Venturi Tube

What a Venturi Tube Is and What Is Being Decided

Every differential pressure flowmeter is a pair: a primary element that creates a measurable pressure difference in the line, and a secondary instrument, almost always a differential pressure transmitter, that reads it and reports flow. The venturi tube is the primary element with the gentlest footprint in the family. Where an orifice plate flowmeter forces the stream through a sharp edged bore and lets the jet tear itself apart downstream, the venturi guides the same acceleration inside cast or machined surfaces, so the flow never separates from the wall until the geometry releases it. That single difference drives everything a buyer specifies: a higher and steadier discharge coefficient, a permanent pressure loss measured in single digit percentages, and a tolerance for solids and scale that would shift a sharp edge calibration within weeks.

The geometry is standardized, not a manufacturer preference. ISO 5167-4 defines the classical venturi as a cylindrical inlet section with the same diameter as the pipe, a convergent cone with an included angle of 21 degrees, a short cylindrical throat, and a divergent recovery cone with an included angle between 7 and 15 degrees. The upstream tap sits in the inlet cylinder and the low pressure tap in the throat, either as single bores or as piezometer rings averaging several ports around the circumference. Total body length lands between roughly 3 and 8 pipe diameters depending on type and size, which is why replacing another element with a venturi is a piping change and not a drop-in swap.

The device carries Giovanni Venturi's name for the eighteenth century laboratory observation that pressure falls where a channel narrows, but the meter itself is Clemens Herschel's engineering: his 1887 registration of the principle as a commercial water meter for municipal mains. That origin still describes the product's center of gravity. A venturi tube is bought where a line runs continuously, where the fluid carries energy the plant pays for twice, and where a service life measured in decades is a plausible procurement assumption. Buyers on intermittent small lines rarely get the capital back and should stop reading here.

The boundary against its neighbors is worth fixing before any comparison table. Against the orifice plate flowmeter, the venturi trades first cost for coefficient stability, pressure recovery, and solids tolerance; the orifice remains unbeatable where the line is clean, the budget thin, and the element disposable. Against the flow nozzle, the venturi keeps a full recovery cone and loses far less pressure, while the nozzle is shorter, cheaper, and built for steam and high velocity duty where its rounded entry resists erosion; on clean dry steam a vortex flowmeter is often the lighter instrument still. And when the commercial settlement is by mass on a small line, a coriolis flowmeter displaces the whole differential pressure question. Within the differential family, the venturi is the only classical geometry that recovers the majority of the measuring differential, which is why this page treats recovery percentage as a headline number rather than a footnote.

What is actually being decided in a venturi RFQ is three commitments. First, an energy commitment: the buyer is paying capital up front to buy back pumping or compression power over years, so the permanent loss figure and its basis must appear in the quotation. Second, a geometry commitment: the beta ratio computed on the as measured pipe bore, the tap arrangement, and the manufacturing class fix the calibration basis for decades, so a late pipe schedule change invalidates the sizing. Third, an installation commitment: the element arrives as a spool that is welded or flanged into the line, so lifting, alignment, and tap routing are part of the purchase. A quotation that does not address all three is incomplete.

How a Venturi Tube Measures Flow

The measurement rests on two conservation laws acting on the stream. Continuity says velocity rises in the narrow throat in inverse proportion to area; Bernoulli says static pressure falls by the kinetic energy gained. Combined, the standard result is that mass flow equals the discharge coefficient, times the throat area, times the velocity of approach factor, times the square root of twice the differential pressure divided by the density. Procurement never runs this calculation by hand, sizing software does, but three of its inputs sit directly on the datasheet: the coefficient C, the throat diameter d that forms the beta ratio with the pipe bore D, and the density at operating conditions.

Beta is the ratio of throat diameter to pipe bore, written beta = d / D, and ISO 5167-4 admits classical venturis from 0.4 to 0.75. Low beta gives a strong differential signal at the price of higher throat velocity and more permanent loss; high beta recovers pressure best but produces a small differential that instruments struggle to resolve at low loads. Most purchased units sit between 0.5 and 0.7. A useful specification habit is to state the expected full scale differential, commonly 25 to 100 kilopascals in water service, and have the supplier prove the beta choice against it rather than the reverse.

