Flash Steam Condenser 2

[sshep]

Friends,

All the plants at the site where I work have atmospheric condensate collection drums. The flash steam from the drum is condensed by either a CW exchanger, or a fin fan. Neither type of system seems to work particularly well as evidenced by steady flash steam plumes. One of these CW based flash condensers is planned for replacement with an alternative (as yet to be determined).

Does anyone have any recommendations or resources for best practice? For example I wonder if a direct contact system might be better- i.e. circulate water through a cooler and spray back into the atmospheric vent, properly designed of course. This would avoid the high tube wall temps that cause CW scaling.

Any help is appreciated,
sshep

 

[CJKruger]

sshep, yes, we have two systems like that.

Flashed condensate enters the drum, steam goes up towards a packed bed and water drops down. Water goes to pump and is cooled in exchangers and is returned through spray nozzles above the packed bed.

Net water leaves on level control, net steam on vessel pressure control.

Problem is we still vent a lot of steam.

 

[sshep]

CJKruger,

Bummer, I thought a properly sized system like that would be almost foolproof (compared to indirect contact). I must give you a star for the info though. Any thoughts as to why your flash steam condensing system isn't working great?

best wishes,
sshep

 

[steamdog]

Is it possible that some of your flash steam is actually live steam from failed steam traps or open bypass valves?? There may be a double benefit of reducing the steam venting and saving some money by investigating further.

 

[sshep]

Hey Steamdog,

We have a pretty good steam trap monitoring program- we basically pay a contractor to do trap monitoring full time.

The problems with cooling water exchanger based condensing systems seem easy to explain because of high tubewall temps and fouling. Fin-fan based coolers are a bit more of a mystery, but the designs have an N2 purge on the vent, and I think maybe there could be blanketing.

Is there anyway to estimate the steam flow from the plume size or other parameter? Assuming I could scaffold to get up near the vent, how can the velocity be measured? I would like to compare to flash steam estimates (using reboiler chect pressures and flow) to see what duties are involved, and to verify there are no leaking or bypassed traps.

best wishes,
sshep

 

[CJKruger]

sshep, I think the reason our systems don't work well is because both (as far as I know) were revamps to the original systems.

To save cost, changes were minimized, and the cooler duties and/or packing surface area are insuffucient.

 

[steamdog]

sshep
I actually found a document from the DOE, I can only guess that this is an estimate (see attached). To calculate the flow you will need the pressure of the steam before the trap and estimated condensate flow and then go to a flash chart.

I have used a vent condenser for this application and it worked, maybe yours is not sized right.

We have also sprayed water directly into the vent of the flash tank, but this consumes water and adds untreated water to the flow which may be unacceptable.

I do not not quite understand the nitrogen purge or why it is needed.

Joe

 

[sshep]

Hey Steamdog,

I envisioned a circulation loop of cooled (against cw)condensate for the spray, so there wouldn't be any contaminate.

Thanks for the DOE paper, which is probably not applicable to a plume (longer than the chart goes) coming from a 10" vent pipe, but it could be handy for other leaks.

As to the N2 purge, it was a mystery to me as well. I thought maybe for air intrusion in the event of a condition of no flash steam generation (shutdown, cold start conditions, etc), however, I made an inquiry to one of our start-up specialists involved with design. He told me it was to reduce a fire hazard at the vent in the case of process side of reboiler leaking into the condensate system. I am not sure I really believe that to be an effective mitigation, but you can make your own opinion.

best wishes,
sshep

 

[katmar]

Does the nitrogen purge pass through the exchanger, or is it introduced downstream? If it is passing through the exchanger then it will emerge saturated with water, and if the atmospheric temperature is slightly below the exit temperature there will be instant condensation and a very visible plume, even if it does not contain much water. Especially on a cold highveld winter morning.

Katmar Software
Engineering & Risk Analysis Software

http://katmarsoftware.com

 

[sshep]

Hey katmar,

The N2 is introduced into the vent pipe downstream of the exchanger.

best wishes,
sshep

 

[bcgord]

sshep:

We have installed a system to condense the flash steam from a mineral processing autoclave, about 150 kg/min of flash steam containing some sulphur vapour. We use two Komax Scepter mixers in a tank of water with about 6000 L/min of cooling water flow. It works well with no venting of any steam. The sulphur does slowly plug the mixers (which are basically an in-line mixer). Check it out at

http://www.komax.com/drawings/scepter_tank_heater.html

This is my first post, sorry if I've violated any conventions.