Cogeneration Explained: How It Works and What It Earns

Cogeneration is one machine doing two jobs. It burns fuel to turn a generator, and instead of venting the exhaust and the cooling water, it pipes that heat somewhere useful — process steam, hot water, space conditioning, or an absorption chiller. The electricity and the heat come out of the same fuel, and that is the whole of the idea.

The term dates to the era when a mill made its own power and heated its own buildings from one boiler. The arrangement has not changed. What makes it still pay is a fact about how electricity is normally produced.

What cogeneration is

Cogeneration is the simultaneous production of electricity and useful thermal energy from a single fuel input. Combined heat and power, abbreviated CHP, is the same thing under a different name; the two are used interchangeably in engineering practice and in regulation.

Three conditions have to hold before an installation counts:

• One fuel stream drives both outputs. Two separate machines in the same building are not cogeneration.
• The thermal output has a genuine use. Heat dumped to a cooling tower does not count, however much of it there is.
• The generation and the thermal load are close enough to connect. Heat does not travel far.

That third condition is why cogeneration is a site technology rather than a utility one. Electricity moves hundreds of miles down a wire at modest cost. Hot water does not. So the plant goes to where the heat is wanted, which puts cogeneration at buildings and factories rather than at generating stations.

Why it exists: the heat a power plant throws away

A conventional power station turns a minority of its fuel into electricity and rejects the remainder as heat — up a stack, into a river, through a cooling tower. The EPA's CHP Partnership puts the average efficiency of fossil-fuelled generation in the United States at 36 percent, with a further 5.3 percent lost on average carrying the power through transmission and distribution. Add a separate boiler raising heat on site, and the combined fuel efficiency of the two systems together comes to 50 to 55 percent.

A cogeneration plant does not have a better thermodynamic cycle than a power station. Its electrical efficiency is usually worse. What it has is somewhere to put the reject heat, and that one difference takes total system efficiency to 65 to 80 percent, with some installations approaching 90.

The arithmetic is the argument. Nothing about cogeneration is clever. It simply stops discarding two-thirds of the fuel.

How a cogeneration plant works

Four parts, in order:

• A prime mover burns fuel and turns a shaft.
• A generator on that shaft produces electricity, usually consumed on site rather than exported.
• Heat recovery captures energy from the exhaust and, depending on the machine, from the jacket cooling water and the lubricating oil as well.
• A thermal loop delivers the recovered heat as steam, hot water, or chilled water by way of an absorption chiller.

The prime mover sets the character of the installation more than any other single choice.

Reciprocating engines

Large piston engines running on gas. They dominate the small and middle of the market, from a few hundred kilowatts to around 10 megawatts, because they start quickly, tolerate part load, and give up heat from several places at once. Most of what they recover arrives as hot water rather than high-pressure steam, which suits hospitals, apartment blocks and leisure centres.

Gas turbines

Simpler, larger, and better suited to steam. A turbine's exhaust is hot and continuous, so a heat-recovery steam generator placed behind it raises process steam at useful pressure. Turbines want to run at or near full load and lose efficiency sharply when throttled, which makes them a fit for continuous industrial duty and a poor fit for a building with a daily cycle.

Microturbines and fuel cells

At the small end, microturbines of 30 to 300 kilowatts and fuel cells trade capital cost and electrical efficiency for low maintenance, low emissions and quiet operation. Fuel cells in particular produce notably clean power. One caution for Pennsylvania: the resource classification that applies to a fuel cell is not the same as the one that applies to a gas-fired engine or turbine, so confirm which tier your output falls into before modelling any credit revenue.

Steam turbines and bottoming cycles

Where a process already throws off high-temperature waste heat — a kiln, a furnace, a flare — a steam turbine placed downstream can make electricity from energy that would otherwise be lost. This is a bottoming cycle, and it inverts the usual order: heat first, power second.

Cogeneration examples, by sector

The textbook answer to what counts as an example of cogeneration is a district heating plant. The useful answer is the set of sites that actually carry a thermal load all year, because that is the only kind of site where the economics hold up.

Hospitals

Continuous demand for hot water, sterilisation steam and space conditioning, no seasonal shutdown, and a standing reason to value power that survives a grid outage. Engine-based systems from the high hundreds of kilowatts into the low megawatts are typical. We have written up the Pennsylvania credit position for this sector in alternative energy credits for hospitals and health systems.

