- Try to avoid agitated tanks even when the agitator is below the surface of the material. Agitation can produce whirlpools or cavitation, which may attenuate the signal or cause it to bounce off a vessel wall. In some cases, the agitation may be so extreme that the measurement you are trying to make is "what the vessel level would be if the agitator was turned off" (Figure 2). This is not a real measurement, and it may not be possible to make it with any degree of confidence or accuracy.
- Sparged tanks, where air or another gas is introduced into the vessel by means of diffusers or spargers, can cause bubbles or foam to form on the surface of the material. It is good to avoid this application. A layer of bubbles or foam can attenuate the signal either entirely or partly. If it attenuates the signal entirely, there will be no echo return. It is more insidious if it only attenuates the signal partly. A false echo can occur from somewhere in the foam layer, rather than either the surface of the foam or the surface of the liquid below the foam (Figure 3).
- Avoid foam. Foam can do three things to the accuracy of the level measurement, and all of them are bad.
1. It can attenuate the signal so that there is no echo or only an intermittent echo. Intermittent echo can sometimes be dealt with using a sample-and-hold circuit or algorithm in the transmitter so that the level doesn't change until the next good echo. Sometimes, however, that can be dangerous, as in the case of a vessel where the level is quite near the maximum fill point.
2. Foam can provide a false reading of the true level. You can get a reading from inside the foam layer, instead of the actual level.
3. Foam clumps can cause the echo to be deflected away from the vertical, and the sensor may receive an echo that has made one or two hops against the side of the vessel, yet still be a high enough signal to fool the transmitter.
- Avoid volatile liquids. Back when I was in sales, I sold an ultrasonic transmitter to a major northeastern United States utility for the measurement of level in huge bunker oil tanks. The sensor was installed in early November, and it worked acceptably well until mid-May of the following year, when the customer reported that the sensor was insisting that the level in the tank was several feet higher than it actually was. This "ghost level" phenomenon is a function of the volatile liquid in the tank. As the ambient temperature rose, the vapor blanket on top of the bunker oil began to become more dense and increased in height. The ultrasonic sensor picked up the top of the vapor layer, instead of the actual oil level in the tank. By late June, the sensor was regularly reading 80% to 100% because the early summer heat had caused the vapor blanket to fill the tank. We replaced the ultrasonic sensor with a FMCW radar sensor, which worked correctly, and I learned something.
- In solids and powders, you may have to aim the sensor at a point that is not 90 degrees to the level surface (perpendicular to the vertical axis of the vessel). You may want to aim the sensor because of rat-holing and angle-of-repose issues at the top, midpoint or bottom of the angle of repose. Try to have the transmitter calculate what the actual level might be. At least one vendor has developed a multiple sensor array that can scan the angle of repose and determine what the actual filled volume of the vessel is.
- Avoid pressurized tanks. The speed of sound changes with temperature and density, and pressurizing the vapor space above the level can affect the density of the vapor space and, therefore, the speed of sound.
Ultrasonic Open-Channel Flowmeters
One of the most important applications for ultrasonic level sensors and transmitters is measuring open-channel flow (figure 4).
Most of the same caveats apply to ultrasonic level sensors used as flowmeters as apply to ultrasonic level sensors used as tank level measurement devices. There are a few more:
- Avoid wind and sun. Wind can blow through the vapor space and attenuate the signal or blow it off course. Sun can raise the temperature of the sensor housing itself beyond the operating temperature range of the device-and higher than the ambient temperature.
- Make sure that there isn't foam on the surface. This can happen often in nitrifying wastewater discharges.
- Make sure that there is not too much turbulence or ripples (or if the flume or weir is large enough, wave action) on the surface.
- Above all, make sure that the flume or weir is installed correctly. Many problems blamed on the ultrasonic transmitter are actually problems that are caused by the flume not being installed level both horizontally and vertically, as well as front to back through the measurement zone.
- Make provisions to keep ice from forming on the sensor in the winter or dripping condensation in the summer.
The One-Trick Pony--Not!
Ultrasonic sensors are simple to understand, easy to install and inexpensive. It's easy to go to them as the unthinking sensor of choice for level applications, just as many people go to differential pressure level sensors. Yet, as many users have found, ultrasonic sensors and transmitters are tricky beasts. As with any other field instrument, applying an ultrasonic level sensor too far outside the manufacturer's recommended application envelope is destined to fail, and sometimes fail spectacularly. But, if you follow these basic guidelines, you will have successful ultrasonic level installations.