Lager Tank Cooling issue

[triple]

I am a private contractor working at a local micro brewery to implement a new Brew kettle and 7 new Lager / Ferment tanks. I built the controls and wired the complete system as well as the PLC program for controlling it.
We are using a Glycol solution for cooling and recently ran into an issue where we are getting a huge temperature difference between the top of the tank and the bottom. These tanks have separate dimple wall baffles for the top and bottom. We feed the coolant into the bottom sides of the baffles and exit at the top side where the control valve is located. One valve per tank controlling 2 separate baffles. We have a similar system in place for 15 years and it does not do the same thing as this new system.
The beer is kept at 36 degrees and the measurement is being collected by a single thermocouple that is located even with the access door. When the tank is emptied, the temperature will drop to 24 degrees. There is no ice in the tank and the freezing appears to be isolated to the top baffle. The tank is never full, maybe 180 barrels or ¾ full.
I realize there are many technical factors involved with the system but I don't believe it is a control issue (electical valve and temperature control) as much as it may be mechanical.

A point in a coupel directions would be helpful.
Signed,
Frosty

[BrewinLou]

Answered in another thread.

[dick murton]

As noted in the first response - there has been much discussion about this before. In simple terms - if fermenting, you can use a low level probe. If cooling neer to freezing point you must use a level probe near the top of the liquid. water (and beer) is densest about 4 degrees C, and above and below this temperature is lest dense and therefore rises. By using a low level probe, you are measuring the densect liquid only, not the coldest.

This is how fish and other creatures survive winter in deep water - the temperature is above freezing at the bottom of the pond / lake

You need a new high level temperature probe.

[Ted Briggs]

I have a simmular problem and hae found that crashing to 40f overnight then to 35 the next day helps.