Urban Drainage Modeling

Conventional urban drainage models deal with rainfall-runoff analysis under the effects of local storms. Additionally, urban areas are also vulnerable to local floods caused by surface flows of the anthropogenic origin (main bursts for example) which also need to be modelled. In most of the cases commercially available models are relatively reliable when dealing with free surface and pipe flow separately. However, when the system becomes surcharged, reliability of models becomes strongly dependent on the quality of modelling the interactions between surface and underground sub-systems. At present, the systems for modelling these interactions reliably as a part of urban flooding analysis do not seem to exist. The paper addresses this problem, presents some results of the ongoing research in development of a methodology for such modelling, placing an emphasis on data pre-processing and points out the future research needs in the area.

We proposed in this paper a RTC strategy allowing surcharged flows in combined interceptors in order to minimize volumes of untreated water discharged to receiving waters during rain events. The strategy uses a newly developed hydraulic model to compute the impact of the controlled variables' values on flows and water depths in the sewer. The selection of the controlled variables' bounds and the formulation of the objective function to be minimized guarantee the convexity of this last function. Consequently, use of a robust local optimizer is adequate to find its global minimum. Application of the elaborated control strategy on the combined interceptor of the city of Laval, Canada confirmed that allowing surcharged flows in the interceptor leads to considerable decreases in total overflow volumes. It was also shown that, assuming proper locations of the controlled regulators, a global reactive control strategy was adequate when controlling a surcharged sewer. Many operational benefits can be gained from the application of a reactive rather than predictive control. Among others, reactive control eliminates the need to resort to rainfall predictions when planning regulation actions; problems related to the acquisition of these data and to their associated uncertainties can then be eradicated. However, surcharged flows are seldom allowed in every pipe of a sewer network to control and some sections always remain under the free surface regime. More studies should be conducted on such sewers to evaluate the impact of the prediction horizon length on overflow volumes and on compliance with operational constraints when free surface flows occur. Conclusions of this paper regarding global reactive control of surcharged interceptors should also be corroborated by optimizations performed on other combined interceptors and using different rain events. A strategy to determine the optimal number and position of the controlled regulators could then be defined according to the results obtained.

The Runoff Block of the EPA Storm Water Management Model (SWMM) has been modified to permit alternative flow pathways during overland flow that permit routing of flow from impervious areas over pervious areas, for instance. The modifications for quantity and quality are described in this paper. These changes should make it easier to evaluate new stormwater management technologies, such as low impact development.

Continuous and multiple event simulations with long time series of rainfall were carried out to investigate the effects if distributed rainfall data are used instead of a single uniform rainfall input. The data basis for this investigations are continuous time series of rainfall data measured over 40 years at 8 rain gauges in an area of about 8 x 10 km. Two large distributed catchments have been modelled and simulated, one with a hydrological model for the calculation and evaluation of combined sewer overflow volumes, the other with a hydrodynamic model for the assessment of surcharge frequency. The analysis of the results of the simulations have been focused on two questions: what is the variance of the results when the time series of different but neighbouring stations are used and how do the results with distributed rainfall differ from those with uniform rainfall. Compared to the mean results from the uniform rainfall calculations, the yearly overflow volumes from distributed rainfall are not significantly lower. With the hydrodynamic calculations for the surcharge frequency, however, the results showed that surcharge occurrences are significantly lower when distributed rainfall data are used.

A local grids-refined method has been developed for a two-dimensional surface inundation simulation with the help of the parallel computing technique to solve the detailed inundation depths and velocities in a mild-slope alluvial plain. The Miau-Li Station of the Taiwan High Speed Rail System (THSRS) is selected as the study area. Two numerical resolutions are adopted in the model are solved by two individual processors. The coarser grid is used for global domain computation, and the finer one is for local region computation around the station. The processors compute the domains in parallel and communicate each other through the Message Passing Interface (MPI). Under the same numerical result, the computation time is efficiently shortened when parallel computing is employed than the sequential computing.

A broken-back culvert has one or more breaks in profile slope and is most often used in areas of high relief and steep topography. The usual broken-back configuration begins with a relatively flat slope followed by a steep slope and a very flat "runout" section. Because of the steeply sloped portion of the culvert, outlet velocities in the absence of a hydraulic jump are very high. Past designs have simply assumed that a hydraulic jump will form in the runout section. Until recently there have been no adequate design tools to address this unique culvert. A computer program called BCAP (Broken-back Culvert Analysis Program) allows for the design of such culverts. BCAP is based on the energy equation (direct step method) and momentum equation (sequent depth equation) for classical jumps. Algorithms compare the subcritical sequent depth inside the culvert to the tailwater elevation — if the tailwater elevation is higher, a hydraulic jump will occur. If a jump forms, the location and length jump length are summarized. BCAP is a Windows-based program that uses the same equations found in HY8 for predicting headwater elevations. Water surface profiles are plotted for the entire culvert length for ten discharges. Rating curves for headwater, outlet velocity, and outlet depth allow the designer to determine the range of flows for which hydraulic jumps occur. If desired, the material in the runout section may be altered (e.g. to increase the frictional losses) to increase the likelihood of hydraulic jump formation. An optional feature is to link directly to the HY8 energy dissipation module, usually necessary if a hydraulic jump does not form. BCAP results have been tested in a laboratory flume and applied to several projects in the Midwest.

Calibration of models is mostly focussed on obtaining an ‘optimal set of parameter values'. Analysis of the properties of these parameters and the analysis of the differences between model and reality after parameter optimization are in most cases neglected. In this paper some techniques for residue analysis and their applications are discussed and demonstrated by an example.

Geographic Information Systems (GIS) technology has become a valuable resource in stormwater modeling and planning. GIS technology has evolved to the point where it is becoming commonplace on the engineering desktop. Much of the functionality that was limited to specialists only a few years ago is now available through both readily accessible public domain tools as well as commercial packages. This paper describes how GIS was used in the analysis of a closed stormwater basins in Ocala, Florida, using a combination of public domain tools and standard GIS software with no additional programming or customization. Hydrological applications included using digital topographical data to delineate subbasins and generate flowpaths and slopes, obtaining percent imperviousness from orthophotos, and determining soil type percentages for each subbasin. Hydraulic applications included generating the components of a stormwater model (e.g., stage-area relationships and connectivity) and mapping the flooded areas based on model results. By way of an example application, this paper demonstrates how GIS tools have evolved to point where they may be used as a standard engineering desktop application to improve the efficiency, accuracy, and clarity of stormwater modeling and planning projects without the need for highly specialized GIS training.

In order to investigate the relationship between infiltration variation in the upper soil layer and rainfall intensity variation, a rainfall simulation study was carried out under unsteady rainfall conditions at the site of a weeded garden. The infiltration curves for unsteady rainfall were simulated using the simple conceptual infiltration model proposed by Diskin and Nazimov. This model takes into consideration the rainfall intensity and rainfall intensity variation. The results show that the simulated infiltration curves using this model approximately reproduce the observed infiltration curves. In conclusion, this study reveals that this model can be used to calculate the infiltration capacity variation in the upper soil layer under unsteady rainfall conditions in the field.

This paper deals with urban snow surveys in the Risvollan research catchment (20.1 ha) in Trondheim, Norway. The measurements are described and the results from 46 days in the period 1994 – 2000 are presented. The purpose of conducting the snow surveys is to better understand the snow distribution in urban areas and to improve urban snowmelt models. This is important because urban snowmelt influences frequency of flooding and combined sewer overflow through the winter season. The Risvollan catchment has ca 30 % roads and roofs compared to the areas with grassland and park area. Three different snow courses (10 points in each course) are defined for the catchment. One goes through housing, road and kindergarden playground area and is referred to as the urban snow course (USC), the two other courses go through grassland and park area and is referred to as the rural snow course (RSC). Human activity is strong in the USC and the results show that the data from the USC is not normally distributed. This is a result of redistribution of snow through ploughing. It is shown that the data from the RSC is statistically normal distributed. This research shows that there is a representative point in the Risvollan catchment that relatively often has an SWE close to the average of the rest of the point measurements. A time series of SWE through the winter could be generated if a snow pillow was installed in this representative point or SWE was measured frequently. A filtering method called "normalizing and sorting" was used on the SWE data to examine the skewness of the snow distribution. The results show that during the winter period the effect of wind on the snow distribution is relatively small and constant in the Risvollan catchment. This is believed to be a result of similar wind effects during snow precipitation events. The snow survey data can be used to find parameters for an areal depletion curve that could be used in an urban snowmelt model for the Risvollan catchment. A method on how to distribute precipitation in an urban energy balance snowmelt model is illustrated.

In this paper a procedure to solve tree type or looped sewer networks considering a steady gradually varied flow approach is presented. The method considers pressure and free surface flow in both sub and supercritical conditions, including the hydraulic jump location. A simple application is showed, considering a looped network with the different flows that can be represented. Extension to consider pressure flow in some reaches is included. The numerical procedure is explained in full detail.

As part of its Long Term Combined Sewer Overflow Control Plan (LTCP), the Philadelphia Water Department (PWD) has proposed implementing a number of technology-based capital improvement projects to the City's sewerage to further increase its ability to store and treat combined sewer flow, decrease CSO volumes, and improve receiving water quality. A major component of the proposed capital improvement projects is the installation of two inflatable dams operated under real time control (RTC) in the City of Philadelphia's Northeast Drainage District (NEDD). The RTC structures would provide the PWD with the ability to optimize the use of available in-system storage during wet weather and to minimize overflows to sensitive receiving streams. Hydraulic and hydrologic models were used to develop a control logic for each of the inflatable dams to optimize storage utilization and minimize potential flooding impacts of the projects. A continuous simulation EXTRAN model was modified to incorporate the inflatable dams and their control logic. In order to characterize the benefit of the inflatable dam projects, continuous simulations were performed to quantify the CSO reductions on an average annual basis. A 2.3 million cubic meter (12%) reduction in average annual CSO volumes to the Tacony Creek is expected through the implementation of these capital projects. This paper documents the control logic development of the inflatable dams to optimize utilization of in-system storage and reduce CSO discharges, the development of the continuous simulation models used to quantify the average annual CSO reductions, and model simulation results.

Most current work regarding the outputs from CSOs and storage tanks is concerned with the treatment of aesthetics and gross solids separation. However, finer suspended and soluble pollutants are more important when considering water quality criteria. Work has been carried out to investigate the retention time and dispersion characteristics of a high side weir CSO structure in controlled conditions by employing a fluorescent solute tracing technique. Data from this work have been analysed to predict coefficients for input to ADE and ADZ numerical techniques, that could easily be incorporated into both river impact and sewer quality models. This would allow more accurate predictions of temporal variations in quality of the sewage in both CSO spill and continuation flows.

