A prover detector switch is the sensor that trips as a sphere or piston passes a fixed point inside a prover, marking one end of the calibrated volume. Two detectors, one at each end of that volume, tell the flow computer exactly when to start and stop counting meter pulses. The certified volume between them is the reference the whole prove is built on, so the switches must fire at the same physical position, run after run. This page explains what detector switches do, how their timing defines the counted pulses, and how switch wear or misadjustment corrupts a prove.
Prover Detector Switch in one line: A prover detector switch is a position sensor built into a prover that trips the instant the displacer, a sphere or piston, reaches a fixed point along the calibrated barrel. A pair of them bracket the certified reference volume, and the meter pulses counted between the two switch trips are compared against that volume to compute the meter factor.
A prover works by pushing a known volume of liquid ahead of a moving displacer and comparing that certified volume against the meter pulses accumulated while the displacer travels. The detector switches are the devices that define where that certified volume begins and ends. As the sphere or piston reaches the first detector, the switch trips and the flow computer opens a counting gate; when the displacer reaches the second detector, that switch trips and the gate closes. Everything between the two trips is the counted pulse train for that run.
The switch itself is usually a mechanical plunger that the displacer physically depresses, or an optical or proximity sensor that senses the displacer without contact. On a bidirectional pipe prover the same pair of detectors serves both directions of travel, so each detector acts as a start point one way and a stop point the other way. On a compact or small volume prover the displacer is a piston, and an optical detector senses a flag on the piston shaft as it sweeps the calibrated length. Whatever the mechanism, the detector's job is the same: fire at one exact, repeatable location.
Because the certified volume is fixed by where the two detectors sit, the switches are not just triggers, they are part of the calibrated reference. When the prover is water drawn or otherwise calibrated, the base volume that gets stamped on the certificate is the volume swept between those specific detector positions at those specific trip points. Move a detector, change how deep the plunger has to be pushed before it trips, or let the optical threshold drift, and the effective volume changes even though the certificate still reads the old number.
The accuracy of a prove depends on the switch firing at the same displacer position every single pass. Meter pulses arrive continuously, so a small error in when a detector trips translates directly into extra or missing pulses in the count. On a small volume prover the total pulse count between detectors may be low enough that the flow computer uses double chronometry, interpolating fractional pulses from precise timing, which makes consistent detector timing even more critical because a fraction of a pulse now matters.
Repeatability is the practical test of detector timing. The flow computer runs several passes and checks that the pulse counts, or the resulting meter factors, agree within a tight tolerance before it accepts the prove. A healthy detector produces counts that cluster closely because it trips at the same point each pass. A detector that trips a little early on one pass and a little late on the next scatters the counts, repeatability fails, and the prove will not close no matter how good the meter is. So the switches are often the real reason a prove will not repeat.
Two things can bias a prove without ever failing repeatability, which makes them more dangerous. If both detectors are consistently offset by the same amount, the counts repeat beautifully but the effective volume is wrong, and the meter factor is quietly biased. And if a detector trips consistently but at a position that no longer matches the volume on the certificate, because the switch was replaced or readjusted without recalibrating, the prove looks clean while shifting the factor. Repeatable is not the same as correct, and detector position is where that distinction lives.
When a prove will not repeat, the detector switches are one of the first suspects, and the symptom pattern helps localize the cause. Erratic counts that scatter randomly point to a sticking or bouncing mechanical plunger, a contaminated optical lens, or debris on the detector face. Counts that are stable but slowly trending over weeks point to mechanical wear, a loosening detector mount, or a shifting optical threshold. A prove that fails only in one direction on a bidirectional prover points to one specific detector rather than both. Diagnostic steps run from the outside in: confirm flow was stable and the displacer sealed, review the per pass counts and their spread, inspect and clean the detector faces, verify the detectors are mechanically tight and unmoved, and only then question the meter.
Detector health is not something an operator can watch continuously from the barrel, but its effects show up clearly in data. Every prove records the per pass counts, the spread between passes, the number of passes it took to close, and the resulting factor. Trended over many proves, a detector that is beginning to stick or drift reveals itself as widening repeatability spreads or a slow shift in factor long before it causes an outright failure. That history is exactly what a single on site prove cannot show and what a fleet wide record can.
A cloud SCADA platform such as Merobix reads prove results, repeatability figures, and meter factors from the station flow computers over an industrial protocol and trends them across every prover in the fleet. Measurement staff can see which provers are closing cleanly and which are taking more passes each time, catch a detector degrading before it strands a prove, and dispatch a technician to the right barrel with the right symptom already known. The physical proving still happens on site, but the pattern that fingers a failing detector becomes visible from any browser rather than buried in stacks of individual prove reports.
A prover has at least two detector switches, one to start the pulse count and one to stop it, bracketing the calibrated reference volume. A bidirectional pipe prover uses the same pair for both directions of displacer travel, each acting as start one way and stop the other. Some designs add extra detectors to define more than one calibrated volume or to provide redundancy.
A common cause is a detector switch that no longer trips at the same displacer position on every pass, which scatters the pulse counts and breaks repeatability. A sticking mechanical plunger, a dirty optical lens, debris on the detector face, or a loose detector mount will all do this. Inspect and clean the detectors and confirm they are mechanically tight before suspecting the meter itself.
Yes, and this is the more dangerous case. If both detectors are offset by the same consistent amount, or a detector was replaced or readjusted without recalibrating the prover, the counts can repeat perfectly while the effective volume no longer matches the certificate. The factor is then quietly biased, so any detector work should be followed by recalibration of the prover volume.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.