Water infrastructure is arguably one of the most important lifelines in our society. The function and scale of water infrastructure has grown with us, from a single well to thousands of miles of pipeline supporting populations millions large.
However, it is easy to forget the important role water infrastructure plays in our lives.
Compared with telecommunication and transportation assets, utility managers of water assets experience the tremendous challenge of taking appropriate action to keep our water supply safe and reliable without being able to see their assets with the naked eye. As a result, managers need access to relevant, accurate, and actionable data to help them make the right decisions.
Data-driven decision making is not a new concept to water utilities. Historically, asset management practices among water utilities was dominated by capacity analysis. Because urbanization and population growth were large drivers during boom times, most water distribution system master plans focused on areas of growth for network expansion, with little attention paid to preventing distribution pipe failure. Utilities did not have an active rehabilitation or replacement schedule because the pipes were relatively new compared to their designed lifespan, making flow capacity data the primary source of information for managing water distribution pipelines.
Capacity-driven planning has gone through its own evolution, from pencil and paper, to punch cards, to computer-based network simulation. The time to identify capacity constraints through iterative network balancing calculation was reduced from months to minutes, while the time and budget to produce a comprehensive distribution system master plan remained roughly the same. The breadth and depth of analysis covered by a master plan significantly increased because of the fast, accurate information powered by computer simulation.
More recently, discussions in asset management have pivoted toward identifying and managing asset failures, as well as improving reliability. Though population growth has slowed in many areas, the ramifications of intense water usage, generated by periods of urbanization, caused pipes to deteriorate and fail earlier and at higher rates than expected. Since water main assets are buried and out-of-sight, the challenge to mitigate these risks has proven difficult for utilities, sometimes with severe consequences.
However, an array of options exists for utilities to reduce the risks associated with buried assets. Proactive utilities started evaluating and building redundancies within their networks to accommodate unexpected outages. Utilities with adequate annual capital budgets began replacing pipes based on year of installation, material of construction, and areas of frequent breaks. Some even created leak detection programs designed to rotate through their service area over various periods of time. These new approaches, and how utilities choose to combine them, contribute to the reduction of asset failure and the resulting consequences.
Just as pipe capacity data transformed, the way in which utilities collect, manage, and evaluate main breaks and pipe condition data is undergoing its own metamorphosis. Utilities are investing more time and attention to their systems and are moving their paper-based records and spreadsheets to digitized methods of record keeping, such as Geographic Information Systems (GIS) and different databases to more efficiently manage their assets. As digital record-keeping becomes the norm, utilities are better able to monitor hotspot areas and apply different asset management practices.
The most proactive utilities realize their digitized records can be used to bring their asset management practices to the next level. When combined with environmental data, Artificial Intelligence (AI), and Machine Learning (ML) tools, such as those offered by Fracta, these records are used to predict failure patterns and correlations at a much more accurate level than the traditional age, material, and break-based models. With more accurate information, utilities are empowered to pinpoint hotspot areas and sections within their network to conduct further condition assessment or take appropriate action. More informed decision-making results in more effective risk mitigation strategies as well as efficient capital expenditure.
As technology continues to develop and utilities become more proactive, objective, data-driven models od decision making for water utilities will surpass the traditional models. The application of AI/ML to assess water networks will continue to empower engineers and utility managers to identify areas that weren’t accessible to them in the past, allowing utilities to achieve more with less. With digitization of data and asset information, water mains are no longer “out-of-sight”, allowing utilities to supply our most valuable resource in a safe and reliable manner.