Why Condensate Is the Expensive Half of a Steam System

Why this matters

Owners look at a condensate return rate the way they look at a filter change: worth doing, not urgent. That reading survives because the only loss anyone quantifies is the sensible heat in the water, and that number comes out modest enough to shrug at. The heat is the slowest of the three charges a lost pound of condensate incurs, and the only one denominated in the same unit as the fuel bill. It also charges you a pound of already-treated water, and then it charges you again through blowdown, because make-up carries dissolved solids that condensate does not, and the boiler has to throw water away to keep up. That third charge compounds the first two, and it is the one that never appears on anyone's estimate.

Blowdown is the sharpest hazard in this article, so it goes first

Manual bottom blowdown discharges boiler water at boiler pressure and temperature. Do not operate one unless the discharge is piped to a blowdown separator or blowdown tank vented to atmosphere and sized for the boiler, never to an open floor drain, and never to a sewer connection that has not been checked against the discharge temperature limit your local authority sets.

Operate it from beside the valve, never standing over it, with face protection under 29 CFR 1910.133, sleeves down and gloves rated for hot liquid contact. On the usual two-valve arrangement, open the sealing valve nearest the boiler fully first, then crack the outer blowing valve; to stop, close the blowing valve first and the sealing valve last. That order keeps the closing shock and the erosive flow off the seat of the valve that has to hold pressure, and reversing it is how a sealing valve gets wire-drawn and starts to weep.

For any work that opens the water side of a boiler or a return line, isolate and lock out under 29 CFR 1910.147, vent to atmosphere, and confirm zero gauge pressure on a readable gauge before a joint is broken. Water in a return line at line pressure is above its atmospheric boiling point and flashes out of the opening you make.

The three charges, named

Charge one: sensible heat. Returned condensate arrives at the feed tank far hotter than make-up, so the boiler starts from a higher enthalpy. A sibling article works this out from a steam table and lands at roughly 16 percent of a boiler's heat input sitting in condensate leaving a 15 psig terminal, against 60 F make-up.

Charge two: treated water. Every pound not returned is a pound of make-up that has to be softened, dosed and deaerated. Condensate is essentially distilled water. Make-up is not.

Charge three: blowdown, and this one multiplies. Boiler water concentrates as steam leaves and solids stay behind, so the boiler must dump concentrated water to hold its solids limit. How many times the feedwater can be concentrated before that limit is reached is the cycles of concentration, and it is set by the feedwater's dissolved solids against the boiler's allowable. Feedwater is a blend of clean condensate and dirty make-up. Cut the return rate and the blend gets dirtier in direct proportion, cycles fall in direct proportion, and blowdown rises. The chemistry of cycles and limits is owned by a water treatment article and is not re-derived here; what this article owns is the mechanical consequence, which is that blowdown is a second, larger heat and water loss stacked on top of the first one.

The gate

Return the condensate unless returning it puts the boiler at risk. Where it does, recover the heat without recovering the water. That single rule resolves both cases below in opposite directions, and it is the whole decision.

Case one: an ordinary building heating load

Take a boiler holding a 100 psig header, terminals on 15 psig, condensate leaving them at roughly 219 Btu/lb, make-up arriving at 60 F and roughly 28 Btu/lb, saturated steam leaving the boiler at roughly 1,190 Btu/lb, and boiler water at roughly 309 Btu/lb. Steam demand is 1,000 lb/hr. Assume the make-up on its own would support four cycles of concentration against the boiler's solids limit, and that returned condensate contributes essentially none.

At 80 percent return - that is, 80 percent of the feedwater is returned condensate - the feedwater is one fifth make-up, so cycles run at four divided by 0.20, which is 20, and blowdown is one twentieth of feedwater, 5 percent. Feedwater is therefore 1,000 divided by 0.95, about 1,053 lb/hr: 842 lb/hr returned, 211 lb/hr make-up, 53 lb/hr blown down. Blended feedwater enthalpy is about 181 Btu/lb. Heat into the water is steam out plus blowdown out minus feedwater in: about 1,190,000 plus 16,400 minus 190,300, or roughly 1,016,000 Btu/hr.

At 40 percent return, the feedwater is three fifths make-up, cycles fall to four divided by 0.60, about 6.7, and blowdown rises to about 15 percent. Feedwater is 1,000 divided by 0.85, about 1,176 lb/hr: 471 lb/hr returned, 706 lb/hr make-up, 176 lb/hr blown down. Blended feedwater enthalpy is about 104 Btu/lb. Heat into the water is about 1,190,000 plus 54,400 minus 122,900, or roughly 1,121,000 Btu/hr.

