Obviously, the flow profile would be affected by the piping layout and fittings both upstream and downstream of the orifice. Distortion of flow profile affects flow measurement. It is relatively easy to visualize this concept.
There is detailed information in API MPMS Chapter 14.3. See also
if we won't provide the suffice horrizantal run up stream to the orifice flow meter ( considering fluid is passed through bends and fittings up stream to orifice) the flow profile would get distorted and flow element won't show accurate results.But how come flow profile down stream to the orifice will effect the accuracy of measuring?
The flow is distorted downstream also if you have bends etc. To get accuracy the flow needs to be un-distroted upstream and downstream of the measuring device.
Flow at any point along a pipe (at the meter) is a function of BOTH upstream pressure AND downstream pressure. And both of those need to be kept as constant as possible while measurement is in progress.
Think of it like the follow through in a golf shot. If you don't do it right you get lousy results even though the ball is already gone when it starts.
I've always thought that the downstream distances in 14.3 were far too long, but without any downstream straight pipe (e.g., attaching a 90 elbow to the outlet of the meter) the stagnant flow areas in the downstream annular space convert to von Karmaan Streets that do result in an increase in the downstream pressure which makes flow read low. We tested this at CEESI in the 1990's but never got permission from the lawyers to publish it. That test was a tiny part of a much bigger test and we didn't have the time to experiment to find out where the influence stopped. It likely that 2D downstream of the meter the elbow wouldn't have mattered, but we didn't test that.
