An ultrasonic level meter is a non-contact, top-mounted instrument that measures the level of a liquid or bulk solid by timing a sound pulse as it travels to the surface and back. Because nothing in the tank touches the medium, it is a low-maintenance workhorse in water and wastewater treatment, chemical day tanks, and bulk silos. The same device is often called an ultrasonic level transmitter when it carries signal-conditioning electronics and a standardized 4-20 mA HART output.
This guide separates the physics that make ultrasonic reliable in open, benign atmospheres from the conditions that quietly defeat it: vapor, foam, vacuum, and steep temperature gradients. It decodes the spec-sheet parameters that decide a project (blind zone, beam angle, range, accuracy, and process limits) and ends with a practical selection sequence.
This guide is written for industrial purchasing engineers and design engineers. It covers 6 chapters, from the time-of-flight principle and transducer types through blind zone, beam angle, and spec decoding to the selection decision, with 7 selection FAQs and manufacturer comparisons. Performance terminology references the public IEC 61298 and IEC 60770 instrument-evaluation standards, with IEC 60529 ingress protection, IEC 60079 for hazardous areas, and ISO 1438 and ISO 4359 for open channel flow.
Chapter 1 / 06
What is an Ultrasonic Level Meter
An ultrasonic level meter is a continuous, non-contact level instrument mounted at the top of a tank, sump, channel, or silo. A piezoelectric transducer emits a short burst of ultrasound, the pulse travels through the gas above the medium, reflects off the surface, and returns to the same transducer. The instrument measures the round-trip time of flight and computes the distance to the surface from the known speed of sound. Subtracting that distance from a configured empty-vessel reference yields level, and a linearization table converts level to volume or to open channel flow.
The category sits within the broader family of continuous level technologies that also includes non-contact radar, guided wave radar, hydrostatic pressure, capacitance, and magnetostrictive measurement. Ultrasonic occupies a clear niche: it is non-contact like radar but far cheaper, it needs no mechanical float or wetted probe, and it is unaffected by the dielectric constant or conductivity of the medium that constrains capacitance and some radar applications. Its defining weakness is equally clear: it depends on a clean, predictable gas path, because the carrier is sound rather than an electromagnetic wave.
Structurally a modern ultrasonic level meter has three parts. First, the transducer, a piezoceramic disc behind a chemically resistant face of ETFE, PVDF, or filled polypropylene that both transmits the pulse and receives the echo, with an embedded temperature sensor for sound-velocity correction. Second, the echo-processing electronics, a microcontroller running digital signal processing that separates the true surface echo from false reflections off internal structures using stored echo profiles and rule-based filtering. Third, the output stage, which conditions the result into a 4-20 mA loop with HART, and increasingly a digital bus such as PROFIBUS PA, Modbus RTU, or IO-Link.
The technology matured in the 1970s and 1980s as low-cost microcontrollers made real-time echo processing affordable, and it spread quickly through municipal water and wastewater because it tolerates dirty, coating, or aggressive liquids that would foul a contacting probe. Today every major process instrument maker offers an ultrasonic line: Siemens SITRANS Probe LU240, Endress+Hauser Prosonic, VEGA VEGASON, Emerson Rosemount 3107 and 3108, and KOBOLD NUS-4 are representative. The intelligence that distinguishes a good instrument from a cheap one now lives almost entirely in the echo-discrimination firmware, not the transducer itself.
Four engineering realities govern whether ultrasonic is the right tool for a given vessel: the gas path must be reasonably free of vapor and dust, the surface must be reasonably flat and non-foaming, the process must stay within modest temperature and pressure limits, and the required range must comfortably exceed the blind zone. Where all four hold, ultrasonic delivers accurate level at a fraction of the cost of radar. Where any one fails, the answer is usually a different technology rather than a better ultrasonic meter.
Chapter 2 / 06
Configurations and Variants
Ultrasonic level products differ less by sensing principle, which is uniform, than by packaging, range class, and application focus. Understanding the configuration families prevents the common mistake of buying a compact short-range probe for a tall silo, or a high-power solids transmitter for a small water cistern. The table below maps the main variants to their typical envelopes.
