Further analysis discloses that as a rule the sensors affected are those that are required to measure a cold product in very moist ambient conditions. In other words in operations such as are often to be found in breweries or dairies. What is the reason for these failures? To answer this question it is necessary to understand the principle of hydrostatic filling level measurement. In the first instance a sensor is installed in the floor of the tank. The membrane of the sensor is loaded by the height of fluid in the tank as well as the air pressure. As is generally known air pressure alters with the height of the measurement location above sea level, but very significant variations in air pressure also occur as a result of weather conditions. For this reason these variations or differentials must be continuously compensated for. The following example makes clear the necessity for this so-called relative compensation: A 3 metres high milk tank, vented to atmosphere, produces a pressure loading on the membrane of 300 mbar. Between meteorological conditions of high and low air pressure a pressure difference of as much as 50 mbar can occur, which here can lead to an error of more than 16 %.There are two types of sensors that are independent of climatic conditions:
- a pressure measurement cell with a compensation capillary to compensate for atmospheric pressure, and
- a double membrane, in this case fitted, however, with a Gore-Tex filter as a moisture barrier.
In the case of the so-called relative pressure measurement cells with a compensation capillary a thin tube directs the atmospheric pressure from the environment on to the rear face of the measurement membrane. Since the air pressure is now exerted on both faces of the membrane, only the hydrostatic pressure that is actually of interest, that of the medium in the tank, contributes to the resultant measurement signal.In the other method using a sealed relative pressure measurement cell and an in-built Gore-Tex filter one pressure membrane measures hydrostatic pressure and the other measures atmospheric pressure.Both methods have however a crucial disadvantage: the diffusion of gaseous water vapour through the membrane cannot be prevented. In particular for the operating conditions already described where there is a cold medium in the tank and moisture outside the tank a problem now arises: vapour passes through the Gore-Tex filter element and condenses at the coldest location as soon as the temperature falls below the dew point. This coldest location is unfortunately the measurement cell itself.Since the condensation process develops a momentum that is, so to say, self-perpetuating, a not insignificant quantity of water collects in the measurement cell. The moisture inside the measurement cell leads to sensor drift, fluctuations in measured data, and eventually to the destruction of the cell.The simplest solution, one that is not very helpful for the facility operator, however, is to limit the permissible ambient moisture in the technical specification to the typical value of 80 % relative humidity quoted by some manufacturers, including some leading names. Better advice, although still not very practical, is the suggestion that the capillary line should terminate in a dry room space.