Tek-Tips is the largest IT community on the Internet today!
Members share and learn making Tek-Tips Forums the best source of peer-reviewed technical information on the Internet!
-
Congratulations cowski on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!
How do you purge a vessell? 5
- Thread starter cougarfan
- Start date
Montemayor
Chemical
dideyjohn:
Yes, in most cases it does matter. Your vessels were obviously not designed for efficient purging - or not designed with accurate specifications. No matter. You can make the purging more efficient by the judicious use of dip pipes (flanged) mounted on your top nozzles, in the form of spargers. That way you can inject your purge gas (N2?) as far away from your outlet nozzle in order to have the longest and most efficient trajectory - ensuring that it has the best "sweeping" effect inside your vessel.
In the above manner you can avoid making mechanical modifications to a stamped vessel - which is what I assume you want to avoid.
Art Montemayor
Spring, TX
Yes, in most cases it does matter. Your vessels were obviously not designed for efficient purging - or not designed with accurate specifications. No matter. You can make the purging more efficient by the judicious use of dip pipes (flanged) mounted on your top nozzles, in the form of spargers. That way you can inject your purge gas (N2?) as far away from your outlet nozzle in order to have the longest and most efficient trajectory - ensuring that it has the best "sweeping" effect inside your vessel.
In the above manner you can avoid making mechanical modifications to a stamped vessel - which is what I assume you want to avoid.
Art Montemayor
Spring, TX
dideyjohn, Montemayor is correct, it does matter. You need to look at initial start pressure/temp. and it's also important to know the vessel contents. In some cases, the contents cannot be released to atmosphere or into work areas and must be collected or flared off. There is also the problem of vessel contents and purge gas reactions. I won't go into the full Environmental and Safety aspects. As for N2, beware, it's an O2 displacer and likes to collect in low areas. If after purging, crews need to enter the vessel, this is a confined space and requires full safety/health controls over entry, work and exit.
Hope this helps.
saxon
Hope this helps.
saxon
- Thread starter
- #4
Montemayor /Saxon- Thanks to both of you for your reply. Our main reason for purging with nitrogen is to bring the oxygen level low enough so the contents added to the vessel will not combust. We typiaccly purge and relase 3 times and wonder if we are doing it correctly. Montemayor - you mentioned dip tubes for more efficenct mixing - if we do not do it that way , will we still get the same results.
Saxon - we are not planning on entering the vessel once we purge it and in most cases when we purge, there is very l9ittle material in the vessel.
Saxon - we are not planning on entering the vessel once we purge it and in most cases when we purge, there is very l9ittle material in the vessel.
-
3
- #5
Dideyjohn,
I do not see any problem with the way you are purging the vessel. What you describe seems to be "pressure purging": you add nitrogen under pressure, and after the gas has sufficiently diffused, vent it to atmosphere (or flare etc.).
Montemayor, on the other hand, describes the method called "sweep purging": the nitrogen is introduced into the vessel at one opening, and withdraw the mixture at another opening and vent to atmosphere (flare etc.).
A third method is called "vacuum purging", where you draw a vacuum (e.g. by using a vacuum pump), replace the vacuum with nitrogen, draw a vacuum again, etc.
Yet another method is "siphon purging", where you fill the vessel with liquid, and subsequently replace the liquid by nitrogen. I used this method to inert a 80,000 m3 cavern before it was used for storage of propane. Advantage was that water is much cheaper than nitrogen!
A more detailed description can be found in:
G.R. Kinsley Jr, "Properly Purge and Inert Storage Vessels", Chemical Engineering Progress, Februari 2001, pages 57-61.
Carl Branan, "Rules of Thumb for Chemical Engineers", 2nd edition, page 272-273.
P. Blakey, G. Orlando, "Using Inert Gases for Purging, Blanketing and Transfer", Chemical Engineering, May 28, 1984.
In your case, pressure purging or vacuum purging seems to me the best options. Sweep through purging consumes more nitrogen, but is commonly used when the vessel is not rated for pressure or vacuum.
