Top Water Level Sensors for Accurate Measurement

Introduction

Accuracy in water level measurement is not optional – it is the primary reason a sensor exists. An inaccurate reading in a municipal reservoir causes incorrect inventory data. An inaccurate reading in a process vessel causes overfilling, under-filling, or failed batch records. An inaccurate reading at a flood warning gauging station puts downstream communities at risk. Accuracy determines whether a sensor earns its place in the system, or whether it generates expensive problems instead of useful data.

This article reviews the water level sensors most widely specified for accurate measurement in 2026, explains which sensor technologies deliver the best accuracy, what accuracy specifications to look for in a datasheet, and how to match sensor accuracy to application requirements. The focus is on continuous level measurement – where accuracy is the primary performance criterion – with notes on point-level sensors where they are relevant.

Understanding Sensor Accuracy Specifications

Accuracy, Repeatability, and Hysteresis

Accuracy (measurement uncertainty) is the maximum deviation of the sensor reading from the true value, expressed as a percentage of full scale (% FS) or in millimetres. A sensor with accuracy of +/- 0.5% FS on a 10-metre range has a maximum error of +/- 50 mm at any point in the range. Accuracy is the primary specification for applications where the absolute level value matters – inventory accounting, flow calculation, or process control.

Repeatability is the ability of the sensor to return the same reading when measuring the same quantity under the same conditions. A sensor can have excellent repeatability but poor accuracy – it reads consistently, but that consistent value is wrong. Repeatability matters most for alarm and control applications where the sensor triggers an action at a setpoint.

Hysteresis is the difference in sensor output when the same level is approached from above versus from below. It is particularly relevant for point-level switches where the switch point differs depending on whether the level is rising or falling.

When reading a datasheet, check whether accuracy is expressed as % FS, % MV (measured value), or millimetres. A % FS specification is more conservative and easier to verify. Ask for the linearity specification: a sensor with good accuracy but poor linearity accumulates error across the range.

Technologies and Their Accuracy Performance

Radar Level Sensors: The Accuracy Benchmark

Radar level sensors are the benchmark for accuracy in water level measurement. The best FMCW (frequency-modulated continuous wave) radar sensors achieve accuracy of +/- 0.5 mm to +/- 2 mm, or +/- 0.05% to +/- 0.1% of full scale, whichever is greater. FMCW radar measures the frequency difference between a transmitted and received frequency ramp, a technique that is inherently more accurate than simple time-of-flight pulse measurement.

For water level applications, the two most relevant radar frequencies are 26 GHz (K-band) and 80 GHz (W-band). 80 GHz FMCW radar offers the highest accuracy and narrowest beam angle, preferred for tanks with restricted openings and applications where wall reflections must be minimised. 26 GHz FMCW radar offers a wider beam and longer range, well suited to large open-channel applications and wide tanks.

Key accuracy specifications to check on a radar datasheet:

Accuracy: +/- mm or +/- % FS

Repeatability: typically 0.1-1 mm

Resolution: smallest detectable change in level, typically 0.1-1 mm

Near-range dead band: the minimum distance from the sensor at which measurement is possible

Temperature stability: accuracy drift with temperature, typically % FS per degree Celsius

Ultrasonic Sensors: Good Accuracy at Lower Cost

Ultrasonic sensors are a lower-cost alternative to radar, with typical accuracy of +/- 2 mm to +/- 5 mm, or +/- 0.25% to +/- 0.5% of full scale – approximately two to five times less accurate than FMCW radar, but adequate for many water storage and monitoring applications.

The primary accuracy limitation of ultrasonic sensors is their dependence on the speed of sound in air, which varies with temperature. An uncompensated ultrasonic sensor in a tank where air temperature varies by 20 degrees Celsius will have an accuracy error of approximately 1.2% – a 12 mm error on a 1-metre range – solely from the speed of sound effect. Modern ultrasonic sensors address this with built-in temperature compensation.

With active temperature compensation, ultrasonic sensors achieve rated accuracy in environments where temperature variation is moderate – indoor tanks, covered sumps. In outdoor environments with large diurnal temperature swings, accuracy can degrade significantly. Wind and surface turbulence also affect ultrasonic accuracy; in open-channel outdoor applications, radar is generally preferred.

