Best way to throttle Plate and Frame Heat Exchanger

[AdverseAdventure]

We have sized our heat exchanger to meet the full demand of our process equipment, but for 75% of the time we will need to run at 1/3 capacity. We have a closed loop 45F chill water supply and a closed process water loop that operates three machines with a 250gal open atmosphere tank. Each machine outputs 200,000 BTU/h at full load. So, basically, we will have a 100,000 - 600,000 BTU/h variable load.

Our vendor for the heat exchanger knows nothing about controls equipment, they can only punch numbers in their calculator and spit out a resultant size. Their numbers are OK as they match our approximations for # of plates and surface area.

The real question:
Can I throttle the cooling capacity of the heat exchanger by using a PID controlled globe valve that reacts based on my process water temperature? Is it a good idea to PID control the chill water as well? I would rather not have the heat exchanger condensing water out of the air when heat load from machines is minimal. I need to maintain a specific temperature range +-20F. Too cold and the machines will not warm up, too hot and they will overheat. What are the implications of this? I would install a bypass on the process side so that when the globe valve closes it will divert the water so the full flow is always dumping back into the 250g tank.

 

[SwinnyGG]

I have a hard time seeing why this arrangement would not work. Seems very reasonable.

Things to look out for- make sure your selection of controls components results in a system that is reliable enough; you don't want a failure to result in no cooling water flow.

You may want to investigate a spring return valve or some other NO arrangement so that your system defaults to full cooling flow in the event of a failure. (based on the assumption that running your machines too cold is better than allowing them to overheat)

You may want to investigate flow smoothing or other in-line features in your lines, so that discrete valve positions don't result in wildly different effective cooling due to cavitation or flow disturbances caused by the valve being partially open.

 

[georgeverghese]

Your description is still a little vague - post us a clear sketch of this system with notes on how you think it should be controlled.

 

[IRstuff]

I don't see how controlling the coolant flow changes what happens at the chiller. It seems to me that you want to throttle the chiller, not the coolant flow.

TTFN (ta ta for now)
I can do absolutely anything. I'm an expert!

https://www.youtube.com/watch?v=BKorP55Aqvg

faq731-376 forum1529 Entire Forum list

http://www.eng-tips.com/forumlist.cfm

 

[EngrPaper]

Sounds very similar to what we do with heat transfer oil. We use a temperature transmitter to control a three-way diverting valve to maintain process temperature on the zone and then just send back the excess flow.

 

[BronYrAur]

Sounds to me like you are concerned about the cold heat exchanger surface condensing water from the air. The heat exchanger should be insulated to prevent this. Otherwise reset your water temp if the process allows. I would only throttle the chilled water side - not both. If you are only bypassing the heat exchanger on the process side and dumping to the tank, I don't see a real value to a 3-way there. You are still pumping the full process flow. A 2-way on the chilled water side is the way to go provided you aren't starving the chiller. We need to know a little more.

 

[MintJulep]

Trying to control both is a good way to ensure that you have constant control oscillations.

 

[GT-EGR]

The chilled water side is the side you should have the control valve on - to maintain the temperature of your process loop. Then all your process water side operational control is just based on system flow demands, and independent of direct temperature functionality.

It sounds like you are having a semi valid concern about having to select a valve that is very oversized for your normal operation. Without getting into your exact single valve selection, and how much turndown you really need, a typical solution to this is installing a 1/3 2/3 control valve arrangement on the chilled water side. It just takes a little more advanced control as far as when to stage each valve on/off - but it will then give you a much larger range of control, since you won’t be constantly using an oversized valve, and if your calcs are correct it will just be the 1/3 valve operating the majority of the time.

As far as condensing water out of the air, you need to insulate the heat exchanger, which is always recommended on any chilled water system component.

 

[sarahrad]

To some degree you can throttle plate and frame exchangers to control the duty. They are designed for a minimum wall shear-stress value to achieve a high heat transfer coefficient and prevent fouling. Plate and frame exchangers are not designed with fouling factors nor do they allow for much fouling due to the small channels and gaps between plates. API 662 Part 1 recommends 50 Pa for minimum wall shear-stress and most vendors will have their own minimum wall shear-stress values that ensure they have a "clean" efficient design. When the flow is throttled to reduce the duty (on either side), this directly impacts the wall shear-stress and the reliability of the performance reduces. Since you have three machines, ideally three plate and frame exchangers would be best for the range of operation specified but two exchangers would also work. This way each exchanger will be sized and operated for the precise range required. Another advantage of have multiple plate and frames is you have the flexibility to clean and maintenance one exchanger and keep the other on-line without needed to take the machines off-line as long as you are not needing 100% capacity.