1. General overview of urban drainage principles and practice by Cedo Maksimovic
1.1 Introduction This chapter chapter is meant to serve as a common introduction to UNESCO’s three publication dealing with urban drainage in three particular climate zones:
volumes
• ASA - Arid and Semi Arid • HT – Humid tropical and subtropical and • CC – Cold climate zones It’s structure reflects the need need to underline the similarities similarities of urban drainage problems problems in particular climate zones and to address the need for breakthroughs in both research and application of the adequate tool in in these region. The chapter provides an introduction to the contemporary state of the art in analysis, modelling , design and management of urban drainage systems indicating indicating that the particular aspects are covered in separate volumes for each zone . The principles of “fitting” the urban drainage solutions solutions into integrated catchment management management plans is introduced. Two principal components components of the integrated integrated flood mitigation mitigation solution solution such as: structural and non-structural measures measures applied in two parts of a river basin (catchment) i.e.: i.e.: urban and suburban zone and the rest of the catchment – rural and natural natural areas are presented. The differences of the situations situations in developing and developed countries countries have been highlighted. highlighted. The concept of natural drainage within the broader broader framework of sustainable solutions solutions is re-iterated and its major components will be presented by distinguishing between the rehabilitation of aged aged systems and the construction of new ones. The scope dependent nature of storm system modelling is presented in the form “an appropriate tool for each task”. The major types of modelling concept (quality, quantity, interactions, integrated) are briefly analysed, by placing placing an emphasis emphasis on data needs needs and data reliability as well as on the need need for development of a new generation of modules modules that will be able to cope with particular aspects of specific climates.
1.2 General characteristics of urban drainage and sustainability concept Water in urban areas, and urban storm drainage as a part of the urban infrastructure, are topics which are gaining in importance in recent years. Cities now house 50% of the world population, consume 75% of its resources, yet occupy only 2% of the land surface. By the middle of the next 9
century, it is confidently predicted that 70% of the global population will live in urban areas. The number of mega cities ( > 10 million inhabitants) will increase to over 20, 80% of which are in developing countries (Niemcynowicz, 1996). Properly designed and operated urban drainage systems with its interactions with other urban water systems are crucial element of healthy and safe urban environment. The concept of sustainable development is provoking a profound rethinking in our approach to urban water management (ASCE/UNESCO-IHP, 1998). Sustainable development is that which “meets the needs and aspirations of the present generation without compromising the ability of future generations generations to meet their own needs” (WCED, 1987). So, sustainable solutions solutions have a “now” and a “then” component, and improvements though necessary in the present must not be carried out at the expense of future needs and situations. An alternative definition (IUCNUNEP-WWF, 1991) asserts that sustainable development is that which “improves the quality of human life while while living within within the carrying capacity capacity of supporting ecosystems”. ecosystems”. Here, the emphasis is placed on mankind’s demand for and impact upon earth resources and the environment. Finally, Agenda 21 behoves behoves us to “think global, global, but act local”. local”. Public participation participation becomes important important and demands individual individual responsibility. responsibility. Sustainable services must be environmentally friendly, socially acceptable and financially viable into the next millennium (Butler & Maksimovic 1999). The sustainability concept calls for overall rethinking and this implies paying attention to particular situations in the local area. Learning about natural and man made processes that that affect the runoff runoff quality and quantity is of prime importance. importance. This publication is thus expected to point out the most important issues that affect the way that we analyse, design, build and operate operate our storm drainage systems in a nature friendly fashion. Our current knowledge about the physical processes involved is far from satisfactory, even in temperate climates where the most of research has been carried out out in the past. Knowledge about processes affecting urban storm drainage systems in particular climates covered in this publication (arid and semi-arid, semi-arid, humid tropical and subtropical and cold) cold) is far from satisfactory. However the publication is aimed at providing an up to date look at solutions to flooding and water quality problems. The concept concept of sustainability sustainability calls for amenity amenity and resources recycling recycling to be taken into account as well. The authors are aware of the fact that many issues raised here require further studies, research and development and that the issues raised will provoke further refinements. In densely populated developed countries (UK, Germany, some parts of the USA, Japan, etc.), urban drainage consumes a high proportion of the investments into urban infrastructure. The reasons for this are the obvious need for an integrated approach to urban water management, and raised public awareness of the pollution caused by urban effluents, which affect both the urban areas themselves and the receiving water bodies. The situation in developing counties counties is also changing rapidly in the sense that all parties involved involved in planning, design, management management and maintenance maintenance as well as funding ( World Bank, aid aid agencies agencies etc.) are becoming aware that storm drainage can not be ignored. On the contrary, it has to be incorporated into integrated urban infrastructure projects with their mutual interactions encompassing not only the conventional conventional problem problem of flood mitigation but also health hazard reduction (water quality concerns) and problems of urban amenities and resources management (Figure 1.1). Although cities are in contact with water from various origins (ground water, streams flowing through or near the city etc.), the major concern of urban drainage systems is water originating in the city area itself, i.e. water from local rainfall (urban storm runoff) and its interaction with the water originating from the rest of the river basin. The change of the role of urban storm drainage (USD) and developments of information processing technology have imposed a need for new tools and products to be used in the problem solving procedure. Methods for flood protection by local storms and for assessment of the effects of pollution transported by storms on receiving waters have been significantly 10
improved during the past two decades with the introduction of computer based simulation, design, optimisation, real time control and management. The achievements of modern informatics (i.e., a higher level of information processing) have made a significant impact on all aspects of problem solving. However, despite significant development achieved, there is still a big gap to be bridged since a compact and reliable package that adequately predicts dynamics and spatial distribution of urban floods and that incorporates source control measures does not seem to exist in the world.
