Computational fluid dynamics (CFD) simulation is a powerful tool for modelling three-dimensional flow from first principles.

Understanding the distribution of flow within a tank and separation of water, chemicals or sludge in water and wastewater treatment is increasingly necessary for the modern engineer to optimise the design and operation of engineering solutions.

To this effect, The Fluid Group have been working on a number of projects which combine the opportunities that CFD can achieve with Microbiological and Chemical Process Simulation. Amongst a raft of interesting results, one simple conclusion has been identified again and again… “Assume nothing… ”

With all modelling in the process Industry (for example, Hydromantis’ GPS-X or EnviroSim’s BioWin for Wastewater Treatment Simulation) a key requirement is the effective understanding of boundary conditions, namely the influent characteristics of the water or wastewater to be treated, flow and volume of the process units.

To the untrained eye, the simplest concept to grasp is volume.

You have an existing plant or a design on paper. It has physical properties of length, width and depth. Therefore you have a volume.

However, most process software assumes a perfect flow characteristic (plug flow or complete mixing). The concept of dead volume or short-circuiting is accepted but is considered outside the scope of the software. Even worse, any discrepancy between simulation and experiment is assumed to be a failure in parameters describing the process (because that’s complex), not the physical volume (because that’s simple).

In a recent example sited by The Fluid Group, the effective volume of a key process unit was approximately three times (2.8) less than the measured volume at high flow (short-circuiting restricting the effectiveness of flow distribution). Clearly the key driver to improve process performance is to understand chemistry, but here the flow performance is as important, if not more so. CFD models complement process simulation and the ability to consider the problem from one or more different angles with separate techniques can be essential for real understanding.

It’s a salutary lesson that on occasion it is the things you think you know best, that actually you know least. In short, sometimes in simulation, it is best to ‘assume nothing’.

The Fluid Group is based in Oxford, UK. www.thefluidgroup.com