Modeling chlorine residuals in Water Distribution Network Using EPANET 2.0 software.

Transcription

1 Modeling chlorine residuals in Water Distribution Network Using EPANET. software. Ms. Kashfina Kapadia* *Reader in Acropolis institute of technology and research Indore. 1. Abstract: In this paper water quality modeling in water distribution network is done using EPANET. software. Two networks one with single loop network and the other one with two source two looped networks are discussed and steady state chlorine concentration at different nodes are determined for maintaining a particular concentration of chlorine. Chlorine is added at the reservoir only and Chlorine concentration at the nodes is determined.. Keywords: Water quality, Water distribution network, chlorine concentration.. Abbreviations: PBA= Particle back tracking algorithm. EDM= Eulerian Finite-Difference Method DVM =Eulerian Discrete Volume Method TDM =Lagrangian Time-Driven Method EDM =Lagrangian Event-Driven Method 4. Introduction: The quality of treated water after leaving the treatment plant deteriorates as water travels through a water distribution network. In recent years there is a growing interest in the multi quality water systems. The interest in quality comes from three types of circumstances: 1. Use of waters from sources with different qualities in a single system, which serves to mix and convey them.. Concern over quality changes, as water flows through the network, due to decay and growth of various chemical and biological constituents.. Accidental entry of low-quality water into drinking-watersupply systems. Some of the factors that influence the changes in water quality between the treatment plant and consumer tap include (1) Chemical and biological quality of source water. ()Effectiveness and efficiency of treatment process.() integrity of the treatment plant, storage facilities and distribution system (4) age, types design and maintenance of distribution network (5) quality of treated water and (6) mixing of water from different sources within a distribution network and other hydraulic conditions. The filtered water which is obtained from filters may contain some harmful disease causing bacteria in it. These bacteria must be killed in order to make the water safe for drinking. The chemicals used for killing these bacteria are known as disinfectants and process is known as disinfection. This process of purification is the most important because the bacterially contaminated water may lead to spread of various diseases and their epidemics, thus causing disaster to public life. The disinfection can be done with the help of bromine, iodine, Ultra violet rays, potassium permanganate or treatment with excess lime, but chlorination is most commonly used method of disinfection in community water supply schemes because it is cheap, easy to use, and effective in killing the bacteria and can be maintained as residual. Presence of residual chlorine in detectable quantity at consumer taps assures water free from bacteriological contaminations. The ultimate objective of distribution network is to provide the sufficient quantity of water at required pressure. The quality of water in distribution network is the most important criteria. End user should get the good quality water. 5.1 Chlorine Transport: Advection (movement in the direction of flow) and dispersion (movement in the transverse direction due to concentration difference) are two important mechanisms for transportation of residual chlorine. The basic equation describing advection-dispersion transport is based on principle of conservation of mass and Fick s law of diffusion. For a non conservative substance, the principle of mass conservation within a differential section of a pipe i.e. control volume can be written as given in the equation below: Cxt (, ) Cxt (, ) Cxt (, ) = u + E ± K [ (, )] R C x t t x x

