Large amount of water is lost every day in water distribution networks (WDN) through leakage. Pressure Management by means of optimal pump operations is one vital way of leakage reduction in water distribution systems. Pumping in WDN can be of two types the first being pumping towards a storage tank which supplies the system by gravity and the second being pumping directly towards the network from storage reservoirs. Several researchers have studied the technique of pump scheduling as an option to minimize leakage in the first kind of systems. However, the inclusion of pressure-driven demand analysis for Real time control optimization in water distribution systems where pumping is done directly towards the pipe network is limited in the current studies. This research addresses this issue by investigating how effective it is to reduce leakage through optimization of pump operation in such systems using a methodology that first determines the location of leaking nodes according to traffic load, diameter of pipes and historical data of failure. These nodes are used during the leakage modeling of the network as points of leakage flow. Another set of nodes, critical nodes, were also determined that are expected to experience large drop in pressure. The constraints of the optimization are designed in such a way that the pressure at these nodes is always above the minimum service pressure in the network. The experimental design includes three categories of tests totaling 13 runs. The research was applied in the case study area of Braila, Romania water distribution system. Multi-objective genetic algorithm (NSGA II) was used to search for optimal pump scheduling in such a way that the total leakage volume during the simulation time and energy consumption by pumps would be minimal. Pressure driven demand analysis using the EPANET software together with WNTR Python library were utilized to do the hydraulic analysis of the model. Flow emitter properties of EPANET junctions were used to model leakage points in the network as orifices. Results show that the total leakage volume in the network resulting to the optimal operation of pumps has a reduction of 12 % relative to the existing leakage rate due to the customary operation. Energy consumption by the considered pump has also shown a 9% reduction. The resulting optimal pump schedules indicated that, for this specific network, the pump can be turned off only during night time where the demand is off peak so as to satisfy the minimum service pressure while minimizing leakage volume.