Condensate Returns Turning the Color of Rust
A 350 hp, 150 psig steam boiler returns about 70% condensate from a process plant. Makeup is softened and tests 0.5 gpg total hardness after regeneration, within design. Recent analysis: condensate pH 6.2, iron 2.5 ppm, copper 0.3 ppm; boiler water 3,500 µmho/cm conductivity, 700 ppm M-alkalinity, pH 11.0, sulfite 25 ppm. Steam is used in jacketed vessels, and some is injected directly into a food process. Over the last quarter makeup has averaged about 30% of feedwater against a design of 22%, and the condensate receiver runs at 190°F where the design is 210°F. Traps and condensate lines show rust-colored deposits, and a condensate pump impeller has a shallow pit. The plant proposes two fixes: add a filming amine to stop the corrosion, and raise boiler water pH to 12 to neutralize the acid. Management wants both approved before the next budget cycle.
1. Where does the acidity in the condensate come from, and why has the softener not prevented it?
Read the worked answer
The acid is carbonic acid, and it is generated inside the boiler, not in the condensate system. Bicarbonate and carbonate alkalinity in the feedwater breaks down under heat and pressure: bicarbonate converts to carbonate plus CO2 plus water, and carbonate converts further to hydroxide plus CO2. That CO2 is volatile and leaves with the steam. In the condensate it redissolves and forms carbonic acid, dropping pH into the 6s. The signature is exactly what is in the log - low condensate pH with elevated iron and copper, because carbonic acid attacks mild steel and, once the pH falls far enough, copper alloys as well. The pitted impeller and the rust-colored deposits at the traps fit.
Softening does not help because a sodium-cycle softener only exchanges calcium and magnesium for sodium. It converts calcium bicarbonate into sodium bicarbonate, which is more soluble and more thermally unstable. The alkalinity passes straight through and becomes CO2 in the steam. In fact, a softener can worsen the problem by removing the hardness that would otherwise have laid down scale while leaving all the bicarbonate in solution.
The excess makeup is a second driver. At 30% makeup against a design of 22%, the plant is importing roughly a third more alkalinity than the design case, and the extra water is itself a symptom worth chasing - the 190°F receiver is 20°F below design, which suggests a trap passing steam or a hot condensate line losing heat or leaking.
2. Evaluate the two proposals - raising boiler water pH to 12 and adding a filming amine - and state what you would do first.
Read the worked answer
Raising the boiler water pH to 12 does not address the mechanism. The CO2 is produced by thermal breakdown of alkalinity, and driving the boiler pH higher does not stop that breakdown; if anything, pushing hydroxide higher on a water that already carries 700 ppm M-alkalinity and 3,500 µmho/cm raises the risk of carryover, foaming, and caustic attack on the boiler metal. The boiler water pH at 11.0 is already inside the normal recommended band. Reject this proposal as written, and note that the conductivity and alkalinity are the numbers to watch, not the pH alone.
A filming amine is a legitimate tool, but it is the second step, not the first. It forms a hydrophobic barrier that protects the metal regardless of CO2 level, but it does not remove CO2, it can plate out in the presence of the high iron already in the system, and it must not be used where steam contacts food unless the product is approved for that service. Since this plant injects steam directly into a food process, a neutralizing amine program with food-contact approval is often the better route, with the boiler water and steam monitored accordingly.
First step: cut the alkalinity at the source and cut the makeup. Fix the condensate losses - survey the traps for passing steam, check the receiver and return lines for leaks, and find out why the return temperature is 20°F low. Then review dealkalization options (reverse osmosis, split-stream softening, or a dealkalizer) against the measured alkalinity load. Only then select an amine program and set condensate targets, typically pH 8.5–9.5 with iron below 0.1 ppm and copper below 0.05 ppm.
3. Write the monitoring plan you would put in place for the next quarter and the trigger points that would send you back to the chemistry.
Read the worked answer
Daily: boiler water conductivity, sulfite, and pH; softener effluent hardness after each regeneration; makeup and condensate flow or percentage so the 30% figure is tracked rather than inferred. Condensate pH and iron daily at the receiver, with copper weekly. Weekly: M-alkalinity and total alkalinity on boiler water and feedwater, plus a total hardness on feedwater, because a softener that passes hardness intermittently shows up in feedwater before it shows up as scale. Monthly: a full analysis including silica, chloride, and phosphate, and a visual inspection of the return lines and a trap survey with an ultrasonic or temperature instrument. Keep the steam purity check going if any steam contacts food.
Trigger points that reopen the chemistry review: condensate iron above 0.1 ppm or copper above 0.05 ppm on two consecutive samples; condensate pH outside 8.5–9.5; makeup above 25% for more than a week; softener effluent hardness above 1 gpg at any time; boiler water conductivity or alkalinity drifting outside the program limits recommended by the water treatment supplier and the boiler manufacturer; or visible new deposits anywhere in the returns.
Also record the decisions, not just the numbers: which amine, at what feed rate, who approved it for food-contact service, and when the trap repairs were completed. A quarter of data plus a repair list is what turns this from a chemical purchase into a solved problem.