Heat transfer problem with 50/50 water glycol solution

[vonbad]

We have a kettle (with agitator) of 700 gallons of canola oil with our target temp of 180°F. Using a 50/50 water glycol mix and low pressure (<40psi). Heat transfer is great till oil reaches 140°F (rise about 1°/min in 1-1 1/2 hours). Then it takes 3-4 hours for the oil to go from 140 to 180. At this point, the glycol is running 218 to process and basically 218 on the return side and the heaters are only running about 10%. It seems the heat transfer has basically stopped. Why? the only thing I can reason is that we are creating a vapor with the water, thus, lossing the heat transfer as the energy is being used to create the vapor.

 

[GenB]

Water boiler point under pressure is the same at 210deg only if you let it loose to boil.
Your problem may be in the flow path in within the kettle,
the speed of the liquid seems very high,
you may have to change your system to 15 psi steam,
for the required temp of your product, steam gives better results,
glycol-water mix works better for higher temps.
but it can do the job if you can control your flow, rid of air and the temp sensor are in the righ place,
you mention that the temp sensors shuttoff even if the temp is not reached!,
something is wrong with the temp sensors or the position of them, they should not trip unless the product has reached the desired temp.
GB

 

[lilliput1]

Heat transfer with jacketed kettle would not be as good as that in a shell & tube heat exchanger that you can have the fluids flow counterflow.

 

[25362]

For Heat Transfer in Jacketed Vessels read an article with this heading by Robert F. Dream, in the ChE issue of January 1999, p. 90+, with a worked example.

Formulas for the time taken for heating or cooling are given. The usual one for the case when the jacket fluid [&Delta;]T is greater that 10% of the LMTD (possible at SOR). And the other one when it is lower (EOR). The time correction factor in that example is ~ 4.1, meaning that the SOR process could be about 4 times quicker than for the EOR !


If you cannot get hold of the article let me know your fax and I'll arrange for it to be sent to you.

These formulas are based on the assumption that the value of U, the OHTC, doesn't vary.

SOR: start of run. EOR: end of run.

 

[CRG]

Break out a good heat transfer reference book and put together a calculation set that models the configuration of the kettle.

Match results with field data. These types of calculations are difficult (at least for me); however, I think you should get a result + or – 40%. Empirical data is more accurate; however, most engineers do not have access to that type of data. Once you review your data, you should be able to get a handle on whether this is a design problem or a service type problem.

To others out there who have done these types of hand calculations, what types of accuracy do you end up with without empirical data?

Lastly, if you do not have experience in heat transfer calculations, you might find the hand calculations awkward and time consuming. And in addition, once you came up with a solution, I suspect you might not trust it. If this could be the case, I recommend that you pass this off to someone in your firm with heat transfer experience.

All of the above advice given by others above seems good; however, without knowing if this kettle was designed properly to start with, it is impossible to determine if this is a repair or redesign problem.

 

[vonbad]

Thank you all for the input and possible solutions. I will should be getting some data back with regard to some of the changes that were made.