Recovered Steam with No User Is Just a Different Vent

A flash recovery saves only the steam a user takes, hour by hour. Past jobs show four ways the user goes missing, and one sheet that catches all four before anyone sizes the hardware.

Steam drifting over condensate vessels and pipework in a plant yard
On this page
  1. The saving is the smaller of two numbers, every hour
  2. 1. The user stops, and the safety valve becomes the vent
  3. 2. The saving is credited at the user’s demand, not at the flash
  4. 3. The user is too hot, or runs at another time
  5. 4. The user is a heat sink with its own ceiling
  6. The check: flash and users on one sheet, hour by hour
  7. Where the user was there first
  8. What we need from you

A flash recovery that goes in without a steady user does not stop the vent. It moves it: to a safety valve, a dump line, a drain, or a feed tank that runs at boiling point. The steam is recovered on the saving sheet and lost on the plant.

The agrochemical flash-recovery post puts this in one line: a user took the recovered steam every hour. This post is about how that condition fails in practice. The lessons come from past jobs, with the names taken off, and each one is a different way of losing the user.

The saving is the smaller of two numbers, every hour

In any hour, the steam a recovery saves is the flash produced or the user’s demand, whichever is smaller. If the flash is larger, the surplus has to leave somewhere, and wherever it leaves is the new vent. If the demand is larger, make-up steam covers the rest.

So a saving sheet that multiplies one figure by 8,000 hours is right only if that figure is the smaller one in every one of those hours.

1. The user stops, and the safety valve becomes the vent

One plant fitted a flash vessel holding about 1.2 bar g, with a reducing station topping up the low-pressure header. The flash was about 83 kg/h. The only users on that header were two preheaters taking 89 kg/h. On paper, a neat match.

When one preheater shut down, nothing took the flash. The pressure climbed to about 2 bar g and the safety valves lifted. The proposed fix was a ½-inch line to dump the surplus into the boiler feed tank. At 83 kg/h that line would run at about 96 m/s, and dumping the full flash would heat the feed tank by about 22 °C, to near 100 °C. That would have brought back the feed-tank venting the job had just removed.

There was a second lesson in the same plant. The earlier audit had planned to move 701 kg/h of users onto the low-pressure header. Only the two preheaters could actually go: the other exchangers needed the higher pressure to do their duty. The list of users shrinks when each exchanger is checked at the lower pressure.

2. The saving is credited at the user’s demand, not at the flash

In a bottling plant, 400–600 kg/h of bottle-washer condensate flashed at 0.5 bar g gives 30–50 kg/h of flash. The proposed user was a set of crate washers that need water at 40 °C and took 75–125 kg/h of live steam.

The payback sheet credited 100 kg/h of live steam saved per line: the crate washers’ whole steam use. Over two lines that is 200 kg/h claimed against 60–100 kg/h of flash available, an overstatement of two to three times. The user was real. It was simply bigger than the flash, and the saving can never be bigger than the flash.

3. The user is too hot, or runs at another time

Too hot. Flash at 0.5 bar g condenses at about 111 °C. One audit proposed it for press dies that work at 150–155 °C. It could pre-heat them; the rest had to come from live steam.

Out of step. Another scheme used flash from a separator to pre-heat a batch melting tank to 100 °C, after which live steam took the product to 150 °C. At the changeover, the design dumped all the flash and condensate to drain. The flash had a user only during the pre-heat window, and even that depended on a cooling step elsewhere producing the flash at the same time. For the rest of the batch, the separator was the vent.

A batch source feeding a continuous user, or a continuous source feeding a batch user, needs the same question: do they run in the same hours?

4. The user is a heat sink with its own ceiling

A boiler feed tank takes flash only until the feed water reaches boiling point: feed flow × (100 − make-up temperature) kcal/h. Five ten-minute checks shows a proposal that claimed two and a half times that ceiling. A cooling-water condenser recovers the water but not the heat, unless the warm water goes somewhere useful. When flash goes to water instead of a steam user, the water is the user, and it has a limit (flash vessel, Jetomat or vent condenser sets out that choice).

The check: flash and users on one sheet, hour by hour

This is the steam balance drawn for one stream.

  1. The flash, by hour or by shift, from the condensate flow at the trap-inlet pressure. Do a normal day and a low day.
  2. Each user that could take it: its pressure, its temperature, its demand in each hour, and when it stops (shutdowns, batch steps, weekends).
  3. Each hour, take the smaller of the two. The sum over the year is the saving.
  4. Each hour the flash is larger: where does the surplus go? That is your vent. Give it a route sized for the full flash, or lift it with a Jetomat into a header that takes steam every hour, or recover less.
  5. Each hour the demand is larger: size the make-up for the user’s whole demand, because the flash can fall to nothing.

Where the user was there first

In the Bharuch agrochemical plant, the 3.5 bar process header was already fed through a reducing station from the 16.5 bar main. The Jetomat replaced part of what the reducing station passed, so every kilogram it delivered had a taker. At an industrial laundry in Spain, a 4 bar g user took all its steam from the boiler; with the flash lifted to it, boiler steam to that user fell from 326 to 208 kg/h.

In both, the user existed before the hardware did. Find the user first, then size the recovery to it.

What we need from you

  • The condensate flow by hour or shift, and the trap-inlet pressure of each source.
  • Each candidate user’s pressure, temperature and hourly demand.
  • When each user stops: shutdowns, batch steps, weekends.
  • Where the surplus goes today: a vent, a safety valve or a drain.
  • Any header fed through a reducing station, with its flow.

FAQ

What if the flash is sometimes more than the user takes?

Then the surplus needs a route sized for the full flash: a modulating back-pressure or spill valve to a sink that can use it, or a Jetomat that lifts it to a header with a steady demand. A safety valve or a small dump line is just a vent with a different name.

How do I work out the real saving?

For each hour, take the smaller of the flash and the user's demand, and add those up over the year. Multiplying either figure alone by the running hours overstates the saving whenever the other one is smaller.

Is the boiler feed tank a safe default sink?

Only up to what the feed water can absorb: feed flow × (100 − make-up temperature in °C) kcal/h. Past that, the tank vents, and on an open tank near 100 °C the feed pumps can struggle as well.

Find out what your steam system could save

Share a few details about your plant. Our engineers will review your steam and condensate system and come back with a no-obligation assessment of the savings and payback.