Digital Twin II

Impact of corona measures and drought on water demand

A digital twin is a digital replica of a living or non-living physical entity. In the context of technical systems this often involves a digital representation of the physical system, which mirrors as closely as possible the behaviour and results of its real twin. This mirroring can achieve satisfactory fidelity thanks to the input of data from a variety of sources. The linking of different models is also common.

Over the last months we have experienced extraordinary circumstances: a large part of the population stayed at home because of the corona measures, and many companies and organisations closed for long periods during the lockdowns. Holiday practices also changed. More people remained in the country, while fewer tourists visited from abroad. Furthermore, in the spring of 2020, for the third year in a row, we were affected by problems related to drought. This is only a first indication of the impact of climate change. It is expected that drought periods could become even longer and more intense in the decades to come.

Both elements – the corona measures and drought – present a challenge to drinking water utilities: people are not where they usually are, and water consumption behaviour has changed. This refers to both the total consumption, as well as its pattern over the course of the day, as well as the geographic spread.

We want to use a digital twin to gain an understanding of how these changes take place. For example, on the basis of mobile telephone data (and possibly other complementary data) from the last few months, to acquire insight about the presence of people at home and/or about specific water-use points. This information will make it possible to improve the modelling of water demand.

Water demand also depends on changes in the water-use behaviour resulting from the corona measures. When people spend more time at home, they will for instance shower or use the washing machine at different times. And in drought periods they will probably water their gardens more often; in this regard, weather data are also important for the digital twin.

It may well be that the changes in people’s presence and behaviour will persist even after the corona measures are relaxed. Working-from-home will probably be more common than in the past. This would mean that changes in water demand patterns on working days would continue over the long term. Digital twins can therefore assist drinking water utilities anticipate such developments.

Digital twin for Eindhoven

In the earlier Digital Twin Demonstrator project (BTO 2020.029), the first steps were taken in the development of a digital twin of the drinking water distribution network of Eindhoven. Mobile telephone data were used to gain insight about the presence of people in an area (on an hourly and daily basis). The water demand (per hour and per day, for different types of days: regular days, holidays and hot days) was modelled on this basis, making it possible to determine the performance (flow volumes and pressure) of the distribution network.

In the present project, we aim to extend this further by:

• collecting mobile phone data, and possibly other complementary data, to establish the presence of people in different (sub) areas during the lockdown, summer period and autumn of 2020;
• collecting readings of water consumption for the same period;
• modelling the industrial and non-household consumption and the impact of drought on water demand;
• modelling the water demand and performance of the distribution network for the regular situation, the lockdown (spring), summer and fall of 2020.

Distribution network performance under extraordinary circumstances

The objective of the project is to provide drinking water utilities, by means of a digital twin of a drinking water distribution network, with insight into the impact of drought and of historically exceptional situations, such as the corona measures, on water demand – absolute increase or decrease, geographic distribution and spread over time – and into the performance of the distribution network. This will enable the identification of possible problem areas, such as zones where pressure is too low.

A digital twin can additionally be used to test possible improvements in the operation and/or design of distribution networks, and for all kinds of scenario studies.

The expected outcomes of this project are:

• A determination of the added value of different data sources [for example, holiday staggering from Statistics Netherlands (CBS), traffic data from Rijkswaterstaat, or use of public transport] to establish the presence of people in an area (and thereby the water demand).
• Insight into the modelling of water demand for historically exceptional situations.
• Extension of the digital twin (developed in the Digital Twin Demonstrator project BTO 2020.029) with a view to modelling non-household consumption and the impact of drought.
• Determination of the impact of corona measures, long-term behaviour changes and drought on water demand and pressure distribution for Eindhoven (building on the Digital Twin Demonstrator project BTO 2020.029).