Clean energy groups bring models to a modeling fight

In states with a weak renewables standard, or none at all, a key battle over an electric utility’s solar deployment plan is fought in a regulatory proceeding triggered by the utility’s filing of an integrated resource plan (IRP)—at least in the 33 states that require utility IRPs.

In theory, to develop an IRP, a utility uses an optimization model that spits out the least-cost mix of future generation resources to meet projected demand, plus a reserve margin.  As prices for solar, wind, and storage fall, you would expect to see these plans reflecting an increasing percentage of renewable generation.  Yet a utility controls the inputs to the model, and selects which scenarios to consider, so it can skew the model results in ways that disadvantage solar, wind and storage.

Few state regulatory agencies have the budget, expertise, or data to run their own model.  But in North Carolina at least, two clean energy groups have hired consulting firms to run models, and have presented the results to North Carolina regulators through a regulatory proceeding.

Both groups found that the least-cost resource mix for North Carolina would involve much more solar than the utilities’ plans projected, with savings coming largely from coal unit retirements and from building only those natural gas units already under construction.

Both groups flagged a key concern with the utilities’ modeling: the two Duke Energy utilities did not let their model “turn off” coal plants, whereas the clean energy groups let their respective models decide whether retiring some coal units would lead to lower utility bills—and found that it would. (North Carolina is served by three utilities: two subsidiaries of Duke Energy, plus a subsidiary of Dominion Energy.)

The North Carolina Sustainable Energy Association (NCSEA) reported that reaching solar generation of 18% by 2033 would help reduce customer bills, compared to just 8% total renewable generation by 2033 as proposed by the two Duke Energy utilities  (as shown in a Duke filing, Figure 12-F).

Similarly, modeling results submitted by the Natural Resources Defense Council (NRDC) showed that adding 8 GW of solar in North Carolina by 2025 (compared to 3 GW in Duke Energy’s plans), plus 1.3 GW of storage by 2030, would yield the lowest customer bills.  That “low bill” scenario also involved retiring 4 GW of North Carolina coal generation by next year.

Both groups hired consulting firms to run the models.  NCSEA hired Synapse Energy Economics to run the EnCompass model, licensed by Anchor Power Solutions, while NRDC hired ICF Consulting to run that firm’s IPM model.

Synapse ran the EnCompass model in both the capacity expansion and production cost modeling modes, said Ms. Jones of NCSEA, while ICF’s IPM model is a capacity expansion model.  (For comparison, the Duke and Dominion utilities used capacity expansion and production cost models for their IRPs; Dominion also made limited use of a power flow model.)  The different types of modeling can achieve progressively more refined results, as shown here:

Both models, EnCompass and IPM, come complete with data on existing electric generating units.  Both Synapse and ICF used cost projections for solar and wind power from the National Renewable Energy Laboratory’s Advanced Technology Baseline report.  For storage, Synapse used cost projections from Lazard’s Levelized Cost of Storage report, while ICF used NRDC’s mid-case cost trajectory synthesizing four data sources.

NCSEA’s and NRDC’s regulatory filings presented different interpretations of Duke Energy’s IRP treatment of existing coal units.  The Synapse report submitted as part of NCSEA’s filing indicated that Duke had “must-run designations for coal units.”  A report by Applied Economics Clinic (AEC) jointly filed by NRDC, the Sierra Club and the Southern Alliance for Clean Energy stated that Duke did not forecast the fixed costs of existing units; AEC noted that “future fixed O&M costs are avoidable if the plant retires.”