Hydraulic

Orifice plate sizing in pumping systems and the ΔP at the worst-case flow

An orifice plate measures flow rate through differential pressure and, in pumping systems, imposes a permanent head loss that must be added to the pump total head (TDH) — especially at the worst-case highest-flow condition.

When to use

Use it when the project requires cheap, rugged flow measurement with no moving parts on a pump discharge line — process water, raw water, oil, or condensate. The orifice plate is the default choice when there is enough straight pipe run upstream and downstream and when the permanent head loss it introduces is acceptable within the system energy balance. For pump sizing, the plate's permanent ΔP adds to the total head (TDH); the critical point to watch is therefore the highest-flow scenario, where the loss grows with the square of the velocity and can eat into the design margin.

What an orifice plate does in a pumping system

An orifice plate is a metal disc with a central sharp-edged bore, installed between pipe flanges. By throttling the flow it accelerates the fluid and creates a measurable pressure drop between the upstream and downstream tappings. This differential pressure (ΔP) has a deterministic relationship with the flow rate, which makes the plate the most widely used primary flow element in industry: cheap, with no moving parts, standardized, and auditable against a published code.

In a pumping circuit, however, the plate performs two roles at once that the designer must keep mentally separate. As an instrument, it converts flow rate into a ΔP signal. As a hydraulic fitting, it introduces a permanent head loss that the pump must overcome. Confusing the measurement ΔP with the real loss is the source of the most common sizing error — which is why this guide treats both sides with equal rigor.

Theoretical basis: from Bernoulli to the ISO 5167-2 equation

Applying the energy equation between the upstream section (D) and the throat (d) together with continuity yields the mass flow rate:

qm = (C / √(1 − β⁴)) · ε · (π/4) · d² · √(2 · ΔP · ρ₁)

The term 1/√(1−β⁴) is the velocity-of-approach factor (E): it corrects for the fact that the fluid already arrives with a non-zero velocity. The discharge coefficient C absorbs everything the ideal model fails to capture — jet contraction (the vena contracta), the velocity profile, and local friction. ISO 5167-2 provides C through the Reader-Harris/Gallagher equation, a function of β, the Reynolds number Re_D, and the pressure-tapping type. For liquids, the expansibility factor ε ≈ 1, since there is no appreciable compression.

The most important practical consequence lies in the square root: ΔP ∝ Q². Doubling the flow rate quadruples the ΔP. This nonlinearity is what makes the worst-case highest-flow scenario the design point that governs both the transmitter full scale and the pump margin.

How to size it, step by step

Define the design flow at the worst case — the maximum flow the system can actually reach, not just the nominal value.
Fix the fluid and conditions: ρ₁ and μ at the operating temperature.
Choose the full-scale ΔP compatible with the available cell (often 25–100 kPa).
Iterate on β until the target ΔP is matched: the equation is inverted to solve for d (and therefore β = d/D), with C evaluated at Re_D. Because C depends only weakly on β, the solution is iterative but converges in a few steps.
Check validity: 0.10 ≤ β ≤ 0.75, Re_D above the standard’s minimum, and enough straight run upstream/downstream.
Compute the permanent loss Δϖ with the ISO 5167-1 correlation and convert it to meters of head to add to the TDH.

Practical design considerations

β is a trade-off, not a free number. Between 0.30 and 0.60 lies the best balance. A smaller β improves measurement resolution at the cost of more permanent loss; a larger β eases the pump but degrades accuracy and tolerance to distorted profiles.
Straight run is a requirement, not a recommendation. Upstream bends, valves, and reducers distort the velocity profile and invalidate C. When space does not allow the required 10–44 D, use a flow conditioner qualified by the standard.
The permanent loss is what matters to the pump. Of the measured kPa, only the fraction Δϖ/ΔP (from ~0.8 at β=0.2 to ~0.3 at β=0.7) is real loss. Add only that portion to the total head.
Materials and edge condition. The upstream sharp edge must stay sharp; erosion, deposits, or reversed installation shift C and introduce a systematic error that is hard to detect.

Tie-in with the standards

The method described here is exactly that of ISO 5167 (and the equivalent ABNT NBR ISO 5167-2): Part 1 establishes the general principles and the permanent pressure loss equation; Part 2 details geometry, pressure tappings (corner, flange, D and D/2), validity ranges, and the C equation. For custody-transfer gas metering, the parallel reference is AGA Report No. 3 / API MPMS 14.3. Following the standard is what gives the number traceability: it defines the conditions under which the published discharge coefficient is valid and the expected uncertainty. Stepping outside the normative range — β outside 0.10–0.75, low Re_D, insufficient straight run — means abandoning the statistical basis that guarantees accuracy, and the result is no longer defensible in an audit.

