ERCOT 4CP · Demand Response · Houston, Texas

Four 15-minute windows
price your transmission
for an entire year.

Every summer, ERCOT's four coincident peaks decide what your building pays in transmission charges for the next twelve months. BOT forecasts the risk daily, pre-cools and coasts through the window, and verifies the avoided kW at 15-minute resolution — on our own buildings first, then on yours.

36%
peak demand cut, one building
372→239 kW
the same cut, in kilowatts
R² > 0.96
M&V baseline · IPMVP Opt. C
15 min
verification cadence
Sec. 01 · The rule

What 4CP actually is —
and why one hour prices twelve months.

Each month from June through September, ERCOT records the single highest 15-minute interval of system-wide demand. Those four intervals are the four coincident peaks — the 4CPs. Under ERCOT's published 4CP methodology, your building's average draw during those four windows sets its transmission cost responsibility for the entire following year.

Read that again, because it is the strangest pricing rule on a Texas commercial electric bill: roughly one hour of operation, spread across four summer afternoons, prices twelve months of transmission charges. A building that happens to be running flat-out at 4:45 PM on the wrong August day pays for that coincidence until the next season resets the number.

The flip side is the opportunity. Commercial buildings burn roughly 40% of U.S. electricity, and most of them treat the transmission line on the bill as weather — something that happens to you. It isn't. If your building simply draws less during those four intervals, your share of the grid's transmission costs falls. No enrollment, no aggregator contract, no rebate paperwork. It is a pure operating play, which is exactly why it rewards buildings that are actually operated. For the full primer on the mechanics, see our explainer, ERCOT 4CP, explained.

Sec. 02 · Know which game you're playing

NCP vs. 4CP — two demand charges,
two different games.

"Demand charge reduction" is really two disciplines that get lumped together. One is set by your worst 15 minutes; the other by the grid's. They respond to different tactics, and confusing them wastes effort.

 NCP — Distribution demand4CP — Transmission demand
What sets itYour building's single highest 15-minute interval each month, regardless of grid conditions.The grid's highest 15-minute interval in each of June–September; your draw during those four windows.
Billing horizonResets monthly — one careless morning costs you thirty days.Sets your transmission ratio share for the entire following year.
When it happensWhenever your equipment stacks up — classically the 8:00 AM all-at-once start.Hot summer weekday afternoons, typically late in the day, June–September.
Can you see it coming?Yes — it's your own load. Meter it and you know.Not with certainty. Nobody knows which afternoon set the peak until the month closes. Forecasting is the whole game.
How you fight itStaggered starts, demand limiting, startup-surge trimming, pre-cooling into the morning.Daily peak-risk forecast → pre-cool ahead of the risk window → coast through it drawing less.

Both games run on the same instrument: interval data. If you only see your load as a monthly kWh total, you are playing neither. More on the monthly game in demand charge reduction for commercial buildings, and on where these lines sit on the bill in commercial electricity rates in Texas.

Sec. 03 · The forecast and the playbook

Nobody knows the window
until it's over. So we forecast it.

Here is the honest problem with 4CP: the peak interval is only declared after the fact. Curtail every hot afternoon and you burn comfort, equipment hours, and tenant goodwill on days that didn't matter. Curtail none and one ordinary Thursday sets your transmission bill. Most buildings resolve this by doing nothing, which is a choice — just an expensive one.

A daily risk read, not a guess

BOT's ERCOT 4CP program publishes a daily probabilistic 4CP-risk outlook for the system peak window, June through September, driven by load and weather data. We publish risk levels with confidence intervals — not pretend-certain calls — and we verify the forecast after every season against ERCOT's declared peaks. Each morning, Ada, our optimization intelligence, consumes that risk read and folds it into the day's operating plan, with our operators reviewing the call. Nothing here runs unsupervised.

Pre-cool, then coast

The playbook itself is old-fashioned building physics. On a high-risk day, the building pre-cools ahead of the risk window, banking cooling in its thermal mass while grid demand is still moderate. When the window opens, it coasts — compressors backed off, the space riding its stored cooling through the critical intervals. Tenants feel a well-run building; the meter shows a building drawing less power during exactly the fifteen minutes that price next year.

