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 MintJulep on being selected by the Eng-Tips community for having the most helpful posts in the forums last week. Way to Go!
air purging 2
- Thread starter Barhp
- Start date
-
1
- #2
You didn't say what pressure the tank is currently at (or how much pressure the "tank" can hold). You also didn't say if you were planning to do a Clearing Purge, a Displacement Purge, or a Dilution Purge. It matters.
If the tank can take the pressure, then a dilution purge is the easiest to calculate, but the pressure is pretty high.
If you have a good flow path that avoids bypassed flow you can do a clearing purge, but this is rare in a tank and the chance of not having bypassed flow is vanishingly small.
David
If the tank can take the pressure, then a dilution purge is the easiest to calculate, but the pressure is pretty high.
If you have a good flow path that avoids bypassed flow you can do a clearing purge, but this is rare in a tank and the chance of not having bypassed flow is vanishingly small.
David
-
1
- #3
Compositepro
Chemical
CO2 is considerably denser than air so if you can add air at the top and exhaust at the bottom it will be considerably faster and more efficient that the opposite, or adding and venting at locations close together.
moltenmetal
Chemical
Is this purge required to permit entry into the tank? If so, you won't be relying on merely on flow and duration- you'll need to measure the actual O2 concentration in the tank, and not just at the tank vent, before you send people in there.
If it were a perfectly mixed tank, the solution to your estimate is a simple differential equation. The trouble as zdas04 points out is that real tanks are neither perfectly mixed nor perfectly "plug flow" from inlet to outlet, nor are these the only two options since unmixed portions may exist and persist. The potential for stratification or other unmixed areas in the tank depends on tank geometry, location of the inlet and outlet, velocity of the inlet etc., and cannot be discounted even when your starting material is 98.5% N2 and your end point is 80% N2.
If it were a perfectly mixed tank, the solution to your estimate is a simple differential equation. The trouble as zdas04 points out is that real tanks are neither perfectly mixed nor perfectly "plug flow" from inlet to outlet, nor are these the only two options since unmixed portions may exist and persist. The potential for stratification or other unmixed areas in the tank depends on tank geometry, location of the inlet and outlet, velocity of the inlet etc., and cannot be discounted even when your starting material is 98.5% N2 and your end point is 80% N2.
Compositepro has raised a point that I have often wondered about, and have come to the conclusion that gases do not sink through each other like that. Has anyone ever actually measured this layering effect? If it were true we would find O2 in greater concentrations near the floor of an enclosed room (and N2 near the ceiling), but we do not.
If we take the analogy with liquids then obviously with immiscible liquids (e.g. oil and water) then the lighter oil floats to the top. However, with perfectly miscible liquids (e.g. ethanol and water) they remain mixed. If this were not so then the first glass poured from a bottle of wine would be stronger than the last. Surely gaseous CO2, O2 and N2 are as miscible as ethanol and water?
I would say that by far the majority of engineers that I have come across (including professors) think that gases layer in this way. Is there anyone who thinks the way I do?
I have seen dry ice (solid CO2) stored in a pit where the CO2 gas remained in the pit, causing extreme danger, but I believed the CO2 atmosphere remained there because of the cold and the constant generation of gas from the sublimation of the blocks.
Katmar Software
Engineering & Risk Analysis Software
If we take the analogy with liquids then obviously with immiscible liquids (e.g. oil and water) then the lighter oil floats to the top. However, with perfectly miscible liquids (e.g. ethanol and water) they remain mixed. If this were not so then the first glass poured from a bottle of wine would be stronger than the last. Surely gaseous CO2, O2 and N2 are as miscible as ethanol and water?
I would say that by far the majority of engineers that I have come across (including professors) think that gases layer in this way. Is there anyone who thinks the way I do?
I have seen dry ice (solid CO2) stored in a pit where the CO2 gas remained in the pit, causing extreme danger, but I believed the CO2 atmosphere remained there because of the cold and the constant generation of gas from the sublimation of the blocks.
Katmar Software
Engineering & Risk Analysis Software
I sure would like to hear an answer to Katmar's queston -
As a purveyor of gas sensors, I have tested many spaces -- and found that both CO2 and CO concentrations do not vary with elevation, even in some pretty tall rooms with stagnant air. CO2 should fall (specific gravity = 1.5+) and CO should rist (specific gravity = 0.96 or so).
The only way we get them to separate is to alter the temperature of the gas introduced to the air, but even then, after some period of time, it becomes homogenous...