You can easily calculate how many purge cycles are required. For example, if you purge Oxygen from 21 to 1% and use pressure purging to 5 atm, two cycles will do the job since (1/5)^2=0.04 and 0.04*21=0.84% Oxygen.
Please note that for hydrocarbons (e.g. propane), it is normally sufficient if you purge to 8% oxygen, since you cannot get a flammable HC/O2/N2 mixture if the O2/N2 fraction is below 8/92!
For pressure or vacuum purging it doesn't matter that all your nozzles are at the top of the vessel. There is no advantage in using dip pipes.
Good luck.
I do not see any problem with the way you are purging the vessel. What you describe seems to be "pressure purging": you add nitrogen under pressure, and after the gas has sufficiently diffused, vent it to atmosphere (or flare etc.).
Montemayor, on the other hand, describes the method called "sweep purging": the nitrogen is introduced into the vessel at one opening, and withdraw the mixture at another opening and vent to atmosphere (flare etc.).
A third method is called "vacuum purging", where you draw a vacuum (e.g. by using a vacuum pump), replace the vacuum with nitrogen, draw a vacuum again, etc.
Yet another method is "siphon purging", where you fill the vessel with liquid, and subsequently replace the liquid by nitrogen. I used this method to inert a 80,000 m3 cavern before it was used for storage of propane. Advantage was that water is much cheaper than nitrogen!
A more detailed description can be found in:
G.R. Kinsley Jr, "Properly Purge and Inert Storage Vessels", Chemical Engineering Progress, Februari 2001, pages 57-61.
Carl Branan, "Rules of Thumb for Chemical Engineers", 2nd edition, page 272-273.
P. Blakey, G. Orlando, "Using Inert Gases for Purging, Blanketing and Transfer", Chemical Engineering, May 28, 1984.
In your case, pressure purging or vacuum purging seems to me the best options. Sweep through purging consumes more nitrogen, but is commonly used when the vessel is not rated for pressure or vacuum.
You can easily calculate how many purge cycles are required. For example, if you purge Oxygen from 21 to 1% and use pressure purging to 5 atm, two cycles will do the job since (1/5)^2=0.04 and 0.04*21=0.84% Oxygen.
Please note that for hydrocarbons (e.g. propane), it is normally sufficient if you purge to 8% oxygen, since you cannot get a flammable HC/O2/N2 mixture if the O2/N2 fraction is below 8/92!
For pressure or vacuum purging it doesn't matter that all your nozzles are at the top of the vessel. There is no advantage in using dip pipes.
Good luck.
-
1
- #6
When using a pressure purging technique in which the vessel is pressurized to a given pressure with nitrogen and then depressurized back to atmospheric conditions, the number of purging cycles to reduce the oxygen concentration depends on
the final oxygen concentration yf required, the pressure obtained when pressurizing with nitrogen Ph, and the final and original pressure, supposedly atmospheric.
The number of cycles N is estimated from the formula:
yf/yo= (14.7/Ph)N
Where yo=0.21 is the first concentration at 14.7 psia.
Assuming the pressure level of the purge is 30 psig, the first purge would result in an oxygen composition of
y1=0.21[14.7/(14.7+30)]=0.069=6.9%; the second,
y1=0.069[14.7/(14.7+30)]=0.0227; the third,
y1=0.0227[14.7/(14.7+30)]=0.0075, and so forth.
If the final oxygen concentration is fixed at 1% = 10-2=0.01 mol fraction the number of purges N is taken from
(0.01)/(0.21)=[14.7/(30+14.7)]N and
N= ln(0.01/0.21)/ln(14.7/44.7)=2.73. So, 3 cycles would have to suffice.
But if the final required oxygen concentration would be, say, 10 ppm = 10-5 the result would be:
N=ln(0.00001/0.21)/ln(14.7/44.7)=8.9, or 9 purging cycles.
Please note also that the higher the "pressurizing" pressure Ph, the lower the number of cycles needed.
the final oxygen concentration yf required, the pressure obtained when pressurizing with nitrogen Ph, and the final and original pressure, supposedly atmospheric.
The number of cycles N is estimated from the formula:
yf/yo= (14.7/Ph)N
Where yo=0.21 is the first concentration at 14.7 psia.
