Design of overflow drainer for deionized water tank

[zhangbd]

Our deionized water storage tank has a overflow drainer installed to avoid overflow when level switch fails, which is a U-shape PVC pipe leads to ambient, you know, generally, that's always one U-pipe full of water. However the question is that how and when do you know there is full of water in U-pipe? I think that some kind of detector should be installed or that a new overflow drainer should be designed to resolve this? Who chould give me one hand?

 

[dtreacy]

A promblem with an overflow line, is that it is a potential area for biofouling. If the tank comes under a negative pressure it could pull back all the stagant water in the U pipe and contaminate the tank and system.
An alternative to this is to install a sanitay rupture disc on the tank that will relive any overflow or over pressure.
Also if you are worried about level switch failure, install a redundant level switch that can be used as a conformation of high level or can be positioned as above the existing switch to act as a safety mechanism on a high high level signal.

Hope this is of help.

Regards
Daragh

 

[dtreacy]

A problem with an overflow line, is that it is a potential area for bio-fouling. If the tank comes under a negative pressure it could pull back all the stagnant water in the U pipe and contaminate the tank and system.
An alternative to this is to install a sanitary rupture disc on the tank that will relive any overflow or over pressure.
Also if you are worried about level switch failure, install a redundant level switch that can be used as a conformation of high level or can be positioned as above the existing switch to act as a safety mechanism on a high-high level signal.

Hope this is of help.

Regards
Daragh

 

[zhangbd]

Dear Daragh,
Thanks very much for your kindly help and, you're right this drainer is a potential area bio-fouling. The existing U-shape section is to separate the tank side from outside, and the pressure inside actually is at ambient pressure due to another breather filter installed. Auditor's advice is that we must make sure that water in the U-shape section of drainer is full, if it's empty, bio-fouling will occur. Therefore, our purpose is to avoid emptying the water in U-shape.
You mentioned that rupture disc may be helpful, but I don't know if that could be intalled in drainer pipe? I think that will be a protection mechanism for application. If this mechanism(rupture disc) is available to drainter pipe, How and where I could get that?

In other hand, if we keep the water in drainter in flowing all time via a slim pipe from tank to prevent emptying inside, is this disposal useful?
Kindest regards

zhang baodong

 

[greg87]

The best approach is to eliminate the open overflow and provide a sanitary rupture disk on the top of tank with a discharge line to the floor/drain. This will prevent overfilling, or overpressure due to any problem. A high level switch will still help avoid high level problems and prevent the disk from breaking in all but the most unlikely cases. The tank will always be sealed, breathing only thru the vent filter.
HOWEVER, I don't know the type of tank you have and what, if any, pressure rating it has. If it is a very low rating, getting a sanitary rupture disk rated to burst before the tank does may be expensive or problematic.
Another option, if you have a tall tank, is to put a rupture disk on the bottom end of your overflow line. Regardless of the rating of your tank, if the waterlevel goes high and runs into the overflow line, it will generate a head on the disk equivalent to the height of the line. If the line is 12 feet vertical heigth, the column of water will produce 5 psig on it. A 5 psig burst pressure is reasonable. Note that you would not have, or want, a U bend in the overflow line in this case. Note that this approach is only for ambient DI water in an indoor location - so that condensation would not occur in the overflow line and accumulate above the disk.
Your last idea of a small continuous flow of water into the liquid seal would work and would be an inexpensive method to maintain a seal in the existing system.
The issue of tank going negative and sucking water in from the seal could be prevented by oversizing the line, and in most cases would be very unlikely. What would be likely using this type of vent seal ia that, due a problem with the vent filter or a much higher draw rate out the the system than you have predicted, you get a level of vacuum that exceeds the height of water in your U bend. You would then get outside air bubbling thru the seal and into your tank. The air would represent a source of contamination. More importantly, you probably would not even be aware that this was happening.

 

[MedicineEng]

Dear All:

This is my oppinion:
1. U bend overflow pipe is unnaceptable in terms of GMP wise and you will have a lot of problems to explain that design to a picky auditor;
2. I've worked with PW systems for some time and the way that it was found to cope with this issue of the overflow is any of the above refered:
My system has three level switches, lets call HSL (high switch level), HHSL ( High high switch level) and LSL (Low switch level). In normal operation, when tank reaches the level HSL, the PW skid will go into recirculation and will not put any more water into the tank until the LSL is reached. Only then the 3 way valve is put again in the charging tank position. If even so, HHSL is achieved, PW skid goes into alarm and stops, the buzzer sounds and the operator has to go see what was the cause for alarm. I think that the probability of having two faulty level switches at the same time is quite remote.
I don't know the size of your tank, but my switches are set at 50% (LSL), 80%(HSL) and 90%(HHSL). And up to now I never had any problem with overflow.

Hope that helps.

PR

 

[dtreacy]

Hi all,

I agree, a U bend overflow poses too many problems in terms of contamination.
We have a similar set up to the one described by PauloRibeiro.
The only additional comments that I can state are as follows:
1)A level transmitter could be used instead of level switches it is less intrusive then float or reed switches. The LT we use is a differential pressure LT. The pressure differential is measured between the base of the tank and the top (head space) of the tank. This differential is converted into head [m] and therefore % tank level. The bottom tank connection is made via NA Novaseptic type connect, which is self-draining and flush with the tank wall. Note that this method does not have the same safety level as the multi level switches but is less intrusive. Depending what the water is used for one or the other is applicable.

2) The “fail state” of all valves is also important.
The valve feeding water to the tank should fail to a closed state and the water diverted elsewhere. This would stop an over fill situation in the event compressed air or power failure.

3) The rupture disc should be sized for the pressure rating of the tank as stated earlier. A sensor should be installed to detect disc breach and shutdown the system.
Direct overflow to a safe area.

Regards
Daragh