process control across two separator system

[gelp78]

I am trying to design a control system across 2 oily water separators, tying into dedicated pumps and the outlets of which are tying into a joint header before going into a desanding unit, the problem is the level control d/s of the desander and minimum flow control tie-in. The two options are single level control with a hand switch going back to the two separators and using the control from one or another separator or accepting that during low level there is a possibility of sand getting into the downstream equipment, which includes a plate separator and 6 water injection pumps.

Any suggestions would be gratefully received as to whether anyone has any other ideas on this system.

Cheers Gill

 

[monaco8774]

If you have a hand switch to select level control then I presume you are only running one vessel at once, otherwise when you switch to one vessel you will have nothing controling the level in the other.

Possibly a foolish question but why do you have two oily water seperators, if you only had one the problem would go away, and in my part of the world anyway we dont design 100% redundancy into oily water seperators ( we put bypasses and isolation in, especially since you stated you have plate seperator downstream)) as they dont tend to go wrong, if its a capacity issue, one larger vessel is generally smaller and cheaper than two smaller vessels.

 

[gelp78]

The issue is a brownfield one. These oily water separators both run at the same time, there is no redundancy. We are trying to retrofit a PWRI system into an existing design. If we only had one separator as you say we would not have a problem. There appears to be an error in my original text also, it is not a plate separator, it is a plate exchanger downstream, which obviously we do not want sand to get into the plates and require maintenance cleaning every two weeks. Thank you for your response monaco8774, however we have come to the conclusion following study work that this original idea was not one that would ever work. I was looking for alternative ideas to the option given to us from our client (the one control handswitched between the two separators). I agree with you also in my part of the world we don't tend to design 100% redundancy into separators either.
Thanks again.