3 replies to this topic

I am trying to evaluate our chilled water system that is used to cool down the air from our blowers before pneumatically transferring our product. We have 2 chillers running in parallel that cool the water down to 50 degF. The water is pumped from a return tank using a small centrifugal pump at 55-60 psi through the chillers to a 3 in. line. This line goes to two main areas - line splits about 20 ft from the pump with a 3in line staying on the ground floor and a 1.5 in. line heading vertically about 65 feet to the top floor. Both of these lines come back together for about 20 ft before dumping into an atmospheric tank from which the water is pumped out of to form a loop. The return water is usually around 60-65 degF.

Both lines supply water to several air coolers - incoming air is around 170 degF and we use fin type air-to-water heat exchangers. The goal is to keep the air below 90 degF, preferably around 70-80 degF.

Some issues I've seen are a few specific HX don't seem to cool the air very well. It seems like the ones at the end of the line are the worst in terms of heat transfer but the water going in and out of the HX is cold. Could this just be that the HX is likely fouled or could the pressure be too low? This is happening on the ground floor too. There is virtually no instrumentation on these lines for me to check pressures/temps around the HXers.

Also on the upper floor, I've found that if a valve on the supply side of the chilled water line is opened, it will suck air into the line... which will cause the water to stop flowing (air locked?) until the system is shut down and restarted. Why would the water suck air in rather then push water out the open valve? Is possible that by going to this high of elevation, that the return water is siphoning the water through the supply side?

I know that calculating pressure in a piping system can be very difficult without software, is it OK to just add up the pressure drop due to length of pipe, equivalent lengths for bends, elevation change, loss from HX, etc. to find out if the pump is providing enough pressure?

Would it be better if this system was actually closed, rather then using an atmospheric return tank? It seems to me like we are wasting the head pressure from the water returning from the top floor by running it into an open tank.

Any feedback would be greatly appreciated,
Thanks!

[Art Montemayor]

Kayaker:

Please review the attached compressed Excel spreadsheet and the included sketch. Include any basic data that I may not have picked up from your post or that you left out.

What we need is a fully, accurate, and detailed schematic of your system - including the various line sizes, loops, and configurations. What you describe is a general description. With the elevation changes and flow differences, it is not strange to find that you are having problems. These problems can be easily guessed at just from the brief description. However, with specific piping layout descriptions, we can get closer to the real, hard facts of what is taking place and why.

You can add additional information on the piping and flows on my sketch or you can add your own. As long as the information is clear, concise, and accurate we will have a better chance of getting to the basic problem(s) faster and with accuracy.

You are basically correct in suspecting that you are losing something in your open, recirculation system. Most - if not the greatest majority of these chilled (& hot water) water systems are designed to be 100% closed - and for good reasons. One of these reasons is to have an expansion tank or "bladder" in the system that will absorb any flow changes, expansions, and shocks. The enclosed system also avoids any inclusion of air and its harmful oxygen content that only presents rusting potential of your system.

I will await your response and hope that this spreadsheet helps the other Forum members to better analyze your system and identify any needed basic data that is lacking or required.
 Chilled_Water_System_Design.zip   224.65KB   169 downloads

[pawan]

It seems that ur case is a good example of pressure unbalance. Probably U have same size exchangers at bottom & top floor & so need same water but line size is half of the ground floor piping. Tell me why would water travel thru the top floor when a good path is available from bottom floor. Thats why U have air suction at the supply header.

Ur system will have pressure drop only due to Chillers & exchangers in the bottom floor. So out of total 55-60 psi If I equally distribute it, It will be roughly 30 Psi at the outlet of chillers. Now U have 65 feet vertical height which is ~30 psi equivalent roughly coupled with other fitting losses, It wont go to top. In this condition U will have negative draft in supply line at the top. But this must be confirmed by the hot condition of air in top exchangers.

Even if U dont have any measurement U can confirm this by touching ur pipelines. Dont bother about Softwares for DP calculation, No software was available in old days & people were still doing good calculations based on fundamental equations.

Now do few things.

1. Prepare a skectch of lines with length measurement & fittings indicated.
2. Manually calculate pressure drop in each branch.
3. Change the line size till U get same [ressure at meeting points of two branches.
4. Start from lower size for bottom floor piping & higher size for top floor.
5. Confirm the flow situation by sensing air side pipelines manually at top floor.
6. Send us the sketch we can suggest a better way.

Thanks for the feedback. I'm working on gathering more info.