Case study 1: Robust and resilient drainage
Drainage and wastewater management plans (DWMPs), which were introduced by the UK government in 2018, compel water and wastewater companies to set out how they intend to ensure their drainage and wastewater systems are robust and resilient from 2025 until 2050.
Under the plans, companies must quantify and present current and future risks, including possible flood risk to sewerage assets from climate change, and prioritise upgrades based on the level of risk in the short, medium and long term.
The problem is that most water and wastewater companies manage thousands of assets, which all need to be assessed to identify those that are most at risk from flooding and the costs involved in making them more resilient.
Wessex Water asked Mott MacDonald to help assess its 6,000 assets, including water mains and sewers, for flood risk from rivers, sea and surface water, to develop flood resilience solutions and to estimate the costs for implementing them for each asset.
The solution
Collecting and analysing information on each of these assets would normally be a hugely resource-intensive, time-consuming and expensive task, but Mott MacDonald worked collaboratively with Wessex Water and developed and implemented lean processes and digital innovations. These included a bespoke app for site survey and data collection, and a high level of automation for data screening and processing and for option and solution development.
The result
The technology and digital innovations enabled the team to deliver a huge programme of work to an extremely tight timeframe and budget and with limited resources. Wessex Water now has a clear idea of the flood risk for all 6,000 assets and knows which are most at risk, where investment needs to go to make the network more resilient, and the indicative costs.
Wessex Water now has a clear idea of the flood risk for all 6,000 assets
The challenges
The biggest challenge was to develop processing tools and adopt a lean and robust process that would enable information from different sources and in varying formats to be combined and analysed to identify levels of risk rapidly. The analysis tools also had to be flexible enough to be used for a range of flood conditions, levels and return periods and to accommodate any new or additional information that might become relevant in the future.
Mott MacDonald developed both the data collection app and the analysis tools in-house, benefiting from the fact that the water team included expert flood modellers and programmers who not only understood the capabilities of the technology but also what the end result needed to look like. The team focused on developing tools that would clearly show the client which assets and equipment were at risk and to what level, with a clear visual display to enable decision-makers to understand the risks at a glance.
The next steps
The technology developed for this project is designed to be flexible so that information from a variety of sources can be collected and analysed, including photos, Lidar survey data, weather and climate forecasts and maps, as well as data collected onsite using the bespoke app. That flexibility should ensure that other types of information can be incorporated in future if required.
In addition, while the system was developed initially for assessing flood risk to water company assets, it could be used just as easily for other forms of infrastructure assets, including roads, rail lines, bridges and power and energy networks.
Information from a variety of sources can be collected and analysed
Case study 2: Enhanced drinking water supply
The Valencia metropolitan area in south-east Spain is home to 1.7 million people, who are supplied with drinking water from two treatment plants, both about 10km from the city.
Each plant offers different guarantees in terms of supply and water quality at source, as well as different production costs and varying storage capacity at different elevations.
The water from both plants is pumped before it flows into the network, then distributed by gravity to the city and 51 municipalities that constitute the metropolitan area through a 200km main system and 2,500km of secondary distribution networks.
Global Omnium, which manages Valencia’s water supply, faces a range of challenges, including increasing demand due to population growth; resource shortages; ageing infrastructure; and optimisation of daily network operations.
Valencia, Spain, which has a population of 1.7m
The solution
Modelling has been used to plan and optimise water supply operations in Valencia for almost 30 years, culminating in the development of a full digital twin for the drinking water network.
The digital twin is now operational, replicating the behaviour of the real water distribution system in a virtual model and allowing the operator to predict how the network will respond under any circumstances.
*Physical systems such as GIS, sensors, supervisory control and data acquisition (SCADA), smart metering and a computerised maintenance management system (CMMS)
The three elements work together to replicate the real system behaviour in a virtual model, which enables the operator to analyse it as a whole, simulate its behaviour under different conditions and improve efficiency and performance.
The result
Since implementing the digital twin, based on the GoAigua Water Twin Ace platform and undertaking Global Omnium’s digitalisation journey, the utility has reduced non-revenue water by more than 30% (a saving of more than 4.5bn litres); improved customer satisfaction by 60%; saved 15% of the energy used for water treatment processes; and reduced maintenance operating expenditure by 20%.
The challenges
Developing a digital twin of a system is not easy, as the twin must be capable of faithfully reproducing the reality it represents. The main challenges are:
- The model must be updated and capable of accurately reproducing any system operation context, whether in the present (real-time), the past or the near future (next 24 hours), for both real and fictional scenarios.
- The twin has to be able to integrate digital information from different sources, such as GIS and sensor data. The Valencia digital twin uses Idrica’s GoAigua platform to bring this information together and standardise it for use in the model.
- The hydraulic model should be properly built and calibrated. The model for Valencia and its metropolitan area is extremely reliable, with error rates of about 2% for pressure and 4% for flow.
- Results should also be displayed in the most useful way for specific users – for example, as both a synoptic SCADA display for use in control rooms and a geographical display for planning tasks.