Variant
Typical Liquid Range
Typical Blind Zone
Best Fit
Compact integral probe
0.25 to 5 m
0.25 m
Small tanks, sumps, day tanks
Mid-range two-wire
0.3 to 8 m
0.3 to 0.5 m
Process tanks, wet wells
Long-range remote sensor
up to 15 to 25 m
0.4 to 0.6 m
Tall silos, large reservoirs
Open channel flow set
0.3 to 8 m head
0.3 m
Weirs and flumes
By electronics location. An integral instrument combines transducer and electronics in one head, the simplest and lowest-cost option, suited to accessible tanks at moderate temperature. A remote, or separate, configuration places a rugged transducer (an FDU sensor in Endress+Hauser terms) at the vessel and a wall-mounted transmitter or controller, such as the SITRANS LUT400, several meters away. Remote mounting protects the electronics from heat, vibration, and washdown, and lets one controller serve multiple sensors and run pump or alarm relays.
By output and wiring. Two-wire loop-powered devices draw their power from the same 4-20 mA pair that carries the signal, minimizing cabling and easing intrinsic-safety design; they are the default for tank level. Four-wire devices have separate power and signal, allowing higher transmit power for long ranges and dusty solids, plus on-board relays for pump control. Bus-native versions speak PROFIBUS PA or Modbus RTU for digital integration, and compact OEM probes increasingly offer IO-Link.
By medium and application. Liquid-optimized transducers assume a flat, reflective surface and reach the longest ranges. Solids-optimized transducers use higher transmit power and aggressive echo filtering to cope with the weak, scattered echoes from powders and granules, where the angle of repose, dust, and a sloping surface all reduce returned energy; rated solids ranges are typically far shorter than liquid ranges on the same hardware. Open channel flow sets pair a standard transducer with weir or flume linearization and totalization firmware, which is covered in Chapter 4.
One configuration choice deserves early attention: frequency class. Lower-frequency transducers, near 18 to 40 kHz, carry energy farther and tolerate dust but have a wider beam and a larger blind zone. Higher-frequency transducers, toward 70 to 200 kHz, give a narrow beam and a small blind zone for short, cluttered vessels but attenuate quickly and cannot reach long range. Most product lines publish several frequency or range variants for exactly this reason, and selecting the wrong one is a frequent cause of unstable readings.
Chapter 3 / 06
The Time-of-Flight Principle
The physics of ultrasonic level reduces to one equation: distance equals the speed of sound multiplied by the round-trip time, divided by two. The transducer fires a pulse, a timer starts, the echo returns, the timer stops, and the firmware halves the elapsed time because the pulse covers the gap twice. Everything that makes ultrasonic accurate or inaccurate flows from how well the instrument knows the speed of sound and how cleanly it can identify the true echo. The table below summarizes the parameters that the principle exposes to the engineer.
Parameter
Typical Value
Engineering Consequence
Speed of sound in air, 0 C
331.5 m/s
Reference for time-to-distance conversion
Temperature coefficient
approx 0.6 m/s per K
About 0.17% error per degree if uncompensated
Transducer frequency
18 to 200 kHz
Higher frequency: smaller beam, shorter range
Beam angle (minus 3 dB)
5 to 12 deg
Wider beam clips obstructions, false echoes
Blind zone
0.1 to 0.6 m
Ring-down region, no measurement possible
Speed of sound and temperature. The speed of sound in dry air is about 331.5 m/s at 0 degrees Celsius and rises roughly 0.6 m/s for each kelvin, reaching about 343 m/s at 20 degrees Celsius. That sensitivity, near 0.17 percent per degree, means an uncompensated 20 degree swing would shift a reading by more than 3 percent of the measured distance. Every credible instrument therefore embeds a temperature sensor in the transducer and continuously recalculates sound velocity. The compensation is exact only when the gas column is at a uniform temperature; a vertical gradient (warm headspace over cold liquid, common in outdoor sun-exposed tanks) bends the effective velocity and leaves a residual error that single-point compensation cannot remove.
Blind zone and ring-down. After firing, the piezoceramic continues to vibrate for a brief interval. Until that ring-down decays, the receiver cannot tell its own residual motion from a returning echo, so any surface within the blind zone (also called the dead band or blocking distance) is invisible. Blind zone scales with transmit power and frequency, from about 0.1 m on short high-frequency probes to 0.6 m on long-range units. The mounting rule follows directly: position the transducer so the highest expected level always remains below the blind zone, with margin, and confirm the lowest level stays inside the rated range.
Beam angle and false echoes. The emitted energy forms a cone whose half-power width, the minus 3 dB beam angle, is commonly 5 to 12 degrees; high-frequency probes such as the KOBOLD NUS-4 narrow this to 5 to 7 degrees. The cone widens with distance, so at long range it can intersect tank walls, ladders, agitator shafts, weld beads, or an inlet stream, each of which throws a false echo. The instrument footprint diameter is approximately twice the range times the tangent of half the beam angle; keep that circle clear of obstructions, mount the transducer perpendicular to a flat surface, and use the device's echo-mapping function to teach it which reflections to ignore.
