PUC flags forecasting, AMI data and EV load assumptions in PSCo DSP
Get AI-powered insights, summaries, and transcripts
Subscribe
Summary
Commissioners and advisors criticized Public Service Company of Colorado’s forecasting methods and urged stronger use of AMI data, clearer EV/heat‑pump load profiles, and scenario testing before approving large GMAC recovery.
Commissioners at the Public Utilities Commission hearing sharply questioned how Public Service Company of Colorado modeled future demand from electric vehicles and building electrification, and they urged more transparent, testable use of advanced metering (AMI) data and scenario analysis ahead of any large, expedited cost recovery through the GMAC.
The issue: PSCo’s forecast combined several inputs — customer “capacity‑check” applications, projected EV and beneficial electrification (BE) adoption curves, and other program growth — in a load‑forecasts tool (referred to in the record as LoadSeer). Advisors and parties told the PUC the company did not provide a record that allows independent scenario or sensitivity testing of those assumptions.
Advisory staff recommended the company run a separate LoadSeer scenario using capacity checks alone so the commission and parties could see how much of the proposed capacity spending rests on customer applications rather than broader BE adoption. Dan Greenberg said the sweep/NRDC approach (prioritizing by how quickly load approaches a feeder’s continuous rating) “might delay capital investment” and focus spending on the most time‑sensitive projects.
Commissioners and parties pressed several specific modeling and data improvements they want before future DSPs or the next GMAC review:
- Differentiate BEV (battery electric vehicle) and PHEV (plug‑in hybrid) adoption curves and charging profiles in modeling; model Level‑2 and DC fast charging (DCFC) separately and geographically rather than assuming a uniform public‑charging shape.
- Use AMI data sets at scale to derive household and retrofit heat‑pump usage patterns rather than extrapolating from small samples. Advisors recommended the company either commit sufficient internal staffing to perform statistically valid AMI analyses or hire an independent consultant and make the consultant’s work available for cross‑examination in future proceedings.
- Report separate scenarios isolating capacity‑check requests from other adoption assumptions; produce hosting‑capacity analysis (HCA) maps and make anonymized AMI data available to parties for validation.
Why it matters: inaccurate or non‑testable load shapes feed directly into whether the DSP requires wide capacity expansion (and therefore large near‑term customer charges) or whether demand‑side management, managed charging or nonwires alternatives could materially reduce capital needs. Chair Eric Blank said the modeling was “generally opaque” and that the company did not present alternatives testing sufficient to approve broad GMAC recovery.
Next steps: advisors recommended directives and data‑access requirements for the company’s next DSP and suggested, where necessary, independent AMI analyses that can be tested by parties. Commissioners asked staff and counsel to propose specific requirements and timeline language for inclusion in future orders.