Hi Xiaoxu,
I hope I am actually helping you, because I do not have experience with such a problem. If you have constant pressure in the permeate region I do not think you have to include it in your model. I would think that your membrane has a certain flow resistance: the water flux through the membrane at any point on the membrane would be proportional to the local pressure difference over the membrane. For the particles you would impose a no flux boundary at the location of the membrane I would think.
I hope I am actually helping you, because I do not have experience with such a problem. If you have constant pressure in the permeate region I do not think you have to include it in your model. I would think that your membrane has a certain flow resistance: the water flux through the membrane at any point on the membrane would be proportional to the local pressure difference over the membrane. For the particles you would impose a no flux boundary at the location of the membrane I would think.