Most quotation requests we receive are missing at least two of the five numbers needed to answer them. The result is a week of email ping-pong before anyone can give a real price. Here is what actually determines the model, and why each item changes the answer.
One: the fluid, and what it does to a sensor
Not the product name. The physical behaviour.
Does it conduct electricity? This decides whether an electromagnetic meter is even possible. Roughly 5 microsiemens per centimetre is the floor. Mains water is far above it. Deionised water, oils and most solvents are below it.
How viscous is it? Below 10 cSt, turbine and vortex meters behave. Above 20 cSt, look at positive displacement or Coriolis.
Is it clean? Solids and fibres foul electrodes and scatter ultrasonic signals. A full-bore magnetic meter handles them.
Is it single phase? Entrained air breaks almost every technology. Steam is its own category.
Two: three flow rates, not one
Minimum, normal and maximum. If you only send the maximum, any meter we quote will be sized for a condition your plant rarely runs at.
The minimum is the one that matters most and the one most often missing. It determines whether the meter stays inside its linear range during night-time or turndown operation, which is exactly when nobody is watching the reading.
Write the three numbers down before you contact anyone. Without them, any quotation is a guess dressed up as a price.
Three: pressure and temperature
Working pressure sets the flange rating and, for gas, the density used in the calculation. Temperature sets the liner and electrode material on a magnetic meter, and the electronics rating on everything else.
Give the range, not the normal value. A meter rated for 80 degrees Celsius that sees a 150 degree steam-out cycle twice a year will fail, and the datasheet will not warn you.
Four: pipe size, material and connection
Nominal bore is the start, but the internal diameter is what the calculation uses. On an existing line, measure the wall thickness rather than assuming the schedule, especially for clamp-on ultrasonic where the bore error goes straight into the flow reading.
Material matters for grounding (plastic and lined pipe need earthing rings on a magnetic meter) and for the acoustic path (concrete and lined pipe are not suitable for clamp-on).
Five: what the signal has to talk to
Pulse, 4 to 20 mA, RS485 Modbus, HART, or a relay contact. Match this to whatever receives it: a PLC input card, a flow computer, a batch controller or a plant historian.
This is the item most often added after the order is placed, and it is the one that causes the most commissioning delays.
Two more that change the answer
Can the line be shut down? If not, clamp-on ultrasonic is the only serious option and the rest of the discussion is academic.
How much straight run is available? Vortex meters need 15 to 30 diameters upstream. Electromagnetic need 5. On a tight skid this single constraint often eliminates the technically better meter.
A worked example
Cooling water, conductive, clean, 5 to 45 m3/h, 6 bar, 35 degrees Celsius, DN100 steel, 4 to 20 mA to a PLC, line can be shut down, 8D of straight run available.
Conductive and clean points at electromagnetic. The 9:1 turndown is comfortable for a magnetic meter. 8D of straight run is enough for the 5D requirement. DN100 at 45 m3/h gives about 1.6 m/s, inside the usual 0.9 to 2.5 m/s band for liquids. The answer is a DN100 flanged electromagnetic meter, and it took less than a minute because all five numbers were present.
You can check that velocity figure yourself with the flow rate and pipe velocity calculator.
Related reading
For the technology comparison behind step one, see electromagnetic vs ultrasonic flow meter. For what happens when a meter is sized wrong, see why a flow meter reads wrong.