Blow-off of storm/combined manhole covers may be attributed to the severe pressure transients which can occur during the transition from free surface to pressurized flow in a sewer system. Observations from a physical sewer model with a submerged outlet indicate that the pressure transients were composed of an initial low frequency water hammer type pressure surge and a subsequent high frequency pressure fluctuation due to the release of trapped air at the upstream manhole. A mathematical model, which was based on the assumption of rigid water columns and a compressible air bubble, was derived to simulate the pressure transients. The frequency of pressure transients predicted by the calibrated mathematical model is in close agreement with that recorded during the laboratory experiments. The attenuation of the pressure fluctuations, however, is underestimated. This may be attributed to the superposition of various air release processes observed during the experiments and the steady-state friction factor used in the mathematical model. This new approach for modelling pressure transients in storm/combined sewers provides an efficient alternative to the traditional Preissman's slot method for mixed flow analysis.

This paper will look briefly at the development and implementation of Real Time Control (RTC) within the sewer system in North West Water (NWW), including the benefits and problems it has brought to the company, environment and community. It will look at several large urban catchments as case studies, varying from the largest tourist resort in Europe (Blackpool on the Fylde Coast) to a large industrialised inland town (Bolton). It will detail systems implemented in previous years, recent improvements and future plans.

An innovative method for planning a Sanitary Sewer Overflow (SSO) management program is described. Heuristic optimization techniques are coupled with performance reliability information to give a least cost set of design solutions that meet probabilistic performance constraints. Long-term hydrologic simulation is used to estimate Rainfall Derived Inflow and Infiltration (RDII) generation based on rainfall records and recent monitored system response to RDII. This information is used in a design simulation model of upstream and downstream SSO controls. A Genetic Algorithm (GA) is used to find least cost design solutions for long-term performance criteria, and a Monte Carlo analysis is performed on each solution to evaluate probabilistic performance constraints. In addition to the single optimal solution found using this risk optimization procedure, a final set of near-optimal solutions and associated reliability information is found to guide design and regulatory decision making. This information may be used to guide an SSO management plan to satisfy regulatory constraints in a least-cost and reliable manner.

A hydrologic and hydraulic modeling system was applied to a mixed combined and separate sewershed to support system expansion planning and design, and to support Combined Sewer Overflow (CSO) regulatory compliance. The model was developed in two stages, or tiers. Tier 1 results, based on a detailed representation of CSO regulator and interceptor hydraulics and a simplified representation of system hydrology, aided in the planning phase for facilities needed to meet combined sewage capture requirements. The refined Tier 2 results, based on more detailed representations of both hydrologic and hydraulic elements, supported the final design stage of a new relief sewer. In addition, initially Tier 1, and now Tier 2 models are used for quarterly discharge monitoring reporting required by a National Pollutant Discharge Elimination System permit for CSO discharges.

Laboratory experiments were carried out to determine minimum self-cleansing velocities for preventing deposition of sediment in upward sloping and vertical pipes of inverted siphons. Tests were made using sediment sizes of 0.78mm and 4.3mm in a pipe of 150mm diameter for eight angles of inclination between 0° and 90°. The criterion for self-cleansing was defined as the minimum velocity needed to prevent the formation of separated dunes in the pipes. For a given sediment concentration the self-cleansing velocity was found to be a maximum at pipe slopes between about 30° and 37.5°. Analysis of the data enabled the results to be described by a single formula valid for any slope of pipe between horizontal and vertical.

As a result of the findings of recent research undertaken by the Cooperative Research Centre for Freshwater Ecology (CRCFE) and others on the ability of pollution control ponds and constructed wetlands to improve stormwater quality, the water industry in Australia identified the need for improved pond and wetland models to assist industry to analyse, design and implement stormwater ponds and constructed wetlands that are better able to improve the quality of stormwater discharged to receiving waters. These improved models needed to: (i) better reflect the range of pollutant pathways and transformation processes, (ii) be capable of simple adaptation to local conditions, and (iii) be accessible to a wide range of users. In response to this need the CRCFE has developed separate pond and wetland (spreadsheet) models that comprise interactive physical, chemical and biological process submodels that include sedimentation and sediment redox processes in the case of ponds, and biofilm uptake processes in the case of wetlands. The application of the pond model to the investigation of pollution interception and water quality in a proposed urban lake in northern tropical zone of Australia, is presented to demonstrate the model capability. A particular concern associated with the proposed lake was the potential for nuisance algal and macro-plant growth over the extended tropical dry season. It is concluded that the case study demonstrates the ability of the CRCFE's pond model to assess linked physical, chemical, biological and microbial processes and to determine the resulting water quality and its changes over time in pond/wetland systems. It is also concluded that designers of new ponds and wetlands to treat stormwater runoff need to apply models like the CRCFE pond or wetland models to gain an understanding of the dominant chemical, biological and microbial processes and to refine the design of ponds and wetlands to meet water quality objectives.

The Las Vegas Wash receives runoff from the Las Vegas Valley which includes the City of Las Vegas, City of North Las Vegas, City of Henderson, and other outlying areas. The annual precipitation in this arid region is approximately four to five inches. In this study, intense spatial and temporal sampling of urban runoff was performed in washes located throughout the Las Vegas Valley. The source of these flows comes primarily from irrigating landscapes, shallow groundwater, and rainfall excess. The purpose of this study is to evaluate the quality of runoff at these locations and to characterize the temporal variability of urban runoff during wet and dry periods. This is critical since there is a lack of comprehensive data for arid regions. Preliminary dry weather runoff results show a clear difference between locations that are influenced by shallow groundwater and locations that have runoff generated only from urban areas. During wet weather, the water quality analysis suggests that the peak concentrations in many constituents (e.g., chemical oxygen demand, Nitrogen, and Phosphorous) occurs about one to two hours prior to the peak water level in the storm channels. This is important when suggesting solutions for minimizing the adverse impacts to water quality. There is also an increase in total suspended solids (TSS) during rainfall events which is important for determining the potential of contaminants being transported downstream. Contaminants can attach to suspended solids and travel to the receiving water body. Lastly, the phase of solids changes from predominately dissolved during dry weather flows to suspended during rainfall event.

Decentralization of stormwater management is necessary in order to achieve sustainable urban storm drainage. This research deals with evaluating long-term performance of infiltration basins by modeling their structural and environmental evolution. System modeling is carried out with a LCA (Life Cycle Assessment) point of view. In the first place, system performance is defined. Performance can not be restricted to a simple technical assessment. Sustainable infiltration basins are also environmentally friendly and socially acceptable with a good economic performance. For this reason, technical, environmental, social and economical aspects are included in this definition. Each performance will be assessed with performance indicators. This requires system modeling and on-field observations. A specific model procedure is built for each performance; it is representative of the structural evolution of the technical system.

The West Branch Paint Creek, a 4-mile long watercourse with a watershed area of 5 square miles located in Ypsilanti, Michigan, is a tributary to one of the last remaining cold water streams in southeast Michigan. The channel has been subjected to an increased frequency of bank full events due to changing agricultural practices and increases in impervious area within the watershed. This has resulted in incising of the channel and bank failure throughout the lower reaches of the watercourse. These conditions have adversely affected the aquatic ecosystem and channel bank vegetation. Riparian landowners have also been affected by increases in the frequency of flooding. As part of this project, funded through the office of the Washtenaw County Drain Commissioner (WCDC), the physical characteristics of the West Branch Paint Creek and its watershed were evaluated to determine a series of permanent measures to mitigate bank erosion and reduce the frequency of flooding in the riparian corridor.

Long-term runoff sampling was carried out at an urban drainage area of 67 ha. Suspended solids in the runoff samples were classified in two fractions; i.e., fine (smaller than 45 μm) and coarse (larger than 45 μm) fraction. The runoff behavior of both fractions was apparently similar, however, quantitatively different pollutant load patterns were observed. Serial rainfall monitoring data was used for the stormwater runoff and quality analyses to investigate the pollutant wash-off process on impervious surfaces such as roofs and roads. The modified wash-off model was applied to discuss the dynamic behavior of fine particles, in combination with the well-calibrated model results of stormwater runoff. The parameter values given for wash-off analysis of total suspended solids did not work well to express the behavior of fine particles. Therefore, a newly calibrated set of parameter values was proposed for the fine fractions and was verified with six events. Highly satisfactory results were obtained with appropriate wash-off coefficients and critical runoff rates for roof and road surfaces. The estimated initial deposits on roofs and roads were correlated with the dry weather periods. A very good correlation was observed in the case of roof deposits, while the initial deposit on roads seemed to be influenced by residual deposit after the antecedent rainfall.

The quality of catchment runoff and of discharges from point and distributed sources including sewer overflows and the impact of these discharges on receiving waters is of increasing concern throughout Australia. When preparing management plans for catchments undergoing land use change three issues of concern are consistently identified by the community. These issues are flooding and drainage, water quality and riparian vegetation. In particular, the impact of urbanisation on water quality is now viewed as a central issue. The XP-AQUALM catchment based water quality modelling package is being increasingly used in Australia to investigate water quality issues when preparing catchment management strategies and plans. The features of the XP-AQUALM packages are overviewed. The use of the package to assess water quality impacts of uncontrolled and controlled development in the urbanising Oxley Creek (54 km²) catchment is presented. The resulting strategy for stormwater pollution control is also outlined. It is concluded that the case study demonstrates the ability of the XP-AQUALM catchment based water quality package to assess the impacts of land use change and point source and non-point source pollution in order to formulate management plans for urbanising catchments that are environmentally sustainable.

Surface water quality in Sedgwick County, Kansas was studied. The data used was historical with continued sampling beyond 1990 and was obtained from these four sources: Kansas Data Access and Support Center (DASC); Better Assessment Science Integrating Point and Non-point Sources (BASINS) modeling program created for the EPA; Sedgwick County & City of Wichita GIS offices; and Kansas Department of Health and Environment (KDHE). GIS (Arc/View) was used to investigate the relationship between pollutants in surface waters at 14 different locations, and, land use, soil type, hydrology, and topography of Sedgwick County. Queries of 24 pollutants against BASINS surface water quality threshold values were done in the GIS to determine degraded surface waters in Sedgwick County. Overall, numerous surface water locations of Sedgwick County were found to be degraded with herbicides, metals, TDS, nutrients, and bacteria at one time or another. For example, the Arkansas River at Derby had a historic total phosphorus maximum of 3.9 mg/L and a mean of 1.1 mg/L, and, a historic mean fecal coliform count of 80077 / 100 mL. The Little Arkansas River at Valley Center had a historic mean fecal coliform count of 5014 / 100 mL. An attempt was made to determine trends between these degraded areas, the pollutants found there, and the surrounding land use and natural features. Herbicide runoff was present in the surface waters of Sedgwick County in levels higher than the federal allowable limits. Nitrate and Phosphorous were present in 40% of the sampling points. The only sample location where there were high levels of most of the herbicides was downstream of a park and open land area.Metals were found in the surface water in Sedgwick County, more than one would expect for a predominantly agricultural area. While Lead and Chromium amounts increased as water passed through the city of Wichita, all other levels of metals decreased. Thus, making it seem that the most likely cause of metals pollution is not commercial and industrial areas, but rather the soils and/or agricultural land uses. Lake Afton was found to be impaired by eutrophication and Cheney Reservoir occasionally had a high concentration of chlorophyll. The Arkansas River was found to be impaired by Total Dissolved Solids. Fecal Coliform pollution seemed to follow normal trends of higher concentrations downstream of larger populated areas, probably from wastewater treatment facility discharges. Many surface water quality improvement activities have taken place over the last 8 years. Improvements in Sedgwick County's surface water are expected as a result of implemented BMPs.

