17.1 Flooded vs DX Advantages and Charge

17.1 Flooded vs DX Advantages and Charge

CIRO Liquid Overfeed and Flooded Systems is 20 scored items (15% of the sitting). Area 5 already tested DX solenoids, TXVs, and EEVs. Area 6 is why most of the ammonia plants you supervise do not run dry-expansion evaporators as the plant standard: they run flooded vessels or liquid-overfeed coils. You must contrast the two feed methods, name the heat-transfer and charge consequences, and explain why industrial anhydrous ammonia prefers a wet coil.

What flooded actually means

A flooded evaporator keeps the heat-transfer surface wetted with boiling liquid. The coil, plate pack, or shell is not trying to dry the refrigerant out before it leaves. Two-phase mixture — liquid plus vapor — is present over most of the circuit. Heat from the load (air, glycol, process fluid) goes into nucleate boiling and forced-convection boiling, not into superheating a dry vapor film on the last third of the tube.

That wetting is the point. Liquid refrigerant contacting metal has a much higher heat-transfer coefficient (HTC) than superheated vapor. A dry-out patch on the inside of a tube is almost an insulator compared with a boiling film. Flooded and overfeed designs keep those dry patches from forming, so the evaporator can hold a smaller temperature difference (TD) for the same capacity, or more capacity at the same TD.

Leaving the evaporator, flooded and overfeed circuits send a wet mixture toward a separator. Superheat at the coil outlet is near zero. The vapor that finally goes to the compressor is dried in the surge drum / low-pressure receiver, not inside the last few feet of coil. If you clip a thermometer on a wet suction header and subtract saturation temperature, you are not looking at a DX superheat that is supposed to be 8–12°F. You are looking at two-phase fluid that should be near saturation.

Three flooded architectures (same wetting idea)

Operators mix the words. The exam cares that all three keep the surface wet:

• Flooded shell-and-tube (or plate) chiller. Refrigerant boils in the shell (or in a flooded plate pack). A level control keeps tubes submerged. Makeup liquid comes from a high-side float or a level valve. Vapor leaves the top to suction. This is flooded vessel operation.
• Gravity-flooded / thermosiphon coil. A surge drum at the evaporator feeds liquid down; two-phase returns up by density difference. Elevation and pressure drop limit how far the coil can sit from the drum. No recirculation pump on that coil.
• Liquid overfeed (liquid recirculation). A mechanical pump or gas pump forces extra liquid through remote coils. Leftover liquid returns in a wet suction to a plant LPR / recirculator. This is the warehouse penthouse workhorse. Section 17.2 owns the ratio; this section owns why you bother.

Overfeed is not a third refrigerant. It is flooded-coil operation with extra liquid mass so every circuit stays wet even when the evaporated portion is small.

What DX (dry expansion) does instead

Direct expansion (DX) — also called dry expansion — meters liquid so the refrigerant finishes evaporating inside the coil and leaves as superheated vapor. A thermostatic expansion valve (TXV) or electronic expansion valve (EEV) holds that superheat. There is no dedicated wet-return header dumping two-phase fluid into a plant recirculator for that coil. A liquid-line solenoid ahead of the valve is on/off for pump-down, isolation, and defrost (Chapter 14).

DX advantages CIRO still expects you to name:

• Less refrigerant charge in the evaporator. The outlet end is mostly vapor. Inventory is a liquid film plus vapor fill, not a flooded tube.
• Simpler piping on small systems: liquid in, superheated suction out, no recirculation pump, no wet suction main sized for two-phase flow.
• Superheat is a compressor-protection signal. If the TXV/EEV is healthy, liquid should not leave the coil. An accumulator is a backup, not the primary separator of a recirculation plant.

DX penalties on large industrial ammonia:

• The outlet end of the circuit is vapor. HTC collapses where quality approaches 1.0. You pay with higher TD, more coil surface, or both.
• Distribution through a distributor and many circuits is harder with ammonia's high latent heat and small mass flow per ton. Some circuits starve while others flood.
• A failed-open TXV or lost superheat slugs the compressor unless an accumulator is in the way. Flooded plants put a vessel and high-level cutout in that path on purpose.
• Oil return on DX depends on leaving-vapor velocity. Ammonia and mineral oil do not mix well. Large plants usually manage oil at the separator and oil pots instead (Chapter 18).