Quality
Quantity
Amenity and resources
Figure 1.1 Stormwater quality, quantity and amenity and resources management of equal importance In modern societies, the status of urban drainage as a part of the integrated infrastructure system varies from one country to another, depending primarily on the level of development and the society awareness of the importance of this problem. In general, the importance of the system increases with the level of development, but there are also exceptions. The awareness of the wetweather pollution potential has rapidly increased in recent years. The systems, which used to have a simple function of collecting storm water and conveying it to the nearest point of disposal as soon as possible, have gradually evolved and are being replaced by the integrated systems which are gaining in importance. Their role has changed and now in addition to covering urban flood protection, pollution control and management they are starting to cater for improvement of the quality of life by bringing water features – creating urban amenity in the city. Additionally, storm water is considered to be a precious resource, which can be retained near the source to be reused, recharged to the underground for aquifer replenishment or to create habitat for the return of wildlife to designated urban areas etc. Conventional urban drainage systems are separate such as shown in Figure 1.2 or combined in which case both waste and stormwater share the same pipe. During dry weather, water is directed to treatment plant (if existing) and during wet weather, part of the mixed water in combined sewers diverts to receiving stream via Combined Sewer Overflows (CSO). If the city is served by a wastewater treatment plant, CSOs may be one of the major point sources of receiving water pollution. In practice, separate systems rarely remain fully separate; there is always some storm water in foul system and waste water in storm systems. In most cases they behave like two combined systems with various degrees of waste water dilution. Treatment plant suffers from intermittent overload during storm periods. Increased environmental concern has lead to development of the concept in which, at least in developed countries, conventional storm drainage systems are gradually being replaced by the systems based on runoff quantity and quantity control. The system consists of several techniques that aim at controlling the problem as near to the source as possible – thus the term “source control”. They all attempt to mimic the natural processes involved. The techniques include storage, treatment and infiltration, by a “water management treatment train” (Figure 1.3), that results in significant reduction of 11
peak and volume of runoff, improved water quality and a possibility of using storm water as a resource and as an element of urban amenity. However, the means of implementing the element and principles of this technology in urban drainage in particular climates is an art still to be mastered despite significant achievement in some countries – for example Sweden, Stahre (1999) in cold climate, city of Curitiba, Brazil, in tropical, several cases in Israel etc in arid climate conditions as resented by Simon (1996). However, in order to reach greater sustainability in both conventional and innovative urban drainage systems, better understanding of the physical processes, interactions between the systems and environment in particulate climatic conditions is needed. This publication is supposed to cover part of missing information and to address the problems that need further investigations.
Surface flow
Inlet
Industrial
Sanitary
Treatment plant
Outflow
Treated water
Figure 1.2 A conventional separate foul and surface water drainage
Conveyance Evaporation
Conveyance
Discharge Prevention
Discharge
Source control Site control Regional control Infiltration Receiving water
Figure 1.3. Surface drainage management train – likely sustainable solution (CIRIA555)
12
1.3 Urban drainage system as a part of an integrated river basin water management – principles of design and operation It is well known that the river basin has been considered an entity that determines both the range and reach of human activities with respect to water in both ancient and modern societies. The catchment is used as a unit for planning and management of not only water, but also of other resources, as well as human and economic activities. In the case of urban drainage of a particular city, the relevance of a catchment is greater for smaller catchments and decreases as the size of the catchment increases, in the sense that the relative effect of the quantity and quality of runoff water generated by that particular drainage system diminishes with the size of the catchment and with the distance from the point of storm water disposal. However, the integrated effect of all storm drainage systems contributing to the balance of surface water and to the flux of suspended sediment and other pollutants has to be taken into account at the level of river basin or sub-basin upstream of the point under consideration, especially in densely populated areas. The interaction of storm drainage systems with downstream municipalities and water users is strong in those cases when the drainage peak flow uses up the capacity of the river channel, so that no capacity is left for downstream runoff. In these cases, the downstream-upstream relationships and links have to be analysed in order to either share the existing capacity or to share the costs of its enlargement. Small river basins in densely populated areas are therefore more sensitive to this problem and shall be analysed in the following discussion. On the other hand, the rivers carrying water from large catchments serve as receiving waters for both solid and dissolved pollutants, and the effect of urban storm water disposal has to be analysed from the point of view of its pollution and contribution to the silting of downstream water, including reservoirs. Alterations to the natural water balance within the catchment area can have both positive and adverse effects on upstream and downstream water users. In that respect, integrated planning and design of urban drainage systems requires that both effects are analysed and an unbiased assessment is made in all phases of the planning and management process. Figure 4. outlines an approach which integrates catchment wide, metropolitan/municipal as well as local area planning and management considerations.
Figure 1.4. Urban Storm Water Master Plan as a part of the Catchment Management Plan
13
The general goal of integrated water management is a sustainable utilisation of water resources respecting the social, economic and environmental interests. Considering the close interrelationship between the society and economy, the first two groups are usually aggregated into socio-economic issues. It also includes institutional issues. It should be recognised that the goals and objectives of integrated water management are formulated at various spatial scales, involving all three components. According to Butler and Maksimovic (1999) the institutional aspects cover the following:
• Development of improved informatic support tools for planning, design and • • • • •
operational management based on improved quantity and quality of data. Incorporation of more (relevant) components and stakeholders into the decisionmaking process (e.g. sustainability, public attitudes). Development of methodologies to evaluate the uncertainty and risk associated with future water management strategies. Decision on how to consult and educate the public concerning the importance of urban water issues. Devising suitable organisational/institutional structures to incorporate the integrated, holistic system management we advocate. Enacting appropriate supporting water legislation and standards.