A literature review of the practices used by municipal and local governments to guide land development activities indicates that the controls, including the much-touted smart growth, are not clearly understood and implemented. The literature cite frequently that basic land use practices, such as zoning and site design standards, which direct when, where, and how development will occur, are not being adequately implemented. Data are often lacking to demonstrate the benefits to receiving water quality that could be realized by redevelopment and infill development. This paper examines hypothetical test cases and compares their performance in terms of the performance of minimal controls, extended detention ponds, and LID using bioretention basins. The preliminary findings of the study indicate that innovative storm water best management practices, such as low-impact design using bioretention basins, are better than extended detention basins at delivering higher quality storm water runoff, peak shaving, infiltration, and potentially streambank erosion protection. The study also shows that for the sample case, controlled growth using a mixture of site design plans such as cluster development and the increased use of disconnected impervious areas permits higher density development while maintaining better receiving water quality than other development options.

Continuous monitoring for runoff pollution was conducted at an urban drainage. Stormwater sampling was extensively done for suspended solids (SS) measurement during initial 8 hours of rainfall focussing on first flush phenomena as well as monitoring rainfall and water level in conduit. Reasonable simulation of runoff quantity was obtained using a commercially available distributed model, considering distribution of roof and road as impervious surface. There were good agreements of water level between observation and simulation after adjustment of values on initial loss and single linear reservoir constant for roof and road surfaces. The Sartor and Boyd model was applied to investigate SS wash-off process from impervious surface. Although simulations with the calibrated model parameter values captured general trends of pollutographs, they did not work well to express dynamic pattern. The disagreement seemed to be caused by different behavior of pollutant wash-off from roofs and roads. It was necessary to allocate different wash-off, coefficient for each surface. In addition, the concept of critical runoff rate was also introduced to initiate delayed wash-off from road surface. The modified model for wash-off process explained well dynamic SS runoff pattern in several events with different rainfall characteristics. It was found that distinction of roof and road surfaces for runoff pollution was essential to explain first flush phenomena in cases of weak rainfall.

Sediment washed from construction sites into lakes, rivers, and waterways is a major source of pollution of many lakes and streams that drain small watersheds in which development is occurring. Golf course construction may cause adverse environmental impacts on the surrounding surface and ground-water systems due to the large-scale alteration of the natural conditions. Little Kitten Creek watershed (5.6 square kilometers) is located in the suburb of Manhattan, Kansas, where an 18-hole championship golf course was built by Kansas State University in cooperation with Jim Colbert, the PGA TOUR and the Golf Course Superintendents Association of America (GCSAA). Water quality monitoring indicated a large increase in sediment yield during construction. In order to simulate the change of sediment yields before and during construction, Agricultural Non-Point Source Pollution modeling (AGNPS 98) was conducted through the integrated use of Geographic Information System (GIS). The AnnAGNPS mode was used to simulate the annualized sediment yields. Readily available raster type DEM files were downloaded from the related website and imported into Arcview GIS. A GIS-assisted computer program (TOPAZ with an interface to AGNPS 98) was used to develop terrain-following cells with all necessary hydrologic drainage areas and connections calculated from the DEM files. The input editor was used to initialize, complete, and/or revise the input data. AnnAGNPS was used to generate the outputs. Measured data of sediment yields during construction were used to calibrate the model. AGNPS 98's predictions of sediment yields for two sub-watersheds (SLK and CLV) before and during construction were compared. AGNPS 98 predicted a sediment yield of 240 tons/km²/year under pre-construction conditions, which is typical on rangeland northeastern Kansas. At SLK, where about 20.4% of the total surface area was disturbed, the sediment yield during construction was 470 tons/km²/year, about 2 times the pre-construction amount. At CLV, where about 44% of the total surface area was disturbed, the sediment yield during construction was 915 tons/km²/year, about 4 times the amount of pre-construction. Although its applicability on a daily basis may have limitations, AGNPS 98 does an excellent job at predicting sediment yields on long term basis (monthly or annual).

The US EPA mandates municipalities with populations greater than 100,000 to collect representative water quality data for their urban storm drainage basins and to subscribe to National Pollution Discharge Elimination System (NPDES) permitting procedures. Event Mean Concentrations (EMC) are determined by monitoring selected subcatchments that represent the dominant land uses. EMCs are multiplied by runoff volumes (e.g. from the Rational Method) to get loads for the entire drainage basin. A case study was conducted in watersheds in the Jefferson and Orleans Parishes of Louisiana to help answer the question: "How reliable are load estimates that are extrapolated from land use based monitoring of small subcatchments?" Both basins have separated sewer systems, pumped drainage, high subsidence (possibly causing cross-flow between sanitary and storm systems) and similar land use and topography. However, Orleans is an older area than Jefferson. NPDES monitoring studies were conducted at residential, commercial and industrial subcatchments in both Parishes. The EPA SWMM4.3 model was calibrated for each of the land use based subcatchments in Jefferson Parish and then these land use based parameters were applied to a large drainage basin in the adjacent Orleans Parish to check the transferability of these calibration parameters. Yu measured the load entering Pumping Station No. 4 (PS#4). It was found that hydrologic factors could be transferred using land use characteristics; however, this was not the case for nutrient loads. An explanation for this failure could be the difference in the age of the development of the two study areas. To account for the possible age effect, the SWMM model was recalibrated using the NPDES monitoring data for Orleans Parish and gave better agreement with measured nutrient loads. A rain chemistry study in the study area showed that a large fraction of the nutrient input is attributable to atmospheric deposition.

Considering different flow components of the urban drainage an enormous variety in quantities of important quality parameters (COD, ammonia, copper and lead) can be observed. This is due to different influence factors which are divided in objectives of the study and in influence factors concerning the investigation methodology and site specific conditions. An important chemical quality parameter in urban drainage is the COD. One advantage is the relatively easy determination which leads to a good data basis and provides a good transferability of data in different flow types. A disadvantage of COD and other sum parameters used in urban drainage monitoring programs is, that it is only to a minor extent possible to formulate statements about ecotoxic potential for receiving waters. Therefore recently alternative to the sum parameters other pollutant parameters have been discussed and are presented in this paper. Beside standard (waste water related) chemical organic parameters new lately discussed organic pollutants are listed and a first evaluation of an environmental potential is given. Groups of substances, like pesticides and nitrophenols, which are often found in surface runoff were taken into account as well as pollutants originating from dry weather flow, e.g. human medicines. Especially in the group of the endocrine disruptors certain substances (e.g. nonylphenol) are already listed in several environmental guidelines and also have a potential of occurrence in different flow types of the urban drainage system.

This paper aims to present the flood warning model — MOPEH which has been used in the São Paulo Flood Warning System, a real time system for rain and flood forecasting. This model uses empirical concepts and establishes hydrological conditions for many areas subject to floods due to different drainage problems. São Paulo has more than 500 flood critical areas and this empirical approach was necessary considering the lack of information and the complexity of the system. Some recent results showed the good performance of the model.

Flooding in urban areas is an inevitable problem for many cities in Asia, in the year 2000 Bombay, India and Phnom Penh, Cambodia had serious problems related to urban flooding. In Bangladesh, Dhaka has faced water logging for the last years. Even a smaller rain events may cause that parts of Dhaka have water standing in the streets. The situation was highlighted in September, 1996 when residences experienced ankle to knee-deep water on the streets. Daily activities in parts of the city were almost paralysed. Heavy traffic jams occurred due to stagnant water on the streets. With the advances in computer technology, many cities manage local and small flood problems with the computer-based solutions. But at present, a few studies of urban flooding with the simulation of both a surcharged pipe network and extensive flooding on the surface are available. In this study, modelling of urban flooding with the interaction between the pipe system and the surface flooding, including a description of the interaction between street and pipe network are carried out.

The Estero-Imperial-Cocohatchee (EIC) watershed located along the southwest coast of Florida experienced a number of severe floods in 1992 and in 1995. Recent studies have revealed that parts of the watershed hydrology were poorly understood. The agricultural and urban development in the area has changed the original flow ways and due to insufficient conveyance capacity and under sized hydraulic structures, the wetland system now acts as a large retention basin, which after long periods of rainfall overflows and inundates the coastal areas. The intensity and impacts of the floods has called for measures to be taken to mitigate future flood events and an integrated hydrological/hydraulic model was developed for the watershed with the main objective of simulating flooding. The model is based on an integrated, physically distributed hydrological model (MIKE SHE) including overland flow, unsaturated zone flow and groundwater flow dynamically coupled to a river hydraulics model (MIKE 11). Compared to traditional flood models, the integrated approach provides a physical representation of the flow processes as opposed to the lumped rainfall-runoff model approach. In the EIC watershed with slowly increasing water levels, the response of the entire hydrological system will affect the critical flood levels and the integrated model thus offers better possibilities for analyzing the causes and effects with respect to flood management. The model has been calibrated against field data including river stages and groundwater levels for a period from 1990–95 with specific focus on matching river/surface water levels during the flooding in 1995. The GIS based flood mapping capabilities of MIKE11 (MIKE11 GIS) is utilized to map flood extents and flood duration. In the paper the model tools, model approach, study objectives and simulation results, are presented. Although the model was primarily developed for flooding purposes on event basis, the modeling system is versatile and may be used for continuous simulation, investigating water resources aspects and environmental impact assessment on the wetlands. Furthermore the model will form a basis for setting up a surface water quality model, which will assist local governments and the South Florida Water Management district (SFWMD) in assessing water quality impacts of various water management scenarios.

Flood risk in urban areas invoked by urbanization and climate change poses a unique problem for the individual citizen, the community, and the insurance companies and requires a flexible information system, which is configurable into an effective management system. This paper discusses a proactive approach to urban drainage management for dealing with flood risk. The approach and modelling system are based on a conceptual framework that is to be developed within the European research project RISURSIM — Risk Management of Urban Drainage Systems.

This paper highlights current issues in the development and submittal of model results for use in U.S. Federal Emergency Management Agency floodplain mapping studies. Case studies for three watersheds in the U.S. are also presented, including the Anclote River in Florida, the Pudding River in Oregon, and the Napa River in California. For the Anclote River watershed, peak flood stages were predicted using the U.S. Environmental Protection Agency's StormWater Management Model (SWMM) and MIKE SWMM, a pre/post processor from DHI Inc. MIKE 11 GIS was used to generate a three-dimensional grid representation of the watershed terrain and to determine the resulting floodplain inundation. For the Napa River and the Little Pudding River watersheds, peak flood stages were predicted using MIKE 11, a one-dimensional hydrodynamic program for modeling riverine systems. MIKE 11 GIS was used in these studies for floodplain mapping and model development to evaluate channel and floodplain maintenance and management strategies.

