Correction on Monday's posts: RT3 is technical constraints, not tertiary reserve

Brief follow-up to Monday's newsletter on Spain's RT3 cost ×10 since 2020 and the post-blackout step.

An expert reader — Sergio Bertolín, who works in commercial electricity retail — emailed to flag a precision issue: what we labeled "RT3 = tertiary reserve" is actually "RT3 = technical constraints in the day-ahead market". He's right, and the correction makes the story sharper.

What stands: all data, ratios, and the renewable correlation. compodem.RT3 did grow ×10 since 2020, the post-blackout ×2.5 step is real, the correlation with PV+wind share (r=+0.76) holds.

What changes: the mechanism description. RT3 isn't tertiary reserve activations — it's the consumer-side cost of REE redispatching the OMIE clearing to keep synchronous units online for system security (what the Spanish grid jargon calls restricciones técnicas del mercado diario). The post-blackout ×2.5 step is now precisely explained: REE introduced Operación Reforzada after April 28, 2025, requiring more synchronous capacity (mainly CCGTs) online for grid stability.

Independent corroboration: PwC published a February 2026 report on Operación Reforzada that pins December 2025 technical-constraints cost at 11.50 €/MWh — our compodem.RT3 for that month is 11.78 €/MWh. Match to the decimal.

Both Monday's blog posts and the newsletter web archive have been updated with the corrected interpretation and a visible erratum block crediting Sergio.

Why this matters: if your business depends on forecasting Spanish electricity system costs, the line item to model is technical constraints (compodem.RT3 column), not tertiary reserve. Under Operación Reforzada it's the fastest-growing component of the consumer bill.

What we learned: better data conversation > defensive data conversation. Sergio's correction made the analysis stronger. The renewable-integration story is the same — actually sharper — and now grounded in the right mechanism.

Re-read the corrected pieces at datons.com/newsletter/2026-05-18 — both posts (RT3 structural growth + 12-month blackout comparison) are now updated.

Thanks Sergio. And thanks to everyone who reads, replies, and pushes for precision.