Document Type

Thesis

Abstract

This honors thesis extends five years of research by undergraduate geology students at the College of William and Mary, evaluating the performance of wet detention ponds in James City County, Virginia. In this study, field measured peak inflows, peak outflows, centroid lag times, and runoff constants from two ponds were compared to design predictions and state and local regulations. HydroCAD models of each site were utilized to explore design elements providing uncertainty in predictions of pond inflows and outflows. Modeling results were used to interpret field performance data and provide recommendations for design improvement. Mulberry Place pond, designed using the rational runoff method, did not exhibit peak inflows, outflows, centroid lag times, or runoff coefficients consistent with design predictions. This pond also likely fails to reduce peak outflows of the 2 yr, 24 hr storm to pre-development values (state standard MS-19). Longhill Grove pond, designed with the SCS TR-55 method, did not perform in the field as designed with respect to peak outflows and centroid lag times. Both ponds likely fail to meet the local James City County standard requiring retention of the lyr, 24 hr storm for 24 hours. Field performance problems at each site resulted from three major design issues: 1) uncertainty in predicted and actual curve numbers and runoff coefficients, creating large differences between predicted and actual inflows and even larger differences between predicted and actual outflows; 2) a simplified “kerplunk” method for calculating 24 hour drawdown and orifice sizing, creating ponds that typically retain water for less than 12 hours; 3) pond volumes 2x and 2.5x smaller than needed to adequately reduce the effect of the first two design problems.

Date Awarded

2008

Department

Geology

Advisor 1

Gregory S. Hancock

Download

Share Feedback

Share

COinS