Vortex flowmeter sizing is a velocity-window problem, not a pipe-size problem, and the four inputs that drive every calculation are fluid type, operating pressure (or temperature for superheated steam), pipe inner diameter, and min/normal/max mass or volume flow [S1]. A typical industrial vortex body is offered in flange sizes from 1/2 in. to 12 in., with wafer bodies the lowest cost and flanged bodies preferred for hazardous or high-temperature service [S2].
The working velocity window for the STLU-G series vortex flowmeter is 2 to 76 m/s, with an extended low end down to 1.98 m/s [S1]. Outside that band the sensor either stalls or generates excessive pressure drop, which is why the meter is almost always ordered one to four sizes smaller than the upstream pipe, a practice that recurs across steam, gas, and low-flow liquid duty. A 1.98–76 m/s window, combined with the 10:1 to 15:1 turndown published for inline vortex bodies [S3], defines the entire operating envelope buyers have to work inside.
The four numbers you cannot skip before any calculation
For steam service the four mandatory inputs are steam type (saturated vs superheated), operating pressure in MPa(g) or bar(g), pipe inner diameter in mm, and flow range expressed in kg/h or t/h at min, normal, and max load [S1]. Saturated steam density is a function of pressure only: 0.4 MPa(g) saturated steam sits at about 1.39 kg/m³, while 1.0 MPa(g) saturated steam climbs to about 2.95 kg/m³, which is why two steam lines at the same volumetric flow can carry very different mass flow [S1]. Superheated steam needs both pressure and temperature resolved against a superheated table or the IAPWS-IF97 formulation [S1].
For liquid service, the equivalent inputs are fluid density and viscosity (passed to the sizing software), the pipe ID, and the volumetric flow range; the manufacturer-supplied sizing routine handles the unit conversion internally once those values are entered [S2]. Where real flow data is missing, the boiler-rating fallback is 80% of nameplate capacity as normal and 110% as maximum, a coarse but usable starting point [S1].
Velocity window, turndown, and why the meter is almost never line-size
The first arithmetic step is converting mass flow to volumetric flow with Qv = Qm / ρ, then computing velocity with v = Qv / (3600 × A), where A is the pipe cross-section in m² derived from the inner diameter [S1]. That velocity is then compared against the meter's published working range, which for STLU-G is 2 to 76 m/s standard and down to 1.98 m/s extended [S1]. Below 2 m/s the practical remedy is to step the meter down one size, accept the extended low range, or change technology, while above 76 m/s the fix is to step up a size or accept the higher permanent pressure loss [S1].
Turndown is the second gate and is often the binding constraint. A published rule-of-thumb for inline vortex bodies is roughly 10:1 turndown with a low-flow cutoff near 1/10th of maximum flow [S3]. In a worked grinding-oil example, the operating range 150 to 2300 L/min on an ND150 (DN150, 162.3 mm ID) line gives a turndown of 15.3:1, and the calculated velocity band is 0.121 to 1.852 m/s, which sits below the typical vortex minimum of 0.5 to 1.0 m/s and only at the very low end of the paddle-wheel range of 0.15 to 0.3 m/s [S3]. At 150 L/min, around 6.5% of maximum, many vortex bodies will simply read zero, regardless of actual velocity, so this is the case where a Coriolis, magnetic, or differential-pressure alternative starts to make sense [S3].
Vortex vs paddle wheel vs orifice vs Coriolis: a four-criteria comparison
The decision matrix for a clean industrial liquid is short. Vortex inline bodies are typically ±0.5–1% of full scale, paddle wheels ±1–2% of full scale, with vortex bodies carrying no moving parts while paddle wheels use exposed bearings that fail in dirty service [S3]. Vortex units need 20+ diameters of straight upstream pipe versus 10–15 diameters for paddle wheels, an installation cost that often decides the call on tight retrofit skids [S3].
