Radar is unaffected by density, temperature or pressure. It is affected by tank geometry and by the product's dielectric constant.
Radar level measurement gained ground because it removes the variables that trouble other technologies: changing density, temperature variation, vapour pressure. It measures distance from the time of flight of an electromagnetic wave, and that is all.
But it has limits of its own, and they are geometric.
Dielectric constant of the product
This is factor number one. The higher the dielectric constant, the stronger the reflected echo. Water (≈80) reflects very well. Light hydrocarbons (≈2) reflect poorly, and in a tall tank the return signal can fall below the detection threshold.
For low-dielectric products there are high-sensitivity versions — or you use guided-wave radar, which concentrates the energy along the probe instead of spreading it into free space.
Obstructions in the emission cone
Internal ladders, agitators, heating coils and the nozzle itself generate false echoes. The instrument can lock onto a fixed obstruction and report a constant level.
The fix is in the installation, not the configuration: place the nozzle away from the wall and clear of internal structures, and run false echo mapping with the tank empty.
Nozzle and antenna
A nozzle that is too long or too narrow attenuates the signal and creates internal reflections. The antenna must extend past the lower face of the nozzle. A badly sized nozzle is the most common cause of a radar that "never worked properly" since start-up.
Turbulence and foam
An agitated surface scatters the signal. Foam absorbs it. In both cases the echo weakens. A stilling well solves turbulence; persistent foam sometimes calls for a different technology.
