Water is a scarce resource which is becoming increasingly inaccessible. It is therefore necessary, in most parts of the world, to capture, transport and allocate it efficiently and thoughtfully. The implementation of monitored water distribution networks is often expensive. The purpose of this project is therefore to monitor leakage and consumption in a non-pressurized agricultural irrigation system using only inexpensive and easily installed pressure sensors. We modeled the water network to automatically simulate a leak randomly through the network. These simulated pressures serve as a dataset to train, test and validate a Random Forest algorithm that detects the leaks. Through pressure measures, the model can locate the junction closest to the leak with an accuracy of 96.24%. This approach therefore allows leaks detection in a water distribution system without the use of expensive flow sensors.