David A. Reckhow, Emily Kumpel

Publication Date

Spring 2021


Rainwater harvesting systems often include quality control systems such as a diverted first flush volume to improve the collected water quality. The first flush volume has traditionally been defined as a set volume of rain based on the first 1-2 millimeters of rain that falls on a roof. Diverting a volume of water can be seen as a waste when rainwater is a main source of potable water, sometimes leading to lack of implementation, and thus contaminating the final collected water. Understanding the variability of first flush volume required due to environmental parameters can be used to develop an optimized first flush system. This study evaluated rainwater catchment first flush volumes by assessing the rainwater quality over volume and time. To study these effects, we built a rainwater collection system on a test site in Amherst, Massachusetts. We performed a tracer study with the rainwater collection system to model the first flush volume required to wash out a dissolved contaminant. We collected four rain events using a fractionation first flush design. We measured water quality parameters in the atmospheric rain, first flush, and collection tank samples for each rain event. Our first flush samples resulted in elevated dissolved organic carbon (DOC) concentrations up to 40 mg/L, although there was high variation between the rain events. UV 254, DOC, and conductivity all trended together within each rain event, demonstrating a uniform wash off, of contaminants. Indicator bacteria up to 200 MPN/100 mL within rain event 1 and 2, indicates the need for disinfection if the water is to be potable. The high levels of DOC and SUVA characterization presented a concern for disinfection by-products (DBP) potential if the water were treated with chlorine. Higher intensity storms seem to increase roof wash-off deposition in the first flush. The majority of contaminants washed off in the first flush seemed to originate from roof wet and dry deposition, demonstrating the need for variable first flush volumes. Hydraulic parameters that affect wash-off, such as rain intensity and collection location, also led to varied first flush volumes. Considering these factors in the first flush volume required, could decrease treatment needs, system maintenance, and concern from treatment by-products.