Most ammonia evaporators in an industrial refrigeration plant are carbon steel, and for a dry engine room that is the right material. An ammonia coil goes to stainless when it sits inside a tank whose contents attack steel, such as a calcium chloride brine chiller, an ice bank or a glycol tank in a food plant, or when it lives in a washdown area where chlorinated sanitizer runs over it every shift. In each case the refrigerant side is the easy side. Ammonia leaves stainless alone, and austenitic 304L stays ductile at -40°F and below without the impact testing that carbon steel often needs under ASME Section VIII at these temperatures.
The rule that overrides everything else on an ammonia coil is copper. Ammonia with a trace of moisture attacks copper, brass and bronze fast, and the corrosion product then circulates through the system. So an ammonia coil from HeatX is stainless tube, stainless headers, stainless or steel valves and fittings, and no brazed joint of any kind. Every tube joint is welded.
Ammonia itself does not attack stainless, so 304L carries the refrigerant side; the coil moves to 316L when the tank outside it holds brine or chloride sanitizer. See Material & Construction for the grade comparison.
Talk to an Application SpecialistAmmonia is the easy side of this coil: the brine or sanitizer outside the tube is what decides whether it is 304L, 316L or duplex.
No copper, no brass, no braze, anywhere on the circuit.
An immersion coil in a tank is nearly always flooded or pumped-liquid fed. The coil fills with liquid ammonia, the liquid boils off the heat coming through the tube wall, and the vapor returns to a surge drum or accumulator. The wetted surface is the whole coil, so the evaporating coefficient is high and the coil can be sized close to the duty. Direct expansion through an expansion valve works on small coils but leaves the last turns dry and superheating, so the end of the coil does nothing and the brine stratifies.
Tell us which feed the plant runs, the suction pressure at the coil and whether the coil is one circuit or several in parallel. Parallel circuits need lengths matched within a few percent, or the short one takes the liquid and the long one starves.
Compressor oil carries over with the ammonia and, being heavier than liquid ammonia and immiscible with it, settles in the lowest turn of the coil. A coil with no oil drain loses surface turn by turn, so we put a drain stub at the true bottom of the circuit. The coil design pressure follows the system; low sides are often designed to 250 psig, and 300 psig covers hot-gas defrost or a plant that pumps down hot. A stainless coil shrinks about 0.2 in. per 20 ft going from a 70°F shop to a -20°F brine, so the headers are guided, not fixed, where they pass the tank wall.
Calcium chloride brine is the hardest service an immersion coil sees. At 20–25% concentration it carries well over 100,000 ppm chloride, two orders past the roughly 1,000 ppm where 316L begins to pit at ambient, though the sub-zero temperature slows the attack a great deal and inhibited brine held at pH 8–9 does run on 316L in many plants. We quote 316L on an inhibited, cold, pH-controlled brine and 2205 duplex, PREN about 35, where the brine is uninhibited, warms during shutdowns or has already pitted a 316L coil. Propylene glycol, ice water and most process liquids are 304L service; the sanitizer outside a washdown tank is the reason to go to 316L in a food plant.
The ammonia coil failures we are asked to replace are almost never the coil. They are a brass sight glass, a bronze-trimmed valve or a copper gauge line that someone fitted to a stainless circuit because it was on the shelf. Within months the brass goes green and porous and the circuit leaks ammonia in to the room. We ship the coil with stainless stubs and a note that says what may be bolted to them. Steel valves are fine on ammonia; any alloy with copper in it is not, and that includes the solder in a gauge fitting.
No. Calcium chloride brine at the 20–25% strengths used for sub-zero chilling carries over 100,000 ppm chloride, and 304L pits at a few hundred. The cold temperature and a well-inhibited, pH-controlled brine let 316L survive in many plants, and 2205 duplex is the usual answer when the brine is uninhibited or ever warms up. We size the coil in whichever grade the brine chemistry supports.
Compressor oil that carries over into the ammonia is heavier than the liquid refrigerant and does not mix with it, so it collects in the lowest turn of the coil. Each turn that fills with oil stops evaporating ammonia and the coil loses capacity without any alarm. A drain stub at the true low point, valved outside the tank, lets the operator pull oil on a schedule and keeps the whole coil wetted.
To the plant's ammonia system rating. Most ammonia low sides are designed to 250 psig and many plants standardize on 300 psig so that a coil isolated full of liquid on a warm day is covered. We pressure test to the code multiple of that pressure and dry the coil before shipment, supply material test reports where specified and stamp the coil to ASME Section VIII where the plant specifies it.
Yes, as long as they are steel or stainless with no copper-bearing trim. Brass bodies, bronze seats, copper gauge lines and soldered fittings all fail on ammonia and are the usual cause of a leak on an otherwise stainless circuit. We supply the coil with stainless stubs, flanged or butt-weld, and will quote stainless isolation valves and the oil drain valve with it if the plant does not stock ammonia-rated steel valves.
Helical and serpentine coils made to your tank drawing in 304L or 316L, with the tube size, pitch and nozzle layout chosen for the duty and the agitator.
+ Learn More
Cooling coils for transformer and rectifier insulating oil, sized on the laminar oil film and arranged so a pinhole leaks oil out, not water in.
+ Learn More
Fully welded coil assemblies in 304L or 316L, with welded headers, support frames and sloped runs that drain, for steam, water and glycol in tanks, ducts and flues.
+ Learn More