You are here:
Precipitation Nonstationarity Effects on Water Infrastructure and Risk Management
Citation:
YANG, Y. J. Precipitation Nonstationarity Effects on Water Infrastructure and Risk Management . Presented at Workshop on Nonstationarity, Hydrologic Frequency Analysis and Water Management, Boulder, CO, January 13 - 15, 2010.
Impact/Purpose:
To inform the public.
Description:
The non-stationary precipitation regime, as increasingly recognized, affects the engineering basis and service functions of drinking water, wastewater, and stormwater infrastructures in urban centers. Small, yet significant rates of temporal precipitation change and diverse spatial distribution of its impacts over the contiguous U.S. affect the technical and management approaches in infrastructure adaptation. Several engineering and management attributes such as the impending actions on aging water infrastructure assets, traditional engineering conservatism, and the adaptation cost, further add complexity to the discussion and development of a practical risk management framework. In this presentation, major precipitation nonstationarity effects are described and adaptation for risk management based on the existing engineering practice is discussed. Precipitation variability, particularly for extreme precipitation events, is an expression of the evolving climate state and the relatively static local-regional climatic factors. The synoptic climate systems have been quantified in model simulations. In our research, the local-regional climatic factors are identified in a systematic investigation including wavelet and statistical analysis of long-range historical precipitation measurements obtained from 1207 climatic stations across the contiguous U.S. The precipitation datasets have registered the interactions between the synoptic systems and local-regional factors for the past>100 years. The analysis results led to the belief that extreme precipitation events, either the high-intensity or low-intensity, have strong periodicity regulated by local-regional factors in six delineated hydroclimatic provinces: (1) Florida and Southeast, (2) Lower Mississippi – Ohio River valley – New England region, or LONE, (3) Great Plains and Midwest (4) Basin-and-Ranges, (5) West Coast, and (6) Great Lakes. At the first level, the provinces follow major physiographic provinces in the U.S. and each has its own characteristic variations in precipitation and perturbation in high frequencies. The high-frequency, high-intensity precipitation clearly differentiates the Florida and Southeast province from the vast regions to the north and northwest. Arguably these local-regional climatic factors will interact with climate states in the future, a factor that deserves representation in downscaling efforts. That extreme precipitation regulation differs in space has implications for hydraulic and water quality engineering of water infrastructure. The effect of non-stationarity on design storms is described by others in this workshop and discussed in recent publications (e.g., Mailhot et al., 2007). A series of “domino effects” on downstream hydrologic processes and water infrastructure operations can be readily discussed. In the remainder of this presentation, we will focus on the water quality aspect of non-stationary precipitation in reference to drinking water plant operations. A scenario-based engineering analysis was conducted on the Greater Cincinnati Water Works’ Richard Miller Treatment Plant that takes raw water from the nearby Ohio River. High river flow, responding to high-intensity precipitation, is known to be responsible for high turbidity in surface water. For the plant, this increased level of turbidity is statistically related to high contents of total organic carbon (TOC) and other water quality variations. In a detailed process engineering analysis, we found that future scenarios will likely result in a large increase in disinfection by-products (DBPs) in finished drinking water and lead to greater probability of violations of the Safe Drinking Water Act regulations. A process adaptation using granular activated carbon (GAC) filter in the Miller Plant would cost around $5 million/year for a 90% probability of compliance. Other levels of risk management are associated different cost as indicated by the cost cumulative density function (CDF) curve. Our investigation so far has pointed to the need of water infrastructure adaptation to non-stationary precipitation changes. The effects are tangible for all three types of infrastructure. It is important that the adaptation framework adapts to current water engineering practices, for which a possible approach of iterative assessment-adaptation-monitoring procedure is discussed.