Industrial ammonia is mostly flooded and liquid-overfeed, not DX. DX still appears on small process evaporators, some packaged units, and a few penthouse air units. Do not tell the exam that ammonia DX does not exist. Tell it that large industrial NH3 prefers wet coils.

Charge: why flooded holds more pounds

Charge is the mass of refrigerant sitting in the circuit. A flooded tube, shell-and-tube chiller, or overfeed coil kept full of boiling liquid holds far more ammonia than a DX coil of the same physical volume, because liquid density is orders of magnitude higher than vapor density.

That extra inventory is not wasted decoration. It is the buffer that keeps the surface wet when load spikes, when a circuit is slightly starved, or when a pump is ramping. It is also a PSM/RMP inventory issue. More pounds in evaporators and surge drums can matter at the 10,000 lb anhydrous ammonia threshold (OSHA 29 CFR 1910.119 Appendix A; EPA 40 CFR 68). Flooded/overfeed plants often have large low-pressure receivers. DX packaged systems often do not. The operator's job is not to minimize charge at all costs; it is to know where the pounds live and why the design put them there.

Worked contrast (order of magnitude, not a RETA constant): a small DX freezer coil might hold a few dozen pounds of ammonia in the tubes. The same room on liquid overfeed, with a wet return and a recirculator serving several penthouses, can hold hundreds to thousands of pounds in the LPR plus the coils and wet suction. CIRO will not ask you to invent a catalog charge. It will ask whether flooded/overfeed increases evaporator charge relative to DX. It does.

Heat transfer, TD, and leaving state

Evaporator TD is the difference between entering air (or fluid) temperature and refrigerant saturation temperature. Because flooded/overfeed keeps the inside wet, you can often run closer to the space, or the same refrigerant temperature with more capacity. On a CIRO screen, a flooded coil showing a high TD with a coil that is not fully frosted to the header is not proving that DX superheat is healthy — it is often underfeed, oil film, frost, or lost recirculation. A DX coil should show superheat at the bulb. A flooded coil should not be judged by DX superheat rules.

Control philosophy, not just hardware

Flooded vessels typically hold level. A high-side float, low-side float, or modulating level valve keeps the boiling surface covered. Liquid-overfeed coils typically do not try to hold a level in the coil. A pump (or gas-powered recirculator) overfeeds liquid; the coil stays wet; leftover liquid returns in the wet suction to the separator.

DX control is superheat. Overfeed control is wetted surface plus separator level. If you apply a DX diagnostic (chase 10°F superheat at a flooded coil outlet) you will starve a wet system or miss a real dry-out. If you apply a flooded diagnostic (expect wet returns) to a DX coil, you will accept liquid in the suction that the compressor cannot swallow.

A supervisor who throttles overfeed balancing valves until a suction header shows DX-like superheat has just converted a 4:1 coil into a starved DX coil without a TXV. Capacity falls. TD rises. That is a commissioning error, not a savings.

Why ammonia, specifically

Anhydrous ammonia has a large latent heat and a relatively small mass flow per ton. DX distributors designed around halocarbons do not automatically behave well with that small mass flow. Recirculating extra liquid multiplies the mass in the tubes so every circuit stays wetted even when the evaporated portion is small.

Ammonia's toxicity also argues for robust compressor protection: a surge drum / LPR with a high-level trip is a more forgiving last barrier than hoping a TXV never fails open on a 300 HP screw. Oil is another ammonia-specific reason. Mineral oil used with ammonia is immiscible and heavier. In a flooded vessel it tends to log in the bottom or in a drain pot, not ride out with the vapor the way it can in some miscible HFC DX circuits. Large plants concentrate oil management at the recirculator, oil pots, and rectifiers. Flooded/overfeed is the architecture that makes that oil strategy possible.

Exam traps

• Flooded does not mean the whole plant is full. It means the evaporator surface is flooded with boiling liquid. High-side receivers and condensers have their own levels.
• Overfeed is a type of flooded-coil operation, not a third mystery refrigerant. Liquid is pumped at a multiple of the evaporated mass so the coil stays flooded.
• Low leaving superheat on a flooded coil is normal. Low leaving superheat on a DX coil is a flood-back warning.
• More charge is a flooded/overfeed trait, not a defect, unless inventory was never accounted for in process safety information and RMP.
• Do not claim a numeric HTC from memory if the stem does not give one. Claim the direction: boiling liquid is far better than dry vapor.