19.4 Closed Intercoolers and Two-Stage Troubleshooting
Key Takeaways
- A closed intercooler cools booster discharge in a coil (or tubes) inside a bath of boiling intermediate-pressure ammonia, without mixing the streams.
- Closed designs are chosen when oil, circuit separation, or a drier high-stage suction is more important than the slightly better approach of an open flash.
- High low-stage discharge temperature is a ratio, superheat, oil-cooling, or intercooler (starved/flooded) problem until proven otherwise.
- A starved intercooler leaves booster gas hot into the high-stage; a flooded intercooler sends liquid to high-stage suction.
- Non-condensables on the high stage raise high-stage discharge pressure and compression ratio and can shove interstage off its target.
19.4 Closed Intercoolers and Two-Stage Troubleshooting
Quick Answer: A closed intercooler is a coil in a bath. Intermediate-pressure ammonia boils on the shell side. Booster discharge (and often high-side liquid to be subcooled) flows inside the tubes and does not mix with the bath. Use it when oil or circuit isolation matters. When low-stage discharge temperature is high, work the list: high ratio, high suction superheat, failed oil cooling, intercooler starved or flooded, high-stage non-condensables — in that physical order, not in the order the loudest alarm appears.
Open flash is the usual textbook picture. Closed intercoolers show up on exams and in plants that needed a heat exchanger rather than a mixing drum. If you only memorized “bubble the discharge,” a closed-intercooler item will feel like a different system. It is still two-stage ammonia. The pressure levels are the same. The heat-transfer path is different.
Closed intercooler hardware
Bath / shell: A vessel is maintained at interstage pressure by the high-stage suction (and by how much liquid is fed to the shell). Liquid ammonia in the shell boils at T_sat(P_int). That boiling liquid is the cold sink. Makeup liquid typically comes from the high side through a level-control valve, so the shell is still a flash inventory — the flash vapor from that feed still goes to the high-stage. What is “closed” is the booster-discharge path, not the idea of flashing high-side liquid.
Coil / tubes: Booster discharge flows through a coil immersed in that boiling bath. Heat crosses the tube wall. The gas is desuperheated toward bath temperature but will leave with some approach — often several degrees to a couple of tens of degrees above the bath, depending on load, fouling, and circuiting. That is the thermodynamic penalty versus bubbling through: you cannot beat direct contact on approach.
A second coil is common: high-side liquid through a subcooler circuit in the same bath, so freezer feed is cooled toward interstage temperature without dumping that liquid into the same vessel the booster gas mixed with. You can still get the subcooling benefit in a closed design; you just bought it with UA instead of with open flash of the entire liquid stream into the booster’s mixing space.
Why specify closed instead of open
- Oil and separation: Booster discharge carries oil mist. In an open drum that oil washes into the intermediate bath and high-stage suction. A closed coil keeps the booster’s oil in the booster discharge / oil-separator world unless a separate oil return is provided. Plants that run different oils, have messy booster oil carryover, or want to keep high-stage suction cleaner lean closed.
- Circuit isolation: Multiple boosters, dirty service, or a need to isolate a booster without dumping its discharge into the shared liquid bath.
- Drier high-stage suction: You still must control shell level so boiling liquid does not carry over, but the booster gas itself is not sparging through the liquid, so the vapor path can be designed as a conventional exchanger outlet rather than a bubbling drum.
What you give up: slightly warmer high-stage suction for the same bath temperature, more surfaces to foul, more pressure drop on the booster discharge, and another exchanger to leak (tube leak mixes high-pressure gas into the bath or vice versa — a real MI item). Closed is a choice, not a universal upgrade.
Two-stage troubleshooting: high low-stage discharge temperature
High booster (low-stage) discharge temperature is a CIRO favorite because it has several honest causes that look similar on a noisy panel. Use pressures in psia, then temperatures, then oil, then the intercooler, then the high side.
1. High compression ratio on the booster
CR_low = P_int / P_suc in psia. If this number climbs from ~4 toward 7–10, discharge temperature will follow even with a perfect intercooler downstream (the intercooler cannot cool gas that has not been compressed yet).
Suction too low: freezer iced, starved HEV / overfeed, fan failure, product pull-down, or a suction valve not fully open. Vacuum deeper than 8.7 in Hg at a coil that should be −40°F means the evaporator is not at design, or the load vanished and the booster is still pulling. Deeper vacuum → lower P_suc → higher CR → higher Td.
