control valve location 2

[steveb1111]

As a rule of thumb, for a

A bubble point liquid feed from a separator is preheated in an exchanger before entering a stabilizer tower. There is a control valve in the feed line where pressure is letdown from separator pressure (20 bar) to stabilizer pressure (8 bar).

Should the control valve be located upstream or downstream of the exchanger.

 

[sshep]

Steve,

By literal definition, the liquid coming from the seperator is at bubble point at 20bar. I think you are therefore really targeting some small amount of vaporization in order to get the feed to the bubble point of your stabilized product. As the amount of vaporization will affect your size and cost (piping, valve, HX) this is a good thing to quantify.

Although we as process engineers often like to see the control valve on the outlet of the exchanger to better know the feed thermal condition based on the temperature prior to letdown across the valve, the more frequent truth is that the cost will be lower to place the valve upstream of the exchanger. Specifically: the pressure rating of the exchanger will be lower; your temp driving force (HX size) may be significantly better if the exchanger is operated at lower pressure; you may be able to avoid an additional thermal relief valve if an always open path is designed to the column; your control valve may be smaller and in less severe service (i.e. less flashing/cavitation); and finally if the heating medium is column bottoms (typical) then a leak will be into tower feed instead of product. I have plenty of experience where the latter reason alone has made the upstream placement of a letdown valve a slam dunk call.

I would love to get more opinions, but barring compelling reason which I have not thought of, I think this is one where the economics alone may convince you that one way is better than the other.

best wishes,
sshep

 

[djack77494]

I think the best approach to this problem is one of determining the most practical way of configuring the valves and equipment. Would you find it acceptable to have a two phase mixture entering a control valve? Personally, I would abhor such a solution and would do everything possible to avoid it. I'd like to think that you could come out of (downwards) the source vessel and have enough static head to ensure 100% liquid going into the control valve. You'll flash through the valve and will have to introduce a two phase stream into your heat exchanger. I'm not fond of doing that either, but I think it's the lesser of two evils.
Design carefully and good luck,
Doug

 

[Samiran]

My experience is that when you have such a high delta P and flashing across a control valve, the line downstream undergoes heavy vibration and frequent damage occurs on its supports which can compromise the rates through it. Hence, the most desirable location would be to locate the control valve as close as the downstream vessel. But, as always, other factors may take preceedence and location may change.