The fundamental qualities of integrated water management are its holistic nature, which recognises the system complexity and inter-connectivity of its elements, demonstrated by exchange of information, energy and matter, and the style of planning actions. The holistic approach also equally involves local/municipal and regional authorities, engineers and natural scientists, environmentalists and decision makers, politicians of all parties, governing and in opposition, as well as the people affected (Geiger, 1994 and Geiger & Becker, 1997). Sustainable water management ensures that no matter is accumulated or energy is lost, by recovery and reuse techniques. This approach requires novel, environmentally sound technologies. In the urban drainage field it calls for a wider application of source control. In the context of urban and industrial water resources, the most pertinent water uses are water supply (safe, reliable and equitable), drainage and flood protection (affordable), sanitation with maximum reuse, recreation (protecting public health), aesthetic and cultural values, and ecosystem health. Solutions applied at urban catchment level have to be analysed in terms of it upstream and downstream interactions. The conditions may vary in various climate conditions and these will be analysed in the main chapters of this publication . Contributions of urban storm drainage projects to the conflicts and uncertainty in water resources plans at a river basin level, can be analysed by taking into consideration the ways in which the existing urban structures, their features, and the newly planned drainage elements affect both water balance and quality in a particular urban area. In this respect, the major difference between urban and rural (or natural) part of a river basin, is the reduced infiltration potential of urban areas and the fast response in generation of surface runoff. A mutual interaction of urban runoff and flows in adjacent steams is shown in Fig. 5. Water running from the upstream parts of catchments flows either through the city’s regulated stream, or through its system of urban drainage infrastructure. The major difference in approaches to integrated solutions is indicated by the ratio of the urban peak flow to the flow in the receiving stream, at the downstream end of the urban area. The forms of urban flooding caused by other man made and natural disasters such as storm surges that usually coincide with heavy rainfall , dike break (Iwasa, Inoue, 1987) have also to be taken into account. We take the Danube, as a large river flowing through the large cities of Vienna, Bratislava, Budapest, Belgrade, etc., as an example. In the most extreme events of heavy storms over these cities, the local runoff contributes only a very small proportion of the flow in the river, and one 14
can claim that the management of urban storm drainage system in these cities will not significantly affect the flow in the Danube. This effect is diminishing as one travels form the spring to the river mouth. In that respect, and with reference to water quantity, these systems do not affect the peak flow in the receiving water and can be designed independently. However, small streams near large cities (i.e. small value of the factor Ψ - equation 1 ) are strongly affected by the urban runoff peak flows, and implementation of source control could be strongly recommended. In some cases it is the only solution. In many cases, the peak flow generated by urban runoff is comparable to the conveyance capacity of the receiving stream. In these cases, the management of urban drainage has a significant effect on the receiving water and its downstream reaches. Consequently, the solution for the particular storm drainage problems has to be developed at the catchment level, and in an integrated way. However, growing concerns over the quality of surface runoff require that the interaction of particular city’s pollution load is addressed in conjunction with other pollution contributions from both upstream and downstream urban areas. The difference in the capacity of the main receiving water calls for classification of the concepts of storm drainage solutions, depending on the ratio of the peak flows, likely to occur at the point of disposal (end of the pipe), to the average discharge in the receiving stream (shown in Figure 1.5.)
ψ= where : Q R
QR 0 − QR
(1.1)
QR
= inflow at the upstream end of the urbanised area; Q R = outlet flow at the 0
downstream end of the urbanised area.
Figure 1.5. Classification of urban sub-catchments and interaction between urban runoff and the adjacent river. In the cases when the receiving urban stream reaches its full capacity under the effect of a given return period flood wave. Source control solutions will be strongly dependent on the value of the factor Ψ and of the urban catchment location (part of the city at high elevations and storm water drained by gravity, or part of the city at low elevations which can be flooded by receiving water). A high value of the factor Ψ , means the high capacity of the receiving water, thus the storm runoff from the local urban area might not affect water level in the river. Thus the implementation of source control may be less beneficial to the for the part of the city located on 15
the left river bank situated in the higher elevations (Figure 1.2.) than to the portion of the city located on the opposite side of the river. In the latter case, the implementation of source control measures could result in significant savings in both construction and operation management costs. Both the cities of Budapest and Belgrade are good examples in which the portions of the city on opposite sides of the receiving waters (Danube in the case of Budapest and both Sava and Danube rivers in the case of Belgrade), have completely different flooding vulnerabilities and different source control suitability. Thus the portions on different river banks would benefit differently if source control measures would have been applied. As a conclusion to this section one can say that there are no universal rules of thumb for implementation of source control techniques. The most appropriate solutions have to be firstly sought through the resolution of conflicts between land and water users at catchment level and than at the level of municipal storm water plan. Both structural and non-structural measures have to be analysed in terms of the suitability to an application of source control and benefits that can be reached. Stormwater quality issues, which were almost ignored in the past, have to be addressed properly in terms of their spatial and temporal distribution and effects on receiving water bodies. A possible approach in the evaluation of suitability by GIS support is given in the paper of Macropoulos et al ( 1998) and will be discussed later in the section 05).
