When a vapor recovery unit misses its emission limits, the cause is usually not one broken part but one diluted, fouled or drifting element in an otherwise healthy system, and the fastest path back to compliance is a structured check rather than a parts cannon. In the field, the same handful of suspects accounts for most misses: air getting into the stream and diluting it, a process stage that has fouled or saturated, liquid where vapor should be, and controls that have drifted so gradually nobody noticed. This article walks through those causes in the order they should be checked, what each one looks like in the data, and how to fix them without shutting the site down longer than the problem deserves.
Start with the stream, not the equipment
The most common reason a healthy unit reads badly is that it is being fed an unhealthy stream, so the first check is upstream of the skid. Air entering the vapor path through a loose connector, a cracked hose, an open sump or an unsealed riser dilutes the hydrocarbon concentration and pushes the process off its design point; the unit then works hard on a stream that is mostly air, and the outlet reading suffers. Inventory anomalies, faint odor near a specific run, or readings that sag after rain all point toward ingress. The piping and its joints deserve the same scrutiny as the equipment, because a vapor path is only as tight as its least convincing connector, and the tightness testing discipline used on fuel lines applies to the vapor side in exactly the same way.
The pattern of symptoms and what each one means
Misses rarely arrive without a pattern, and the pattern is the diagnosis. Before replacing anything, put the recent readings beside the duty they accompanied and find the row that matches:
| What the readings show | Most likely cause | The confirming check |
|---|---|---|
| Outlet rises steadily over weeks | Stage fouling or bed saturation building up | Compare inlet and outlet across one week of similar duty |
| Misses only during deliveries | Liquid carryover reaching the process stage | Inspect the knockout vessel drain and the line low points |
| Misses after rain or at dawn | Air ingress through a wet or cooled joint | Test connectors and watch for bubble trails |
| Outlet fine, recovery volume low | Controls drift on pressure or flow setpoints | Verify sensor readings against a reference gauge |
| Misses on one fuel only | Stream composition outside the design case | Log which product was moving during each miss |
The usual hardware suspects
Once the stream is cleared, the process stages come into focus, and each has a characteristic way of aging. Membrane stages foul slowly when liquid or particulate reaches the modules, and the symptom is a gradual loss of separation margin rather than a cliff, which is why trend lines matter more than single readings. Activated carbon beds saturate on schedule by nature, and a bed pushed past its economic endpoint lets hydrocarbons through in exactly the pattern a slowly rising outlet shows. Condensation stages lose ground to fouling on their coils and to refrigeration performance that has quietly slipped. Knockout vessels that are not being drained faithfully pass liquid into stages that were never meant to see it. None of these failures is exotic; each is a maintenance item wearing the costume of an emissions problem.
Check the controls before blaming the process
A disheartening share of misses are instrument stories, and they deserve their own pass. Sensors drift, setpoints get adjusted during some long ago upset and never restored, and control loops that cycle or hunt can push the process outside its comfortable band during exactly the hours the tests sample. The checks are cheap: verify readings against a reference instrument, walk the alarm history for oscillation patterns, and confirm the setpoints against the original commissioning sheet rather than against memory. If the outlet trace breathes in rhythm with a pump or a valve, the process may be perfectly healthy and merely being steered badly. Fix the instrumentation first when it is the culprit, because no process adjustment can compensate for a controller that is lying about the state of the system.
An order of checks that finds the cause
Site teams who clear misses quickly tend to run the same sequence, which is worth adopting as a card next to the panel:
- Pull the trend, not a snapshot, and mark exactly when the misses happen relative to deliveries, weather and fuel slate;
- Clear the stream first: walk the vapor path for ingress points and confirm the knockout vessel is drained and honest;
- Check liquid management, because carryover masquerades as stage failure more often than any other single cause;
- Inspect the process stages in duty order, looking for fouling, saturation or frost, guided by the trend shape rather than by habit;
- Verify instruments and setpoints against references, and only then reach for parts replacement, with the trend as the acceptance test afterwards.
The order matters because each step either clears a suspect or points at the next one, and the parts budget survives the investigation intact.
Turning a miss into a maintenance rhythm
The best outcome of a missed limit is not just a fix but a rhythm, because almost every cause on this list announces itself in the data before it reaches the limit. Fouling shows as a slow climb, saturation as a bed nearing its replacement hours, ingress as readings that answer the weather. Sites that log outlet readings against duty and review the trend weekly catch these stories early, and their compliance checks become confirmations rather than anxieties. The same habit applies to odor complaints, which usually share a root cause with the misses. A unit that is watched in trend lines rarely surprises its operator, and that calm is worth more than any single piece of hardware on the skid, which is why the vapor recovery unit ships with monitoring points designed to be read, not just wired.
Frequently asked questions
Can a unit be temporarily out of limits without being broken?
Yes, and this is why the trend matters. A rich delivery slug, a hot afternoon or a partial load can push a healthy unit toward its limit for an hour. What separates a healthy system from a failing one is the shape between events: healthy units return to their normal band when duty normalizes, while a real fault holds the reading high or climbs steadily regardless of duty.
How long does a typical investigation take?
With trend data available, most causes are found within a day of structured checking, because the sequence above clears the biggest suspects first. Investigations stretch out when only snapshots exist or when parts are replaced before the stream and liquid management have been cleared, which is exactly the order this article argues against.
Should the unit be upsized if misses keep happening?
Only after the investigation closes clean. Undersizing is a real cause, and there are sites where the duty has genuinely outgrown the unit, but on most sites the misses trace to maintenance and stream issues that a larger unit would inherit untouched. Size against the measured load once the system is honest, not against the frustration of a recurring miss.
Steady emission performance starts with a steady stream, and the double layer composite pipe systems from the sister line keep the vapor path tight from the tank to the unit inlet.
Luoyang Wohong Petrochemical Equipment Co., Ltd. builds vapor recovery and membrane separation packages for fuel terminals, depots and gas stations. Contact us for system sizing, configuration advice and spare parts support.