Data （959）for SP-Pinns

The dataset contains 959 samples obtained from simulations of a seawater reverse osmosis membrane system. Each sample includes 12 variables describing spatial location, structural characteristics, flow information, and transport properties.The first three variables represent the spatial coordinates of the sampling point. Variables 4–7 describe the physical and geometric structure of the membrane channel; these parameters characterize the structural configuration but are not considered as prediction outputs in the present study.Variables 8–10 correspond to the velocity components in the three spatial directions, representing the local flow field. Variable 11 represents the pressure drop at the corresponding spatial location.Variables 11–12 represent the solute concentration and the mass transfer coefficient, respectively. These quantities are computed using the convection–diffusion equation, where the convective terms depend on the velocity components obtained from the flow field and the pressure drop determined by the governing fluid dynamics.Overall, the concentration and mass transfer coefficient are derived by coupling the convection–diffusion equation with the flow information described by the Navier–Stokes equations, where the velocity components and pressure drop provide the necessary hydrodynamic conditions for the transport calculation.