4-port chilled water tank mixing

[232gulfstream]

Suppose I have a 1500 gallon tank with 4 piping connections on it. On the left side I have a set of pumps and my chiller. On the right side I have a second set of pumps sending fluid to a process line in the field. I have 2 different flow rates and 2 different temperatures on each side of this tank. It is easy to calculate the blended temperature of the water returning to the tank but how would I calculate the overall mixed tank temperature after lets say 5 minutes time.

A brief sequence of operations of how this system works:
Chiller pumps (800 GPM) and chiller run and cool tank water to 42 degrees F.
Process pumps energize (660 GPM) and take the 42 degree water to a process line and return it to the tank at 56 degrees F.
Process pumps run for 5 minutes.

What is the tank temperature after 5 minutes? Is there any software that can model this and see how the tank temperature behaves over certain time periods?

 

[LittleInch]

Sounds like a CFD analysis to me.

Five minutes probably isn't long enough to get to steady state so this will be a constantly changing number.

Also presumably it takes a certain amount of time for the return flow from the process unit to get to 56F?

if the chillers are idling / shut down because there is no DT, they will take time to ramp back up.

Lots o transient things going on here.

Just curious or is there a problem?

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

 

[232gulfstream]

There are no problems I was just curious to see if there was a long hand way of determining this. I believe this is probably a calculus problem but I was hoping maybe someone could point me toward an example that I could use as a template. I don't have access to any CFD software.

 

[BronYrAur]

I'm curious at your wording. Why would you have blended water "returning" to the tank? I'm probably just reading too much into it.

With a total flow of 1,460 GPM, you are basically emptying the tank every minute. Depending on baffling, etc. the mixed temp calculation would be tricky. If you took it to the extreme and replaced the tank with just a primary/secondary bridge, you have 140 GPM more primary water hitting the bridge than you have secondary. So 140 GPM of 42 deg water will bypass through the bridge and mix with the 56 deg secondary return water. If there is ideal mixing, the return water back to the chiller should be 53.6 deg.

If you really only have a 5 minute process, you will probably have issues with cycling. It will take longer than 5 minutes for the chiller to get down to temperature. It may short-cycle as it is trying to reach setpoint.

 

[EnergyProfessional]

I would start with a schematic of how that looks like inc. the inside. Some tanks have baffles for example. You get something between plug-flow and completely mixed.

 

[LittleInch]

You might be able to do a simplified approach using excel and use a heat energy approach in 30 second intervals.

SO if you start with 1500 gallons at one temperature you have energy X

Then you add in 330 gpm of water at 56F which has added in energy Y

So energy in the tank is 1170 at 42 and 330 at 56. Blend.

Find out what temperature the water is.

For the next 30 seconds repeat, but this time start with your blended temp, add your 330 gals again, remove 400 gallons of blended temp water and replace with 400 gal of water at 42F

New blend is 330 gal @52, 400 gal@ 42 and 770 gal at your previous blend.

Rinse and repeat





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

 

[Compositepro]

Your process is thermodynamically inefficient. It is similar to placing a heat exchanger between the supply and return lines in a chilled water loop. The load will be receiving warmer water and the chiller will be receiving colder water than if there were no exchanger (or tank). Using the heat exchanger analogy, you could arrange baffles in the tank to form a counter-flow heat exchanger, in which case you would almost decouple the chiller from the load. In effect, you would be adding an insulated wall and trying to transfer heat through it. The only heat that would actually flow would be due to lack of insulation.

Why would you do that?

If the tank your desire is to store cold water, pump chilled water into the bottom of the tank and supply the load from the top of the tank. Return water from the load directly to the chiller.