I have heard of three alternate methods (vs the traditional one consisting of allowing 50 to 25% of the system frictional pressure drop excluding itself to the control valve) to determine the allowable pressure drop across a control valve during the design stage:

1/ Connell's one using a formula that everyone knows

2/ A method consisting of assigning a minimum pressure drop (10 or 15 psi) at its maximum expected design flowrate and at a upper opening limit 80%. This method has been presented by Frank Yu in "Easy way to estimate realistic control valve pressure drops" issued August 2000 in HP.

3/ Also, in section 3.7 - chapter 3 of the very good free ebook available at http://www.lightmypump.com/pump_book.htm
It is recommended to assume a pressure head drop across the valve of 10 ft of fluid at its maximum expected design flowrate and at a upper opening limit 90%.

My feeling is that the Connell's one is more rigorous as the notion of controllability is incorporated in the equation, but in the article corresponding to the second method, Mr YU concludes that his method is better than the Connell's one (from an economical point of view namely)...

A/ Then i would like to know if some experienced engineers could tell me which one they would recommend and why?
AND their feeling on the third method?

B/ Also, I would like to know the origin of the methods 2/ and 3/, i.e. why 10-15 psi or 10 ft of fluid? why these values?

The application may dictate some design issues.  For example a steam letdown valve might not follow any common guidelines.

Method 3 MIGHT work in selected applications, but it sounds more like drop expected with an on-off valve.  

1/3 the system pressure drop at the valve at max flow is the rule-of thumb I was taught. That gives effective control. And "System pressure drop" is almost NEVER the same as the system pressure.  JLSeagull's reference to a steam blowdown valve being the exception, where upstream is pretty much constant and downstream is atmospheric.    

Note that As the flow is dropped back the system pressure drop decreases as the square of the flowrate, and the pressure backs up on the pump curve,  so you're not backing up the flow much before the valve is dropping essentially ALL of the system flow.  

Yet I see data sheets every day where the inlet pressure is specified as the same pressure regardless of flowrate, and the outlet pressure is specified as the same pressure regardless of flowrate, and only the flowrate changes. Those numbers cause me concern because they will NOT reflect reality.

MORE NEWS