In recent years, owing to the rapid process of urbanization and the frequent occurrence of extreme storm weather, combined sewer overflows (CSO) pollution has gained intense attention all around the world. This type of combined drainage system is employed as the main drainage method in most urban areas. Unfortunately, various pollutants can be carried into the water body when this combined drainage system overflows during heavy storm events, which will lead to serious water pollution problems, such as the deterioration of water quality, eutrophication, and abnormal growth of aquatic organisms [1
]. Moreover, the CSO can result in other kinds of pollution to the river as well. For example, the research of the Suzhou industrial park showed that the overflow of stormwater was the source of heavy metal pollution [4
]. Therefore, the ever-increasing CSO pollution need to be urgently addressed by the improvement of the stormwater management standards and wastewater treatment plant (WWTP) [5
]. In order to minimize the impact of combined sewer overflows pollution, many cities have taken measures to build or expand the interception and storage facilities. However, most measures are expensive and ineffective, and the interception of stormwater mixed into the sewage system may lead to a reduction of pollutants concentration in the WWTP, greatly reducing the efficiency of biochemical treatment processes [7
]. Therefore, in order to effectively ensure the safety of the urban water environment, it is necessary to put forward corresponding strategies to solve the urban combined sewer overflows pollution problems.
With the development of computer technology and the Internet, many cases of real-time control (RTC) technology have been introduced into the urban drainage system to solve the problem of large cost and decentralized management [9
]. For example, Parolari et al. (2018) [12
] used a water level control method based on the steady-state probability density function to control the discharge of a stormwater retention pond. The results showed that this technology can not only increase the water storage capacity and hydraulic retention time but also reduce the increase in peak flow and pollutants caused by urbanization [12
]. Kroll and his colleagues utilized a water level based RTC control technique in Flanders, Belgium, and the results showed that the application of RTC can reduce the total overflow by 20%–50% [13
]. However, RTC technology based on flow rate or water level exhibits a low correlation between pollution flow and concentration, which cannot achieve the desired results [14
]. Thus, many scholars put forward the concept of pollution based real-time control (PBRTC) rule and corresponding intelligent water dispatching to realize the accurate analysis of pollutant concentration variation of combined sewer overflow. The PBRTC rules can transmit real-time data of water quality of CSO through sensors by optimizing online regulation rules and accurately monitoring the mixed-flow pollution process curve. Then, the controller is used to simulate the rainfall forecast, flow rate and water quality, optimize and generate control rules, thus achieving a better performance of real-time scheduling. With the development of water quality sensors and modeling tools, more and more attention has been paid to this control strategy, which is more effective than RTC based on flow or water level [16
]. Ly et al. (2019) [16
] compared the treatment effect of water-quality based RTC strategy QBR using the mass-volume (MV) curve with that of hydraulics-based RTC (HBR) in small scale wastewater networks (HBR). The results showed that the QBR was able to provide CSO load reduction from 3% to 43% compared with HBR for more than one-third of storm events [16
]. Weinreich et al. (1997) [18
] used the newly developed PBRTC simulation tool to reduce the total phosphorus (Ptot
) and NH4+
-N load of the urban drainage system by 48% and 51%, respectively [18
]. In general, it has been proved to be a feasible strategy to optimize and control urban storage facilities by PBRTC rules. However, in current practical application, the PBRTC rules are mainly used for improving the original intercepting-storage facilities by installing the corresponding intelligent control components. These methods rarely make use of the digital drainage model to acquire the systematic construction plan during the design and construction process, thus limiting the application of PBRTC rules to some extent [11
]. Therefore, in order to systematically control urban combined sewer overflows pollution, it is critical to select appropriate PBRTC rules to optimize the current digital drainage model, which would finally make most use of the roles of the digital drainage model and PBRTC in urban drainage system. Currently, there are successful cases of coupling RTC technology and drainage models. For example, Campisano A. et al. (2016) used the flow rate based RTC technology and drainage model, which can reduce CSO volume [20
]. However, water level or flow rate based RTC technology have poor correlation with pollutants in the stormwater, which obtain the undesirable results in some study. Therefore, coupling the PBRTC rule and drainage models have been tried out. For example, Sharior et al. (2019) compared the total suspended solid (TSS) based RTC with water level based RTC, which can obtain a 40% reduction in the occurrence of CSO [17
]. Nevertheless, TSS is not the only pollutant in the stormwater. It is very meaningful to explore the effect of the use of PBRTC based on other pollutants. As a result, the combination of PBRTC rule based on chemical oxygen demand (COD) value and digital drainage model was used in our study.
In this study, the stormwater pipe network system in Fuzhou is used as an example. By using InfoWorks ICM and ArcGIS, the two-dimensional drainage model is established, validated, and calibrated, which is used to optimize the traditional interception-regulation scheme and evaluate its chemical oxygen demand (COD) interception capacity and overflow COD concentration. Based on the PBRTC control rules, a double-gate scheme based on water quality was proposed, which used the COD as the control index and combined the local pollutant control specifications to confirm the specific COD value, and its intercepting capacity and overflow situation were simulated and evaluated. Under the same designed rainfall, the traditional scheme was compared with the double-gate scheme, including the facility size, COD interception rate, and COD concentration over 50 mg/L overflow volume. Furthermore, the operation performance of the double-gate scheme and the traditional scheme beyond current design condition, including the interception COD rate, the average COD interception concentration, and the over-standard volume of the overflow stormwater were further studied.