Controlling reduced iron and manganese in a drinking water reservoir by hypolimnetic aeration and artificial destratification

Highlights

• WLAs replenished aquatic oxygen by bottom aeration and artificial destratification.

• WLAs can decrease the reduced Fe and Mn during seasonal hypoxia and rainfall events.

• Impact of flood on water quality was minimized by discharging at reservoir entrance.

• The diffusive flux of Mn was increased by operations of WLAs while Fe decreased.

• Influences of WLAs on biogeochemical cycles in sediment should be evaluated.

Abstract

The concentrations of iron (Fe) and manganese (Mn) in the water column have extremely important effects on the water quality of drinking water reservoirs; however, reservoirs often experience problematic Fe and Mn levels during seasonal stratification and rainfall events. Water-lifting aerators (WLAs) were deployed in the Jinpen Reservoir to control these issues with Fe and Mn at the source via bottom aeration and artificial destratification. In this study, variations of Fe and Mn concentrations in the water column, porewater, and sediments, were used to characterize behaviors of reduced Fe and Mn under the conditions of hypolimnetic aeration and artificial destratification during periods of hypolimnion hypoxia and rainfall events. The results showed that replenishing aquatic oxygen levels by aeration can effectively decrease the dissolved Fe and Mn in the water column thereby increasing the sedimentation rate and the diffusive flux of Fe and Mn at the sediment-water interface (SWI). The dissolved Fe was significantly chemically oxidized and the concentration remained relatively low in the water column during WLA operations, while dissolved Mn persistently accumulated in the near-sediment regions because of its complex kinetics. Our in situ profiles of labile Fe and Mn in the sediments demonstrated that the diffusive flux of Mn (J_Mn) was largely increased by the increased concentration gradient at the SWI, while the diffusive flux of Fe (J_Fe) decreased. The sediments were observed to rapidly become anoxic and release Fe and Mn after WLA deactivation; this emphasized the importance of appropriate operations linking the artificial and natural mixing periods to prevent SWI hypoxia and the release of reduced substances.

Introduction

In recent decades, with the exhaustion of groundwater resources and increasingly stringent prohibitions on resource exploitation, reservoirs and lakes have become the most important water resources for urban drinking water supplies at global scale (Godskesen et al., 2013; Hendrickson and Bruguera, 2018). Compared to the closed and cold storage conditions for groundwater, the surface water quality in lakes and reservoirs is severely threatened (e.g. by hypolimnion anoxia, climatic changes, and flood events) worldwide because of open storage conditions (Fink et al., 2016; Khan et al., 2015; Muller et al., 2012). Handling source water without pre-treatment can be very complex and expensive for water treatment plants. Therefore, approaches for improving in situ water quality are urgently needed by the managers of lakes and reservoirs (Beutel and Horne, 1999; Singleton and Little, 2006). Alternative facilities that address hypolimnion anoxia in lakes and reservoirs have been developed to replenish oxygen in the water column, including hypolimnetic oxygen systems (HOx), artificial destratification systems and water-lifting aerators (WLAs) (Debroux et al., 2012; Gafsi et al., 2015; Li et al., 2018). The depletion of oxygen in the hypolimnetic layer can cause the release of a considerable amount of reduced substances such as dissolved Fe and Mn, and other compounds from sediment that degrade the source water quality and complicate water treatment (Beutel, 2003; De Jonge et al., 2012; Granina et al., 2004).

Iron and manganese are the abundant transition metals that their biogeochemical cycles at the SWI greatly influence the water quality in aquatic ecosystems (Lovley, 2004). The aesthetic problems and health risks associated with excess Fe and Mn in drinking water systems have received much attention worldwide, and multiple guidelines regarding their concentrations have been established (Cerrato et al., 2010; Hafeman et al., 2007) and in China, the limitation of total Fe and Mn concentration in surface water are 0.3 and 0.1 mg/L, respectively. Although dissolved Fe is easily removed through biotic oxidation, Mn is a nuisance contaminant because of its complex redox kinetics, and it requires large amounts of oxidizers during the water treatment processes (Gantzer et al., 2009). However, chemical oxidizers can react with naturally occurring organic matter to form disinfection by-products, which increase risks to public health (Davies and Morgan, 1989; Morgan, 2005).

Water-lifting aerator and other in situ oxygenation utilities have been proven to efficiently decrease levels of dissolved Fe and Mn in the water column by replenishing dissolved oxygen (DO) through biotic/abiotic oxidation processes (Cover and Wilhm, 1982; Ma et al., 2015). Nevertheless, most of the published research on aeration systems has primarily focused on the oxygenation of and the efficiency of contaminant removal in the water column (Beutel and Horne, 1999; Li et al., 2018), but comprehension of the biogeochemical behaviors of Fe and Mn at the SWI with the use of various oxygenation utilities remains incomplete (Beutel and Horne, 1999). Some studies have demonstrated that once the dissolved Mn is oxidized and precipitated as oxide, it settles and then be redissolved by reducing conditions in the sediment (Bryant et al., 2011b). This accumulated dissolved Mn could be more easily released into overlying waters because of the large concentration gradient. A few in situ studies based on the HOx have emphasized that biogeochemical cycling must be considered in assessing the effectiveness of oxygenation operations (Gantzer et al., 2009; Munger et al., 2016). However, the effects of oxygenation systems on complex biogeochemical processes have not been comprehensively evaluated because of the low resolution of in situ data and the necessity for long data-collection periods (Bryant et al., 2011b).

Underflows can transport large quantities of dissolved and particulate Fe and Mn into reservoirs during rainfall events. Several studies have investigated the impacts of flood matter on the source water quality of reservoirs (Fink et al., 2016) but few solutions have been proposed because of the complicated intrusion of density currents (Huang et al., 2014). Moreover, settled particles in density currents could further exacerbate the excess sediment oxygen demand (SOD) and the release of reduced substances (Herzsprung et al., 2010; Huang et al., 2014). During rainfall events, the operation of oxygenation systems is cautiously ceased in case of sediments resuspension, thereby preventing studies of oxygenation systems on external pollution. In our previous study, we speculated that releasing density currents at the mouth of a reservoir would avoid underflows plunging into the main basin and decrease external pollution as much as possible (Huang et al., 2014). Based on the processes of discharging the flood at the mouth of reservoir for decreasing particulate contaminants, the effects of WLAs on the external reduced Fe and Mn under low-turbidity conditions (e.g. with small inflows or residual flood branches) must be understood.

The study presented here is based on in situ water column, sediment, and porewater data, which was collected to investigate the influences of WLAs on: (1) the accumulation of dissolved Fe and Mn during summer anoxia, (2) the external pollution of reduced Fe and Mn during rainfall events, together with the operations of discharging floods at mouth of reservoir, (3) in situ microprofiles of labile Fe and Mn in porewater in response to environmental variations and (4) the corresponding diffusive flux of Fe (J_Fe) and Mn (J_Mn) from sediments. The results of this work provide new insights for improving the water quality in drinking water reservoirs by using WLAs.