Overflow for Pressure Vessel?

[keschnipp]

Is there any way to protect the inlet of a pressure vessel from backup?

We have hot DI water (240F) used in two places throughout our plant. One section involves a contactor with a solvent and the other is for final purification before crystallization. A week or so ago, the supply pumps for the DI water failed and the solvent backed up the header into the DI water system. When the pumps came back on, the solvent got into the final product loop. Five days production contaminated. Needless to say, upper management was not happy.

To fix, one idea was to put in a pressure vessel (to maintain the water temp) but to include a way to prevent the back up, similar to an overflow on an atmospheric tank. Is there any way to do this? We have other ideas in development, but this was upper managements original idea.

Thanks for the help.
Kurt.

 

[friartuck]

Perhaps i am missing the point, but have you ever heard of a non return valve, double check valve or a RPZ reduced pressure zone valve. All of these are used to prevent backflow.

If fitting one risks pressure vessel over pressurisation, then fit an expansion vessel.



Friar Tuck of Sherwood

 

[smckennz]

Do a web search for backflow preventers. They are a bit like two check valves in series with a drain between them.
They are commonly used in potable water systems where there is risk of backflow contamination.

Cheers

Steve

 

[Montemayor]

keschnipp:

There are several ways to avoid back-flow in your situation. First, I'm going to have to make assumptions over some basic data that you didn't furnish:

1) Both of your target DI applications are vessels that are pressurized at a level above that of the suction to your pump(s);

2) The DI introduction into each target vessel does not incorporate an "air break" - i.e., a virtual separation of the introduced DI from the liquid phase in each vessel; (why, I wouldn't know because this is a common way to avoid back-syphoning of the liquids);

3) You are not incorporating check valves in the DI nozzle feeding each of your vessels (again, I don't know why but this must be your case from what you describe).

The ways to prevent back flow are:

1) The most obvious and the least expensive is the mechanical employment of a double-check valve installation, in series.
2) The second way to mitigate the event is to employ instruments: the sensing element can be a back flow switch or a pressure switch (where the set pressure is slightly lower than that normally feeding the vessels. The switch must be rapid-acting and the connected block valve must be preloaded to act rapidly to shut off the backflow. You can also electrically tie-in the pump's motors' failure to actuate the block valve.

Depending on the economics or criticality, I would be tempted to employ both the above methods because in a Hazop the check valves normally are not accepted as functioning every time without failure. So, this belt-and-suspenders type of logic helps to ensure that backflow is checked everytime since a failure would constitute double jeopardy.

I hope this helps you develop a system to avoid what must be an expensive failure.

Art Montemayor
Spring, TX

 

[keschnipp]

First, thanks for the replies.

Second, we do have swing check valves installed in two places going to the injection point with the solvent. Both failed. It is not possible to have air break at the system as we need to maintain the temperature of the water (220 to 240 F) for optimal process conditions. If there was an air break, we would flash down to 212 and have to repressurize and heat back up with steam (current heat source is waste heat from air compressors).

As solutions, we are invesigating check valves better suited to the temperature, as we think the springs in the checks became both brittle and susceptible to breakage. We are also looking into the instrumentation you suggested, Art.

 

[jalvarez]

Kurt:
Based on your description I suggest to use a double shut-off and purge valve. You need to locate a point in your process, at atmosferic pressure, to send the small amount of liquid between the two shut-off valves when the system trips, automatically or manually.
If avoiding continue leakage while the system is closed is considered important, you may include a glass flow indication (or simply a small purge valve) in the purge line. If you see liquid flowing, you need to repair one of the shut-off's. The purge line would be small (tipically 1" or less). Specify the shut-off's as Class V or VI depending on your hermeticity requirements.
With this arrangement, you can obtain an air gap as Art suggested.
Locate the valves in order to minimize the extra heating requirements. I assume that the best place is close to the vessel itself, in this way you may recirculate the fluid upstream the valves to maintain the temperature.
By the way, remember one of the Murphy's: If everything fails... try what your boss suggested.
Have a safe and healthy day
J. Alvarez