1.4 Basic principles of rainfall-runoff and pollution modelling and outlook for their application in particular climates 1.4.1. Water quantity aspects
Modelling in urban drainage serves various purposes such as overall assessment of the catchment response as a part of strategic and master planning to detailed network and ancillary elements design, assessment of pollution, operational management, real time control and analysis of interactions among sub-systems. The type of model applied depends on the goal of modelling, spatial coverage, data and technology availability but most often on the knowledge, skills and experience of the modeller. Once familiarised with a certain model, the user tends to apply it even in the cases in which that particular model is not appropriate. In principle the simple lumped models (black box and similar) in which the whole catchment is treated as an entity, can provide reliable results and good fit with measurements obtained on the same point from which the data have been used for calibration. One cannot expect to get realistic results for the points within the catchment ( network) unless the measurement is performed on that point and “new model” obtained by calibration against that data sets. The only reliable approach is to obtain more reliable data on a catchments physical characteristics and then develop and apply physically based model in which uncertainty is reduced by replacing the role of physically meaningless parameters with these characteristics. The general principles of conceptual and physically based models of both water quality and quantity have been known for several decades Maksimovic, Radojkovic (1986) , Yen (1986), O’Laughlin et al (1996). Detailed description of modelling principles in temperate climates in which these models have been developed, and from which the data were collected for model calibration, is beyond the scope of the present chapter. More details on the attempt to use these models in specific climates will be given in the separate chapters. The conceptual models are based on assumptions such as constant runoff coefficient, SCS curve numbers, rational formula, time-area, unit hydrograph etc. originating years before the computers reached the level of development that allowed their broader application in daily practice. Although they were developed for application in natural and rural catchments, they continued to be used in urban areas, where the conditions are significantly more complex, spatial variability of soil and impervious areas require much finer spatial resolution, and man made 16
object require detailed specification of infrastructure system and their interaction with the flow pattern. When properly calibrated against measurements, these models can produce seemingly logical results especially if one is modelling the whole urban area as a single catchment and model calibration performed against data in one point. This could be useful for example in design of centralised storage facilities, inflow to treatment plants and similar cases when the response of the whole catchment is considered. However, for detailed runoff modelling of complex features such as trunk systems with broad sub-catchment areas, street drainage systems with detailed property drainage components and sub-catchments, models of this nature generate results of high level of uncertainty. Physically based models in which a more detailed presentation of the catchment characteristics are made and distributed modelling is applied should theoretically be less sensitive to subjective assessment of model parameters. In the simplest terms, the whole catchment is divided (delineated) in smaller sub-catchments which, depending on the purpose of modelling, can vary in shape and size arbitrarily as to accommodate, the most realistic model presentation of flow pattern (Figure 1.6). The temperate climate approach considers the following element of modelling :
• Rainfall as an input: single storm, series, historical rainfall, etc. • Interception (surface depression) • Infiltration (steady, unsteady, unsaturated soil, simple solution or Richard’s equation • • • •
based solutions) Surface runoff Gutter flow Flow in ancillary structures Pipe flow
When it comes to runoff modelling in specific climate conditions it is evident that this approach needs further upgrades as to accommodate features like:
• • • • • • • • •
Different forms of precipitation (snow fall in CC) Different forms of interception (HT) Different forms of runoff formation (snow melt in CC) Effect of different cultural, planning, building and other effect on interception (ASA and other) Strong interaction of surface runoff with sediment transport (ASA) Lack of proper infrastructure and interactions with solid waste and waste water ( developing countries, low income habitats) Interaction with ground water (infiltration, exfiltration - all) Interaction with source control features Interactions with real time and other control structures.
It is evident that these features require separate attention, although not necessarily a completely new model. Most of them can be accommodated into reliable, well conceptualised physically based model. This does require more knowledge, reliable data, proper interpretation. Some of the shortcomings of the models of this nature as presented in the paper of Maksimovic et al (1999) are:
• Concepts dating back many decades. The development of contemporary information technology’s computing power has not always been mirrored by improvements in the models, 17
• The outdated concepts are often hidden behind powerful graphics and presentation glamour • Modelling of urban water interactions are almost non-existent and integrated modelling is in its infancy
Figure 1.6. Summary of physically based approach requiring a reliable catchment delineation
• Many models lack modularity, transparency and transportability (automatic "scaling up and down"),
18
• Data quality and completeness is usually not properly addressed by software • • • •
•
developers, its users often lack the knowledge of basic assumptions on which models are built, Data acquisition and processing are not compliant with model structure and complexity, or models are not capable of producing proper results from available data base (DRIPS Syndrome - Data Rich Information Poor Systems Complete digital data on the urban infrastructure and on the spatial distribution of basic urban environment features (land use, DEM- digital elevation models etc.) is rarely found at an appropriate horizontal and vertical resolution, Thorough testing against high quality data sets is often exercised neither by developers nor by users, High level of independent, international verification of new products is rarely performed. In-house verification tends not to reveal the weak points of the products, A proper educational component is often missing.
