There are generally two approaches to grounding pressure sensors. Both work just fine, as long as they are used in isolation. However, when both are used together, problems can occur.
The first is to use a chassis, or case, ground. This means that the manufacturer has designed the transducer to make contact to earth ground through the case/housing of the sensor. This approach allows the transducer to be naturally grounded during the installation process.
The second approach is to use a drain, or ground, wire that is connected away from the actual sensor; the distance depends on the cable length and the specific application. Generally, the lead wires, including the ground wire, are installed into a controller box of some sort. This is frequently the case when a common ground is needed between the controller, sensor, and other equipment operating as a system.
Either method by itself is an appropriate way to ground a transducer. However, when both are used together on the same transducer, problems can arise. Submersible pressure sensors are a good example. The ground potential of the fluid where the sensor is submersed could easily be a different potential then a control box mounted 30’ or more away. This difference in potential generates a voltage and acts like a battery. It can cause performance problems, and in more extreme cases, can even cause corrosive electrolysis (see photo above).
The best practice is to use one method or the other. Many sensors are designed to have both options available. If you use the case ground approach, then simply do not connect the ground wire to anything. If you use the ground wire approach, then be sure the case has not been grounded by the manufacturer or the application method.