Open Channel Flowmeter

What is an Open Channel Flowmeter

An open channel flowmeter is a flow-measurement system used wherever water moves under gravity in a conduit that is not pressurized and not flowing full: an irrigation lateral, a Venturi-shaped concrete channel, a sewer flowing at half depth, or the rectangular outfall from a treatment plant. The defining condition is a free water surface open to the atmosphere. Because the cross-sectional area of the water changes with depth and the conduit is not bounded on top, the inline principles that dominate pressurized pipes, electromagnetic, Coriolis, and turbine, cannot be applied. The open channel method substitutes a calibrated hydraulic structure for the missing pipe wall.

The system always has two parts. The primary device is a fabricated structure, a flume or a weir, that constricts the channel into a precisely defined shape. By the physics of critical flow, that shape forces a fixed, repeatable relationship between a single measured water depth, called the head, and the volumetric discharge. The secondary device is an instrument that measures only that head, typically a non-contact ultrasonic transmitter, a bubbler, or a submerged pressure transducer, and a transmitter that converts head to flow using the device's published equation or a lookup table. No velocity is sensed in this classic arrangement; flow is inferred entirely from depth and geometry.

A second, newer family dispenses with the primary device. Area-velocity meters place a submerged sensor in the channel that measures both the wetted level and the mean velocity, usually by acoustic Doppler, then multiplies area by velocity. These meters fit existing pipes and channels with no civil works and keep reading during surcharge or reverse flow, but they trade away the geometry-only calibration that makes flumes and weirs defensible for regulatory reporting. Most permit-grade installations still specify a primary device.

The engineering scale is broad. A small 1-inch Parshall flume meters from about 0.005 cubic feet per second (roughly 0.14 litres per second), while a 50-foot throat handles up to about 3,280 cubic feet per second (roughly 93 cubic metres per second). The same device family therefore covers a pilot bench rig and a river diversion, but no single size or geometry serves the whole range; selection is the act of matching channel hydraulics to one specific primary device and one head sensor.

The lineage is well documented. Sharp-crested weirs were standardized for hydraulic measurement in the 19th and early 20th centuries, and the V-notch weir remains the reference for clean low flows. The flume that bears Ralph Parshall's name was developed from 1915 onward, when Parshall of Colorado Agricultural College and the USDA modified the earlier subcritical Venturi flume by adding a throat contraction and a drop in the floor to force critical flow and shed sediment. That self-scouring throat is the reason the Parshall flume, and the U-shaped Palmer-Bowlus flume that followed for sewers, displaced the weir in dirty-water service.

Primary Device Families: Flumes and Weirs

The primary device is the heart of an open channel flowmeter, and the single most consequential selection decision. All primary devices fall into two families: weirs, which dam the channel and force water to spill over a calibrated crest, and flumes, which contract the channel to accelerate flow through a throat. Weirs are simpler and cheaper and resolve low flow well, but they pond sediment and demand significant head loss. Flumes are self-cleaning and economical on head, which is why they dominate wastewater. The table below compares the main types.

Rectangular thin-plate weir. A vertical plate with a sharp-edged rectangular notch, either suppressed (full channel width) or contracted (narrower than the channel). Water springs clear of the crest as a ventilated nappe, and discharge follows the head over the crest. ISO 1438 governs its geometry and the Kindsvater-Carter method that corrects for effective width and head. It is accurate and inexpensive but needs a free overfall and a stilling approach, so it suits clean, fairly steady channels.

90-degree V-notch weir. A triangular notch concentrates low flow into a deeper, more measurable nappe, giving the best low-end resolution of any primary device. The fully contracted 90-degree notch is a USBR and ISO 1438 standard. Because head rises steeply with flow, a small discharge still produces a readable head, which is why V-notch weirs are the reference for laboratory and clean-water rating. They are unsuitable for water that carries solids, since sediment ponds against the plate and shifts the approach velocity.

Cipolletti weir. A trapezoidal notch with 4:1 (vertical:horizontal) side slopes. The flaring sides are designed to offset end contraction so the discharge coefficient stays effectively constant with head, simplifying the rating to a near-rectangular form. It is common on irrigation canals where ease of field calculation matters more than ultimate accuracy.