Additionally, O’Loughlin et al (1996) claim: “ Despite this availability of information, tools and guides, and the success of rainfall-runoff models in providing generally acceptable basis for design of infrastructure works, there are limitations to the modelling of rainfall-runoff processes . They identify four major reasons for this:
• • • •
Insufficient data Variability of rainfall inputs Insufficient temporal detail Model incompatibility
Concerning the level of detail they point out that engineers have long been skilled at idealising or conceptualising systems, to produce manageable models involving typically 10 to 100 elements to represent a complex urban drainage network. Now that there is a capacity to work with more detail, it is necessary to look at appropriate levels for various tasks and the relationships between models of various scales. It has already been mentioned that for studies concerning general response (in the terms of both quality and quantity) of the catchment or sub-catchment of a considerable size, it may suffice to apply a lumped approach in which spatial variability of catchement characteristics as well as of precipitation is ignored. Providing that reliable measurements at the end of catchments are available, the results of input-output correlations are used instead. Some models of this nature will be discussed in the particular chapters of separate volumes. 1.4.2. Quality aspects
Storm water runoff becomes polluted when it washes off concentrated and diffused pollution sources spread across the catchment. An example of the average concentrations found in storm runoff is presented in Table 1.1. (Source: Xanthopoulos and Hahn 1993 and Cordery 1977). In addition to soil erosion caused by raindrop impacts and shear stress action, two major sources contribute to storm water pollution in temperate climate zones: a. diffused sources (Figure 1.7) originating primarily from atmospheric fallout and vehicle emission, additionally spread by the vehicles and wind and b. concentrated sources originating mostly from human activities – bad housekeeping (industrial wastes, chemicals spread in urban areas – gardening for example) exposed to and widespread by wash-off by storm runoff. 19
Both of the processes generate soluble and suspended material. Throughout the process of transport, depending on hydraulic conditions, settling and re-suspension takes place on the surface and in pipes, as well as biological and chemical reactions. These processes are often considered to be more intense in the initial phase of the storm (first flush effect), however, due to temporal and spatial variability of rainfall and flowing water, first flush effects are more pronounced in pipes rather than on surfaces Deletic (1998), where high concentrations of pollutants can be expected throughout the runoff process. The success of runoff quality modelling exercise is strongly dependent on the quality of model (its reliability to realistically reproduce processes taking place in nature), and the reliability of data against which the model has been calibrated Table 1.1. An example of average concentrations of pollutants in storm runoff Quantity Mean Concentration Conductivity (µS/cm) 108 - 470 BOD (mg/l) 7.3- 15 TOC (mg/l) 26 – 28.3 NH4(mg/l) 1.92-2.75 Pb (µg/l) 160-525 Zn(µg/l) 320 - 2000 Ni(µg/l) 35-57 pH 6.47-6.78 COD (mg/l) 47-146 DOC(mg/l) 3.1.-5.1 P(mg/l) 3.1.-5.1 P(mg/l) 1.6-2.95 Cd(µg/l) 2.8 – 6.4 Cu(µg/l) 23-184 Coliforms (/100ml 2.2 – 5.6 (10*6) Similar to quantity modelling, storm runoff quality modelling can be undertaken at various levels of complexity, starting again with simplest input - output relationships. More advanced models deal with spatial distribution of diffused pollution sources and analysis of unsteady process of incipient of solid particles motion, bringing them to suspension, transport along the paved areas, deposition in grassed areas (Deletic 1999), transport through the pipes and disposal either into receiving water body or into treatment plant. In order to enable the comparison of modelling approaches between the models being used in temperate climates with those in development or in need to represent the conditions in other climate conditions, the basic principles of quality modelling are briefly summarised. Most of the models in current practice model the runoff quality by correlating the concentration of pollutants to the concentration of particles of suspended solids which are modelled in the phase of build-up and wash-off. The most common approach in build-up modelling is based on the assumption of an exponential relationship between the amount of solids available on the surface, M , and the duration of antecedent dry weather period, tdry. This equation was adopted in the model of Deletic at al (1977) – Figure 1. 8: M( T ) = M o (1 − e
−k( t dry + t')
)
(1.2)
where M [g/m2] is the amount of solids available on the surface, T [day] is the time elapsed from the start of the first rainfall in the series, tdry [day] is the duration of antecedent dry weather 20
period, and t' [day] is the virtual time, M 0 [g/m2] is the maximum amount of solids expected at the surface, and k [day-1] the accumulation constant. The virtual time is calculated by assuming that deposition is zero at t’ days before the start of the antecedent rainfall, as indicated in Figure 1.8. A spatial distribution of solids is modelled, based on r ecords from the literature, a different approach to prior models, which all assume that sediment is distributed evenly over the modelled surface. It should be noticed here that this approach build-up modelling could be successfully used in ASA climates where most of the solids accumulated are either atmospheric deposit or are transported by wind. However, in cold climates where a great deal of pollution is experienced in the snowmelt period from de-icing activities, which are not uniformly distributed over the entire catchment, alternative methods have to be applied (for example GIS supported spatial distribution of salt used in de-icing). In this respect a critical evaluation of other models used in both quantity and quality modelling in particular climate conditions should be made as for their suitability for application in specific climates.
wind
Rain and snowmelt traffic
industrial wastes and landfills
soil erosion
polluted runoff
human activities
Figure 1.7. Diffused pollution sources in urban area
21
activities
The reliable modelling of suspended solids wash-off has to be combined with surface and pipe flows to which the solids entrainment module has to be attached. The approach applied in Deleti} et al (1997) will be used as an illustration. In this approach, the solids wash-off one dimensional model contains the following sub-blocks: 1. overland flow; 2. solids entrainment; 3. suspended solids transport by overland flow. Overland flow is modelled using the kinematic wave equation, which has been used before for the modelling of surface runoff. Solids entrainment is assessed by a new method, developed by the first author, which considers independently rainfall and overland flow effects on amount of material lifted from the surface. The rainfall effect is assessed by means of the kinetic energy of rain drops, while the effect of flow is expressed by shear stress. One calibration coefficient is needed for this method. The general principles of modelling will be described in more detail enabling thus the comparison to be made between the commonly applied approach and the one that could be used in presenting the specific aspects wash-off in ASA, CC and HT climates. Physically based modelling deals with mass and momentum conservation principles which are that simplified or adjusted for the specific features of the particular catchments ‘characteristics, boundary conditions internal and external local climate induced boundary conditions. For a unit width of the road surface (Fig. 1.9.a) the continuity equation, Eq. 1.3, and the full momentum equation, Eq. 1.3, can be written as:
∂ h ∂ q + =i ∂ t ∂ x e
(1.3)
τ +τ ∂q ∂(q 2 / h ) ∂h + + gh − ghS s + b i = 0 ∂t ∂x ∂x ρ 1 2 3
{
1.
1 4 24 3
1 2 3
2.
3.
4.
1 4 2 4 3
5.
