Date of Award

Winter 2020

Project Type

Thesis

Program or Major

Civil Engineering

Degree Name

Master of Science

First Advisor

Thomas P Ballestero

Second Advisor

Joseph Poythress

Third Advisor

James Houle

Abstract

Stormwater is known to be a major source of water pollution. In addition to increased flows resulting from increased impervious area in watersheds, stormwater is known to carry pollutants such a suspended solids, hydrocarbons, and nutrients into receiving water bodies. Because of this, stormwater treatment systems, referred to as ‘best management practices’ (BMPs), are utilized to dampen peak flows, reduce total runoff volume, and remove pollutants. While many studies have investigated the effectiveness of various BMPs in the removal of pollutants, fewer studies have investigated the effects of numerous storm variables on the effectiveness of treatment systems. This thesis investigates the effects of storm event variables on the removal efficiencies of different BMPs for different stormwater pollutants, with the goal of drawing broader conclusions on the effects of event variables on system performance.Using data from the University of New Hampshire Stormwater Center (UNHSC) and International Stormwater BMP Database, this thesis investigated the effects of variables such as event duration, precipitation depth, runoff volume, peak rain intensity, peak runoff flow, and the antecedent dry period on the percent removal (%RE) of pollutants such as solids, hydrocarbons, Nitrogen, Zinc and Phosphorus in BMPs such as bioretention systems, swales, retention ponds, subsurface gravel wetlands, and sand filters. Apart from investigating these specific relationships, these analyses were conducted to evaluate if event variables can serve as adequate indicators of BMP performance. Using statistical analysis techniques including regression analysis and principal component analysis, this study investigates relationships between storm event variables, influent event mean concentrations (EMCs), and removal rates across different pollutants and BMPs. Most relationships were not determined to be statistically significant at the 95% confidence level. Although few relationships were significant at the 95% confidence level, there were additional relationships found to be significant at lower levels of confidence. A trend of note was a group of positive relationships between several influent EMCs and removal rates. It is worth noting that antecedent dry period had no statistical bearing on the influent concentration and removal efficiency, when it is has been found by previous studies to effect pollutant loadings. This lack of bearing would be expected because of the use of EMCs masking first flush characteristics. These analyses were also used to examine the proportions of relationships effecting removal rates by system. These results indicated swales were these systems whose performance was most effected by changes in event variables, followed by sand filters, bioretention systems and subsurface gravel wetlands, and retention ponds being he east effected. Although analysis yielded a number of statistically significant trends indicating relationships effecting system performance, the results suggest that relationships effecting system performance are either not of sufficient statistical confidence, or are a result of additional intervening factors effecting performance. This indicates that hydrologic factors are not a sufficient indicator or predictor of system performance. Overall, these results substantiate performance in Green Stormwater Infrastructure (GSI) to be mostly independent of event variables because of the use of static design and regulated outflow. This backs up that GSI static design standards are adequate for the removal of the pollutants studied.

Download

Share

COinS