how fast to empty a tank ?

[Attchoum]

good morning everyone.

I read a article few years/month ago about how fast a tank will be empty in a boom senario. Now im teaching a deep diving course and I cannot find the info anymore...

the article was about different boom senario like: burst disk failure, lp hose failure, 2 stage freeflow, Ip hose failure, and so on...

if anyone have he info I would really appreciated.

thx
Etienne

[sskasser]

Coincidentally, this was posted about 5 minutes after your question: http://www.cavediver.net/forum/showthread.php?t=10779

[netmage]

I believe this is the post you were thinking of.....

http://thedecostop.com/forums/showth...ain+tank+empty

-Tim

[Attchoum]

nice coincidence...

thx for the info

etienne

[SmittyGoesDeep]

You guys sure about this?

I'll fess up to being a newbie and all, but I distinctly remember killing time (lots of it) hanging around dive shops in my OW days waiting for even partially filled tanks to drain so they could be VIPed. The guy would usually crank the valve all the way open, drape a towel over the valve and let 'er rip in a back room. I'd wait ... and wait.

I'm willing to be convinced, but that data about a full 80 emptying in seconds really flies in the face of my experience. How could that much air empty through such a tiny opening as a valve in such a short time? I mean are the physics right on this?

That said, does anybody have thoughts on how to quickly get to your manifold for isolation? Again, I'm only Basic, so not a ton of experience. But I've tried to practice just such a scenario and have always had a challenge contorting myself to reach around and grab the knobs. Am I configuring my gear wrong, or is this a challenge for everyone who dives backmount?

I'd Appreciate your thoughts — Smitty

[aainslie]

The key thing to realise is that the flow rate is related to the pressure in the cylinder. let's say it's directly proportional to pressure (this probably isn't quite true but it'll help make the point). Then the flow out would be exponential.

So while it may "hiss" for a long time, that's basically getting rid of the last couple of hundred PSI.

To show how this exponential stuff works - let's say the tank takes 20 seconds to drop from 3000 psi to 1500 psi. Then the numbers would be like this:

300-1500 20 secs
1500-750 20 secs
750-375 20 secs
375-187 20 secs
187 - nothing Another minute or so.

So the thing is making a helluva racket for a little under 3 mintues or so - but it lost half its gas in 20 seconds!!

I'll bet that most of you when looking at Curt's numbers were thinking linear, e.g. if it takes 70 secs to go to empty, then it must take 35 secs to go from 3000 to 1500. This isn't the case! It'll be closest to exponential as the exit path is least constrained (such as a burst disk) and closest to linear when highly constrained (such as a HP hose going, where there's a small orifice to sort of act like a supersonic orifice).

So two takeaways:

1) In a lot of cases things are WORSE than Curt's numbers initially lead one to believe,
2) Listening to how long cylinders hiss in a shop isn't a good indicator. A lot of that hissing is after the pressure has dropped below 500 psi or so, effectively useless for us.

[chimie007]

You guys sure about this?

I'll fess up to being a newbie and all, but I distinctly remember killing time (lots of it) hanging around dive shops in my OW days waiting for even partially filled tanks to drain so they could be VIPed. The guy would usually crank the valve all the way open, drape a towel over the valve and let 'er rip in a back room. I'd wait ... and wait.

Sure about this ? You go and try it..... make sure you wear ear protection.... I bet you the dive monkey wasn't opening the valve all the way.

[aainslie]

Sure about this ? You go and try it..... make sure you wear ear protection.... I bet you the dive monkey wasn't opening the valve all the way.

Raphael,

Do you ahve any idea how often you and I answer a post simultaneously? It's scary really...

[Attchoum]

I remember busting a busrt disk on a huge fill tank.

in about 40 seconde air was out, and we are talking about a 5 pieds tall tank

in a backmount configuration, i can close my isolation valve and my right valve in less that 15 seconde

isolator and left take my aabout 22

see you

etienne

[aainslie]

For everyone except Raphael, the Canucks and any Louisianans wandering through the door - a "pied" is a foot.

Burst disks are one of the best routes out of a tank. The path through the valve is pretty convoluted, and the burst disk shortcuts some of these.

Hmmm... back on the linear thing, I'll bet that losing anything on the LP side of things is closer to linear than exponential, slowing down the initial bad effects quite a bit. But I'm guessing, it'd be interesting to set up a tank with a disconnected LP hose, open it up and time the loss of each 500 psi as the tank drains.

Anyone sitting bored at CEE reading this? How about trying an extension of Curt's stuff, and redoing those scenarios with the addition of timing the loss of each 500 PSI of gas, i.e. fill to 3000 and take times at 2500, 2000, 1500, 1000 and 500 psi? Try it for:

- A burst disk (just shove a wrench on the holder and take it off)
- An HP hose (remove the gauge and turn the valve on. you'll need a gauge on a second HP port)
- A "broken" LP hose (take off the second stage)
- A freeflowing reg (this'll be highly dependent on the tuning of the reg but I'm betting gas loss will be pretty linear).