A stormwater master plan study was completed for the Blue River Watershed in southeast Johnson County, Kansas. The study focused on developing a stormwater management plan that addresses existing problems and provides guidance for future growth. Traditional US Army Corps of Engineers Hydrologic Engineering Center (HEC) models were applied with innovative mapping techniques using HEC-GeoRAS with ArcInfo and ArcView. These techniques allowed for the efficient development of spatially referenced watershed models covering 207 square kilometers (80 mi²). The GIS databases were efficiently used to develop input parameters for the HEC-1 and HEC-RAS models. ArcInfo coverages of land use, soil type, slope, elevation, and Manning n values were used to delineate 368 subareas and determine basin flow path, slope, and drainage area. Literature values were used in conjunction with GIS data, to develop values for key subarea input parameters such as subarea curve number and time of concentration. Both lag time and curve number for each subarea were calculated with a customized spreadsheet that used spatial data exported from ArcInfo. For each subarea the customized spreadsheet calculated a composite curve number based on land use and soil type while the lag time was determined by dividing the flow path into sheet flow, shallow concentrated flow, secondary channel, and primary channel. For each subarea, the longest flow path to the subarea outlet was automated in ArcInfo. The HEC-1 and HEC-RAS models were linked with the Modified Puls routing method which requires a storage volume verses outflow table for each routed reach. HEC-RAS was used to generate storage volume verses outflow for the HEC-1 model. The iterations between the two models were automated with a customized spreadsheet that linked the HEC-1 and HEC-RAS models. The integration of the HEC-1 and HEC-RAS resulted in a high quality model that accurately represented the hydrology as well as the hydraulics of the watershed. Channel and small lake storage was accurately accounted for with this routing approach. Over 182 kilometers (115 miles) of stream channel was modeled through 368 subareas and 164 hydraulic structures. A total of 1,309 cross-sections, 43 inline weir structures, 90 culverts, and 31 bridges were coded into the HEC-RAS model to accurately represent the hydraulic characteristics of the watershed. HEC-GeoRAS in conjunction with ArcInfo was used to prepare a geo-referenced based HEC-RAS model with stream network, cross-section geometry, and reach lengths. HEC-GeoRAS was used to generate HEC-RAS input data and process HEC-RAS output water profiles into spatially accurate watershed floodplains. Water surface profiles and cross-sections modeled in HEC-RAS were imported into ArcInfo using the HEC-GeoRAS software providing a cost-effective procedure to geographically link the model results to the GIS. Floodplain polygons were then overlaid onto existing GIS coverages in ArcView to provide an effective planning and decision making tool. In addition, ArcView scripts were effectively used to display, quarry, and export survey and cross-section profile data for use in HEC-RAS.

Traditionally the rainfall data measured at one or more rain gauges is used to define the input data in urban runoff simulations. In practical simulation tasks the areal precipitation is estimated by applying the point measured data on the area around this location assuming uniform spatial distribution. If data from more than one rain gauge is available, a mathematical interpolation can be used to take into account the spatial distribution. Analysing spatial intensity distributions measured with weather radar systems show that both attempts based on point measurements are only weak assumptions of the real spatial distributions. The paper describes a study where different approaches for modelling the spatial distribution were carried out in order to simulate urban runoff in a large urban drainage system.

Lincoln Creek has a long history of flooding and other beneficial use impairments. This continuous tributary of the Milwaukee River is approximately 14.5 kilometers (9 miles) in length and drains a 5370 hectares (20.7 square miles) entirely urbanized subwatershed of the Milwaukee River. Land use in the subwatershed is a mixture of commercial, industrial, and residential properties. More than 600 structures were at risk of direct flooding in the 100-year event. Lincoln Creek has over 200 storm sewer outfalls that discharge to the stream. The entire length of the stream has been channelized to accommodate floodwaters from storm water runoff, which has been ineffective in controlling flooding. Lincoln Creek flows over six constructed sills and drop spillways. Approximately 25 percent of the channel is lined with concrete to assist in draining runoff volumes. Masonry and rock walls line about 10 percent of the channel. The remainder of the creek is in very poor hydraulic condition. The Milwaukee Metropolitan Sewerage District (MMSD) and the Wisconsin Department of Natural Resources (WDNR) are implementing flood controls and other resource enhancements for Lincoln Creek to reduce flooding problems, improve aquatic and terrestrial habitat, and enhance recreation. A Flood Control Management Plan was completed for the MMSD that identified specific improvements to Lincoln Creek using a watershed approach. The estimated cost of the project at that time was $70 million. Ten individual reaches on Lincoln Creek were identified and specific improvements were suggested for each reach. A team led by Camp Dresser & McKee Inc. (CDM) recently completed the design of improvements for all 10 reaches.

The New Orleans metropolitan area is at or below sea level. Which means that its storm water drainage from an average annual rainfall of approximately 60 inches must be pumped (or lifted) for discharge. Most metropolitan areas around the world use gravity flow for discharge of urban drainage. Because of the elevation of this area, hurricanes and the flooding they can cause are of great concern to water resources planners. This paper presents the simulation of flooding that would have been caused by the transposition of hurricanes Fran and Danny. Hurricane Fran was a category 3 hurricane that made landfall along the southeastern North Carolina coast on September 5, 1996. Hurricane Danny was a slow moving Category 1 storm that made landfall in the vicinity of Mobile Bay in mid July 1997. The simulation was based on the flooding that would be caused by the combination of hurricane-associated rainfall and diminished pumping capacity due to storm tides. Neither of these storms would have produced a storm surge that would have overtopped the hurricane protection levees. The study area was the watershed in New Orleans drained by Drainage Pump Station No. 4 (DPS #4). DPS #4 drains some 4,800 acres by discharging the storm water runoff into Lake Pontchartrain. In the past one hundred years, the watershed has been impacted by 14 hurricanes that caused significant damage and by 28 other hurricanes and tropical storms that have caused minor damage. The discharge head(s) for the analysis of DPS #4 was based upon hurricane storm surge data from the National Weather Service's Sea, Lake, and Overland Surges from Hurricanes (SLOSH) model. The flooding of the watershed was estimated by use of the U.S. EPA Storm Water Management Model (SWMM) for routing the precipitation. The results of the simulation showed that the lower category, slow moving storm, Danny, would produced significantly greater flooding.

The island nation of Taiwan was hit hard and caused tremendous damages and casualties by Typhoon Herb storming from July 1 through August 1 of 1996. Flooding in densely populated urban areas of Taipei County, situated in the northern Taiwan, drew special attentions and concerns. This study employed simulation models to reconstruct the 1996 flooding event occurred in Tu-Cheng, Pan-Chiao, Chung-Ho and Yung-Ho areas of Taipei County, and examined different situations which would have alleviated the extent of flooding. These results will be helpful in planning non-structural measures for flood protection and flood mitigation in the future. This study employed the HEC-1 model to simulate urban drainage runoff discharges, the Charima model to simulate river stages in the Tan-Shui River system into which the urban drainage discharges, the 2-D overland flow model developed by the Hydrotech Research Institute of National Taiwan University to simulate urban flooding phenomena and back flows through floodgates at Kuang-Fu and Tu-Cheng. The results of the flooding simulation indicate that back flows through the floodgates amount to 19.8% of the total water volume (including back flows) in the drainage basin studied while pumped-out water only amounts to 12.2% as a result of improper gate operations. In addition, influential effects of flood-releasing discharges from upstream reservoirs of Shih-Men and Fei-Tsuei and the piled-up garbage mountain in the floodplain on the flood event were investigated. These results are very helpful in flood management and operations in the future.

A numerical simulation model of an idealized drainage basin of 1 km² was developed. The Kinematic Wave method was used to formulate the rainfall-runoff process. The simulation results show that for basins with moderate to high rate of infiltration loss, the critical rainfall duration associated with the maximum peak discharge may be shorter than the time of concentration of the basin, t_c. This implies a storm with a spatial coverage of only part of the basin may generate the maximum peak discharge. For basins with low or negligible infiltration loss, the critical rainfall duration approaches the value of t_c.

Side weirs are flow-diversion and flood control devices widely used in urban drainage, land drainage, and irrigation systems. The De Marchi equation is widely used for computation of side-weir flow. However, in this equation the variation of flow depth with longitudinal distance is expressed in implicit form. Many investigators focused on the relation between the coefficient of discharge and the upstream Froude number and other flow and/or geometric parameters. Presented herein is a general design method for side weir. Also presented is a concept that for subcritical flows over a side weir the flow characteristics are governed by downstream flow conditions and the geometric parameters of the channel and the side weir. The study is based on De Marchi's equation. The analysis is conducted through non-dimensionization of the parameters in the equation. The study results indicate that the side weir flow is a function of the downstream flow depth and Froude number, and the ratios of weir height to downstream depth and channel width to the product of weir length and discharge coefficient. Charts were produced and procedures were given for engineering design.

The US Army Corps of Engineers, Engineer Research and Development Center, has selected the two-dimensional, square-grid, physics-based hydrologic model CASC2D as its premier two-dimensional watershed model. CASC2D has been applied in many non-urban watersheds. This paper focuses on calibration and verification results using multiple rainfall events in an urbanized watershed on the campus of die University of Connecticut in Storrs, Connecticut. Studies underway on the Vermillion River watershed in Lafayette, Louisiana are also discussed. Results of recent research applying CASC2D on an urbanized catchment indicate that the model is superior to standard-practice lumped-conceptual rainfall-runoff models because of its distributed hydrodynamic formulation. This result has important implications for the development of advanced techniques for runoff modeling and flood prediction in urbanized areas.

Most of numerical procedures used in hydraulic design are able to compute free-surface or pressure flow. But mixed flow is a problem that appears in some hydraulic structures and different procedures have been developped to deal with this problem. The problem that has usually focussed the atention of many researchers is the one surge front, moving upstream or downstream. Different methods have been tested to reproduce the pressure field values, surge celerity, and so on. But another problem that can appear in a sewer network is the transition from free-surface to pressure flow at both ends of a pipe. With the purpose to analyse this phenomenon, an experimental facility was set-up to carry different tests in order to gain knowledge about the problem (qualitative objectives) and reproduce by means of numerical modelling the observed results. Different numerical approaches were tested and finally, an approach based on the full unsteady flow equations, both for pressure and free-surface flow, considering the MOC (Method of Characteristics) provided the best results.

This paper discusses the representation of long term, highly attenuated runoff in collection systems, focusing on rainfall dependent groundwater infiltration. Traditionally, the contribution to surface water systems, or the surface water element of a combined sewer system, has been represented by a combination of overland runoff and a constant baseflow. It is recognised that there are types of contribution to urban drainage systems, such as stream ingress or the contribution from large permeable areas, which do not fall easily into either the runoff or baseflow categories. In addition, the increasing prevalence of Sustainable Urban Development (SUD) structures is promoting a need to represent highly attenuated groundwater contribution. Until now, representation of this type of contribution has been attempted with varying degrees of success, by attenuating the direct overland runoff. It has been identified that new mechanisms are required. A feature of model verification and application to design, at least in the UK, is the use of long duration rainfall data. In the past a model would be verified against three storms recorded during a short term flow survey and then applied to design using standard ‘normal distribution' storms. It has now been identified that there is a need to verify against one continuous flow survey data set and then to apply to design using several years of recorded data or standard annual time series rainfall events. An important element of the groundwater infiltration module is in the application to long term data. This paper presents existing overland runoff models and discusses their limitations in the representation of groundwater contribution. The groundwater infiltration model as implemented within InfoWorks CS is then detailed. The potential for application of the groundwater model to sustainable urban development structures is then briefly described. It is recognised that the inclusion of the groundwater infiltration model in InfoWorks CS is a recent innovation, the paper therefore concludes with a recommendations of further work to enhance the model and improve its application in engineering design studies.

