Understanding and predicting customer consumption to help reduce DMA leakage

With our Managing Director, Stuart Mawditt, and Technical Delivery Manager, Edwin Williams, heading to Rio de Janeiro next week to present at IWA Water Loss 2026, this month’s company blog shares the abstract of their paper on the Paradigm Club Project – highlighting how improved understanding of customer consumption can support more effective DMA leakage reduction.

SME Water has been working with a number of UK water companies to improve their understanding of customer demand and leakage. This has culminated in the creation of a standard approach for demand analysis, called Paradigm. By analysing entire DMA flow profiles over an extended period, rather than just using minimum night flows and total daily flows, Paradigm delivers significant insight into how water is really being used within distribution networks. This is a valuable tool for water companies looking to separate demand from leakage as they continue to pursue increasingly challenging leakage reduction targets.

Customer behaviour is both complex and constantly changing and the project has recognised that this solution cannot be developed in isolation. To meet this challenge SME Water have partnered with a group of UK water companies (full current list in the references section) to implement and develop the analysis. In its fourth year Paradigm is forming the cornerstone of these water companies’ leakage reduction and demand analysis strategies.

Paradigm analyses all DMAs within each of the water companies’ distribution areas, allowing companies to target customer demand and network anomalies, which they otherwise might have treated as leakage within their analysis and regulatory reporting. In parallel to this, Paradigm provides weekly analysis of DMAs to allow leakage targeting to be undertaken proactively and with consideration of predicted customer behaviours.

Over the past three years the club project has brought together previous analysis concepts to create an analysis approach consisting of three main phases

1. Demand forecasting at DMA level using a combination of water company data and open-source data.
2. Historic validation of DMA behaviour against demand forecast to build up a confidence rating.
3. Weekly validation of DMA behaviour against demand forecast to identify emerging issues and ensure the correct response of either a Data, Leakage or Integrity investigation.

The analysis delivers significant additional insights compared to traditional MNF and operability tests, allowing the simple identification of breached DMAs, missing demand and incorrect data. An example of this is shown in Figure 1.1 where a comparison of the expected behaviour (orange line) gives a clear indication that the actual net flow for the area (blue line) is not representative of how the area should be performing. In this case the network operator was quick to undertake boundary valve checks and identified a breach, returning the DMA to its expected behaviour.

The presentation will focus on the lessons learnt from the UK industry and how data has been used to create the expected profile for a DMA. This is an approach that can be taken away by practitioners and used to validate their own DMAs in a simple manner.

There is a barrier to wider implementation of the current Paradigm model en mass outside of the UK, as UK customer profiles are unlikely to be the same as demand profile in other countries, due to a range of factors affecting demand.

A positive outcome from the presentation would be to generate enough interest from international water companies (outside of the UK) to explore a similar (but lighter touch) club project to take the lessons learnt and develop a set of modelled behaviour profiles for different international environments. This could form a simple but very powerful tool for practitioners in middle and lower income countries to be able to quickly validate monitored flow data in their distribution system.

The Paradigm concepts were previously presented by Stuart Mawditt in an IWA Webinar entitled “Understanding and predicting customer consumption to help reduce DMA leakage” in October 2024. https://www.youtube.com/watch?v=hnZmfcF9c5E.