2 Considering the first order reaction of chlorine with other substance the conservation of mass is given by the equation in which C(x, t) is concentration of chlorine at location x at time t, u= velocity in the pipe and E= coefficient of longitudinal dispersion, K R [C(x, t)]= a reaction rate constant in which (-) sign is used for reflecting decrease in concentration due to decay rate. The first term in this equation represents the rate of change of chlorine within a differential section of pipe, the second term accounts for advection flux of the chlorine in which (-) sign reflects an increasing concentration within the control volume if the mass inflow rate is greater than the outflow rate and third term accounts for dispersive flux of chlorine and fourth term accounts for chlorine reaction and it is negative. Dispersion of chlorine is negligible if the velocity is high and dispersion coefficient is very low in such a case the third term can be neglected. 5. Chlorine decay: The mechanism of chlorine decay in pipe has two dimensions. The first dimension is the reaction of chlorine with substances present in water. This decay of chlorine is known as bulk decay. The second dimension is the reaction of chlorine with substances present on pipe wall. Pipe wall get frequently coated with variety of scales whose compositions depend on pipe type source water quality and treatment technique. This decay of chlorine is known as wall decay. The wall decay in distribution network may be predominant where significant corrosion is present Bulk decay rate constant: Bulk decay is measured by recording the chlorine concentration, at time intervals, from glass bottle that have been previously filled with the sample water. Bulk decay rate constant is mostly assumed to first-order kinetics. Bulk decay coefficients for chlorine depend very much on the nature of the source water and the treatment it has received. "Typical" values range from.1 to 1. (per day). 5...Wall decay rate constant: Wall decay of residual chlorine is due to disinfection of bio-film and reaction with corrosion by-product. Chlorine decay rate caused by bio-film is normally much smaller that caused by corrosion by products. Wall decay rate constant is a function of velocity, pipe diameter, pipe length, diffusivity, and viscosity. Wall decay rate constant can differ significantly from pipe to pipe even in the same network. 5...Overall decay rate constant: A simple method to define the overall decay rate constant as sum of bulk and wall decay rate constant. 5..Mixing in distribution network: Mixing at junction is assumed that complete and instantaneous and detention time inside the junction is assumed to be negligible. C i x= = j I k Q C j j x= L j s s j Ik Q j + Q + Q C Where i = link with flow leaving node k; Ik = set of links with flow into k; Lj = length of link j; Qj = flow in link j; Cs = external source concentration entering node k; and Qs = external source flow entering node k. Cj= Existing concentration of chlorine. 6. Static model of Cl concentration: Chlorine concentration reaches a steady state after a certain time period. This time period is different at different nodes. In static or steady state model it is assumed that steady state is reached at all nodes and thus initial variation of chlorine concentration is ignored for predicting chlorine concentration at different nodes. Initially hydraulic analysis of the network is carried out to obtain pipe discharge. From these discharges velocity of flow and travel time in each pipe is computed. Concentration at upstream of a pipe: C xu = C j for x= 1 X. Then chlorine decay equations are written and then mass conservation equations and written. In steady state condition, booster chlorination stations can be provided in water distribution network to boost the chlorine concentration to the specified level or by specified amount. Thus at the booster point either final chlorine concentration is known or amount of chlorine added is known in analysis and accordingly equations are modified to obtain unknown chlorine concentrations Dynamic Model of Cl concentration: Dynamic model for predicting chlorine concentration determines movement and spread of chlorine under time varying conditions. Since water demands and operation of various elements in the network are time dependent, dynamic model provides more accurate prediction of chlorine concentration. s Eulerian Finite Difference method:

3 Eulerian FDM approximates the derivatives in equation with their finite difference equivalents along a fixed grid of points in time and space Eulerian discrete volume method: In this method each pipe is divided into series of equally sized, completely mixed volume segments. At each successive water quality time step the concentration within each volume segment is reacted according to reaction law and then transferred to adjacent downstream segment Lagrangian Time driven method: This method tracks the concentration and size of series of non overlapping segments of water that fills each pipe of the network. The size of the most upstream segment in a pipe increases as water enters the pipe while equal loss in the size of the most downstream segment occurs as the water leaves the pipe Lagrangian Event driven method Lagrangian Event driven method is similar to TDM except that it updates the network only when leading segment in some pipe is consumed at the pipe s downstream node Particle Back Tracking Algorithm. Particle Back Tracking Algorithm is a Lagrangian model developed for chlorine transport in networks without storage tank. 8.1 ABOUT EPANET. Software. EPANET performs extended period simulation of hydraulic and water quality behavior within pressurized pipe networks. A network consists of pipes, nodes (pipe junctions), pumps, valves and storage tanks or reservoirs. EPANET tracks the flow of water in each pipe, the pressure at each node, the height of water in each tank, and the concentration of a chemical species throughout the network during a simulation period comprised of multiple time steps. In addition to chemical species, water age and source tracing can also be simulated. Running under Windows, EPANET provides an integrated environment for editing network input data, running hydraulic and water quality simulations, and viewing the results in a variety of formats. These include color-coded network maps, data tables, time series graphs, and contour plots. EPANET was developed by the Water Supply and Water Resources Division (formerly the Drinking Water Research Division) of the U.S. Environmental Protection Agency's National Risk Management Research Laboratory. 8. Water quality modeling capabilities In addition to hydraulic modeling, EPANET provides the following water quality modeling capabilities: Models the movement of a non-reactive tracer material through the network over time. Models the movement and fate of a reactive material as it grows (e.g., a disinfection by-product) or decays (chlorine residual) with time. Models the age of water throughout a network tracks the percent of flow from a given node reaching all other nodes over time. models reactions both in the bulk flow and at the pipe wall. uses n th order kinetics to model reactions in the bulk flow uses zero or first order kinetics to model reactions at the pipe wall accounts for mass transfer limitations when modeling pipe wall reactions Allows growth or decay reactions to proceed up to a limiting concentration. employs global reaction rate coefficients that can be modified on a pipe-by-pipe basis allows wall reaction rate coefficients to be correlated to pipe roughness allows for time-varying concentration or mass inputs at any location in the network models storage tanks as being either complete mix, plug flow, or two-compartment reactors. By employing these features, EPANET can study such water quality phenomena as: Blending water from different sources, Age of water throughout a system. Loss of chlorine residuals. Growth of disinfection by-products. Contaminant propagation events. 8. Water quality analysis by EPANET: EPANET s water quality simulator uses a Lagrangian timebased approach to track the fate of discrete parcels of water as they move along pipes and mix together at junctions between fixed-length time steps. These water quality time steps are typically much shorter than the hydraulic time step (e.g., minutes rather than hours) to accommodate the short times of travel that can occur within pipes. However, as with hydraulics, results are only reported at the end of each userspecified reporting time step. 8.4 Water quality sources: Water quality sources are nodes where the quality of external flow entering the network is specified. They can