An exfiltration trench splits an inflow hydrograph into two components. The outflow hydrograph is attenuated by the storage in the voids of the clear stone fill and the balance of the inflow is transmitted to the ground water through the pervious base and walls of the trench. Such a facility can provide an innovative approach to stormwater management with several advantages over conventional end-of-pipe solutions. Details of the trench design are described, some points concerning construction procedure are reviewed and recommendations for post-construction monitoring are presented. Experimental procedures for testing the facility are also described.

For the past 40 years researchers have been questioning Horton's model of infiltration excess for surface runoff generation, especially in humid regions. Researchers have been studying the concept that storm runoff is often generated by only a portion of the watershed and that runoff occurs first around streams or areas where the water table and soil depth readily cause surface saturation. This study evaluated the effects of inflection angle (a measure of contour curvature; defined as the angle a contour line makes with itself when crossing a drainage-way) and the relative location (the position of a land area in relation to a drainage-way) on the rate and volume of runoff. In this study, zero-order watersheds were separated into four distinct location types, which were based on their relative location to larger order drainage-ways. Traditionally-computed curve numbers for hypothetical watersheds (land areas without distinct natural drainage-ways or outlets) were shown in this study to be typically from 10 to 40 curve number values too high when used in the NRCS CN runoff model. Minor draws off of the main watershed drainage-way were shown to have both runoff curve numbers and peak runoff rates that were highly dependent on the average watershed inflection angle. NRCS runoff model error estimates were shown to be distinctly correlated (r² = 0.64) to the inflection angle. Watersheds that were located upslope from the direct extension of a stream appeared to be dependent on the relative distance of the watershed outlet to an active stream.

Bed shear stress is a relevant factor to determine erosion and deposition of sewer sediments. Experiments have been undertaken in a laboratory flume and in two sewer trunks of "Le Marais", in Paris, to compare different methods to calculate local bed shear stress (based either on the log-law of the wall or on the Reynolds shear stress distribution) and to calculate mean bed shear stress (with τ_m = ρgR_hJ). Measurements were performed with a micro-propeller and an Acoustic Doppler Velocimeter during dry weather flow. Values of local bed shear stress calculated with the different methods were comparable but measurements with ADV revealed a great dispersion of shear stress close to the bed. It was found that in sewers were b/h < 5 the spatial variability of local shear stress was very low, and so local bed shear stress in mid-flow was equivalent to mean bed shear stress. The relation τ_m = ρgR_hJ_c was found to be applied with the energy slope because data plotted on maps are very imprecise and lead to big errors.

Sewer flow quality modelling software uses equations derived from uniform granular sediment studies. Actual sewer sediments are mixtures of organic and granular material and interactions between these fractions may account for the poor performance of current models. Laboratory experiments were carried out using surrogate sewer sediment mixtures to examine interactions between the fractions. Impaction of the bed surface by saltating granular particles increased the erosion of fine grained organic sediments. Changes in the composition of the bed surface over the duration of a test resulted in change in the availability of fine grained sediment. An existing model was further developed to attempt to account for these mechanisms, using a empirically based correction factor.

The presence of a granular layer in the inverts of pipes has been reported in a number of sewer systems in France, the United Kingdom, Belgium, and Germany. Given the typical range of flow conditions found in combined sewers and typical particle characteristics of inorganic in-sewer solids the mode of transport of many of these mineral sediments may normally be considered as traditional bed load transport. Current commercial software packages modelling the erosion of sediments and the transport of pollutants in sewer pipes still utilise well-known, or modified versions of, bed load or total load transport equations that were derived for transport of non-cohesive material in alluvial streams. In this paper the predictive performances of the equation of Ackers and White (originally developed for the transport of sediment in rivers) and the equation of May (a formula derived from laboratory pipe experiments) are examined against two separate data sets. The first data set is obtained from laboratory erosion experiments on sewer sediment at the Hydraulics Laboratory of the University of Ghent. This sediment was collected from the experimental catchment "Le Marais" in Paris. A second data set is derived from bed load transport rate measurements recorded in the inlet pipe at the National CSO test facility in Wigan, United Kingdom. These formulae were selected for study because of their widespread use in the prediction of in-sewer sediment transport both in commercial software and in the latest UK design guidance for new sewers. It was seen that neither formula performed well. It was seen that each of the traditional formulae have significant difficulties in predicting the erosion thresholds and transport rates for individual grain size fractions from such non-uniformly sized in-sewer deposits.

The movement of large organic solids in sewers has not been studied until recently. New European Regulations require that these solids are controlled so as not to be seen in the environment once escaped from combined sewer overflows. Hence there is now a greater imperative to understand how these solids move and are discharged from sewer networks. A study over the last 5 years has investigated all aspects of the origins and movement of these solids in real sewers in Scotland. Computational modelling is essential to predict the behaviour of these ‘gross' solids in sewer networks. A stochastic model has been developed to simulate their movement by predicting the probability of transition of any given solid among three different states (suspension, near bed and immobile state). This new modelling approach has been developed using continuous time Markov Chain and Kolmogorov system equations. Simulation has been carried out using the InfoWorks hydraulic model and tool box MATLAB/SIMULINK. This paper presents an application of the stochastic approach to simulate the gross solids movement in a sewer network using this probability based approach.

The paper describes a comprehensive feasibility study carried out to resolve a principal conceptual dilemma concerning the Ischia's (Italy) wastewater problem: Should the wastewater be treated and disposed on the island or should the concept of remote wastewater treatment on the Italian mainland be applied? The study was based on modelling of the impact from the continuous wastewater emissions and combined sewer overflows (CSO) on the water quality in the sea surrounding the island. The simulations were carried out by the application of integrated generic simulation tool — Integrated Catchment Simulator (ICS) — for addressing the environmental impact from urban catchments on the receiving waters. The pollution loads from the present and future sewer network was simulated by use of a deterministic pollution transport model for urban drainage networks (MOUSE TRAP). This provided a detailed information on hydraulic and pollution loads at any point of interest in the island's sewer network. The highly variable pollution emissions from the wastewater system, together with the recipient dynamics (sea currents, wind, and tides) and the water quality processes in the sea (dispersion and fate of Coliform bacteria and dispersion and settlement of suspended solids) create a very dynamic situation concerning the impacts in the surrounding sea. Description of these effects required a 2-D hydrodynamic and water quality model (MIKE 21) to be setup for the waters around Ischia. ICS was applied in order to perform easy and consistent analyses of the entire system, facilitating the transfer of the urban pollution loads to the sea. Simulations were carried out for the present situation and for alleviation schemes, including local wastewater treatment in the island's municipalities, centralised treatment in one large WWTP on Ischia, and with the transport of wastewater to mainland.

During the last years many countries have begun with the connection of existing software packages simulating sewer systems and processes in wastewater treatment plants in order to compare different scenarios in the overall urban wastewater system. Because the original software packages were developed for other purposes, the question still remains whether they can be applied for integrated modelling as well. Therefore the aim of the study focuses on possible problems brought up by linking modules of the sewer system and the wastewater treatment plant. The study focuses on modules with different sets of state variables, where conversion factors are necessary at the interface. Sensitivity analysis has been carried out in order to find out which state variables should be determined quite exactly and which are not very sensitive and can therefore be estimated more roughly. It can be seen that in the described catchment and at the WWTP the consideration of different conversion factors for dry and wet weather has no great effect on the WWTP effluent. The sensitivity analysis revealed that the parameter set of the buildup/washoff model as part of the sewer model seems to be of high importance for the WWTP effluent. Based on the study the conclusion may be drawn that it is necessary to distinguish suspended and dissolved COD fractions in the sewer model.

This paper presents the results of a recently completed Ph.D. study by Willems, in which a methodology has been worked out to model the current and future states of surface waters considering all sources of water and pollution in the hydrographic catchment in an integrated/holistic way. Both aspects of water quantity and quality have been considered. In this way, the methodology is applicable for river management, both from the perspective of flooding problems and water pollution problems. These problems have internationally been the subject of increasing concern, especially during the last two decades. To solve the problems, water management stategies have been worked out by authorities all over the world. Each of these strategies is often set up only from the perspective of one water or pollution source. In this way, dikes have been built to limit of risk of river flooding and urban drainage systems have been improved to solve urban flooding problems. At the same time, waste water treatment plants (WWTP's) have been built to reduce the cumulative domestic emission loads to the receiving rivers and improved overflow structures, storage and sedimentation basins have been constructed to limit the occasional overflow emissions. Also emission standards have been set up in combination with the latter two types of investments and additional emission standards have been defined for industrial pollution sources and fertilization standards to limit the agricultural pollution. Although all these measures and investments have improved the surface waters' states already significantly, the improvements often did not reach the expected level and/or the measures caused additional problems. In Belgium, this became undeniably clear the last few years by the increased number of floodings in both rural and urban areas. The increasing number of sewer systems and connecting areas and the isolation of natural flooding areas from rivers by dikes partly explain this increase. By the growing insight in the global water system, it has becoming clear that the efficiency of measures and investments based on the analysis of only one water or pollution source is often not optimal. The different water subsystems in a hydrographic catchment have to be considered in an integrated way. This means for instance that immissions (the impact of the emissions on the receiving river) have to be analysed instead of emissions (the discharges and pollution loads from urban drainage systems, industry and agricultural activities). The mathematical modelling of immissions by the holistic modelling of all subsystems causing an increase of water and pollution in the receiving rivers was the subject of the study.

The Paper outlines the changes in modelling approaches arising from scientific research and decision support tools. Models are emerging as both a research tool (where applied to complex large natural resource systems) and as a decision support tool, enabling the application of complex information to management decisions in a simple and accessible manner. The Paper lists the major drivers of change and the modelling response in terms of strategies, partnership and packaging. The Paper draws on three case studies to illustrate these changes.

Protecting the natural environment while fostering local development is one of the main challenges that engineers face today. New species are added to the federally threatened and endangered species lists as their habitats are destroyed by residential and commercial development. The bog turtle (clemmys muhlenbergii) is threatened by new development that damages not only its wetland habitat, but the upland areas that drain into the wetlands. The bog turtle was listed as federally threatened in 1997. Protection of bog turtle habitat is a concern to many state and local governments because its habitat is often located near prime land slated for commercial, residential, and transportation related development. Until recently, there was little information available about the water sources that feed these wetland habitats. The Maryland State Highway Administration, in conjunction with other Maryland and federal agencies, has formed a biological assessment team in order to develop a protection plan for one such bog turtle habitat by investigating the hydrology associated with bog turtle habitats. The primary objectives of the hydrologic investigations are to: 1) identify and monitor the major hydrologic inputs/outputs; 2) identify and monitor the major recharge/discharge zones; 3) develop recommendations for future development and stormwater management; and 4) use the results of this analyses as a guideline for developing future protection plans for the bog turtle. Although there has been much recent research on the habitat, genetics, and movements of bog turtles, this is the first large scale study of the hydrology of the bog turtle habitat. An extensive field monitoring network was installed throughout the 8.8 km² watershed. The field data was used to set up and calibrate groundwater and surface water flow models. Surface and groundwater models were developed for the watershed and the results were used to develop a water budget for the bog turtle habitats. The model results provided state and local planners with estimates of seasonal average groundwater table elevations and streamflows. A series of proposed condition models were developed to predict changes in seasonal average levels in order to assess the impacts of the development on the local groundwater table and the wetlands water supply. This information will be used to develop the first major bog turtle preserve in the state of Maryland. The information should be valuable to other Mid-Atlantic states who are facing similar declines in their bog turtle habitats and populations.

