Re the February 2015 Control Talk column on the commissioning of control valves (How to Avoid Problems in Electrical and Instrument Installations): How does one go about proving that the actuator is sized large enough to handle actual process conditions? For example, it may be 31 Bar G at 400 °C upstream and 1 bar G at 150 °C downstream. The stroke testing methodology you describe will not show you. Then when the plant is live, it may be too late. The actuator may not be sized sufficiently to close against such a high differential pressure.
Greg McMillan replies:
Good question. The problem of an undersized actuator is most apparent when the valve tries to close and open at operating pressures. Notify the supplier of the requirement to have 150% of the thrust/torque required for shutoff and opening at the maximum pressure drop. You can also ask for test results in the supplier’s pressure-flow lab.
I advocate doing stroke testing not only at the normal operating point, but also at the closed position at worst-case pressure drop. Ideally, this pressure can be simulated in a shop setup and during water batching. With the valve closed, 0.5% increments should be applied, and the results monitored for stick-slip and overshoot. Once the valve opens, then 0.5% decrements should be applied, and the results again monitored for stick-slip and backlash (dead band).
This testing at the closed position is often not done. Suppliers are reluctant to do it because stick-slip is greatest at the closed position. At startup and for low operating rates, the valve needs to smoothly open and close.
Some tight shutoff valves used for surge control would overshoot to 25% open when asked to open just a few percent due to changes in a downstream user flow, resulting in a severe upset to the compressor and to all users. pH reagent valves are notorious for riding the seat due to the extreme rangeability requirements from the titration curve nonlinearity.
The overhead pressure and condenser temperature control systems of batch reactors have extremely low flow requirements at the start of the batch cycle, resulting in control valves riding the seat. One of the best things I did for one protégé in the ISA Mentor Program, who was experiencing poor overhead system batch reactor control from the use of a tight shutoff valve for throttling, was to recommend the tight shutoff valve be used for isolation and a throttle valve with low friction and oversized actuator be added for control.