Often when commissioning or troubleshooting automatic control valves, there’s a discovery that the control valve, even though fully closed, doesn’t fully shut off the flow of process fluid through the plug and seat.
Although closed, there is an “allowable leakage rate” as part of each control valve’s specification. The following article aims to clear up some confusion about control valve shutoff.
Considering control valve shutoff capability
An automatic control valve opens and closes to increase or decrease the flow of process fluid through it to keep a measured process variable at its desired setting. While these types of valves are very important and perform the job of controlling a process very well, they don’t always close tightly enough to completely shut off the flow of process. Nor should they be expected to.
The components in a control valve that modulate the flow are called “the trim”. There are different types of control valves have different components that serve the purpose of “trim” and play a part in control valve shutoff. For instance, in a sliding stem globe valve the trim is made up of a plug and stem, a seat ring, and usually some type of plug guiding or flow characterizing device called a cage. In a V-notch, segmented ball valve the trim is made up of a ball and shaft, and a seal ring or seat ring. In a butterfly valve it’s the disc and seal ring. These trim components are designed to best throttle the flow as smoothly and efficiently as needed. Asking these same components to completely and effectively shut off the flow of process fluid can often be a challenge, since that’s not what they were designed to do.
Although some manufacturers of automatic control valves will claim a “tight shut off”, on some of their valves it requires special parts, materials, and manufacturing assembly. The American National Standards Institute (ANSI) recognizes and classifies a range of acceptable seat leakage allowed in a control valve, see Table 1. That’s right; the American National Standards Institute acknowledges “a maximum allowable seat leak rate” for given requirements. Furthermore, there’s the American Petroleum Institute (API) Standard 553 for Refinery Control Valves that references the ANSI/FCI Standard for allowable leakage.
But, if the control valve is closed, should it not leak? I’ve had a mantra of sorts for years that goes like this; “Control valves control, shut off valves shut off. And neither one does the other’s job very well.”
|
Leakage Class
|
Maximum Seat Leakage
|
Test Procedure
|
|
Class I
|
See Para. 4.2.1 of ANSI/FCI 70-2
|
No test required provided user and supplier so agree
|
|
Class II
|
0.5% of rated valve capacity
|
Type A (See Para. 5.1 of ANSI/FCI 70-2); pressure applied to the valve inlet, with outlet open to atmosphere or connected to low head loss measuring device, full normal closing force provide by the actuator.
|
|
Class III
|
0.1% of rated valve capacity
|
Type A (See Para. 5.1) ); pressure applied to the valve inlet, with outlet open to atmosphere or connected to low head loss measuring device, full normal closing force provide by the actuator.
|
|
Class IV
|
0.01% of rated valve capacity
|
Type A (See Para. 5.1) ); pressure applied to the valve inlet, with outlet open to atmosphere or connected to low head loss measuring device, full normal closing force provide by the actuator.
|
|
Class V
|
5 x 10-4 ml per minute of water per inch of seat diameter per psi differential
5 x 10-12 m3 per second of water per mm of seat diameter per bar differential
4.7 standard ml per minute of air per inch of orifice diameter
11.1 x 10-6 standard m3 per hour of air per mm of orifice diameter
|
Type B (See Para. 5.2)
Type B (See Para.5.2)
Type B1 (See Para. 5.3)
Type B1 (See Para. 5.3)
|
|
Class VI
|
Leakage per Paragraph 5.4.4 as expressed in ml per minute versus seat diameter
|
Type C (See Para. 5.4)
|