The City of Columbus, OH, Division of Sewerage and Drainage (DOSD) intends to construct a sanitary trunk sewer in the vicinity of the Jackson Pike WWTP to replace an existing segment of the Scioto Main Sanitary Trunk Sewer. The new Scioto Main Trunk Sewer Replacement (SMTR) will eliminate all overflow structures at the Renick Run area as part of the West Columbus Local Protection Project (WCLPP) administered by the U.S. Army Corps of Engineers. Due to the complex hydraulics in the vicinity of the Jackson Pike WWTP a detailed dynamic model was developed to assess the sewer system performance under different operating conditions. The model used the sewershed approach to maximize system storage without affecting the current services and to satisfy the regulatory conditions to operate the sewer system. The developed model investigated sewer system performance using 6-month, 1-, 2-, 5-, 25- and 100-yr storm events to optimize the operation of the collection system. Initial model results indicated potential overflow at the Rhodes Park overflow using the 5-yr storm and above. After a series of model runs to optimize system operation and boundary conditions potential overflow at Rhodes Park is anticipated for storms above the 25-yr storm. The sewershed approach was used to satisfy the following objectives: 1) Provide primary and secondary treatment for combined and separate tributary flow due to dry weather flow and wet weather flow up to the 6-month storm, 2) Provide primary and secondary treatment at Jackson Pike WWTP for separate flow up to the 25-yr storm, 3) Provide primary treatment at Whittier Street Storm Tanks for combined flow due to the 1-yr storm frequency and above, 4) Maximize bypass flow from Jackson Pike WWTP to Southerly WWTP through the flow control structure to avoid backup in the Scioto Main Replacement Sewer, 5) Maximize in-line storage in the 156-inch Interconnector Sanitary Trunk Sewer, 6) Maximize pumping rates at the Interconnector Pump Station to its limit, 7) Regulate tributary flow from Scioto Main Trunk Sewer using the Hydrovex regulator to 60 MGD, and 8) Provide in-line storage upstream the Hydrovex regulator to a maximum elevation of 720.00 feet to avoid overflow at Rhodes Park overflow. The sewershed approach and the developed model allowed the determination of the optimum operating conditions by maximizing in-line storage while meeting all required constraints (e.g., overflow, pumping capacity, control flow devices, sewer capacity).

This paper will present results of a study about the effects of an increased combined sewage influent to wastewater treatment plant (WWTP). Within this project, the effects on the total emissions from a sewer network and a sequencing batch reactor (SBR) were calculated by using simulation models. The results were used for an optimized design of the treatment plant. Furthermore, a short outlook on the objective to realize an integrated operation of sewer system and an existing SBR, will be given.

The water quality simulation model FGSM of the German Association for Water, Wastewater and Waste (ATV/DVWK) has been set up to simulate major water quality parameters of a small urban stream. The stream is receiving the treated waste water of two treatment plants and combined sewage overflow (CSO) of a total of four storm water storage tanks with overflow (SWST). First the model was calibrated for dry weather conditions. Based on the calibrated model a typical CSO event was simulated. The impacts of CSO at one SWST was compared to CSO at all four tanks and dry weather conditions. A particular focus was set on accumulative effects. The model was able to simulate dry weather conditions well. The results for the CSO events appear plausible for both hydraulic and pollution impacts. The model was able to simulate high peak flows without damping. It was shown that events, which would have little impact when taking place a single site, have a profound impact when occurring at numerous sites in the stream system due to accumulative effects. Therefore the model is suitable for evaluating the acute effects of CSO events. The conclusion of the study is that the design and operation of storm water tanks should take into account the accumulative effects of CSO of further tanks in the stream system.

Snohomish County has developed an integrated approach to evaluate the impacts of current and future development upon surface water resources within urbanizing areas of the county. This approach is intended to provide decision makers with critical information related to land use alternatives, the cumulative impacts of development, the effectiveness of regulatory and capital improvement alternatives, and priorities for the spending of limited capital funds. Through this approach, the County is better able to define their level of service for surface water facilities and to move towards clear compliance with the Growth Management Act of Washington State. This approach has gradually evolved since Snohomish County first created the Master Drainage Plan (MDP) Program in 1998. It has been applied to some of the urbanizing areas of the county to evaluate the major streams and drainage systems in these areas. Major river systems are evaluated under a separate program at Snohomish County.

The City of Greensboro, North Carolina, USA is developing and implementing a proactive Municipal Stormwater and Watershed Management Program that includes unique Geographic Information System (GIS) database and application tools for a high level of GIS integration for urban stormwater management. For example, customized GIS technology allows the City to prioritize stormwater infrastructure and drainage system maintenance, assist with implementation and tracking of the municipal NPDES stormwater discharge permit, track water quality data and improvements, enhance the local floodplain management program, and facilitate multi-objective, watershed-based modeling and urban stormwater management master planning. Advanced technology coupled with dedicated staff and stakeholder involvement, clearly defined goals, and a watershed-based focus are allowing the City to develop optimal plans for urban stormwater management, watershed protection, and watershed restoration. This paper provides an overview of the City's Stormwater Management Program with a focus on an integration of interactive hydrologic, hydraulic, and water quality models with a robust GIS database to create a tool that provides a sound technical basis for informed watershed-based management and restoration decisions by local officials. A foundational aspect of the City's program includes ongoing stakeholder and community input as staff attempt to balance science and engineering with local values related to natural resources and urban watershed management. Once legal requirements are satisfied for designated water resources, community values provide an ultimate measure of success for local stormwater and watershed management programs. The current work is referred to as the "Dynamic Watershed Management" (DWM) project and represents initial steps at the local level towards development of a holistic and sustainable urban watershed management program. Due to the flexibility of the GIS tools and databases, the DWM program could be expanded in the future to include comprehensive urban water resources management such as water supply system optimization, sanitary sewer system inventory and modeling, and advanced water quality modeling to support Total Maximum Daily Load (TMDL) programs, pollutant load allocations, discharge limits, and pollutant trading programs for local watersheds. The GIS-based applications and modeling tools are therefore helping the City of Greensboro effectively meet today's stormwater management requirements while also preparing for the comprehensive watershed-based management challenges and opportunities of the future.

Camp Dresser & McKee is currently developing a stormwater master plan for the Lake Jesup Basin in Seminole County, Florida. The overall objective of this stormwater management plan is to provide an affordable level of service for flood and water quality control for the systems the County is responsible for operating. The information presented in the stormwater management plan can be integrated into the efforts being undertaken by the St. Johns River Water Management District in developing the Lake Jesup Restoration Program. The first phase of this study is a data collection effort which relies heavily on GIS applications for mapping, management and characterization of essential data required for a complete hydrologic and hydraulic analysis of the basin as well as a public participation program. An engineering analysis will then be performed, which is comprised of stormwater model development, water quality investigations, alternative flood mitigation solutions with public input, and the development of an overall stormwater management master plan for the basin. Results presented in the master plan can be combined with other information pertaining to sensitive issues such as water quality, circulation patterns, sediments, and land management practices of Lake Jesup to form a holistic approach to stormwater management.

Low impact development (LID) achieves stormwater management controls by fundamentally changing conventional site design to create an environmentally functional landscape that mimics natural watershed hydrologic functions (discharge, frequency, recharge and volume). This is accomplished in four ways. First, minimizing impacts to the extent practicable by reducing imperviousness, conserving natural resources /ecosystems, maintaining natural drainage courses, reducing use of pipes and minimizing clearing and grading. Secondly, recreate detention and retention storage dispersed throughout a site with the use of open swales, flatter slopes, rain gardens (bioretention) and rain barrels. Thirdly, maintain predevelopment time of concentration by strategically routing flows to maintain travel time. Fourthly, encourage property owners to use effective pollution prevention measures and to maintain management measures. The purpose of this paper is to introduce and describe Low-Impact Development (LID) hydrologic analysis and computational procedures used to determine LID stormwater management requirements. The hydrologic analysis presented is based on the Soil Conservation Service (SCS) TR-55 hydrologic model. LID development site planning and integrated management practices (IMPs) are defined and categorized into components of Lid objectives. Computational procedures for determining IMP requirements are demonstrated through design examples.

Management of stormwater at an operating industrial facility can be difficult and challenging considering the regulations for pollution prevention. However, managing stormwater at a closed industrial facility undergoing a comprehensive cleanup, restoration, and redevelopment program is an even more challenging task. For some closed industrial sites, stormwater runoff may be "impacted" by site contaminants and, as such, may require treatment prior to discharge. For those parties responsible for the site remediation, the management and segregation of "impacted" and "clean" runoff becomes a major design, construction, and operations issue. Water management at such sites typically involves three components: (i) impacted water; (ii) sediment-laden clean runoff; and (iii) post-remediation (clean) runoff. Because impacted water generated must be collected and treated prior to discharge, the need for water treatment is anticipated to continue until such time when the remedial activities are complete and water is no longer considered impacted. As such, developing a site-wide water management strategy is an essential component of remediation. This paper focuses on the design of the impacted water management system and the development of the overall site-wide water management strategy for the remediation of the Avtex Fibers Superfund Site in Front Royal, Virginia. At this site, the remediation schedule and the water management strategy are interdependent. In order to understand the interdependences at the site, the overall remediation project was modeled with respect to water management issues. The model incorporated the following time dependent items: (i) the remediation schedule; (ii) the potential for impacted water generation; (iii) the available impacted water storage capacity; and (iv) the rate of impacted water treatment. The modeling effort resulted in the development of an overall site-wide water management strategy that is closely related to the remediation schedule.

Among the most urgent problems of environmental monitoring, note should be taken of simulation of pollutants transfer in water objects. This problem is especially urgent for urban conditions as there exist lost of pollution sources there. It solution is of great ecological and economical significance allowing to investigate the influence of different pollution sources on an ecological object, both separately from each other and jointly, to predict the results of such influence and the consequences of nature protective decision made against the pollution sources. By using mathematical models, minimization of technical facilities, in particular, measurement equipment indispensable for control and management of each pollution source is achieved. They also are needed for large plants and factories with biochemical purification of waste water, for their designing and ecologically secure exploitation, and also for of those sites, shops that are guilty of waste water pollution over the norm.

