The highly instantaneous nature of turbulent multiphase flows gives rise to interesting phenomena, which are expected to significantly affect the fate of microorganisms in natural water bodies. However, current state-of-the-art models for water quality prediction often make strong simplifications with respect to turbulence due to its complexity. Because of this, heterogeneities in space and time within these complex systems are usually improperly captured, and, therefore, large uncertainties related to quality predictions exist. Using massively-parallel computer simulations, we are aiming to reveal the relative importance of the various transport and conversion subprocesses to provide important insight into the mechanisms through which the heterogeneities of the complex flow affect the selfpurification capability of a contaminated water body. This will pave the way for a general improvement of water quality assessment methods, which is a crucial requirement to ensure an acceptable water quality in many environmentally-relevant processes.