It is not uncommon for the predictions of a pulsation or vibration analysis to disagree with field measurements on the installed system. When this happens, the first question is always: which one is right? The answer matters because corrective action based on the wrong information will either be ineffective or will make the problem worse.
Sources of Disagreement
The most common sources of disagreement between analysis and field measurements fall into three categories. First, as-built geometry differences: the installed piping does not match the geometry used in the analysis. Even small differences in pipe lengths, elbow radii, or support locations can shift acoustic and mechanical resonances by meaningful amounts.
Second, gas composition differences: the actual operating gas composition differs from the design basis used in the analysis. Since acoustic velocity depends on gas molecular weight and temperature, a different gas shifts all acoustic resonance frequencies.
Third, modeling simplifications: all acoustic models include simplifications. Piping fittings, valves, and instrumentation are modeled with approximations that introduce some error. In most cases these errors are small, but in sensitive systems near resonance they can be significant.
The Reconciliation Process
When field measurements disagree with analysis predictions, the correct approach is to update the model to match the as-built conditions and re-run the analysis. This process — often called model updating or field correlation — reconciles the analytical model with the measured behavior.
Once the model accurately predicts the measured vibration, it can be used with confidence to evaluate corrective options. An un-correlated model should not be used as the basis for corrective action recommendations.
TTI Field Engineering performs field measurement campaigns specifically to support model correlation efforts. The combined capability — field measurement and analytical modeling under one firm — allows us to close the loop between analysis and reality efficiently.
