4 replies to this topic

[sandeepk2022]

Hi everyone, I am looking for some guidance and industry experience regarding an issue we are facing with a depropaniser reboiler condensate return system. We have a depropaniser reboiler where LP steam is used as the heating medium. Steam flow to the reboiler is controlled through column temperature control. The condensate generated in the exchanger drains to a condensate pot, and from the condensate pot it is routed to the LP condensate header. The bottom of the reboiler is located above the condensate pot, and the condensate level in the pot is controlled through an LC valve.

We suspect that the exchanger is over-surfaced, which results in very low steam chest pressure inside the exchanger. Because of the low pressure, the condensate pressure is not sufficient to push the condensate to the LP condensate header. Due to this, the exchanger starts getting flooded with condensate until enough pressure builds up to push the condensate out. In our case the condensate needs about 1.5 kg/cm²(g) pressure to move to the header, while the condensate temperature is around 70°C. This intermittent discharge of condensate is creating severe hammering issues in the system.

There was no equalization line between the exchanger and the condensate pot in the original system. We tried installing a temporary equalization connection, but this did not resolve the problem. As a trial, we injected nitrogen (a non-condensable gas) into the exchanger and pressurized the condensate pot to around 1.5 kg/cm²(g). After doing this the system started working smoothly, condensate was flowing continuously to the header, the exchanger did not flood, and the hammering problem disappeared.

Based on this observation there is now a proposal to install a permanent nitrogen connection with a PCV to maintain around 1.5 kg/cm²(g) pressure in the condensate system. However, I am not very confident about implementing this as a long-term solution because I could not find much literature or industry references supporting the practice of injecting nitrogen into a steam condensate system. I am concerned about potential issues such as accumulation of non-condensable gases, possible impact on heat transfer in the exchanger, steam trap performance, two-phase flow instability, or other operational risks over the long term.

Has anyone experienced a similar situation or used nitrogen injection to maintain pressure in a condensate return system? Is this something that is practiced in industry, or would it generally be considered only a temporary workaround? Any insights, experience, or references would be greatly appreciated.

[katmar]

It sounds like your reboiler is seriously over-sized. Injecting nitrogen into the reboiler will decrease the heat transfer coefficient and this means that you need to increase the temperature difference by increasing the steam pressure.  Basically what you are doing is varying the overall heat transfer capability of the reboiler by varying the heat transfer coefficient.

 

I have not seen it done this way but it is fairly common to vary the heat transfer capability of a reboiler by varying the heat transfer surface area.  This is done by varying the level of the condensate in the reboiler itself, rather than in the condensate pot.  It would require a more complicated control setup with the condensate level cascaded to the column temperature to control the steam flow. I'm sure you will find many references to this method, but here is one to get you started.

[breizh]

Hi,

What about using a condensate pump to transfer condensate from column pot to condensate header?

Pressure Powered Pumps for condensate | Thailand | Spirax Sarco

Good also to share a PID .

Breizh 

[Pilesar]

The reason the chest pressure is low is the exchanger is more efficient than required. A depropanizer reboiler will likely build up a fouling layer over time which will eliminate the problem. Even when tubes are clean, this is usually self-correcting for process control like you have witnessed. The condensate system downstream is the annoyance. The best fix depends on cost. I have recommended N2 addition for reboiler in a similar situation also as something to try since it was incredibly cheap to apply. I did not see my recommendation implemented and am pleased it is working in your case. I have not seen this done in real life but it seemed to me like an elegant solution also. Just a small N2 tubing connection with a manual valve was my idea. Do you need a continual flow of N2? I thought perhaps just a shot of N2 once or twice a day would be adequate if there were no automatic inerts bleed. Or leave a very small manual N2 bleed in place if necessary that the operators could tune. You have quite a range of acceptable total steam pressures. By adding N2, the partial pressure of the steam would be the same since that is controlled by the existing control system. You are only affecting the total pressure.

Ideas I considered at the time in no particular order in other circumstances:

1) add N2 manually as needed (see above).

2) intentionally foul the exchanger tubes some way.

3) reduce surface area by physically cutting and plugging many of the tubes.

4) putting condensate pot at the elevation of the exchanger bundle (for horizontal exchanger only).

5) pressure powered pump using steam pressure to remove the condensate. 

6) putting a gooseneck in the condensate piping so that the apex of the pipe before the condensate pot is about mid-way of the exchanger bundle to force a permanent condensate level. I did implement this one for a vertical exchanger. 

   See the sketch below. I had several cross pieces at different elevations with valves so that the condensate level could be tuned.

   I did equalize the top of the loop to the exchanger pressure. My idea was that if the bottom valve were opened, the exchanger would be at maximum surface area. Closing the bottom valve would force the condensate to rise in the loop as well as in the exchanger. With no valve at the top of the loop, there would be no way to block in the condensate completely.

   This was put into the field, but I don't think the operators ever closed any of the valves to try it. I suspect they just laughed at the crazy engineer who would design condensate piping like this.

      ------   

     /       \

    |         |

    |---X-- |

    |         |

  -----X------> to condensate system

[Pilesar]

The steam/N2 mix inside the exchanger flows to the tube surface. The steam condenses which leaves the N2 portion at the surface of the tube. More steam/N2 mix rushes in to fill the empty space left by the condensing steam. The vapor flow is always toward the tube. The result is the N2 tends to collect at the tube surface in greater proportion than in the rest of the exchanger and blanket the tubes. Controlling the quantity of N2 needed might be tricky. Adding inerts may be much easier than removing inerts. My preference would be to experiment to see how quickly the pressure inside the steam chest changes after adding N2. If the pressure stays high for some time, manual N2 addition may be adequate and easier to control.