Process Safety regarding gas seal water supply 3

[Iomcube]

Cl2 & H2 should never mix ...but I have a situation where I have H2 compressor, H2 seal & Cl2 seal being fed by same water line.

What if x2 events happen simultaneously:
1] Supply water tank become empty so water line is empty
2] H2 seal is CS (carbon steel) & the immersed line can develop pitting holes & small H2 can be present in the top ullage of H2 seal

...in case 1st event is true this H2 will be sucked into Cl2 seal via empty water line (bcz Cl2 seal is under suction)

 

[1503-44]

Where is the PID?
An alarm should wake up the operators while the process shuts down and valves to the tanks close.

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[Iomcube]

@ 1503-44
I have yet to employ feasible safeguards ...thats the question

 

[1503-44]

You asked "what if", not how to control it.

It would seem that when any seal water tanks go low, you need an alarm and when it reaches low-low, or if the previous low alarm is not acknowledged by the operator, a process shutdown sequence needs to activate. Closing one valve does not appear to be sufficient.

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[Iomcube]

I think a U-Bend at feed water line can be regarded another safeguard

 

[LittleInch]

First rule is eliminate the risk.

So the CL tank should be fed by a separate tank or a separate water line. Only if there are very good reasons not to ( and cost isn't one of them)

The tank has a pretty low margin any way (400mbarg CL pressure vs 460 mbarg at 5m with your 40 mbarg vacuum on the top of the water column).

Actually maybe less than that as the 5m shown is water level to top of that external tube, but it's the depth of submergence of the pipe which is crucial. What length is that? Exactly 5m as well?

Loose less than a metre of water and your seal is gone.



Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Iomcube]

So the CL tank should be fed by a separate tank or a separate water line. Only if there are very good reasons not to ( and cost isn't one of them)
It is not possible

What length is that? Exactly 5m as well?
As you can see in the illustration it is (dip leg) exactly 5m

Loose less than a metre of water and your seal is gone.
Yes...thisnloss of water only occurs if initially seal breaks bcz of overpressure

 

[LittleInch]

"It is not possible"

Sorry, but at concept phase that is really not the right attitude to take and even if you think it is you need to document why and be able to justify it possibly to an external body.

The illustration is exactly that - an illustration, not a detail drawing and it doesn't specify that distance. We can only see what you tell us.

I get the impression I'm annoying you and your answers are coming across as a bit "spiky". You might not like having your design being questioned, but we're all just trying to help here. If you don't appreciate my comments I'll shut up and let you get on with what I personally think is not a well thought out design. IMHO.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[1503-44]

And maybe if you did that a total process shut down would not be so pressing of an issue.

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[Iomcube]

@ LittleInch

I get the impression I'm annoying you and your answers are coming across as a bit "spiky". You might not like having your design being questioned, but we're all just trying to help here.
Haha. Sir English is not my native language so I can write what I think is correct but its positiveness & or negativeness; I am not well aware of. Sorry if my comment appeared 'spiky'. Because of space limitation only x1 tank can be utilised.

The illustration is exactly that - an illustration, not a detail drawing and it doesn't specify that distance. We can only see what you tell us.
I did specify linear & elevation distances in meters.

@ 1503-44
And maybe if you did that a total process shut down would not be so pressing of an issue
So right now we have x3 feasible solutions:
1] Add some digital Level indication to feed tank
2] Piping U-bends
3] ...maybe individual check valves at feed water lines to each seal

On the 3] point. Check valves maybe a problem at small feed flows especially if tank supply is gravity assissted alone

 

[1503-44]

U bends may get their liquid traps blown out if velocity is not well controlled.
Check valves are known for unreliable sealing.
I wouldn't count on either for critical applications. Valve backup would be preferred.
I'd do the alarms with automatic valve closure at low-low, shut down sequence continues as necessary. Tank levels are always nice to know.

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[Iomcube]

U bends may get their liquid traps blown out if velocity is not well controlled.
I am unable to understand this

 

[LittleInch]

So what's wrong with separate pipes.

Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[Iomcube]

So what's wrong with separate pipes.
Accepted

 

[1503-44]

Prevention is always best.

Liquid traps are pressure sensitive, can also trap unwanted air/gas reducing flow rate or causing vapor locks, and subject to unwanted siphon, need a sight glass and you're never really sure if they are there when you need them, especially on startups, when you probably need them the most. Save those for the kitchen sink.

--Einstein gave the same test to students every year. When asked why he would do something like that, "Because the answers had changed."

 

[georgeverghese]

Dip legs acting as safety seals should always be made of some corrosion resistant material. SS316 may be okay for H2, but for wet chlorine, some kind of superduplex stainless steel or titanium.

 

[MJCronin]

This particular thread is another excercise in hostile attitude ... IMHO

The OP comes to a group of strangers, asks for help, then acts with hostility when answers do not suit them...

I have never understood this and never will

This has been going on for decades on eng-tips..

(It is important to note that Iomcube has yet to use "Please" or "Thank you" ... in this thread )

MJCronin
Sr. Process Engineer

 

[Iomcube]

@ MJCronin
I must reply to your allegations

First this thread has x3 great posts ...& I starred them! Thats a quantitative thanks you.

Second go through this whole reply [the bold part]

@ LittleInch

I get the impression I'm annoying you and your answers are coming across as a bit "spiky". You might not like having your design being questioned, but we're all just trying to help here.
Haha. Sir English is not my native language so I can write what I think is correct but its positiveness & or negativeness; I am not well aware of. Sorry if my comment appeared 'spiky'. Because of space limitation only x1 tank can be utilised.

The illustration is exactly that - an illustration, not a detail drawing and it doesn't specify that distance. We can only see what you tell us.
I did specify linear & elevation distances in meters.

@ 1503-44
And maybe if you did that a total process shut down would not be so pressing of an issue
So right now we have x3 feasible solutions:
1] Add some digital Level indication to feed tank
2] Piping U-bends
3] ...maybe individual check valves at feed water lines to each seal

On the 3] point. Check valves maybe a problem at small feed flows especially if tank supply is gravity assissted alone