Simulation of Effluent Biological Oxygen Demand and Ammonia for Increasingly Decentralized Networks of Wastewater Treatment Facilities

Abstract

There is growing interest in decentralization of wastewater collection and treatment systems, and with increasing numbers of small facilities there is greater need to predict effects of networks of treatment facilities on a watershed. A network analysis of effluent biological oxygen demand (BOD) discharge and variability for a gradient of decentralization was conducted by simulating many individual facility discharges using Generalized Linear Models. Simulations were performed for synthetic networks of varying numbers of facilities and of varying capacity. For a total flow of 400,000 m3/day, simulations indicated that a network of 1,000 small facilities would produce 80% higher median effluent BOD than a single large facility; however, a single facility had much greater variability, so the 95th percentile BOD discharge was 58% higher than for the network of small facilities. This methodology was also applied to a case study of wastewater facilities discharging in the St. Vrain watershed in Colorado. As with the synthetic network simulations, for ammonia discharges in the St. Vrain watershed, a centralized network produced lower median effluent ammonia than either the existing system or a decentralized network. The 95th percentile effluent discharge for both the decentralized network and the centralized network was 14% lower than for the existing network with a wide variety of facility sizes. Decentralized networks are able to reduce aggregate variability through diffusion of risk across many facilities, while centralized networks achieve improved median treatment through better individual facility performance.