Treatment of sanitary sewer overflow using fixed media bioreactors

Fixed media bioreactor (biofilter) is a proven technology used for wastewater treatment in unsewered rural areas. As an on-site treatment system, it provides high treatment efficiency with relatively low cost and maintenance. Sanitary sewer overflow (SSO) has characteristics of high flow rate, low and varying pollutant concentrations, and low frequency of events. Abatement of pollution from SSO is one of the most challenging areas in the environmental engineering field. No previous research has been published on SSO treatment with fixed media bioreactors. This study expands the application of fixed media bioreactor technology by testing its feasibility in the treatment of SSO wastewater at high hydraulic loading rates.;Five types of fixed media bioreactors---sand, textile peat, textile/sand, and peat/sand---were set up in the laboratory and used to treat SSO once a month. Wastewater from a simulated 25-year peak flow was loaded to the surface of bioreactors at a high hydraulic loading rate of 0.2 m/hr for 6 hr. One influent sample and three effluent samples were collected for each bioreactor at each 6-hr loading. Biochemical oxygen demand (BOD5), total suspended solids (TSS), ammonia nitrogen (NH4-N) and total phosphorus (TP) were measured as waster quality characteristics. The actual influent concentrations were 40--125 mg BOD5/L and 91--200 mg TSS/L.;Sand bioreactors showed the best system performance for treating SSO with effluent concentration of 14 mg BOD5/L, which met 7-day effluent limit for stream discharge from conventional treatment technologies in Ohio's antidegradation rule (3745-1-05). All bioreactors removed >85% TSS in the SSO wastewater and three achieved >90% reduction. The effluent concentrations of TSS in all bioreactors were lower than 18 mg/L, the discharge standard of the Ohio antidegradation rule. These data suggested excellent performance of fixed media bioreactors in filtering solids at high hydraulic loading rates.;The bioreactors had the highest efficiency in the NH4-N and TP removal for the first 2-hr composite samples, indicating that bioreactors can be efficient in nutrient removal for a relatively short SSO event. The treatment capacity decreased with higher influent concentration during 4 hr and 6 hr loadings. Linear regression models were developed (r2 = 0.72--0.94) for the nutrient discharge at 4--6 hr loadings. Although influent concentrations and time course of loadings have negative impact on the treatment of TP, fixed media bioreactors showed constant efficiency with 2 mg TP/L in SSO wastewater, which is TP concentration in wet weather SSO estimated by USEPA (2004).;The treatment efficiency was also evaluated in two sand bioreactors with different resting period between two SSO events. The length of resting periods (1, 2, 3 and 6 months) between two SSO events did not show any significant impact on the treatment performance. Therefore, sand bioreactors should be able to treat the wastewater at long-term low frequency of loadings, which is a common characteristic of SSO events.;Different dosing strategies and loadings were observed in six sand bioreactors used for the SSO treatment at four high hydraulic loading rates. The result showed that performance declined with high levels of organic matter and at extremely high hydraulic loading rates. Ponding on the top and slow drainage were noted as operational problems. These problems were resolved by distributing the SSO in small frequent doses in pilot-scale experiments.;Linear regression models and Bayesian models were applied for the analysis of organic matter removal with same possible predictors. Linear regression models did not effectively build the relationship between effect factors and response but they presented some significant impact of possible factors on the treatment of BOD5 and COD. Bayesian methods produced good simulation for the experimental data of organic matter removal. This method can accommodate unbalanced and complex data in one model and was deemed an appropriate analysis to explore possible factors in this study. The results showed that the type of bioreactor media was the most significant factor in removal efficiency of organic matter in SSO.