Ask a utility demand response manager to list their flexible assets and you will hear the same three answers: smart thermostats, EV chargers, batteries. The second-largest electric load in the American home almost never makes the list. It sits in a basement or a garage, runs on decades-old resistance technology, and stays invisible until the hot water runs out.

That blind spot is about to become untenable. A federal efficiency rule taking effect in May 2029 will do something no utility program has managed in twenty years of trying: put a grid-connectable flexibility asset into millions of American homes every year, automatically, at the point of sale. The question utilities should be asking is not whether water heaters become a demand response asset class — it is whether their enrollment machinery will be ready when the fleet arrives.

That was the central argument of a conversation we hosted last week with Roch Naleway, Market Development Manager for Utilities at A. O. Smith — one of the “big three” US water heater manufacturers. The full recording is below, and it is worth the hour for anyone who runs a residential flexibility portfolio. Here is the case, and what we think it means.

A Fleet Built by Regulation, Not Persuasion

Today, only around 3–5% of the electric water heater market is addressable for demand response without significant retrofits. That single number explains why water heaters have stayed a pilot-program curiosity while thermostats and batteries became program staples: there was never enough connectable hardware in the field to justify serious program investment.

The US Department of Energy’s 2029 efficiency standard changes the supply side by force. From May 6, 2029, new electric storage water heaters over 35 gallons — the most common residential sizes — will effectively need heat pump technology to comply, and heat pump water heaters ship grid-ready: nearly every unit carries the EcoPort, the standardized CTA-2045 connection port that accepts a communication module and a defined command set. DOE projects the rule will lift the heat pump share of newly manufactured electric storage water heaters from roughly 3% today to more than 50% — and in the over-35-gallon class, grid-ready units become close to the only option.

Then the replacement math takes over. Water heaters last only 10–15 years, so a steady slice of the installed base is replaced every year — by Naleway’s own back-of-the-envelope estimate, on the order of 6–7%, which would mean a couple of million electric units annually arriving grid-capable whether anyone enrolls them or not. As he put it: with anything water-heater related, “it’s not an if this is going to happen — it’s a when.”

No other demand response asset class has ever had its hardware problem solved by regulation. Thermostat programs spent fifteen years and significant incentive budgets buying their way into homes one device at a time. Water heaters will get there as a side effect of an efficiency standard.

The Objections Have Already Been Field-Tested

The standard utility objections to water heater demand response — customers will notice, the hardware is unproven, nobody has done it at scale — have quietly already been answered.

Oregon and Washington have required CTA-2045 connectivity on electric water heaters sold since 2018. That is seven years of installed, connectable units sitting in two states, most of them never enrolled in anything. The Pacific Northwest also hosts the most thorough validation work in the country: Bonneville Power Administration and NEEA ran a major CTA-2045 demonstration across both heat pump and electric resistance units specifically to prove out load shifting as a pathway to utility adoption.

On customer experience, the field data is unambiguous. Across the deployments A. O. Smith has observed, complaints come from a fraction of one percent of customers — and most of those arrive during the 30–60 day baseline period, before a single control event has run. The water heater is, in Naleway’s framing, “essentially a battery” that stores energy in water: control it outside of shower and laundry windows and the customer never knows. Hot water is the product; flexibility is invisible. We made a similar argument about why residential demand response programs struggle — friction and perception, not physics, are what kill participation.

There is even proof that residential demand-side response moves grid-scale numbers when it is asked to: during a California grid emergency, a single utility text message shed roughly 4,000 MW of load within an hour. The demand side shows up. The infrastructure to call on it routinely is what has been missing.

The Real Gate Is the Delivery Chain, Not the Device

Here is the honest part of the case, and the reason this asset class rewards careful program design: an individual grid-connected water heater is worth roughly 0.1–0.5 kW of utility-side capacity, against 0.5–1 kW for an HVAC system. The value per device is real but small — which means, in Naleway’s words, the whole delivery chain “has to be dirt cheap.” Communication module, program management, enrollment, dispatch: every layer has to be lean, because there is no margin to waste between the device and the grid.