Parshall flume. The workhorse of wastewater and irrigation. An hourglass plan of converging walls, a contracted throat, and a diverging outlet, with a deliberate drop in the floor through the throat. The contraction forces critical flow, and a single upstream head defines discharge. Twenty-two standard throat sizes span roughly 0.005 to 3,280 cubic feet per second. The throat scours grit, the head loss is modest, and the device tolerates partial submergence before correction is needed, all reasons it is the default for municipal effluent. ASTM D1941 is the test method.

Palmer-Bowlus flume. A U-shaped insert designed to drop into a circular sewer pipe, available for lines from about 4 to 48 inches. It needs very little head and installs in an existing manhole or channel, making it the standard for in-sewer flow surveys. Its rating is more sensitive to approach conditions than a Parshall, so a straight settled approach run is important.

Long-throated and Venturi flumes. ISO 4359 covers rectangular, trapezoidal, and U-shaped critical-depth (standing-wave) flumes. The long throat produces a stable hydraulic jump and a computable rating from geometry alone, so these can be designed for non-standard channels and tight head budgets without physical calibration.

Level Sensing and Area-Velocity Technologies

Once a primary device fixes the head-discharge relationship, the meter's accuracy is limited by how precisely the secondary instrument measures the head. Four secondary technologies dominate: non-contact ultrasonic, bubbler, submerged pressure transducer, and the structure-free area-velocity meter. Each has a distinct error budget and failure mode. The table below compares them on the parameters that matter at the procurement stage.

Non-contact ultrasonic. A transducer mounted above the flume or weir fires an acoustic pulse and times the echo from the water surface. It never touches the dirty stream, needs little maintenance, and runs on low power, which suits remote sites. Its limits are physical: a blanking distance (dead band) immediately below the sensor where echoes cannot be resolved, sensitivity to the speed of sound (which forces temperature compensation, since sound speed in air changes roughly 0.17 percent per degree Celsius), and false echoes from foam, heavy fog, steam, or steep sidewalls. Endress+Hauser Prosonic and Siemens Echomax sensors are representative.

Bubbler. A small compressor meters air through a dip tube submerged in a stilling well; the back-pressure needed to push bubbles out equals the hydrostatic head above the tube opening. Bubblers deliver roughly four times finer resolution than ultrasonic, about plus-or-minus 0.005 ft (1/16 inch) against plus-or-minus 0.02 ft (1/4 inch), and they ignore foam, steam, wind, floating debris, and air-water temperature differences entirely, since the sensing element is hydrostatic, not acoustic. The price is an air compressor, periodic tube purging, and a stilling well. They are the preferred choice for low flows where the head signal is small and for foul, foaming headworks atmospheres.

Submerged pressure transducer. A vented gauge sensor placed at the channel invert reads hydrostatic head directly. It is simple and immune to surface conditions, but the wetted sensor is exposed to fouling and abrasion, the vent tube can clog, and any change in liquid density biases the depth. It is common in temporary surveys and in level-only level transmitter applications.

Area-velocity meters. These abandon the primary device. A submerged sensor measures wetted depth (by pressure or upward-looking ultrasonic) and mean velocity (by acoustic Doppler shift off particles and bubbles), then multiplies area by velocity. Teledyne ISCO's 2150 and the Signature meter's submerged Doppler and LaserFlow options are widely used. They install in existing pipes and channels with no head loss, keep reading under surcharge and reverse flow, and suit retrofit and temporary monitoring. The trade-off is intrinsic accuracy: the velocity reading depends on profiling assumptions and is degraded by silt, fouling, and low particle content, and many discharge permits still require a geometry-calibrated flume or weir for the reported number.

Head-Discharge Equations and Standards

The open channel flowmeter is only as trustworthy as the head-discharge equation behind it, and those equations are published, standardized, and device-specific. Every primary device reduces to a single empirical relationship of the form Q equals C times H raised to the power n, where Q is discharge, H is the measured head, and C and n are constants fixed by the device geometry. Getting flow right therefore means installing the right device, measuring H at the correct gauge point, and applying the matching equation. Below are the standard forms.

