Selective bioclogging targets the biofilm growth in highly permeable zones of reservoirs or aquifers to divert water into low permeability zones. It alters the hydrodynamics of the subsurface flow systems to favorable performance conditions. Applications may include microbial-enhanced-hydrocarbon-recovery (MEHR) and bioremediation. Despite its success at the laboratory scale, application of bioclogging at the reservoir scale is hindered by the lack of understanding and advanced modeling and prediction tools. To understand controls of bioclogging processes at the reservoir scale, a Reactive Transport Model (RTM) has been developed in this work for in situ biostimulation of L. mesenteroides. This fermenting bacterium produces the biopolymer dextran in the presence of sucrose. As a first step, we considered the flow, transport, and bacterial growth and dextran production reactions in a single phase fluid (water) system, because most reactions occur either in the water phase or at the water-solid interface. Parameters for biomass growth and dextran production were obtained from column experimental data. The numerical experiments were carried out using the spatial distribution of porosity and permeability extracted from open-hole well logs collected at a characterization well near the King Island gas field in Southern Sacramento basin in California. The numerical experiments suggest that there exists an optimum range of injection rates (between 543 and 1,195 bbls/day). The volumetric injection rates need to be sufficiently fast to overcome microbial growth and clogging at the vicinity of the bore wells. They also need to be low enough to allow sufficiently long residence times for dextran production. Results show significant dextran formation and the associated porosity and permeability alterations to divert water into low permeability zones. The bioclogging effectiveness, measured by the percentage of the water diverted into the low permeability zones, varied between 10 to 75% depending on injection conditions. With the same total mass injection rates of sucrose, increasing flow rate is more effective in selectively bioclogging highly permeable zones than increasing sucrose concentration. Other processes, including the attachment of biomass to the solid surface without being washed out, are also important. The developed model offers a powerful tool to optimize injection conditions for effective bioclogging in naturally heterogeneous reservoirs.