Figure 1.8: The concept used in modelling of solids build-up at the surface
Figure 9: a) Road surface flow; b) Gutter flow 22
(1.4)
where, h [m] is the water depth, q [m3/s/m] is the unit overland flow, ie [m/s] is the effective rainfall intensity, Ss [-] is the surface slope (the natural slope of the street surface), tb [Pa] is the bed shear stress, ti [Pa] is the additional shear stress due to rainfall drops, x is the spatial coordinate, and t is the time from the start of rain. It should be noted here that the source term on the right hand side of the equation 1 is based only on the contribution from direct rainfall. In CC conditions for example this term has to be modified as to include the effects of snow melt and freezing, which have to incorporate the temporal variations their thermodynamic properties. Similarly in HT and ASA conditions it might be necessary to include the evaporation term which has not been included here. The initial and boundary conditions are given below, h( x ,0) = 0 ;q( x ,0) = 0 ;q(0, t ) = q up ( t )
(1.5)
where qup is the unit overland flow at the end of the upstream section. The effective rainfall intensity ie was calculated by Linsley's equation, i e = i(1 − e − P / yd )
(1.6)
t
∫
where i [m/s] is rainfall intensity; P = idt [mm] is the total amount of precipitation up to time t, 0
and yd [mm] is the retention coefficient and is dependent on surface type. The bed shear stress , tb was defined as:
q ρ τ b = C τb ( ) 2 2
(1.7)
h
The friction coefficient, C tb depends on the flow type :
C1 Re Re ≤ Re b , laminar flow Cτ b = C 2 Re > Re b , turbulent flow Re C3
(1.8)
where Re=q/ν is the Reynolds number and ν is the water kinematic viscosity. C 1, C 2, and C 3 are constants that depend on surface type, and Reb is the critical Reynolds number between laminar and turbulent flow. The effect of rain drops was modelled by an additional shear stress , τ i which is difficult to define separately. Therefore, the total shear stress, written as,
τt = τb + τi
(1.9)
which incorporates both phenomena was used. The local and convective terms (marked as Term 1 and Term 2 in Eq. 1.4), as well as the pressure gradient term (marked as Term 3) are much smaller then the remaining two and are usually neglected and only gravity and friction terms (Term 4 and Term 5) were kept within the dynamic equation. The resulting equation is well known as the kinematic wave equation, and has been used for the modelling of both overland and gutter flow . Consequently, for modelling of gutter flow the following equations can employed (Fig. 1.9.b).
23
∂A ∂Q + = q w + q r + ieLg ∂t ∂x Q=
0.375
n
H 8 /3
Sg tgϕ
(1.10) (1.11)
where Q [m3/s] is the gutter flow; qw [m2/s] is the unit inflow from the sidewalk; qr [m2/s] is the unit inflow from the road surface; ie [m/s] is the effective rainfall intensity L g [m] is the gutter width, A [m2] is the cross section area, H [m] is the water depth by the curb, n [m-1/3 s] is the Manning roughness coefficient, S g [-] is the longitudinal slope of the gutter, ϕ [0] is the transverse angle of the gutter. Eq. 1.11 is known as Izzard's (1946) formula which differs slightly from Manning's expression, but gives better results for the shallow flow in a triangular cross section channel . In specific climate conditions the right hand side of the equation 1.10 can be modified as to include additional terms the contribute to water balance. It was assumed that there is no flow at the beginning of a rainfall event (the initial conditions). The inflow from the upstream reach was used as an upstream boundary condition. Furthermore, the solids entrainment, pollution transport by overland flow and gutter flow are modelled by making use of kinetic energy of rainfall drop impact, carrying capacity of surface runoff and principle of turbulent transport and diffusion in open channel flow (Deletic et al. 1997). Although these principles are universal thus applicable in other climate conditions, the appropriate modifications have to be made in transport and diffusion equations in order to incorporate their specific conditions, mainly in the source and sink terms of mass conservation and transport equations. Some of these principles are discussed in the main body of the text, however it should be noted that they are to be further investigated, tested and checked against reliable data. In this respect this publication is to be seen as a source of information on both current practice and need for further investigations in order to realistically reflect the conditions in particular climates. Additionally, it is noted that the above considerations have only dealt with suspended solids.
1.5 Common UD models and needs for their improvements and update Physically based models are based on the analysis of processes on the surface and in networks, and is performed by taking into account detailed features on the surface (topography, soil characteristics, land use, connectivity between elements etc.) and of the networks and ancillary structures. This section will mention just a few (more detailed presentations are given in the other chapters ) of the existing models available either freely or commercially:
• SWMM (US EPA’s Storm Water Management Model) – Huber (1995). This is one of the first models developed, with a high degree of physically based principles incorporated. Its initial versions (still in frequent use in its original main frame version) have served as a basis for development of the other models which have taken advantage of later development of personal computer technology. • Hydroworks (HR Wallinford –Wallingford Software) The latest versions of the package are user orientated and can be used for matching with data sources and in composition of reports. • MOUSE (DHI –Danish Hydraulics Institute-1990) Broadly used internationally. The developers have made an effort to incorporate some of the developments of PC 24
technology (for example data base management in network simplification). A discrepancy is noticed between versatile pipe flow model and surface runoff one which would benefit from upgrading and proper matching with GIS and surface flooding routines. • Hystem Extran (ITWH – Fuchs and Scheffer ( 1990) • Bemus (IRTCUD, Djordjevic et al 1998) However the models seem to have reached a level in which most of the model developers seem to have lost enthusiasm for further upgrade and improvement of models’ capability in dealing with complexity of urban environment. Adding powerful graphic and colourful images does not contribute to the reliability of modelling as long as the upgrade of the physical background is not improved. In addition to the above specific particular features of particular climate regions, the following aspects need to dealt with in either model development or customising for application in particular climate conditions:
• Capturing, filtering, compaction and processing of high spatial resolution data (primarily obtained by remote sensing. These data would enable a better representation of terrain and land use) and its matching with GIS tools, the use of which could enhance the analytical power of the models. • Analysis of the effect of maintenance and management practices (de-icing, sewer flushing, gullyspot cleaning, street sweeping and of the other storm runoff and quality relevant activities) on water quality • Analysis of the effects of source control practices • Surface flooding (interactions of surcharged underground network with superficially flooded areas, flood risk analysis) These new incentives seem to be needed for a significant breakthrough to be made. This publication aims to provide some material which could serve as a guideline for development of new generation of models or improvement of the existing ones.
