(本文)Lately, various problems originated from combined sewer overflows (CSOs) during wet weather events have been becoming social issues. CSOs occur by wet weather events in the main, when wet weather flows (WWFs) exceed the capacity of a combined sewer system (CSS) or a wastewater treatment plant. When CSOs occur, contaminated stormwater flows into receiving waters directly without any appropriate treatment in general, and it has the high potential to affect adversely those aquatic environments. Urban catchments are particularly problematic due to their landuse property with high rate of runoff containing a great deal of surface pollutant deposits accumulated in antecedent dry weather period (ADWP).

Usually WWFs of CSS mainly consist of surface pollutants, domestic wastewater and in-sewer deposits. Moreover, the generation and transport of pollutants in CSS during wet weather events are very complex because of various media, space and time scales. Therefore, in order to understand the behavior of CSOs, it is one of the best approaches to observe the runoff behavior of these constituents separately. In particular, the drainage characteristics of dry weather flow (DWF), the behavior of particulate pollutants in CSS and in-sewer deposits should be investigated thoroughly.

The concentrations and loads of pollutants in WWFs depend on several factors such as rainfall conditions, washoff of surface pollutants, runoff of in-sewer deposits and diurnal fluctuation patterns of DWF. It is considered that, whilst the surface pollutants and in-sewer deposits, whose total loads change due to ADWP, affect the first flush stage mainly, discharge loads originating from DWF influence the whole of WWF duration with different time-series variation patterns for every pollutant.

Numbers of researches on the behavior and risks of pathogenic microorganisms in CSOs including bacteria and viruses as well as the impact by overall pollutants and nutrients have been reported. A large number of research results have pointed out that in-sewer deposits in CSS are closely related with the behavior of pathogenic microorganisms. Pathogenic microorganisms with considerable concentrations are transported in the adhering form to solid materials, deposited and concentrated in the sediment of CSS in dry weather, and during wet weather events such sediment is flowed out rapidly. In addition, it has been also reported that there is fair possibility of repopulation within the deposits in CSS especially in the case of bacterial indicators and coliphages.

In order to assess the impacts of CSOs, to protect the public health and to enhance the environmental quality of receiving waters, it is crucial to investigate the runoff characteristics of WWFs and to develop a detailed model for the simulation of the runoff dynamics of each pollutant and microorganism in WWFs. Therefore, the specific objectives of this study were (1) to monitor the diurnal fluctuation patterns of each pollutant in DWF of CSS and to evaluate their characteristics, (2) to validate the drainage properties of microorganisms in DWF, (3) to grasp the behavior of particulate materials in CSS and the relationship between particulate materials and microorganisms, (4) to investigate the washoff behavior of in-sewer deposits and to elucidate the role of in-sewer deposits for the runoff of pollutants and microorganisms in WWFs, (5) to develop a model for the simulation of the runoff dynamics of WWFs based on the measured data of the study area.

To study the diurnal fluctuation patterns of each pollutant and microorganism in DWF, 24 hour monitoring was conducted two times at a combined sewer catchment in dry weather. Multivariate analyses including correlation analysis, cluster analysis and factor analysis were carried out using the 24 hour time-series monitoring data. Moreover, artificial field flushing experiments which could separate the washoff of in-sewer deposits from surface effluents in WWFs were carried out in order to investigate the washoff behavior of in-sewer deposits in CSS. The time-series washoff characteristics of each pollutant, bacterial indicator and enteric virus were investigated. To grasp the details of the washoff behavior of each pollutant and microorganism, time-series analysis on the runoff concentration and load curves was conducted, and all kinds of indices on the first foul flush were utilized. Furthermore, in order to simulate the dynamics of pollutants and microorganisms in WWFs, a sewer runoff model was developed based on on-site monitoring data. The models were especially aimed to describe the runoff behavior of in-sewer deposits and the diurnal fluctuation of DWF in CSS. Lastly, the runoff behavior of each pollutant and microorganism was simulated for various rainfall conditions using the model developed in the study. The behavior and contributions of each constituent such as surface pollutants, in-sewer deposits and DWF were simulated respectively.

The results of 24 hour monitoring of the time-series concentration and load variations of DWF at a small subcatchment in CSS were discussed in Chapter 4. The ranges of diurnal load variations of each water quality parameter reached ten to several tens times within a day. It also became obvious that the 24 hour concentration and load fluctuation characteristics of each parameter differ from each other. The patterns of concentrations and loads of SS, VSS and COD showed their peaks at late night and in the morning; and TN, DTN and NH3-N in the early afternoon and in the morning. Bacterial indicators were discharged rather uniformly from daytime to midnight representing the highest peaks in the morning, after continuous decrease till early morning. On the other hand, coliphages and enteric viruses revealed far intense variations of concentration and load.