4 represent the main treatment works, a well-head or satellite treatment facility, or an unwanted contaminant intrusion. Source quality can be made to vary over time by assigning it a time pattern. EPANET can model the following types of sources: A concentration source fixes the concentration of any external inflow entering the network at a node, such as flow from a reservoir or from a negative demand placed at a junction. A mass booster source adds a fixed mass flow to that entering the node from other points in the network. A flow paced booster source adds a fixed concentration to that resulting from the mixing of all inflow to the node from other points in the network. A set point booster source fixes the concentration of any flow leaving the node (as long as the concentration resulting from all inflow to the node is below the set point). The concentration-type source is best used for nodes that represent source water supplies or treatment works (e.g., reservoirs or nodes assigned a negative demand). The booster-type source is best used to model direct injection of a tracer or additional disinfectant into the network or to model a contaminant intrusion. 8.5 Parcel Tracking Algorithm: EPANET tracks the change in water quality of discrete parcels of water as they move along pipes and mix together at junctions between fixed-length time steps. The following actions occur at the end of each such time step: 1. The water quality in each parcel is updated to reflect any reaction that may have occurred over the time step.. The water from the leading parcels of pipes with flow into each junction is blended together, along with any external inflow to the junction, to compute a new water quality value at the junction. The volume contributed from each parcel equals the product of its pipe s flow rate and the time step. If this volume exceeds that of the leading parcel then the leading parcel is destroyed and the next parcel in line behind it begins to contribute its volume.. New parcels are created in pipes with flow out of each junction. The parcel volume equals the product of the pipe flow and the time step. The parcel s water quality equals the new junction value computed in Step. To cut down on the number of parcels, Step is only carried out if the new junction quality differs by a user-specified tolerance from that of the last parcel in the outflow pipe. If the difference in quality is below the tolerance then the size of the last parcel is simply increased by the volume of flow released into the pipe over the time step with no change in quality. The updating of water quality in storage tanks receives special treatment (see Tank Mixing Models). Initially each pipe in the network consists of a single parcel whose quality equals the initial quality assigned to the upstream node. Whenever there is a flow reversal in a pipe, the pipe s parcels are reordered from front to back. 8.6 Modeling water age: Water age is the time spent by a parcel of water in the network. It provides a simple, non-specific measure of the overall quality of delivered drinking water. New water entering the network from reservoirs or source nodes enters with age of zero. As this water moves through the pipe network it splits apart and blends together with parcels of varying age at pipe junctions and storage facilities. EPANET provides automatic modeling of water age. Internally, it treats age as a reactive constituent whose growth follows zero-order kinetics with a rate constant equal to 1 (i.e., each second the water becomes a second older). 8.7 Source tracing: Source tracing tracks over time what percent of water reaching any node in the network had its origin at a particular node. The source node can be any node in the network, including storage nodes. Source tracing is a useful tool for analyzing distribution systems drawing water from two or more different raw water supplies. It can show to what degree water from a given source blends with that from other sources, and how the spatial pattern of this blending changes over time. EPANET provides an automatic facility for performing source tracing. The user need only specify which node is the source node. Internally, EPANET treats this node as a constant source of a non-reacting constituent that enters the network with a concentration of Water quality reactions: The water quality module of EPANET can track the growth or decay of a substance by reaction as it travels through a distribution system. In order to do this it needs to know the rate at which the substance reacts and how this rate might depend on substance concentration. Reactions can occur both within the bulk flow and with material along the pipe wall. Bulk fluid reactions can also occur within tanks. EPANET allows a modeler to use different reaction rates for the two zones of reaction Bulk Flow Reaction Rates: Bulk flow reactions are reactions that occur in the main flow stream of a pipe or in a storage tank, unaffected by any processes that might involve the pipe wall. EPANET models these reactions using n-th order kinetics, where the

