Pressure drop through a conventional jacket with baffles

[daniel27ad]

Hi,

The final goal here to know the flow of chilled water that passes through the jacket of a certain tank. There are no flow meters available to obtain it though. Neither is there a path from where I can divert the flow in order to put it on recipient or something. Neither do I have any calculation memories available, or at least that's what the people who could have them have told me.

So I thought that maybe I could have the chance of calculating it based in the pressure drop through the jacket, as I happen to have the inlet and outlet pressure measurements (25 and 12 psi respectively). However, by thinking this I was assuming there is a method to calculate the pressure drop through a vessel jacket based on the flow and other information, so that I could use the method backwards and obtain the flow.

I think that I have done quite a few research on the subject and I have not been able to find any source to tell me specifically how to do it, except for this question that someone posted in another forum.

http://www.cheresources.com/invision/topic/2132-pressure-drops-through-vessel-jackets/

Any feedback on whether I am trying to figure this out the right way?

Does the solution posted in that other forum seem to have sense? Can I get any suggestions otherwise? Or can someone recommend any books
that explain how to calculate the pressure drop through a jacket?

I am still to find the internal and external diameters of the jacket. It is a conventional jacket with spiral baffles for what I´ve been told.

I appreciate the help in advance.

 

[MikeHalloran]

That's a lot of pressure drop for flow through a jacket.
Perhaps the flow is very high.
Perhaps there's a metering orifice somewhere between the pressure gages.
We can't tell from here.

We'd have a chance of estimating the flow if you transcribed the device's label, so we could look it up. ... or you could.



Mike Halloran
Pembroke Pines, FL, USA

 

[georgeverghese]

Another way of getting this chilled water flow is to do a heat balance

Heat picked up by chilling water = Heat lost by tank

Assuming that this heat transfer doesnt suffer any losses to ambient.

Are there temperature readouts for chilled water inlet and outlet lines

Would this work ?

 

[daniel27ad]

Thank you both for your suggestions.

Mike, since we are talking about tanks, I wouldn't expect there to be much info about it if I look the plate. This being that the design is usually generic, and it is usually in the documentation that the vendor provides where more custom stuff such as this might be included. In this case, I asked the department and they said they have nothing.. Now, I might be right or wrong, but there is nothing to lose if I give it a try, it won't surely steal much time for me.

George, I am trying to give a shot to your suggestion. I measured the inlet and outlet temperatures of the water using an infrared temp sensor we had available, along with some black tape in the pipe surface to get a good emissivity. I guess it is fair to say that the pipe surface is in thermal equilibrium with the fluid. Then I have the temperature at which the product usually comes into the tank, and the low limit at which cooling is removed. Talking with the operator he gave an idea of how long it takes to cool down the product to the desired temperature.

The only thing is that the cooling starts as soon as the tank starts filling. So we have the period of cooling where the tank volume is changing and then the period where it stays the same level until it reaches the temperature. I cant recall exactly how you deal something that is neither a closed system nor a control volume, which is what I think we have when it is filling. Just a matter of dusting off the Thermodynamics book probably. Still, I don't think this first period will affect much though, as it seems that the period with fixed volume is several times greater than the period with the tank filling.

 

[georgeverghese]

It would be much easier to look at things when the level is fixed in the tank, and not when it is filling.

Is this a batch cooling operation? What happens after the tank reaches a fixed level and has cooled down with the coolant running through the jacket ?

Is this tank insulated ? Where are all the other heat loss sources ? Tank roof, bottom, shell ? Do you have these quantified ?

Do you have temp readouts as a function of time of how the tank contents cool down as the coolant is flowing through the jacket in this fixed volume mode?

 

[georgeverghese]

The other complication is with temperature gradients that would exist in "stagnant tanks" - the temp reading for the tank may not be the same as the weighted mean temp for the whole tank contents.
So this approach also has its challenges.
If you cannot install or cut into this with an RO with taps anywhere on this coolant inlet or exit line for some reason, would this be of interest ?

http://www.flexim.com/en

 

[daniel27ad]

Ice cream mixture arrives the tank after being pasteurised. When coming into the tank, the temperature is about 6 C. As soon as the tank starts filling, the ice water cooling is allowed through the jacket. The readings I got for Inlet and Outlet temperature of Ice Water are 1.5 C and outlet is 3 C respectively. The tank is 4000 l.

If the tank temperature reaches 1.5 C, they will stop the Ice Water flow. They were not able to tell me how long it takes for the mixture to reach that temperature with the tank full though, as they usually start taking product out before it reaches 1.5 C. They committed to take note of this time next weekend, as it happens that on weekends the product stays for longer inside the tank.

It is not insulated, so my bad for not considering that. The jacket covers the sides and bottom of the tank, not the top, so there is another non quantified heat loss.

The answer to your last questions is yes, I do have a log of the temperature inside the tank as a function of the time.

Regarding your last post:

I do not know exactly where the temperature sensor is, but I think it is a fairly good approximation of the product temperature, even more considering that there is an agitator so I think that makes the product temperature more uniform.

I spoke to my superior about getting a clamp on ultrasonic meter already, he agreed with me in the importance of it but told me to not expect in short term.

 

[georgeverghese]

Presume you have the Cp value for this concentration of pasteurised milk at about 3-4degC? There might be half a chance of getting a stab at this if you have temp readouts of milk with time at constant level over the weekend then.
BTW, why do you need this flow information - is it because you've got unusually high dp on the coolant side? What is this high dp doing to the coolant system ?
Presume the feed / exit nozzles on this coolant jacket allow the jacket to be fully liquid flooded - exit nozzle is close to the highest elevation of the jacket? Else you would get high dp even at low coolant flow with air trapped in the jacket.