Runoff has important meaning in the context of the performance of post-construction best management practices (BMP) and compliance with National Pollutant Discharge Elimination System (NPDES) storm water permits. To illustrate, this work uses continuous simulation modeling to estimate water balance components from bioretention and compares the results to those of a representative natural condition for the St. Louis, Missouri, area. Three different bioretention scenarios were modeled, each with different capture volumes and design features. The water balance of the representative natural condition is based on a prairie landscape in clayey silt soil over limestone bedrock. If runoff is defined as all discharge originating from the BMP, then none of the scenarios achieved the target natural condition. This is because evapotranspiration (ET) is the dominant process removing runoff in a natural condition, deep infiltration (deep recharge) in the St. Louis area is limited by geology, and ET is a small part of runoff removal in bioretention. If runoff is defined as the sum of overflow and underdrain flow, then any of the three bioretention scenarios could be adequate. However, there is a significant difference in BMP size, and by correlation, construction and land costs. This work demonstrates why an assessment of the capability of BMP performance in the context of local conditions is needed before setting BMP performance standards into NPDES permits. The BMP water balance indicates that in geology typical of St. Louis, Missouri, mimicking or restoring the pre-development runoff volume is not achievable with BMPs that primarily rely on infiltration to reduce volume. Bioretention is predicted to increase stream base flow, but the overall volume of water discharged to sewers, creeks, and rivers would be greater than the natural condition. DOI: 10.1061/(ASCE)EE.1943-7870.0000625. (C) 2013 American Society of Civil Engineers.