Assuming the pressure level of the purge is 30 psig, the first purge would result in an oxygen composition of
y1=0.21[14.7/(14.7+30)]=0.069=6.9%; the second,
y1=0.069[14.7/(14.7+30)]=0.0227; the third,
y1=0.0227[14.7/(14.7+30)]=0.0075, and so forth.
If the final oxygen concentration is fixed at 1% = 10-2=0.01 mol fraction the number of purges N is taken from
(0.01)/(0.21)=[14.7/(30+14.7)]N and
N= ln(0.01/0.21)/ln(14.7/44.7)=2.73. So, 3 cycles would have to suffice.
But if the final required oxygen concentration would be, say, 10 ppm = 10-5 the result would be:
N=ln(0.00001/0.21)/ln(14.7/44.7)=8.9, or 9 purging cycles.
Please note also that the higher the "pressurizing" pressure Ph, the lower the number of cycles needed.
Another reference is the NFPA 69 on "Explosion Prevention Systems", especially chapter 5 and Annexes B, C and D are relevant.
Again, your system and procedure of pressure purging seems to be OK. You may save some nitrogen by calculating how many cycles you need to reach a safe atmosphere. Note that after pressuring the vessel with nitrogen, you need to wait long enough for the nitrogen to diffuse sufficiently.
Vacuum purging could be quicker, but then you need a vacuum pump and the vessel must of course be rated for vacuum.
Again, your system and procedure of pressure purging seems to be OK. You may save some nitrogen by calculating how many cycles you need to reach a safe atmosphere. Note that after pressuring the vessel with nitrogen, you need to wait long enough for the nitrogen to diffuse sufficiently.
Vacuum purging could be quicker, but then you need a vacuum pump and the vessel must of course be rated for vacuum.
- Thread starter
- #8
Okay , fellow helpers and experts -
I have done this without any sparge tube or line. I presurized from the top up to 10 psig , released pressure - took a sample of the gases in the kettle at the bottom, middle and top of the vessel and then did this agin for 2 more times. I came up with the following values:
1st test 2nd test 3rd test
Top: 13 8.3 5.4
Middle 13.6 8.6 5.4
Bottom 15 10 6.3
Questions: I was hoping to get below 5 % oxygen after three purges - no luck - also I noticed that the oxygen level was always higher near the bottom of the kettle... should I use a tube to force nitrogen down to the bottom of the kettle?
I have done this without any sparge tube or line. I presurized from the top up to 10 psig , released pressure - took a sample of the gases in the kettle at the bottom, middle and top of the vessel and then did this agin for 2 more times. I came up with the following values:
1st test 2nd test 3rd test
Top: 13 8.3 5.4
Middle 13.6 8.6 5.4
Bottom 15 10 6.3
Questions: I was hoping to get below 5 % oxygen after three purges - no luck - also I noticed that the oxygen level was always higher near the bottom of the kettle... should I use a tube to force nitrogen down to the bottom of the kettle?
-
1
- #9
Montemayor
Chemical
cougarfan (or dideyjohn):
My response to your original query was based on various assumptions I had to make, since your basic data is sparce to say the least. I assumed:
1) The vessel is a batch, kettle reactor;
2) Time in between batches is valuable and should be reduced to a minimum;
3) The purging is done in-between batches;
4) The kettle is rated at 75 psig MAWP and has no vacuum rating;
5) The residual vapors in the kettle are heavier than air (or N2);
6) N2 supply is available at 75 psig;
7) All purging is done to a low pressure flare header;
8) If additional speed is beneficial, a blower can be used to expel the purge gases faster (without pulling a vacuum).
Depending on how expensive the down time is for you and on the N2 cost, I would still handle the purging in the manner I described earlier. I've been in this pickle before and have consistently found that the down time (dead production time) is far more expensive than the N2 required and that evacuation speed is of the essence. I have consistently purged kettles faster with the "sweep" method than with the Pressure or Vacuum method described by Guidoo. The pressure and vacuum methods are basically the same; they both depend on the time to diffuse the inert and the number of cycles.
