This project covers the development of a controller for a laboratory system that emulates a sewer network. A sewer network receives wastewater from residential areas, industries, runoff rainwater and transports them to a wastewater treatment plant (WWTP). In one particular case, at Fredericia, there are some heavy industries discharging occasionally large amounts of wastewater over a short time. Such fluctuating flow and/or contaminant concentrations causes a stress on the WWTP. In this research project, we worked on a laboratory setup imitating a sewer network. We use a buffer tank to hold back water, a gravity sewer pipe to transport water, pumps to continuously generate a typical pattern of water flow and a overflow tank to collect water from sewer pipe outlet. Now, this led us to a problem statement: How can a laboratory setup that mimic a real sewer network be assembled so that we can later utilize MPC, along with disturbance predictions and a storage tank that results in stable working conditions for the wastewater treatment plant. Controlling the output flow from the buffer tank into the top of sewer pipe gives the possibility to smooth out the flow of water at the end of gravity sewer pipe. With a simplified model, prediction of flow disturbances, performance function and constraints, a Model Predictive Controller (MPC) was developed and the final results were promising.