도시지역 LID 최적화를 위한 강우유출모델의 적용

다국어 초록 (Multilingual Abstract)

Particulates and heavy metal constituents (e.g. lead, zinc and copper) included in the stormwater runoff could vary from locations due to the variation in the nature and intensity of human activities. On the other hand, organic compounds could be found in higher concentrations at the elevated urban impervious areas (e.g. rooftops and ledges) because of the surrounding vegetation and airborne-related activities. Low impact development (LID) is a system developed for improving the altered water circulation and simultaneously reducing the pollutants involved in the generated runoff. Direct sampling of these facilities is laborious, expensive and time-consuming. For these reasons, computer simulation models were developed for modelling, estimation and evaluation of various treatment systems for both runoff quantity and quality simulations. The Environmental Protection Agency’s (EPA) Stormwater management model (SWMM) is a comprehensive hydrological and water quality simulation model developed for urban areas. The main objective of this study was to optimize the physical characteristics of three developed LID facilities and predict the possible volume and pollutant reduction capability of the mentioned systems with respect to varying design conditions thru simulation using SWMM. The three LIDs considered in this research were infiltration/filtration system, tree box filter and rain garden. The infiltration/filtration system was treating paved road runoff and consisted of a pre-treatment tank, media zone and final effluent tank. Meanwhile, the tree box filter was treating the runoff generated from a rough asphalt parking lot and composed of three stacked media layers namely top layer woodchip, middle layer sand and bottom layer gravel. The rain garden was treating runoff from impervious concrete rooftop and has a mixture of soil and gravel as the main filter media having corresponding depths of 1.5 m and 0.80 m for the middle and sides, respectively. These treatment facilities were calibrated and verified using the Box’s complex optimization algorithm which was performed using Matrix Laboratory (MATLAB) language. The total suspended solids (TSS) mass loads were considered as the main target pollutant in the simulation. The results showed that the TSS mass load was significantly correlated with total lead (TPb), total zinc (TZn) and total copper (TCu) for having a Person coefficient ranging from 0.87 to 0.95 (all p value less than 0.05). Furthermore, the volume and TSS removal efficiency of the three LIDs for treating paved road, parking lot and impervious rooftop runoff were ranging from 60 to 84%, 73 – 83% and 70 to 84% respectively. In order to optimize the physical design characteristics of each LID, the respective storage volume and surface area ratio (SV/SA ratio) and depth was adjusted with an increment of 0.25% both in increasing and decreasing trend. It was found out that the infiltration/filtration system and tree box filter have an improved reduction performance by an approximate of 70 to 90% on both volume and particulate removal for a -0.25% adjustment of SV/SA ratio and depth. Moreover, various design installation configurations were simulated to the three LID sites namely centralized system, parallel and series distributed system. Based on the results, the centralized system could maintain high reduction performance despite of doubled catchment area and lesser added maintenance and operation costs. Compared to the centralized system, the two distributed system have significantly improved the volume and pollutant reduction with an approximate range of 5 to 40%. However, these distributed systems required additional maintenance and operation costs. The reduction performance of each LID site with varying urban land uses (e.g. paved road, parking lot and impervious rooftop) were also simulated and evaluated. Based on the findings, the infiltration/filtration system has maintained a volume and pollutant reduction by at least 75% on the three urban catchment areas. Lastly, the three LID facilities were connected into series and formed a complex LID system with secondary treatment facility. The infiltration/filtration and rain garden were treating the respective urban land use however, the effluent was directly delivered to the tree box filter which served as secondary LID system. At the final discharge, the volume and TSS concentration were reduced by an approximate of 50 and 80% respectively.

국문 초록 (Abstract)

