Orifice Plate Sizing

An orifice measures flow by trading pressure for velocity. Force the stream through a smaller hole and it must speed up; by Bernoulli, that rise in velocity shows up as a drop in static pressure. Measure the pressure difference across the plate and you can infer the flow, because the differential rises with the square of the flow rate. That square-law relationship is why an orifice has limited turndown — at low flow the differential becomes too small to read accurately.

The real flow is always a little less than ideal Bernoulli predicts, because the jet keeps contracting just past the plate to a narrowest point called the vena contracta, and because friction takes a cut. The discharge coefficient C bundles those effects into a single multiplier near 0.6. ISO 5167 gives C through the Reader-Harris/Gallagher equation, fitted to thousands of calibration runs, as a function of the beta ratio, the pipe Reynolds number, and where the pressure taps sit.

Because C depends on the Reynolds number, an orifice calibrated at one flow drifts slightly at very different flows, and below a minimum Reynolds number it is no longer validated at all. For most liquid and gas services that run comfortably turbulent this is a non-issue, but for viscous fluids or very low flows it is the first thing to check. The square-law and the Reynolds dependence together define the usable measurement window.

Everything about an orifice plate is governed by one number: the beta ratio, the bore divided by the pipe diameter. It is the single design choice that sets the differential pressure, the permanent loss, the measurement uncertainty, and the straight-run requirements all at once, which is why ISO 5167 brackets it so tightly.

Push beta low — a small hole in a big pipe — and you get a large, easy-to-read differential and a coefficient that is stable and installation-tolerant. The price is a steep permanent pressure loss, because most of that differential is never recovered, plus a bore small enough to foul or erode. Push beta high — a hole nearly as big as the pipe — and the permanent loss falls away, but the coefficient becomes touchy about the upstream velocity profile, the required straight runs lengthen, and the uncertainty climbs.

The 0.2 to 0.6 band most engineers design within is the compromise where none of those penalties is severe. Within it, you tune beta by choosing the differential-pressure range of the transmitter: a wider DP range at the same flow gives a smaller bore and a lower beta. That is the practical lever — pick a sensible DP full-scale, size the bore, check where beta landed, and nudge the DP range if it strayed outside the band.

An orifice only measures correctly if the flow arriving at it is fully developed and swirl-free. That is why ISO 5167 specifies minimum straight lengths of pipe upstream and downstream, and why those lengths depend on both the beta ratio and the nature of the nearest fitting — a single elbow needs less run than two elbows in different planes, which generate swirl. Where the straight run is not available, a flow conditioner can substitute for much of it.

The plate itself has to stay sharp. The square edge of the bore is what fixes the discharge coefficient; once it rounds off from erosion or corrosion, the coefficient drifts and the meter reads low. In dirty or abrasive service this is the dominant maintenance concern, and it is another argument against very small beta ratios, where the edge sees the highest velocity. Plates are routinely pulled and inspected for exactly this reason.

Finally, remember what the orifice costs you in energy. The permanent pressure loss — roughly (1 − β¹·⁹) of the measured differential — is pumping or compression work spent purely to make the measurement. On a large continuous line that loss runs all year, and the lifetime energy bill can justify a lower-loss device such as a venturi or an averaging pitot. The orifice wins on capital cost and code support; on operating cost it does not always.