The discharge coefficient is where the venturi earns its price. ISO 5167-4 assigns C = 0.984 to the as-cast convergent type, C = 0.995 to the machined convergent type, and C = 0.985 to the rough fabricated welded type, with published uncertainties of 0.7, 1.0, and 1.5 percent respectively. The numbers are high because there is no sharp edge and no free jet: the flow follows guided surfaces, so little energy is lost between the taps and the coefficient barely moves with Reynolds number inside the valid range. A sharp edged plate, by contrast, works near C = 0.60 with a coefficient that walks as its edge dulls or fouls.

The recovery cone is the economic organ of the device. Downstream of the throat, the divergent section decelerates the flow across its 7 to 15 degree included angle and converts kinetic energy back into static pressure; near 7 to 8 degrees the boundary layer stays attached and recovery is best, while 15 degrees shortens the body and gives up a few more points. Across the family, the permanent pressure loss, meaning the pressure that never comes back, lands between 5 and 15 percent of the measuring differential. A sharp edged plate of equal duty loses 50 to 60 percent and a flow nozzle about 30 to 40 percent, numbers that chapter 5 turns into kilowatts and money.

The secondary instrumentation deserves equal discipline. The differential pressure transmitter is ranged to the full scale differential with turndown held in reserve for future load growth, and its zero is proven against a manifold that equalizes both sides. Because the flow equation consumes density, gas and steam installations add a pressure transmitter on the line plus a temperature input for density compensation, and liquid installations near a pump or a control valve benefit from a pressure sensor logging static pressure so the manifold is never driven over range. Impulse lines between taps and instruments are small bore piping with real failure modes, and chapter 6 treats them as part of the meter rather than as tubing belonging to somebody else's scope.

Types, Construction, and Duty Boundaries

ISO 5167-4 sorts classical venturis into three classes by how the flow passages are made, and the class sets diameter range, coefficient, and uncertainty in one stroke. As-cast convergent venturis pour the inlet, convergent, and throat sections integrally in molds and machine only the tap faces and end connections; they span nominal 100 to 800 mm. Machined convergent venturis finish the inlet, the convergence, and the throat to fine tolerances in cast or fabricated bodies and top out near 250 mm. Rough fabricated welded venturis roll plate into cones, weld the assembly, and serve the large water and effluent mains from 100 mm to 1,200 mm and beyond.

The as-cast type is the volume product of water works. Its cast surface meets a roughness limit rather than a finish specification, and the standard rewards it with C = 0.984 and 0.7 percent uncertainty across Reynolds numbers from 200,000 to 1,000,000 and beta from 0.4 to 0.75. Cast iron and ductile iron bodies dominate potable and raw water, and epoxy or fused coatings protect steel variants in effluent. Buyers should note that the tightest published uncertainty in the differential pressure family belongs to this mid market type, which is why calibrated accuracy claims on water mains usually trace back to as-cast geometry.

The other two classes buy different advantages. The machined convergent type buys coefficient headroom with machining time: C = 0.995, the highest in the ISO 5167 family, valid from Reynolds 100,000 to 500,000 in pipes of 50 to 250 mm, which suits skid mounted hydraulic circuits, test stands, and lines too small for a stable casting. The rough fabricated type accepts C = 0.985 with 1.5 percent uncertainty from Reynolds 200,000 to 2,000,000 and beta 0.4 to 0.7; on a 1,000 mm main the body is rolled steel plate, the ends are butt weld or flange, and a few tenths of a percent of uncertainty matter less than the civil works of installing the spool at all.

One relative belongs in this chapter as a paragraph, not as its own entity. The venturi nozzle, standardized in the nozzle part of ISO 5167 rather than with the classical venturis, carries a rounded convergent profile into a shortened throat and adds a compact divergent outlet that recovers more pressure than a bare nozzle, though less than a full classical venturi body. Specifiers meet it on medium liquid and steam lines where a classical venturi will not fit the footprint and a plain nozzle gives up too much pressure. It carries its own beta, diameter, and Reynolds envelope and its own coefficient treatment, and a buyer who needs one should specify it by that clause text rather than as a venturi variant.

Construction details finish the type map. Taps are either single bores in the inlet and throat or piezometer rings averaging three to eight ports, and ring averaging earns its cost on large diameters where flow distortion from an upstream bend would otherwise bias one side of the throat. End connections divide between flange joints, typical to about 600 mm, and butt weld spools above that. Materials follow the fluid: cast iron and epoxy lined carbon steel for water, WCB and 316L trim for treated and brackish duties, duplex grades where chloride attack rules. The duty boundaries are equally concrete: nothing useful exists below 50 mm, settling slurries demand a velocity review against throat erosion, and on conductive water below the diameters where a venturi pays back, an electromagnetic flowmeter is the honest alternative priced in the same study.