That economic constraint is exactly why standards dominate this space. A manufacturer shipping millions of units a year cannot absorb bespoke integrations — as Naleway put it, “we would not be able to accommodate 50 different utilities coming with their own standard.” The viable architecture is the one the industry has been converging on: standardized device control at the bottom (EcoPort and CTA-2045, which we unpacked in our piece on how OpenADR and CTA-2045 enable practical demand response), and open protocols — OpenADR today, with IEEE 2030.5 conversations developing — connecting fleets to aggregators and utilities above it.

This is the layer where our own work with A. O. Smith sat. Codibly supported their integration with one of the leading demand response aggregators through an OpenADR 2.0b implementation — a single standards-based pathway into many programs, rather than fifty custom ones. The general lesson from that project applies to any OEM looking at this market: protocol certification is the market-access ticket, and the aggregator route spreads one integration investment across an entire market — the economics of which we covered in our analysis of how demand response aggregators make money.

Where Programs Already Run — and What They’re Waiting For

Water heater flexibility is not hypothetical. Programs and demonstrations are live today, concentrated — predictably — where energy costs are highest:

Region Program / Initiative What it covers
Washington Puget Sound Energy — Flex Water Heaters Voluntary residential program for electric tank water heaters (resistance and heat pump)
Oregon / Washington Pacific Power residential demand response Water heaters included in the residential DR offering
Pacific Northwest Bonneville Power Administration + NEEA demonstration Large-scale CTA-2045 deployment on heat pump and electric resistance units to validate load shifting
Southeast / Mid-Atlantic Dominion Energy — Water Energy DR Program Demand response for eligible heat pump water heaters
Southeast Duke Energy virtual power plant Water heaters connected within Duke’s VPP build-out
New York NYSERDA, NYPA, Con Edison / Orange & Rockland demonstrations CTA-2045 direct load control; NYPA case highlighted winter peak shaving with resistance units upstate
California SMUD × TECH Clean California Heat pump water heater incentives conditioned on enrollment in a qualifying SMUD demand response program
Hawaii Utility flexibility activity High energy costs make water heater flexibility economic; active grid-flexibility programs
Live water heater demand response programs and demonstrations in the US, June 2026. Activity concentrates in high-rate, coastal markets — the standard diffusion pattern for new DR asset classes.

The pattern in that table is the diffusion curve of every demand response asset class: coastal and high-rate markets move first, the middle of the country follows once the economics are proven. FERC Order 2222 keeps the federal door open by letting aggregated distributed resources bid into wholesale markets.

What is missing is scale commitment. Naleway’s test for when this market truly turns: a utility that commits to connecting 100,000 water heaters a year — not a 500-unit pilot. The technology pathways exist today, the 2029 rule guarantees the hardware supply, and the regulatory access is in place. The remaining variable is a program owner willing to treat water heaters as a portfolio asset rather than an experiment. Given that regulators increasingly push utilities toward the lowest-cost resource — and a paid-for appliance that is already installed is hard to beat on cost — the first movers are likely to set the template everyone else licenses.

Build the Enrollment Machinery Before the Fleet Arrives

The 2029 rule means utilities don’t have to buy this fleet — they only have to enroll it. That inverts the usual program economics, and it rewards whoever is ready first. For utilities, readiness means a program design where enrollment happens at the point of sale or installation, not through mailers chasing customers years later. For OEMs in water heating and adjacent appliance categories, it means doing what A. O. Smith did: pick the open protocols, integrate once through aggregator pathways, and be present in programs before your competitors’ fleets are.

Three years sounds like a long runway. Against a utility program-design cycle — regulatory approval, vendor selection, integration, baseline periods — it is barely enough. The manufacturers have done their part; the standards exist; the demonstrations are conclusive. What happens between now and May 2029 determines who owns the cheapest flexibility asset on the residential grid.

If you are an OEM or a utility working out your own pathway into demand response — protocol selection, aggregator integration, program architecture — our white paper on the business case for demand response is the place to start, or talk to our team about how we approach OEM enablement.