Flow Measurement in Partially Filled Pipes: Doppler, Area-Velocity and Level Methods

Measuring flow in a full pressurised pipe is a solved problem. Measuring flow in a partially filled pipe — a sewer, a storm drain, a discharge channel — is not, because there are two unknowns instead of one: the velocity of the water and the cross-sectional area it occupies. This guide explains why that matters, how the main methods work, and how to choose.

Why Partially Filled Flow Is Hard

Discharge is velocity multiplied by cross-sectional area. In a full pipe the area is fixed, so measuring velocity is enough. In a partially filled pipe both terms vary:

• Area varies with depth — and in a circular pipe the relationship is not linear.
• Velocity varies with depth — the same channel flows faster when it is deeper.
• Velocity varies across the section — fastest near the centre and surface, slowest at the walls and bed.
• Sediment changes the geometry — a deposited bed reduces area and alters the roughness.
• Flow can be unsteady — surcharge, backwater effects and diurnal patterns all distort the profile.

Any method must address both terms, either by measuring them or by inferring one from the other.

The Main Methods

Area-Velocity: The Direct Approach

Measure velocity and level independently, then calculate area from level using the known channel geometry, and discharge from area and velocity.

• Velocity measurement: Doppler ultrasound, or electromagnetic point velocity.
• Level measurement: Ultrasonic, radar or pressure.
• Strengths: Works across a wide range of conditions; handles surcharge and reverse flow; no need to modify the channel.
• Weaknesses: Requires knowledge of the cross-section; sediment affects accuracy; velocity is measured at a point and must represent the mean.

This is the dominant method for sewer and manhole monitoring, and it is why compact Doppler velocity instruments such as the MT-D800B, with low power draw and RS485 Modbus RTU output, are widely deployed in inspection chambers and discharge outlets.

Doppler Velocity: How It Works

A Doppler flow meter transmits an ultrasonic signal into the water and measures the frequency shift of the signal reflected by suspended particles or bubbles moving with the flow. The shift is proportional to velocity.

• Requires scatterers — particles or bubbles must be present to reflect the signal. Very clean water can be difficult.
• Measures velocity, not level — level must come from a separate sensor or be entered as a fixed value.
• Compact and low power — well suited to battery and solar installations.
• Sensitive to mounting position — the sensor must be in the flowing stream, not in a stagnant zone or buried in sediment.

Level-to-Flow with a Primary Device

Install a weir or flume with a known, stable relationship between upstream level and discharge, then measure level only.

Primary device	Best for	Notes
Triangular (V-notch) weir	Low flows	Good accuracy at small discharges
Rectangular weir	Moderate flows	Simple construction
Parshall flume	Wide range, dirty water	Self-cleaning, low head loss, tolerant of solids

• Strengths: Once installed and verified, only a level sensor is needed; accuracy can be excellent.
• Weaknesses: Requires civil works; the structure causes head loss and can trap sediment upstream; accuracy degrades if the device is damaged, silted or submerged (drowned flow).

Electromagnetic Flow Meters in Full Pipes

For genuinely full pressurised pipes, electromagnetic meters are accurate and reliable. In partially filled pipes they require the pipe to remain full, which is why some installations use a flume or an inserted reduced section to maintain full flow at the measurement point.

Method Comparison

Method	Measures	Civil works	Handles surcharge	Sediment tolerance	Typical use
Area-velocity (Doppler + level)	Velocity and level	None	Yes	Moderate	Sewers, manholes, outfalls
Doppler velocity only	Velocity	None	Yes	Moderate	Fixed-geometry channels
Weir / flume + level	Level, converted to flow	Yes	No — drowned flow invalid	Low to moderate	Open channels, treatment works
Full-bore electromagnetic	Velocity in full pipe	Yes	Yes	Good	Pressurised mains, dosing lines

Choosing a Method

Situation	Recommended approach	Reason
Existing sewer, no civil works possible	Area-velocity	No modification needed, handles variable depth
Industrial discharge outfall with a flume	Level with flume curve	Accurate and verifiable once installed
Battery-powered remote site	Low-power Doppler	Minimal power demand
Clean water with few particles	Consider electromagnetic or level-to-flow	Doppler needs scatterers
Channel with heavy sediment	Flume with level, or area-velocity with sediment compensation	Sediment distorts area and velocity

Installation Factors That Decide Accuracy

• Representative location: Straight reach, steady flow, away from bends, junctions, drops and pumps.
• Sensor mounting: Securely fixed in the flow path, clear of the bed, oriented as specified by the manufacturer.
• Cross-section survey: For area-velocity, the geometry used in the calculation must match reality — measure it, do not assume the design drawing.
• Sediment monitoring: Track the bed level; a rising bed silently corrupts the area calculation.
• Verification: Compare against an independent method — dilution gauging, tracer, or a temporary reference instrument.
• Cable and power: In manholes, plan for ingress protection and safe access for maintenance.

Common Sources of Error

• Assuming a design cross-section that does not match the built one.
• Sediment accumulation silently reducing area.
• Sensor mounted in a stagnant zone or buried in silt.
• Drowned flow at a weir or flume, where the downstream level submerges the device and invalidates the curve.
• Very clean water giving weak Doppler return.
• Turbulence and air entrainment near drops and pumps.

Conclusion

Partially filled flow measurement comes down to handling two unknowns properly. Area-velocity instruments measure both and suit existing sewers and outfalls with no civil works; primary devices with level measurement give excellent accuracy where construction is possible and flow stays within the device’s range. Whichever you choose, the accuracy of the result depends less on the instrument than on siting, a real cross-section survey, and periodic independent verification.

Frequently Asked Questions

Can a Doppler meter work in clean water?

It needs something in the water to reflect the ultrasonic signal — particles or bubbles. In very clean water the return signal may be too weak for reliable measurement, and an electromagnetic or level-to-flow method is usually a better choice.

Do I need a separate level sensor?

For true area-velocity measurement, yes — Doppler gives velocity, and level must come from another sensor. In a channel of fixed, known geometry with stable depth, some installations enter level as a constant, but this is only valid if depth genuinely does not change.

Why does my flow reading drift over months?

Sediment is the most common cause: a rising bed reduces the cross-sectional area while the instrument continues to use the original geometry. Regular bed-level checks and updating the geometry in the instrument resolve it.

What is drowned flow?

When the downstream water level rises enough to submerge a weir or flume, the level-to-flow relationship changes and the standard curve no longer applies. Instruments should detect and flag this condition rather than reporting a value.

How often should flow meters be verified?

At commissioning, then periodically — annually at minimum for most applications, and after any event likely to have changed the channel, such as a storm, sediment clean-out or structural work.