Water Measurement Question

Emerson and Rosemount both come to mind. What you want is commonly done and I suggest talking directly to an engineer or product specialist.

Would a simple float switch work instead?

Can you do the measurement inside the mixing container (ultrasonic measurement), or does it have to be done before or as the water is added?

Was going to suggest the same..

I would couple this to a scale measuring tank weight for accuracy...

I don't have a suggested supplier, but I have some brain candy for you.

Fuel dispenser mechanism. The faceplate reports fuel dispensed, (presumed) accurate to 0.01 gallon. I know the current bowser versions can be set to dispense a specified volume, then stop.

My first thought was load cells. Also considered a tank of known value (100 Liter) when full, that used to fill your batch tank. However it sort of sounds like volumes change, so likely not practical.

15gpm?

As Lyn suggested in #1, Rosemount or other water meter suppliers should be suited.

"Domestic" water meters (dial type) that we install for reticulation are guaranteed +/- 2% when new.

We have a few electronic ones (Reset when close cover) that we use for calibration at +/-0.2% at around $500 AUD.

We also have a calibrated (traceable certificate similar to USA NIST) at around $7k. (Probably 2k for meter and 5k for lab fees)

Perhaps a better approach would be to set up a 100L tank on an electronic scale. Zero the scale with the tank on it, fill the tank to weight and then pressurize the tank with air to quickly unload it or use a large cross section gate valve to gravity dump. Faster, more accurate and the batch fill can be recorded for record purposes if the scale has a logging output feature.

I was thinking along similier lines in using a 100Lt tank which can be filled to an overflow point then dumping the contents into the containers assuming the tank can be positioned to suit. But your idea of using air pressure would work easier just open and close appropriate valves. No weighing would be needed.

FLOW METER TURBINE, WITH SHUT-OFF DEVICE

17.007.-

purports to have +/-3% accuracy, depending how well you calibrate your pulse counting circuit.

"The combination of a flow meter and a solenoid valve in one compact unit enables, with suitable electronics, the accurate metering of a liquid together with the ability to interrupt the flow. A servo controlled valve avoids water hammer when connected to the mains supply. This unit can be used for measuring, feedback control or dosing as it ensures accurate measurement of liquid volumes. The simple internal pulse transmitter offers a long life time without risk of failure. To avoid air entry causing inaccurate measurements, the flow meter is sited upstream from the solenoid valve. The flow meter's output are pulses, where the frequency or the number of pulses are equivalent to flow rate or volume. Depending of the orifice used different measuring ranges can be achieved."

See: http://www.akmueller.de/en/products/product-detail/17007-flow-meter-turbine-with-shut-off-device/

Data sheet: http://www.akmueller.de/fileadmin/user_upload/Produkte/PDFs/17_007_en.pdf

Your lab's electronics technician or engineer should be able to make the pulse counter and control interface, and display, with a simple discrete logic circuit or microcontroller board.

You will need to calibrate the pulse count to your local water utility supply parameters. You may need to consider a coarse fill tank upstream of the flow meter assembly if your utility's supply parameters vary a lot.

Just curious. What is the problem if you get it wrong?

Is it to avoid 'giving away' an expensive product, or to avoid over-spill in production (eg, a hazardous substance) - or if 'short' - how will your customers know.

I buy petrol at the pump but I have no idea how accurate the gauge is. I just have to accept the pump is calibrated to traceable compliance standards.

On the other hand, I buy a bottle of milk or beer and can see instantly whether I get a pint or not. Assuming the milk bottle or beer glass is guaranteed '1 pint' of course.

What if you used weight to fill the containers. Put your container on a scale, tare out the container then add 100 kg of water. Should be highly accurate depending on the scale.

Even if your scale is not calibrated, measure precisely 100 l of water into the container on the scale one time and record the scale reading. Then fill to that reading every time afterwards.

Turn your volume problem into a mass problem.

Weigh it.

Bazzer

I would suggest trying a graduated 100 liter container mounted above the filling station and let gravity do the work. A mark on the tank would be your reference point and you should be able to get close to your 1% target. Ryan Herco has such polypropylene tanks that are translucent. Some have covers which might help reduce airborne contamination issues. With your present system, I would find it very hard to get 100 L repeatedly over the course of a day, to anything better than 5%.

I want to thank all who responded for your time and consideration. I always seem to be looking for something that isn't available yet.

Like I commented before, weigh the water. Scales are definitely available and highly accurate.