In Chapter 5, multivariate analysis was performed using the 24 hour load variation data in order to characterize the diurnal fluctuations of each pollutant and microorganism in DWF. All of the water quality parameters were divided into three groups according to their origins and existence forms: The nitrogen parameters (TN, DTN and NH3-N), conductivity, phosphate and bacterial indicators belonged to group 1, which was mainly considered to originate from the toilet wastewater (urine, feces); washing wastewater-related materials such as LAS and fluorescent agents (DSBP, DAS1) belonged to group 2; and organic solid materials such as SS, COD and turbidity categorized into group 3. Therefore, in order to monitor the behavior of WWFs appropriately, it should be taken care enough that the discharge patterns of each water quality parameter, for instance bacterial indicators, may differ from those of overall water quality parameters greatly.

The washoff characteristics of in-sewer deposits in CSS were investigated in Chapter 6. The washoff patterns of each parameter resulting from in-sewer deposits by an artificial field flushing were to be categorized into typical 3 groups; strong first foul flush group including SS, VSS and COD. This group was estimated to be caused mainly by the erosion of upper deposit layer containing organic solids abundantly; partial first foul flush group including bacterial indicators. This group was considered to be influenced by a few factors, i.e. the earlier first foul flush-like washoff, the latter "delayed" washoff by additional erosion of the residual deposits and the inflow of DWF; and no first foul flush group including enteric viruses and DOC. The order of these groups signifies the strength of the first foul flush and the runoff priority as well. In view of practical management of CSOs, much portion of SS, VSS and COD runoff loads can be controlled by relatively small interception of WWFs. However, this countermeasure appropriate for strong first foul flush group would not work well for bacterial indicators and enteric viruses. If our concerns are laid on the management of pathogenic microorganisms, new strategy should be explored.

In Chapter 7, a sewer runoff model composing a distributed runoff model was developed. The values of key parameters, i.e. potency factors of each pollutant in DWF, attached ratios of microorganisms to suspended substances, particle size distribution, representative size and specific gravity of particles, were experimentally determined. Instead of present simulation using one set of representative size and specific gravity, it was confirmed that the simulation considering two particle groups with different representative sizes and specific gravities was reasonable. Furthermore, while discussing the expected problems by direct input of 24 hour profiles surveyed at a specific point, new methodology which estimates the source profiles using the data measured downstream was established. To understand the transformation characteristics of flow rate and each water quality parameter in the transport process in CSS, runoff factors composed by time delay factor, removal efficiency factor and flattening factor were proposed and applied to estimate the source profiles. As a result, the simulated profiles using the estimated source profiles represented good agreement with the measured profiles. In particular, the accordance of 24 hour SS profile was greatly improved.

In Chapter 8, the runoff dynamics of each pollutant and microorganism were simulated for various rainfall conditions using a distributed runoff model, InfoWorks CS ver. 7.5. The origins of each constituent in WWFs were classified into surface deposits, coarse and fine particles in DWF and dissolved substance in DWF, and simulated individually. The runoff dynamics of each water quality parameter in WWFs were different from each other according to their main origins. SS, COD and TOC represented strong first foul flush resulting from the erosion of coarse particles in in-sewer deposits. TN and bacterial indicators showed partial first foul flush. Coliphages revealed no first foul flush. From the viewpoint of runoff load, the earlier first foul flush originating from in-sewer deposits and the latter prominent inflow of surface pollutants were to be divided distinctively; the latter was observed in SS, TOC and TN. The loads of bacterial indicators and coliphages from surface deposits were negligible. After the first flush stage, the runoff concentrations and loads of WWFs were directly governed by the time-series variation of DWF. Hence, in order to manage CSO-related measures, it would be significant to consider the time-series variation patterns of DWF as well as the first (foul) flush behavior of each water quality parameter.

Therefore, it could be concluded that this study provides new information and the simulation tool for better understanding of the runoff dynamics of each pollutant and pathogenic microorganism in WWFs considering in-sewer deposits and DWF, which will be truly useful to apprehend the runoff behavior of CSOs and to explore the measures for CSO issues.

