Recently, Mackenzie Mailhot, our Intelligence Analyst at Citylitics, wrote an article for BCWWA, looking at AI in BC’s drinking water industry.

As  climate change and extreme weather events become more and more prevalent every year, the conservation of our drinking water resources and sustainability within the water industry is more important than ever before. Preserving our clean drinking water resources can drastically reduce energy and chemical usage in water treatment plants, protect districts from water shortages and ensure that more Canadians (especially those living in disadvantaged communities) have access to the clean water they need. But as important as it is to emphasize the conservation of our drinking water, every year a substantial portion of our drinking water is lost to the water distribution system. In fact, in 2019 alone, British Columbia lost approximately 92,600,000 cubic meters of water from distribution systems (Statistics Canada, 2021), approximately 11.7% of all of the province’s drinking water.

This loss can be attributed mainly to the province’s aging water infrastructure. As a result, water pipes and fittings become outdated and prone to major leaks. Since this infrastructure is all underground, it is difficult to determine the location of a leak until water has reached the surface and has already become a serious issue. There are methods to maintain and monitor hundreds of kilometre long water systems to find leaks, but most are both inefficient and expensive. From having water operators manually watch CCTV footage of pipes to check for leaks to just waiting to see a leak appear on the surface, wasting valuable water resources in the process, neither solution is particularly effective if the goal is conserving water and energy.

Luckily, there are new methods which show promise in the effort to conserve water and reduce British Columbia’s 11.7% water loss value. BC  municipalities have been working to solve their water loss issues by utilizing an innovative solution – artificial intelligence. This article will use market intelligence and data from BC  municipalities to explore new and developing artificial intelligence trends in the province’s water industry. The data was sourced from Citylitics’ Market Intelligence Platform which uses machine learning to aggregate the mountains of public documents and data generated by cities, utilities and public agencies, and transform it into critical predictive intelligence, providing an early-stage view of infrastructure projects within the drinking water, wastewater, stormwater, transportation, telecommunications, power and energy, and many more industries. By aggregating and classifying municipal documents using Citylitics’ predictive intelligence tools, we can identify plans for upcoming innovative water projects and get a sneak peek into the future of water management. This article leverages data from over 4,000 recent projects in BC’s drinking water industry from 158 unique BC  municipalities to uncover trends and themes around artificial intelligence in water.

BC’s Aging Water Infrastructure

Pipe breaks and water loss will occasionally occur in any distribution system and are impossible to prevent completely. The risk of leaks within the distribution system is usually attributed to the pipe material, system water pressure and the age of the infrastructure along with numerous other factors. Even seemingly minimal climatic and soil quality conditions can have a greater than expected influence on the degradation of water pipes.

According to the 2016 Canadian Infrastructure Report Card (University of British Columbia, 2016), 29% of Canada’s drinking water infrastructure is approaching or currently is at the end of its useful life, and the data tells a similar story. Water loss is a frequent concern for municipalities across the province. Twenty-eight current drinking water projects from British Columbia mentioned issues with high water loss or initiatives to help reduce water loss within the distribution system. In addition to water loss projects, 99 BC entities also referenced upcoming water conservation projects and 114 sources explicitly mentioned aging infrastructure within their water systems. These early indications of problems within the municipality’s water system are often an indicator that the municipality will soon be searching for a new solution and that their current leak detection methods are not working properly.

Rethinking Leak Detection

Common distribution system maintenance and monitoring methods involve installing  CCTV cameras to monitor their water pipes and assigning a technician to manually watch hundreds of hours of footage for potential areas of concern. This process is generally inefficient and time-consuming, and it’s no wonder most water systems are seeking out new solutions.

One trend that repeatedly appeared  in the market intelligence was that municipalities are looking for ways to utilize new technologies to improve water system efficiency and sustainability. Of 130 drinking water projects that mentioned “leak detection” projects/programs:

• Six projects cited ‘real-time’ or ‘early-stage’ leak detection
• Three initiatives included ‘leak detection software’
• Three insights mentioned a ‘data-based’ or ‘smart’ monitoring system
• Three specified the use or incorporation of ‘artificial intelligence’ within the project

One planned leak detection project in BC mentioned a specific type of artificial intelligence – Artificial Neural Networks or ANN. ANN can be used in leak detection and water system monitoring to scan CCTV footage while adapting to detect and diagnose pipeline leaks by recognizing inconsistencies in water flow patterns. This software greatly increases the efficiency of the manual monitoring and leak detection process, reducing the time required to scan footage and increasing the detection accuracy.

What is clear from these recent water projects listed in municipal capital improvement plans and meeting minutes is that BC  municipalities are beginning to actively seek out ways to incorporate machine learning into their distribution system in the coming years.

AI Inside the Treatment Plant

Artificial Intelligence has also begun to appear in water treatment plant operations as well as in water distribution systems to increase efficiency, safety and save energy. The intelligence from BC ’s municipal documents indicates that smart technologies are being incorporated to manage and prioritize pump runtimes to increase plant energy efficiency and are being used in conjunction with asset management software to monitor plant equipment and operations. Another larger theme seen in upcoming water treatment plant initiatives is the integration of digitization and machine learning into asset management and monitoring systems. The market intelligence from BC municipalities showed a total of 100 upcoming Supervisory Controls and Data Acquisition (SCADA) projects, 67 of which were either system upgrades/improvements or feasibility study initiatives. An AI-based technology called ‘digital twins,’ which can be combined with SCADA systems, was mentioned by two BC  water utilities. A digital twin is a machine learning technology which can use real-time data to simulate a virtual model of the water treatment operation systems and aid in proactive and predictive maintenance within the plant.

BC Wants Smart Solutions

Whether it’s within the distribution system or the water treatment plant itself, BC . municipalities have been turning to artificial intelligence to optimize their water system operations. Their overarching priorities and goals, as shown in Figure 1, are clearly stated in the types of projects they pursue: data-driven, sustainable and efficient. British Columbia’s provincial government has also supported innovations in environmentally-friendly water infrastructure by providing $10,000 grants to municipalities for feasibility studies related to sustainable drinking water projects (Government of B.C., 2023), incentivizing municipalities to dedicate resources to sustainable solutions.

Artificial intelligence still is not a perfect solution for the industry.  It may not be able to replace human decision-making for water utilities entirely, but it has been shown to be a powerful tool in the goal of conserving potable water and reducing energy consumption in the water industry. From becoming more proactive about pipe maintenance to avoid large-scale water leaks to optimizing treatment plant operations, the province of British Columbia is turning to digitization and artificial intelligence to meet their sustainability and efficiency goals.

Author: Mackenzie Mailhot is an Intelligence Analyst at Citylitics, Inc. Citylitics delivers predictive intelligence on local utility and public infrastructure markets in North America. The company’s data engine transforms over a billion documents buried in 31,000+ city & utility data sources into high-value intelligence for targeted business development.

References:

Government of B.C. (2023). Infrastructure planning grant program. The Official Website of the Government of British Columbia. https://www2.gov.bc.ca/gov/content/governments/local-governments/grants-transfers/grants/infrastructure-planning-grant-program

Statistics Canada (2021). Potable water use by sector and average daily use. Statistics Canada. https://www150.statcan.gc.ca/t1/tbl1/en/tv.action?pid=3810027101&cubeTimeFrame.startYear=2017&cubeTimeFrame.endYear=2019&referencePeriods=20170101%2C20190101

University of British Columbia (2016). UBC research helps small municipalities manage aging water systems. UBC Okanagan News. https://news.ok.ubc.ca/2016/02/16/ubc-research-helps-small-municipalities-manage-aging-water-systems/