Key specifications to check:

Accuracy with temperature compensation: +/- mm or +/- % FS

Temperature compensation range: the temperature range over which the stated accuracy is maintained

Beam angle: narrower beam reduces wall reflection errors but requires more precise mounting

Dead zone (near range)

Hydrostatic Pressure Sensors: Accuracy in Immersion Applications

Submersible hydrostatic pressure sensors achieve excellent accuracy in immersion applications. The best instrumentation-grade hydrostatic level sensors achieve accuracy of +/- 0.1 mm to +/- 0.5 mm, or +/- 0.05% to +/- 0.1% of full scale – comparable to FMCW radar. Industrial hydrostatic sensors typically specify +/- 0.25% to +/- 0.5% of full scale.

The accuracy of pressure-based level measurement depends on three factors: the accuracy of the pressure sensor element, the accuracy of temperature compensation, and the accuracy of the water density assumption used in converting pressure to level. The P = rho g h relationship requires a known water density value. If actual water density differs from the assumed value, the level reading will be systematically in error. Advanced hydrostatic level sensors address density variation with dynamic density compensation: a second temperature sensor and a density correction algorithm applied to the pressure measurement.

For well, borehole, and tank applications where immersion is acceptable, hydrostatic pressure sensors offer the highest accuracy per unit cost.

Key specifications to check:

Accuracy: +/- % FS, with full scale clearly defined

Total error band (TEB): a combined accuracy and temperature error specification that is more representative of real-world performance than basic accuracy alone

Long-term stability: drift in calibration over time, typically % FS per year

Temperature compensation range

Float Switches: Accuracy for Point-Level Applications

Float switches are point-level sensors, not continuous level sensors. A well-designed float switch has a switching level repeatability of +/- 3 mm to +/- 10 mm in stable conditions – adequate for pump control and alarm applications but not suitable for continuous level monitoring or flow calculation.

The key accuracy-related specification for float switches is hysteresis: the difference in water level between the rise and fall switching points. A float switch with 20 mm hysteresis will energise the pump at 200 mm and de-energise at 180 mm – a dead band that must be accounted for in pump control design.

Comparing Sensor Technologies for Accuracy

Table: Accuracy Comparison of Water Level Sensor Technologies

FMCW radar at 80 GHz achieves the best continuous-level accuracy across the widest range of conditions. Hydrostatic pressure sensors achieve comparable or better accuracy in immersion applications, particularly at instrumentation grade. Ultrasonic sensors offer good accuracy at lower cost in controlled environments. Float switches serve point-level applications adequately.

Top Applications Where Accuracy Matters Most

Open-Channel Flow Measurement

Flow rate through a weir or flume is calculated from measured water level using the rated hydraulic equation for the structure. A 1 mm error in level measurement at a Parshall flume can translate to a 2-5% error in calculated flow rate, depending on flow depth and flume geometry.

For open-channel flow measurement, radar sensors are preferred because of their immunity to wind, temperature variation, and surface conditions. FMCW radar at 80 GHz offers the best accuracy and narrowest beam for narrow channels and small flumes. 26 GHz FMCW radar is well suited to wider channels and rivers where the wider beam provides reliable surface illumination.

Inventory Management in Water Storage Tanks

Municipal and industrial water storage tanks require accurate level data for inventory management, delivery scheduling, and loss detection. An accuracy of +/- 10 mm on a 10-metre tank represents an inventory uncertainty of approximately 0.1% of full scale – acceptable for most inventory management purposes.

FMCW radar sensors at 26 GHz or 80 GHz achieve this accuracy reliably in open or vented tanks. Ultrasonic sensors can achieve comparable accuracy in climate-controlled indoor tanks.

Industrial Process Control

Water level control in industrial processes – boiler feedwater, cooling tower basins, CIP (clean-in-place) rinse tank monitoring – requires accuracy sufficient to maintain level within the required operating band. A process that requires level to be maintained within +/- 20 mm of the setpoint demands a sensor with repeatability of approximately +/- 2 mm and accuracy of +/- 5 mm or better.

