Vulnerability of Urban Water Systems to Climate Change and Mitigating the Potential for Cascading Failures

Abstract:

As average temperatures and occurrences of extreme heat events increase in the Southwest, the water infrastructure that was designed to operate under historical temperature ranges may become increasingly vulnerable to component failures. Critical components in water systems may be more likely to fail under future conditions, and it is possible that the combination of these failures could cascade to system-wide failures compromising water quantity, pressure, and quality delivered to customers. To model the propagation of probability of infrastructural component failures to decreases in system-wide performance under future climate conditions, an event tree failure analysis framework is used Arizona water systems are modeled as the case study.

For each major component in the infrastructural systems involved in water extraction, treatment, and transmission, the probability of failure is estimated for a range of possible temperatures. Probability of component failure is calculated by using lifespan derating information to estimate the mean of time to failure probability distributions for different temperatures. The probability distributions are then integrated over selected time intervals to calculate the probability of failure under present and future conditions.

Under worst case scenario conditions in Arizona, the probability of failure of two critical components in all water infrastructure systems, motors and electronics, will increase 4% and 10% under an average summertime ambient temperature increase from 104^oF to 113^oF, where the probabilities of failure over their lifetime at 113^oF are 67% and 89% respectively. At the temperature of 122^oF, the probability of failure of motors increases another 12% to become 79% likely to fail, and the probability of failure of electronics increases another 7% to become 96% likely to fail.

The probabilities of system-wide failures are then calculated by inputting events and their probabilities into event trees that show potential pathways to system failure or recovery. We find that though heat has the potential to significantly decrease the reliability of the urban water systems in Arizona, proactive governance and strategic improvements to maintenance practices can effectively offset these decreases, thereby minimizing risk.