twin heat exchangers 1

[Jack Nicholson]

Dear all.
Hello.
I have a question. Would you please help me?
In attached shell and tube twin parallel heat exchanger, Natural gas (NG) enters in tubes and propane enter in shell. Although both exchangers are completely the same, but propane level in exchangers are not equal, So the PRP-out temperature are not the same too. Why???

And surprisingly although PRP level in E-A is lower than E-B, but NG-out temperature in E-A is lower. Why???

One suggestion to overcome this issue is to connect both shell drain of heat exchangers, so they are like connected drums. What is your idea?

 

[Latexman]

E-A is transferring more energy. To find out why that is, will take a deeper dive into the data and exchanger's

Good Luck,
Latexman

 

[don1980]

Gas-side heat transfer (film coefficient) is highly sensitive to turbulence (Nre). Since you're obviously getting better heat transfer in E-A, one possible explanation is that the NG side of E-A has more pressure drop (resulting in higher Nre, and better HTC) due to the friction loss in the NG supply line to E-B.

Another possibility is that you have some debris obstructing NG flow into the E-B tubes.

 

[georgeverghese]

Exit temps on HX-A are not realistic - NG exit temp should be higher than C3 vap exit temp. Check calibration of TTs'. Local TG cannot be relied on.

Level variation may be due to difference between actual density and value input at DCS for translation of HX-B dp cell readout to level. Local readout at HX-A may be the correct value(assuming this is a local LG and not a dp cell type unit), so multiply the current density at control LT in DCS by 10/7, roughly. Then level readouts at HX-B LT will be 70% also. Actual C3 density in these kettles may have further increased from simulation values due to accumulation of trace C4 (probably iC4)in feed C3. No need for level equalisation line since there is no possibility of unequal accumulation of C4 on shellside between these units.

 

[Jack Nicholson]

"Local readout at HX-A may be the correct value(assuming this is a local LG and not a dp cell type unit), so multiply the current density at control LT in DCS by 10/7, roughly."

I couldn't get your point. But I must say level of both heat exchangers are based on local gauges.

 

[georgeverghese]

Okay, so readouts are from local magnetic level guages - these are also known for producing erroneous readouts.

What does LT - LIC on HX-B show ?

 

[Jack Nicholson]

LT on HX-B shows 69%

 

[Latexman]

Attaboy, georgeverghese!

Good Luck,
Latexman

 

[georgeverghese]

Since mag LI and dp cell readouts on HX-B match, suggest mag LI on HX-A be removed and sent for maintenance or replaced at next turnaround if there is time (not a critical maintenance activity, since this is not in a process control or process safety function).