Water Quality Management in Rivers Using Water Quality Models for Assessment and Prediction

Water Quality Management in Rivers using Water Quality Models for Assessment and Prediction

Sudhira H. S.1, Praveen Gurukar S.2, Anurita3 and Lokesh K. S.4

1 Energy and Wetland Research Group, Centre for Ecological Sciences, Indian Institute of Science, Bangalore –560 012

2 JSS Consultants, SJCE Campus, Manasagangothri, Mysore – 570 006

3 Aspire Communications, SJCE Campus, Manasagangothri, Mysore – 570 006

4 Department of Environmental Engineering, Sri Jayachamarajendra College of Engineering, Mysore – 570 006. Karnataka. India.

Water quality assessment in rivers has attained considerable importance in recent years. This is because of the deterioration of river water quality due to pollution from different sources attributed to human activities on the pretext of economic development.Water quality modeling and subsequent prediction can be one of the most important potential tools for water quality management in rivers. In this regard, the current paper highlights the importance of water quality assessments in rivers with a thrust on management.

This paper presents a study on water quality assessment and modeling studies undertaken for river Cauvery at Srirangapatna. The water sampling was done at six transects along the river stretch over a distance of 5 km. Water quality monitoring was done over a period of three months extending from February 2002 to April 2002. The water quality parameters and upstream flow characteristics were analyzed for review and validation of available water quality model under steady-state conditions.

Water quality model developed by Gurudatt (1989) was applied in the present study. In particular the model MIXPIPOX for single effluent discharge is used here. The river water quality and hydrological characteristics were used as inputs for this program. The low river flow analysis was also carried to determine the design flow for the least flow in the river representing the critical flow conditions. These data were further used to calibrate and validate the MIXPIPOX computer model for the conditions prevailing in the river. The non-dimensional diffusion factor value (was arrived at by using the conservative parameters (conductivity, TDS) for model calibration and validation was done using non-conservative parameter (BOD).

The decay rates, reaeration coefficients and  values for each transect was used to compare the observed and predicted values by plotting graphs for the same. From these plots, it was found that observed and predicted values correlate well for three transects and agree with the trend for the rest of the transects. Finally, the study evaluated the viable engineering option of setting up effluent treatment plant for river Cauvery near Srirangapatna considering the  rates, decay rates and reaeration coefficients. This was carried for the options of providing primary and secondary treatments only. From the water quality assessments the extent of pollution in the river is not very significant. The paper discusses some aspects accounting for this phenomenon as well as other options available for safe and efficient disposal of wastewaters.

Introduction

Environmental pollution control has evolved as one of the major themes among decision makers and planners, while posing a greater challenge to engineers and scientists to understand and analyze various phenomena involving this. In general it is hard to define and quantify many of the important physical, chemical, biological, economic and social interrelationships, between the many components of any environmental pollution control system. Constraints in manpower, technical know-how and economic resources have led to lesser understanding of such systems.

River systems are subjected to undue pollution loading in the form of industrial and domestic discharges. Prevention and control of pollution to rivers in India often follow the obsolete “end-of-pipe” treatment methods. In order to assess the impacts of wastewater discharge on the receiving water body’s quality, computer aided mathematical models, once calibrated for existing river quality conditions can be used to generate future scenarios of water quality due to increased pollution loads, and also simulate alternative wastewater treatment scenarios, thus forming an integral part of the decision making system for water quality management.

Water quality models are thus evolved in order to predict and assess various conditions that may prevail in the water body due to effluent discharges. Water quality models are mathematical constructs, which integrate a number of complex phenomena such as water transport, reaction kinetics, and external loadings. There are two basic reasons for constructing mathematical representations of natural aquatic ecosystems. First, there is a need to increase the current level of understanding regarding the cause-effect relationships operative in all aquatic environments. Secondly, models provide a synthesis of understanding, which is increasing in the policy arena. Mathematical models, which incorporate the characteristics of river channel, outfall and pollutants of concern, are utilized as tools in such analyses.

