Waste heat recovery in hard-to-abate plants: what actually pays back

Start with where the heat actually goes

In cement, around 75% of a typical dry process plant’s primary energy arrives as kiln fuel, and pyroprocessing takes 93 to 99% of total fuel consumption. In glass, more than 80% of site energy is consumed at the hot end, where the batch is melted. Everything else on site, grinding aside, is small by comparison. So when a hard-to-abate plant asks where to start, the honest answer is rarely a technology. It is a heat balance.

Waste heat recovery sits at the top of most wish lists in this sector because the losses are visible. You can stand under a preheater tower and feel the exhaust duct radiating. The harder question, and the one that decides whether a scheme pays back, is not whether heat is leaving the stack. It is whether that heat is genuinely surplus, at what temperature, in what quantity, and during which operating states.

The two streams that matter in cement

Cement waste heat recovery targets two streams: preheater exhaust, typically at 300 to 400 C, and clinker cooler vent air. Both are large, both are continuous whenever the kiln runs, and both are routinely described in vendor proposals as free energy.

They are not free. On most plants the preheater exhaust is already doing work: it dries raw meal in the raw mill and fuel in the coal mill before anything else gets a look at it. Raw meal moisture is the single variable that most determines how much heat is genuinely available for recovery. A plant grinding wet raw materials may have very little surplus exhaust heat while the raw mill runs, and far more when it stops. Any feasibility study that quotes one recovery figure for all operating states has not been done properly, and we have reviewed plenty that do exactly that.

Heat to heat usually beats heat to power

The instinct is to generate electricity. Steam Rankine and organic Rankine cycle systems are proven in cement, and IFC analysis shows waste heat recovery can supply up to 30% of a cement plant’s own electricity needs. On the right site, with the right heat balance, that is a real and bankable outcome.

But power generation is also the most capital-intensive way to use recovered heat, and it converts high-grade heat at modest cycle efficiency. Before sizing a turbine, check the heat-to-heat options: drying duties for raw materials, fuels, aggregates or secondary products, space and process heating on or near the site, and preheating of combustion air or feed streams. Displacing fuel directly typically pays back faster than generating power, needs less plant, and every displaced gigajoule of fossil fuel now avoids carbon cost as well as fuel cost. Power should be what you do with the heat that is left after the cheaper uses are served, not the default answer.

Glass: the regenerator is the heat recovery system

Glass melting furnaces typically run at 50 to 60% thermal efficiency, with flue gas losses of 25 to 35%. That is why regenerator condition dominates glass energy audits: the checkerwork is the heat recovery system, and its health decides how much preheat the combustion air receives. Furnace energy intensity commonly drifts upward by 2 to 3% per year over a campaign as regenerators age and checkers foul, so a furnace that looks fine against last year may be well off its commissioning performance. Tracking specific melting energy against campaign age, regressed against pull rate, is the only way to see it clearly.

Two further levers sit beside the regenerators. Every additional 10% of cullet in the batch cuts melting energy by roughly 2.5 to 3%, because remelting glass takes less energy than reacting raw batch. And batch and cullet preheating, which puts flue gas heat back into the cold charge, saves around 15% of melting energy where layout and cullet quality allow it. Neither requires waiting for a rebuild.

What pays back, in order

Plant by plant the ranking shifts, but across the cement, glass and building materials sites we work with, the pattern is consistent enough to state as a default sequence.

• Measurement first. Map every heat stream by temperature, mass flow and operating state, and build a site heat balance. This is days of engineering, not months, and it stops capital chasing the wrong stream.
• Combustion and losses. Excess air and O2 trim, kiln shell loss and refractory management, cooler performance. Commonly worth 2 to 5% of specific heat consumption in cement, at little or no capital.
• Heat to heat. Recovery to drying and heating duties, which displaces fuel directly and usually carries the shortest payback of any capital scheme on the register.
• Glass hot end housekeeping. Regenerator recovery, cullet ratio, batch preheating. Campaign-time measures that buy margin while rebuild capital waits.
• Heat to power. Steam Rankine or ORC on what remains, sized from the measured surplus rather than the nameplate exhaust flow.
• The quiet quick wins. Compressed air for IS forming machines, combustion air fans, bag filter fans and annealing lehrs typically yield 5 to 15% savings from leak management, pressure optimisation and variable speed drives.

Carbon changes the maths

For years, waste heat recovery economics were argued on fuel price alone, and marginal schemes stayed on the shelf. The EU ETS has changed that arithmetic, and the change accelerates from 2026 as free allocation phases out and CBAM phases in, with cement among the first six sectors covered. Every gigajoule of fossil fuel displaced now carries an allowance value on top of its fuel value, and that value compounds across the life of the scheme.

Practically, this means registers built three years ago are stale. Measures ranked by fuel-only savings need re-ranking with carbon included, and schemes that missed the hurdle rate at the last review may clear it comfortably now. We model energy and carbon together as standard, because that is the number the investment committee is actually deciding on.

How to start

Begin with a thermal audit that maps the heat honestly: every stream, every operating state, drying duties respected, surplus quantified. Our energy audit and compliance team delivers exactly that, with a ranked and costed register as the output rather than a description of your own plant. Where the register points at capital schemes, our design and projects engineers take recovery systems from feasibility through detailed design and commissioning, and verify the result to IPMVP.

Waste heat recovery in this sector rewards discipline and punishes optimism. Measure first, displace fuel before generating power, and let the heat balance, not the brochure, choose the scheme. There is more on how we work with kilns, furnaces and the rest of the hot end on our cement, glass and building materials page.