Best Practice Float Traps and Check Valve 1

[sshep]

Gentlemen,

While I was making some midnight rounds on a start-up, I found an operator draining a large amount of condensate from a waterlogged exchanger to grade. He asked why it was taking so long. The exchanger had a float trap lined up to the condensate system but no check valve, so the condensate was merely backflowing from the condensate return system. I have previously noticed that it is generally not recognized by operators (and some engineers) that most float traps do not have an integral check valve- if there is level in the trap, liquid flow can pass either direction depending on the pressure.

At another company I worked at it was standard detail to include a check valve after the float trap. This is also shown in the Spirax-sarco hook-ups manual. The standard drawings at this site for exchanger drain traps does not include a check valve.

Does anyone have advice on best practices for exchanger process trap drain piping? Are there any special start-up or loss of steam hazards that can exist because having (or not having) a check valve?

best wishes,
sshep

 

[Joesteam]

It sounds like you need a condensate pump because if you have higher pressure in the condensate line than at the inlet to the steam trap. Typical heat exchanger 'stall'.
Yes, a check valve will prevent back flow, however, you still will not be able to drain your condensate due to lack of differential pressure.
Many systems with modulating control valves cannot handle ANY backpressure, and may even run under a vacuum, even with positive pressure entering the heat exchanger. This problem is often solved by isolating the heat exchanger from the rest of the condensate system - using a pump.

Joe Lambert

http://www.control-associates.com/

 

[insult2injury]

sshep: I think that if you reference most manufacturer reference material (such as Armstrong, etc.), they show a check valve installed downstream of the trap. I do believe that I've even seen it in Spirax literature.

Joesteam: I believe that the heat exchanger was not in service and they were draining it prior to startup.

I2I

 

[insult2injury]

It should also be noted that for the majority of applications, the check valve is not required due to the drain pipe routing which is typically in a downward direction.

I2I

 

[sshep]

Thanks! Insult2injury was correct, the question wasn't about exchanger stall or the choice of pumping traps, etc. The stall condition was just a fact of the exchanger and process side of the reboiler being cold, with only a warm-up steam flow going in which the operator intended to drain to grade.

I guess I should have clarified that this was a trap discharging to an elevated condensate return header. The operator had lined up the trap and was draining at a low point before the trap, not realizing that there could be back flow from the return header. I was wondering what is best design practice regarding a check valve, and what other start-up hazards might exist if someone did not recognize the possibility of backflow.

best wishes,
sshep

 

[MJCronin]

I believe that there might be another issue besides the check valve, which of course is necessary.

Some of the standard details for steam traps show a discharge into the top of the condensate return header.

Could your piping be connected to the bottom of the overhead header ?

-MJC

 

[sshep]

MJC,
As always you make a good point (as you usually do) about the piping. This is a detail that should provide some mitigation, and that I didn't think about. In this case it looks right (entering header at top and then header sloping down to the drum), but even so the return is a very long 10" line climbing and running back to the main header so it could pocket alot of condensate. Thanks for the reminder.
best wishes, sshep