Choosing an improperly applied sized or improperly sized control valve can have serious consequences on operation, productivity and most important, safety. Here is a quick checklist of basics that need to be considered:
- Control valves are not intended to be a an isolation valve and should not be used for isolating a process.
- Always carefully select the correct materials of construction. Take into consideration the parts of the valve that comes in to contact with the process media such as the valve body, the seat and any other “wetted” parts. Consider the operating pressure and operating temperature the control valve will see. Finally, also consider the ambient atmosphere and any corrosives that can occur and effect the exterior of the valve.
- Put your flow sensor upstream of the control valve. Locating the flow sensor downstream of the control valve exposes it to an unstable flow stream which is caused by turbulent flow in the valve cavity.
- Factor in the degree of control you need and make sure your valve is mechanically capable. Too much dead-band leads to hunting and poor control. Dead-band is roughly defined as the amount of control signal required to affect a change in valve position. It is caused by worn, or loosely fitted mechanical linkages, or as a function of the controller setting. It can also be effected by the tolerances from mechanical sensors, friction inherent in the the valve stems and seats, or from an undersized actuator.
- Consider stiction. The tendency for valves that have had very limited travel, or that haven’t moved at all, to “stick” is referred to as stiction. It typically is caused by the valves packing glands, seats or the pressure exerted against the disk. To overcome stiction, additional force needs to be applied by the actuator, which can lead to overshoot and poor control.
- Tune your loop controller properly. A poorly tuned controller causes overshoot, undershoot and hunting. Make sure your proportional, integral, and derivative values are set). This is quite easy today using controllers with advanced, precise auto-tuning features that replaced the old fashioned trial and error loop tuning method.
- Don’t over-size your control valve. Control valves are frequently sized larger than needed for the flow loop they control. If the control valve is too large, only a small percentage of travel is used (because a small change in valve position has a large effect on flow), which in turn makes the valve hunt. This causes excessive wear. Try to always size a control valve at about 70%-90% of travel.
- Think about the type of control valve you are using and its inherent flow characteristic. Different types of valve, and their disks, have very different flow characteristics (or profiles). The flow characteristic can be generally thought of as the change in rate of flow in relationship to a change in valve position. Globe control valves have linear characteristics which are preferred, while butterfly and gate valves have very non-linear flow characteristics, which can cause control problems. In order to create a linear flow characteristic through a non-linear control valve, manufacturers add specially designed disks or flow orifices which create a desired flow profile.
These are just a few of the more significant criteria to consider when electing a control valve. You should always discuss your application with an experienced application expert before making your final selection.
Source : msjacobs
In item 8 linear control valve characteristics are prefered. From my experience, most globe valves have an equal percentage curve. Itis the opposite than a quick opening characteristic. With an equal percentage curve, the valve opening in the beginning gives less flow trough than at the end. The purpose of this ids, that it will result in a lenear flow trough behavior since the delta-P over the valve will also change in most cases. Normally more than 80% of industrial control valves have an equal percentage curve.
If we keep flow sensor upstream of the control valve.. how do we get feedback.