Now read what moved. Halving the return rate raised boiler heat input by about 10 percent, which is the number an owner shrugs at. It raised make-up from 211 to 706 lb/hr, a factor of about 3.3, and it raised blowdown from 53 to 176 lb/hr by the same factor. Softener regenerations, chemical dosing and deaerator duty all scale with make-up, so they all went up by roughly that same 3.3.

Split the 10 percent to see the compounding. Hold blowdown at its original 5 percent and drop only the return rate, and heat input rises to about 1,096,000 Btu/hr, an increase of about 80,000 Btu/hr. Let blowdown rise as it actually does and you add a further 25,000 Btu/hr. So about a quarter of the heat penalty is the blowdown that a straight "lost sensible heat" calculation never counts, and that quarter arrives with three times the treated water and three times the chemistry attached to it. The heat number is not wrong. It is simply the only one of the three denominated in the same unit as the fuel bill, which is why it is the only one anybody prices.

Case two: the same rule, the opposite answer

Now a shell-and-tube exchanger heating a process fluid, with the process side running at a higher pressure than the steam side. A tube leak here pushes process fluid into the condensate, not the other way round, and that condensate goes to the boiler.

The gate says return it unless returning it risks the boiler, and here it plainly does. Depending on what the process fluid is, contamination can mean oil that blankets heat transfer surface and causes carryover, or dissolved solids that overwhelm the treatment, or something that attacks the boiler metal directly. The result is a boiler out of service for a chemical clean and a jurisdictional inspection before it can be returned to service, measured in days of no heat, against a heat recovery worth about a tenth of the fuel input.

So the answer flips, and it flips in a specific direction rather than to "do nothing":

• Monitor and dump. Continuous conductivity monitoring on that condensate stream with an automatic dump valve that diverts to drain on an out-of-range reading, and a manual dump the operator can hit. This keeps most of the recovery and buys an alarm.
• Or recover heat without recovering water. Run the suspect condensate through a heat exchanger that preheats make-up, then send the condensate to drain. You keep charge one and accept charges two and three.
• Never blend a suspect stream into a clean return upstream of the monitor, because a blended stream dilutes the very reading you are relying on to catch the leak.

Notice what did not change: the rule. The same sentence produced "return it" for a heating coil and "monitor or divert" for a process exchanger, because the risk term is the only thing that moved.

What would move either answer

Distance and elevation. Returning condensate from a remote building may need a pump and a line, and where the run is long enough that the recoverable enthalpy is small, the honest call is to recover heat locally and return nothing.

Make-up quality. The 3.3x multiplier above came from make-up that supports four cycles on its own. Soft, low-solids make-up supports more cycles and blunts charge three considerably; hard or high-solids make-up sharpens it. That is exactly why the same lost return rate hurts two plants differently.

Flash recovery. Condensate dropping from 15 psig to a vented receiver flashes about 4 percent of its mass to steam, and from a 100 psig header roughly 13 percent. That flash is part of what you are trying to keep, and a flash vessel that puts it back into a low-pressure main is a separate recovery from the water itself.

Measuring your actual return rate without fooling yourself

You need two of the three flows: steam produced, condensate returned, and make-up. Take the two you can measure best and derive the third.

Give each meter its error basis and its character before you do the arithmetic. A meter specified as a percentage of reading holds that percentage across its range; a meter specified as a percentage of full scale is worth a fixed quantity anywhere on the dial, which is why an oversized make-up meter chosen for fill rate is nearly useless at running rate. If your two meters are independent instruments with independent random spreads, those spreads combine in quadrature rather than adding, so two meters each good to 2 percent of reading give a ratio good to about 2.8 percent, not 4. If both readings come off instruments sharing one calibration reference, part of the error is a common systematic offset that partly cancels in the ratio, and the honest statement is that you know the ratio better than you know either flow.

The cheap cross-check that needs no meter at all: log softener regenerations per week and chemical drum changes per month over a season. Both scale with make-up volume, both are already recorded somewhere, and a step change in either is a return-rate change that happened at a knowable date.

References

• ASME Steam Tables (IAPWS industrial formulation) for the enthalpy values used above
• 29 CFR 1910.147 for isolation of pressure and thermal stored energy; 29 CFR 1910.133 for eye and face protection during blowdown
• ASME Boiler and Pressure Vessel Code, in the edition adopted by your state's boiler law, for blowdown piping and separator requirements, which bind the installation and reach you through the jurisdictional inspection
• See related: Cycles of Concentration in Plain Terms; The Water Chemistry That Attacks a System; How to Read a Steam Table Without Memorising It