This project seeks to explore the way in which calibration of water quality models is made in WDS studying determinants such as water age and residual chlorine concentration. In this approach, we search for physical and chemical properties that help the calibration process, preventing it from being a simple numerical adjustment. There are two reasons why substance concentration varies throughout the network: reaction with bulk water and reaction with the conducts wall. These processes are associated with bulk coefficient (k(b)) and wall coefficient (k(w)), in each pipe. For the first one, kinetic models with constant decay coefficients have been applied for almost 30 years now. However, in recent times, these models have been considered inadequate due to their limited capacity to estimate residual chlorine decay with high precision. In this way, we take actual ideas of substance reaction in water to calibrate WDS based on models that are more rigorous in qualitative terms. It is used a variable bulk coefficient from Hua, Vasyukova and Uhl (2015), that contemplates water mixing from diverse sources. For k(w) we consider pipe material and age. This way, the parameters to calibrate are alpha and beta for each water source in the network and k(w) for each coherent group of pipes. Initially, we use hypothetical networks with different calibration scenarios, and then, as case of study, it is performed a water quality model calibration in the trunk network from Bogota, Colombia.