Maintenance teams don't usually go looking for a compressed air leak survey. They notice a symptom first (the compressor sounds like it's working harder, a machine at the end of the line is underperforming, the power bill crept up) and then have to work out what's actually driving it. Below are the four symptoms that most commonly trace back to leak load, and what each one is actually telling you.
The compressor rarely unloads
A fixed-speed rotary screw compressor is designed to load and unload: running fully loaded when demand is high, then unloading (running unloaded, or shutting down entirely on some control schemes) once the receiver reaches its set pressure and demand drops. If your compressor spends most of its time loaded, even during periods when production isn't drawing much air (overnight, on a lunch break, over a weekend), that's the clearest single signal of a base leak load. A system with no leaks should be able to build pressure and then coast on receiver capacity while nothing downstream is consuming air. If it can't, something downstream is consuming air continuously, and leaks are the most common reason.
This is worth checking specifically outside production hours. A compressor that loads heavily during a shift could just be meeting real demand. A compressor that's still cycling load at 2am with the plant empty is very likely feeding leaks, not production.
Pressure sags at the far end of the line
If pressure at the point of use (a tool, a cylinder, a process valve) is noticeably lower than what's showing at the compressor's outlet gauge, and it's not a new problem tied to a recent change in demand, distribution losses are the usual suspect. Some pressure drop across a distribution network is normal and expected. A drop that's grown over time, or that's worse at the far end of a long run than it used to be, usually means the network is carrying more total flow than it was designed for, because leaks along the way are adding continuous demand that competes with the legitimate demand at the end of the line.
This symptom is often first noticed as a quality or performance problem (a pneumatic tool running slower, a cylinder not fully actuating) before anyone connects it to the compressed air system itself. If a point-of-use problem traces back to "not enough pressure" rather than a fault with the equipment itself, and turning up the compressor's set pressure is being used as the fix, that's a sign the real issue is upstream loss, not undersupply.
| Symptom | What it usually indicates |
|---|---|
| Compressor loaded outside production hours | Base leak load with nothing to explain it |
| Pressure sags at the far end of a run | Distribution losses competing with genuine demand |
| A second (standby) compressor cutting in more often | Total demand has grown past what one machine should be carrying |
| Power bill drifting up with no production change | Continuous energy draw with no corresponding output |
A second compressor is cutting in more than it used to
Plants with a lead-lag compressor setup (one machine handling normal load, a second cutting in during peaks) often treat an increase in lag-compressor run time as a capacity problem: production has grown, so of course the second machine is needed more. Sometimes that's true. But if production volumes haven't meaningfully changed and the lag compressor is running more than it used to, the more likely explanation is that total system demand has crept up, and since nothing about the production process itself is asking for more air, that creep is coming from leak load accumulating on the network over time.
This one is easy to miss because the symptom looks like success: the standby system is doing its job, keeping pressure up. It's also one of the more expensive versions of this problem, since running two compressors instead of one, even at partial load each, is rarely as efficient as one machine running well within capacity.
A second compressor cutting in more often isn't proof you've outgrown your system. It's often proof your system has been quietly leaking for longer than anyone's checked.
The power bill has drifted, and nothing else explains it
Compressed air typically accounts for a meaningful share of a manufacturing site's total electricity cost, and it's one of the few loads that runs continuously rather than tracking production directly. If a site's electricity spend has drifted upward over a period where production volume, shift patterns, and other major equipment have stayed roughly constant, compressed air is one of the few places that continuous, unexplained draw can hide: a leak, unlike almost every other electrical load in a plant, keeps consuming energy at exactly the same rate whether the line is running or not.
What these symptoms don't tell you
None of the four symptoms above can tell you where a leak actually is, or how many there are. That's their real limitation. A compressor that won't unload confirms something downstream is consuming air continuously: it doesn't say whether that's one large leak on a header or forty small ones scattered across tool stations, and it doesn't tell a maintenance planner which fitting to send someone to first. The symptom is a reliable signal that a problem exists; it's not a diagnosis, and treating it as one (assuming, for example, that a single obvious leak explains the whole pattern) often means the underlying load barely changes after that one repair.
This is also why symptom-based troubleshooting tends to plateau. A team that fixes the two or three leaks they can hear or see by walking the floor will see some improvement, then find the compressor is still loading more than expected: the leaks that are audible or visible at floor level are rarely the majority. The rest sit overhead, behind guarding, or simply below the threshold of what a person can detect without instrumentation.
What these symptoms have in common
None of the four above prove a leak on their own: a genuine increase in production demand can produce some of the same signs. What they share is that they're all consistent with the same underlying mechanism: continuous, unaccounted-for air consumption that doesn't correspond to anything productive happening on the plant floor. A compressed air network doesn't develop these symptoms overnight; they build gradually as individual leaks accumulate, which is also why they're easy to write off as "just how the system runs" rather than investigated properly.
If one or more of these symptoms sounds familiar, the next useful step isn't guessing at a cause: it's finding out where the actual leaks are. See how an acoustic survey locates and sizes them without shutting the line down, or run a rough cost estimate for what a base leak load in this range is likely costing using the ROI calculator.