Kazi Rashid

ORCID Identifier(s)


Graduation Semester and Year




Document Type


Degree Name

Doctor of Philosophy in Civil Engineering


Civil Engineering

First Advisor

P Andrew Kruzic


ABSTRACT: This dissertation explores the effectiveness of a complex system employed to capture the first flush (FF) generated from industrial activities occurring at an airport. The First Flush System (FFS) utilized by the DFW International Airport is comprised of multiple in-line bypassing fuel separators and located in many parts of the Air Operations Area (AoA) of the airport as well as maintenance areas. The system collects and conveys a portion of the rainfall runoff generated from the drainage areas of the industrial land-use portion of the DFW airport to a Stormwater Pre-treatment Plant (SWPTP) and then to Trinity River Authority (TRA) Water Treatment Plant. The runoff which exceeds the capacity of the FFS system is bypassed to nearby outfalls and water bodies. Airport areas hosting activities such as air and ground traffic, operation and maintenance of air and ground vehicles and equipment, cleaning, fueling, deicing, etc. are considered as industrial areas requiring National Pollutant Discharge Elimination System (NPDES) permit. Industrial areas are significant sources of nonpoint source (NPS) pollutants, washed off by rainfall and carried to water bodies. The FFS at DFW airport is assumed to capture First Flush (FF) which is the initial portion of the runoff containing major portion of the washed-off pollutants, potentially enhancing treatment and cost efficiency. The study investigates the performance of the FFS using historical data analysis to estimate rainfall runoff capturing capacity of the system and to investigate the influence of potential factors such as rainfall depth, peak intensity, time duration, lag time, and average total baseflow volume on the percentage of the rainfall runoff captured by the system. Multiple Linear Regression (MLR) modeling was conducted to establish a relationship between the capture rate of the system and potential factors for predicting the performance of the system. Sampling activity, bypass analysis, water quality (WQ) analysis, and FF analysis were conducted as part of the investigation of the system. The FFS is complex due to varied flow sources and drainage areas. Historical data analysis of 60 rainfall events suggests that the system performs better in capturing rainfall runoff of long duration rainfall events of small to medium magnitude. Rainfall depth, and peak intensity were found to be strongly related to capture rate of the system, while time duration, lag time, and average total baseflow volume showed no strong trends. Regression analysis was conducted to establish a relationship between fraction of rainfall runoff captured by the system and the primary parameters selected in the historical data analysis. The model found that the system performed well for low to medium rainfall depths and peak intensity occurring over a longer period of time which is consistent with the findings of the historical data analysis. Another extended model which included lag time and average total baseflow was also investigated. It was found from the regression analysis of the extended model that the predictive capability of the model increases if lag time and average total baseflow volumes are included in the model, indicating that these two parameters could represent the complex hydraulic characteristics and field conditions of the system. Bypass analysis at the separators indicated that bypassing occurs during intense rainfall events, primarily influenced by rainfall intensity and timing. Findings suggested that the bypassing time of the separators were longer for rear-loaded rainfall events where the rainfall intensity causing bypass occurred at the end of the rainfall event. The cumulative rainfall depth at bypass was found to be less than 0.08 in which is significantly less than 0.25 in identified as FF rainfall depth by the DFW airport. WQ analysis of both SWPTP and Separator H1O revealed that the pollutant concentration levels were low in most cases except for Chemical Oxygen Demand (COD) concentrations at separator H1O. No sampled pollutant concentration exceeded the Texas Pollutant Discharge Elimination (TPDES Permit limit at SWPTP. The inflow at SWPTP was found safe to be discharged at outfalls instead of TRA water treatment plant at times when glycol from deicing activity was not being released. At the separator H1O, except for one oil/grease (O/G) concentration all O/G concentrations were below TPDES permit limit. The COD concentrations for the wet weather sampling at separator H1O was found to be high at the latter portion of the runoff. The baffle in the separator was found to have very little impact in retaining pollutants inside the separator, especially the pollutant concentrations of O/G found at different depths. It was observed that FF generally did not occur at the SWPTP except for a weak FF for O/G for one sampling event, indicating that the observation of FF in such a system is unlikely. The absence of FF occurrence is substantiated by the WQ and FF investigations. The complexity of the system, including varying lag times and large drainage areas, contributes to the challenges in detecting FF. Since the separators have their own designated sub-catchments as contributing drainage areas, it can be assumed that the complex flow conditions thought to occur at the SWPTP is less prevailing at any individual separator. Hence the possibility of observing a first flush at a separator is assumed to be higher. However, the first flush was also not observed from the one sampling event conducted at separator H1O, which could be due to the rainfall event being of very small magnitude, and therefore not being representative. Overall, the study provides valuable insights into the performance and limitations of the FFS emphasizing the importance of site-specific considerations in stormwater management. The capture rate analysis showed that the system is capable of capturing a good portion of the rainfall runoff volume for small to medium events, but because of bypassing phenomenon it is uncertain if that portion of flow volume is the FF flow, and if the major portion of pollutants is captured within that flow. The findings underscore the need for further research to comprehensively evaluate FF and its impact on pollutant capture in complex stormwater management systems.


First flush, Industrial activity at airport, Rainfall runoff capture rate, Water quality analysis, Bypassing fuel separators, Best management practices, Regression analysis


Civil and Environmental Engineering | Civil Engineering | Engineering


Degree granted by The University of Texas at Arlington

Available for download on Friday, August 01, 2025