The short version
- A last service report lists the driver and compressor end make and model, run hours on each major part, what was replaced, the gas and load, and recorded temperatures and pressures.
- Compare run hours against typical failure intervals for that make, model, and part to flag what is due.
- What was replaced last time points to what usually fails next, since failures travel in patterns.
- A rising discharge temperature or a falling suction pressure at the same load is an early warning worth ranking high.
- EverSense automates this read against 30 years and about 25,000 real field repair reports, so nothing slips.
Your last service report is a prediction, not a receipt
The last service report from a compressor unit is usually filed and forgotten. That is a missed chance. Every report holds enough detail to point you at the next likely failure on that unit, if you know how to read it. This guide walks through what the report contains and the logic to turn it into a ranked watch list for both the driver and the compressor end.
Remember that a gas compression unit is two machines bolted together. The driver, often a Caterpillar or Waukesha engine, and the compressor end, often an Ariel frame or an integral unit like an Ajax. Both wear. Both fail. Drivers are a major source of downtime, so a good read covers the whole unit, not just the cylinders.
What a last service report actually contains
Before you can read a report, know its parts. A complete last service report for a gas compression unit usually lists:
- Make and model of the driver and the compressor end, plus serial numbers. This is the key that unlocks the right failure intervals.
- Run hours on the unit and on each major part: valves, rod packing, piston rings, rider bands, bearings, and the engine top end.
- What was replaced or rebuilt last time, and the reason. Look for notes like suction valve replaced, found broken plate.
- The gas and the load: gas composition (sweet or sour, wet or dry), suction and discharge pressure, and the load step the unit runs at.
- Recorded temperatures and pressures: suction and discharge temperature per cylinder, discharge pressure per stage, jacket water and oil temperature, and any packing or bearing temperatures.
- Oil analysis results if a sample was pulled, showing wear metals like iron, copper, lead, and tin.
If any of these are missing, that gap is itself a finding. A report with no gas analysis or no per cylinder temperatures leaves you guessing.
Match run hours to typical failure intervals
Start with run hours against the known life of each part for that make and model. Wear is not perfectly linear, so treat these as windows, not exact dates. Rough guides from the field:
- Compressor valves: often 12 to 18 months on clean dry gas, but as short as 3 to 6 months on wet or sour gas at high load. In hours, a valve might run tens of thousands of hours on clean air at half load, or fail near 6,000 hours on wet sour gas at 90 percent load.
- Rod packing and piston rings: planned life is often near 8,000 hours, with an early look around 4,000 hours to check the actual wear rate on that unit.
- Engine top end: often near 20,000 hours. An in frame overhaul runs near 40,000 hours and a major overhaul near 80,000 hours, longer on clean fuel and sound maintenance.
Now compare. If the report shows valves at 5,000 hours on sour gas, they are near the end of their window. If the engine sits at 19,000 hours on the top end, that is due soon. The gas and load on the report tell you which end of each range to use. Dirty or sour gas and high load push every number shorter.
Read what was replaced last time
What a crew replaced last time is a strong clue to what comes next. Failures travel in patterns.
- If a discharge valve was replaced for a broken plate, look hard at the matching suction valve and the same valve on the other cylinder. High discharge temperature that cooked one valve tends to cook its neighbors.
- If rider bands were changed, the piston rings on that cylinder are usually not far behind, and a scored rod or cylinder may follow if the wear ran long.
- If packing was replaced for high vent flow, check whether the rod was measured. A worn or scored rod chews new packing fast.
- On the engine, if one cylinder head was rebuilt, valve stem projection and oil consumption on the rest of the bank are the trend to watch.
The reason noted next to a repair matters as much as the repair itself. Replaced on schedule is routine. Replaced, found scored means the root cause may still be there.
Watch the recorded temperatures and pressures
The numbers written on the report are early warnings if you read them against load. The rule: a change with no change in operating conditions is the signal.
- Rising discharge temperature on a cylinder, at the same load and suction, is a classic early sign of a leaking valve or worn piston rings. Hot gas is leaking back and being recompressed.
- A suction temperature that climbs toward discharge temperature on one cylinder points to a suction valve leak.
- Falling discharge pressure or dropping capacity at the same speed suggests blow by past worn rings or rider bands.
- High or rising packing and bearing temperatures flag friction and coming failure. A wet, oily, or high flow distance piece vent is packing in distress.
If a PV card (a pressure volume card) was taken, it is the clearest read of all. A discharge valve leak drops the re expansion line. A piston ring leak shows the measured curve crossing the ideal curve in both chambers.
Do not skip the oil analysis
If the report includes an oil sample, the wear metals name the part that is wearing:
- Iron points to cylinders, liners, gears, and the crankshaft.
- Copper points to bearings, bushings, and cooler tubes.
- Lead and tin point to babbitt bearing overlay wearing through.
One high number is a flag. A rising trend across two or three samples is a diagnosis. Compare this report to the last one. The direction matters more than the single value.
Turn the paper report into a ranked watch list
Now put it together. For each part on the unit, write one line with three things: how close it is to its interval, whether the last repair or the recorded numbers point at it, and what a failure would cost in downtime and collateral damage. Then rank.
- 01Rank first anything with two signals lining up. Example: valves past their hours window on sour gas and a discharge temperature creeping up. That is your next failure.
- 02Rank next single strong signals, like an engine top end due on hours or a rising wear metal trend.
- 03Rank last the routine items that are simply approaching their interval with no other sign.
The output is a short, ordered list of what to inspect or stage parts for at the next shutdown. That is a real predictive read from one piece of paper.
How EverSense does this for you
Reading one report by hand is doable. Reading every report across a fleet, every month, against the failure history of dozens of equipment makes is not. That is the work EverSense automates. It takes your last service data, matches the make and model of the driver and the compressor end to 30 years and about 25,000 real field repair reports across 38 equipment makes, and checks every diagnosis against the unit's own OEM manuals. It returns the same ranked watch list this guide describes, with the likely failure, the evidence behind it, and a draft investigation for your crew.
EverSense works on day one from your last service data, with no sensors required. It only advises. Your team still makes the call. If you want to see your own reports turned into a ranked watch list, book a demo.
Common questions
Can I really predict a failure from one service report?
One report gives you a snapshot. You get a real prediction when you read run hours against the known interval for that make and model, then check what was replaced last time and the recorded temperatures. Two or more of those pointing at the same part is a strong call. A trend across several reports is stronger still.
What if my report is missing the gas analysis or per cylinder temperatures?
That gap is a finding on its own. Without the gas you cannot pick the right failure interval, since sour or wet gas shortens valve and packing life sharply. Ask the crew to record gas composition and per cylinder suction and discharge temperatures on the next service so the next read is sound.
Do I need sensors or a PLC feed to do this?
No. The whole method runs on the paper or digital service report you already have. Inline oil sensors and a live PLC feed can sharpen the trend later, but they are an add on, not a requirement. EverSense works day one from your last service data.
Does this work for the engine, or just the compressor end?
Both. The driver, usually a Caterpillar or Waukesha engine, is a major source of failures. Read its top end hours, oil consumption, and valve stem projection the same way you read the compressor valves and packing. A full read covers the whole unit.
How is this different from a fixed maintenance schedule?
A fixed schedule changes parts on the calendar whether they need it or not. Reading the report is condition based. It uses the actual hours, gas, load, and recorded numbers on that unit to move the likely failures up or down the list, so you spend effort where the risk really is.