Shell and tube HEX U and flow rate 6

[Renoyd]

I am working on a project involving an old gas plant built in 1970's. The UA's of the Shell and tube exchangers need to be determined. The only UA data availabe are from the plant performance test conducted a year ago. Now the plant inlet flow will be reduced significantly. Can anyone help to determine the relationship between the UA and the flow rate? Thanks!

 

[pmover]

fyi, only the U will change for varying flow rates, not the A (area).

i do not have my GPSA databook with me, when i get home, i will investigate. if you have one, start looking.

also, you could:
- contact a heat exchgr mfg and inquiry with them - speak with the engineer and not the sales folks.
- search the wonderful web world (plenty of useful sites)

meanwhile, good luck!
-pmover

 

[djack77494]

If this is anything other than a very small extrapolation from existing operations, I would suggest that you run (or pay a consultant to run) some professional quality software that can adequately model your exchanger's performance. The "standard" in my opinion is HTRI.
Doug

 

[Renoyd]

Pmover,

I didn't find much content about the relationship between U and flow rate in GPSA section 9 although U is discussed. If you find any useful information, please let me know. Guess this is more of an empirical issue?

Doug,

HTRI definately has authority in heat transfer issues. We just haven't decided if we need to go that further. Thank you for the suggestions.

 

[vpl]

Renoyd

U is a function of the heat transfer rate (Q) which in turn is a function of the flow rate.

Any heat transfer book has the equations; there is a free one on the web that has been referenced on this forum multiple times. I would recommend "Heat Transfer - Professional Version" by Thomas ONLY because in the Appendicies it has a step-by-step process for figuring out the heat transfer rate of a shell & tube heat exchanger.

Having said that, the basic formulae are:

Q=m[sub]dot[/sub][Δ]T and
UA= Q/LMTD where LMTD is ([Δ]T1 - [Δ]T2)/ln([Δ]T1/[Δ]T2)
and [Δ]T1 = T[sub]h,i[/sub] - T[sub]c,o[/sub] and [Δ]T2 = T[sub]h,o[/sub] - T[sub]c,i[/sub]


Patricia Lougheed

Please see FAQ731-376: Eng-Tips.com Forum Policies for tips on how to make the best use of the Eng-Tips Forums.

 

[chusker15]

A good rule of thumb is found in Lieberman's book, "Process Design for Reliable Operations" he brings up the relationship that U=v^0.7. U is in Btu/hr/ft2/F, v is ft/s.

I agree with djack, if it's a large change in velocity, I'd use a detailed rating program like Aspen's HTFS.

 

[pmover]

Renoyd,

i keep equipment/process data from various sources in my GPSA databook, in addition to the GPSA databook documentation (something i learned from a saavy, seasoned engr). at one time, i remember obtaining good documentation relating to your question, but unfortunately, i was not able to locate the documentation as i remembered. naturally, thy brain cells are a bit rusty too . . .

the other postings are useful in conducting your search. remember, when inspecting the equation for U, it boils down to the convective coefficients; thus flow and fluid properties.

something to ponder, conducting a test on exchanger at different flow rates may yield satisfactory results.

sorry, good luck!
-pmover

 

[dcasto]

Fig 9.9 in GPSA gives you some typical U's . If you have the data from the test, you can use the Fig 9-10 to re approximate the U's based on the different variables, like viscosity, mass rate, Cp.

Other than that, do it by hand or use a software program that'l calculate new U's. I assume you have area or can calculate that from U1A's or data sheet.

 

[pmover]

Renoyd,

a response from a heat xsfer specialist:
"Unfortunately there are no thumb rules that will address the problem.

The simplest and accurate way would be to have someone run the problem in HTRI and check if the exchanger would work under the new conditions. A TEMA Data sheet for which this exchanger was designed and the new operating data will be required to check the performance of the exchanger for the revised conditions."

hope this clarifies the situation, which is what other responses lead to.

good luck!
-pmover

 

[Renoyd]

Thanks a lot, Guys. Your inputs are of big help in different aspects.

chusker15,

I am very interested in the formular proposed (I will try to get the material and do some study):

U=v^0.7

So I can put it this way:

U=Ut*(V/Vt)^0.7

(Ut & Vt are test data which we have)

Then we can get the U's under different flow rates.

I will try to input it into the model to calculate the UA's of the HEX's with varying flow rates and thus the duties. Or we will use some software as introduced. Hope this works.

 

[pmover]

Renoyd,

perhaps this website may be of assistance to you. I would highly recommend comparing calcs of a known set of design conditions to the results this site returns.

http://www.freecalc.com/hxfram.htm

hope this helps!
-pmover

 

[25362]

The 0.7 exponent for gas flow convection coefficients is an approximation for fully turbulent flows, it may smoothly drop to about 0.5 in the laminar flow region. A drop in linear velocities may also affect the fouling factors.

Convection heat transfer coefficients for liquids are also much influenced by changes in the Prandtl number. An example taken from J.P. Holman's Heat Transfer McGraw-Hill for a machine oil of 20 cSt at 100[sup]o[/sup]C: Pr at 20[sup]o[/sup]C: 10,400; at 120[sup]o[/sup]C, 175.