Water quality simulation models are used extensively in waste load allocation studies, environmental impact investigations and for analyzing the cause and effect relationships in a water body. A common feature in most of these models is the use of rate parameters to describe processes occurring in the water body. The reliability of the model is a function of, among other things, how well these parameters reflect the processes they are intended to describe.

The processes involved in modeling are model calibration, model verification, model application and model post audit. Calibration is the process of identifying appropriate values for model parameters and is a difficult task because of uncertainty in the mathematical expression of the system (model structure identification), the inability to control environmental experiments, and uncertainty in field data (Beck 1983). The expert system or the expert advisor projects undertaken by the Environmental Protection Agency (EPA) at the Center of Exposure Assessment Modeling recognized calibration as the single most difficult step in modeling water quality.

Models of water quality have evolved over the course of time in response to various issues. This evolution has included increased complexity in the number of aquatic processes that have been included like – nitrogen and phosphorus cycling and interaction with primary and secondary trophic levels, toxic chemical fate and transport processes of solids partitioning and biodegradation and volatilization, and aquatic food web chemical bioaccumulation models. But there has been more significant development over time that has resulted from the management issues of controllability. From the perspective of the need to incorporate external causative inputs into the overall modeling framework, three different stages of water quality modeling are identified (Thomann, 1998).

One Dimensional DO Modeling in Rivers

Dissolved Oxygen (DO) is an important water quality parameter affecting the health of a river and hence, great deal of importance is attached to maintain the DO at desirable level. DO is important to aquatic life because detrimental effects can occur when DO levels drop below 4 mg/L or 5 mg/L, depending on the aquatic species. Suspended solids influence the water column turbidity and ultimately settle at the bottom, leading to possible benthic enrichment, toxicity and sediment oxygen demand. Nutrients can lead to eutrophication and DO depletion. Thus, in order to evaluate the effects of wastewater discharge on instream DO level, it is necessary to understand the relationship between pollutant characteristics and stream environment.

The important factors and processes affecting DO include the following:

• Decay of Carbonaceous and Nitrogenous BOD
• Atmospheric Reaeration
• Photosynthesis and Respiration
• Sediment Oxygen Demand

The one dimensional DO analysis in rivers are based on several assumptions (Thomann and Mueller, 1987). The important assumptions pertaining to the DO model are:

• Steady state conditions prevail with the exception of photosynthetic activity, which is assumed to follow a sinusoidal curve over any 24-hour period.
• The stream flow is mainly unidirectional (occurring in longitudinal direction).
• Complete mixing of input loads (wastewater, tributary) with the stream waters occurs instantaneously at the inflow locations.
• Longitudinal dispersion effects are negligible.

Streeter Phelps Equation

The earliest and most familiar DO balance relationship is the Streeter Phelps equation, which considered the BOD decay and the atmospheric reaeration mechanisms. The equation is given by –

D = Do exp(-Kat) + {KdLo/(Ka-Kr)}{exp(-Krt) – exp(-Kat)} …Equation 1

where,

D = DO Deficit = Cs – C, mg/L;

Cs=Saturation DO Conc., mg/L; C= Ambient DO Conc., mg/L.

Ka = Reaeration Rate; Kd & Kr = Decay Rate Coefficients

Lo = Upstream BOD in mg/L.

The distribution of DO with downstream distance, given by this equation is termed as “Oxygen Sag Curve”. The DO sag curve is as shown below.

Figure 1: Dissolved Oxygen Sag Curve

It is seen that the DO concentration, C, reaches a minimum at a location termed as critical location. At this point, oxygen uptake by BOD is just balanced by the oxygen input from the atmosphere.

Generalized Steady State Model

As outlined by Thomann and Mueller (1987), the DO balance equation can be written as follows for steady state conditions by using the daily average photosynthetic rate, Pav (instead of the diurnal DO variation):

dD/dt = KdL + KnN – KaD – Pav + R + Sv …Equation 2

However over the years analysts have used the general steady state model for water quality modeling studies using computer programs for analysis.

