nonetheless remains clear to me that the PSC did not err by rejecting Vote Solar’s extraordinary 36% capacity calculation as erroneous under the SPP, and that the PSC’s 6.1% determination is the most reasonable one of record. While it is correct that NorthWestern also has a high summer peak demand for capacity that should be acknowledged, that does not cure the fact that there is little solar capacity during NorthWestern’s greatest demand times during winter, and assigning a 36% solar capacity will result in a significant over-crediting error, and an enormous cost shift to the ratepayers.

¶99 The proxy method of determining avoided energy costs has been long used by the PSC and is not disputed. However, here the District Court’s conflation of the factors came through as it faulted the PSC for ostensibly failing to consider the cost of additional generating units NorthWestern planned to install in 2019 for increased capacity. However, the PSC did consider the additional generation units as capacity resources in its avoided capacity assessment, rather than as energy resources. The arguments of the parties and the reasoning of the Court, addressed above, have not demonstrated to me that the PSC’s consideration of the new units—in accordance with individual determinations of energy and capacity that is required under 18 C.F.R. § 292.304(e)—is arbitrary, particularly under the guiding principle of ratepayer indifference. I would thus reverse the District Court’s determination.

¶100 Much was made by the District Court, and by the Appellees, of a Commissioner’s comment about the potentially deleterious impact the PSC’s determinations could have on QF development. Without regard to the question of bias, the comment itself captures a

62

reality in the development of alternative energy—that as a matter of economics, alternative energy is more expensive to produce, and development cannot be sustained upon the QF buying program alone. Many years of experience under PURPA’s buying program have proven that “avoided cost is often insufficient to fund renewable energy.” Michael Dorsi, Clean Energy Pricing and Federalism: Legal Obstacles and Options for Feed-in Tariffs, 35 Environs Envtl. L. & Pol’y J. 173, 176-77 (2012). “The claimed benefits of renewable energy are not that it is cheaper to produce, but that it is a better deal once social costs are considered.” Dorsi, 35 Environs Envtl. L. & Pol’y J. at 176. Nonetheless, under the limits imposed by PURPA, avoided costs “are not intended to compensate the QF for more than capacity and energy,” Covanta Energy Group, 105 F.E.R.C. at 61,007, and, at bottom line, “Congress’s intent [is] that utility customers be financially indifferent.” Cal. Pub. Utilities Comm’n, 134 F.E.R.C. at 61,160. Of course, development of alternative energy is encouraged by other state and federal programs and incentives, but the point here is that courts are not permitted to further advance the laudable purposes of PURPA’s buying program by shifting greater costs to the consumer than permitted by Congress. “When states attempted to include externality costs in their avoided cost rates, FERC ruled that only those costs which the utility faces may be considered in setting avoided cost.” Dorsi, 35 Environs Envtl. L. & Pol’y J. at 176. The duty of the courts is to strictly enforce Congress’s express granting of protection to ratepayers.

¶101 In conclusion, I would hold that the District Court erred by reversing the PSC’s determinations regarding the carbon adder, contract length, avoided capacity, and avoided energy. While I would have reversed the Court’s failure to remand to the PSC to set new

63

rates, that concern is mooted by my other conclusions. I would affirm the District Court’s reversal of the PSC on contract symmetry.

/S/ JIM RICE

Justice Laurie McKinnon joins in the dissenting Opinion of Justice Rice.

/S/ LAURIE McKINNON