Echo processing intelligence. In a real vessel the transducer hears many echoes: the true surface, multiple bounces, structural reflections, and noise. Modern firmware (Siemens calls its version Process Intelligence) stores an echo profile, applies time-varying gain to compensate for attenuation with distance, and uses rule-based discrimination to lock onto the genuine level echo even when it is weaker than a fixed false reflection. The quality of this algorithm, more than the raw transducer, determines whether an instrument holds a stable reading through agitation, filling turbulence, and changing surface conditions.
Chapter 4 / 06
Media, Process Limits, and Standards
Because the carrier is sound traveling through gas, ultrasonic compatibility is governed less by chemical corrosion than by the physics of the gas path. The transducer face still has to resist the medium and its vapors, but the decisive questions are whether the pulse can reach the surface, return cleanly, and travel through a predictable atmosphere. The table below maps common conditions to suitability.
Condition
Ultrasonic Suitability
Preferred Alternative
Clean liquid, open tank
Excellent
None needed
Bulk solids, dusty
Fair, use high power
Non-contact radar
Heavy vapor or steam
Poor
Radar or guided wave radar
Foam on surface
Poor
Guided wave radar
Vacuum or near-vacuum
Not possible
Radar or pressure
Hot, above approx 90 C
Limited
Radar or pressure
Pressurized, above approx 3 bar
Limited
Radar or pressure
Process temperature and pressure. The polymer transducer face and the need for a predictable gas column set practical limits. Many ultrasonic level meters cap process temperature around 70 to 90 degrees Celsius and process pressure near 3 bar (about 43 psig), well below what radar tolerates. High temperature also widens the gas temperature gradient that degrades sound-velocity compensation, and pressure changes the gas density and therefore the speed of sound. When a vessel runs hot or pressurized, the engineering answer is a different measurement principle rather than a higher-spec ultrasonic.
Surface and atmosphere defeats. Foam absorbs and scatters sound, so a foaming surface returns little or no echo. Heavy vapor, steam, and dense dust attenuate the pulse along its path; condensate on the transducer face does the same at the source. Strong agitation or filling turbulence tilts and roughens the surface, scattering the reflection. Vacuum removes the carrier entirely and makes measurement physically impossible. None of these are firmware problems; they are limits of the medium, and the radar versus ultrasonic decision usually turns on exactly this list.
Transducer materials. The wetted, or vapor-exposed, face is typically ETFE or PVDF for chemical resistance, or filled polypropylene for general water service. ETFE and PVDF tolerate a wide range of acids, bases, and solvents in the vapor phase, which matters in chemical day tanks even though the liquid never contacts the sensor. For sanitary applications the relevant concern shifts to the process connection and cleanability rather than the transducer chemistry.
Open channel flow and standards. An ultrasonic level meter installed above a calibrated weir or flume becomes a non-contact flow meter: it measures head, and the transmitter applies the standard head-to-flow relationship and totalizes volume. ISO 1438 governs rectangular and V-notch thin-plate weirs, and ISO 4359 governs rectangular, trapezoidal, and U-shaped flumes. For instrument performance and ingress, IEC 61298 defines how functional and performance characteristics are tested, IEC 60770 covers transmitter evaluation, and IEC 60529 defines the IP enclosure ratings (commonly IP66 or IP67). Hazardous-area versions carry ATEX, IECEx, or NEPSI certification against the IEC 60079 series, typically as Ex ia intrinsically safe.
Chapter 5 / 06
Key Specification Parameters
A datasheet may list twenty or more lines, but a handful decide whether an instrument will work in your vessel. The seven parameters below are the ones to read first: measuring range, blind zone, accuracy and resolution, beam angle and frequency, process temperature and pressure, output signal, and ingress and certification. Each is explained with representative published values.
Measuring range. The rated range is the maximum distance the transducer can reliably resolve, and it differs sharply between liquids and solids on the same hardware because solids return weaker, scattered echoes. Representative liquid ranges span a compact VEGASON 61 at up to 5 m, a Siemens SITRANS Probe LU240 at 6 or 12 m, an Endress+Hauser Prosonic T FMU30 at up to 8 m, and a long-range KOBOLD NUS-4 at up to 25 m in liquids and 10 m in solids. Choose a range that exceeds the deepest measurement by a comfortable margin, since echo strength falls off near the limit.