90-degree V-notch weir. The Bureau of Reclamation gives, for a fully contracted 90-degree V-notch sharp-crested weir under free flow with 0.2 ft < H < 1.25 ft, the relation Q = 2.49 × H raised to 2.48, where Q is in cubic feet per second and H is the head over the notch vertex in feet. The exponent near 2.5 is why the V-notch resolves low flow so well: head changes sharply for a small change in discharge. ISO 1438, ASTM, and USBR all recommend the Kindsvater-Shen formulation for V-notch weirs of arbitrary angle.

Rectangular thin-plate weir. Discharge varies with the crest length L and the head as roughly H raised to 1.5, refined by the Kindsvater-Carter method of ISO 1438, which replaces L and H with an effective width and effective head plus a discharge coefficient that accounts for contraction. A Cipolletti (trapezoidal) weir is rated almost like a rectangular weir because its 4:1 sloping sides are designed to cancel the end-contraction correction.

Parshall flume. The free-flow rating is Q = C × Ha raised to n, where Ha is the upstream head and both C and n change with throat width. The table below lists published coefficients (USBR / Wikipedia values, US customary units, Q in cubic feet per second and Ha in feet) and shows how the constants migrate across sizes.

Submergence correction. Each equation above is valid only under free flow, where a hydraulic jump downstream isolates the throat from the tailwater. When the ratio of downstream to upstream head exceeds the device limit (50 to 80 percent for a Parshall flume depending on size), the flow is submerged, a second downstream head must be measured, and a correction is applied. USBR is explicit that this degrades accuracy: submerged-flow head imprecision can add 4 to 20 percent error on top of the 3 to 5 percent free-flow uncertainty, and designing for submerged operation is no longer considered good practice. The engineering rule is to size and set the device so it always runs free.

Standards map. ISO 1438 covers thin-plate weirs (rectangular and V-notch), ISO 4359 covers rectangular, trapezoidal, and U-shaped flumes, ISO 4360 covers triangular-profile (Crump) weirs, and ASTM D1941 is the Parshall flume test method. The USBR Water Measurement Manual is the field handbook that ties these together with installation, gauge-location, and approach-condition guidance.

Key Specification Parameters

Comparing open channel flowmeter quotes means reading two spec sheets at once: the hydraulic spec of the primary device and the instrument spec of the secondary sensor. Seven parameters drive the selection decision: flow range, system accuracy, head measurement resolution, head loss, free-flow submergence limit, approach and installation requirements, and output and logging. Each is explained below.

Flow range. The primary device sets the measurable span. A given Parshall throat has a published minimum and maximum discharge, and the family as a whole covers roughly 0.005 to 3,280 cubic feet per second across 22 sizes. The operating point should sit well inside the device's rated band; too small a flow produces a head below the minimum measurable head and the reading collapses into sensor noise.

System accuracy. The total uncertainty is the combination of the primary device rating and the secondary sensor error, not either alone. A properly installed Parshall flume in free flow is typically quoted at 3 to 5 percent of discharge; a clean V-notch weir can do better at low flow. That figure assumes free flow, a correct gauge location, and a settled approach. Submerged operation, a poorly located gauge, or surface-wave error at the gauge point can each add several percent.

Head measurement resolution. Because discharge scales as H raised to 1.5 to 2.5, a small head error amplifies into a larger flow error, so resolution is decisive at low flow. Bubblers resolve about plus-or-minus 0.005 ft (1/16 inch), ultrasonic about plus-or-minus 0.02 ft (1/4 inch). On a V-notch weir where flow goes as H raised to 2.48, a 1 percent head error becomes roughly a 2.5 percent flow error, which is why low-flow metering favours the finer bubbler.

Head loss. The energy the primary device consumes. Weirs need a substantial free overfall and a 2x to 5x head allowance; flumes lose far less, and Palmer-Bowlus and long-throated flumes least of all. Where the channel has little available fall, this parameter alone can rule out a weir.

Free-flow submergence limit. The downstream-to-upstream head ratio above which a single upstream head no longer defines discharge: roughly 50 to 80 percent for Parshall flumes by size. The device must be set so the tailwater never crosses this limit, or a downstream sensor and correction become mandatory and accuracy falls.