I've found that every time you shut off the N2 and then re-initiate the pressurization again, the process takes a lot of operator procedural time - which costs $$$. That's why I am partial to the faster and more predictable "sweep" method. Of course, as Guidoo points out, there are other techniques you can employ (if you have the time): filling the vessel with water or another liquid; or pulling a vacuum (if your vessel can take it).
If your residual vapors are really heavy, you can try pushing them up (& out) through a dip pipe. In other words, do the reverse of what I originally suggested. In a "sweep" method, the direction makes no difference - as long as it is in a positive direction OUT of the system.
You are in the best position to evaluate since you possess all the basic data. It's your decision as to what you consider optimum conditions for your situation.
Lot of luck.
Art Montemayor
Spring, TX
My response to your original query was based on various assumptions I had to make, since your basic data is sparce to say the least. I assumed:
1) The vessel is a batch, kettle reactor;
2) Time in between batches is valuable and should be reduced to a minimum;
3) The purging is done in-between batches;
4) The kettle is rated at 75 psig MAWP and has no vacuum rating;
5) The residual vapors in the kettle are heavier than air (or N2);
6) N2 supply is available at 75 psig;
7) All purging is done to a low pressure flare header;
8) If additional speed is beneficial, a blower can be used to expel the purge gases faster (without pulling a vacuum).
Depending on how expensive the down time is for you and on the N2 cost, I would still handle the purging in the manner I described earlier. I've been in this pickle before and have consistently found that the down time (dead production time) is far more expensive than the N2 required and that evacuation speed is of the essence. I have consistently purged kettles faster with the "sweep" method than with the Pressure or Vacuum method described by Guidoo. The pressure and vacuum methods are basically the same; they both depend on the time to diffuse the inert and the number of cycles.
I've found that every time you shut off the N2 and then re-initiate the pressurization again, the process takes a lot of operator procedural time - which costs $$$. That's why I am partial to the faster and more predictable "sweep" method. Of course, as Guidoo points out, there are other techniques you can employ (if you have the time): filling the vessel with water or another liquid; or pulling a vacuum (if your vessel can take it).
If your residual vapors are really heavy, you can try pushing them up (& out) through a dip pipe. In other words, do the reverse of what I originally suggested. In a "sweep" method, the direction makes no difference - as long as it is in a positive direction OUT of the system.
You are in the best position to evaluate since you possess all the basic data. It's your decision as to what you consider optimum conditions for your situation.
Lot of luck.
Art Montemayor
Spring, TX
- Thread starter
- #10
Montemayor _ your basic assumptions are correct -
1) The vessel is a batch, kettle reactor - Yes
2) Time in between batches is valuable and should be reduced to a minimum - Yes and no - few minutes estra o=is okay...
3) The purging is done in-between batches - Yes
4) The kettle is rated at 75 psig MAWP and has no vacuum rating - typically at least 75 psig and most are rated for vacuum
5) The residual vapors in the kettle are heavier than air (or N2) - sometimes yes and sometimes no....
6) N2 supply is available at 75 psig - typically 80-90 psig
7) All purging is done to a low pressure flare header - vented to atmosphere
Can you explain further about sweeping purging versus pressure purging? Does it use less N2 typically? As I mentioned earlier, a few minutes extra doesn't affect us.
How is sweeping purging done?
I just changed my username to cougarfan in place of dideyjohn
1) The vessel is a batch, kettle reactor - Yes
2) Time in between batches is valuable and should be reduced to a minimum - Yes and no - few minutes estra o=is okay...
3) The purging is done in-between batches - Yes
4) The kettle is rated at 75 psig MAWP and has no vacuum rating - typically at least 75 psig and most are rated for vacuum
5) The residual vapors in the kettle are heavier than air (or N2) - sometimes yes and sometimes no....
6) N2 supply is available at 75 psig - typically 80-90 psig
7) All purging is done to a low pressure flare header - vented to atmosphere
Can you explain further about sweeping purging versus pressure purging? Does it use less N2 typically? As I mentioned earlier, a few minutes extra doesn't affect us.
How is sweeping purging done?
