By Luke Brazier, Product Owner, Arcadis Gen
Delivering clean water to the people of the United States is a significant accomplishment, especially with a total pipe network of almost 2.2 million miles of underground pipes with a total asset value of almost 2 trillion dollars.
However, in maintaining this service, water organizations are faced with challenges from their local infrastructure as assets deteriorate over time, and they must make critical decisions on how to invest their capital before service interruptions occur.
Historically, asset managers and engineers have worked with consultants and relied on age-based analysis to develop optimal investment plans. As the water industry evolved, more modern technologies have been used to help manage assets and prevent failures.
Now, cameras and sensor devices are playing a greater role in helping water utilities gain more control of their assets. In addition, the rise of analytics has provided organizations with useful insights and smarter ways of gathering information.
Let’s explore how data analytics and machine learning support organizations with their investment decisions, as well as demonstrate how factors such as likelihood of failure (LoF), consequence of failure (CoF), and risk are essential parts of the planning process.
See the Impact on a Mid-Size Water Utility
To better illustrate how a potential mid-size water utility can incorporate data analytics to improve, streamline and optimize its operations, we’ve created an example using a fictitious water company, Future Water, Inc.
Future Water, Inc. (FWI) is a water company that owns more than 60,000 pipes spanning more than 200 miles, and experiences up to 100 pipe failures in an average year. These failures are usually caused by corrosion, water velocity, clogging, movement, and extreme temperatures.
FWI operates in four-year planning cycles. It’s currently at the beginning of a new planning cycle, and planners are expected to include several pipe replacement initiatives which include assets labelled as high Risk, or of a particular age.
Experience the Benefits of Likelihood of Failure, Consequence of Failure and Risk in Pipe Replacement Programs
Currently, FWI’s pipe replacement criteria is age based, so it doesn’t include the previously mentioned causes of pipe failure. However, data analytics and machine learning can provide greater insight about each asset and allow more informed decisions to be made.
Before discussing the benefits or quantitative risk measurement and how it can help water companies and utilities to help demonstrate return on investment (ROI), we need to define Likelihood of Failure, Consequence of Failure, and Risk.
- Likelihood of Failure (LoF) analysis predicts the probability that an asset will fail based on associated properties like age, length, materials, diameter, location, soil type, corrosivity, depth, pressure, and velocity – all correlated to failure history.
- Consequence of Failure (CoF) is a quantifiable value based on potential financial, environmental, and social impacts – ranging from public safety to environmental contamination to costs related to collateral damage caused by the failure.
- Risk is the combination of LoF and CoF – just multiply the LoF of any asset by its CoF – adjusted for any risk mitigation measures currently in place. Mitigation measures reduce the overall impact of failures by modifying LoF and CoF.
Now that we’ve defined all three, let’s look at how they’re used in the real world. Imagine you’re flipping a coin. The probability of it landing on heads is 50 percent, so on two flips you’d expect it to fall on heads at least once. Engineers apply that same logic to pipe LoF.
So, if a group of five assets has a 20 percent LoF, you would expect at least one of those assets to fail in a year. This approach is useful as it helps with estimating costs and planning for investment budgets.
However, uncertainty also plays a role. While we expect – and have budgeted for – one asset to fail each year, in reality two or more could fail. As the group of assets grows larger it’s less likely that the number of failures will exceed the average, and the impact is minimal.
Considering CoF, one way to determine the value is to estimate the cost of a failure, or monetized consequence. Asset managers assign a cost to a type of failure, usually based on previous failure data, and costs incurred for those failures.
Use LoF, CoF and Risk to Create an Investment Plan
FWI can now calculate the potential cost of failure for a group of assets, so engineers can manage assets more efficiently – prioritizing investment on assets with the highest LoF and CoF, and opting not to replace the assets with lower risk.
Additionally, FWI can run multiple LoF scenarios with different associated properties for more accurate risk assessments, leading to more informed reporting and recommendations when presenting to stakeholders, or when negotiating budgets with municipalities.
Achieve More Accurate Predictions with Technology
Software solutions that combine data analytics and machine learning, as well as LoF, CoF and risk; can help water companies and utilities not only become more independent, but also save time, costs, and work more efficiently.
Now organizations can analyze their entire installed asset base and associated failure history, and then forecast likely future failure data to be used in future planning cycles for asset management and investment.
When building your own asset management and investment plan, it’s good practice to include historical asset and failure data. This provides validity and will reassure both stakeholders and decision makers that the information presented is based on accuracy.
This article is sponsored content and was contributed by Luke Brazier, Product Owner, Arcadis Gen.