인간과 자연활동의 다양성에 의하여 강우유출수에 함유된 납, 아연 및 구리와 같은 중금속과 입자상 물질은 토지이용별로 다양하다. 반면에 유기물질의 경우, 불투수면이 높은 고도화된 도시지역에서 주변 식생과 대기침적의 영향 때문에 높은 농도로 나타난다. 저영향개발(LID)은 자연적 물순환을 구축함으로써 강우유출수의 유출저감과 함께 오염물질의 저감을 꾀할 수 있는 기술이다. 이러한 저영향개발 시설에서의 직접적 모니터링은 시간과 비용측면에서 매우 노동집약적이다. 이러한 이유로 컴퓨터 시뮬레이션 모델은 다양한 처리시설의 유출유량 및 수질의 모의, 추정과 평가를 위하여 개발되었다. 미국 환경청(US EPA)의 강우유출관리모델(SWMM)은 도시지역에 적용가능하며 수문학적 유출량 및 수질모의를 종합적으로 가능하게 하는 모델이다. 본 연구의 목적은 SWMM을 이용하여 다양한 디자인 상태에서 3종류의 LID 시설의 물리적 특성을 최적화하고 시스템내의 유출 및 오염물질 저감량을 예측하기 위해서이다. 3가지의 LID 시설은 침투/여과 시스템, 나무여과상자와 빗물정원이다. 침투/여과 시스템은 포장된 도로로부터의 강우유출수를 처리하고 있으며 전처리조, 여재층과 최종 방류조로 구성되어 있다. 한편 나무여과상자는 거친 아스팔트로 포장된 주자창의 강우유출수를 처리하기 위하여 설치되었으며, 3가지의 여재층으로 구성되었 있으며 위로부터 우드칩, 모래 그리고 자갈로 구성되었다. 빗물정원은 콘크리트 포장된 지붕 빗물 유출수를 처리하기 위하여 조성되었으며, 주요 여재로는 중앙부 깊이 1.5m와 바깥쪽 깊이0.8m에 토양과 자갈을 사용하였다. 이러한 처리시설은 박스 복합 최적 알고리즘(Box’s complex optimization algorithm)을 이용하여 검보증이 이루어졌으며, MATLAB을 이용하여 수행되었다. 모의에서 주요한 목표오염물질은 도시지역에서 흔히 문제가 되는 TSS 부하량을 선정하였다. TSS는 Total Pb, Total Zn 그리고 Total Cu와 피어슨 상관계수(모든 p 값은 0.05보다 낮음)가 0.87-0.95를 가질만큼 의미 있는 상관성을 보였다. 또한 도로, 주차장 및 지붕유출수를 처리하는 3개의 LID 처리시설에서의 유출량과 TSS 제거효율은 각각 60-84%, 73-83% 그리고 70-84%를 보이는 것으로 나타났다. 각 LID 시설의 물리적 디자인 특성을 최적화하기 위하여 각각의 저류량과 표면적 비(SV/SA비)와 깊이는 증가와 감소 측면에서 0.25%의 증분으로 조정되었다. 침투/여과시스템과 나무여과상자는 SV/SA비와 깊이가 -0.25%의 조정될 때 유출량과 입자상물질의 제거측면에서 약 70-90% 정도 저감율을 보였다. 또한 3가지 LID 시설을 중앙집중식, 병렬 및 직렬시스템으로 각각 연계될 때의 디자인 시설의 형상에 대하여 모의기 진행되었다. 연구결과에 의하면, 중앙집중식이 두배의 유역면적과 낮은 유지관리 및 비용에도 불구하고 높은 저감효율을 보였다. 이와 대비하여 다른 두개의 분산시스템은 5-40%까지 유출량과 오염물질 저감량을 증가시켰다. 그러나 이러한 분산시스템은 추가적인 유지관리 비용과 운영비를 필요로 하는 것으로 나타났다. 다양한 도시적 토지이용(주차장, 도로, 지붕 등)에서 각 LID 시설의 저감효율을 산정한 결과 침투/여과시스템은 3가지의 토지이용에서 최소 75% 정도 유출량과 오염물질의 저감량을 보였다. 마지막으로 3가지의 LID 시설을 상호 직렬로 연계할 경우, 유출량 저감과 TSS 저감량은 약 50%와 80%로 나타났다.

목차 (Table of Contents)

• Nomenclature x
• I. INTRODUCTION 1
• 1. Background 1
• 2. Research goals and Objectives 5
• II. RELATED LITERATURE 7
• 1. Urbanization impacts to natural hydrology 7
• 2. Urban stormwater runoff 10
• a) Catchment Area 10
• b) Water quality parameter in urban runoff 14
• 3. Nonpoint source mitigation strategy 20
• a) Nonpoint source (NPS) management measures in Korea 21
• b) Low impact development for NPS management 24
• 4. Computer modelling approach 29
• a) Prediction models utilized in urban stormwater management 29
• b) Stormwater management model (SWMM) 38
• c) Modelling of the pollutant process in urban runoff 44
• III. RESEARCH METHODOLOGY 48
• 1. Monitoring of the three developed low impact development (LID) sites 48
• a) Site location and physical design characteristics 48
• b) Water sampling scheme for the three LID sites 55
• c) Water quality analysis of the runoff samples 56
• 2. Calculation and data analysis of the observed water quantity and quality of the urban runoff 59
• a) Pollutant concentration analysis 59
• b) Statistical and descriptive analysis 60
• 3. Application of SWMM modelling approach 61
• a) SWMM modeling set-up for the three LID sites 61
• b) Simulation scenarios 64
• c) Calibration and verification process 66
• d) Scope of limitations in the SWMM modeling approach 73
• IV. RESULTS AND DISCUSSIONS 75
• 1. Data generation and model calibration 75
• a) Urban runoff characteristics from respective monitored events 75
• b) Hydraulic and water quality calibration 83
• c) Model verification results 100
• 2. Physical design optimization and performance evaluation of each LID sites 103
• 3. Effective installation configuration for each LID 108
• a) Volume and particulate reduction performance 109
• b) Maintenance and operation requirements 110
• 4. Land use substitution for performance evaluation 112
• 5. Series-connected system of the three LID sites for urban stormwater management 115
• V. CONCLUSION 117
• VI. REFERENCES 119
• VII. APPENDICES 133
• 1) Standard values for SWMM component parameter 133
• a) Depression Storage 133
• b) Manning’s n – Overland Flow 133
• c) Manning’s n – Closed conduits 134
• d) Manning’s n – Open channels 134
• 2) MATLAB calibration code 135
• Abstract* 140
• 국문초록 143
• Acknowledgment 145
• Curriculum Vitae 147