Recirculating Plate Chiller Calcs

Australia & New Zealand Homebrewing Forum

Help Support Australia & New Zealand Homebrewing Forum:

argon

firmitas, utilitas, venustas
Joined
8/5/09
Messages
2,994
Reaction score
125
I like the idea of a Recirculating Wort Chiller similar to Jamil's Whirlpool Immersion Chiller, particularly the idea that the wort that has just been chilled is added back to the kettle to increase the rate at which the entire volume loses heat.

I wanted to understand (or at least have an estimation of) how long it would take to chill the entire volume to my desired temperature based on the variables of the following:
- Recirculation flow rate
- Wort Out Temperature
- Chiller Out Temperature
- Chiller Water Temperature
- Wort Volume

A couple of knowns;
1. My ground water through the chiller gets my boiling wort down to 27C
2. To achieve the most efficient cooling i run the wort through the chiller at 3L/min

So i wrote a couple of little equations and plugged them into excel. This then generates a dynamic schedule that tells me how long it would take to get the entire volume of wort to my chiller water temp.

TF1 = Temp in Kettle @ flameout
TF2 = Temp in Kettle after each chilled addition
T1 = Temp
T2 = Temp
V1 = Hot wort out of kettle
V2 = Chilled Wort out of chiller

TF1 = [V1/VF x (T1-T2)] + T2 which changes to;
TF2 = [V1/VF x (TF1-T2)] + T2 after first chilled addition

In my circumstance, with my constants, the estimated time it takes to drop 23L of the wort from boiling to approx 27C by way of the recircing chiller is approx 4 mins.

Wheras the 'passive' chiller, where the hot portion of the wort remains in the kettle, the estimated time it takes to drop the 23L of wort from boiling to approx 27C is approx 8 mins

In simplistic terms, what is happening is, that instead of chilling wort that is at a constant temp (ie sitting just below boiling), the recircing chiller is chilling wort that is actively dropping. Thereby, creating a very fast efficient chiller.

So i'm yet to test this and i expect that i'll need to stir the kettle whilst chilling to make sure that the chilled portion of wort is properly homogeneous throughout the hot portion.

I still think, in practice, it will take a little longer due to equipment heat and ambient temperature fluctuations, but it should give me a broad estimation of how long it may take. Next time i get some free time i'll do a hot water run and log actual times and temps and see how close the calculator is in application.

I've attached the excel sheet for anyone that thinks it may be interesting to have a look at. Everything is editable, just change the red cells to see their effects in regards to time and temp. Feedback on how it may be improved (or why it may not work) is appreciated. Hope someone finds it useful or interesting.

View attachment Recircing_Plate_Chiller_Calculations.xls
 

[email protected]

Well-Known Member
Joined
16/5/09
Messages
187
Reaction score
17
Argon,

At work at present so can't go into detail.

One thing missing is that heat exchangers don't maintain the same heat rejection rate with changing inlet and outlet conditions. The heat transfer is dependent on the log mean temperature difference (LMTD). What this means as the temp of the wort comes down, the amount of energy/heat transferred is reduced and the rate of cooling will reduce. It will be faster to cool the batch once through than recirculate; however, recirculating will be better at locking in hop aroma and reducing isomerisation. The best option for hop aroma for a plate would be hopback as the cooling is instant!

I will have a look at the spreadsheet later. Do you have measured temps of your water side in and out as well?
 

argon

firmitas, utilitas, venustas
Joined
8/5/09
Messages
2,994
Reaction score
125
Thanks for the info mate :icon_cheers: .

What i experience is unwavering 27C at wort out side when i've set the pump flow rate.

Not sure if the below graph shows what you've explained (LMTD), but you can see that with the lowering wort temp going into the chiller the differential does taper off.

Recircing_Plate_Chiller_Calculations.jpg

Edit: Blue being recirc and Yellow being single pass
 

proudscum

Well-Known Member
Joined
17/9/10
Messages
471
Reaction score
2
At what temp does isomerisation start to slow down stop as i am thinking of using my plate chiller as a prechiller for an immersion chiller which will recirc through a hopback into the kettle?

cheers SCum
 

donburke

Well-Known Member
Joined
18/1/09
Messages
1,395
Reaction score
107
Thanks for the info mate :icon_cheers: .

What i experience is unwavering 27C at wort out side when i've set the pump flow rate.

Not sure if the below graph shows what you've explained (LMTD), but you can see that with the lowering wort temp going into the chiller the differential does taper off.

View attachment 52082

Edit: Blue being recirc and Yellow being single pass
does your formula allow for the re-heating of the cooled wort that is returned to the kettle ?

i would say your curve should look less linear, as the cooling rate will decrease the closer you get to the temperature of the cooling water

something like this ...

