5 replies to this topic

[prasad1901]

Does anyone have any experience dealing with the design on a half pipe coil jacketed reactor?The heating medium that we have is hot oil.How do we estimate the amount of oil required to heat up the reactor?And at what speed should the pump work at to ensure the smooth flow of hot oil around the coil?Hope to hear your opinions asap.Thank you.

[Art Montemayor]

 Half_pipe_Coil_Heat_Transfer.doc   97KB   628 downloads
Prasad:

It doesn't sound like you have much experience with forced convection heat transfer, so I'll address your questions directly:

1) Yes, I've done jacketed reactor heating applications - for half-coil and full jacket.
2) You estimate the amount (actually, the flow rate) of oil circulated through the coil by calculating the amount of heat transfer it will do and the coil pressure drop you can tolerate.
3) The pump speed has nothing whatsoever to do with forced convection heat transfer. I am totally confused by this question. You can run your oil circulation pump at a variety of speeds and obtain the same heat transfer - for 60 Hertz frequency, you can have 3600, 1800, 1200, & 900 rpm normally.

A reactor jacket does not serve well as the main heatup tool. It is limited in capacity and in efficiency. All a jacket is intended to do is to offer token heat transfer - at best. A reactor size is selected on its production, not on its heat surface capabilities.

Read the attached copy of an old thread on this subject. It may help you out.

[prasad1901]

Thank you Montemayor for your prompt reply.

I am having problem finding the right correlations to estimate the flow rate of oil required. I found some information on half-pipe jacket design equations at this thread:
http://www.cheresour...l_design2.shtml.

However, it stops at the calculation of heat transfer coefficient. How do I proceed on after that?

Also, please note that my reactor is not cylindrical but somewhat like a diamond shape.

Hope to hear from you asap.thank you.

[Art Montemayor]

Prasad:

You have confirmed my opinion that you do not have much experience with forced convection heat transfer when you ask what to do with the heat transfer film coefficients.

You also failed to read the information I submitted; otherwise, you would know that the over-all heat transfer coefficient I recommend is approximately 70 Btu/Ft^2-hr-oF.

The author of the article you found on this site fails to state the obvious: his equations are theoretical. He also fails to advise you of everything else I cite in the thread I submitted.

I don't know what you mean by "diamond shape".

My strongest recommendation is to not depend on jacket heating (or cooling) as a reliable design factor for reactors. Jacket heating is nothing more than a token attempt to use what is there - and nothing more. Rarely does the jacket furnish enough surface area to justify any reasonable attempt at controlling reactor heat (or cooling) requirements. Take what is given to you and design your heating requirements based on another method - like an agitated, internal coil or an external heat exchanger.

[prasad1901]

Dear Montemayor,
Thank you for your reply. I am aware that the over-all heat transfer coefficient you recommend is approximately 70 Btu/Ft^2-hr-oF but that is for cooling water.I am using hot oil(thermal oil) as my heating medium.I came accross this link where they created a model to simulate heat transfer in several types of jackets, namely half pipe, annular jacket and dimple jacket.Is this model a good estimate?

http://www.uts.com/R...atedVessels.pdf

And, I also can't change the design as the client wants it that way.No questions asked. What I meant by diamond shape is the reactor shape.I think it will be a lil tough putting those coils around.Hope to hear from you soon.

Thank you.

[prasad1901]

Dear Sir,

The situation here is, the reactor has been designed based on optimum dimension for reaction efficiency.

The client wants us to use half pipe jacket on this reactor for heating purposes.We are using thermal oil as the heating fluid and also as the process fluid.Process fluid is to be heated to 320 deg C with the heating fluid at 360 deg C.

Due to difficulty in rolling these half pipe around the reactor, it is decided to use a square bar around the jacket.But, I'm not sure if that is possible.

However, please find attached my calculations for the flowrate required to heat up the reactor based on the lowest heating time. Please take note that the flowrate is the manipulated variable.


Hope to hear from you soon. Thank you.

Attached Files

•  Half_Pipe_Calculation.xls   58.5KB   610 downloads