Gas Flow Measurement
If flow is your thing, we got lots to show you! "AGA" refers to
the American Gas Association. They developed these standards, or
at least they publish them.
AGA 3 is American Gas Association's Report No. 3, which deals with the
proper design and calculation methodology of deriving flow from an
orifice meter.
Basically, an orifice
meter is a pipe with a restricting plate inside it. It's
cheap, it's robust and it's a very common way of measuring fluid
flow. The plate has a hole of a known size in the middle.
When a fluid flows through the "orifice plate", as it is called,
there's a pressure difference between the plate's two sides. By
measuring this differential pressure, if you know the density of the
fluid, you can calculate the rate of flow. Click on the "AGA 3"
link to do a sample calculation. The page only deals with gas
flow, but orifice meters can be used for liquids too.
Note: this calculation will give you the flavour of the calculation but
there's a small problem with the discharge coefficient, so don't take
the results of this page to the bank. In fact, the whole page is
here for educational purposes only and while I'm quite happy if you
learn more about flow measurement here, don't come crying to me if you
use these pages for some other reason and find out they're
flawed. Of course, if you do find out there's a problem in one of
the pages, do please tell me.
AGA 3
Now, AGA 7 deals with gas meters that are generally more accurate but
are also more finicky than orifice meters. Any kind of meter that
produces a known count of pulses per unit volume flowed can be
compensated back to a base pressure and temperature with AGA 7.
This includes the bellows meter (the big gray gas meters people have in
their houses) and the turbine meter, which is basically a propeller
inside the pipe. The calculation is much simpler than AGA
3. To do a sample AGA 7 calculation, choose the link below.
AGA 7
Back in the old days, scientists couldn't measure gases accurately
enough to see the weird things that gases do under pressure, so Boyle,
Charles and Avogadro got together and cooked up the "ideal gas law"
-- PV = nRT (not literally -- see the link). You probably
remember that gem from high school chemistry. Now that people
make big money shipping gas under high pressure around the countryside,
there's money riding on figuring out exactly how much gas went from A
to B. In the process of figuring this out, the scientists
discovered that real gases aren't ideal at all! They almost
always take up less volume than the ideal gas law says they
should. The proportion of actual volume to ideal volume is known
as the "compressibility"
of the gas. It depends on what is in the gas as well as the
pressure and temperature of the gas.
Since a gas is a big collection of molecules with electrons on the outside and protons on the inside,
bouncing around against each other, those molecules aren't as indifferent to each other as
the ideal gas law assumes them to be. They attract each other
to some extent, either just a tiny bit in the case of fairly inert
gases, or quite a lot in the case of gases with a strong dipole
moment, like carbon dioxide or hydrogen sulphide. If you've ever
held two magnets close together, you'll know that they will
automatically align so as to attract each other. The same thing
happens with adjacent molecules, so they cozy up a little bit compared
to the ideal gas law's predictions. This results in net density
increase, and the amount of this increase is what AGA8 calculates.
AGA 8, below, is basically the "industry standard" for calculating
compressibility. For one thing, it literally is an industry
standard, having been created and published by AGA, an industry
body. It's a long winded calculation but it's interesting.
It's just one of a number of calculations that do the same thing.
NX-19 is another simpler but less accurate calculation. These
calculations are known as equations
of state. The page below does the detail characterisation
calculation.
AGA 8
If you just want to crank out a quick orifice plate calibration, you can use this page -- it performs both AGA3 and AGA8.