12.3 Thermosiphon Oil Coolers and Oil-Temperature Trends

12.3 Thermosiphon Oil Coolers and Oil-Temperature Trends

Industrial ammonia screws dump a large fraction of input power into oil. Oil seals the rotors, lubricates bearings, and leaves the machine near discharge temperature. That heat has to leave the oil before the oil goes back in. CIRO screens show oil in, oil out, and often a refrigerant temperature on the cooler. Those three numbers diagnose the cooler faster than waiting for a bearing-metal alarm.

What a thermosiphon oil cooler is

A thermosiphon (also spelled thermosyphon) oil cooler is a heat exchanger — often a plate pack or shell-and-tube — that cools compressor oil with high-side liquid ammonia. There is no refrigerant pump. Liquid from the condenser or high-pressure receiver flows down a liquid leg into the cooler (the cooler sits below the liquid source). Oil heat boils some of that liquid. The lower-density two-phase mixture rises back to the condenser or receiver vapor space. Gravity and density difference keep the loop moving.

Contrast with liquid injection (refrigerant dumped into the compressor, which raises mass flow and can cost kW/ton) and water or glycol oil coolers (depend on tower or process water). A typical CIRO plant description is thermosiphon oil cooling plus an evaporative condenser: the oil cooler is riding the same high-side liquid the condenser just made. If Chapter 12.1 left you with a dirty or air-bound condenser, this chapter is where that heat shows up as hot oil.

Temperatures you should see

• Oil inlet (from the separator / discharge end) sits near discharge temperature — often in a 160–185°F neighborhood at load, depending on ratio and injection design. It is supposed to be the hot number.
• Oil outlet (back to the compressor) is cooler. A healthy screen example is 136°F. That is a normal oil-out class for many loaded screws on thermosiphon.
• 154°F oil on the same style of screen is abnormal. An 18°F hike is not a thermocouple shrug. Treat 154°F as a high-oil-temperature flag, especially at 100% slide valve.

Refrigerant leaving the oil cooler should be liquid or two-phase, as the manufacturer piped it. A thermosiphon that is fed and circulating does not superheat the ammonia on the way out; there is still liquid present to boil. If the refrigerant outlet is superheated, the loop boiled dry and then the vapor kept taking heat. That is a starved or blocked thermosiphon, not extra-good cooling.

Starved or blocked loop — refrigerant outlet superheated

Walk the loop before you blame the compressor:

• Liquid feed valve throttled or closed
• Strainer or nozzle plugged (construction debris, rust, oil)
• Receiver level too low — no static head to push liquid into the cooler
• Equalizer, vent, or return riser blocked or vapor-bound
• Shared thermosiphon header: when one compressor cycles off, its cooler can fill with liquid and starve the running machines on the same header
• Cooler elevation or reversed riser after a rebuild

Oil may still look tolerable for a few minutes while remaining liquid boils off; then oil-out climbs and the refrigerant outlet goes superheated. That pairing — hot oil plus superheated refrigerant out — is almost never high wet-bulb only. Fix feed and circulation.

If refrigerant out is still two-phase and oil is hot, the limitation is oil-side fouling, too much heat (ratio and load), or a hot condensing reference (the liquid you are boiling against is itself too warm).

High oil temperature at 100% slide with high discharge

Full slide (100%) is maximum swept volume and usually maximum oil-heat load. High discharge pressure and temperature means high heat of compression dumped into oil and discharge gas. The exam pairing is deliberate: this is when a weak cooler shows.

Three supervisor causes, in the order you can actually check:

1. Cooling water or condenser sump high. Thermosiphon liquid is only as cold as the high side that produced it. A hot evaporative-condenser sump, a high wet-bulb, or a water-cooled condenser running hot raises the boiling temperature in the oil cooler. Less ΔT, hotter oil. Non-condensables and dirty condensers therefore heat the oil even though the cooler is ammonia-cooled.
2. Fouled oil cooler. Oil-side varnish and coke, or refrigerant-side film, cut UA. Oil-in stays high; oil-out does not drop the design 30–50°F. Approach on the cooler itself (oil-out minus refrigerant sat T) opens up.
3. Overcompression. High compression ratio — low suction, high discharge, or both — raises discharge temperature and oil temperature even with a clean cooler. Fix the ratio (unloading, another stage, head pressure, suction problems) rather than opening an oil-cooler bypass as the first move.

136°F versus 154°F on a trend at the same slide and similar wet-bulb is a maintenance ticket, not a new normal. If 154°F appears only at 100% on the hottest afternoon, still check those three causes; do not live with it because the slide is loaded. Bearings see the oil that is actually injected, not the nameplate.

Do not copy a memorized good oil temperature from another plant as if RETA published a single legal value. Use the pattern: oil must leave cooler than it entered, refrigerant must not go superheated, and a step change from 136°F to 154°F at high load is a diagnostic. CIRO on-screen references include operating screens; you are being tested on whether 136°F and 154°F mean normal versus abnormal in that loaded, thermosiphon-cooled context.

Reading the oil row on a CIRO-style screen

A useful mental checklist:

1. Slide-valve position and whether discharge is high for today's wet-bulb.
2. Oil-in near discharge? If oil-in is cool, you may have injection or measurement issues, not a hero cooler.
3. Oil-out 136°F class versus 154°F class.
4. Refrigerant out: two-phase or liquid versus superheated.
5. Condenser liquid T versus condensing P (non-condensables heat the whole high side).

Thermosiphon oil cooling belongs in Monitoring System Performance because oil temperature is a leading indicator: it moves before a bearing alarm, and it explains kW/ton when head pressure and slide look acceptable but the machine is pumping heat into a dead cooler. Tie it to approach and TD: a plant that ignores 154°F oil while chasing only room temperature will pay in power and in the next teardown.