Failure forecasting allows generating a Likelihood of Failure ranking for each pipe as well as predicting the number of breaks for each pipe and each future year, and generate the aging curve, and a risk-based Effective Useful Life. This is done using utility data analyzed with advanced statistical approaches such as multi-variable regression models and Machine Learning.
At the end of the Failure Forecasting task, the following results will be delivered to the utility:
- Ranking of Likelihood of Failure for each pipe for each year in the future
- Aging curve (break rate by age) for each pipe or groups of pipes and aging speed
- Risk-based Remaining Useful Lives
Pipes are replaced or rehabilitated for a multitude of reasons. First, to control the break rate for which risk-based criteria are defined. Other (non-risk) criteria also come into play. Risk-based criteria require that we compute the Likelihood and Consequence of Failure of each pipe. They can also be used to determine which pipes should be inspected in priority.
At the end of the Projects Prioritization Plan task, the utility will have a list of pipes candidates for inspection, or rehabilitation & replacement ranked in order of priority. The Projects Prioritization Plan will be defendable thanks to quality and utility-specific data, and optimal thanks to advanced analytical approaches.
The pipes candidates for rehabilitation & replacement are subjected to various scenarios (which pipes, how much and when) in order to evaluate the resulting:
At the end of the Long-Term Rehabilitation Plan task, the utility will have selected the scenario that best meets its objective. The preferred Long-Term Capital Plan will be defendable thanks to quality and utility-specific data, and optimal thanks to advanced analytical approaches.
There is unfortunately no such thing as
"data in...results out" or quick deployment. Data must first be reviewed as predictions generated with insufficient or compromised data may lead to erroneous and non defendable solutions. We first diagnose data issues (missing, incoherent, structural) using an extensive library developed over our 25 years of experience. We then remedy those issues using a variety of cutting-edge approaches including Machine Learning and a proprietary algorithm. Improved data quality goes a long way beyond asset management. Many other utility functions and deliverables benefit from better data.
At the end of the Data Cleaning task, the utility data will be on the path to improved quality. The upcoming Rehabilitation & Replacement plan will be more trustworthy.
We generate the most extensive suite of statistical results designed specifically for water systems in the form of :
They can be downloaded and included in a report or consulted within the platform. Strong statistics can be used for many other utility functions beyond asset management, and be integrated in other reports in addition to the Asset Management Plan.
At the end of the Statistics task, the utility will get a better understanding of the system performance including program of rehabilitation, and will be in a position to interpret output results from the advanced analyses, priority ranking and planning scenarios.
Copyright © 2018 infraPLAN - All Rights Reserved