I just changed my username to cougarfan in place of dideyjohn
The method of inerting a vessel can be selected from a variety of options: vacuum purging, pressure purging, sweep-through purging or siphon purging. The selected one depends on factors such as time and equipment availability, cost of nitrogen, available inert gas pressures, complexity of operation, operators' skill, and, of course, level of inertness required.
From the test results submitted by cougarfan when using pressure purging, it appears that the mixing and diffusion of gases is not complete and shows about a 15% deviation from top to bottom. Using that 15% as a SF we could estimate with simple arithmetics, the number of cycles and/or the oxygen concentration required as follows:
Assume starting concentration of oxygen and pressure are 21% and 1 ata, respectively.
After pressurizing to, say, 2 ata with pure nitrogen, (i.e., PR=2) the resulting concentration would drop to 21:2=10.5%, adding 15% as a SF, the vessel average concentration after depressurizing back to 1 ata becomes 12.1%.
Any successive cycle under the same pressurizing/ depressurizing conditions, would result in an oxygen concentration equal to the previous one divided by 2 plus 15%.
So,
the first: (21/2)(1.15)=12.1%
the 2nd: (12.1/2)(1.15)=6.9%
the 3rd: (6.9/2)(1.15)=4%
the 4th: (4/2)(1.15)=2.3%, and so forth.
Combining the above into a single formula one gets for final oxygen concentration:
yf=21(1.15/PR)N
where PR is the pressure ratio (pressurizing to previous depressurizing) conditions assumed to be equal in all N cycles. By applying natural logarithms to the formula one gets N, the number of cycles.
In the example given, after N=4 cycles and a 15% SF:
yf=21(1.15/2)4=2.3%
Meaning that, under the above conditions, 4 cycles are enough to bring down the oxygen concentration to below 2.5%; one can always add one more cycle for safety reasons.
If the PR is not constant, one must use the actual PR per cycle to estimate the oxygen dilution effected in that particular cycle. If the inert gas itself contains some oxygen, as it happens with nitrogen obtained from air by PSA, then the dilution effect should be correspondingly adjusted. In cases as these, in which conditions vary from cycle to cycle, it is natural to do hand estimates as above, rather than using formulas.
From the test results submitted by cougarfan when using pressure purging, it appears that the mixing and diffusion of gases is not complete and shows about a 15% deviation from top to bottom. Using that 15% as a SF we could estimate with simple arithmetics, the number of cycles and/or the oxygen concentration required as follows:
Assume starting concentration of oxygen and pressure are 21% and 1 ata, respectively.
After pressurizing to, say, 2 ata with pure nitrogen, (i.e., PR=2) the resulting concentration would drop to 21:2=10.5%, adding 15% as a SF, the vessel average concentration after depressurizing back to 1 ata becomes 12.1%.
Any successive cycle under the same pressurizing/ depressurizing conditions, would result in an oxygen concentration equal to the previous one divided by 2 plus 15%.
So,
the first: (21/2)(1.15)=12.1%
the 2nd: (12.1/2)(1.15)=6.9%
the 3rd: (6.9/2)(1.15)=4%
the 4th: (4/2)(1.15)=2.3%, and so forth.
Combining the above into a single formula one gets for final oxygen concentration:
yf=21(1.15/PR)N
where PR is the pressure ratio (pressurizing to previous depressurizing) conditions assumed to be equal in all N cycles. By applying natural logarithms to the formula one gets N, the number of cycles.
In the example given, after N=4 cycles and a 15% SF:
yf=21(1.15/2)4=2.3%
Meaning that, under the above conditions, 4 cycles are enough to bring down the oxygen concentration to below 2.5%; one can always add one more cycle for safety reasons.
If the PR is not constant, one must use the actual PR per cycle to estimate the oxygen dilution effected in that particular cycle. If the inert gas itself contains some oxygen, as it happens with nitrogen obtained from air by PSA, then the dilution effect should be correspondingly adjusted. In cases as these, in which conditions vary from cycle to cycle, it is natural to do hand estimates as above, rather than using formulas.