Approach and installation requirements. Every primary device assumes a straight, settled, subcritical approach of several channel widths, a level crest or floor, and a gauge taken at the prescribed point (for a weir, the head is read upstream at a distance of three to four times the maximum head so surface drawdown does not corrupt the reading). Poor approach hydraulics are the most common field cause of error.

Output and logging. The transmitter interface to the control system or data network. Common options are 4-20 mA, relay or pulse totalizer outputs, Modbus RTU, and on battery-powered survey loggers (such as the ISCO Signature and Hach FL900) internal datalogging with cellular or radio telemetry. For regulatory effluent monitoring, totalized volume, time-stamped logging, and MCERTS or equivalent approval may be required.

• 4-20 mA: Two-wire analog flow signal to a SCADA or PLC analog input. The most common output for fixed plant installations.
• Pulse / relay totalizer: Emits a contact per unit volume for mechanical counters and sampler pacing.
• Modbus RTU / digital bus: Carries flow, level, velocity, and diagnostics over a single serial pair to a data concentrator.
• Internal datalogger + telemetry: Battery survey meters store level, velocity, and flow and report by cellular or radio link for temporary or remote sites.

Selection Decision Factors

To turn the preceding chapters into a specific purchase, follow the decision sequence below. Most open channel metering failures trace not to a faulty instrument but to a primary device chosen or installed for the wrong hydraulics, so the order of these steps matters. The list doubles as an RFQ template.

1. Characterize the stream first: clean or solids-laden, expected minimum and maximum flow, and whether grit or sewage is present. Solids push you toward a flume; clean low flow toward a V-notch weir. This single judgement eliminates half the device families before any spec is read.
2. Measure the available head loss: survey the fall available between the upstream channel and the downstream water level. A weir needs a large free overfall; if the channel is flat, only a flume (Parshall, Palmer-Bowlus, or long-throated) will fit.
3. Size the primary device to the flow range: pick the throat or notch so the normal operating point sits inside the rated band and the minimum flow still produces a head above the sensor's minimum measurable head. Verify the maximum flow stays in free flow against the device submergence limit.
4. Check tailwater and submergence: estimate the downstream depth at peak flow and confirm the downstream-to-upstream head ratio stays under the device limit (50 to 80 percent for Parshall flumes). If it cannot, either drop the device elevation or accept a two-sensor submerged correction.
5. Choose the secondary sensor: bubbler for low flows and foul, foaming atmospheres where resolution is critical; non-contact ultrasonic for clean spans, remote low-power sites, and minimal maintenance; submerged transducer or area-velocity meter where no civil works are possible or surcharge must be captured.
6. Confirm approach and gauge location: provide the straight settled approach run the standard requires and place the level sensor at the prescribed gauge point. Budget for a stilling well if the surface is turbulent. Poor approach is the leading field error.
7. Match standards and certification: specify the governing standard for the device (ISO 1438, ISO 4359, ISO 4360, or ASTM D1941) and, for regulatory effluent reporting, any required environmental approval such as MCERTS, plus totalizing and time-stamped logging.
8. Total cost of ownership (TCO): primary-device civil works + sensor + transmitter + power (a bubbler needs a compressor) + calibration verification + maintenance (tube cleaning, sensor de-fouling). A structure-free area-velocity meter saves installation cost but may not satisfy a permit that demands a calibrated primary device, in which case the cheaper option cannot be used at all.

One last commonly overlooked dimension is manufacturer serviceability and supply integration. A primary device and its level transmitter must share a verified head-discharge table, and a mismatch silently biases every reading. Teledyne ISCO (Signature, 2150, LaserFlow), Endress+Hauser (Prosonic FMU90 and FDU92), Siemens (SITRANS LUT440 OCM with Echomax sensors), Hach (FL900 series), Nivus, and KROHNE supply matched transmitter-and-sensor systems, while flume and weir fabricators such as Openchannelflow, Tracom, and Plasti-Fab supply pre-calibrated fibreglass devices. Buying the structure and the instrument as a verified pair, with the certified rating table in hand, removes the most common source of long-term measurement drift.