Bipolar membranes (BPMs) offer significant potential for electrochemical technologies, especially in CO2 electrolysis. Their unique ability to create pH gradients provides optimum chemical microenvironments at the cathode and anode for efficient, cheap and sustainable CO2 electrolysis. However, state-of-the-art BPMs are limited in their performance due to phenomena such as membrane degradation and membrane dehydration leading to low ionic conductivity, and hence their complete potential for various electrochemical technologies is yet to be realized. Previous studies have identified water transport as one of the main factors limiting the performance of BPM. A multi-physics model of BPM can serve as a useful tool to gain a comprehensive understanding of different performance limiting factors and thereafter propose solutions to overcome these limitations. The goal of the present study is to provide model based design guidelines for a BPM that facilitates efficient water transport, making it a suitable candidate for integration into a zero-gap CO2 reduction electrolyzer operating at industrial-grade performance levels.We have built a 1-D continuum model of a BPM kept in an aqueous (bi)carbonate electrolyte solution, incorporating physical phenomena such as water-dependent transport of ions and neutral species, field-enhanced homogeneous reactions (water-dissociation and acid-base reactions) along with CO2 phase transfer reaction at the BPM-aqueous electrolyte interfaces. The continuum model employs Poisson-Nernst-Planck equations to resolve local chemical species concentration. The effect of the electric field on the equilibrium dissociation constants of various charge-generating homogenous reactions is incorporated. This becomes important especially at interfaces within the domain where a sharp increase in electric field is observed. Unlike previous modeling studies, we also incorporate mass-conservation equation for water transport to resolve local water concentration inside the BPM. Local water concentration influences the transport of species in the BPM through their water-dependent activity and diffusivity. The electric potential distribution is resolved by incorporating water-dependent local permittivity of the BPM.Our model investigates both the reverse and forward bias operational modes of the BPM. In alignment with the experimental observation, in the reverse bias regime, the model identifies a distinct initial phase where current density is limited, with the majority of the current carried by protons and hydroxide ions at low overpotentials. As the potential increases, there's a rapid increase in current density, indicating a breakdown regime caused by an electric field-enhanced water dissociation reaction at the BPM interface. At even higher potentials, the polarization curve enters a third phase where the current density saturates characterized by water transport limitation. This aspect of the BPM polarization curve is uniquely captured by our model. Conversely, in the forward bias mode, as expected we do not observe any limitations due to mass transport, since water is generated at the bipolar junction. Here, the model forecasts a saturation of CO2 in the electrolyte, leading to gas bubbles forming at the BPM interface. To address the issue of water transport limitation, we conduct various parametric studies focusing on different membrane characteristics, such as individual layer thicknesses and their ion exchange capacities. The outcomes of these analyses provide crucial guidelines for optimizing these parameters and enhancing water management in BPM-based systems.