Capture.JPG

85 litres wort, 22 degree tap water @ 4 litres/min

your system appears to be a sophisticated form of immersion chiller, just located outside of the kettle instead of in it
 

argon

firmitas, utilitas, venustas
Joined
8/5/09
Messages
2,994
Reaction score
125
does your formula allow for the re-heating of the cooled wort that is returned to the kettle ?

Yep, this one
TF2 = [V1/VF x (TF1-T2)] + T2

As the chilled wort is added back in (TF1) the wort in the kettle (TF2) lowers it's temp. This happens in a logarithmic nature with diminishing rate of cooling.

i would say your curve should look less linear, as the cooling rate will decrease the closer you get to the temperature of the cooling water

something like this ...

View attachment 52084

85 litres wort, 22 degree tap water @ 4 litres/min

your system appears to be a sophisticated form of immersion chiller, just located outside of the kettle instead of in it
Good to see some actual data.
Are you recirculating then? From the graph it would seem as if you are.
Is this an immersion or plate chiller?
Does the 4 litres/minute relate to water flow or wort flow?
If plate what temp does the wort come out at?
Cheers
:icon_cheers:

Edit: just checked your gallery... SS immersion chiller?

Edit Edit: holy crap nice rig :icon_drool2:
 

donburke

Well-Known Member
Joined
18/1/09
Messages
1,395
Reaction score
107
Yep, this one
TF2 = [V1/VF x (TF1-T2)] + T2

As the chilled wort is added back in (TF1) the wort in the kettle (TF2) lowers it's temp. This happens in a logarithmic nature with diminishing rate of cooling.



Good to see some actual data.
Are you recirculating then? From the graph it would seem as if you are.
Is this an immersion or plate chiller?
Does the 4 litres/minute relate to water flow or wort flow?
If plate what temp does the wort come out at?
Cheers
:icon_cheers:

Edit: just checked your gallery... SS immersion chiller?

Edit Edit: holy crap nice rig :icon_drool2:

i recirculate the wort via march pump,

my chiller is immersion 24 metres of 8mm OD 6mm ID SS coil

4 litres/min is the water flow (best i can get given internal resistance)

wort flow is ummmmm .... no idea but i guess around a similar rate

using a plate chiller and then returning the wort to the kettle is the same principle as an immersion chiller, instead of the wort being cooled when in contact with the coil and mixed, its cooled externally in the plate chiller and then mixed back in

i dont think you would have any improvement over a conventional immersion chiller, unless;

1) you get better heat transfer between the metal in the plate chiller than you do with the coil
2) you have more surface area in the plate chiller than you do with your coil

if water flow rate, surface area etc are the same then i reckon it would take the same amount of time to chill using either method

if you want to get below isomerisation temps then recirculate your wort through the plate chiller until you get to approx 80 deg then go straight to fermenter from then on

it will definitely take you longer to chill your batch if recirculating than just a single pass through the plate chiller
 

[email protected]

Well-Known Member
Joined
16/5/09
Messages
187
Reaction score
17
Argon,

Just did the calc considering the non-linear heat transfer nature and calaculate out that it will take about 17 mins to cool to 30 degC which is in line with I think Jamil's cooling performance. I think he states around 15 mins.

If you do it once through it will take you approximately 8 minutes as you have stated into fermenter.

The calculation is more simple than I thought.

First calculate the heat to be removed from the wort and then calculated the heat rejected at 10 degC steps given that you state that you have a constant 27 C outlet.

bulk wort heat accumulative
temp transferred time
degC kJ/min min
100 864 0.0
90 756 1.1
80 636 2.3
70 516 3.7
60 396 5.5
50 276 7.8
40 156 11.2
30 36 17.1

Total 17.1min
 

kieran

Well-Known Member
Joined
29/1/06
Messages
258
Reaction score
8
.. I whirlpool via Therminator, and one thing I like about it, is that most of the trub stays in the kettle.
 

argon

firmitas, utilitas, venustas
Joined
8/5/09
Messages
2,994
Reaction score
125
To clarify are you saying;

Single Pass Chill to 30C = 8 mins
Recirculation Chill to 30C = 17.1 mins

?
 

[email protected]

Well-Known Member
Joined
16/5/09
Messages
187
Reaction score
17
To clarify are you saying;

Single Pass Chill to 30C = 8 mins
Recirculation Chill to 30C = 17.1 mins

?
Yep.

It is slower to cool with recirc as you are gradually dropping the incoming temperature to the heat exchanger resulting in less heat being dumped to the cooling water.

When you run straight through you maintain a high differential between the wort and the water and are able to dump a consistantly high amount of heat.

I would do as DB said - cool to 80 C for aroma etc and then run straight through or buy a hop back.

Edit: added spreadsheet calc

View attachment plate_hx_calcs.xls
 

Latest posts

Top