Computer Programs for Water Quality Modeling

Computer programs have been developed to aid in the data analysis, and for model validation and application to evaluate viable engineering and management options. The programs written in FORTRAN include algorithms to consider the effect of background concentration of pollutants. These are based on the concept of mixing zones in rivers and methods developed by Gowda (1980, 1984a and 1984b). The program MIXPIPOX has been developed to simulate the lateral and longitudinal distributions of CBOD and DO in river channels receiving effluents from pipe out fall at bank or in river channel (Gurudatt, 1989). There is only one outfall in the study stretch of the river and hence, the computer program MIXPIPOX applicable to DO evaluation in rivers having single outfall is utilized for the DO model evaluation.

Objectives

This paper brings out the study to assess the quality of river water and quantitatively predict the effect of polluting discharges on river quality by evaluating the existing model using field data. This paper explores the application of MIXPIPOX model for predicting DO in river Cauvery, at Srirangapatna where the river is polluted by domestic wastewater from a pipe outfall. The specific objectives of the study were:

• To monitor and evaluate river water quality due to effluent discharges into river Cauvery at Srirangapatna.
• Low flow analysis for the design conditions.
• Review and validation of available water quality models under steady-state conditions.
• To calibrate and validate the model using field data.
• To apply the model for evaluating viable engineering options.

River Cauvery is an important source for agriculture, drinking water supply and other purposes. For the present study river Cauvery was chosen near Srirangapatna town, since at present there is no wastewater treatment facility in this town. The domestic wastewater is directly discharged into the river, due to which the river water is experiencing significant pollution.. Hence, the stretch of river Cauvery near Srirangapatna town has been selected for a detailed water quality assessment and modeling studies.

Description of Study Area

The domestic wastewater from the Srirangapatna is discharged into the River Cauvery. Srirangapatna is an island formed by the branches of river Cauvery. The geographical location of Srirangapatna is 76o 41’ longitude and 12o15’ latitude. The Sriranganatha temple and Nimishamba temple are located on the banks of this river. This attracts lot of tourist all round the year and hence there is a high floating population. There is a Bathing Ghat close to the Sriranganatha temple at the north branch of the river. The sewage effluent of the town is also let into north branch of the river. Further, downstream of the river, due to the presence of Bathing Ghat close to the famous Nimishamba temple, the river is again subjected to contamination. The study stretch of the river is shown in Figure 2. This necessitated a detailed water quality evaluation both upstream and downstream of the river for an assessment and model application.

The sampling sites were selected considering the following points:

• Objectives of the study
• Sources of pollution
• Physical characteristics of the stream
• Accessibility of the site
• Equipment availability

Water sampling was therefore done at upstream and downstream of bathing ghats and at various places along the river until the two branches of river confluences at Sangam. The water quality parameters (DO, BOD, Chlorides, pH, Temperature, Specific Conductivity) and upstream flow characteristics were all analyzed for review and validation of available water quality models under steady state conditions. These data were also used to calibrate and validate the model for the conditions prevailing in the river.

Figure 2: Study Stretch Showing the Drain and Transects for River Cauvery at Srirangapatna

The field surveys were designed to collect data on outfall discharge and background water quality characteristics and on the transverse distribution of various parameters of interest at six transects. A general field study procedure is described below.

The location of transects can be based on preliminary field measurement of a conservative parameter (e.g. conductivity, TDS) at selected access points to establish the approximate longitudinal boundary of the mixing zone. The selected locations of transects were marked on a map as shown in Figure 2. Cross sectional depth was measured at a minimum of 10 points at known lateral distance measured from a reference first outfall bank. Velocity measurement was also done by standard stream flow gauging procedure.

The following parameters were measured insitu.

1. Temperature: Temperature of the water body was measured by dipping a sensitive thermometer.
2. Dissolved Oxygen Demand (DO): Insitu measurement of DO content was made by dipping the probe of portable DO kit.
3. Physical Characteristics of the River: Width of the river at each transect was noted. Depth of the river at several points along individual transect was measured by dipping a graduated rod while traveling on a raft. Velocity of the flow in the river was measured using a rubber ball as a float, which was allowed to float between two points in the river. The distance between the two points and time of travel of the ball were recorded and the velocity of flow was calculated. Average velocity was determined by repeating the process two to three times.
4. Wastewater Flow: The flow of wastewater that enters the river from the Srirangapatna town was measured.
5. River Water Flow: Flow of river Cauvery during the study period was noted from the records of NationalRiver and Lake Conservation Directorate (NRLCD).

