Chuck Gadd wrote:

On Tue, 11 Feb 2003 10:30:53 -0800, Dave Millman
wrote:

However, about 3 weeks ago my dual gage regulator-needle valve system ran out of
gas. It did an end of tank dump. pH dropped from about 6.4 to a bit under 6
before I noticed fish behavior and turned it off.

So a needle valve does not prevent an empty tank pressure dump.

What pressure do you run out of your regulator? Does your system
have a check-valve? Where is the check valve located?

CO2 tank side reads 800 psi, needle valve side set at 10-11 psi.

I have a German check valve on the hose from the regulator to the reactor.

The whole system came from M3 about 15 months ago.

I've seen cases where the regulator pressure was set VERY low (2psi)
and the resulting increase when the "tank dump" happened caused in a
CO2 overdose. But if the regulator pressure is set higher, the
needle-valve needs to be closer tighter, and when the output pressure
rises, the needle-valve still limits the CO2 and prevents problems.

Additionally, if there is a check-valve located after the
needle-valve, the needle-valve often has to be set at a higher rate in
order to overcome the resistance of the check-valve. Some
check-valves cause problems, others don't.

This is really interesting. In between my last post and this, I was up at a fish
store that told me about a fish die-off they attributed to not turning off the CO2 at
night. I looked at their tank: needle valve, no cover, lights suspended 8 inches
above the tank, no chance of CO2 being trapped above the water and displacing O2. I
told them it was unlikely that the night time ph drop killed the fish, and asked for
more detail. They told me it happend three weeks ago, pH measured below 6 when they
saw the dead fish. Then they said they had just about been ready to change the CO2
tank, because it was almost empty before the die off. This is exactly what happened
in my tank, except nothing died (they did stop eating for a day, though!).

So I have two separate experiences that the CO2 rate through a needle valve system
can rise when a bottle gets empty. What does the CO2 triple point diagram say? Does
pressure increase when the last liquid is gone?