So you are measuring compressed air flow? Great! Now you can generate some serious savings. But wait! You have installed three meters downstream of a main flow meter, and you add them all up? and you see that the sum of these is not equal to the main meter? in that case, read on as you are not alone. And we have some great tips on how to minimize this effect.
Causes of deviation
There are a few things that can cause deviation between sensors in general, and flow meters in particular.
The specified measurement uncertainty of the productsThis means that two or more sensors are always deviating from each other. For example, when the measurement uncertainty of both sensors is +/- 5% of reading, there is a chance that two flow meters deviate 10% max. There can also be random variations when you look at the flow meter display, as a result of small fluctuations in the actual flow. This makes simultaneous readout of multiple displays "by the eye" a very demanding task. If you want a proper comparision, always log the output to a file.
The rangeabilty of the sensor - full scale error
The rangeability, or dynamic range varies per sensor. A vortex meter has 1:15 range. A DP meter 1:10. A thermal mass flow sensor 1:100 or more. So if you are comparing one to the other, and one of them is in the low part of the range, you might find large deviations. Most manufacturers state "% full scale" which means if you have a 1% full scale sensor, it will deviate up to 10% of reading at 10% of the range. in case of insertion probes, always check the velocity, to relate readings to full scale.
The diameter might be incorrect
When you enter the diameter, you sometimes don't know what it is, as the pipe can be old, and you cannot look inside the pipe. Rust, or fouling inside the pipe can change the inner diameter over time. Use of an ultrasonic thickness gage can help to minimize this problem.
Up- and downstream length
Wrong installation (not enough length/not inserted in the middle/not aligned with the flow direction) will have a large effect on readings. Check the piping table in the user manual for details. As an example, a double elbow can create a large distortion, and you will need at least 40 x D to stabilize the flow profile.
Dryers purge (air loss over the dryers)
Often overlooked: Dessiccant driers purge. Ideally between 5% and 10%. But we have seen cases with 20% purge. Check this always before concluding that the flow meter is wrong.
Lagging/ leading readings due to system capacity
The entire compressed air system is a network of storage and pipes. The storage (which may be the pipe itself) can cause lagging of signals. A peak detected by a sensor upstream can be ahead of one downstream, or not seen at all due to the buffer in between. When comparing flow meters, always allow enough time to average out these effects.
What to do next?
Some tips and tricks to reduce the uncertainty and maximize the field accuracy of your flow meters:
- Pay some respect to nature: Up and downstream length (after a bend it needs to be 30-40D before
- Make sure the sensor's measurement range fits the application well
- Check the insertion depth, make sure the probe is IN the pipe, and in the middle
- Align the probe properly with the flow direction. Make sure it is not reversed
- Configure the inner diameter. Is this correct? Did you measure the wall thicknes?
- Check if the dryer has a purge or perhaps air losses in the drain.
Further reading: See the user manuals of the respective flow meters for additional information on the required pipe length.