8 replies to this topic

[jrtailor09]

Dear All,


Currently, I am working on proposal of Ammonia flare system project for one of the client based in GCC. I have an experience on the HC flare system.

There are 2 flare header coming to the ammonia flare system as below.
Refer attached sketch for easy understanding.

• PSV discharge having liquid + gas stream
• PSV discharge having only Gas stream

The PSV discharge having liquid + gaseous stream will be transferred to the Flash tank for separation of liquid & gas stream. The separated liquid stream will be sent to ammonia storage tank & Gas stream connect with the main gas flare header & transferred to Ammonia KOD.

I have following question regarding the system, may u consider silly also

• Why water seal drum is not considered to in this design? HC flare system have a separate KOD & water seal drum for adequate protection against flash back form the flare tip while Ammonia have only KOD integral with flare stack.
• What is reason for providing the steam tracing to the flash tank?

Attached Files

•  Sketch.xls   31KB   126 downloads

[ankur2061]

Jatin,

Some helpful links for ammonia flares are provided below:

http://www.eng-tips.....cfm?qid=186352

http://www.zeeco.com...s_Project_3.pdf

http://www.gasflare..../Flare_Type.pdf

Hope this helps.

Regards,
Ankur

[kkala]

• Operating pressure just upstream flare tip is close to 0 barg in the majority of flares, so liquid ammonia temperature would approach –33 oC at that point. This would create the risk of ice formation on any seal drum and downstream piping, potentially blocking the piping. Seal drum (if any) would have to be filled with (ethylene glycol) antifreeze solution instead of water, see similar LPG case in http://www.cheresour..._1#entry51916"]http://www.cheresour...h__1#entry51916, post by snalawi on 10 Oct 2011.
  Water seal drum is not judged mandatory. Its flash back protection can be (partially) replaced by the molecular seal placed in the flare stack upstream of the tip. Purge gas flow is also anticipated.
• A reason to install heat tracing in the flash drum could be "help to evacuate it from liquid" when the installation shuts down for maintenance. Pump would hardly suck liquid below drum LLLL due to risk of vapor ingression, or cavitation.
  But is it certain that only flash drum is heat traced? If all piping and flash drum to flare is heat traced, reason could be the fact that outside humidity would get frozen during a discharge, which could block some valves. I have seen this freezing on ammonia pipes continuously operating in a fertilizer plant, even though they were not heat traced at that time (late 1970s). As if I remember operators closing and opening the small valves (numerous in parallel, no problem if one closed at a time) every now and again, to avoid block by freezing.
  Comments on heat tracing usefulness and additional explanations on the intent of installing it would be welcomed.

Edited by kkala, 26 November 2011 - 05:02 AM.

[jrtailor09]

Dear Ankrurji ,

Thank you very much for your useful link for flare system.

Article by ZEECO for ammonia flare system is very good & intresting.

Regards,

[jrtailor09]

Dear Kkala,

Thank you very much for spending your expensive time for above reply.

Our minimum outlet temperature for the liquid ammonia is 6 °C only. We are using aqueous ammonia only for the seal drum not Ethylene glycol. They are operating current flare system with this only as discharge of KOD will be send to the waste ammonia collection tank.

I am agreeing that water seal drum is provided for flash back protection only to maintain back pressure of 100-300 mm wc height. Same can be implemented to HC flare system also but I didn’t come across any HC flare system without water seal drum but in ammonia & acid flare found that vendor is not proposing the water seal drum.

Why only flash tank is steam traced? Why not all lines are coming to the flash tank? I had asked all reputed vendor of flare system vendor for one of my previous project but none of the vendor has given proper reply.

[kkala]

Minimum temperature = 6 oC is noted, corresponding to an anhydrous NH3 pressure of 4.3 barg. Pressure is anticipated to get close to 0 Barg, as fluid approaches flare tip. Probably I miss something.
Does "liquid" concern aqueous NH3 discharged to flash tank? "Flash tank" and "ammonia storage tank" do not favor this assumption.
A reason against water seal drum can be freezing temperature, also met in LPG flares (indicated in web reference, post by snalawi). But 6 oC minimum temperature would not freeze water, nor external moisture, so water seal drum could be installed and heat tracing on lines is not needed (unless ambient temperature can be quite low).
If there were aqueous NH3 into the flash tank (unlikely), a reason for heat tracing it could be to heat its content before maintenance, to promote NH3(g) removal.
Vendor, if willing, could clarify points. We usually describe our understanding and ask vendor to comment or advise, so that "escape" may not be easy.

[ankur2061]

Jatin,

Some additional web resources on ammonia flares available at the following links:

http://www.usbcd.com...80413/8387.html

http://www.purgit.co...-flare-service/

http://www.nao.com/ammonia_flares.htm

Regards,
Ankur.

[jrtailor09]

Dear all,

Please find below reason for my queries for your information & reference.

) When liquid ammonia is discharged to a lower pressure environment (like a flare header), it will rapidly vaporize, absorbing heat from that environment. The evaporating temperature can easily drop to -40 degrees, which would freeze any water seal rapidly potentially blocking the flare system.
2) Steam tracing provides a heat source to vaporize any liquid ammonia that results in the flash tank and remains after the relieving event stops.

[kkala]

Consequently the ammonia to flare is anhydrous, ending to -33 oC at atmospheric pressure (close to it in flare header). Previous posts referring to anhydrous ammonia are valid, vendor does not consider pipe heat tracing necessary (only in flash tank). As said before, liq NH3 lines in the fertilizer plant did not have heat tracing (but valves were opened / closed from time to time).
Note: -40 oC was read as impact test temperature for atmospheric refrigerated NH3 tanks, see http://www.cheresour...a-storage-tank/, attachment to post by jrtailor09, para 3.4 (Design and materials of construction).