Orifice plates, venturis, flow nozzles, and averaging pitot tubes all look different, but they measure flow the same way underneath: put a restriction in the line, and the pressure drops across it by an amount that grows with flow. The link between that pressure drop and the flow is not a straight line - it is a square-root relationship that comes straight out of Bernoulli's principle. This guide explains the shared physics behind every differential-pressure meter, why the square root makes DP flow inherently nonlinear, and why that nonlinearity limits turndown on the gas measurement a SCADA system depends on.
DP flow principle in one line: The differential-pressure (DP) flow principle measures flow by placing a restriction in the pipe and reading the pressure drop it creates. From Bernoulli's principle, as fluid speeds up through the restriction its pressure falls, and the resulting differential pressure is proportional to the square of the flow rate. Flow is therefore proportional to the square root of the measured differential, which is why every DP meter needs square-root extraction and why its usable range at the low end is limited.
Bernoulli's principle says that for a flowing fluid, where velocity rises, pressure falls, because the fluid's total energy is conserved between its pressure and its motion. A DP flow element exploits this by narrowing the flow path. To pass the same amount of fluid through a smaller opening, the fluid must speed up, and as it accelerates through the restriction its pressure drops. Downstream, where the pipe opens back up, much of the pressure recovers. The instrument measures the difference between the pressure upstream of the restriction and the pressure at the throat, and that differential is the raw measurement.
The relationship between that differential and the flow is the crucial part, and it is not linear. Because kinetic energy scales with the square of velocity, the pressure drop scales with the square of the flow rate. Turn it around and flow is proportional to the square root of the differential pressure. Double the flow and the differential quadruples; run at ten percent of full flow and the differential collapses to just one percent of its full-scale value. Every DP meter therefore emits a signal that must be square-rooted to become a linear flow reading, an operation called square-root extraction.
This one square-root law is what unifies the whole family. An orifice plate creates the restriction with a sharp-edged hole, a venturi with a smooth converging-diverging taper, a flow nozzle with a rounded contraction, and an averaging pitot by sensing the impact and static pressures across the pipe - but every one of them produces a differential that goes as the square of flow. The devices differ in how much permanent pressure they cost, how much straight pipe they need, and how they handle dirty fluid, yet they share the identical governing physics. Learn the square root once and it applies to all of them.
The square-root law is the source of DP flow's defining weakness: limited turndown, meaning the ratio between the highest and lowest flow a meter can measure accurately. The problem lives at the low end. Because differential falls as the square of flow, low flows produce tiny differentials that shrink into the noise, drift, and resolution limits of the pressure transmitter. At ten percent of full flow the differential is only one percent of full scale, and at that level a small absolute error in the pressure reading becomes a large percentage error in flow once the square root magnifies it.
The square root also stretches errors unevenly across the range. Near full flow, where the differential is large, a given uncertainty in pressure translates into a small flow error. Near the bottom, where the differential is minuscule, the same pressure uncertainty translates into a large flow error, because the derivative of the square root is steep near zero. This is why a DP meter's accuracy is usually quoted over a limited span rather than the full range, and why pushing a single orifice run much below roughly a three-to-one or four-to-one turndown erodes confidence in the low-flow reading.
Gas measurement feels this acutely. Gas flow at a well or gathering point can swing widely between high production and near-shut-in, and the low end is exactly where the square-root law is least forgiving. Operators extend the range by staging multiple runs, switching transmitter spans, or choosing a different technology entirely for wide-ranging service, but none of that changes the underlying square root - it only manages around it. Understanding that the nonlinearity is baked into the physics, not a defect of a particular meter, is what makes those trade-offs make sense.
In a SCADA system, DP flow rarely arrives as a single flow number - it arrives as its component measurements, and often the computation happens close to the physics. A flow computer or RTU reads the differential pressure across the element together with the static line pressure and the fluid temperature, applies the square-root extraction, and corrects for gas density using the pressure and temperature. That combination is what turns a raw differential into a compensated mass or standard-volume flow suitable for allocation and custody transfer, and it is why a DP gas meter is really a cluster of tags rather than one.
Historizing those inputs, not just the final flow, is what keeps the measurement trustworthy. A cloud SCADA platform such as Merobix can trend the differential pressure, static pressure, and temperature alongside the computed flow, so an engineer can see whether a low or erratic flow reflects real conditions or a differential that has sunk into the transmitter's noise floor at the bottom of the square-root curve. When a plugged tap or a frozen sensing line distorts the differential, the raw DP trend usually shows it before the flow number looks obviously wrong.
The nonlinearity also shapes how alarms and thresholds should be set on these tags. Because low flows compress into a narrow band of differential, an alarm placed too near zero may sit inside the region where the meter can no longer resolve flow reliably. Setting SCADA thresholds with the square-root behavior in mind - and keeping normal operation up on the part of the curve where the meter is confident - keeps the flow data meaningful across the wide swings a gas well or gathering system routinely sees between full production and shut-in.
Because the pressure drop across a restriction grows with the square of the flow rate, a consequence of Bernoulli's principle. That means flow is proportional to the square root of the measured differential, not to the differential itself. Every DP meter therefore needs square-root extraction to convert its raw signal into a linear flow reading.
The square-root law makes the differential collapse at low flow - at ten percent of full flow the differential is only one percent of full scale. At those tiny differentials, transmitter noise and drift become a large percentage of the reading once the square root magnifies them, so accuracy falls off sharply at the low end. This limits a single DP element to a modest ratio between its highest and lowest reliable flow.
Yes, at the level of principle. All of them create a restriction that produces a pressure drop proportional to the square of flow, so all obey the same square-root relationship. They differ in how much permanent pressure loss they cause, how much straight pipe they need, and how well they tolerate dirty fluid, but the governing physics is identical.
Merobix reads your field devices into a cloud SCADA - the real thing behind these terms, live in days from any browser.