The same machinery handles the monthly NCP game. Demand-response modeling on one of our buildings, 11111 Wilcrest Green, built on its real daily peaks, shows startup-surge trimming alone is worth roughly $3,500–$8,000 per year depending on cap depth. And it executes cheaply: the Atenea BAS runs on edge microcontrollers with zero licensing — we built it for under $50 per controller, while other landlords pay $50–200K a year in BAS licensing to do less.

Sec. 04 · Verified, not asserted

Demand response you can
take to an auditor.

A curtail call that isn't measured is a story. After each season we document the avoided coincident-peak kW, and that documentation feeds the same M&V record as everything else we do: continuous AI-driven measurement and verification at 15-minute resolution, IPMVP Option C, with baseline models holding R² above 0.96. Most M&V in this industry is a $10–25K study performed once a year; ours runs automatically every 15 minutes, so the peak we shaved on an August afternoon shows up in the record that same day.

We proved this on our own money first. We own and operate 4 buildings totaling ~421,000 SF in Houston, with 19+ years of below-CPI operating expenses. At 800 Wilcrest, annual energy use fell roughly 50% from 2015 to 2025 (time-normalized, not weather-normalized). At 11200 Richmond — acquired September 2025 — we verified a 22.4% energy reduction about eight months in. And on one building, peak demand dropped 36%, from 372 kW to 239 kW. The platform behind it, RMS, also serves ~600 K-12 schools and Harris County, with 44 validated M&V reports across 15 buildings.

44
validated M&V reports across 15 buildings — the verification habit, on the record.
22.4%
verified energy reduction YTD at 11200 Richmond, ~8 months after acquisition.
~600
K-12 schools plus Harris County served by the RMS platform.
Sec. 05 · Straight answers

ERCOT 4CP & demand response — FAQ

What is ERCOT 4CP and why does it matter for a commercial building?
4CP stands for the four coincident peaks: the single highest 15-minute system-wide demand interval in each of June, July, August, and September on the ERCOT grid. Under ERCOT's published 4CP methodology, your building's average draw during those four intervals sets its share of transmission cost responsibility for the entire following year. Roughly one hour of operation prices twelve months of transmission charges — which makes 4CP management one of the highest-leverage demand-charge plays available to Texas commercial buildings.
Do I have to enroll in a utility demand response program to reduce 4CP charges?
No. 4CP avoidance is not an enrollment program — it is an operating discipline. If your building draws less power during the four peak intervals, your transmission cost responsibility falls the following year. The hard part is that nobody knows which afternoon will set the peak until the month closes, so the practical requirements are a daily peak-risk forecast, a pre-cool-and-curtail playbook the building can actually execute, and post-season verification against ERCOT's declared peaks.
How much can demand-charge reduction actually save a commercial building?
Results are building-specific, but they are measurable. On one building in our own Houston portfolio, peak demand dropped 36% — from 372 kW to 239 kW. Demand-response modeling on another building, using its real daily peaks, shows startup-surge trimming alone is worth roughly $3,500–$8,000 per year depending on cap depth. Because the non-kWh lines — demand, transmission, and power factor — make up nearly half of a Texas commercial electric bill, peak management often rivals consumption measures in dollar impact.
How do you prove the demand response actually worked?
With measurement and verification, not screenshots. We document avoided coincident-peak kW after each season against ERCOT's declared peaks, and we run continuous M&V at 15-minute resolution under IPMVP Option C with baseline models at R² above 0.96. Most M&V is a $10–25K once-a-year study; ours runs automatically every 15 minutes, so a curtail call and its result are both on the record.
Sec. 06 · Start before June

Find out what your four
worst hours cost you.

Send a recent utility bill — we'll show you what's hiding in your interval data: where your demand and transmission money is going, and what a pre-cool-and-coast season would hold onto. Free RMS trial — all it needs is one recent utility bill.

CompanyBuilding Optimization Technologies, LLC · G&W Holdings
Base800 Wilcrest Dr, Houston, TX 77042
Request a 4CP risk read