Hi Marlon,
I tried to modify the example the way you showed in the screenshot. The model had difficulty initializing and won't run. Did you need to do anything else to simulate the model and get zero speed?
I am actually not surprised to see an error with just a speed sensor connected to the shaft. Now the turbine has no external load or resistance, AND no inertia! If you add a rotational damper into the model, parallel to the sensor, you will see some shaft rotation again.
Now back to why the "load" subsystem is set up the way it is. I believe it's representing a dynamometer setup. Some dyno can maintain a constant speed (through its internal feedback system I suppose) and measure the resulting torque output from the system, which in its idealized form is modeled by a velocity source and a torque sensor. You can defintely replace that with a more realistic mechanical load.
Hope this helps.
Yifeng
