Read on Twitter
Thread on ERCOT reliability: I run a small resource adequacy consulting firm that analyzes reliability across the country. Our clients among others include ERCOT, SPP, and MISO. All three entities are dealing with reliability issues today. 1/x
*what follows are my thoughts only. Not speaking for my clients* 2/x
It's probably not much solace but electricity suppliers plan to keep these types of issues to less than once per decade. This is the first event in ERCOT in 10 years. It's been much longer for others. 3/x
Planning to keep the lights on entails carrying adequate reserves for 3 key uncertainties: extreme weather, generator outages, and economic growth uncertainty. 4/x
The most extreme winter weather in ERCOT was expected to add 10% to normal peak. This equates to 70GW. This is predicated on experiencing 1989 temps when Dallas hit -1F. Dallas was 13F on Sunday evening and the load was already 70GW. 5/x
Generator outages are generally higher during extreme weather, but rarely exceed 10% of conventional generation supply. Economic uncertainty is not relevant to today's issues since it is a long term risk. 6/x
ERCOT entered this winter with a 31% reserve margin (reported values are higher but ERCOT's official peak load forecast is lower than what's expected with 50% probability). They should have had >7GW of reserves even during a 1 in 40 year event. 7/x
But ERCOT had 34 GW of generation unavailable. Outages increase during cold weather, but nothing on that scale. Projecting historical cold weather outages, today's outages could have been 5-7GW, not 34GW. Culprit is likely fuel unavailability.
There is ~30 GW of wind capacity in ERCOT but it's only given credit for 6GW of reliability and was still producing higher during Sunday evening's peak. I'm not convinced there is a material problem with the assessments of renewable energy contributions to reliability. 9/x
We have a lot to sort out, but the efficient answer isn't likely more capacity in ERCOT. It's probably a combination of weather hardening generators, and improving fuel supply adequacy. Market design may not appropriately incentivize generators to perform those functions. 10/10