Key Specification Parameters

The datasheet lines that decide whether a quoted venturi matches the application are few, and the table below pairs each with what it settles. The paragraphs after the table cover the clauses that weak quotations omit.

Two rows deserve expansion. The discharge coefficient clause must name the ISO 5167-4 manufacturing class and state whether the value is the standard figure or a wet calibration; laboratory proving on water lifts the accuracy claim from the 0.7 to 1.5 percent family band toward half a percent and is bought only where custody or fiscal settlement justifies it. The bore clause must pin the as measured pipe internal diameter at the meter station, schedule and all, because beta is computed on that bore, and a schedule change after design silently moves beta, the differential, and the indicated flow. The orifice plate flowmeter shares exactly this bore trap, which is why experienced specifiers write the measured bore and its tolerance into the datasheet for both elements.

The comparison table places the venturi against the two primary elements it is most often bid against. Read it by rows, not by columns: the venturi wins the loss and solids rows, the sharp edged plate wins first cost and delivery, and the nozzle holds the steam and high velocity middle ground. The straight run row deserves emphasis in retrofit work, because 0 to 5 diameters upstream is frequently the difference between fitting a meter into an existing pump room and re-routing the pump room.

The ordering traps collect in one paragraph. Nominal diameter is not measured diameter, and at 600 mm a schedule change can move the bore by more than 10 millimeters, enough to shift beta by about 2 percent and the calibration with it. Tap orientation is a fluid question, horizontal for liquids and top mounted for gases, and it must be stated so the spool arrives drilled correctly. Gaskets follow the pipe flange standard, and finally the differential range must be agreed with the instrument supplier, because a venturi sized for 100 kilopascals full scale paired with a transmitter ranged to 10 will never resolve low loads no matter how good the element is.

Selection Decision Factors and RFQ Language

The energy argument is arithmetic, not opinion. Take a 600 mm raw water main moving 4,000 cubic meters per hour. A venturi with a 10 kilopascal permanent loss absorbs about 11 kilowatts of hydraulic power; a sharp edged plate on the same duty at 60 kilopascals absorbs about 67 kilowatts. At 75 percent wire to water efficiency and 8,400 running hours a year, the difference is roughly 630,000 kilowatt hours per year, which at typical industrial tariffs pays back the venturi premium in one to three years and keeps paying for the decades after. Run the same arithmetic on any continuous line before accepting the cheaper element; on intermittent lines the arithmetic runs the other way, and honesty about load factor is the first selection skill.

The second selection driver is what the fluid carries. A venturi has no sharp edge to blunt and no cavity upstream of the bore where solids can pack, so raw river water, treated effluent, soft slurries within velocity limits, and scaling cooling water are all legitimate territory. The throat is the erosion point, and the specification answer is a replaceable or hardfaced throat insert plus a velocity ceiling agreed with the supplier. Where the fluid is conductive water without a custody implication, an electromagnetic flowmeter delivers similar maintenance peace at lower installed cost on moderate diameters, and honest selection practice prices both before committing.

The third driver is space. ISO 5167-4 asks for little: zero to 5 pipe diameters upstream depending on the disturbing fitting, and about one diameter downstream, against the 10 to 40 upstream diameters a sharp edged element can demand. That asymmetry makes the venturi the retrofit element inside buildings, on skids, and in existing galleries where pipe runs were never allocated for metering. For verification without cutting the line, a clamp-on ultrasonic flowmeter can be hired for an audit week and its readings folded into the venturi calibration record; several utilities keep exactly this pairing as standard practice on their large mains.

Honest selection also names the duties where other technologies win. Saturated and superheated steam above the pressure class of a lined venturi is classic flow nozzle and vortex flowmeter territory: the nozzle for high velocity and wet tolerance, the vortex flowmeter where the steam is clean and dry and turndown matters more than recovery. Clean hydrocarbon custody on smaller lines belongs to the turbine flowmeter with its long proving record. When the settlement is by mass rather than volume, a coriolis flowmeter on a small line or density compensated metering on a large one beats any differential element on uncertainty, because it removes the density estimate from the chain.