本研究は、「Characterization and Modeling of Wet Weather Pollution Dynamics in Combined Sewer considering In-Sewer Deposits and Dry Weather Flow」と題して、9つの章から論文を構成している。

第1章では、研究の背景と目的、および論文の構成を述べている。

第2章では、合流式雨天時越流水(CSO)問題やそれに伴う水質汚濁現象、越流水中に含まれる汚濁物質や病原微生物に関する問題点について詳細に整理している。また、都市域の雨天時汚濁流出現象をモデル解析する手法についての取りまとめも行っている。

第3章では、研究対象地域とした千葉市内の合流式下水道排水区の状況や下水採取地点の特徴、下水管渠からの下水サンプリングや流量測定方法、下水中に含まれる汚濁物質の化学分析方法、培養法および分子生物学的手法に基づいた微生物の定量方法が示されている。

第4章では、対象排水区(面積67.4ha、人口5518人)において実施した晴天時下水水質の24時間変動調査(2回)に基づき、汚濁負荷量変化は10倍から数十倍あること、その負荷量変動パターンが水質項目ごとに異なることを明らかにしている。そして、朝方と夜に負荷量ピークを示すグループ(SS、VSS、COD)、朝方と昼過ぎにピークを示すグループ(T-N、DTN、NH3-N)、朝方に高いピークを示すグループ(大腸菌群等、エンテロウイルス)、さらに特定のパターンを示さないグループ(コリファージ、ノロウイルス)に分類することができることを示している。また、特定のパターンを示さない微生物指標の挙動を詳細に評価するために、その一部排水区(面積1.1ha、人口110人)で5分間隔の採水調査を実施した結果、細菌類やエンテロウイルスはその濃度変動が比較的安定しているのに対して、アデノウイルスやノロウイルスについては、5分程度の短時間においても大きく変動することを確認している。

第5章では、4章における24時間水質変動データをもとに、相関分析、クラスター解析、主成分分析などの統計解析を実施して、負荷量パターンの類似度を定量的に評価している。その結果、家庭排水や管路内堆積物の巻上げにより増加すると想定される水質項目(濁度、SS、VSS、COD)、主としてトイレ排水が起源と想定される水質項目(栄養塩類、健康関連細菌指標群)、洗濯排水が起源と想定される水質項目(界面活性剤、蛍光増白剤など)といった3グループに分けられることを示している。このことは、CSO対策による水質モニタリングにおいて、グループごとに連続測定しやすい代表指標を選ぶことで、汚濁負荷変動の全体像を簡易に把握できる可能性を示唆している。

第6章では、従来からCSO汚濁負荷に大きく寄与すると指摘されている管路内堆積物の雨天時流出特性を評価するために、人為的に大量の水を管渠に投入するフラッシュ実験を現場で実施した結果を報告している。水質項目ごとの流出パターンの相違を評価するために、平常時の汚濁負荷量に対する流出汚濁負荷量の比の大小、濃度と流量との相関図における時系列変化パターンの相対的な比較を行うことの有効性を指摘している。その結果、SS、VSS、CODなどは初期フラッシュ現象が顕著であること、健康関連細菌指標群は初期フラッシュ現象から部分的に遅れた負荷量ピークが観察されること、DOCやウイルス類については初期フラッシュ現象がなく、緩慢な負荷量ピークを示すことを示している。また、初期フラッシュ現象が顕著であったSSを基準として、平常時と流出時の水質項目ごとの負荷量の比を時系列的に求めることで、管路内堆積物の雨天時流出挙動との関連について議論している。

第7章では、家庭排水中の汚濁物が晴天時には管渠に堆積する現象を考慮して、晴天時汚水中の粒子状物質の挙動を再現できる流出解析パラメータの設定を行っている。実下水に対して適用したイメージアナライザによる観察と湿式での篩い分け試験の粒径分布結果から、細粒と粗粒の二つの代表粒径にそれぞれ比重を設定すると共に、粒子に付着する微生物の割合を実験的に決定することで、晴天時における下水管路流下現象をモデルで再現できることを示している。

第8章では、7章で求めた晴天時における下水汚濁物質の堆積状況を予測したのち、典型的な降雨条件を与えて汚濁物質の流出過程を推定した結果を示している。各水質項目の推定結果は、人為的なフラッシュ実験において観察された流出パターンと類似していることを報告している。また、汚濁負荷の挙動を起源別に追跡することで、CSOにおける汚濁負荷のうち内訳についても考察を行っている。

第9章では、上記の研究成果から導かれる結論と今後の課題や展望が述べられている。

以上の成果では、雨天時合流式下水道越流水に伴う汚濁負荷量を解明する上で役立つ貴重な調査データやその解析結果を提供しているだけでなく、水質項目ごとの流出挙動・起源を考察することで、汚濁対策に有用な提言を行っている。さらには分布型モデルによる雨天時汚濁負荷流出解析の基礎的な手順手法を提供しており、それらの知見は都市環境工学の学術の進展に大きく寄与するものである。

よって,本論文は博士(工学)の学位請求論文として合格と認められる。