Team

The most secure access to clean drinking water is by means of a connection to a water distribution system. In cities of economically developed countries, the population is almost entirely dependent on these large-scale systems to fulfil their vital needs. Urban water supply systems are therefore considered critical infrastructure. Increasing urbanisation, water scarcity, extreme weather events and the need to save energy as a result of climate change as well as ageing components are exposing these systems to hazards and stress. Digitalisation offers opportunities to meet these challenges, but brings with it new threats from cyber attacks. A failure of the drinking water supply can have catastrophic consequences.

How can systems cope with the uncertainty caused by such diverse hazards?

One approach is to design them according to the principle of resilience. Resilient technical systems are able to maintain a minimum level of functional performance even during critical events and subsequently regain their full functionality.

In my research, I am investigating how urban drinking water distribution systems can be designed and operated so that they become resilient. To this end, I consider the hypothesis that decentralisation increases resilience. By using valves to divide a water distribution system into self-sufficient, stable subsystems, each with its own potential source (pump, tank), the effect of critical events is to be contained within the affected subsystems, while energy efficiency is increased under undisturbed conditions and the monitoring of the subsystems is simplified. Decentralisation is achieved not only in terms of system topology, but also by equipping the pumps and valves with their own controllers, so that system-wide control results from the interaction of the individual components instead of the entire system being dependent on a central control unit that represents a single point of failure.

A particular focus of my work is on developing and testing the controllers. These should fulfil the four functions for achieving resilience: monitoring, reacting, learning and anticipating. Using methods of time series analysis, the current state of the system is monitored and checked for the presence of a fault. If a fault is detected, the controller adjusts the setpoint based on the fault in order to counteract the influence of the fault. Analysis after a fault has occurred allows the controller to learn from the fault and adjust the threshold values for detecting the presence of a fault. New faults can be anticipated through prediction based on empirical values.

In addition, I consider water distribution systems as socio-technical systems. This means that I also consider the role that consumers can play in the resilience of the system.

I use a test rig for the experimental validation of my research.