The geometric and hydraulic data are given in Table 1. The water samples transported to the laboratory were analyzed according to the procedures given in Standard Methods (1992) as mentioned below:

1. pH – Measured using a digital pH meter
2. Conductivity – Measured using a digital conductivity meter
3. TDS – Measured by digital TDS meter
4. BOD – By standard dilution technique
5. COD – Reflux method

The observed water quality data for the month of April 2002 are presented in Table 2.

Table 1: Geometric and Hydraulic Data of Study Stretch of CauveryRiver

Transects / A / B / C / D / E / F
Distance ‘x’ meters / 50 / 80 / 730 / 830 / 930 / 4930
Discharge in m3/sec / 70.54 / 70.54 / 70.54 / 70.54 / 70.54 / 70.54
Upstream River Flow m3/sec / 72.48 / 72.48 / 72.48 / 72.48 / 72.48 / 72.48
Effluent Flow Rate m3/sec / 0.06 / 0.06 / 0.06 / 0.06 / 0.06 / 0.06
Width ‘b’ meters / 75 / 80 / 65 / 65 / 95 / 80
Depth ‘h’ meters / 0.5 / 0.45 / 0.65 / 0.60 / 1.29 / 0.85
Velocity ‘v’ m/sec / 0.42 / 0.40 / 0.38 / 0.40 / 0.5 / 0.75

Table 2: Water Quality Parameters of CauveryRiver for the month of April 2002

Details / Distance
In meters / Temperature o C / pH / DO
mg/L / TDS
mg/L / COD
mg/L / BOD
mg/L / Conductivity
 mhos/cm
Effluent Discharge / - / 28.9 / 6.9 / 2.5 / 320 / 60.8 / 18.0 / 662
Upstream / - / 27.9 / 7.6 / 7.47 / 190 / 12.8 / 1.4 / 432
Transect A / 50 / 28.0 / 7.3 / 3.28 / 250 / 40.6 / 16.7 / 535
Transect B / 80 / 28.0 / 7.2 / 4.67 / 240 / 23.8 / 12.3 / 520
Transect C / 730 / 28.2 / 7.6 / 7.2 / 210 / 15.7 / 3.6 / 463
Transect D / 830 / 28.0 / 7.5 / 7.4 / 220 / 30.2 / 2.3 / 456
Transect E / 930 / 28.2 / 7.6 / 7.2 / 220 / 28.0 / 1.0 / 452
Transect F / 4930 / 28.4 / 7.8 / 6.8 / 210 / 19.2 / 1.4 / 452

Stream Flow Analysis

Changes in water quality parameters and the assimilative capacity of a river are essentially dependent on the flow of the river at any given time. Critical water quality condition occurs during low flow periods. The low river flow analysis was carried to determine the design flow for the least flow in the river representing the critical flow conditions. This critical flow is used to determine the impact of wastewater on the river. Hence, it is essential to conduct a stream flow analysis exercise to determine the minimum flow of the river.

The average minimum flows in the River Cauvery during 2001-2002 are presented in Table 3. In the present study the most critical flow conditions were determined by subjecting the average minimum flows over the period mentioned above to probability analysis. The data were arranged in ascending order and ranked. The probability of occurrence of each flow (p) was determined from:

P = m/(n+1) …Equation 3

where; ‘m’ is the rank of each flow and n is the total number of observations. The values of probability (p) were plotted against the corresponding values of discharge (Q), shown in Figure 3. From this plot, the flow with 90% probability of occurrence was obtained to be 72.48 cumecs just upstream of the effluent outfall. This represents the low flow in Cauvery river at Srirangapatna. In this study, flow increase due to seepage from the catchment area is ignored which means that the low flow considered will be a conservative estimate.