Moisture in the tank?

[all4scuba05]

Getting a little confused reading some of these threads.
1-If the pressure in a tank is higher than ambient pressure, can ambient air still push its way in? If so, explain.
2-If a tank has 4 psi worth of air in it and the valve is closed how does moisture get in?

 

[The Chairman]

At 33fsw the pressure surrounding the tank is 14.7 psi. Nature abhors a vaccuum, even a relative one.

Also, moisture can be introduced during the filling process. This would point to a incompetently set up and/or maintained compressor.

 

[miketsp]

At 33fsw the pressure surrounding the tank is 14.7 psi. Nature abhors a vaccuum, even a relative one.

Also, moisture can be introduced during the filling process. This would point to a incompetently set up and/or maintained compressor.

??
At the surface the pressure surrounding the tank is 14.7psi.
At 33fsw the pressure surrounding the tank will be twice that.

The OP doesn't make it clear if the measured tank pressure is absolute or 4psi above ambient.

 

[all4scuba05]

The threads I've been reading are pertaining to tanks that are out of water.

 

[all4scuba05]

when folks say that a tank has 3000 psi in it. They mean psig. So lets stick with that language for simplicity sake

 

[rjack321]

Water almost always gets into tanks during the filling process or when swapping cylinders. Water is often in the valve area and the fill whip blows it into the tank if the tank valve isn't cracked and blown free of water first.

 

[The Chairman]

Water almost always gets into tanks during the filling process or when swapping cylinders. Water is often in the valve area and the fill whip blows it into the tank if the tank valve isn't cracked and blown free of water first.

As I first said: Incompetence is the reason for this to happen. There is NO WAY that a competent blender/filler will allow water to get into a tank. NO WAY.

 

[Luis H]

As long as air under pressure comes out nothing can go in. I normally recommend not to let your tank to go below 100 PSI just as a safety margin. With 100 PSI in the tank, the pressure inside the tank is higher than the pressure at any recreational diving depth. Therefore, no one can argue that water could have gotten in during the dive.


The most common way of getting moisture into the tank is during your gas fill. This is the only time when something is being pushed into the tank (by the filling gas pressure). If the connection is not totally dry, any water droplets in the valve or the filling hose will be pushed into the tank. A few water droplets is a lot more moisture than any respectable compressor air filter will allow to pass.

I have visual hundreds of tanks, and whenever I see any light surface rust, the pattern is always in narrow streaks, like from water droplets running down the inside.

I don’t care if you are filling in a cooling water tank or not, I always blow dry both sides of the connection (tank valve and filling hose) just before I connect them. Many of my steel tanks are over 40 years old and I intend to keep for another 30 years or more.


Wow...5 posts in the time that it took me to type that.

 

[Leadking]

If ambient is 14.7 and pressure gauge is 4 psi we have 4 psi positive or 18.7 absolute!

 

[miketsp]

Well now that we've had the usual answers let me throw a spanner in the works.

Let's start with a pressurised cylinder at 3000psi with perfectly dry air.
Water vapour is for all intents and purposes treated as a gas that obeys Dalton's Law of Partial Pressures.
So we have inside the cylinder a 0PP of water vapour.
Everytime you open the valve to bleed air out or to pressurise your reg you will agree that externally the PP of the water vapour is above zero.
So there will in fact be a backwards diffusion of the water vapour into the cylinder even though the other gases via their PPs total 3000psig and are flowing out.

I remembered reading about this effect but I didn't want to make the post until I located some backup.

If you look at
http://www.eng-tips.com/viewthread.cfm?qid=123404

you will see some discussion on reverse diffusion back into HP systems.