Re: Measuring water flow

```
On Sat, 13 Nov 1999, Duane wrote:

> I know this is not a plant question but I am hoping someone can help me.  I
> would like to measure the total flow rate of a few different filters at given
> times.  What device do I need to do this and where can I get it?

I've never seen a hobby-scale device for general use in measuring filter
flows.  Commercially available flow sensors and their necessary circuitry
will probably run you more than \$1000 US.  There are some types of flow
sensors (you might see some built into powerheads) that cost much less,
but I haven't seen those for sale recently for general use.

In the absence of a useful flow meter you need to come up with other
methods of measuring flow.  This is a science fair - type undertaking.

There was a time - years ago now - when I wanted to quantify every aspect
of my tanks.  I spent lots of time measuring various sorts of flows.  It
helped to have an engineering degree and courses in fluid mechanics.

Most of the methods for measuring flow will not measure flow under the
exact conditions you are interested in; any change in inlet or outlet
elevations, constrictions or impediments to flow will change the flow.
If you can't measure flow under exactly the conditions you're interested
in then you may be able to measure unobstructed flow at different outlet
elevations then project or interpolate the flow rates to conditions you
are interested in.

The simplest measuring method is to let the water flow into a container
and measure the time it takes to get a known amount of water (or measure
the volume delivered in a known period of time). The flow rate is then the
volume of water divided by the time required.  This can only be used in a
few instances.

The more general principle behind flow measurement is to measure the
average velocity of water passing through a cross section of known area.
The method you need varies from situation to situation and I think they
mostly require some physics.

One approach that's fairly easy to use (sometimes) is to set the outlet
tube level and measure the horizontal distance traveled while the water
falls through a known drop.  Then you calculate the velocity needed to get
that far before the acceleration of gravity causes it to fall throught the
drop.  The stream of water is usually a little narrower just outside the
tube outlet then it is inside the tube.  If there is a noticable
constriction at that point then you'll need to measure the cross sectional
area of the stream at the constriction and use that in your calculation.
Otherwise, just use the internal cross section of the tube.

For slower velocities you might be able to inject a small amount of food
dye or some other tracer into the flow and measure the time necessary for
the tracer to move a known distance.  This works pretty well for
estimating flow in ditches and small streams, where the tracer might be
nothing more than a blade of grass.  I tried the food dye method in a
vertical tube once and found that the dye was enough heavier than water
that it gave me bad results.

I have used pitot tubes to estimate water velocity at outlets and measured
the pressure drop from the free surface in a tank to inside a tube leading
from the tank to estimate the velocity in the tube.  These methods give
pretty good results, but they depend on bernoulli's equation for fluid
flow and require corrections for various losses.  Venturis provide another
variation on the same theme.  You probably need an engineering background
to use any of these methods

For Aquaclear power filters there should be a way to use the water level
inside the filter to measure flow.  I haven't worked out the details.  I
think it could be based on the standard method of measuring stream flow
with an ogee weir.  Otherwise to measure flow through the box filters you
would need to measure the flow on the intake side.

Have fun.
Roger Miller

```