How much of a home fuel cell's waste heat is actually useful?

The short answer: waste heat counts only when the house needs it at the same time

How much of a home fuel cell's waste heat is actually useful depends on timing: heat recovered from the stack only has value in the hours when the house wants heat at the same moment the fuel cell is running to make electricity. Heat made when the hot water tank is already full gets rejected or simply raises tank losses.

That sounds obvious, but cogeneration is often credited with all of its recovered heat. A 2026 open-access study in the journal Energies argues that this overstates the benefit, and shows that the useful share swings widely with which heating loads are connected and how the fuel cell is told to run.

How residential fuel cell cogeneration works

A home fuel cell combined heat and power unit makes electricity by combining hydrogen with oxygen from the air and captures the heat from that reaction for hot water. Japan's ENE-FARM program is the best-known example: those units make hydrogen from the natural gas or LPG already piped to the home, and the hot water storage tank is built into the power unit. Japan's government reported more than 200,000 installed units as of May 2017 and a power-generation efficiency of 52 percent for one sub-1 kW model.

Every cogeneration machine has a heat-to-power ratio: how much useful heat it makes for each unit of electricity. An MIT thesis on micro-CHP notes that small Stirling engines sit at the extreme, around 6 to 1, while fuel cells make comparatively more electricity per unit of heat. The same thesis points to a basic problem in single-family homes: thermal and electrical loads often do not occur at the same time, which is why storage tanks or a grid connection are usually needed.

What the heat-to-power matching study found

The Energies study simulated a small off-grid house in Aomori, Japan, hour by hour in TRNSYS 18. Solar panels charge a battery and run an electrolyzer that makes hydrogen; a 700 W PEM fuel cell turns that hydrogen back into electricity when solar falls short, and its heat goes to a stratified hot water tank that also takes heat from evacuated-tube solar collectors.

The authors compared the fuel cell's own heat-to-power ratio, which rose from 0.39 at a quarter of rated load to 0.667 at full load, with the house's demand ratio in every hour. With hot water as the only heat load, the fuel cell was potentially heat-rich in 88.7 to 92.3 percent of the hours it was needed for electricity, meaning it made more heat than the house could use. Once space heating was added, that fell to 2.2 to 3.7 percent, because the house's residual heat demand was 12.6 times its electrical shortfall.

Delivered heat was not all a gain either. Compared with a run with the fuel cell switched off, only 25 to 61 percent of the heat it put into the tank reduced unmet heating demand; 22 to 54 percent displaced heat the solar collectors would have supplied, and 15 to 22 percent went into extra tank losses. Under a fixed calendar schedule with hot water only, 75 percent of the recovered heat could not get into the tank at all.

Why the dispatch rule matters

A dispatch rule decides when the fuel cell runs: on a calendar, following the electrical load, following the heat demand, or a combination. In the study, following the electrical load raised the house's self-sufficiency from 88.9 to 96.3 percent. Calendar dispatch, and heat-led dispatch with space heating, would have needed 2.8 to 5.0 times more hydrogen than the system could produce.

Hydrogen storage is a lossy battery. In the baseline case the solar-to-hydrogen-to-electricity round trip was 29.1 percent efficient. Electrolyzers from 40 to 110 square centimeters of electrode area gave identical service, while a 25 square centimeter unit could not keep up and its hydrogen inventory fell every year.

Limits of the result and what to take from it

This is one simulated house in one climate. The authors note that the stack model was checked only against catalog ratings, not a commercial unit, that some heat exchanger values were assumed, and that their hour-by-hour bounds ignore heat carried over between hours in the tank.

The practical lesson holds more broadly: when judging any home cogeneration system, ask what the heat will be used for and when. A unit serving only hot water will often have heat to spare, while one tied into space heating can use nearly all of it, and the operating rule can matter as much as the hardware.