Performance of exchangers in Preheat train

[lgmarti]

I was given the task of identifying which exchangers in the crude unit preheat train where fouling. I was lucky becuase I have plenty of thermocouples and flowmeters to calculate the U coefficient of each bank of exchangers. There were 3 banks of exchangers where the U coefficient had deteriorated over the past 3 years. Well, we shut them down and pulled them. Voila! THEY WERE NOT DIRTY! When they were put back in service, the inlet temperature to the furnace only improved by 5 degrees. Where did I go wrong? Q=UA(LMTD). I thought that solving for U and monitoring Uactual/Udesign was enough to identify fouling exchangers. I even simulated the exchangers using HETRAN. The film coefficients did not appear to be the limiting factor...However, they were very closed to the fouling contribution to the resistance. Please help.

 

[25362]

I am given to understand that the deterioration was indeed progressive over a three years lapse albeit no change on type of crude, thermal loads, fluid rates and temperatures of hot fluids have taken place. That is, no process change has occurred in the last three years, and the other units in the pre-heat train perform as expected.

I also understand that no pressure gages were installed to verify any change in friction drops.

When pulling the tube sheets have you looked to the baffles and other H/E sections, besides looking for fouling ?

I may surmise that internal one-sided leaks may be happening: across the head baffles of the tube side, or across the baffles in the shell side.
Leaks at the pass-partitions may be caused by bending of these due to differential pressures, by erosion or corrosion or by permanently or "temporarily" bowed tubesheets due to thermal stresses. Other leaks may be appearing around the tube expansions. In short: leakage and maldistribution of fluids.

Questions may arise on whether there is some pre-flashing taking place that would show a smaller temperature change on the crude side. Of course, the warm side heat balance, should help in showing whether the transferred heat justifies this assumption. If there were inerts accumulating on one of the sides, the h/e would suffer.

As you see, an in depth analysis will, without doubt, help you to find the reasons for the suspected H/T deterioration.

I hope the ideas above may assist you in finding a cure for the plant's ailments.

Good luck!

 

[lgmarti]

25632, Thank you for you input. Becuase of the outcome, it is obvious that I have done several things wrong. I am trying to learn as I go.

We run a cocktail of crudes. Sometimes we blend up to 5 different crudes. The mixture API varies between 17-30 (the average is 21). I have noticed that on some days heat transfer is better than on others. But, I have not been able to say "when the API is 23, the heater inlet temperature is 520F vs 490F when the API is 17". (We base everything on the heater inlet temperature and we have not seen 520F since 2001).

The previous engineer kept a U coefficient graph of all the exchangers in the preheat train. Some of the ones we open showed a U coefficient decrease over time. Some didn't.

We do not perform a periodic pressure survey. However, we have indication between the crude feed pump and the desalters; and, from the desalters to the heater inlet. I just checked all the data, sorted out by rate and API and did not noticed an increase in pressure drop. However, the heater inlet pressure has gone up by 40 psig.

I have never heard of leaks at the pass-partitions so I did not look for them. I don't know if inspections or the fix equipment engineer checked for this.

I am not sure I understand how to check for pre-flashing. Please explain in more detail.

Again, thank you.

 

[25362]

Look at the operating pressures and the crude compositions to estimate whether prevailing temperatures could justify initial vaporization.