For steam specifically, the matrix is vortex (2 to 76 m/s, no moving parts, mid cost) versus orifice plate (turndown about 3:1, limited by differential pressure), Coriolis mass meter (true mass, high accuracy, highest cost and pressure drop), and Annubar/Pitot (10 to 80 m/s, low permanent loss, lower accuracy) [S1]. The vortex vs electromagnetic flowmeter sizing logic decision comes down to conductivity: magnetic meters require a conductive liquid and offer excellent low-flow performance, while vortex meters handle steam, gas, and non-conductive liquids but lose accuracy below roughly 0.5 m/s. For most clean, conductive liquids above about 0.5 m/s, the magnetic meter is the cheaper, lower-turndown answer; for steam, gas, or non-conductive hydrocarbons, the vortex body is the default.
Body materials, sizes, and the wafer-vs-flanged call
Standard vortex body construction is 316 stainless steel with Hastelloy as the common upgrade for corrosive chemistry, and a typical size range runs 1/2 in. to 12 in. flanged, with wafer (flangeless) bodies the lowest-cost option and flanged bodies specified whenever the process fluid is hazardous or the operating temperature is high [S2]. Bluff-body geometry is essentially standardized: the body width must be a large enough fraction of the pipe diameter that the full flow participates in shedding, the upstream face needs sharp protruding edges to fix the separation lines across the flow range, and the body length in the flow direction must be a defined multiple of the body width [S2].
Sensor choice inside the body is now almost universal: piezoelectric or capacitance-type detectors read the pressure oscillation around the bluff body, and linearity, low-Reynolds behaviour, and sensitivity to velocity-profile distortion vary only slightly with bluff-body shape [S2]. For a chemistry-skid retrofit on aggressive media, the practical spec is flanged 316SS body, Hastelloy bluff body, remote-mounted electronics, and 20+ diameters of upstream straight run, which is the same envelope described in the vortex flowmeter spec map for steam, gas, and liquid duty.
Five real selection traps and the remedies that actually work
Trap one: "DN100 pipe, send a DN100 meter." A DN100 saturated steam line at 0.4 MPa(g) carrying 800 kg/h normal flow has a density of about 1.39 kg/m³, a volumetric flow near 575 m³/h, and a velocity around 2.0 m/s, which sits right at the lower edge of the 2 to 76 m/s window, so a DN80 body is the safer call [S1]. Trap two: ignoring superheat. A superheated steam line at 1.0 MPa(g) and 300 °C has markedly lower density than the saturated case at the same pressure, so the velocity climbs and the meter can overspeed even at the same mass flow.
Trap three: trusting the boiler nameplate. The 80% normal / 110% max fallback gives a usable envelope but masks seasonal swing, so a saturating-day minimum below 2 m/s will stall the body [S1]. Trap four: dirty liquid on a paddle wheel. Exposed bearings fail quickly in oily or particle-laden service, which is why vortex is preferred over paddle wheel for grinding oil at the cost of higher minimum velocity, around 0.5 to 1.0 m/s versus 0.15 to 0.3 m/s for paddle [S3]. Trap five: undersized upstream straight run. A vortex meter installed with less than 20 diameters of upstream pipe distorts the velocity profile, degrades the Strouhal relationship, and pushes the indicated K-factor off calibration [S3].
Where the vortex flowmeter is the wrong tool
Vortex is not the right tool when the fluid is a clean, conductive water-based liquid with substantial low-flow demand, where an electromagnetic flowmeter sized into the same line will hold accuracy down to near-zero flow at lower installed cost. It is also the wrong tool for two-phase flow, slurries with large solids that can momentarily stall shedding, or any application where minimum velocity stays under roughly 0.5 m/s for sustained periods, which is the regime that breaks the published 10:1 to 15:1 turndown envelope [S3]. For very low flow on a large line, the standard remedy is to reduce the meter body two to four sizes below the line, accepting a moderate to significant pressure-drop increase that must be checked against pump head or steam-supply pressure [S3].
Trackable signals for the next planning cycle: the STLU-G series volumetric flow tables on silverinstruments.com (saturated-steam mass flow by DN class) and any vendor update to the extended low-velocity limit below 1.98 m/s; published minimum-velocity revisions for inline vortex bodies on grinding-oil and similar low-flow liquid service; and the IAPWS-IF97 reference implementations used by OEM sizing software for superheated steam.
Spec-level background on the components involved: linear guide, and crossed roller guide.