THE CHEMISTRY
CRISIS IS ROLLING,
NOT CLOSED.
The 2018 Texas event was the first regional shock. Twenty-three more coal units shut down in 2025 alone. Demand for fly ash rose in 2024 while supply fell. Then three major utilities moved to keep coal burning longer under demand pressure. The pattern is not historical. It is structural, ongoing, and federal-regulatory-cycle sensitive.
The crisis is rolling, region by region.
The frame is consistent across PSIP institutional surfaces. It anchors the national doctrine on the chemistry crisis and threads through every state record where the regional supply story is in motion.
Supply contracted. Demand expanded.
EIA reports that only 2.6 gigawatts of coal capacity actually retired in 2025, because federal emergency orders kept plants running. Total coal-fired capacity is scheduled to fall from 172 gigawatts (May 2025) to 145 gigawatts by the end of 2028.
Simultaneously, fly ash consumption rose. ACAA's 2025 Production and Use Survey documents 14.6 million tons of fly ash consumed in 2024 concrete production, up 2.7 million tons year over year. ACAA has indicated the 2024 figure will be revised upward in early 2026 once ash harvesting and bottom ash recovery data is finalized. Demand is up while supply is constrained. This is the structural setup for the regional fly ash deficit story.
The pattern is not yet visible at the national average. It is visible at the regional level. Wave B's per-state pages document the regional picture in Texas, Tennessee, Florida, Georgia, and North Carolina.
Coal preserved on borrowed regulatory time.
Three utilities have moved to extend coal operation under load growth, and their stated reasons are not the same. TVA cited rapidly increasing electricity demand. Duke Energy cited federal actions that eased restrictions on coal generation. Georgia Power framed its co-firing program around continued compliance with federal environmental regulations. Where an extension is documented in a filed resource plan or a board resolution, the supply preservation is real. The premise underneath it is conditional on a federal regulatory cycle that is itself in motion.
Kingston,
Cumberland
On February 11, 2026 the TVA Board of Directors voted to run both plants past retirement dates TVA had committed to in Records of Decision, Kingston's nine coal units by the end of 2027 and Cumberland's two units by the end of 2026 and the end of 2028. The board cited rapidly increasing electricity demand after a decade of flat demand.
Plant Bowen,
Plant Scherer
On July 15, 2025 the Georgia Public Service Commission approved extended operation of certain coal and natural gas units through at least 2034. Southern Company's annual report lists Plant Scherer Unit 3 running through at least December 31, 2035. The stated driver is substantial new load growth over the coming decade, with data centers driving a significant majority of it.
Belews Creek,
Cliffside, Marshall
Duke asked North Carolina regulators in October 2025 to run three dual fuel coal stations two to four years longer than planned. Belews Creek Units 1 and 2 move to 2040 under both demand scenarios Duke modeled. Cliffside and Marshall move only under the higher one. The Commission has not ruled, and its order is due by December 31, 2026.
A new cement chemistry, poured without warning.
Portland-limestone cement (ASTM C595 Type 1L) now accounts for the majority of blended cement shipped in the United States. The transition was carried out by the cement industry as a sustainability measure. Up to 15 percent finely ground limestone replaces clinker. The reactivity is different. The water demand is different. The behavior with admixtures is different. The behavior in cold weather is different.
The chemistry change was largely not paired with admixture reformulation at the mix design level for most residential concrete. The contractors pouring the slab were not notified. The mix designs they were ordering by habit kept producing what looked like the same concrete on paper. The concrete that arrived on the truck was a different material.
The combination of the Type 1L transition and the fly ash collapse compounds. The cement chemistry changed at the same time as the supplementary cementitious material that traditionally protected against alkali-silica reaction disappeared. Most residential concrete contractors in the United States are pouring a different material than the one they were trained on, with no documentation that the change has occurred, and no protocol for adjusting the mix design to compensate.
Ash harvesting buys time. Logistics shocks tighten the picture.
The industry is shifting toward harvesting fly ash from existing on-site landfills (sub-grade or pond ash storage) to extend usable fly ash supply. This buys time. It does not solve the structural supply decline. Harvested ash chemistry can differ from freshly produced ash; quality control is more variable. Mix designs that perform with fresh ash do not necessarily perform identically with harvested ash.
Railroad labor disputes and freight corridor congestion across 2024 to 2025 disrupted fly ash transportation between regions. A Tennessee ready-mix plant that previously could replace a lost local supply with rail-shipped ash from Wyoming or Pennsylvania can no longer do so cleanly. Regional supply has become more regional in practice, not less. The state pages below document what each region is now operating on.
The crisis, region by region.
Each state below carries its own coal-plant landscape, regional fly ash supply trajectory, case-law context where it has surfaced, and institutional record relevant to residential concrete operating in that state in 2026.
The protocol turns a rolling crisis into a documented pour.
PSIP-certified contractors specify the supplementary cementitious material strategy in writing before the mix is ordered. The batch ticket is verified before the truck unloads. The closeout packet documents the ash source, the chemistry on the day of the pour, and the climate-day conditions the slab will live in. The protocol is structurally independent of every material supplier. The protocol scores. The protocol does not recommend products.