Formulas and fundamentals

Mass flow rate (ISO 5167-2) qm = (C / sqrt(1 - β⁴)) · ε · (π/4) · d² · sqrt(2 · ΔP · ρ₁)

Relationship between mass flow rate and differential pressure. qm [kg/s]; C = discharge coefficient [-]; β = d/D [-]; ε = expansibility factor (≈1 for liquids); d = orifice bore diameter [m]; ΔP = measured differential pressure [Pa]; ρ₁ = upstream fluid density [kg/m³]. The term 1/sqrt(1-β⁴) is the velocity-of-approach factor E.

Diameter ratio (beta) β = d / D

β is the ratio of the orifice bore d to the pipe internal diameter D (same units). ISO 5167-2 restricts 0.10 ≤ β ≤ 0.75. Low β → higher ΔP and higher loss; high β → lower ΔP and lower accuracy.

ΔP at worst-case flow (flow scaling) ΔP₂ / ΔP₁ = (Q₂ / Q₁)²

For fixed β and fluid, the differential pressure varies with the square of the volumetric flow rate Q [m³/h]. Size the plate at the maximum flow (worst case): that is where ΔP and the permanent loss peak and where the ΔP cell must have its full-scale range set.

Permanent pressure loss (ISO 5167-1) Δϖ = [ (sqrt(1-β⁴(1-C²)) - C·β²) / (sqrt(1-β⁴(1-C²)) + C·β²) ] · ΔP

Δϖ is the NON-recovered (permanent) head loss imposed by the plate, expressed as a fraction of the measured ΔP. It is this portion — not the measurement ΔP — that adds to the pump total head. The fraction drops from ~0.8 (β=0.2) to ~0.3 (β=0.7).

Total head increment Δh_plate = Δϖ / (ρ · g)

Converts the permanent loss into meters of fluid column to add to the TDH. ρ [kg/m³]; g = 9.81 m/s². e.g. Δϖ = 46.9 kPa in water → Δh ≈ 4.79 m H₂O.

Pipe Reynolds number Re_D = (ρ · v · D) / μ = (4 · qm) / (π · D · μ)

Re_D sets the validity of C and the flow regime. ISO 5167-2 requires Re_D above minimum limits (a function of β and tappings). v [m/s] = mean pipe velocity; μ [Pa·s] = dynamic viscosity.

Standards & methods

ISO 5167-1 (general principles and permanent pressure loss)
ISO 5167-2 (orifice plates)
ABNT NBR ISO 5167-2
Reader-Harris/Gallagher (discharge coefficient C equation)
AGA Report No. 3 / API MPMS 14.3 (related reference for custody-transfer metering)

Typical reference values

Quantity	Typical range	Note
Recommended beta ratio (β)	0.30 to 0.60	Best trade-off range for accuracy vs. loss; the standard allows 0.10–0.75.
Typical discharge coefficient (C)	0.60 to 0.62	For a sharp-edged orifice at high Re_D; computed via Reader-Harris/Gallagher.
Expansibility factor (ε) for liquids	≈ 1.00	ε < 1 only for compressible gases/vapors.
Pipe velocity (water)	1.0 to 3.0 m/s	Usual discharge range; directly affects ΔP and Re_D.
Permanent loss as a fraction of ΔP	0.73 (β=0.5) to 0.30 (β=0.7)	The higher the β, the smaller the non-recovered portion.
Upstream straight run	≥ 10 to 44 D	Depends on the upstream fitting and on β (ISO 5167-2 table).

Worked example

Plate on a water discharge line (D=100 mm, β=0.5)

Inputs

Design flow (worst case): 50 m³/h
Pipe internal diameter (D): 100 mm
Beta ratio (β): 0.50 -
Density (ρ): 998 kg/m³
Discharge coefficient (C): 0.605 -
Dynamic viscosity (μ): 1.0×10⁻³ Pa·s

Results

Orifice bore diameter (d): 50.0 mm
Pipe velocity (v): 1.77 m/s
Reynolds number (Re_D): ≈1.76×10⁵ -
Measurement ΔP: ≈63.9 kPa
Permanent loss (Δϖ): ≈46.9 kPa (4.79 m H₂O)

With β=0.5 the 50 mm bore produces a measurement ΔP of ~64 kPa (0.64 bar), a comfortable range for a typical ΔP cell (full scale ~100 kPa). The Re_D of 1.8×10⁵ is well above the ISO 5167-2 minimum, validating C. The key pumping takeaway: of the 64 kPa measured, only ~46.9 kPa (a fraction of 0.733) is permanent loss — equivalent to 4.79 m H₂O that must be added to the pump total head. If the system's maximum flow reached 60 m³/h, the ΔP would jump to 64·(60/50)² ≈ 92 kPa and the permanent loss to ~6.9 m H₂O; it is this worst-case scenario that dictates the transmitter full scale and the pump margin.

