37 replies to this topic

[kkala]

Having read the whole thread, remarks (representing personal opinion) are as follows.
1. Request to clarify rationale, on API 2000 excessive thermal inbriething, got a "managerial" reply. It would probably have no response, if addressed to a local organization here. A more managerial answer ("look into the issue, solve it and inform me ") by a local small manager was once witnessed by me.
API could write the rationale of changes on every new edition, and this would be beneficial for them as well as their readers.Probably they write a brief history of changes implemented in the first page of some recommended practices? This has some resemblance to rationale, I remember ASME code has written so in the acceptance test procedure for boilers ( mentioning old PTC 4.1).
2. Cooperation of USA and Europe for practices seems to result in products probably more complete, but also more complicated. For example API RP 500, API RP 505, IP 15, concerning ATEX - Hazardous areas; or API 2000, ISO 28300, EN 14015 concerning tank breathing. Simplicity and clarity should not be disregarded in documents for engineering.
3. API issue recommends practices, needing engineering judgment to implement. The requirements for higher thermal inbreathing are undoubtedly applicable to new tanks. I suppose it depends on our judgment to apply them on existing tanks, or decide when. Probably a tank to the old API 2000 would face no problems during its life, but what about an accident due to breathing? Inspectors can support then that compliance to new API 2000 edition had been necessary. On the other hand such an accident seems remote, according to today's historical data.
4. Contrary to assumed situation as above for API, a practice established by law (e.g decree, directive) has to define whether the old one is still in force, as well as when the new practice is mandatory for existing equipment too. So situation is more clear (but API can supply more flexibility / freedom for the proper way).
5. API do not officially consider themselves accountable for omissions in previous editions. They could support that those editions complied with human knowledge and experience and acceptable safety level on the time of issue. After all "use your judgment". Even codes are developed and change in the course of time.
6. Nevertheless mentioned rationale seems to comply with the "free" spirit of API. May it be included in their future editions. This subject can be a chance to remind API organization of it. Hopefully I am not too optimistic.

Edited by kkala, 05 September 2011 - 12:00 AM.

Hi Guys,
For your information attached I send you a presentaion about "Differences Between API-2000 5th and 6th Editions".

Regards

Attached Files

•  API 2000 5th vs 6th Final.ppt   1.43MB   462 downloads

[Pete74]

Hi there,
I have just come across this post and found it very interesting. I have a situation where I need to provide nitrogen blanketing on 6 very large tanks (5000m3 each) at a refinery.The difference in the new API2000 edition for table A.3 and the equation based approach is concerning (787Nm3/h vs 2525Nm3/h).
The issue for me isn't sizing the vacuum vent valve for the new inbreathing rates as I believe this can be designed for.
However does the the nitrogen capacity need to be sized for this rate. Correctly me if I wrong but I always believed API2000 was to size the tank vacuum/vent size. Nitrogen capacity for the tank could be sized at a reduced rate as long as some oxygen was allowed in the tank. For example ensuring the oxygen content was below 5% for the emergency thermal inbreathing case would normally suffice?
For information Appendix F in API2000 does allow a 50% reduction in nitrogen requirement (1263Nm3/h) for level 3 inert protection (no flame arrestor), however this still gives me a larger nitrogen rate than the table which still seems too high to me.

Any thoughts?.



Thanks
Pete

[ankur2061]

Pete,

I do believe that the API committte has made a hash out of API STD 2000 not withstanding my admiration for this venerable institution which has been in the forefront in establishing standards for design, engineering, operations and safety in the petroleum industry.

API STD 2000, 2009 has created ambiguity and confusion regarding thermal rates by leaving the choice of selecting the thermal rates based on Section 4.3.2.3 or Annex A (Table A.3 and A.4) to the reader without being specific when and where these two different methods are applicable.

All along my argument has been that were the tanks designed for thermal rates as per the earlier API 2000 (now appearing in Annex A of the standard) incorrectly designed and what are the recorded failures of tanks designed for thermal rates based on the Tables 2A and 2B in the older API 2000 and now appearing as Table A.2 & A.3 in the newer API 2000.

This is where I believe that the API 2000 committee has left too much open for interpretation by an individual. A standard is not supposed to leave important design aspects open-ended.

Regards,
Ankur.

[Pete74]

Hi Ankur,
thanks for your response. For my project I am going to clearly state in my design basis that the breathing requirements on the tanks were based on Annex A of API2000 and the tankage vents will be sized on this basis.
However I am interested in your opinion regarding nitrogen capacity. Do you believe these rates should also be applied to the nitrogen capacity. Using these rates can lead to extremely large nitrogen requirements for a refinery. I have heard that these rates shouldn't be used for sizing the nitrogen capacity and reduced rates can be used for nitrogen but I have no idea of by what factor?
Can you provide any guidance or experience of this?