1.6 GIS and informatic support Geographical Information Systems are know to deal with acquisition, processing and implementation of data of a spatial nature (Boroughs 1986). Despite significant progress being made in this technology and its application in various water and environmental engineering fields, their application in urban drainage is still relatively limited. Significant progress has been made in the use of GIS based data in creation of data bases linked urban water infrastructure system simulation models (for example AquaBase – Kuby 1998). For the creation of initial data sets (GIS layers) various sources of data can be used (Figure 1.10). The systems are extremely powerful in providing input data to models after the elementary manipulations with layers presenting physical features of the catchment (such as elevation model and land use) and superficial and underground network have been performed. Starting in the late eighties, with some of the first papers on GIS application in urban drainage - Elgy et al (1993), the research group of the present author has developed a methodology for handling arbitrary data sources and automatic creation of input files for storm drainage modelling. An example of data preparation for creation of input files for catchment delineation (Maksimovic (1995) is given in Figure 1.11. The results of application of catchment delineation is presented in Figure 1.12. Figures 1.13 present the results of application of GIS functionalities in the analysis - assessment of the suitability of a catchment of implementation of source control techniques and Figure 1.13 depicts the results of the application of this analysis in the survey of the applicability of source 25
control in the same catchment (Macropoulos et al 1998 and Macropoulos et al 1999). The works of Prodanovic (1999) and Djordjevic et al (1998) provide further development towards GIS assisted physically-based flood modelling in urban areas based on the dual drainage concept. There is a huge unexploited potential of GIS application in particular climates. In the individual chapters, authors present current techniques in data analysis and modelling. Most of the specific features of the urban catchment in particular climates are of a spatial nature which renders them particularly applicable to quantification by GIS (e.g. suitable for application of GIS. It can be used in quantification of both physical features (such as soil propensity characteristics, soil erosion, pollutant potential distribution, snow cover, asphalt temperature, solar radiation exposure). These and other GIS applications are yet to be researched and made a part of the daily routine.
1:2000
1:2000
•• Paper Paper maps digitizing digitizing
•• Satellite Satellite images Satellite images
•• Video Video images Video images
•• GPS GPS data GPS data
•• Photogrammetry Photogrammetry Photogrammetry
•• Digital Digital data from Digital from total stations total stations
•• Dynamic Dynamic positioning & bathimetry Dynamic & bathimetry data bathimetry data data
Figure 1.10. Sources of data for GIS applications
Paper maps Areal photographs etc. existing or DEM -- corrected Land cover + network etc.
Pre processing of primary data and creation of secondary files for subcatchment delineation
Slopes Aspects Slope threshold Flow angles Subcatch. boundaries etc. Input files for models
26
Figure 1.11. Pre processing and post-processing of data for catchment delineation
27
Fig. 1.12. GIS supported catchment delineation (Maksimovic et al 1994)
Fig. 1.13. Suitability of the Klisa catchment for application of infiltration techniques
28
Fig. 14. Reduction in maximum water level for 10 years return period rainfall
1.7. Concluding remarks and acknowledgement The material presented in the present three volumes is result of the team work of numerous specialists gathered around the UNESCO IHP V programme under the theme 7: Integrated Water Management in Urban Areas within the Theme 7.3. Urban Drainage in specific climates. The series of the three volumes has been produced under the co-ordination role of the regional IRTCUD (International Research and Training Centre on Urban Drainage) units for particular climate regions : humid tropical in Brazil, cold in Norway and arid and semi arid in Sharjah. The production of the present volumes would be impossible without UNESCO’s endorsement and co-ordination roles of the key co-editors: Prof. Carlos Eduardo Morelli Tucci (for HT volume), Dr. Sveinnung Saegrov, Mrs. Jadranka Milina (MSc) and Prof. Sveinn T. Thorolfsson (for CC volume) and Prof. Mamdouh Nouh (for ASA volume). Thanks are due to the contributing authors of the chapters in individual volumes. Their names are listed in the relevant volumes. It is sincerely hoped that that publication of these three volumes will encourage further research and development in those regions in which there is still much to be learned about the governing physical processes and in which the most appropriate sustainable solutions can be found to the problems of urban flooding, storm water quality management, amenity development, provision, enhancement and resources recycling.
