Business Process Improvement at Toronto Water

Toronto skyline

By Tim Kruse

The City of Toronto Water Department, Ontario, Canada, (Toronto Water) serves nearly three million customers who rely on more than 1,800 employees to keep operations running at their peak. Toronto Water is responsible for drinking water supply and four water treatment plants, wastewater collection and an additional four wastewater treatment plants, as well as stormwater management.

While managing these assets and the related processes, a group of self-described ‘hardcore’ data users were experiencing challenges with effectively utilizing Toronto Water’s supervisory control and data acquisition (SCADA) and laboratory information management system (LIMS) data on a regular basis to execute critical tasks.

Toronto Water is not alone in its struggle with data. A 2021 report from Dodge Data & Analytics and Bentley Systems, examines the how U.S. water utilities are still in the early stages of digital transformation and provides the results of a survey of U.S. water utilities. In the survey, 87 percent of utilities are collecting digital data, however, 90 percent of those respondents say that their data is isolated in disconnected IT systems, spreadsheets, or paper records. Sixty eight percent of utilities report lack of visibility across stakeholders interferes with effective capital planning and operations.

Methodology

In June 2017, a cross-functional ‘Operational Intelligence Group1 was formed within Toronto Water, comprised of plant engineers, central process engineers, and members of the process controls systems group. The group’s goal was to select the best solution for Toronto Water from a myriad of options.

The market has many solutions and approaches for managing large volumes of data. The main options are:

  1. Data Warehouses – A data warehouse is a type of data management system that is designed to enable and support business intelligence (BI) activities, especially analytics. Data warehouses are solely intended to perform queries and analysis and often contain large amounts of historical data.2
  2. Data Lakes – A data lake is a centralized repository designed to store, process, and secure large amounts of structured, semi-structured, and unstructured data. It can store data in its native format and process any variety of it, ignoring size limits.3
  3. Historians – Typically a historian is a service that collects data from various devices in a SCADA
    network and logs it to a database. Newer historians are sometimes used to collect data from non-SCADA applications.
  4. Middleware or Data Brokers – These types of tools provide direct access to application data and facilitate the sharing of data between applications. Data brokers are often used directly with sensors or other devices that produce data. Middleware and brokers do not store data but simply allow users access to data from applications or devices.

Many people are not familiar with the main differences between data warehouses and data lakes. Warehouses are typically used by business professionals to access processed data that has a specific use (e.g., regulatory or operational reporting). Lakes are filled with mostly raw data (or nearly raw data) and are used by data scientists to answers yet to be asked questions. The data in warehouses is more complicated to setup and change while lakes are easier to create and change.

The group performed market research and became aware that data management software solutions were available that could improve efficiencies and support their complex workflows. The group interviewed vendors and other large utilities from around the world finding that complex data warehouses systems were commonplace, and typically require a high level of effort to sustain. The group also drew from their own understanding of unsuccessful large deployments of enterprise systems to create requirements for a new data management solution.

The city also had a large legacy data management solution that had its own data repository. Users of this system frequently shared concerns over data accuracy and discrepancies with data retrieved directly from source systems. The requirements included a software application that would transform their existing data systems into a single, powerful view of Toronto Water’s operations. The software had to support hundreds of active users. Another consideration was a large IT organization with strict cybersecurity measures in place. Above all, an application that fostered power users was desired, but one that did not require a high level of effort to maintain.

A unique aspect of the process included establishing internal governance, which was endorsed by all managers and department heads involved with the targeted data systems. Unanimous endorsement ensured that sufficient resources would be made available to satisfactorily fulfill the roles and responsibilities of the implementation project – including Lead Business Owner, Supporting Technical Leads, Advisory Group, and Application Owners. After thorough evaluations, the group concluded that nothing could provide the level of functionality of the top solution and in February 2018 the group made a unanimous decision to acquire eRIS, a SUEZ owned application.

