Test-bed evidence versus field evidence

1. What a test bed measures

A valve test bed is a controlled and repeatable rig. It clamps the valve, applies a known pressure, and measures seat leakage with a chosen medium and method. Factory acceptance testing to API 598, with leakage classes described by ISO 5208, is performed on rigs like this. The valve is clean, new or freshly refurbished, and cycled under supervision.

This evidence answers one question well: does the valve meet its specification at the moment of test. It does not answer another: will it still meet that specification after years in live service. None of this is a weakness of the test bed. It is simply a different question. Three features give test-bed results their value:

1. The medium, pressure and temperature are chosen and held steady.
2. The condition of the valve at the time of test is known.
3. The same test can be repeated, so results can be compared.

2. Why the field behaves differently

An installed valve meets the real service, not the rig. The line carries dirt, wax, water and product residues. Temperatures swing through seasons and operating cycles. Some valves sit without operation for long periods, and friction builds in the stem and the seat. Installation stresses add side loads that no test bed applies.

Lubricant films age, wash away or dry out, so operating torque creeps up. Seats pick up damage from debris passing through the line. The result is drift: seats that no longer seal cleanly, stems that bind, and packing that weeps. A test-bed pass is therefore a baseline, not a forecast. It tells you where the valve started, not where it is now. Field evidence exists to close that gap, and it must be read with the same care.

3. How to read valve test bed results

When a supplier offers valve test bed results, read them as a snapshot. Reasonable checks include:

1. The standard and the leakage class the test followed.
2. The medium and the test pressure actually applied.
3. The condition of the valve at test: new, repaired or refurbished.
4. Who witnessed the test and signed the record.

A certificate describes that valve at that moment under those conditions. Comparing results between suppliers only works when the conditions match. A strong result on a new valve says little about an installed population that has run for years without attention. A shop-floor demonstration sits between the two worlds: controlled, but on a valve chosen by the demonstrator. What a live demonstration can and cannot prove sets out those limits.

4. What counts as field evidence

Field evidence starts with screening. Operating torque, the feel of the handwheel, sound across the seat, and pressure behaviour across a closed valve all carry information. Each of them indicates or suggests a condition, and none of them proves it. The confirming test for a passing valve is a seat-leak test under agreed conditions.

Common forms include pressure decay across the closed valve, double block and bleed monitoring, or a witnessed leak test. API 598 and ISO 5208 are factory acceptance references. They define how leakage classes are expressed, but they are not field test results. Records taken before and after treatment, at the same point and in the same way, make a field claim credible. Our field results directory records such outcomes without naming clients. For the standard a credible result should meet, see what a credible before-and-after result looks like.

5. Choosing the next evidence step

Match the evidence to the decision. When you buy or refurbish a valve, a witnessed test-bed result against the specified class is the right instrument. When you maintain valves already in service, the rig cannot help, because the valve cannot leave the line. The next evidence step is a small field trial.

Before any injection of cleaner, lubricant or sealant, the gates come first: fittings in sound condition, a valve design and service that suit injection, and a permit that allows the work. Where the maker's documents exist, consult them for limits and warranty terms. See running a one-valve trial for decision rules, and choosing an online valve maintenance provider for judging suppliers. A valve condition survey can rank a whole population first. Where a valve cannot be tested or treated safely, the honest endpoint is a proper assessment, or replacement.