N2O is different, though its specific gravity is really close to CO2. Releasing hot NO2 (like from a diesel engine exhaust) into a room, you'll immediately find a high concentration at the ceiling -- but it fairly quickly cools and falls to the floor...
Wish I understood the mechanism...
Good on y'all,
Goober Dave
As a purveyor of gas sensors, I have tested many spaces -- and found that both CO2 and CO concentrations do not vary with elevation, even in some pretty tall rooms with stagnant air. CO2 should fall (specific gravity = 1.5+) and CO should rist (specific gravity = 0.96 or so).
The only way we get them to separate is to alter the temperature of the gas introduced to the air, but even then, after some period of time, it becomes homogenous...
N2O is different, though its specific gravity is really close to CO2. Releasing hot NO2 (like from a diesel engine exhaust) into a room, you'll immediately find a high concentration at the ceiling -- but it fairly quickly cools and falls to the floor...
Wish I understood the mechanism...
Good on y'all,
Goober Dave
I have had a bit more of a think about the problem, and even done a quick experiment gently pouring cold colored water into clear hot water, and now have a better idea of what happens.
Compositepro is correct in that if you add a body of less dense gas into the top of a vessel it will tend to float on the heavier gas for a while. But, depending on how gently you introduce the lighter gas, there will be relatively rapid mixing. And once mixed, the gases will not layer out again.
This is exactly in line with DRWeig's experience where hot light gases rise, and then rapidly mix with the rest of the environment.
The difference is that a body of gas can be lighter or heavier than the bulk gas and therefore rise or fall, but once mixed the gravity forces on an individual molecule are probably outweighed by the Brownian Motion collisions keeping the gas uniformly mixed.
Anyway, what does all this theorizing have to do with Barhp's original question?
If you can introduce the air fairly remotely from where the N2 is released then the most conservative assumption is that the vessel is perfectly mixed - i.e. do not assume that you can take advantage of plug flow or layering to flush the N2 out. There was a time when I could have set up (and maybe even have solved) the differential equations for this process, but now I would break it down into 15 second intervals and just copy the perfect mixing formulas down in my spreadsheet however many times necessary until I got the desired degree of dilution.
And then I would take moltenmetal's advice and measure the O2 concentration before entering the vessel. This is probably a statutory requirement where ever you are.
Katmar Software
Engineering & Risk Analysis Software
Compositepro is correct in that if you add a body of less dense gas into the top of a vessel it will tend to float on the heavier gas for a while. But, depending on how gently you introduce the lighter gas, there will be relatively rapid mixing. And once mixed, the gases will not layer out again.
This is exactly in line with DRWeig's experience where hot light gases rise, and then rapidly mix with the rest of the environment.
The difference is that a body of gas can be lighter or heavier than the bulk gas and therefore rise or fall, but once mixed the gravity forces on an individual molecule are probably outweighed by the Brownian Motion collisions keeping the gas uniformly mixed.
Anyway, what does all this theorizing have to do with Barhp's original question?
If you can introduce the air fairly remotely from where the N2 is released then the most conservative assumption is that the vessel is perfectly mixed - i.e. do not assume that you can take advantage of plug flow or layering to flush the N2 out. There was a time when I could have set up (and maybe even have solved) the differential equations for this process, but now I would break it down into 15 second intervals and just copy the perfect mixing formulas down in my spreadsheet however many times necessary until I got the desired degree of dilution.
And then I would take moltenmetal's advice and measure the O2 concentration before entering the vessel. This is probably a statutory requirement where ever you are.
Katmar Software
Engineering & Risk Analysis Software
Compositepro
Chemical
It is exactly the same mechanism at work as cause convection currents in air or liquids. There are two competing effects. Mixing due to diffusion and convection, and stratification due to density differences. Stratification is a powerful but macroscopic effect. Hot air balloons rise due to it and the envelop prevents mixing. The density ratio of CO2 to air is 44/29=1.52. The density ratio between air at 20C and air at 30C is 303/293= 1.034. This is a small density difference but it will create strong convection currents that promote mixing if the heat is at the bottom. It also creates a strong tendency to stratify if the heat is at the top.
Diffusion has a powerful mixing effect on a molecular scale but it is very slow and ineffective on a macroscopic scale. So, once mixed, air and CO2 will never separate. But it will take a very long time for diffusion alone to mix a large vessel containing CO2 on the bottom with air at the top. Convection caused by thermal effects or drafts is far more effective on a macroscopic scale.
