A field data-based method to determine the pressure-burst relationships in urban water distribution networks

One of the challenges facing water and wastewater companies around the world is water loss from water distribution networks following burst pipes and leakage, which imposes high economic, social and environmental costs on these companies. So, every year a large part of the budget of water and wastewater companies is allocated for the repair and rehabilitation of the pipe network. Therefore, knowing the frequency of burst pipes will help in estimating network leakages and selecting appropriate management strategies for dealing with these. Various factors affect the failure of water distribution pipes, one of the most important being water pressure. Therefore, the development of models to predict failure based on effective factors precisely is necessary to achieve optimal leakage management in water distribution networks. In the present study, using a developed model and analysis of pressure and burst field data in the urban water distribution network, the relationship between pressure and burst rate has been determined for a district of Tehran. The study area has 516 km of mains pipelines that include many types of material such as polyethylene, ductile iron, steel, PVC and asbestos cement. Both polyethylene and ductile iron pipes were selected for the investigation because they comprise 93 percent of the network length. After collecting and revising the statistics and information about the bursts and pressures recorded during the years 1386 to 1395, we have calculated the average zone point and the pressure index at this point and assigned it to the whole area. Finally, the pressure-burst relationship was presented individually for each material in the pipes. The prediction model of the failure consists of two parts, namely independent and dependent parts, through which the pressure parameter is linked through a power component to the failure rates. In this study, the maximum daily pressure at the average zone point was used as a pressure index in the pressure-burst relationship. The pressure-burst relationship for polyethylene and ductile iron based on maximum the daily pressure index is presented separately. In the relationships obtained for comparison, both the average of maximum daily pressure and the maximum of maximum daily pressure values were used. The results of this study showed that, in the dependent pressure part, the average of the maximum daily pressure index presents a more accurate result compared with the maximum value of the maximum daily pressure index and has a higher correlation coefficient. The reason for the inappropriateness of the maximum value of the maximum daily pressure can be temporary and non-permanent overload in one or more days of the year. As it may not really have caused a failure but has been involved in the calculation, this index does not have an accurate prediction of burst. Also, the relations are obtained for two conditions with a power pressure of 3 and an unknown situation, which indicates that, in the case of unknown power, higher correlation coefficients are obtained. Thus, for polyethylene the power is equal to 3 and the correlation coefficient = 0.97, while for ductile iron, the power was equal to 2.7 and the correlation coefficient = 0.99. According to the relationships obtained, it can be concluded that the pressure-burst model could predict the number of failure of main pipes in the water distribution networks well. The results also showed that pressure variations more often affect burst frequency in the polyethylene than the ductile iron pipes. The exponent of pressure in the failure prediction model also depends on the pipe material and is larger for polyethylene in comparison with ductile iron material, and the average for the maximum daily pressure index was a more accurate indicator in the failure prediction model. According to the results of this paper, we can improve pressure management and rehabilitation strategies for the reduction in burst frequency. By applying accurate pressure management and awareness of material susceptibility to burst, it is possible to reduce failure rate and, consequently, water loss.