South East Water

HELPING SOUTH EAST WATER PREVENT LOSS OF WATER SUPPLY TO CUSTOMERS

CLIENT

REQUIREMENT

South East Water (SEW) needed a tool that could assist them in understanding the risk of water supply loss in their network and ways this could be anticipated and addressed.

CHALLENGE

Modelling the complicity of SEW’S water transfers was a real challenge, especially as the model had to be fast and present the results in a clear manner.

SOLUTION

Our ROSE tool is able to forecast the effects on the water network of unexpected adverse weather conditions, planned and unplanned maintenance, as well as variations in demand. It allows the water company to find the best way of operating the system to minimise risk of loss of supply.

Beast from the East puts SEW’s network under strain

South East Water supplies water to 2.2 million customers in the south east of England. In 2018, a combination of an unusually cold weather front (the ‘Beast from the East’) with a sudden subsequent rise in temperature, led to freeze-thaw incidents. The resulting burst pipes and leaks increased demand for water by more than 20%. The situation was made worse as planned maintenance work meant the network was already strained. Water levels in service reservoirs fell, resulting in a significant number of customers lose their tap water supply, some of them for several days. This highlighted the need for better tools that could understand outage risk and forecast when a combination of planned service reservoir maintenance work and an unplanned event significantly increases risk. This would allow SEW to manage the risks and be ready to respond to events and avoid loss of supply to customers.

Reducing the complexity of SEW’s network

Modelling the complex interactions within SEW network with the various transfers between reservoirs, water treatment works and consumers was a real challenge. Previous models used complex physics modelling and were too slow to deal with a dynamic critical situation. We worked with SEW to identify the key features of the network and effective ways they could be modelled.

A Rose by any other name

We developed the Risk of Supply Events (ROSE) model. ROSE looks at the level of every reservoir and evaluates all the options available to maintain each reservoir. It then optimises the choice of interventions to ensure all reservoir water levels stay above the required level (so all customers get an uninterruptedsupply of water), and warns if this is not possible. Inputs such as different demand profiles can be varied to simulate challenging what-if scenarios. ROSE is fast, providing timely results, and these matched the detailed models – and the expert judgement of the engineers when we stress tested it – well.

COVID-19 puts ROSE to the test

ROSE provided a significantly faster, more usable tool for SEW, producing results dramatically faster than the original model for each individual water resource zone. This time saving has significantly improved SEW’s ability to understand and anticipate the risk of loss of water supply to their clients. The delivery of ROSE coincided with the Covid-19 pandemic, and the tool was able to simulate these unprecedented circumstances to provide insightful, accurate and useful solutions.

An enterprise capable decision support tool

Our work provided an innovative and successful solution to SEW’s problem. We’re planning to develop it further to allow for cross-zonal optimisation and incorporate live data sources, improving the capability and usability. We’re also planning to transition it into an enterprise capable decision support tool, allowing wider range of users in different contexts, including in day-to-day operations in the control room.

Let’s collaborate

Send us an email, to discuss a new project.

We’re a team of innovators who are excited about unique ideas and help companies to create amazing solutions.

Water