Condensate recovery: what it saves and what it costs to install

Walk the back of most dyeing floors in Savar or Narayanganj and you will find condensate running to drain — hot, chemically treated water that the boiler paid to produce, going into the effluent plant that then pays to handle it. Every kilogram dumped has to be replaced by cold, hard borehole water, softened, dosed and heated again from roughly 30 °C.

Condensate recovery is one of the few steam projects where the savings appear in three separate budget lines at once: fuel, water treatment and effluent. That makes it easy to justify and easy to overstate, so the arithmetic below keeps the three separate.

What a kilogram of returned condensate is worth

Condensate returning at 90 °C carries about 377 kJ/kg. Make-up water at 30 °C carries about 126 kJ/kg. The heat saved is roughly 251 kJ for every kilogram returned — and that is only the first of three savings.

Take an illustrative plant generating 6 t/h of steam, 6,000 hours a year, currently returning 25% of its condensate and aiming for 65%:

Substitute your own marginal fuel cost and water cost — the point of the table is the structure, not the total. The water line alone is significant where a plant buys treated water or pays for effluent volume.

The third saving is indirect. Returning clean condensate lowers feedwater TDS, which lowers the required blowdown rate, which saves fuel again. On the numbers in our blowdown article, lifting return from 20% to 50% cut required blowdown from 9.4% to 5.4%. Do not double-count it: calculate the blowdown saving from the new feedwater TDS, not by adding a separate percentage.

Find out what you are actually returning

Spirax Sarco’s introduction to condensate recovery and its reference on flash steam explain the two-phase behaviour that governs how the return main must be sized.

Most plants over-estimate their return rate. Measure it rather than assume it:

• Meter the make-up water into the feed tank for a full week of normal production. Make-up divided by steam generated gives your true loss rate.
• Check the feed tank temperature. A tank sitting at 45 °C is not receiving much condensate, whatever the drawing says. A well-returned system usually runs 75–90 °C.
• Walk every drain point with a torch. Open drains discharging continuously are either dumped condensate or failed-open traps, and both belong on the same survey.
• List the users excluded from return and why. "It was contaminated once in 2023" is not a current engineering reason.

Building the return system

The most common installation mistake in retrofits is treating the return main as a water pipe. Condensate at pressure flashes as it enters the line; the resulting mixed flow needs far more cross-sectional area than the liquid alone. An undersized main raises backpressure on every trap discharging into it, which stalls drainage on the dyeing machines the project was meant to help.

Building the payback case

Assume the installed cost of the receiver, pumps, 180 m of return main, insulation, monitoring and commissioning for the illustrative plant above comes to 2.2 million BDT. Against a fuel saving of 4.68 million BDT a year, simple payback is roughly six months. Halve the assumed recovery rate and it is about a year.

Present both cases. The recovery rate achieved depends on how many user points can actually be tied in, and the honest version of the business case shows what happens if only part of the plan is delivered. Add the water and effluent saving as separate lines so finance can see which parts of the benefit are certain and which depend on tariff assumptions.

Phase the work by return line, not by budget year. Each completed line delivers its own saving, and the feed tank temperature is a live measurement of progress that anyone can read.

Before you start

Fix failed traps first. Returning condensate through a leaking trap population sends live steam into the return main, pressurising it and stalling every other trap on the line. Survey the traps, repair them, then commission the return system into a network that behaves the way it was sized to behave.

Check that the boiler feed pumps can handle a hotter feed tank. Raising feed temperature from 45 °C to 85 °C changes the net positive suction head available, and a pump that cavitates on hot water will make the project look like a failure for reasons unrelated to the recovery scheme.

Read this with the steam trap audit method and the efficiency calculation methods so the savings land in one consistent energy balance. To survey your return rate and scope a recovery system, contact our engineering desk or see our repair and maintenance service.

Related reading: Achieving LEED certification in Bangladesh via steam and boiler upgrades and Why Bangladeshi groundwater destroys boiler tubes.