I usually run simulations including the volute, but it should be similar to set up. I use it to test a new design to create a part of a performance curve and get an idea of the power usage/ efficiency. I normally set the boundary condtions for the volute and wear plates as real wall - stator. The inlet and outlet I set to total static pressures, with the inlet as 1 atmosphere and the outlet at to a pressure to set the total dynamic head at which point of the performance I want to test.
For goals, I set inlet and and outlet surface goals of mass flow rate for convergence, and total average pressure for non-convergence.
For effeciency, I set another surface goal on the impeller vanes for z-component of torque (rotation about the z-axis).
For convergence, I set the the flow criteria to between 1-5kg/s depending on the impeller and for the torque between 20-50 Nm. You need to turn off the number of iterations for convergence or the simulation will stop before it has actually converged. When the simulation converges the inlet and outlet mass flow rates should be close to equal and opposite. From experience the torque is usually fairly jumpy compared to the flows so I use more loose criteria for it.
You'll need to fiddle around with your model and settings a lot because there are so many things that affect the simulation. I recommend starting with a large mesh setting at first and tighten it once you can get it working. Also, I've found when sometimes having troubles with convergence around BEP, to run a simulation at a lower head value.
Once the simulation converges, you should have all the data to calculate your efficiency. ie. flow, head, rpm & torque.