It is common knowledge among those involved in the management of water distribution systems that the onset of winter brings about an increase in maintenance activities. This phenomenon is obviously more acute in systems situated in northern latitudes. As a matter of fact managers will pay close attention to dropping ambient temperatures, they will mark it on their calendar and make the appropriate preparations for the expected influx of broken water mains. With the advent of supervisory control and data acquistion systems and improvements in instrumentation, the distribution manager has in his disposal a plethora of additional data. As more information becomes available the ability to trend and correlate various sources of data has helped distribution managers to better understanding of various phenomena and their empirical observations. In our attempt to contribute some answers to the titillating question of how can we predict water main breaks we have looked into the data collected from the District of Columbia (District) water distribution system and have arrived at some plausible explanations as to the origin and causes of water main failures specific to pipes made of cast iron material. In studying the data we found earlyon that variation in water temperature rather than ambient temperature was closely related with the frequency of occurrence of broken water mains. This propelled us into examining other elements such as age, material and size of pipe. Includes tables, figures.
