Comment

01 March 2005

Ultrasonic is on the level

Microprocessor-implemented fuzzy logic technology mitigates interference.

Sonic (up to 9500 Hz) and ultrasonic (10 to 70 kHz) level switches operate either by the absorption (attenuation) of acoustic energy as it travels from source to receiver or by the frequency change of a vibrating diaphragm face, oscillating at 35,000 to 40,000 Hz.

Ultrasonic level transmitters operate by generating an ultrasonic pulse and measuring the time it takes for the echo to return. If the transmitter sits at the top of the tank, the pulse travels in air (at a speed of 1087 ft/sec at 32°F, or 331 m/sec at 0°C); therefore, the time of travel is an indication of the depth of the vapor space above the liquid in the tank.

If the transmitter is on the bottom of the tank, the time of travel reflects the depth of liquid in the tank, and the speed of travel is a function of what that liquid is. In the case of water at 77°F (25°C), an ultrasonic pulse travels 4936 ft/sec (1505 m/sec).

To prevent false readings

To understand the capabilities and limitations of ultrasonic instruments, one must understand the conditions that determine the characteristics of sound: temperature, reflection, propagation, and absorption.

Temperature compensation is essential in ultrasonic level measurement because the velocity of sound is proportional to the square root of temperature, and, in the case of air, it changes by about two ft/sec (0.6m/sec) for each degree Celsius change in temperature.

To measure the time of travel of the echo of an ultrasonic pulse, it is essential that some of the sonic energy reflect. Liquids as well as solids with large and hard particles are good reflectors.

Fluff and loose dirt have poor reflecting characteristics because they tend to absorb the sonic pulse. It is also important that the reflecting surface be flat because the angle of reflection equals the angle of incidence.

Therefore, if the sonic pulse reflects from a sloping surface, its echo will not return to the source, and the roundtrip travel time, or time of flight (TOF), will not accurately reflect the vertical distance.

Irregular surfaces result in diffuse reflection in which only a small portion of the total echo travels vertically back to the source.

The travel—propagation—of sound results in its dispersion, loss of intensity. The intensity of sound decreases with the square of distance; therefore, the echo becomes exponentially weaker as the range of the instrument is increased.

The decrease in sound energy is a result of not only dispersion (traveling distance) but also from absorption in the substance through which it travels. For example, an ultrasonic sound wave traveling in dry and dust-free 60°F (20°C) air weakens by one to three dB for each 3.3 ft (1m) of travel.

Clearly, an ultrasonic transmitter is subject to many varieties of interference that will affect the strength of the echo. Many of these physical phenomena are beyond the control of the instrument manufacturers although microprocessor applications help compensate for some of the symptoms of their presence.

All the transmitter can be expected to do is accurately compute the roundtrip time of flight based on the first segment of the echo, provide temperature compensation or heat if condensation in the transducer is a possibility, and provide a strong well-focused ultrasonic pulse. It cannot change the reflection, propagation, or absorption characteristics of the process.

Ultrasonic level devices can be used for both continuous and point measurement. The point detectors for measurement of gas-liquid, liquid-liquid, liquid-foam, or solid-gas interfaces can group by design into damped sensor of on-off transmitter categories.

They also can break down by method of packaging as single-element and two-element units. The continuous level detector designs can fall into the categories under-liquid sensors and above-liquid sensors.

Most designs use a 20kHz or higher (up to 70kHz) oscillator circuit as the ultrasonic signal generator. Some designs incorporate filters, discriminatory circuitry in electronics, or software in microprocessors to prevent false readings that might result from random noise.

Technology on the march

In the mid to late 1990s, a number of new developments occurred in ultrasonic level sensor technology. One such improvement involves the continuous monitoring of the depth of sludge blankets. This ultrasonic sensor works using a dual-head assembly and a microprocessor, and it is useful in the wastewater treatment industry.

Another development was improved filtering capabilities via microprocessor-implemented fuzzy logic technology that mitigated interference reflections and factored in deposit buildups that occur in tanks over time.

Another advancement involves automatic self-calibration of ultrasonic sensors, which can correct for some of the effects of changing vapor space composition or temperature and can provide more than a single calibration target.

The multiple calibrations targets come as precisely located ridges in a sounding pipe and can result in level measurement inaccuracies within 5 mm over a distance of 30 meters.

New techniques in ultrasonic level measurement are revolutionary, utilizing the changes in the speed of sound in the tank wall. This speed does change when the other side of the tank wall is wetted and therefore can work for level measurement in both the transmission and the echo mode of operation. The testing of such units in pilot applications has taken place.

The mid to late 1990s saw the incorporation of serial data communications links such as RS485, and the HART protocol became available on several manufacturers' systems to link the output data to higher-level systems.

In the late 1990s and 2000s, the use of various digital fieldbus technologies became commonplace as the output signal of ultrasonic level transmitters. Foundation fieldbus and Profibus PA and DP are the main fieldbuses currently in use.

Another development in the ultrasonic or time of flight technology is several manufacturers have designed their systems to allow the use of the same transmitter between both ultrasonic and microwave (RADAR) technologies. This allows upgrades or changes in requirements between the two related devices but still allows inventories of transmitters to be cross-utilized between the different but similar sensor technologies. The trend is to utilize the similar functionality of the two related technologies from a hardware and software/firmware point of view that benefits application users.

The embedded microprocessor can compensate for the spurious signals. It can suppress false echoes to prevent any misinterpretation by increasing the detection threshold at one or more fixed points. If the reflected signal is strong enough, the transmitter will follow the true echo, even during the increased detection threshold.

Should the reflected echo signal be smaller than the threshold, the transmitter will look around the increased threshold and will hold the output until the true signal appears again.

Reflective properties density

Ultrasonics is now one of the traditional methods of level measurement. As probe-type level switches, they can give reliable performance even on difficult slurry- or sludge-type services.

The main advantages of ultrasonic transmitters are the absence of moving parts and the ability to measure the level without making physical contact with the process material. In some specialized designs, even penetration of the tank is avoidable.

The reliability of the reading is unaffected by changes in the composition, density, moisture content, electrical conductivity, and the dielectric constant of the process fluid. If temperature compensation and automatic self-calibration are included, the resulting level reading can be accurate to 0.25% of full scale.

In terms of limitations, the ultrasonic level transmitter is just as good as the echo it receives. The echo can be weak as a result of dispersion, which reduces sound intensity by the square of distance, and absorption, which in dry air reduces its energy level by one-to-three dB/m. The energy content of the echo will go down even further if the bin is tall, if the vapor space is dusty, or if it contains foam or other sound-absorbing materials such as water vapors or mists.

In addition to the problem of weak echoes, another potential problem is the reflective properties and density of the process surface. If that surface is sound absorbing like with fluffy solids, sloping, or irregular, causing a diffused reflection of the ultrasonic pulse, the result can be an error, as the roundtrip time of travel might not correspond to the vertical distance between transmitter and level. DT

Behind the byline

Nicholas Sheble (nsheble@isa.org) edits the Control Fundamentals department. The source for this critique is Instrument Engineers' Handbook, Process Measurement and Analysis, CRC and ISA Press 2003, Bela Lipták, Editor.