The Valencia digital twin in operation
The next steps
The GoAigua platform was designed to be scalable for any size of supply system, and is easy to update in line with any changes that are made to the system. It is already being used for a smaller city in Spain, where the local population of 20,000 increases to 100,000 in the summer. In this location, the digital twin, which covers 248km of pipes and 95 sensors, is helping the operator to find the most efficient way of managing these huge seasonal variations.
Digital twins are on the radar of most water supply companies, and many already have the building blocks in place to make a start in the form of data collection – such as pressure and flow sensors installed in their networks, smart meters in homes and computerised maintenance management systems at their control centres – and hydraulic modelling of the network.
Digital twins are on the radar of most water supply companies
Case study 3: An optimised sewer system
Louisville is a river community and the largest city in the US state of Kentucky, with a population of 755,000 and a sewer network consisting of more than 5,100km of underground sewer pipes, pumping stations, flow storage basins, storage tunnels and flow diversion gates.
For many years, operators had limited information about what was happening on the network, and reacted to changes as they occurred, manually turning valves and operating gates to regulate the sewer flows.
Much of the network carries both surface water run-off and sewage. In heavy rain, the city’s largest treatment plant does not have the capacity to treat this combined flow and the excess is discharged into local watercourses at combined sewer overflows.
In the 1990s, when the Louisville Metropolitan Sewer District (MSD) started addressing this issue, about 25bn litres of untreated sewage mixed with rainwater was entering watercourses each year. Building large underground storage tanks to catch and store this overflow would cost hundreds of millions of dollars.
However, the network has some large tunnels that were built for flood control during the 1930s and that were not being fully utilised during small-to-medium rainfalls while the sewers were overflowing. The MSD wanted to optimise flow management in the sewer system to make the best use of this existing capacity and minimise the number of new underground storage facilities required.
Louisville, Kentucky, with a population of 755,000
The solution
In 2000, the MSD began working with Tetra Tech, a global provider of consulting and engineering services, to evaluate the benefits of applying data-driven, advanced real-time control to optimise flow management of the sewer system. The solution, based on Tetra Tech’s Csoft software, creates a set of operational strategies that are used to manage the storage, diversion and pumping facilities to keep the pipes full during a storm and slowly release the stored water to the wastewater treatment plant when the rainfall subsides.
In essence, the solution extends process control into the sewer system and uses it to equalise and balance flows, ensuring a steady inflow to the plant and maximising the available treatment capacity.
The technology uses data from sewer monitoring sensors, rainfall forecasting, modelling and data analytics to deliver intelligent sewer management in real time.
The key elements of the solution are:
- Real-time monitoring of the sewer conditions that collects and centralises data every five minutes
- A digital twin of the sewer to represent the behaviour of the complex pipe system and provide level and flow predictions
- An advanced algorithm to reduce computing time to find the optimal solution by running through millions of alternatives, considering both current and future conditions
- Automated data validation and actualisation so the system can ‘learn’ from real-world conditions
- Built-in robustness to gain operator confidence and acceptance – such as remote operator override, fail-safe operation modes, and system-wide visualisations
- Automated reporting and KPI graphic display to save time for post-event analysis and regulatory compliance
The result
The Csoft real-time control solution and visualisation platform has been providing predictive and system-wide optimisation in Louisville since 2006. In that time, the MSD estimates it has saved US$200m-US$300m in capital costs, by downsizing the need for new storage capacity, and reduced sewer overflows by 15bn litres per year.
An added benefit is the ability to redirect flow to other parts of the system to allow access for maintenance and troubleshooting. The flow monitors and level sensors inside the sewer pipes provide continuous information, which is displayed on control panels with built-in alarms that indicate potential problems. This artificial intelligence allows troubleshooting to occur in a timelier manner and more safely for staff.
The solution has also enabled the MSD to go from reactive to proactive and predictive operation, and be in full control of its sewer and treatment systems.
Excess capacity in the sewer to store wastewater during a storm is spread across more than 5,000km of underground pipes throughout Louisville
The challenges
The result may sound simple: enabling the MSD to use millions of litres of excess capacity in the sewer to store wastewater during a storm. However, this capacity is spread across more than 5,000km of underground pipes throughout the city, requiring Tetra Tech to create a digital twin to represent this complex pipe system and the way it is operated.
The digital solution not only examines how to route the wastewater flows, but also tracks the impacts of different types of storms – for example, long, slow rainfalls or rapid thunderstorms – and how they flow through the system.
Another challenge is that, while the platform finds the solutions that store the most wastewater, it must anticipate unexpected failures and avoid creating risks such as flooding homes and streets.
The solution has enabled the MSD to go from reactive to proactive and predictive operation
The next steps
The Csoft solution is adaptable and scalable as the MSD sewer system expands to serve its growing customer base and to meet changing environmental requirements. As new facilities are added – for example high-rate treatment – more operational scenarios and decisions can be considered.
This solution has also been implemented in Canada and France and for other water management applications.
Sign up to receive news from ICE Knowledge direct to your inbox.