Run your own calculation Create a free account and get this for your own inputs in seconds.

Common mistakes

Adding the MEASUREMENT ΔP to the pump total head. What goes into the TDH is the PERMANENT loss (Δϖ), typically 30–80% of the ΔP, not the full ΔP.
Sizing the plate at nominal flow and forgetting the worst case. ΔP grows with Q²; at maximum flow the loss can exceed the pump margin and saturate the ΔP cell.
Picking a β too high (>0.65) just to cut the loss: measurement accuracy degrades and sensitivity to velocity-profile disturbances increases.
Ignoring the upstream/downstream straight-run requirement and installing the plate right after a bend, valve, or reducer — the distorted profile invalidates C and introduces a gross error.
Applying a constant C (0.61) without checking Re_D. At low flow rates Re_D can fall below the standard's limit, where C varies and the equation loses validity.
Failing to specify the pressure tappings (corner, flange, or D-D/2). C and Δϖ depend on the tapping type; mixing correlations produces a systematic error.

Frequently asked questions

Does the orifice plate ΔP add to the pump total head?

Yes, but only the PERMANENT portion (Δϖ), not the full measurement ΔP. The fluid recovers part of the pressure downstream; the non-recovered loss — typically 30% to 80% of the ΔP depending on β — is what adds to the TDH. For β=0.5 the fraction is ≈0.73.

Which beta ratio should I choose?

Between 0.30 and 0.60 for most projects. A low β gives a large ΔP (good measurement resolution, but high permanent loss); a high β reduces loss but worsens accuracy and robustness to the velocity profile. The standard allows 0.10–0.75, but the extremes call for caution.

Why size at the highest-flow scenario?

Because ΔP and permanent loss grow with the square of the flow rate (ΔP ∝ Q²). Maximum flow is where the ΔP transmitter can saturate and where the loss imposed on the pump is greatest. Sizing at nominal flow and forgetting the peak is a classic mistake that compromises the pump margin.

How is the discharge coefficient C calculated?

Through the Reader-Harris/Gallagher equation adopted by ISO 5167-2, a function of β, Re_D, and the pressure-tapping type. For a sharp-edged orifice at high Re_D, C is around 0.60–0.62. Do not use a fixed value without checking the standard's minimum Re_D.

How much straight pipe run does the plate require?

It depends on the upstream fitting and on β. ISO 5167-2 tabulates typical lengths from 10 D to more than 40 D upstream and about 4–8 D downstream. Nearby bends, valves, and reducers distort the profile and invalidate C — use a flow conditioner if space is limited.

Is an orifice plate suitable for viscous or two-phase liquids?

For Newtonian liquids with Re_D within the standard's range, yes. For very viscous fluids (low Re_D) the equation loses validity; for two-phase or solids-laden flow the sharp-edged plate suffers error and erosion — in those cases evaluate an eccentric/segmental plate or another element (Venturi, nozzle).

Glossary

Beta ratio (β): Ratio of the orifice bore to the pipe internal diameter (β = d/D). The central parameter governing ΔP, permanent loss, and accuracy.
Discharge coefficient (C): Dimensionless factor that corrects the theoretical flow to the actual flow, accounting for jet contraction and losses. Computed via Reader-Harris/Gallagher in ISO 5167-2.
Differential pressure (ΔP): Pressure difference between the upstream and downstream tappings of the plate, read by the cell/transmitter. It is the signal from which flow rate is inferred.
Permanent pressure loss (Δϖ): The portion of the ΔP that is NOT recovered downstream. It is the real head loss the plate imposes on the system and that adds to the pump total head.
Expansibility factor (ε): Correction for fluid compressibility. It is ≈1 for liquids (incompressible) and <1 for gases and vapors.
Pressure tappings: Measurement locations (corner taps, flange taps, or D and D/2). They define which C and Δϖ correlation to apply; they cannot be mixed.