FMCW radar is preferred for demanding process control applications. Hydrostatic pressure sensors are widely used in process applications where immersion is acceptable, offering excellent accuracy at lower cost per sensor than radar.

Flood Warning and Environmental Monitoring

River gauging stations that feed flood warning systems require level accuracy sufficient to detect significant changes in river stage – typically 10 mm or better – with high reliability and minimal maintenance. Solar-powered, cellular-connected FMCW radar level sensors have made continuous, accurate, remote river monitoring commercially practical.

For environmental research applications – groundwater monitoring wells, lake level monitoring, tidal measurement – instrumentation-grade hydrostatic pressure sensors achieve the highest accuracy. These sensors are calibrated against national measurement standards and provide data quality suitable for hydrological research and compliance reporting.

How to Specify Accuracy for a Water Level Application

Specifying accuracy correctly is one of the most consequential decisions in sensor selection. Over-specifying accuracy wastes money. Under-specifying accuracy creates data quality problems that are expensive to fix after installation.

Begin with a clear definition of the required accuracy, expressed as a maximum permissible error in millimetres or as a percentage of full scale, and the conditions under which this accuracy must be maintained – temperature range, measurement range, liquid properties.

For open-channel flow measurement, derive required level accuracy from required flow accuracy using the hydraulic equation for the measuring structure. For inventory management, express required accuracy as a maximum permissible inventory uncertainty and convert to required level accuracy using tank geometry. For process control, determine the required operating band width and add a safety margin for the sensor accuracy and repeatability.

Always verify the accuracy specification in the context of the full measurement range. A sensor specified as +/- 2 mm accuracy on a 10-metre range has a maximum error of 0.02% of full scale – excellent. The same sensor on a 100-metre range has a maximum error of 0.002% of full scale – still excellent – but the absolute error is 10 times larger. Understand how accuracy degrades at the low end of the measurement range, where many applications spend most of their time.

Request a total error band (TEB) specification from the sensor manufacturer. This combines accuracy, temperature effect, long-term stability, and linearity into a single specification that represents worst-case error under all specified operating conditions. TEB is more honest than basic accuracy and makes it easier to compare sensors from different manufacturers on an equal basis.

Installation Practices That Protect Accuracy

A sensor specified to +/- 1 mm accuracy will not deliver +/- 1 mm accuracy if installed incorrectly.

Non-contact sensors (radar and ultrasonic): Mounting the sensor off-centre in a tank produces a false echo from the tank wall that the signal processor may track instead of the water surface. Mounting the ultrasonic sensor in direct sunlight causes local heating of the transducer and degrades temperature compensation accuracy. Verify that the beam angle does not intersect tank walls, ladders, or other internals within the measurement range.

Pressure sensors: Installing the pressure sensor above the lowest point of the tank introduces an offset that must be entered during calibration. Cable deformation and thermal expansion of the cable can introduce additional errors. Use a properly rated submersible cable with a sealed connector, and install the sensor at the lowest practical point.

All sensors: Verify calibration after installation. A two-point calibration check – comparing the sensor reading at empty and at a known reference level – confirms the installation is correct and calibration is intact. Recalibrate or adjust the installation if verification shows errors exceeding the application requirement.

Conclusion

Accuracy is the fundamental measure of value in a water level sensor. FMCW radar level sensors – particularly 80 GHz instruments – deliver the highest accuracy and most consistent performance across the widest range of conditions. Hydrostatic pressure sensors offer comparable or superior accuracy in immersion applications at lower cost. Ultrasonic sensors achieve good accuracy in controlled indoor environments at the lowest cost.

For every application, derive the required accuracy from the application need – flow calculation, inventory management, process control, or flood warning – and select a sensor that exceeds that requirement with margin. Examine accuracy specifications in the context of the full measurement range, temperature range, and liquid properties of the application, and verify by requesting a total error band specification or a post-installation calibration check.

The sensor that delivers the highest accuracy in the datasheet is not always the best choice. The best choice is the sensor that delivers the required accuracy in the specific application conditions, with the lowest total cost of ownership over the expected lifetime of the installation.