Pits and low areas are often classified as confined spaces for safety reasons where heavy gasses may be present.
Diffusion has a powerful mixing effect on a molecular scale but it is very slow and ineffective on a macroscopic scale. So, once mixed, air and CO2 will never separate. But it will take a very long time for diffusion alone to mix a large vessel containing CO2 on the bottom with air at the top. Convection caused by thermal effects or drafts is far more effective on a macroscopic scale.
Pits and low areas are often classified as confined spaces for safety reasons where heavy gasses may be present.
Since I started teaching my Purge class in 1996 (and several hundred times since) I've gobbled up everything I could find on this subject.
The best I can tell from the literature and from some unpublished experiments is that gases at the same pressure resist mixing with gases at different velocities or temperatures. When temperatures equilibrate, mixing begins in earnest and once mixed cannot be mechanically separated without extreme measures (that's what the U.S. government spent tens of billions of dollars at Oak Ridge, TN to do during the Manhattan Project in WWII, it was all about doing mechanical gas on gas separation).
All the Who-Done-It book, movie, and TV plots of someone introducing dry ice into a room and letting the sublimation create an oxygen poor environment are plausible because the CO2 is so cold that it will stratify. The rest of the plot is that after the victim has suffocated, the CO2 warms up and distributes itself through the building becoming nearly undetectable.
When doing a Clearing Purge (for example), you are relying on the no-flow boundary at the edges of your purge gas to drag the gas out of the way. This is actually pretty effective and in pipe or a reasonably sized vessel, about three volumes of purge gas gives you a 99.8% confidence that you have removed all of the gas you don't want. The problem is that there is absolutely no way to get an efficient gas flow within a tank. All of the inlet options are too small. All of the outlet options are too small. Maximum pressure is too low.
Consequently, your only choice for a tank is usually a modified Dilution Purge. This goat-roping goes something like:
(1) pressure the tank to 1 psig with air;
(2) evacuate the tank to -1 psig with a vacuum pump;
(3) Introduce air to raise pressure to atmospheric;
(4) Test tank atmosphere with a remote probe (with a pump) in at least 8 locations each 5 ft of tank height (24 tests for a 15 ft tall tank); and
(5) repeat until tank atmosphere is at the desired concentration.
You'll hit the repeat step about 20-30 times with a 2 psig swing. Plan on it taking a day and a half. You could significantly shorten the process if the tank could take more pressure and/or more vacuum, but blowing the lid off or crushing it like a beer can are really not optimum results.
David
The best I can tell from the literature and from some unpublished experiments is that gases at the same pressure resist mixing with gases at different velocities or temperatures. When temperatures equilibrate, mixing begins in earnest and once mixed cannot be mechanically separated without extreme measures (that's what the U.S. government spent tens of billions of dollars at Oak Ridge, TN to do during the Manhattan Project in WWII, it was all about doing mechanical gas on gas separation).
All the Who-Done-It book, movie, and TV plots of someone introducing dry ice into a room and letting the sublimation create an oxygen poor environment are plausible because the CO2 is so cold that it will stratify. The rest of the plot is that after the victim has suffocated, the CO2 warms up and distributes itself through the building becoming nearly undetectable.
When doing a Clearing Purge (for example), you are relying on the no-flow boundary at the edges of your purge gas to drag the gas out of the way. This is actually pretty effective and in pipe or a reasonably sized vessel, about three volumes of purge gas gives you a 99.8% confidence that you have removed all of the gas you don't want. The problem is that there is absolutely no way to get an efficient gas flow within a tank. All of the inlet options are too small. All of the outlet options are too small. Maximum pressure is too low.
Consequently, your only choice for a tank is usually a modified Dilution Purge. This goat-roping goes something like:
(1) pressure the tank to 1 psig with air;
(2) evacuate the tank to -1 psig with a vacuum pump;
(3) Introduce air to raise pressure to atmospheric;
(4) Test tank atmosphere with a remote probe (with a pump) in at least 8 locations each 5 ft of tank height (24 tests for a 15 ft tall tank); and
(5) repeat until tank atmosphere is at the desired concentration.
You'll hit the repeat step about 20-30 times with a 2 psig swing. Plan on it taking a day and a half. You could significantly shorten the process if the tank could take more pressure and/or more vacuum, but blowing the lid off or crushing it like a beer can are really not optimum results.
David
- Status
Similar threads
- Locked
- Question
- Locked
- Question