Stumped with Plate Chiller

[t_bosak@yahoo.com]

I’m having trouble cooling a 5 gallon batch of wort down to pitching temperature with my new plate chiller. The following data is provided to help with any suggestions.



I use a 40 plate welded heat exchanger 7.5” wide x 4” deep x 3” thick front to back.

Inlet water temperature last use was 55 degrees F and is run through a XL type immersion chiller with ½” diameter copper tubing sitting in 3-4 gallons of ice water. The discharge pressure at the hose bib is 60 PSI and the pressure drop is 20’ of 5/8” garden hose and then the copper tube chiller and another 20’ of 5/8” garden hose. The garden hose is not straight but also not kinked. The plate chiller discharge pressure drop is through 15’ of 5/8” garden hose and maybe 2-3 feet of elevation rise.



The heat exchanger inlet sits about 8” below the boil kettle outlet and I use gravity to move the wort through it. I use a ½” ID tube about 36” long between the kettle and heat exchanger. With the valve wide open I get a pretty good flow. The chiller is cleaned well in both directions after use, so I’m confident it is not clogged.



I have used this chiller 5 times now and each time I vary both the flow of wort and the flow rate of the cooling water, but can not find the correct combination of flows. If I remember my engineering heat transfer class, counter flow surface temperature means everything, so I think if the spaces between the plates are “flooded” then I have maximum surface area contact, but have tried all combinations of flow with no luck.



During the most recent use I was only able to get the temperature down to an average of 95 degrees. I must be doing something wrong, I based on everything I read I should be able to get my pate discharge temperature down below 70 degrees F.



With a mechanical engineering back ground, I’m embarrassed to not be able to figure this out. Any help is appreciated.



Sincerely

Tom from Pgh PA

[ajdelange]

This may seem like a silly question but are you sure you are operating in "counterflow" mode? I ask that because I recall a similar question from this forum (I think) where the poster included a photograph of the setup and it was clear from that photo that wort and cooling water were going in the same direction.

With 55 °F input water (and counter flow set up properly) you should be able to get down within a degree or two of 55 °F if 1) The coolant flow is fast enough and 2) the wort flow is slow enough. A flow ratio of 5:1 (coolant:wort) should do this in most cases. I suggest you experiment with boiling water before you continue with wort.

[t_bosak@yahoo.com]

Yes counter flow set up was verified. I think I will turn chiller 180 degrees, so my hot side inlet in on the bottom and outlet is up top. I think this way I will ensure the space between the plates are fully flooded.

Thanks for your advice

[Kbar]

With my 30 plate I can get to 75F in one pass. I run a whirlpool loop and in 1-2 minutes can get 60F. AJ may have a point. Pics?

Log Mean Temp Diff. I too am a mech engr. BYO just did a nice article on this.

[28paws]

It might help to submerge the plate chiller in a bucket of ice water.

[LordUlrich]

as far as adjusting flow my process is, set coolant to full, and run the wort as fast as possible to hit pitching temp (usually about 75%), if I end up running the wort wide open and I am still below my target then I cut back the coolant flow (this usually happens when i get clogged with hops or something)

[Cliff]

In a water cooled injection molding machine one closes down the output valve on the coolant line to increase the thermal transfer and speed mold cycling.

Try to increase the pressure in the chiller a scosh by restricting the output

Also slow the flow Water is a phase change medium and and absorbs thermal energy differently at different temperatures. Easiest way to see this is to pop a thermometer in a pot and watch the thermometer and a clock plotting the time/Temp on a graph while the water comes to a boil. You'll see that during the last several minutes before a boil the water temp climbs painfully slowly.

[ajdelange]

In a water cooled injection molding machine one closes down the output valve on the coolant line to increase the thermal transfer and speed mold cycling.

In a counterflow chiller the amount of heat transfer depends on many things, among them the ratio of the coolant flow to wort flow rates. If you increase the coolant flow rate and/or decrease the wort flow rate the wort will exit at a temperature closer to the coolant inlet temperature. And conversely. But the relationship between "efficiency" (defined as the drop in temperature divided by the initial difference in wort and coolant temperature) is not a linear function of the ratio. A lot has to do with whether coolant and wort flow are laminar. If either flow is increased to the point where the flow is no longer laminar transfer will be effected. And even where the flow is laminar the relationship between ratio and efficiency isn't linear. Each chiller can be characterized by a parameter I call Q, which depends on the geometry of the chiller. It's units are gallons per minute (or liters per second or whatever you will). A good operating point for a chiller is where the coolant flow rate is Q and the wort flow rate Q/5. This results in 99% efficiency (e.g. wort inlet temp 200, coolant temp 50, wort outlet temp 51.5 °F) with 5 times as much water used as wort is cooled. Increasing the coolant flow beyond this point has little benefit. Reducing it to 0.2*Q results in a decrease in efficiency to about 95% (exiting wort temp 56.5 °F) in the example.

Also slow the flow Water is a phase change medium...

There is no phase change in a properly operated wort chiller.

...and absorbs thermal energy differently at different temperatures.

No it doesn't (at least not appreciably - the specific heat of water changes by less than 1 % over the entire range from freezing to boiling).

Easiest way to see this is to pop a thermometer in a pot and watch the thermometer and a clock plotting the time/Temp on a graph while the water comes to a boil. You'll see that during the last several minutes before a boil the water temp climbs painfully slowly.

That's not because of a change in specific heat but rather because the rate of heat transfer between two systems depends on the temperature difference between them. Thus as the water gets hotter less heat gets transferred to the water from the heat source while at the same time more gets transferred from the water to the surrounding air through the walls of the vessel. Also as high temperature the vapor pressure of water is approaching 1 atm and even though not yet in full ebulition, a lot of heat is being carried off in that vapor: 2270 Joules*/g. Note that this is over 4 times the energy which went into that gram to raise its temperature from just above freezing to 99 °C. Clearly, loss of latent heat through vaporization would be the major reason for the decreased temperature rate in most geometries.

Inside a wort chiller where there is no vapor (unless in turbulent flow and no place for the vapor to go if it is so it can't carry away heat) the temperature of the coolant will change 1 °C for every 4.219 Joules*/L put into it at near boiling and 1 °C for every 4.181 J/L put into it at 20 °C.


*Joule (after whom the unit of energy is named) was the son of a wealthy brewer and did much of his seminal work in a laboratory attached to his father's brewery).