Results

An initial rollout project began in Spring 2019. By selecting a solution that was widely used commercially “off-the-shelf,” and met all required functional needs, Toronto Water started using (and benefitting) immediately. An extensive library of system integrations, alarm capabilities, integrated electronic logbook and a modern, intuitive mobile application streamlined adoption.

The initial phase of the project was completed by the end of December 2021.

Toronto Water initiated the project in 2019 at one water and one wastewater plant, known as Phase One. The Toronto Water and vendor team developed 20 key reports, performed a rollout of plant operational logbooks, and captured manually collected data into a dedicated data repository stored in the city’s own Oracle RDBMS. The teams also created a data validation process to ensure that a small and critical data subset was reviewed and corrected prior to including in certain reports.

The application’s Data Validation process provide users with the ability to rapidly review, flag, approve, and even modify, if necessary, data from source systems like SCADA Historian or manually entered by operators. Both manual and automated business rules can be used during the validation process. Automation can review millions of data points and the remaining flagged values are elevated for manual review by an operator or supervisor. Examples of automated business rules include flagging data that falls below or exceeds desired values such as flow stop-start, values that violate a certain rate of change, frozen values, and data that violates regulatory limits. Data points that need further study are staged for manual review and approval.

Digital water does not need to be as complex as artificial intelligence to be highly valuable to utilities.

One of the first reports developed was the Wastewater Treatment Weekly Director’s Summary. This report was chosen because it uses data from all available sources, namely, all eight Toronto Water SCADA Historians, the city’s laboratory information management system (LIMS), manually collected data by system operator, and finalized data.

Support and training were keys to success. Customized training was developed in an e-learning platform, which was interactive and online, and was customized to Toronto Water’s exact data sources and screen layouts. New users could easily complete the training as allowed by their workloads and shift schedules. Toronto Water leveraged an internal web tool to serve up quick start guides and a wide range of beginner to power user training resources.

Another key to success was the establishment of Platform Governance. The governance plan defines the roles and responsibilities for different parties. The five (5) roles include Lead Business Owner, Supporting Business Owner & Technical Support Lead, Supporting Business Owner, System Advisory Group, and Application Owner.

Subsequently to the initial project, Toronto Water engaged the vendor to assist with expanding the application to include all remaining Toronto Water treatment plants. Eight additional plant and facility logbooks were deployed along with additional plant operational reports. Other projects were initiated to improve data integration and sharing including one project to automatically transfer daily operational data to a regional partner and to a co-developed data repository and website. This solution eliminated the manual daily process with a fully automated, reliable data transfer tool.

Conclusions

Toronto Water had numerous options when selecting their next data management platform. Using their experience and seeking recommendations and advice from other similar utilities, the city selected a platform that did not require the creation of a data warehouse or data lake and yet provided a single, powerful view of Toronto Water’s operations.

By selecting this application, Toronto Water not only improved operation and business decision making, but they also experienced a clear return on investment (ROI). The ability to extract all required process data with enhanced analytics, automated reporting functions, and powerful calculation tools has resulted in faster, easier, and cleaner data and reports. This turnaround has been a key driver in staff-hours saved, which the project owner quantified has quantified is nearly $600,000 CAD. The Phase One Initial Deployment was awarded the Toronto Water Excellence Award for both the value and execution of the project.

With strong support both internally and from Suez, Toronto Water continues to develop content and extract increasing value from the solution over time.

Digital water does not need to be as complex as artificial intelligence to be highly valuable to utilities. Prior to deriving value from complex analytical solutions, utilities would benefit from enhancing their enterprise data management to better understand and trust their data quality and control expectations for different datasets. Key activities such as managing data governance is critical in case of staff turnover and ongoing validation of key operational advisory systems.


Tim Kruse is a vice president and general manager of eRIS, a SUEZ Smart & Environmental Solutions company. eRIS provides intelligent tools and expertise to support the digital transformation including integrating data collection solutions (smart meters, sensors and probes) to provide digital models and real-time applications.

Leave a Reply

Your email address will not be published. Required fields are marked *

*