Community Solar and Value of Solar Under Review in Minnesota

Date: 2 Feb 2018 | posted in: Energy, Energy Self Reliant States | 0 Facebooktwitterredditmail

Minnesota’s leading community solar program is about to go through its first major change, from using net metering retail compensation (plus a credit payment) to using Minnesota’s value of solar payment. Unlike other states that have torpedoed net metering, Minnesota’s alternative compensation actually takes into account the value of solar energy to the utility, the grid, and society (for more on the policy, see our 2014 report).

Below, we share some comments we filed last week on the 2018 value of solar calculation by Xcel Energy, offering some context for how this calculation works and what factors matter most. But first, some background.

The Value of Solar

The value of solar concept was created by legislation in 2013, with an order to the Commission to create a universal and transparent methodology for utilities to fill in the numbers. Below, we show the components of the value of solar and the resulting values for one utility, Xcel Energy, over time.

The first thing to note is that the value hasn’t changed much over time, with consistent components other than avoided distribution capacity. This value, removed after the first year, is appropriately back with more nuance in the 2018 edition, being calculated over 9 distinct planning areas (more on that later).

The two other values that have changed over time are the environmental values (including criteria pollutants and carbon dioxide) and avoided fuel costs. Environmental costs are changing in two ways: first, the Public Utilities Commission recently updated and increased the externality cost of a number of criteria pollutants. The last bar in the chart reflects the update to the NOx value, as distinct from 2018 v2. The second element is how much natural gas fuel use is avoided, as that’s the fuel on the margin. This gets into a wonky calculation called heat rate (the amount of fuel required to generate a kilowatt-hour of electricity from a gas power plant). Our comments address the proposed change from the 2018 v1 to 2018 v2 edition calculation.

The other significant change is in avoided fuel costs, and the falling value in 2018 simply reflects the falling cost of gas due to market conditions.

As it stands, the value of solar is relatively close to the retail electricity rate for residential customers of Xcel Energy in Minnesota. However, in this particular regulatory proceeding, there’s the confounding factor that value of solar will be used for community solar projects, only, which have unique costs (e.g. subscriber management, customer acquisition). Previously, the Commission has provided adders to make projects financeable when the default compensation rate (the net metering rate) was insufficient. It remains to be seen if they’ll do so again.

Three Calculations That Need Scrutiny

With the health of Minnesota’s community solar market on the line, ILSR’s comments focused on three calculations: environmental value, heat rate and fuel costs, and avoided distribution capacity costs.

Environmental Values

For environmental values we addressed two issues. First, to use the updated externality costs for NOx that are now available from the Commission, to reflect the actual cost to society of avoiding NOx pollution from burning gas at power plants. We also asked the Commission to use its externality values reflecting the urban customer base of Xcel Energy, rather than suburban or exurban. Below is the actual text of our comments:

As highlighted in comments from Xcel Energy filed January 4, 2018, the 2018 vintage value of solar should include the updated externality values for NOx. Additionally, given the vast majority of Xcel’s Minnesota customers live in urban areas, the urban values–and not the urban fringe–are most appropriate metric for the societal impact of pollution.

Heat Rate and Fuel Costs

We also addressed the heat rate calculations Xcel offered, because the result in a dramatically different value for solar. Xcel says that they are reverting back to a calculation they first used in 2014, based on the blend of gas plants (combined cycle, CC, or combustion turbine, CT) in operation at solar-producing times. However, their calculations are inconsistent, and crucial documentation of their method is missing:

Xcel’s heat rate calculations are confusingly inconsistent. Xcel first filed a revised heat rate of 8,755 for the 2018 vintage value of solar rate, down from 9,158 in 2017. In their revised 2018 filing, calculations using a CC/CT “blend approach” from 2014 returned a solar-weighted heat rate of 7,482. According to Xcel, this calculation for 2016 and 2017 was done with a six-year average of the solar-weighted heat rate (IR 11). The original 2018 solar-weighted heat rate was done with a single year forecast (and resulted in a small decrease. In the response to IR 11, Xcel says they “believes the solar heat rate in 2018 should have utilized the same CC/CT blend approach” as reported in 2014. The response references a CC/CT blend approach allegedly filed in 2014 (we were unable to find the document the Company references: DOC IR No. 19, Attachment H, and submitted on May 16, 2014 in Docket No. E002/M-13-867).

