SPHASE-Smoothed Particle Hydrodynamics Activated Sludge Engine

Historically, research in wastewater treatment focused on experimental pilot scale studies. Due to the increase in computational power, computational modelling of treatment plants became more important. Following a model description for a single sludge wastewater treatment system (Henze et al., 1987), the concept of activated sludge modelling (ASM) has been validated by a large number of experiments and is considered as a state-of-the-art concept for modelling biochemical processes. However, this model assumes perfect mixture within the reactor such that the effects of the hydrodynamics are neglected. This is an unphysical simplification and hence the research a few groups extended to the application of computational fluid dynamics (CFD) methods (Le Moullec et al., 2010; Zima et al., 2009; Wang et al., 2010) to simulate the hydrodynamics of wastewater treatment. Resolving the encountered complex multiphase problems proved to be difficult with conventional Eulerian CFD methods. In this project, a numerical engine for activated sludge modelling (SPHASE) based on the smoothed particle hydrodynamics method (SPH) is developed. Since SPH is a fully Lagrangian meshless CFD method, it is more suitable for simulating the hydrodynamics of treatment plants than Eulerian methods. SPH directly supports multiphase flow (Colagrossi et al., 2003), can incorporate process rates (Aristodemo et al., 2010) and accounts for transport phenomena (Tartakovsky et al., 2007). Further advantages are the easier consideration of reactor geometries, aeration systems and operational changes as well as the straight access to GPU implementations. Hence, SPH is the ideal method for simulating the hydrodynamics in wastewater treatment. The procedure intended in SPHASE is to develop a physically complete two-phase air water flow SPH model, where the oxygen concentration is governed by an advective diffusion equation. Subsequently, computational routines for key physical properties of biochemical processes (e.g. oxygen transfer rate) are added. Since the ASM model is considered as state of the art, the description of the biokinetics will be based on this well-established model. The key point of the project is to couple the locally resolved hydrodynamics to the ASM model such that local effects are considered in the biochemical processes. The coupling interface in SPHASE is provided by the local soluble oxygen concentration (SO). In summary, SPHASE is a fundamental research project with the aim of developing the SPH method further to allow for simulating biophysical processes in environmental engineering with a special focus on the application to wastewater treatment. SPHASE is the first application of SPH in wastewater treatment and therefore should also establish the method in this field. The project has a highly innovative character and its importance extends from the fields of wastewater treatment/water respectively sanitary engineering to fluid mechanics, hydraulics and computer aided simulations.

The main outcome of this project is a model to resolve the spatial distribution of biological concentrations in wastewater treatment (WWT) processes. This is facilitated by the novel idea of coupling a particle based method for fluid dynamics to a biokinetic model that predicts the evolution of biological compounds in activated sludge. The approach allows to accurately predict the movement of those compounds while simultaneously computing a state of the art model for the evolution of biological processes. Under consideration of external mixing the coupled model was solved for periods longer than 24 hours, which is required due to the slow dynamics of the biological system. The developed model was published in peer reviewed journals and implemented in free and open source software. The implemented solvers were designed to run efficiently on desktop computers and notebooks. For this, either the multicore capabilities of modern CPUs are used or optionally a graphics card. Their processing power - originally devoted to 3D applications - is leveraged to accelerate the scientific computations required to predict the fluid dynamics efficiently. By focusing on the problem domain, the developed software outperforms the current state of the art of solvers as shown in the accompanying journal publication. Based on the interdisciplinary work of physicists, computer scientists, mechanical engineers and environmental engineers, a very efficient model was developed and implemented. All outcomes of this development have been published as open success such that the results can be transferred to other domains that involve biochemical process and fluid dynamics.