5 instantaneous rate of reaction (R in mass/volume/time) is assumed to be concentration-dependent according to R= K b C n q 1, C 1 Demand q where Kb = a bulk rate coefficient, C = reactant concentration (mass/volume), and n = a reaction order. Kb has units of concentration raised to the (1-n) power divided by time. It is positive for growth reactions and negative for decay reactions R= K b (C L -C) C n-1 EPANET can also consider reactions where a limiting concentration exists on the ultimate growth or loss of the substance. In this case the rate expression for a growth reaction becomes where CL = the limiting concentration. (For decay reactions (C L - C) is replaced by (C - CL).) Thus there are three parameters (Kb, CL, and n) that are used to characterize bulk reaction rates Pipe Wall Reaction Rates: In addition to bulk flow reactions, EPANET can model reactions that occur with material on or near the pipe wall. The rate of this reaction can be considered to be dependent on the concentration in the bulk flow by using an expression of the form where Kw = a wall reaction rate coefficient and (A / V) = the surface area per unit volume within a pipe (equal to 4 divided by the diameter). The latter term converts the mass reacting per unit of wall area to a per unit volume basis. EPANET limits the wall reaction order (n) to either or 1, so that the units of Kw are either mass/area/time or length/time, respectively. The parameter Kw appearing in the above rate expression should be adjusted to account for any mass transfer limitations in moving reactants and products between the bulk flow and the wall. EPANET does this automatically, basing the adjustment on the molecular diffusivity of the substance being modeled and on the flow's Reynolds number. (Setting the molecular diffusivity to zero will cause mass transfer effects to be ignored.) 9.1 Single source Network: q4 1 Pipe 1 4 Decay factor Ex = e -kt Q1 4 Q 4 Writing equations for concentration at upstream end of pipes 1 to 4 C 1u = C 1 C u = C 1 C u = C C 4u = C 4 Writing chlorine decay equations for pipe. C 1d = C 1u F 1 C d = C u F C d = C u F C 4d = C 4u F 4 Writing mass balance equations at source and demand nodes -C 1u Q 1 - C u Q = - C 1 q 1 C d Q -C u Q -C q = + = Q CdQ C4dQ4 Cq 1d 1 4u C Q + C Q + C q = q Writing the equations into matrix form and solving them, we get the nodal values of chlorine concentrations.

6 Q1+ Q C1 C1q1 EQ Q+ q C = EQ q EQ 4 4 C EQ 1 1 Q4 + q4 C4 9. Two source four demand looped Network: we can write down the equations in the matrix form as follows, by solving these equations by we get the concentrations at all the nodes. We can solve these simultaneous equations using MATLAB and compare the results by EPANET. Q1+ Q C1 C1q1 EQ 1 1 Q q C + EQ Q6 + q EQ 4 4 C ( Q+ Q4) Cn a = EQ Q4 + Q5 + q4 C4 EQ 5 5 Q7 + q5 C 5 EQ 6 6 EQ 7 7 q6 C 6 1. References: 1. Bhave P.R, Gupta R., (6), Analysis of Water distribution network Narosa Publishing House. Nagpur.. Bhave P.R,() Optimal Design of Water Distribution Networks. Narosa Publishing House.. Clark, R.M., (1998) Predicting chlorine residuals and formation of THMS in drinking water J Environ Engg. ASCE 14(1), Islam, M.R (1997). Inverse modeling of chlorine concentration in pipe network under dynamic conditions. J Environ Engg. ASCE 1(1), Islam, M.R (1998) Modeling of constituent transport in unsteady flows in pipe networks. J Hydraul engg. ASCE 14(11) Rossman L.A.(1994) modeling chlorine residual in drinking water distribution system. J Environmental engg. ASCE 1(4) Conclusions: from this analysis it can be concluded that EPANET. is a very good network modeling package that contain capability of water quality modeling. The water quality model can be used to assist managers to perform a variety of quality related studies. They can be listed as follows. Calibrating and testing hydraulic models of the system through the use of chemical tracers. Locating and sizing storage facilities and modifying system operations to reduce the age of water. Modifying the design and operations of the system to provide a desired blend of water from different sources. Finding the best combinations of pipe replacement, pipe relining, pipe cleaning, reduction in storage holding time and injection rate at booster station to maintain desired disinfection levels throughout the system. Assessing and minimizing the risk of consumer exposure to disinfection by products and Assessing the system s vulnerability to incidents of external contamination. 11.Acknowledgement Authors are thankful to Mr. P. R. Bhave and all the persons who have directly or indirectly helped or motivated to undergo this project work.