Given the substantial impact on the value of solar, we would like to know more about the heat rate methodology employed.

The utility also changed its method for calculating the avoided fuel cost. While we agree that it should reflect market prices, it is unclear if the utility is using a consistent (or legally required) methodology. Our comment:

We would similarly like more information about Xcel’s natural gas fuel cost projections. While the overall forecast aligns with futures markets, we were unclear if they are precisely following the methodology outlined in the 2014 VOS order, shown below:  

  1. First, monthly prices are determined by averaging the 30 days of NYMEX prices for each month, starting with the most recent 30 daily prices and then repeating the same 30- day averaging for every other contract month of the 12 year period. If a utility calculating a VOS rate does not have historical daily NYMEX prices already collected internally they can obtain this data by recording quotes for 30 days. The timing of the data collection should be accounted for in planning the VOS rate calculation.       
  2. Then, the monthly prices are averaged to give a 12-month average in $ per MMBtu, resulting in the first 12 annual prices in the set of 25 annual prices. Prices for years beyond this NYMEX limit are calculated by applying the general escalation rate. An assumed fuel price overhead amount, escalated by year using the general escalation rate, is added to the fuel price to give the burnertip fuel price.
  3. Prices for years 13 through 25 are calculated by escalating the year 12 annual average NYMEX quote by the general escalation rate annually for each year.

Avoided Distribution Capacity Costs

The final focus of our comments was on the avoided distribution capacity provided by solar. This is a calculation meant to represent how solar allows a utility to avoid upgrading lines and substations in local neighborhoods because of solar’s overlap with peak energy use. In this case, the utility’s filing has some wild inconsistencies from year to year in their reported distribution costs, as well as bizarrely low estimates for future distribution costs that are 40% lower than historical norms. The detailed comment follows:

In each of its value of solar filings, Xcel Energy has provided Table 14. Deferrable Distribution Capacity Costs. However, the reported figures for prior years have changed in subsequent filings. We have several questions:

  • Why does the 2018 filing show a $400,000 decrease from the 2017 table?  
  • Why does the amount of distribution costs change for a given year in subsequent filings? For example, the Distribution Project Costs reported for 2010 was roughly $129 million in Xcel’s VOS filing in 2015 and 2016, but then reported as $153 million in their 2017 filing, and just $98 million in 2018.
  • Why are planning area cost forecasts substantially below historical averages? The following table shows that, with the exception of the Northwest planning area, the average annual cost forecast is at least 40% lower than the historical average. If distributed solar enables deferral of distribution capacity costs, then abnormally low forecasts will have a severe dampening effect on the value of solar.

We have two additional questions about Xcel’s methodology for calculating avoided distribution capacity costs:

  • How did Xcel select its planning areas?
  • Why does Xcel use a different methodology for calculating system-wide avoided costs than for planning areas (referenced in comments by Nokomis Partners)?

Wrap Up

Minnesota’s policy was adopted at a crucial time for the discussion of the value of solar, when many state legislatures and regulatory commissions were debating how solar producers should be compensated. We hope it will continue to lead in providing a best practice for a transparent and thorough methodology for valuing distributed solar. We signed off as we usually do, thankful to have the opportunity for public comment, something that doesn’t always happen:

Thank you for considering our comments. We look forward to participating further in this docket, and appreciate that there has been no legislative preemption of this regulatory process.


This article originally posted at For timely updates, follow John Farrell on Twitter or get the Democratic Energy weekly update.

Photo credit: hercios via Flicker (modified)

Avatar photo
Follow John Farrell:
John Farrell

John Farrell directs the Energy Democracy initiative at the Institute for Local Self-Reliance and he develops tools that allow communities to take charge of their energy future, and pursue the maximum economic benefits of the transition to 100% renewable power.