Seat leakage norm for control valves

Do control valves need to shut off

S. Vijayakrishnan - 12 January 2021

On/off valves in use are either kept fully open or fully closed. They are rarely kept partly open. Some designs are not even suitable for use that way.

About the author

Mr S. Vijayakrishnan
Dr. S. Vijayakrishnan is a valve professional with over 30 years’ experience primarily in application engineering and marketing. He has worked with VAAS and with the Indian operations of DeZURIK, Habonim and Bray. He can be reached at
In contrast, control valves are selected and sized to provide a range of flow for a specified pressure drop across the valve. Almost all of them spend their life operating within a minimum and maximum opening to provide the corresponding flows.During normal operation, they are never moved to a fully open or fully closed position.

When an upset condition requires the flow to be fully shut down (or in other cases, fully released), another on/off valve positioned next to the control valve is used.

Considering that a control valve is not used to shut-off flow, the leakage characteristic of control valves in the closed condition should be irrelevant for selection as opposed to on/off valves. However, we find elaborate specifications for seat leakage stipulated in most control valve datasheets.

In fact, most of the seat leakage standards have been set up originally by control valve industry bodies such as the ISA!

Complying with stringent seat leakage norms usually leads to expensive design features and /or manufacturing operations which add to the cost of the control valves. Depending on the class of leakage, the seats are designed to suit and the finish and tolerances of mating parts in seating are selected. Elastomer materials are selected to provide near-zero leakage where feasible; metal-to-metal seating with expensive hard facing and/or finishing by grinding, lapping, polishing, coating, etc.are used. Valves are then tested to confirm that seat leakage is indeed within the norms of the standard.

In general, valves require the maximum operating thrust/torque at a fully closed position to overcome seating forces. Sometimes the required force is further increased by the extra force that may be required to seat complying to seat leakage norms. The actuator selected on this basis could be larger and more expensive than one that is really required. A suitable actuator is required only to overcome the dynamic fluid forces while throttling between, say, 5 percent to 95 percent opening – which is what control valves usually do. In fact, throttling at more than 95 percent closed position may affect the life of seating surfaces and is generally not recommended for control valves.

However, all this additional cost hardly adds any value to the control valve’s real-life performance as the seat tightness capability is totally unnecessary! If the end-user community recognizes the futility of current seat leakage specifications of control valves, they can make effective changes in the standards.

For example, the norms could be relaxed to align with what various types of valves achieve by default.

For example, a normally machined, metal-to-metal seat, single-seated globe control valve may achieve Class II (as per ANSI/FCI 70.2) seat leakage without any additional cost being incurred. In that case, settling the leakage norm to Class II provides a norm to assure minimum quality from manufacturers without adding any cost.

Another approach would be to do away with the seat leakage norm for control valves. A beginning in this direction is already seen with some recent control valve models doing away with the seat leakage norm in their specification- typical examples are the Series 39 Control valve from Bray or the Figure 635 Slurry Control valve from Emerson/Keystone.

The current norms for seat leakage may be continued without change in the case of on/off valves where they are, of course, relevant. However, some modifications may be possible on the variety of standards (Class I to VI in ISA S75.02 for example) with a rationalization to few norms. For example, one norm for elastomer seating, and one or two for metal-to-metal seating may be sufficient.

We should also realize that these valves are tested with clean media in ideal test conditions by the manufacturer before dispatch. The actual seat tightness may usually somewhat deteriorate. Having very stringent norms for pre-dispatch tests may thus be somewhat pointless.

Share this