LMTD manipulation

[Roach]

I have an existing shell and tube HX with the following design values:
190 GPM both sides
Hot inlet 220ºF
Hot out 200ºF
Cold in 165ºF
Hot out 185ºF

I want my cold in and out to be 140 and 160ºF

This increases my LMTD from 31 to 48 (includes correction factor for hybrid design). At the design value I needed 1.8MMBTU/hr, and now I see that I need 2.7MMBTU. Will this work, or does the increased LMTD affect my "U" value, negating the extra BTU/hr that I add on the hot side?

Thanks

 

[25362]

From the data submitted it seems you are speaking of aqueous fluids or similar. At first sight, assuming these fluids aren't vaporizing, condensing, or undergoing a chemical transformation, an increase of 50% in heat duty, would mean a corresponding increase in both flow rates while keeping the same temperature changes on each fluid. Higher linear velocities would tend to improve the "U" value. Higher flow rates would involve spending more energy to overcome expected increased friction drops.

If you could be more specific and submit more information on the HE hybrid design and the fluids concerned, I believe you would facilitate the job of the experts in helping you.

 

[Montemayor]

Roach:

First and foremost, your data is wrong or you have a typo:

Hot inlet 220ºF
Hot out 200ºF
Cold in 165ºF
Hot out 185ºF

You must mean that the Cold out is 185 oF.

But in any case, you can't keep the flow constant (as you infer) and expect to keep the delta temperature of either stream constant while increasing the duty. The equation that holds is:

Q = W*Cp*(T2 -T1)

The LMTD has no bearing on the heat load. It is used in the equation

Q = U*A*LMTD

which is used to determine the heat transfer area.

Something is wrong in the information you furnish or you haven't expressed yourself correctly. Do you intend to increase the flowrate(s)?

25362 is correct. You need to be more specific (& accurate) in your data.

Art Montemayor
Spring, TX

 

[Roach]

MonteMayor,
To be more specific and accurate, yes this is a typo.
Cold out is 185ºF.
And, actually after further thought, I would prefer to not have to manipulate flow rates, although that was the intention originally. My apologies for any inferring.
25362,
Yes both fluids can be assumed to be water, and the Hot loop is pressurized to prevent flashing to steam.
After further review of HXs (I don't run into these problems often) I see that the "U" value and the surface area are basically fixed (assuming low fouling/proper PM), and the one number I can manipulate is LMTD. This being said, the easiest thing for me to do is decrease the hot loop temperatures respective to the desired cold temps, maintaining the design value for the LMTD. Will this work?


i.e.,
Hot in 195º
Hot out 175º
Cold in 140º
Cold out 160º

Thanks for the responses

 

[25362]

We still don't know the fluids' characteristics, such as viscosities, thermal conductivities, etc., that may influence the U at different temperatures. If we consider water on both sides, it is reasonable to assume that the new conditions, same flow rates, same duty of 1.8 MM Btu/h, will be met. Good luck.

 

[25362]

BTW, even for pure water a drop in temperatures as assumed would reduce the U value by up to 10%, this is due to a reduction of the Reynolds numbers and of the thermal conductivities, that cannot be compensated by the increase in Prandtl values.
Thus, I estimate that an increase in the temperature of the hot stream by about 3[sup]o[/sup]F would suffice. {pipe]

 

[25362]

BTW, even for pure water a drop in temperatures as assumed would reduce the U value by up to 10%, this is due to a reduction of the Reynolds numbers and of the thermal conductivities, that cannot be compensated by the increase in Prandtl values.
Thus, I estimate that an increase in the temperature of the hot stream by about 3[sup]o[/sup]F would suffice.

 

[kyong]

Roach,

If you want to fix flow rates of both sides,
to reduce hot side temperature is only way.
And as 25362 wrote, U will decrease a little.
So LMTD should be a little more than 31. 25362 is saying
around 34.
On the other hand, any way, reducing hot side flow rate will also decrease cold side temperature. In that event,
U will decrease. And temperature difference between hot side out and in will increase, which affects LMTD.

 

[tkdhwjd]

Roach,

If your exchanger is designed to pinch out (which I doubt), you will have to adjust flow or temperature of cold/hot media to achieve the desired temperatures.

Slight change in U is expected due to change in thermal properties when operating temperatures are changed. However, you are dealing with liquids (non-compressible fluid), and therefore, I think you can assume it remains the same... unless you really have to meet the target temperatures.

Cold and Hot inlet temperatures are not affected by tne exchanger, but the outlet temperatures are. Heat exchange area appears to be fixed since you are dealing with single phase flow in both sides. Therefore, you have the following options to meet your targets:

1. Retube to lower surface area
2. Throttle hot or cold medium flowrates to control Q-transfer

You will have to do some analysis to see which option is more practical and economical.

 

[tkdhwjd]

Roach,

typo...

I meant to say "unless your exchanger is designed to pinch out, you will have to manipulate flows or surface area to meet the target temperatures without changing the process temperatures."

 

[Roach]

Thanks to all for the input. I intend to decrease the inlet temperature on hot side to maintain LMTD and assume similar U value. After this, I will field adjust until I reach the requirements for the process water.

Roach