Common menu bar links
CONSTRUCTION INNOVATION, Sept. 2009
[ Table of contents | Subscribe ]
Typical water distribution systems comprise hundreds and even thousands of buried pipes. Effective planning for pipe renewal requires the identification of individual pipes that are most likely to experience elevated breakage frequencies. This is challenging because the deterioration mechanisms of water mains are complex and involve numerous factors, both static (e.g., pipe material, size, age, soil type) and dynamic (e.g., climate variations, cathodic protection).
Historical pattern and forecast of breaks for a group of pipes.
Examination and analysis of historical water main breakage patterns using statistical methods provide an effective and inexpensive way to estimate deterioration for individual pipes or groups of pipes. These historical patterns are used to forecast anticipated future breakage rates which are subsequently used to plan and prioritise water main renewals.
NRC-IRC has developed a model that estimates future failure rates by considering both static and dynamic factors that affect deterioration.
The model, implemented in the form of a computer application called Individual Water Main Renewal Planner (I-WARP), will be available through the Water Research Foundation (formerly known as the American Water Works Association Research Foundation) www.waterresearchfoundation.org
I-WARP differs from other statistical models for the analysis of breakage patterns of individual water mains in that it considers dynamic factors in addition to static factors. Qualitative dynamic factors, such as operational changes (leak detection campaigns, pressure zone changes, etc.) that might influence breakage rates can also be considered, provided appropriate information is available. Currently available statistical models capable of considering dynamic factors (e.g., D-WARP, www.nrc-cnrc.gc.ca/eng/projects/irc/renewal-planner.html) are designed to conduct high-level (group-level) water main renewal plans for groups (or cohorts) of pipes.
I-WARP is based on a probabilistic concept called the Non-Homogeneous Poisson Process (NHPP). A Poisson Process describes a stochastic (random) process in which events (in this case, water main breaks) occur randomly, albeit at a given mean rate. This same process is used to model the arrival of cars at a traffic light.
I-WARP application requires inventory and breakage data about individual pipes. Pipes are divided into homogeneous groups - pipes of the same material, same diameter, same vintage - or any other grouping criterion for which data is available. The model is first trained (or calibrated) to discern historical breakage patterns of each pipe group. This calibration exercise provides group-specific parameters, which are then used to forecast future breaks.
Mandatory inventory data for each individual pipe include pipe diameter and year of installation. Optional inventory data include such parameters as pipe material, length, geographic location, cathodic protection and soil type. Mandatory breakage data (for a minimum of five years) includes break date and its association with a specific pipe in the inventory data. Since I-WARP allows for the consideration of dynamic factors, the inclusion of climate data (temperature and precipitation) is beneficial.
I-WARP will enable one to prioritise the renewal of individual water mains within a homogenous group. It will help users plan water main renewal activities at pipe level, thus complementing the D-WARP software developed earlier at NRC-IRC.
For more information, contact Dr. Yehuda Kleiner at (613) 993-3805 or email yehuda.kleiner@nrc-cnrc.gc.ca.