Interstage too high: high-stage not keeping up (slide valve, capacity, trip), intercooler pressure high because the high-stage CR rose, or the plant is holding interstage up for a medium-temp load while the freezer suction collapsed. High P_int with low P_suc is a double squeeze on the booster.
Compute CR before you blame oil. A 250°F discharge at CR 4.5 is a different story from 250°F at CR 8.
2. High suction superheat
Isentropic discharge temperature starts from actual suction temperature, not from T_sat. A booster sucking 0°F gas at 10.4 psia (−40°F sat) has 40°F of superheat and will discharge much hotter than a near-saturated recirculated plant. Causes: DX coil starved or hunting, overfeed ratio too low, liquid line restriction, or a heat source on the suction main. Recirculated plants should run low suction superheat. If they do not, the low-temp liquid supply (including a starved intercooler feed) is a suspect.
3. Bad oil cooling
Screws dump a large fraction of the compression heat into oil. Thermosiphon oil coolers, liquid injection, or water-cooled oil coolers that are fouled, low on thermosiphon level, valved off, or air-bound will show high oil temperature and high discharge temperature even at a normal CR. Check oil temperature and the oil-cooler ΔT, not only the discharge thermocouple. Two-stage does not retire oil cooling; freezer boosters still need it.
4. Intercooler starved versus flooded
These are opposite failures. Do not treat “intercooler problem” as one symptom.
| Intercooler state | What you see | Booster Td | High-stage suction |
|---|---|---|---|
| Starved (low level, feed valve failed shut, empty bath) | Little or no desuperheat | Stays high leaving the IC if you measure after the coil/drum | Hot, high superheat |
| Flooded (level high, float failed open) | Liquid toward high-stage | Booster Td may look normal or cool after the IC | Low or zero superheat, liquid risk |
| Fouled closed coil | Approach blows out | Hot gas leaving the coil | Warmer than bath, like a mild starve |
| Open drum bypass / dip pipe uncovered | Gas not actually bubbling | Same as starved | Hot |
Starved closed coil: tube-side gas never sees enough boiling liquid on the shell. Flooded shell: high-stage gets wet. Know which machine is screaming before you open valves.
5. Non-condensables on the high stage
Air and other non-condensables collect in the condenser / high-side. They raise true condensing pressure above the saturation pressure that matches the condenser temperature (bad approach: pressure says 95°F+ while the coil is cooler). High-stage P_dis rises, high-stage CR rises, high-stage Td rises, and the high-stage may unload or lose capacity so interstage rises. The booster then sees a higher P_int and its CR and Td climb even though the freezer did nothing wrong.
Purge indication is a condenser story: condenser pressure higher than P_sat at the measured condensing temperature. Do not start by opening the intercooler drain. If the high-side pressure-temperature pair is lying, fix non-condensables and condenser fans/spray first.
A practical diagnostic order
- Convert booster suction, interstage, and high-side discharge to psia. Compute both stage ratios.
- Compare suction T to P/T: superheat.
- Compare discharge T to what that CR deserves. If CR is ugly, fix suction or interstage before chasing oil.
- Oil temperature and oil-cooler operation.
- Intercooler level and leaving-gas temperature versus bath / P_int saturation.
- High-side approach and NC (condenser P vs T).
- Only then: valves, check valves between stages, capacity slides, and “the booster is worn.”
Other two-stage complaints that are not Td but ride along: high-stage slugging (flooded IC), low freezer capacity with a hot booster (starved IC and starved low-temp feed), oil in the intercooler (open design, drain it as a managed task), and interstage that will not come down because the high-stage is off while someone still runs the booster — a ratio disaster. Never run a booster against a shut high-stage and call the discharge temperature a mystery.
Closed versus open does not change that tree. It only changes where you look for desuperheat: coil leaving temperature and shell level on closed; dip-pipe submergence and drum leaving temperature on open. The compression-ratio arithmetic is identical.
What distinguishes a closed intercooler from an open flash intercooler?
A two-stage ammonia booster shows high discharge temperature. Interstage pressure converted to psia gives a normal booster compression ratio, suction superheat is low, and oil temperature is normal. High-stage suction, however, is much hotter than intercooler saturation temperature. What is the most likely intercooler problem?
Non-condensables collecting in the condenser of a two-stage ammonia plant most directly raise which of the following, and how can that heat the booster?