- Thread starter
- #12
25362 - I ran the test aagin and waited longer for the sample to be taken. The pressure in the vessel was taken to 10 psig.. The nitrogen is 99.5 % purity. The results were as follows:
1st test 2nd test 3rd test
Top: 12.2(13) 7.6(8.3) 4.9(5.4)
Middle 12.0(13.6) 7.4(8.6) 4.8(5.4)
Bottom 12.1(15) 7.4(10) 4.7(6.3)
The results were more together giving me more time to sample the gas in the vessel. Where is the formula you referenced above come from? When I checked it for the testing I just did, it came pretty close. Is pressure purging the most effective type in a shortest period of time? If not, what is and how is it done?
1st test 2nd test 3rd test
Top: 12.2(13) 7.6(8.3) 4.9(5.4)
Middle 12.0(13.6) 7.4(8.6) 4.8(5.4)
Bottom 12.1(15) 7.4(10) 4.7(6.3)
The results were more together giving me more time to sample the gas in the vessel. Where is the formula you referenced above come from? When I checked it for the testing I just did, it came pretty close. Is pressure purging the most effective type in a shortest period of time? If not, what is and how is it done?
Cougarfan,
Could you give an indication how long you waited before you took a sample (just interested).
Formulas used by 25362 can be found in article I mentioned before:
G.R. Kinsley Jr, "Properly Purge and Inert Storage Vessels", Chemical Engineering Progress, Februari 2001, pages 57-61. Equation nr. (9).
This article describes all methods in detail. You can download it from
As I wrote before, pressure purging seems to me a good option in your situation. Vacuum purging requires a vacuum pump, and vessel needs to be rated for vacuum. For sweep purging you need to install a dip pipe. It may be faster than pressure purging, but requires more nitrogen. It depends on your own situation (e.g. costs/availability of nitrogen, costs of downtime, whether you want to install the dip pipe etc.) what is more economical.
Could you give an indication how long you waited before you took a sample (just interested).
Formulas used by 25362 can be found in article I mentioned before:
G.R. Kinsley Jr, "Properly Purge and Inert Storage Vessels", Chemical Engineering Progress, Februari 2001, pages 57-61. Equation nr. (9).
This article describes all methods in detail. You can download it from
As I wrote before, pressure purging seems to me a good option in your situation. Vacuum purging requires a vacuum pump, and vessel needs to be rated for vacuum. For sweep purging you need to install a dip pipe. It may be faster than pressure purging, but requires more nitrogen. It depends on your own situation (e.g. costs/availability of nitrogen, costs of downtime, whether you want to install the dip pipe etc.) what is more economical.
- Thread starter
- #14
Siphon purging is the procedure that uses least nitrogen, because the reactor is filled with any compatible liquid leaving very little space for gas, and, when combined with the sweeping method, it is very useful to attain very low oxygen levels.
Pressure purging is much quicker than the vacuum purging method but uses more inert gas. Besides, as the vacuum is deeper the capacity of the vacuum system decreases. Thus the final decision on which purging procedure to use is economical, and based on cost and performance (mainly the oxygen level desired). Apparently, pressure purging is the simplest and most practical method, when the final level of oxygen desired is in the range required by cougarfan.
As I explained the formula I presented, is simple a result of arithmetics, by repeating N times (cycles) the same procedure.
Pressure purging is much quicker than the vacuum purging method but uses more inert gas. Besides, as the vacuum is deeper the capacity of the vacuum system decreases. Thus the final decision on which purging procedure to use is economical, and based on cost and performance (mainly the oxygen level desired). Apparently, pressure purging is the simplest and most practical method, when the final level of oxygen desired is in the range required by cougarfan.
As I explained the formula I presented, is simple a result of arithmetics, by repeating N times (cycles) the same procedure.
If the purge is a single operation on empty vessel you can vacuumize it and break vacuum with inert gas few times.
If you are working with flammable liquid it is preferrable to do blanketing instead purge (slight overpressure up to 5-10 mbarg by pilot valve and concervation vent)
If you are working with flammable liquid it is preferrable to do blanketing instead purge (slight overpressure up to 5-10 mbarg by pilot valve and concervation vent)
- Status
Similar threads
- Locked
- Question
- Locked
- Question
- Locked
- Question