3 replies to this topic

[Iarik]

Good day, I have a system consisting of 7 multistage centrifugal compressors (precoolers, intercoolers, aftercoolers). Cooling system is of a closed type with demin water as a coolant, contained in 2x85 m^3 sealed concrete tanks, runs through water/water plate heat exchangers before going to the plant. I have picked up that pressure releave valves on the tanks periodically open, releasing air. That means that cooling system has got air ingress from compressors side, as water pressure is lower than compressed air pressures. As a result, because of a poor heat exchange all temperatures at the plant (compressed air at the compressors, lub oil, cooling demin water system) are above normal value. Has anybody faced this kind of a problem? Any suggestions on the way how to determine the place of air ingress except opening and testing of the coolers (as 2 off the compressors cannot be stopped)? And one more question: is there any difference in how heat exchange will be influenced by air ingress in tube-shell and plate heat exchangers? Which type of heat exchangers will suffer more? thank you

[Art Montemayor]

 

 

Larik:

 

I wish you would have taken the time to develop and submit a detailed sketch of your cooling water system, showing where the concrete tanks are located and how the piping is installed.  If this is a closed loop, indirect cooling water circuit there should normally not be any need for using such large tanks within the circuit – unless there is a broader scope of work than just supplying the air coolers.  You normally have an expansion tank located at an elevated position within the closed loop.

 

I am concerned that you may, indeed, have an air leak developed in one of your tube bundles.  Since you haven’t told us the discharge air pressures involved, I get concerned that this probable leak may be in an aftecooler and the excessive air pressure might break out into a full tube rupture that would immediately pressurize the concrete tanks (which normally can’t withstand much pressure.  Hopefully your pressure relief valve on the concrete tanks is of sufficient size that it can handle the full air capacity of the compressors.

 

To answer your specific queries:

1. Has anybody faced this kind of a problem?  - I have.
2. Any suggestions on the way how to determine the place of air ingress except opening and testing of the coolers?  -I know of no method beyond testing the individual coolers, one by one.
3. Is there any difference in how heat exchange will be influenced by air ingress in tube-shell and plate heat exchangers?  - I can’t imagine how you could about doing this, especially without detailed piping diagrams (especially isometrics).
4. Which type of heat exchangers will suffer more?  - It is not the type of exchanger that suffers; it is the alloys employed, the quality of the water circulated, the mechanical design of the exchangers, and the concrete tanks that are threatened.

 

 

[Iarik]

Hi Art,
 
Sorry for not including the sketch from the beginning. See it attached (no pumps or valves indicated).
 
   cheresources.jpg   102.24KB   3 downloads

In essence the demin tanks are designed in a such way, that it is nearly impossible to overpressurise them. Apart from pressure releave valves each of the tanks has a rupture disc installed, it is 350 mm in diameter.

Therefore, I think the tanks are not the biggest concern. I also do not think that a pipe inside of any of the coolers can rupture fully as the air is on the shell side (tube/shell coolers).
The biggest concern is a reduction of a cooling capacity, which influences performance of the machines, as well as increases temperatures of oil.

I will have to plan for pressurising of each cooler separately...

 

Thank you for the feedback!

[curious_cat]

1. Your performance loss & the observation that the air release valves pop periodically could conceivably be entirely unrelated phenomenon. 

 

2. Can you check delta T or a performance metric across individual HEX circults? If indeed one leak is the cause you might see one HEX with grossly degraded performance. 

 

3. Depending on your piping flexibility you could pairwise route different combinations to different tanks or if you have spare tank capacity even a single HEX to a tank and see if you can isolate which combination gives you the air release.

 

4. How sure are you that this valve popping hasn't existed from Day1?