Universities and campuses

Several buildings on one site, a central plant already standing, and distribution piping already in the ground. Campus plants are often the largest cogeneration installations in a region, and the same central-plant logic extends to the rest of a campus estate — see credits for schools and universities.

Wastewater treatment

Anaerobic digestion produces methane as a by-product of treating sewage. Burning it on site for power and digester heating turns a disposal problem into both outputs at once. It also raises a genuinely interesting classification question, which we worked through in digester gas cogeneration: Tier I or Tier II?.

Manufacturing with process heat

Food and beverage, pulp and paper, chemicals, refining — anywhere steam is an input to the product rather than a comfort measure. These are the strongest cogeneration sites in existence, because the thermal load is as steady as the production schedule. The Pennsylvania angle is covered in process efficiency credits for manufacturing plants.

Data centres and commercial offices

Both are harder. A data centre carries an enormous electrical load and almost no heat demand, so the recovered energy needs an outlet — absorption cooling, or a neighbour willing to take it — before the case closes. Office buildings have the opposite difficulty: a thermal load that collapses for several months of the year.

Cogeneration, combined heat and power, and trigeneration

The vocabulary is more confusing than the technology.

• Cogeneration and combined heat and power mean the same thing. The first is the older term and still the more commonly searched; the second is standard in United States policy and regulation.
• Trigeneration, or combined cooling, heat and power, is cogeneration with an absorption chiller added, so the recovered heat can be turned into chilled water in summer. It is a way of giving the heat somewhere to go year-round.
• A topping cycle generates electricity first and recovers heat afterwards. This is what nearly everyone means by cogeneration.
• A bottoming cycle takes high-temperature process heat and generates electricity from it on the way out. Sometimes called waste heat to power.
• Total system efficiency counts electricity and useful heat together against fuel input. Electrical efficiency counts only the power. A cogeneration plant looks poor on the second number and excellent on the first, and quoting the wrong one is the commonest way these systems get misrepresented in either direction.

The constraint that decides whether it works

Most explanations of cogeneration stop at the efficiency figure. The figure is real, but it is conditional, and the condition is the part a prospective owner needs.

Recovered heat only counts if something consumes it. A total system efficiency in the seventies assumes the thermal output is being taken. In a building whose heat demand disappears in April, the same machine spends a third of the year performing like a mediocre power station.

That has a design consequence which catches people out. A cogeneration plant should be sized to the thermal load, not the electrical one. Sizing to the electrical peak produces a machine too large for the heat demand most of the time, which leaves two bad options: dump the surplus heat, or shut down and restart. Cycling is hard on engines and worse for the return than running a smaller unit continuously.

The practical screen is short. Does the site carry a thermal load that persists through the year? Is that load on the same site as the electrical demand? Will the plant run enough hours to repay the capital? A site that fails the first question is not rescued by the other two.

What the output earns in Pennsylvania

Electricity generated on site displaces electricity that would otherwise be purchased, and that is the first and largest part of the return. The recovered heat displaces boiler fuel, and that is the second.

There is a third, less widely known. In Pennsylvania the metered electrical output of a qualifying cogeneration system can be registered to earn alternative energy credits, which are sold separately from the power itself. That revenue does not depend on exporting anything: the credits attach to generation, including generation consumed behind the meter.

Two things are worth knowing before modelling it, and both are covered properly elsewhere rather than restated here:

• The eligibility route turns on system size, and the calculation, the metering and the registration sequence are set out in how cogeneration generates alternative energy credits.
• Pennsylvania imposes no system-efficiency threshold on cogeneration for this purpose, which surprises engineers who have worked under other states' programmes. The evidence is in Pennsylvania's AEPS program has no efficiency threshold.

One timing point matters more than its length here suggests. The credit clock starts when a complete application is on file — not when the plant was commissioned. Years of unregistered operation build up no retrospective entitlement, and no filing made later reaches back for them.

Where to go next

• Cogeneration and Pennsylvania alternative energy credits — the Pennsylvania position, with a calculator for nameplate capacity, operating hours and capacity factor.
• How cogeneration projects qualify — the statutory eligibility test.
• Measuring cogeneration output — meter placement, parasitic load and what gets counted.
• Renewable energy credits explained — if the credit side of this is unfamiliar.

Frequently asked questions