Thanks
Pete

[ankur2061]

Pete,

I think you can explore the Annex F of API STD 2000 for inert gas blanketing requirements based on the level of flashback protection. Level 1 provides the lowest inert gas blanketing requirement based on a specific combination of flame arrester or in other words for a specific type of flashback protection.

Another method would be to insulate the vessel which would reduce the inert gas blanketing nitrogen by reducing the thermal inbreathing rate in case you do not want to take credit for any flashback protection by a flame arrester. A 1 inch insulation thickness shoud dramatically reduce the thermal inbreathing for the vessel. You can use the spreadsheet available at the link below where there is an option to provide insulation to the tank. You may check out the thermal inbreathing for the tank without insulation and with insulation. I can assure you that by providing insulation to the tank, the thermal inbreathing rate will reduce dramatically.

http://webwormcpt.bl...estimation.html

Regards,
Ankur.

[Pete74]

Hi Ankur,
thanks for your suggestion. I see that Annex F does have reduction factors for inert gas blanketing, however these appear only to be relevant to the new equation based approach for calculating inbreathing and not for Annex A table based method which I was going to use. I would need to use level 1 or 2 protection (flame arrestor) in order to get below the Annex A table based method (787 Nm3/h).
Or do you believe these level factors (0.1, 0.2, 0.5) can be used in combination with the Annex A table. For level 3 protection (no flame arrestor) multiplying the table rate by 0.5 for nitrogen blanketing requirement (787Nm3/h x 0.5).
Alternatively as you suggested I could provide insulation
but I believe that this is also only given for the equation based method and not for the Annex A tables?

Thanks
Pete

[ankur2061]

Pete,

Whatever I mentioned in my last post has nothing to do with the Annex A of API 2000. Annex A is what was there in the main body of the older vesion of API 2000. Do not try to mix and match the Annex A thermal venting values with the blanketing requirement calculations in Annex F. The Annex F calculations are based on the calculations given in the main body i.e. Section 4.3.2.3 of API 2000.

Similarly the insulation credit is only applicable for the method given as per Section 4.3.2.3 of API 2000 and not for the thermal venting tables given in Annex A.

Regards,
Ankur.

[Pete74]

Hi Ankur,
thanks for your response. I thought so too, it was more me being hopeful I could use API2000 to bring down my nitrogen capacity to the tanks.
I still believe API2000 rates should only be used for sizing the tankage vaccum vents and that the Annex A rates are still applicable (as mentioned in Shell powerpoint presentation). However for system nitrogen capacity Annex A in API2000 still seems overly conservative. If it is acceptable to in-breath a certain amount of air through the breather valve (no product purity issues)then I believe nitrogen capacity can be significantly reduced using a combination of nitrogen purge with air inbreathing to ensure the oxygen content remains below 5vol%.

Thanks again
Pete

[vne]

Ankur,

 

Regarding your article at http://webwormcpt.bl...stimation.html, I have a question as follows:

 

The following is quoted from your article:

Inbreathing

Emptying (pump-out) and ambient cooling will lead to normal inbreathing. General equation to determine inbreathing flow :

 

V_in,air = V_pe + CV_tk^0.7 R_i

 

where

V_in,air = Total inbreathing in Nm3/h (Air)

V_pe = Pump-out or emptying in m3/h

V_tk = Tank capacity in m3

 

 

R_i = Insulation reduction factor 
C = Factor subject to vapor pressure, average temperature and latitude (see Table 1)

 

In your formular, Pump-out rate (Vpe) is in m3/h whereas thermal inbreathing and total inbreathing are in Nm3/h. In my opinion, Vpe should be converted from normal air (or blanketing gas) condition to normal condition. Please advise if this approach is correct. 

[breizh]

vne,

 

Consider this resource , even not the last updated , but should solve your query.

 

Breizh

[ankur2061]

vne,

 

Your approach is not correct. The equation correlates liquid volume flow rate in m3/h directly to air volume flow rate in Nm3/h or in other words the volume of liquid displaced (removed) by pump-out is replenished by the same volume of air. This is simply because the liquid temperature and the air that replaces the liquid will never be at the same temperature and the reference conditions for air flow or air displacement are always defined at "Normal" conditions of 0°C and 1 atmosphere absolute (1.01325 bara).

 

Hope this helps.

 

Regards,

Ankur.  

[vne]

breizh and ankur,

 

Thank you for your reply. I've check with the Protego software, the result shows that there is no need to convert the pump out rate to normal condition.