Kate Lamb opened the meeting by introducing ERCOT's updated long-term load forecasting methodology.
She emphasized the growing complexity in forecasting due to increasing large flexible loads (LFLs), such as data centers and cryptocurrency mining operations.
Explanation that for the first 180 days of outage studies, ERCOT would use a base forecast adjusted for electric vehicles, crypto, and photovoltaic reductions.
Beyond the initial 180 days, ERCOT plans to overlay data from the Large Load Interconnection Queue onto the base forecast, moving away from reliance solely on contract officer letters.
This shift aims to provide a more accurate short-term outlook, as officer letters have previously led to overestimations in demand projections.
Sam Morris then led a discussion on the evaluation of officer letters from 2023 and 2024.
The analysis revealed that 55.4% of these letters resulted in energized projects, but only 26% of the projected megawatts were realized, highlighting the need for more reliable forecasting methods.
Sam presented data from Transmission Service Providers (TSPs), noting that they received requests for 230 gigawatts (GW) of load by 2028–2030.
This figure was adjusted to 173 GW, and ERCOT's methodology further reduced it to 112 GW.
For 2031, TSPs projected 340 GW, which was adjusted to 218 GW, and ERCOT's forecast brought it down to approximately 145 GW.
These adjustments are based on actual observed trends and data, such as the interconnection queue, rather than solely on projections from contracts and officer letters.
Sam Morris also discussed the impact of large industrial projects, like data centers, on the forecasts.
These were reduced from 86 GW to 24 GW in ERCOT's adjusted projections, aligning more closely with observations from other ISOs and RTOs.
He emphasized that ERCOT's methodology for the adjusted forecast is evolving and continually updated, especially in terms of the impact of crypto and other emerging loads on the grid.
The conversation shifted to the application of the adjusted load forecast for the Capacity Demand and Reserves (CDR) Report and Regional Transmission Plan (RTP), which would use the adjusted data, while Regional Planning Group (RPG) Projects would stick with the TSP forecast.
Sam Morris clarified the process for outage scheduling, which would use the adjusted forecast for the first 180 days, factoring in large load interconnection queues instead of contracts and officer letters, which were deemed too optimistic.
Following the 180-day period, the forecast would revert to the adjusted forecast, incorporating contracts and officer letters again.
Seth Cochran asked about the usage of TSP forecasts in ERCOT's independent review process for RPG projects.
Sam noted that Robert Golden would be the person to confirm how ERCOT would use these forecasts, as the independent review was still being finalized.
Ian Haley inquired whether the decision to discount forecasts was made due to the Maximum Daily Resource Planned Outage Capacity (MDRPOC) issue or if it was something ERCOT had planned in advance.
Sam explained that the team had realized the disconnect between contract officer letters and the interconnection queue, prompting them to focus more on the interconnection queue for short-term forecasts.
Kevin Hanson asked whether the numbers presented included any LFL “haircuts”, which Sam confirmed they did.
Sam mentioned that the forecast would be updated by March of next year, possibly in February, with new iterations based on refined data from studies on large load forecasts.
He elaborated that the team was working with an outside source for this research, and while they had given timelines, it could take longer to finalize.
Seth Cochran further clarified that the next CRR study would likely use the updated forecasts presented in the discussion, with data adjustments continuing to evolve.
Sam Morris confirmed that the full forecast, including historical data dating back to 1980, was posted on the website.
This included both the TSP and ERCOT adjusted forecasts and went up through 2026.
Kate Lamb clarified that the forecast data was large, so there had been delays in posting additional years, but that it would be updated by next week.
Sam emphasized that the ERCOT team was aiming for full transparency in their processes and would be happy to provide support to anyone needing assistance with reading or interpreting the data.
Pete Warnken then addressed the question about the NERC long-term studies, explaining that NERC would use the adjusted load forecast for their assessment, although they may also request risk scenarios or further analysis if needed.
Sam wrapped up by thanking everyone for their questions and feedback, reminding the group that a lot of this data is available on the website, and encouraged anyone with further questions to reach out.
▶️3 - Handling coal-to-gas conversion projects in the CDR
Current CDR does not account for coal-to-gas conversions in the protocols.
Interim approach being developed to handle these conversions.
The February 2025 CDR included two Unconfirmed Planned Retirement units, Luminant’s Coleto Creek and CPS Energy J. K. Spruce 2, that will be converted from coal to gas fuel.
Expected no net change in capacity for conversion projects initially.
New line item to be created for fuel conversion projects, focusing on coal to gas.
Proposing a status code EXT-OUT for extended outages during conversions.
Two line items for projects like Coleto Creek: original and extended outage representation.
Fuel type change will be indicated in the CDR with a footnote.
Once a conversion project meets planning criteria, it will be marked as "plan repower" with the status code PLAN-RPR.
Official retirement leads to status change to conventional unit, marked as CONV.
Updated details to be included in the protocols and next NPRR.
Coordination with companies required for outage scheduling.
Plants will have a GNIR number post-interconnection request.
Coordination necessary for resource type changes, especially in sixty-day reports.
Request for feedback by mid-next week on the proposed approach.
