Air compression is thermodynamically wasteful. A 37kW compressor might deliver 11kW of useful pneumatic power to the tools. The other 26kW heats the air, the oil and eventually the compressor room. Every hour that machine runs, you're generating heat you've paid for and dumping it outside through the building fabric.
Heat recovery systems capture that waste and put it to work. The payback calculations are often better than the capital investment case for VSD, and far better than most energy consultants will propose first.
Where the heat actually goes
On a typical oil-injected rotary screw, 80-94% of input electrical energy is recoverable as useful heat. The distribution breaks down roughly as:
- 80-85%: in the cooling air stream (air-cooled machines) or cooling water (water-cooled)
- 5-10%: carried in the compressed air discharged to the system
- 2-5%: radiated to the machine room
Water-cooled machines offer the best recovery potential. The heat is concentrated in cooling water at 60-90°C at the outlet, a useful temperature for a range of applications. Air-cooled machines can also be recovered from, but at lower temperature and lower quality.
Air-side recovery: the cheapest route with the most limitations
The simplest approach on an air-cooled machine is to duct the hot cooling air to somewhere useful. A heat recovery kit, insulated ducting, a motorised damper, and a thermostat, routes 40-60°C exhaust air to an adjacent space in winter and diverts it outside in summer.
Capital cost is low: typically £2,000-6,000 installed for a medium-sized machine. Payback can be very fast if the adjacent space has significant heating demand, a workshop, a loading bay, a stores area that currently uses gas or electric heating.
The limitation is temperature. Hot air at 40-50°C is useful for space heating and little else. You can't use it to heat water to a functional temperature or for any process heat requirement that needs reliable supply above 40°C. It's also seasonal, the heat is only useful when the building needs heating, which in a UK industrial environment is roughly October to April.
Water-side recovery: the better economics case
Water-cooled machines, and air-cooled machines retrofitted with an oil-water heat exchanger, produce hot water at 60-80°C. That temperature is useful for domestic hot water preheat, process water heating, rinse tanks, washing facilities, or as a feed into a low-temperature heating circuit.
A 55kW water-cooled compressor running 4,000 hours per year can produce around 165,000-180,000 kWh of hot water energy annually. At current gas prices, replacing gas water heating with recovered compressor heat saves roughly £6,000-9,000 per year depending on your gas rate and how well your loads match.
Capital cost for a full water-side installation, heat exchanger, pump, controls, buffer vessel, pipework, typically runs £8,000-18,000. Payback of 1-3 years is achievable for sites with consistent hot water demand. That's genuinely competitive with almost any other energy efficiency investment in an industrial building.
The oil-water heat exchanger retrofit is available from most major manufacturers for their standard ranges. It doesn't require a water-cooled machine from the outset, though water-cooled machines offer better efficiency and higher return temperatures.
Why most sites don't do this
The reason heat recovery is underused isn't cost or technical complexity. It's that the compressor sits in an engineering budget and the heating system sits in a facilities or utilities budget. Nobody is looking at both at once.
The site I worked with that had the most effective heat recovery, a 75kW GA running into a plate heat exchanger feeding the canteen and shower block, had a facilities manager who asked the question. The compressor had been on site for four years before anyone looked at it.
A genuine heat load is the prerequisite. If your site has no hot water requirement, no adjacent space heating, and no process heat demand, heat recovery doesn't work, you can't store what you can't use. But if you have any of those loads, measure them, compare them to your compressor's recovery potential, and build the case. It's usually there.