Some considerations for these marvels of technology.
Neither technology is very good at handling foam on the top of a liquid. Ultrasonic usually bounces off the top, radar bounces at some indeterminate point depending on the radar technology and the foam density. There's no application chart for foam, it's an empirical test to see what the functionality of either technology is.
Both technologies reflect off sidewalls and objects like ladders, creating false echoes. There are generally algorithms for accounting for permanent structures whose echoes are always the same.
Each can have varying speed-of-response, model/technology specific. Are both fast enough to register the process changes?
The deadband below the transducer fact in which a return signal can't be reliably read has reduced over time for the most recent generation of radar (some brands) but is still close to about 10" or a foot for ultrasonic. Can the installation tolerate the deadband?
Ultrasonic can not handle any vacuum or pressure over atmosphere without significant error or loss-of-echo due to speed-of-sound variations. Is that deal killer?
Ultrasonic is worthless in an atmosphere of Carbon dioxide. Just does not work.
Although some ultrasonics have an adjustment of speed-of-sound in a given atmosphere, that's fine for temperature when the adjustment is made, but there's no accounting for change in speed-of-sound for other temperatures and speed-of-sound varies by temperature. Is the process atmosphere in the vessel constantly isothermal at a fixed temperature?
Boot-up time for initialization on a couple radars I've worked with is about 45 seconds before the initialization 4.0mA output drives to a valid value. Is initialization time a factor?
Radar will shoot through a plastic vessel and bounce off a ladder laid against the outside of the tank, providing false level readings. Ask me how I know.
In an open-to-atmosphere vented tank (a requirement for ultrasonic), high relative humidity will condense on fall or spring days when the temperature drops below dewpoint, depositing condensed water on the face of the transducer or antenna which can affect either technology.