This study aims at developing a pollutant load estimate system by combining a runoff simulation model with GIS database around Lake Biwa watershed and to examining its availability. This system works on a windows computer and has a variety of functions making use of GIS characteristics. Data processing flow in the system consists of four main sections. First, objective basins in Lake Biwa tributaries for estimation are selected using GUI. Second, frame values of the selected basins are calculated from the GIS database using an overlay function of spatial analysis. Third, generated loads, removed loads and remaining loads of pollutants in the basins are calculated using the frame values and basic unit loads in pollutant sources and treatment facilities. And finally, the simulation model simulates pollutant behaviors such as generation, transportation, accumulation and runoff using the data obtained in the preceding processes and time series of precipitation observed in meteorological observatories. By using this system, it has become easier to discuss the pollutant behavior on digital maps in computers for the environmental management of the Lake Biwa Watershed.

Current trends in catchment management and the desire of the community for an enhanced aquatic environment require information about the water environment within the catchment. This information maybe obtained from environmental monitoring programs or from the application of catchment modelling systems. Where an assessment of the effects of potential changes to the management protocol is required, the application of a catchment modelling system is necessary. An important component of the implementation of a catchment modelling system is the definition of values for the system control parameters; these control parameters are the input data to the modelling system. Historically, many of these parameters have been assumed to have an average value across the catchment which has the effect of reducing variability and, hence, inhibits simulation of catchment processes influencing the quality of stormwater runoff. When these parameters are allowed to vary spatially, the number of parameters requiring calibration becomes excessive and, in general, the available data are insufficient to enable definition of a unique set of control parameter values. New developments in information technology, however, can be used to mitigate the problems associated with definition of spatially variable control parameters for catchment modelling systems. Presented in this paper are the results of a study investigating inference of the control parameters for an urban catchment modelling system using information stored in temporal and spatial databases. It will be shown that inference of parameters using this information improves the system performance.

Total Maximum Daily Load (TMDL) developments for the nations impaired streams have become a major focus of activity over the past few years. This concept encompasses the determination of the Load Capacity (LC) of a stream and the associated Waste Load Allocation (WLA) for point and Load Allocation (LA) for non-point source pollutants. Even though guidelines exist and continue to be expanded and changed EPA is not very specific of how the components of the TMDL need to be determined, which has left much of the development at the discretion of the responsible agencies. As a result many different approaches exist to comply with the objectives of an acceptable TMDL. One of these objectives is to ensure the use of state of the art science, while another demands equity among the discharges. It is therefore imperative to include a framework that allows optimization as the omission of the latter causes excess costs that otherwise could potentially be reduced. This paper presents a methodology to effectively compute sensitivities that can be used at a later state to develop an efficient WLA procedure in for the case of CBOD/DO load scenarios. Sensitivities for steady state and unsteady flow scenarios are presented. It is shown that using a 2-dimensional dynamic modeling approach permits identification of a large number of temporally distributed sensitivities for sources along a highly dynamic estuary.

The City of Greensboro is currently developing a modern, dynamic, GIS-based stormwater management system. The first part of the system consists of a complete inventory of the stormwater infrastructure and conveyance system inventory for the entire City. The stormwater and conveyance system inventory has been completed for portions of two principal watersheds in the City (North and South Buffalo Creek watersheds). The current phase (Phase III) of the project includes the remaining areas in the City and is organized and will be completed on a watershed-by-watershed basis. The City of Greensboro is being assisted in the inventory by Dewberry & Davis, Inc. and Concord Engineering & Surveying. Dewberry & Davis is serving as the prime consultant. Using a combination of Global Positioning Systems (G.P.S.) survey and conventional surveys, storm drain structures are surveyed to determine the 3-D coordinates of drainage point features including: curb, grate, and yard inlets; junction boxes and manholes; pipe inlets, outfalls, and energy dissipaters; swales, channel and creek sections; bridges and culverts; lakes and stormwater management facilities (ponds). Information (attributes) is collected for each storm drain structure such as pipe diameters, pipe elevations, pipe slopes, dimensions, materials, structure and surrounding conditions, tie-in's, stilling basin data, flood evidence, damage potential, scour, and connectivity, type of discharge, flow, color, odor and other factors. Numerous measurements are made of headwalls, bridges (including abutments and piers), culverts (including barrels), lakes and ponds (including spillways, risers, etc.). Photos are taken of all storm drainage structures using digital cameras. Cross sections are also being surveyed along major creeks for use in development of computer simulation models (HEC-RAS) for predicting flood hazard areas and water quality conditions. A primary key to the success of Greensboro inventory program has been technology. This technology is currently being used by field crews to ensure the quality and accuracy of the data collected. Survey crews are using Real Time Kinematic (RTK) G.P.S. Receivers with accuracies of +/– 5 cm for both horizontal and vertical coordinates. Feature attribute crews are using state-of-the-art, Pentium-based pen computers and custom developed software to allow for the direct creation and editing of GIS shapefiles in the field where the data is collected. To facilitate the accurate, efficient, and safe (without structure entry) measurement of structure depths and pipe elevations, Dewberry & Davis has developed a one of a kind laser measurement device that measures the vertical difference in elevation from the manhole rim to the pipe inverts. The project also involves a rigorous quality assurance and quality control program to ensure data collection and GIS integration standards.

This model for runoff/flood management offers a practical approach when a quick response is needed for early warning, forecasting, reservoir management, or other situations where water surface elevation downstream is function of the water surface elevation upstream. All required computations are done in a PC, or Laptop in the field, where one velocity needs to be taken in a single location of the cross section of interest. In this cross section, width of bottom and surface, and water depth needs to be known. The dynamic conditions in the cross section of interest will determine how long will take for a wave, or a sustained crest up gradient in the basin to reach such cross section. This method was developed for those situations where no other information exists but an additional head upstream and very general condition of a channel. Under this scheme by computing the channel as a pipe it is possible to obtain results for the range of acceptable errors. The additional head at the inlet of the equivalent pipe will result in knowing the head at the outlet of the pipe, namely the water surface elevation at the cross section of interest. The computer program output gives the extra head in the cross section of interest, as additional water surface elevation and arrival time.

When designing a measuring network several questions must be answered. Basic properties like defining sampling interval and the exact locations for measuring sites are mostly based upon expert's judgement. In this paper a design method for measuring networks aiming at obtaining information for model calibration is proposed. This method is based upon the application of a hydrodynamic model for initial design. Based on a min-max approach for the sampling frequency, a first estimate is obtained. Based on a mathematical analysis of the model parameterization and information content of potential measuring locations it is possible to judge measuring networks objectively. Moreover, by applying some optimization technique (e.g. a genetic algorithm) an automated optimization of the measuring network is possible.

Tipping bucket rain gauges are commonly used in rainfall measurement campaigns. The main disadvantage of these devices is that they underestimate rainfall volumes at the higher rainfall intensities, due to the loss of water during the tipping action of the device. In this paper a new method is proposed to correct recorded rainfall data, without the need for a detailed calibration test. The comparison of the calibration results of 24 gauges (6 different types) shows an interesting relationship between the rain gauge resolution (rainfall depth corresponding to 1 tip) and the slope of its calibration curve. This relationship proves that the tipping time of the tipping mechanism has nearly the same value for every tested device. Therefore the slope of the regression line can be calculated theoretically and, as a consequence, one simple calibration test is enough to determine the complete calibration curve of any tipping bucket rain gauge. This test has to be carried out anyway in order to find the exact resolution of the device. Two monitored storm events, one moderate storm and one heavy storm, are used to study the possible influence of the correction method on calculated cumulative rainfall volumes.

Genetic Programming (GP) has been explored as a flow forecasting tool. A catchment in Singapore with a drainage area of about 6 km² is used for this case study. GP was trained to simulate runoff from a conceptual rainfall-runoff model, Storm Water Management Model (SWMM), which was first calibrated using Shuffled Complex Evolution (SCE) algorithm. Four storms of different intensities and durations are used for training and verification of the GP models. The results show that the runoff prediction accuracy of genetic programming based tool, measured in terms of root mean square error and correlation coefficient, is reasonably high. Thus, GP coupled with a robust optimization scheme such as SCE is a viable complementary tool to traditional conceptual rainfall-runoff models.

The optimisation of complete wastewater systems receives more and more attention, since it has become clear that the current approach of optimising the subsystems of sewerage and wastewater treatment separately often results in sub-optimal solutions. This optimisation, however, shows to be very troublesome in practice due to its complexity. The complexity of a wastewater system optimisation problem is related to the many variables involved (multi-dimensional search) and to the optimisation goal, defined by an objective function. Optimisation with respect to multiple objectives, e.g. overall costs as well as environmental performance, is often suggested but increases the complexity. One of the optimisation or search techniques which is capable of dealing with such a complex surrounding is a genetic algorithm (GA). This paper discusses the potential of the use of genetic algorithms for optimisation of wastewater systems. The definition of the objective function and the characteristics of the GA, especially mutation probability, proved to be key-elements for a successful application of a GA for this type of problems. Finally, it is concluded that GAs are capable of solving the very complex optimisation problems related to improvement of total wastewater systems.

The Urban Pollution Management (UPM) procedure is well established in the United Kingdom as a way to evaluate the performance of the urban drainage system's effect on receiving waters' quality. However, although engineering solutions developed using the UPM procedure have been successful in meeting design criteria for water quality they have been less successful at delivering these benefits at least cost. This paper introduces a novel approach to the design of cost effective solutions to urban water quality problems. The approach combines the concepts of UPM, simplified integrated urban catchment modelling and genetic algorithm optimisation within a software tool named Cougar. The use of Cougar to identify least cost engineering solutions to the pollution problems caused by combined sewer systems is introduced through an illustrative example. The promising results achieved demonstrate the capability of the approach used to optimise designs in terms of cost and water quality performance.

In this paper, a publicly available genetic algorithm (GA) is used to calibrate conceptual runoff models of varying degrees of complexity, using storm hydrograph data from the 200-km² Bushkill Creek watershed, Northampton County, Pennsylvania. The simplest model tested is a linear lumped storage model, with two surface runoff parameters, one subsurface runoff parameter, and no routing. The most complex model uses coupled storage reservoirs (for soil moisture and groundwater storage), includes routing, contains nonlinear surface and subsurface runoff, and has seven parameters. Other model formulations tested fall between these two extremes. The coefficient of determination between the modeled and observed time series is used as the objective function for calibration. The GA calibrated all of the models tested in less than 15 sec using a PentiumIII computer with a 500 MHz processor. In terms of goodness of fit, the addition of routing gave the most significant improvement over the simplest model (r² of 0.92 vs. 0.42), with nonlinear runoff and baseflow generation providing modest improvement (r² of 0.96). The results suggest that a low-dimensional modeling approach will be sufficient to simulate streamflow in the watershed.

The need for sewers to carry sediment has been recognised for many years. Traditionally a minimum ‘self-cleansing' velocity was specified in the UK and, although this approach had been successful in many cases, it was appreciated that a minimum velocity — unrelated to the characteristics and concentration of the sediment or to other aspects of the hydraulic behaviour of the sewer — could not properly represent the ability of sewer flows to transport sediments. In particular, it was already known that a higher flow velocity would needed to transport a greater concentration of sediment, or to transport a given concentration of sediment in a large sewer than in a small sewer. During the 1980s, sediment transport theory had been increasingly applied to the design of sewers, particularly in major interceptor sewer schemes. But, in the absence of any universally recognised guidance, the design methodologies and criteria adopted were developed on a project-by-project basis, building on the designers' increasing experience and understanding of the subject. In recognition of this, a research project was initiated by CIRIA (Construction Industry Research and Information Association) in 1992, to develop a new design methodology for sewers, which would take advantage of the available knowledge on sediment mobility and the effects of sediment deposition on the hydraulic performance of sewers. The project culminated with the issue of CIRIA Report 141. This paper describes the main findings of the project and summarises the design guidance contained in the report.

