CASE STUDIES AND VALUE OF RESEARCH
Developing a Main Replacement Prioritization Tool
There is a tremendous need to invest and upgrade America’s drinking water infrastructure. According to a 2007 USEPA
survey, the 20-year capital investment needed for public water systems is estimated at about $335 billion. There is no
question that aging infrastructure is in need of renewal, but a continuing issue that needs to be addressed is the order
in which pipes are to be renewed. If a pipe is replaced too soon, it may not have reached its full life cycle. Replace
a pipe too late and incur some unavoidable costs for additional repairs and damages caused by main failures and
compromise the level of service provided to the customer. One clear goal of most replacement programs is to reduce
the cost and number of breaks as a program proceeds.
The St. Louis County Water Co., now part of Missouri American Water (MAW), was one of the early experimenters
in the field of main prioritization as one of the subjects of a study through the Water Research Foundation,
Prioritizing Water Main Replacement and Rehabilitation (2002, order #90898/project #459). This project
developed a decision tool, which utilized main break data and cohort analysis as the key criteria that helped MAW
make pipe replacement decisions.
Improved main break data collection and the advent of a geographic information system (GIS) led to a search for a
new tool in 2008. Working with GL Denton Noble (GL), MAW prepared main and main break data for incorporation
with a GIS network designed for the project. The pipe network was sourced from a computer-aided design (CAD)
to GIS translation that was performed by MAW. GL then scrubbed the database to remove any “non-pipe” lines. A
GIS process was then created by GL to assign pipe installation year and material based on annotations from the CAD
database. The GIS was populated with geo-coded main break data segregated by corrosion, joint failures, and breaks.
Following a period of data cleansing to verify and cross check data, GL put their program called Main Replacement
Prioritization (MRP) to work.
The MRP model is a three-step function that reflects the earlier decision tool, but with some notable enhancements
1. The first step utilizes main break history to construct a condition assessment model of all ferrous pipes.
2. The second step looks at a series of possible performance indicators to be used for renewal criteria for pipe
segments and selects weightings to be given for these criteria.
3. The third step constructs a multiyear project program taking into account the annual budget available and the
proximity of other mains that are subject for replacement.
Step 1 performs both a cohort and proximity analysis for families of ferrous water pipe. In other words, pipes of like
material and diameter within close proximity of one another are assigned a break curve based upon past history. In
this way, for example, a block of 4-inch cast iron main lacking a recorded break history near 4-inch cast iron pipes with
numerous breaks are grouped together. In work supported by the Computer Aided REhabilitation for Water Networks
(CARE-W) work in Europe, this approach, while not perfect, has been found to be more predictive of future breaks
than other statistical studies. This tendency towards failure is then incorporated into a broader set of performance
criteria in step 2.
Step 2 allows the utility to select other criteria that it feels are elements for replacement criteria. MAW added to the
condition assessment six other weighting factors including:
1. The presence of critical customers (usually hospitals and schools).
2. If the pipe was galvanized steel — the utility has a goal to phase all of this material out of use, as it is often
associated with water quality complaints and poor flow.