29
Bibliography
BOROUGHS, P. A. (1986) ‘Principles of Geographical Information Systems for Land Resources Assessment’ . Oxford Science Publications, Oxford
BUTLER, D.; C. MAKSIMOVIC (1999) ‘Urban Water Management; Challenges for the next Millennium’ . Progress in Environmental Sciences, Vol. 1 No. 3 p. 213-235 CIRIA (1999), ‘Sustainable urban drainage systems; Design manual for Scotland and Northern Ireland’ Report No. C521, Construction Industry Research and Information Association, London DELETIC, A.; C. MAKSIMOVIC; M. IVETIC (1997) ‘Modelling of Storm wash-off of Suspended Solids from Impervious Areas’. Journal of Hydraulic Research , Vol. 35, No. 1, p. 99-117. DELETIC, A. (1998) ‘The First Flush Load of Urban Surface Runoff’. Water Research, Vol. 32 No. 8, p. 2462-2470. DELETIC, A. (1998) ‘Sediment Behaviour in Grass Filter Strips’. Water Science and Technology, Vol. 39, No. 9, p. 19-136. DHI (1990) ‘MOUSE Modelling of Urban Sewer Systems on Microcomputers; Users Guide and Technical Reference’. Edition 3.0. DJORDJEVIC, S.; C. MAKSIMOVIC ; D. PRODANOVIC (1998) ‘An Approach to Simulation of Dual Drainage’. Water Science and Technology, Vol. 39, No. 9, p. 95-104. ELGY. J., C. MAKSIMOVIC; D. PRODANOVIC (1993A) ‘Matching Standard GIS Packages with Urban Storm Drainage Simulations Software’. Sixth International Conference on Urban Storm Drainage, Niagara Falls, Canada ELGY. J., C. MAKSIMOVIC; D. PRODANOVIC (1993B) ‘ Using Geographical Information Systems for Urban Drainage Hydrology’. Proc. of the International Conference on Application of Geographical Information Systems in Hydrology and Water Resources, HydroGIS 93. K. Kovar and H. P. Nachnebel (ed) Vienna, Austria FUCHS. L.; C. SCHEFFER (1991) ‘HYSTEM-EXTRAN, Version 4.3, Model Description and Documentation, Microcomputer in Urban Drainage’. Inst. fuer Technishe and Wissenschaftlische Hydrologie,Hanover. GEIGER, W. (1998) ‘Principles of Integrated Water Management for the Revival of Old Industrial Areas’. Springer Verlag, ASI Series, p. 57- 106 , Editors: A. G. Buekens and V. V. Dragalov. Geiger, W. F. and M. Becker (1999) ‘Revisiting the Past? New Approaches to Urban Drainage in the Emscher Area’, Proc. of Engineering Foundation Conference: Sustainable Urban Water Resources in the 21 st. Century, Sept. 7-12, Malmo, Sweden, Publ. ASCE, Editors: Ch. Rawney, P. Stahre, L. A. Roesner, pp 123 -135 Haestad Method (1997): Computer Applications in Hydraulic Engineering. 166 p. HUBER, W. (1995) ‘EPA Storm Water Management Model – SWMM’. In V.P. Singh (ed .) Computer Models in Watersheshead Hydrology, Water Resources Publication, pp. 783- 808 IWASA, Y. ; K. INOUE (1987) ‘Comparative Study of Flood-Runoff Analysis in View of Disaster Research’. Proc. US-Asia Conference on Engineering for Mitigation Natural Hazard Damage IZZARD, C. F. (1946) ‘Hydraulics of Runoff form Developed Surfaces’, Proc. of Highway Research Board p. 129-150. 30
KUBY, R. (1998) ‘Common Ground for GIS Support to Water Supply, Sewerage and Drainage (AquaBase)’Training course: GIS in Urban Water, Internal publication of Imperial College and CUWUK, MAKSIMOVIC, C. and M. RADOJKOVIC (1986) ‘ Urban Drainage Modelling’ , Proc. Int. Symp UDM’86 Dubrovnik, Yugoslavia , Pergamon Press 540 p. MAKSIMOVIC, C. , D. PRODANOVIC, J. ELGY; L. FUCHS (1994) ‘GIS (or GIM) in Water Projects – Tools or Toys’. Proc. International Conference of Hydroinformatics Deft, Balkema MACROPOULOS, C., C. MAKSIMOVIC; D. BUTLER (1999) ‘Effectiveness of Storm Drainage in urban Flood Risk Reduction under Source Control Implementation’. Fifth International Mirce Symposium: System Operational Effectiveness 1-3 December (in press) MACROPOULOS, C.,D. Butler and C. MAKSIMOVIC (1998) ‘GIS supported evaluation of source control applicability in urban areas’. Water Science and Technology, Vol. 39 No. 3, p.243-252 MAKSIMOVIC, C., L. FUCHS, D. PRODANOVIC and J. ELGY (1995) ‘Full Scale Application of Standard GIS Packages with Urban Storm Drainage Simulation Software’. Kurier Abwasserung Germany, Book 1, p. 3-13 MAKSIMOVIC, C., D. BUTLEr and N. GRAHAM (1999) ‘Emerging Technologies in the Water Industry’. In Water Industry Systems; Modelling and Optimization Applications, Ed. D. Savic and G. Walters, Research Studies Press, p. 39-64. NIEMCZYNOWICZ, J. (1996) ‘Challenges and interactions in water future’. Environmental Research Forum 3 4, 1-10, Transtec Publications, Switzerland. O’LOUGHLIN, G., W. HUBER and B. CHOCAt, (1996) ‘Rainfall-Runoff Processes and Modelling’, Journal of Hydraulic Research, Vol. 34., No 6., p. 733-752. OSTROWSKI, M.W. and W. JAMES, (1998) ‘Requirements for group decision support systems for urban stormwater management’. Fourth Int. Conference on Developments in Urban Drainage; UDM’98, London PRODANOVIC D. (1999) ‘Unapredjenje metoda primene hidroinformatike u analizi oticanja sa urbanih povrsina’ (in Serbian) (Improvements of application methods in analysis of runoff from urban areas), PhD Thesis presented at the University of Belgrade, Yugoslavia, November 1999, p 196. SIMON, A. (1996) ‘Estimation of the Utilisation of Possibilities of Overland Flow in Urban Areas in Israel’s Coastal Plain’. In Integrated Water Management in Urban Environment Edited by J. Niemczynowicz, Trenstech Publications, p. 221- 232. STAHRE, P. (1999) ‘Ten Years of Experiences of Sustainable Stormwater Management in the City of Malmo Sweden’. Proc of 8 th International Conference on Urban Storm Drainage, Sydney, Australia, p.1087 – 1097. XANTHOPUOLOS, C. and H. HAHN (1993) ‘Anthropogenic Pollutants Wash-off from Street Surface’. Proc. Of the Sixth Int. Conference on Urban Storm Drainage, Niagara Falls, Canada YEN B. C., (1986) ‘Rainfall–Runoff Process on Urban Catchment and its Modelling’. Invited lecture at Int. Symp. UDM’86 Dubrovnik, Yugoslavia , Pergamon Press (Ed. Maksimovic & Radojkovic)
31
