Condensing corrosive vapor streams

Traces in the gas become percent-level concentrations in the condensate, because the first water to condense strips chloride out of thousands of times its own volume of gas. A vent at 50 ppm chloride with 5 percent water can produce a first-film condensate above 1,000 ppm, and at a 150 F metal temperature that is enough to pit 304 or 316 within weeks. Shutdowns make it worse: the film evaporates, salts crystallize under deposits and at tube joints, and the next start-up redissolves them into a far stronger solution. Duplex resists pitting better; titanium is the usual answer where chlorides are the only problem.

Titanium is not a universal answer; it excels against chlorides, wet chlorine and oxidizing acids, and fails against fluorides, dry chlorine and hot reducing acids such as concentrated hydrochloric or sulfuric. A hot sulfuric condensate that would attack titanium is handled by a nickel-molybdenum alloy, and even trace fluoride, a few ppm from an etching step, can take out titanium tubes. Dry chlorine ignites titanium at elevated temperature. Have the condensate analyzed, including trace species, and rank the candidate alloys against that analysis at the warmest metal temperature in the dew-point zone. The ranking changes from one stream to the next.

Acid dew point is the temperature at which an acid solution begins to condense from a gas, and it lies above the plain water dew point whenever SO3 or HCl travels with water. Flue gas with 10 ppm SO3 and 10 percent water, for example, condenses sulfuric acid near 270 F while water alone would not condense until about 115 F. Any metal running between those two temperatures collects strong, hot acid at its most corrosive. HCl behaves differently from SO3: its dew point sits only a few degrees above the water dew point, so the window is narrow and the whole wetted length usually gets the alloy.

A Welded Plate exchanger handles corrosive vapor well and is often the least expensive route to titanium or nickel alloy surface, because plates 1 to 2 mm thick use a fraction of the metal a tube bundle does. It needs clean vapor: channels a few millimeters wide trap solids, and most welded packs cannot be opened for mechanical cleaning. Gasketed Plate and Frame units are a poor fit, since the gaskets themselves swell or embrittle in acids and chlorinated solvents. If the vapor carries catalyst fines, polymer or salt, choose tubes in alloy instead and accept the higher metal cost.

The corrosive stream belongs in the tubes in nearly every case, because then only the tubes and tubesheets need high alloy and a failed tube can be plugged without taking the unit out of service. Shell-side condensing of a corrosive vapor means an alloy shell, baffles, tie rods and spacers, which can double the metal cost and adds crevices at every baffle hole. Tube-side condensing also keeps the condensate moving in a single direction so no acid pools. Ask for the tube ends to be lightly expanded after welding as well, so no gap is left behind the weld where condensate can wick in and sit.

Drain the vapor side completely, flush it with clean water or a mild neutralizing rinse, and dry it with warm air or nitrogen before the unit stands idle. A film of condensate left on the tubes concentrates as it evaporates, and much of the pitting in this service happens during those idle hours rather than in operation. Give the design the connections that make this routine practical: a flush inlet at the top, a drain at the lowest point, and a dry-gas purge connection. Log each rinse and dry-out; an unexplained failure often traces to a weekend when the routine was skipped.