[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

CO2 valve characteristics at near empty tank

***Highly technical post warning*** skip to last paragraph for summary
if you wish.

There are two technical inaccuracies and some of this misinformation has
been repeated so often that it is a serious problem of becoming one of
our aquatic plant urban legends. I hope that any follow up postings will
remain civil and technically accurate. This is not meant to be personal
in any way. I do hope this is new information since I have not been
monitoring the APD closely in recent months. We've had a recent addition
to the family; young Dylan joined us three weeks ago.

The first point is relatively minor: fluid flow is never supersonic
except in the vicinity of aircraft or objects like bullets which are
travelling through a fluid faster than the speed of sound in that fluid.
At that point, there is a standing shock wave which propagates from the
object which is the boundary of the fluids going at different speeds.
The fluid on the objects side of the shock wave is not supersonic
relative to the object and the fluid on the other side of the shock wave
is virtually stationary with respect to the ambient fluid. The fluid in
the shockwave itself is highly compressed, hot and probably forms a
region where the speed of sound of the fluid is much higher locally
because these molecules are travelling faster than the speed of sound in
the ambient fluid. I'm not as conversant with incompressible fluids such
as liquids but I think it would be the same. In valves, I don't think
its correct to say that the flow is supersonic although you should be
aware that the speed of sound in a fluid is a function of temperature,
pressure and density in the fluid. I believe that this fully choked flow
occurs when the inlet pressure is twice the outlet pressure so this will
always occur in CO2 systems where the inlet pressure is above 30 psi or
so (depending if there is any significant downstream resistance).

The second point refers to the so called "empty tank dumping" syndrome.
I don't know of any reason why pressure would ever rise within a CO2
cylinder near its end of life unless it were subjected to a sudden and
dramatic increase in temperature. I can think of one phenomenon which
-might- account for an observation of increased flow near end of tank
but I'm not conversant enough with the actual flow coefficients of the
valves in question to know if this is possible or not. Someone who is
interested should contact the manufacturers and inquire about this.

Here's my proposed theory: in a high pressure system where there is no
regulator or where the regulator pressure is set very high, as the tank
becomes empty there will be a time when the tank pressure falls below
the normal operating pressure at the inlet of the needle valve or other
flow restricting device (such as a sintered glass diffuser) and the flow
rate through the valve diminishes until the energy loss within the valve
is low enough that the flow in the valve transitions from turbulent flow
to laminar flow. In laminar flow mode, the effective valve coefficient
changes (its non-linear remember) and a greater volume of fluid can
transit at a lower pressure. This phenomenon would be exacerbated when
the valve is operated at a much higher range of pressures. If the valve
is always operating in laminar flow mode, there would be no such laminar
to turbulent flow transition. If the valve is in choked mode, i.e.
operating at greater than 30 psi, then the valve is certainly operating
in turbulent flow mode. If the flow mode is turbulent, then there will
always be some point at which flow transitions to laminar; the only
question is how much CO2 is still inside the system when this occurs. I
suspect but cannot confirm that valves operating at 5-10 psi may be in
laminar flow mode. There will be a good deal less CO2 to expel at end of
tank in a system where the inlet valve is operating at a low pressure
than at one operating at high pressure. At 30 psi there is about two
tank volumes of CO2 left to expel, at 60 psi there are 4 tank volumes
left to expel. If there is no pressure regulator in the system, there is
a very significant volume of CO2 left to expel when the flow
characteristics of the valve begin to change. These laminar/turbulent
transitions may be exacerbated in other types of flow restrictive
devices since the pressure/flow graph may be even more non-linear than a
needle valve which is quite good. The manufacturers of these devices
should provide us with the pressure/flow graphs since it is a very
simple matter to determine experimentally with the correct instruments. 

I hope that everyone understands the importance of the pressure
regulator in providing accurate flow control near end of life in a high
pressure system. Pressure regulators control pressure. Needle valves
control flow rates at a constant pressure. Both devices are required to
-accurately- control flow rates. I cannot say whether sintered diffusers
provide sufficient resistance without needle valves but they should be
much safer together with a regulator and needle valve in series. I
venture that too low a flow is not really the problem, too high a flow

Steve Pushak                              Vancouver, BC, CANADA 

Visit "Steve's Aquatic Page"      http://home.infinet.net/teban/
 for LOTS of pics, tips and links for aquatic gardening!!!