Tight gas reservoirs contain a significant potential for dry gas production despite of the very low formation permeability. Tight gas fields are commonly defined as having less than 0.1 millidarcy (mD) matrix permeability and low porosity (less than 10 %). The general approach to increase the gas production is the use of a well stimulation technique (hydraulic fracturing) as these formations are normally not economically producible. OMV is currently active in Pakistan and holding interest in tight gas production. It is a challenging task to produce gas at commercial flow rate in such low permeability reservoirs without stimulation. However, after hydraulic fracturing, in some wells issues were encountered due to the increased water production. Excessive water production is frequently upcoming issue in tight gas wells and is responsible for increasing water disposal cost and leads to a rapid decline in the hydrocarbon production. Production decline has been observed for many wells in this field. This problem has not been understood so far and needs investigation. The purpose of the present thesis is to investigate the nature of the problem and to make suggestions for potential solutions. Therefore, petrophysical and geological data were reviewed and were interpreted. Furthermore, overview of potential formation damages is also accomplished. Numerical models have been developed in order to perform an accurate history match and forecast the level of water production. Finally, possible solutions are suggested on basis of the findings in this thesis. Vertical lift performance curves were generated for the producing wells to allow reservoir simulations, by rebuilding the existing Prosper models. Several simplified and even more realistic reservoir models with sensitivity runs were made in order to simulate accurately the history of the field. To validate the changes in water production, Sawan-4 was chosen to simulate as the FMI did not indicate natural fractures around the well, only some induced fractures. Therefore, the model has been setup as a matrix system. The fractures were modelled explicitly with local grid refinements (LGR) as it is a common simulation method. In order to achieve an appropriate match, several sector models were applied. The size of the sector was varied until an accurate history match was obtained. Without adding any aquifer to the reservoir model, the water production was simulated. The pay zone is realistic, uniform sandstone confined by shale layers. The historical tubing head pressure (WTHP) matches the simulated pressures even using a simplified model. As the evidence for water production was also in the simulation results, a research for different solutions has been done to avoid this issue in the future.