Blind zone. Quoted as a fixed minimum distance, typically 0.25 m on compact probes and 0.3 to 0.6 m on long-range units. This is a hard mechanical constraint: the maximum level must stay below it. Treat the usable span as the rated range minus the blind zone, and design the nozzle height accordingly rather than mounting flush at the maximum fill point.
Accuracy and resolution. Published accuracy is usually the larger of a percentage of the set range and an absolute floor. The Siemens SITRANS Probe LU240 states 0.15 percent of range or 6 mm; the Endress+Hauser Prosonic T FMU30 states 0.2 percent of range or 3 mm; the KOBOLD NUS-4 states 0.2 percent of reading plus 0.05 percent of full scale. Resolution, the smallest detectable change, is finer, commonly 1 mm or better. Read accuracy as percent of the configured span unless a datasheet explicitly says percent of reading, since the two differ greatly near empty.
Beam angle and frequency. These travel together: higher frequency yields a narrower beam and smaller blind zone but shorter range. VEGA VEGASON transducers operate in the 18 to 70 kHz band, while many compact probes run higher, toward 200 kHz. Narrow beams (5 to 7 degrees) ease installation in cluttered vessels; wider beams reach farther in clean ones. Match the frequency variant to both the range and the obstruction map of the tank.
Process temperature, pressure, and ambient. Confirm three separate limits: the process or transducer-face temperature (often up to 70 to 90 degrees Celsius), the process pressure (often up to about 3 bar), and the ambient electronics temperature (commonly minus 40 to plus 60 degrees Celsius, lower with a display). Remote-mount the electronics when ambient heat, vibration, or washdown would exceed the head rating.
Output signal. The interface to the control system. Common options:
4-20 mA + HART: Two-wire loop-powered, the default, with HART carrying remote configuration, echo diagnostics, and a secondary variable on the same pair.
PROFIBUS PA: Digital bus for large DCS projects, multiple devices on one segment.
Modbus RTU: RS-485 serial, common on water and wastewater PLC systems.
IO-Link: Compact OEM and machine-builder integration with digital parameterization.
Relays: On four-wire controllers, for pump alternation, alarm, and totalizer pulse output.
Ingress protection and certification. Enclosures are typically IP66 or IP67 per IEC 60529 for outdoor and washdown duty. For classified areas verify ATEX, IECEx, or NEPSI marking against IEC 60079, usually Ex ia intrinsically safe for two-wire devices. Sanitary applications add 3-A or EHEDG process-connection requirements.
Chapter 6 / 06
Selection Decision Factors
Selecting an ultrasonic level meter is mostly a screening exercise: confirm the application is suitable for ultrasonic at all, then size and specify within that envelope. The ordered sequence below works as a fixed RFQ template, and most field failures trace to skipping step one rather than to a wrong model number.
Confirm ultrasonic is appropriate: Rule out vacuum, heavy vapor or steam, persistent foam, hot service above roughly 90 degrees Celsius, and pressure above roughly 3 bar. If any apply, move to radar, guided wave radar, or hydrostatic pressure before going further.
Medium and surface: Liquid or solid; flat or sloping (solids); calm or turbulent; clean or coating. Solids and turbulence both reduce usable range, so derate accordingly.
Range and mounting geometry: Measure the maximum and minimum distance from the planned nozzle. Pick a rated range that exceeds the deepest reading with margin, and verify the shallowest reading still clears the blind zone.
Beam path and obstructions: Map walls, ladders, agitators, pipes, and inlets. Compute the beam footprint at full range and choose a frequency or range variant whose cone stays clear, or relocate the nozzle.
Temperature and ambient: Check process temperature against the transducer limit and ambient against the electronics limit. Specify a remote-mounted transmitter where heat, vibration, or washdown demand it.
Output and control functions: 4-20 mA HART is the default; add PROFIBUS PA or Modbus for digital systems, and choose a four-wire controller with relays where pump control, alarms, or open channel flow totalization are required.
Process connection and ingress: Thread (G or NPT) or flange to suit the nozzle; IP66 or IP67 for outdoor or washdown; transducer face material (ETFE, PVDF, or PP) to suit the vapor environment.
Certification and total cost: Specify ATEX, IECEx, or NEPSI Ex ia for hazardous areas and 3-A or EHEDG for sanitary duty. Weigh purchase price against commissioning effort, false-echo troubleshooting time, and the cost of a wrong-technology choice that has to be replaced.
One dimension is easy to overlook at the quotation stage: serviceability and commissioning support. Ultrasonic instruments live or die on echo-mapping and false-echo suppression, so a clear local display, a well-documented setup wizard, HART DD or EDD files registered with the FieldComm Group, and accessible application engineering all shorten startup and reduce nuisance trips. Siemens, Endress+Hauser, VEGA, and Emerson maintain support and spare-part channels in China and most major markets, which matters more over a ten-year service life than a small difference in list price.