The purpose of this article is to present the results of the fieldwork conducted in the catchment "El Ensanche" located in the city of Santiago de Compostela (Northwest Spain) and to provide a preliminary analysis of the data collected. The aim of this fieldwork is twofold: first, to carry out a hydraulic analysis and examine the contamination discharged by the occurrence of the CSO's sampled, and secondly to calibrate and validate the hydrological and quality models of the catchment under study. This paper will cover the first of the two objectives in a preliminary approach to solving this problem.

Japan has periods of heavy rainfall twice a year: the rainy season in May and June, and the typhoon season in August and September. However, in colder areas, peak precipitation and drainage occurs in December and January due to snow falling. Methods of elimination for rain and snow run-off vary according to regional climatic conditions. This paper will discuss the problems of the snow run-off and methods for its elimination as well as introduction of regional climatic conditions.

For effective study and analysis of a condition of quality of a water environment, acceptance of the conforming solutions on its improvement the adequate information is indispensable, that is connected to huge number of measurement of different parameters which are carried out with the help of automatic, permanent systems. The environment is fast varying dynamic object, the control behind a condition which one by non computerized methods is hindered and is economically unprofitable. At realization of the analyses more than 3–4 times per day economically are expedient to use the automated systems for status monitoring of environment. In these systems the cost of the information in 2–6 times is less, than at usage of laboratory methods.

Computer models are helpful in analysing complex urban drainage systems. Furthermore evolutionary algorithms may be used for the purpose of simulation-based optimisation to identify desirable system configurations in a realistic manner. In this context the well-balanced combination of theoretical background and practical implementation will result in efficient methods and tools to support everyday planning and management duties. Particularly in view of increasing economic and ecological demands promising solution procedures for the practice of water management are of great benefit.

In February 1998 the Environment Agency (EA) instructed North West Water Ltd (NWWL) to carry out a programme of works to meet a new, demanding discharge consent standard (average of 3 spills per bathing season per outfall) at specific coastal discharge points. NWWL instigated one of the most ambitious, fast track engineering design/construction programmes witnessed within the UK water industry. A programme costing over $105M (£70M) had to be designed, approved by the regulator, procured, built, and commissioned in a period of 15 months. A joint research project between NWWL, Water Research Centre (WRc) and others was reaching a successful conclusion at this time. This concluded that when designing large storage volumes, follow-on storms were of such critical importance that conventional (single event) modelling techniques are rendered inappropriate. Therefore, Continuous Simulation was the only valid way of confidently establishing the additional storage volumes required to meet the new consents. This paper outlines the hydraulic modelling requirements of this programme of works and details how these advanced modelling techniques were applied in particular to the Fylde Coast area that includes Blackpool, the largest holiday resort in Europe.

Sea-dyke is sometimes constructed, so one part of bay has been enclosed. This kind of situation is also observed at Isahaya of Nagasaki in Japan, that is a big project of construction of sea-dyke and reclamation, In order to make clean the enclosed water bodies, it becomes necessary to reduce the untreated pollutant runoff from the watershed as well as to remove the pollutants loaded in the receiving water or to suppress their desorption. Among all, it cannot be avoided to exactly predict and treat the pollutants from the non-point sources, attaining the desirable water environment. In this paper, pollutant runoff mechanism from the non-point sources has been investigated through a field observation. At the same time, a method of exactly predicting the pollutant runoff rates has been discussed for establishing the effective water management.

First flush tanks are supposed to be useful to control the impact of the pollutants contained in combined sewer overflows over the receiving waters. Their design is based upon the idea that the first water in a storm event contains most of pollution caused by the cleaning of the catchment and the conduits. The design of these structures is often controlled by dilution rules which give retention volumes. These rules do not take into account the hydraulic behaviour of the structure itself, which is not well known so far, as these tanks include some hydraulic phenomena which are difficult to evaluate and even to measure. An experimental facility has been developed in the CITEEC Hydraulics Laboratory in A Coruña University in order to measure how the hyetographs and the pollutographs vary at the inlet and at the outlets of the structure. Discharge, accumulated volume and pollutant taxes have been measured in unsteady conditions for different specific volumes (m3/Ha) in the structure. The removal of the pollutant mass has been evaluated, as well as the mixing phenomena inside the tank. The experimental facility consists of a scale model of a first-flush tank similar to those used by the Northern Water Authority in Spain. The hydrographs and the pollutographs have been obtained by direct field measurements in a combined catchment in Santiago de Compostela (Spain). A colorimetric tracer has been used to simulate pollution (neither settling nor reaction has been measured) and different volumes of the structure have been simulated by varying the discharge scales, according with a Freudian model. The result consists of a series of graphs which show the amount of pollutant removed as a function of the tank volume, for some first-flush hydrograph-pollutograph peak distance, and a discussion of the actual efficiency in the removal of pollution in the structures installed in Spain and world-wide.

The aim of this paper is to present the first results of the fieldwork conducted in the catchment ‘‘Fontiñas'' located in the city of Santiago de Compostela (Northwest Spain). The aim of this fieldwork is twofold: firstly, to carry out a hydraulic analysis and evaluate the contamination discharged, by the occurrence of the rainfall events, and secondly to calibrate and validate the hydrological and quality models of the catchment under study. This paper will cover the first of the two objectives in a preliminary approach to solving this problem.

This paper presents the specific problems in urban drainage in cold climate. The climate affects the urban drainage by changing the urban hydrólogical cycle and it becomes more complex in cold weather than in warm-weather. In cold climate the low temperatures and the snowfall causes several problems to the urban drainage systems. The urban runoff is affected by: 1) frozen ground surfaces, 2) frost penetration into the ground, 3) snow on ground, 4) rain-on-snow, 5) snowdrift, 6) natural and man-made snow redistribution and 7) snow removal. Additional problems are caused by frost penetration, frost heaving, freezing in pipes, ice on ground surfaces, clogging of gutters and inlets, icing in manholes and in storm sewers, ice in open watercourses such as urban creeks and rivers, ponds and lakes etc. There are also changes in the transport of urban runoff and stormwater pollutants, the operation of runoff control facilities and sewage treatment plants. Snow may be stored on the catchment and produce runoff during warmer weather. Frozen ground thaws slowly and high runoff rates may occur when rain falls on frozen ground. In maritime cold climate subsequent melting and freezing periods often give significant runoff periods. Other problems are due to flooding, combined sewer overflows (CSO) and overloading of wastewater treatment plants. Pollutants may be accumulated in the snow in streets. When the last 10–20 % most polluted part of the accumulated snowmelts and enters the sewer system, a shock load of pollutants may occur. The sewer system is filled up from previous inflow and a part of the concentrated pollutants may be discharged in overflows to local recipients. The wastewater conveyed to the treatment plants is cold, 2–5 C°, due to high inflow/infiltration in rain-on-snow and melting periods. Problems to the urban surface runoff are caused by the snow redistribution on sidewalks etc or temporary surface water storage because of clogged inlets. In cold climate areas the planning and designing procedures for urban drainage often do not consider the presence of the snow and even not the operation and maintenance procedures and guidelines. A great need for further research and development in urban drainage is therefore needed, followed up by a relevant training and education, including preparation of appropriate educational materials. Such work is going on at the Department of Hydraulics and Environmental Engineering, NTNU.

This paper describes a cost reducing system for intercepting combined sewer systems in cold climate areas with emphasis on reduction in annual CSO. It also describes how the system was realised in practice and describes research, development and experiences on full-scale systems for cost reduction in construction and operation of interceptors for combined sewer systems in some urban areas in Norway and Iceland. The procedure was as follows: 1. Using one joint pumping pipeline for all pumping stations for sewage collection and transportation to a treatment plant. 2. Conduction of a thorough investigation of the runoff conditions and accurate flow measurements in combined sewers during the whole year together with calibration of a computer model. 3. Construction of flow-duration curves for evaluation of the annual inflow to the interceptor. The first full-scale system was implementation in the city of Steinkjer, Norway. Therefore the system was called "the Steinkjer system". The system has been successful where it is in use in Norway and in Iceland. In average the system has proved to be 20% cheaper in construction and 30% cheaper in operation than a conventional interceptor system.

Urbanization makes impermeable area increases very fast. As a result, for the same rainfall the runoff quantity and the peak flow are bigger than that under natural condition. Meanwhile the water infiltrated into stratums decreases and the underground water level falls down. In order to improve urban water environmental situation, the techniques of rainfall infiltration are introduced, which focuses to recover and improve the infiltration condition of urban area. Dalian has 1,340 hm² vegetation area, may cut about 1.27 x 10⁶m³ rainfall water to decrease flood and part of it become underground water a year.

Rainfall is the most important input for many design calculations. Ideally, long historical rainfall series should be used for modelling and a statistical analysis should be performed on the results afterwards. In combination with the detailed models that are commonly used nowadays, this leads to huge calculation times. This research was set-up in order to verify which kind of simplifications can be made with respect to the rainfall input. One must find an optimum between accuracy of the modelling results and calculation effort. This optimum can be different for different applications. The different types of rainfall simplifications which are considered here are composite design storms, short selected rainfall series and modified single storm events. In many cases the optimum is more likely a simplified model in combination with continuous long term simulations. Well-calibrated (physically based) simplified models can reach almost the same accuracy as the corresponding detailed models within a fraction of the calculation time. Furthermore, these simplified models are very useful in order to select or compose the proper rainfall input for detailed modelling.

In 1999, the digitization of a unique rainfall series of 100 years was completed, which was an ideal opportunity to search for trends in the rainfall. Large variations in rainfall extremes over the century have been observed. For small aggregation levels there is a small decrease in extreme rainfall events over the century. For large aggregation levels, between a few days and a few months, there is a more explicit increase in extreme rainfall. Because the rainfall on the seasonal aggregation level is only slightly increased, this increase for aggregation levels between a few days and a few months can only occur due to larger clustering. However, the final conclusion is that these trends are not significant.

As a result of the agreed Asset Management Plan for North West Water (NWW) covering the period from April 2000 to March 2005 (AMP3), there is a need to improve a total of 919 unsatisfactory intermittent discharges to surface waters, either at Wastewater Treatment Works (storm tanks or inlet CSOs) or within the wastewater network. The 919 intermittent discharges are made up as follows: 1) 125 are designated as impacting on bathing Water / shellfish waters, 2) 353 are designated as having an aesthetic impact without river impact, and 3) 441 are designated as having an impact on the in-river quality. (River Impact). Some linking of problems has been carried out to ensure that geographical location is taken into account. From this the following numbers have been identified: 1) 585 river impact and linked aesthetics / bathing waters, 2) 121 stand alone bathing water / shellfish intermittents, and 3) 213 stand alone aesthetic intermittents. The three categories have differing modelling study requirements, as the design standards are different.