Modeling%20Water%20Quality%20In%20Drinking%20Water%20Distribution%20Systems:%20Its%20Potential%20for%20Enhancing%20Water%20Security

1
Modeling Water Quality In Drinking Water
Distribution Systems Its Potential for
Enhancing Water Security
2
Will Discuss

• Concern over the vulnerability of water
  distribution systems to security threats
• Basics of water quality modeling and its
  application in drinking water networks
• Two examples of water quality modeling for water
  security
• The future use of water quality modeling as part
  of a decision support framework

3
9/11 Raised Concerns About Critical
Infrastructure in the US

• Water supply was identified as critical
  infrastructure
• It is now general consensus that the
  vulnerability of drinking water networks systems
  to security threats is a major concern
• Utility industry has also recognized the
  importance of environmental monitoring in
  maintaining water security

4
Drinking Water Systems In The U.S.

• There are 54,000 community water systems in US
  serving 264 million people
• 79 of the population receives drinking water
  from large utilities (serving 10,000 or more),
  representing 14 of the systems
• 21 of the population receives water from small
  utilities (serving less than 10,000 people)
  representing 86 of the systems

5
U. S. Water Supplies Have Common Characteristics

• Water source
• A lake, reservoir, river or ground water from an
  aquifer
• Surface supplies generally have conventional
  treatment facilities and disinfection
• Ground water systems
• May have full range of treatment technology but
  some practice chlorination only or do not
  disinfect at all
• Transmission systems
• Tunnels reservoirs and/or pumping facilities
  and storage facilities
• Distribution system
• Carrying finished water pipes to consumer

6
Distribution System is Most Vulnerable Part of
Water System

• Community water supplies designed to deliver
  water under pressure and most of the system
  capacity is reserved for fire fighting purposes
• Could damage or destroy a tank or reservoir
• Potential for the deliberate introduction of
  contaminants into a distribution system (back
  flow, cross connections)
• Need to be able to predict contaminant transport
  pathways and to measure concentration of
  contaminants in networks
• Cyber attack could also provide a serious threat
  to an utilities operations. However many SCADA
  systems are not connected to the Internet

7
Predicting Contaminant Movement In Drinking Water
Distribution Systems

• Movement of water in distribution systems is
  complex
• The ability to predict movement is still
  relatively crude
• Also need to be able to predict changes in
  concentration of contaminants
• Few attempts to integrate monitoring and modeling

8
Contaminants May Be Conservative, or May
Experience Decay or Growth

• Changes may take place in the bulk phase or at
  the pipe wall
• Quality may be influenced by
• Cross Connections
• Failures at the Treatment Barrier
• Transformations in the bulk phase

9
(No Transcript)
10
(No Transcript)
11

• Water Quality Modeling Principles
• Conservation of mass within differential lengths
  of pipe
• Complete and instantaneous mixing of the water
  enteringpipe junctions
• Appropriate kinetic expressions for the growth or
  decay of thesubstance as it flows through pipes
  and storage facilities
• This change in concentration can be expressed by
  a differentialequation of the form

12
Where Cji is the substance concentration
mass/ft3) at position x and time t in the link
between nodes i and j vij is the flow
velocity in the link (equal to the links flow
divided by its cross-sectional area in ft/sec
kij is the rate at which the substance
reacts within the link (mass/ft3/sec)

13

• Storage tanks can be modeled as completely
  mixed,variable volume reactors where the change
  in volumeand concentration over time are
• Where- Vs is the volume (ft3) of the tank- Cs
  is the concentration in tank s

14
The following equation represents the
concentration of materialleaving the junction
and entering a pipe
15

• Where

16
Model Interaction
Water quality models are generally piggy backed
on hydraulic models.
Hydraulic Model
Flows and velocities
Water Quality Model
Water quality results
17
Will Use EPANET To Illustrate the Need For
Integrating Modeling and Monitoring

• First example will be the application of EPANET
  to North Marin Water District in California
• Illustrates the linkage between monitoring and
  modeling
• Second example is the waterborne outbreak in
  Cabool Missouri in 1990
• Forensic application of modeling

18
Modeling of Contaminants

• First field study using EPANET in North Marin
  California
• Modeled chlorine residual propagation and THM
  formation
• Applied to two source system

19
North Marin Water System

• Located near Novato , California
• Serves over 50,000 people
• Virtually no rainfall during warm summer months
• Uses two sources of dramatically different
  quality

20
(No Transcript)
21
(No Transcript)
22

23
(No Transcript)
24
(No Transcript)
25
(No Transcript)
26
(No Transcript)
27
(No Transcript)
28
EPANET Applied to Waterborne Outbreak in Gideon
Missouri in 1993

• Salmonella contamination occurred in municipal
  tank due to failure of hatches and vents
• Taste and odor complaints caused water officials
  to start flushing program
• Out of population of approximately 1000 people,
  440 became ill and 7 people died
• Used model to track outbreak and identify source

29
Municipal Water System in Gideon Was Old and in
Disrepair

• Tuburculation and corrosion in the distribution
  pipes was a major problem
• Two municipal tanks
• Another tank was located on the property of the
  Cotton Compress which was the major employer in
  the area

30
(No Transcript)
31
Waterborne Outbreak

• On November 29, 1993 Communicable Disease
  Coordinator for the Missouri DOH became aware of
  two high school students with culture confirmed
  Salmonellosis
• Within two days five additional patients were
  hospitalized with confirmed salmonellosis
• Missouri Department of National Resources was
  informed that DOH suspected a water supply link
  to outbreak
• DNR samples were positive for fecal coliform
• City of Gideon was required to issue a boil water
  order

32
(No Transcript)
33
Homes with Cases Between 11/23 11/28 and 11/29
12/10 1994 in Gideon, Missouri
34
(No Transcript)
35
(No Transcript)
36
Comparison of Early Confirmed Cases and
Salmonella Positive Sample Versus Penetration of
Tank Water During First Six Hours of Flushing
Program
37
Current Status of Water Quality/Hydraulic Models

• Increasingly sophisticated
• Applied to exposure studies
• ATSDR study on contaminated ground water
• Much research into modeling changes in water
  quality
• Formation of DBPs and Chlorine Residuals
• Tank Mixing Models

38
EPA Research in Real Time Monitoring Systems

• First EPA effort was development of sensors for
  temperature, chlorine residual, fluoride and
  nitrate data with Battelle
• Asked to assist during MCL violation in
  Washington DC
• Initiated research on development of sensors and
  probes for chlorine residual, pH and temperature
  using pipe loops
• Applied to DC water system
• Future efforts should focus on integrating
  modeling and monitoring

39
Summary and Conclusions

• Water systems have been classified as critical
  infrastructure
• Identified as potentially vulnerable
• Contaminant Propagation Can be Modeled and there
  are various models available
• EPANET is a public sector model that has become
  widely used

40
Summary and Conclusions

• EPA has been conducting research into sensor
  development
• Applied to operation of small package plants
• Extended to chlorine residual monitoring in
  Washington DC system
• Future research will focus on integrating remote
  sensing and water quality modeling