FAQ
What is the difference between an ultrasonic level meter and a radar level meter?
Both are non-contact, top-mounted, time-of-flight instruments, but the carrier differs. An ultrasonic level meter emits a sound pulse (roughly 18 to 200 kHz) that travels through the gas above the liquid, so its accuracy depends on the speed of sound and is degraded by vapor, dust, foam, temperature gradients, and vacuum. A radar level meter emits a microwave pulse that travels at the speed of light, is almost unaffected by gas composition, pressure, or temperature, and works to higher process pressures and temperatures. Ultrasonic is cheaper and adequate for benign atmospheric tanks of water, wastewater, and bulk solids; radar is the choice for vapor-laden, pressurized, or high-temperature service.
What is the blind zone and why can the meter not see into it?
The blind zone, also called the dead band or blocking distance, is the region directly below the transducer where no reliable measurement is possible, typically 0.1 to 0.6 m depending on frequency. After the transducer fires a pulse, the piezoelectric crystal continues to ring for a short time. During that ring-down the transducer cannot distinguish its own residual vibration from a returning echo, so any surface inside the blind zone produces a false or missing reading. Mount the instrument so the highest expected level always stays below the blind zone, and add 10 to 20 percent of margin to the rated range above the lowest level.
How does temperature affect ultrasonic level accuracy?
The meter converts echo travel time to distance using the speed of sound, which in air is about 331.5 m/s at 0 degrees Celsius and rises roughly 0.6 m/s per Kelvin, near 0.17 percent per degree Celsius. An uncompensated 10 degree error therefore shifts the reading by about 1.7 percent of the measured distance. To counter this, transducers carry an internal temperature sensor that continuously corrects the sound velocity. Compensation assumes a uniform gas temperature, so a strong vertical temperature gradient, for example sun-heated headspace over cold liquid, still introduces error that no single sensor can fully remove.
Can an ultrasonic level meter be used in vacuum, steam, or heavy vapor?
No for vacuum and poorly for heavy vapor. Sound needs a gas medium to propagate, so any vacuum or near-vacuum application makes ultrasonic measurement impossible. Heavy vapor, steam, condensate films on the transducer face, and dense dust clouds attenuate and scatter the pulse, weakening or losing the echo. Liquids that flash off vapor layers, hot service above about 70 to 90 degrees Celsius, and pressurized vessels above roughly 3 bar are all outside the comfortable envelope of ultrasonic. For those duties use a non-contact radar or guided wave radar level meter instead.
What does the beam angle mean and how do I size the standoff?
The beam angle is the angular spread of the emitted cone, commonly 5 to 12 degrees at the minus 3 dB points; high-frequency probes such as the KOBOLD NUS-4 reach 5 to 7 degrees. A wider beam at long range can clip tank walls, ladders, agitator shafts, weld seams, and inlet pipes, generating false echoes. Calculate the footprint diameter as twice the distance times the tangent of half the beam angle, then keep nozzles, internal obstructions, and the liquid surface clear of that cone. Aim the transducer perpendicular to a flat liquid surface; tilt and turbulence both scatter the return.
What output signals and accuracy should I expect on the spec sheet?
The dominant interface is a two-wire 4-20 mA loop with HART, with PROFIBUS PA, Modbus RTU, and IO-Link offered on some models. Typical published accuracy ranges from 0.15 to 0.5 percent of the set range, with an absolute floor often quoted alongside, for example the Siemens SITRANS Probe LU240 states 0.15 percent of range or 6 mm, and the Endress+Hauser Prosonic T FMU30 states 0.2 percent of range or 3 mm, whichever is larger. Resolution is finer than accuracy, commonly 1 mm or less. Always read accuracy as percent of the set span, not percent of reading, unless the datasheet says otherwise.
Can an ultrasonic level meter measure open channel flow?
Yes. Mounted above a calibrated primary device, a thin-plate weir or a critical-depth flume, the meter measures head (water height) and applies the head-to-flow equation built into the transmitter. ISO 1438 covers rectangular and V-notch thin-plate weirs and ISO 4359 covers rectangular, trapezoidal, and U-shaped flumes. The transmitter linearizes head to flow rate and totalizes volume. Because the instrument never touches the water, it suits wastewater, irrigation, and stormwater where contact sensors would foul. Accuracy of the overall system is governed mainly by the weir or flume geometry and approach conditions, not by the meter alone.