Utility gases round out the map. A thermal mass flowmeter reads gas flow in standard units directly, with no moving parts and no impulse lines, and it wins on plant air networks where running an impulse bundle back to the transmitter is the expensive part of a differential installation. Local visual duty on a utility takeoff still belongs to the variable area flowmeter, which needs no power and reads well enough for balancing. The discipline in every case is to let the duty choose: the venturi when energy and solids dominate, something else when they do not.

The RFQ block for a venturi tube should be written once and reused. A complete block states: quantity and line size with as measured bore and schedule; fluid, density basis, temperature range, and solids loading with particle size; flow range minimum to maximum with the required full scale differential; manufacturing class with C and its basis, standard or calibrated; tap type and orientation; end connections and the governing flange or butt weld standard; materials for body, throat, and any liner; straight run available upstream and downstream; the instrument package, meaning differential pressure transmitter ranges plus compensation inputs; and the documentation set, calculation sheet to ISO 5167-4, material certificates EN 10204 3.1, and a pressure test certificate. A quotation returned without the calculation sheet is not evaluable, only comparable on price, which is the wrong axis for a meter.

Installation, Standards, Failure Modes, and Maintenance

Installation quality sets the calibration basis. The metering spool is set with its bore concentric to the pipe within the displacement tolerance the standard allows, with gaskets that do not intrude into the bore; a gasket lip of a few millimeters at the throat end is a real coefficient error, not a cosmetic defect. Tap orientation follows the fluid: horizontal or just below the horizontal centerline for liquids so gas clears, top of pipe for gases so condensate drains back, never the bottom where solids collect. Impulse lines run with a continuous slope to the differential pressure transmitter, one to ten percent grade depending on fluid, so bubbles and condensate never trap in a low spot and read as flow.

Commissioning proves the loop rather than the element. Under first flow, the pressure transmitter used for density compensation is checked against the line gauge, a portable pressure sensor at the throat tap confirms that static pressure stays above the cavitation margin computed at maximum rate, and where an audit is required a hired ultrasonic flowmeter runs alongside for the first week so the baseline curve is corroborated independently. The first month of stable data is archived as the reference curve of differential against flow, because every later dispute about drift is settled by comparison against this baseline, not against the datasheet.

The standards frame is compact. ISO 5167-1 states the general principles, vocabulary, and the flow equation shared by every differential pressure element. ISO 5167-4 defines the classical venturi geometries, the coefficient values, their uncertainties, and the installation conditions cited throughout this page. ISO/TR 9464 collects the practical guidance for applying the ISO 5167 series to real fluids, including the density and compressibility treatment that gas and steam sizing depends on. Calibration claims trace to water laboratories operating proving rigs; a datasheet that cites ISO 5167 generically, without naming part 4, has not specified a venturi.

Failure modes in service follow a short list. Tap plugging leads: a single blocked throat tap reads low differential, and the loop responds by opening whatever it controls, so routine impulse flushing or chemical injection on scaling service is prescribed maintenance, not optional care. Erosion at the throat follows solids: it polishes and enlarges the bore, raises the effective beta, and reads flow low at an accelerating rate; the defense is a scheduled throat gauge check on slurry duty. Cavitation in liquid service appears when throat pressure falls to the liquid vapor pressure, audible as crackle at high rate and visible as pitting just downstream; the cure is back pressure, a valve or elevation held downstream, computed at maximum flow. Reynolds excursions below the class limit silently invalidate the published C, and a gasket intruding at a flange joint during a turnaround quietly repeats the original commissioning defect.

Maintenance is calibrated indifference: a well built venturi on clean water needs an annual zero check, an impulse line inspection, and nothing else for years. Slurry and scaling duties add throat gauging and flushing on a schedule the first two inspections set. Instruments carry their own calendar: transmitter recalibration, manifold valve exercise, and early replacement of impulse tubing on corrosion found. Where the station settles custody or fiscal totals, periodic proving against a master turbine flowmeter belongs in the contract rather than in goodwill. Plant air networks measured by a thermal mass flowmeter get their elements cleaned on the same outage calendar, and small branches still read by a variable area flowmeter have floats and scales checked with them, because a drifted branch meter corrupts the balance the main meter is calibrated against. Every one of these tasks costs a fraction of the energy the element saves, which is the closing argument of chapter 5: the venturi is bought once and defended with small money for decades.