Contamination of Water Distribution Systems

Walter M. Grayman, PhD, PE

Abstract

It is widely accepted that our water supply systems are susceptible to contamination – intentional or accidental. When a contaminant enters the water distribution system, depending upon the point of entry, the type and quantity of contaminant, the duration of entry, and the operation and design of the water system, it may travel short distances or may reach large portions of the distribution system. This paper discusses the following issues related to contamination of distribution systems:

  • the movement and transformation of contaminants in the distribution system
  • methods (models) for predicting the movement of contaminants and how these models can be used to mitigate the consequences
  • the use of contamination warning systems as a mechanism for responding to an event and reducing the impacts of the contamination event
  • siting of monitors in the distribution system to detect the contamination
  • design of distribution systems to reduce impacts if a contaminant enters the system
  • decontamination of a contaminated distribution system

Introduction

Water distribution systems are inherently susceptible to both intentional and accidental contamination. As shown in Figure 1, there are numerous points at which a contaminant may enter the distribution system. These range from major components such as treatment plants, pumps and tanks, to individual hydrants and customer connections. Since it is highlyinfeasible (or virtually impossible) to fully protect all potential points of entry, in addition to trying to lower the likelihood of contamination events, a major emphasis has been on managing the distribution system in case of contamination and trying to minimize the impacts of such events.

Figure 1: Example potential entry points for contaminants into a distribution system

Fate and Transport of Contaminants in a Water Distribution System

When a contaminant enters a distribution system it will mix with the water flowing in the distribution system and be transported along with the flowing water. Depending upon the particular contaminant, it may also experience chemical, physical or biological reactions and may interact with other constituents in the water and with the pipe wall. The degree to which the contaminant spreads through the distribution system and potentially impacts customers depends upon the point of entry, the type and quantity of contaminant, the duration of entry, and the operation and design of the water system. Consequently, it may travel short distances or may reach large portions of the distribution system.

Mathematical hydraulic and water quality models of distribution systems may be used to calculate (estimate) the movement of a contaminant in a distribution system (Grayman 2005). Both public domain software (e.g., EPANET [Rossman 2000]) and commercial packages are available. The models first calculate the flows and pressures throughout the system over the course of a day or longer periods and then calculate the movement of the contaminant and resulting concentration within the flow assuming either conservative or reactionary behavior. As with all models,the results are only approximations with the accuracy dependent upon factors such as the level of detail of the model and how well it represents the actual system. These models have been used for a variety of tasks related to the contamination of distribution systems including estimating the vulnerability of a system to contamination, design of monitoring systems, and as part of a contamination warning system for mitigating and reducing the impacts of a contamination system.

In order to demonstrate the movement of a contaminant in a distribution system, a model has been applied to an example system that serves approximately 25,000 people under two separate contamination scenarios. In the first scenario, shown in Figure 2, the clearwell at the water treatment plantis contaminated over a period of several days. As illustrated, the contaminant (highlighted in bold in the map) moves through the water distribution system relatively rapidly and after 24 hours a large majority of the system is contaminated. Eventually the entire system would be affected. In the second scenario (Figure 3), a contaminant is injected into the system along a small diameter transmission line (150 mm) serving a neighborhood in the northwest corner of the system. As illustrated, this results in the entire neighborhood being contaminated with the remainder of the system unaffected.

Figure2. Spread of contaminant in a distribution system – Scenario 1

Figure 3. Spread of contaminant in a distribution system – Scenario 2

The speed at which a contaminant moves through the system is largely controlled by the water demand (water consumption by customers), resulting in more rapid movement during higher demand daylight hours than at night. The concentration of the contaminant at any point within the distribution system is controlled by a set of interacting factors including the volume of contaminant that enters the system, the duration of the injection and the amount of uncontaminated water that dilutes the contaminant. Other factors that affect the movement of the contaminant and its concentration include the operation of the water system and the presence of storage tanks. The model effectively integrates all of these factors in calculating the resulting concentration.

Contamination Warning System

The primary mechanism that is currently available to mitigate or reduce the impacts of a contamination event in the distribution system is a contamination warning system (CWS). A contamination warning system is a combination of monitors, institutionalarrangements, analysis tools, emergency protocols, and response mechanismsdesigned to provide early warning of contaminants in order to minimize customer exposure.

USEPA (2007) describes aconceptual model for contamination warning system operation as follows.

  • Monitoring and surveillance. The basic components of online water quality monitoring, sampling and analysis, enhanced security monitoring, consumer complaint surveillance, and public health surveillance occurs on a routine basis, in near-real time until an anomaly or deviation from the baseline or base state is detected.
  • Event detection and possible determination. Event detection is the process or mechanism by which an anomaly or deviation from the baseline or base state is detected.
  • Credible determination. Credibility determination procedures are performed using information from all contamination warning system components as well as external resources when available and relevant. If contamination is determined to be credible, additional confirmatory and response actions are initiated.
  • Confirmed determination. In this stage of consequence management, additional information is gathered and assessed to confirm drinking water contamination. Response actions initiated during credible determination are expanded and additional response activities may be implemented.
  • Remediation and recovery. Once contamination has been confirmed, and the immediate crisis has been addressed through response (e.g., flushing, emergency warnings, etc.), remediation and recovery actions defined in the consequence management plan are performed to restore the system to normal operations.

On-line monitors or sensors are widely considered as the primary means of detecting a potential contamination event in a distribution system. The role of monitors in a contamination warning system and how they may be applied during a potential contamination event to detect, characterize and confirm a contaminant event is shown schematically in Figure 4 (Grayman 2010). In order to reliably and efficiently detect potential contaminants, the monitors must be sensitive to the presence of a wide range of agents and there must be a sufficient number of appropriate monitors so that detection occurs in a timely manner. Historically, monitoring sites were selected primarily on the basis of informal selection criteria that reflected the representativeness and accessibility of the sites andthe specific purpose(s) of the monitoring system.

Figure 4. Flowchart of a CWSas part of a response to a potential contamination event

The success of a contamination warning system in reducing impacts of a contamination event is highly dependent upon the length of time of the various steps shown in Figure 4. In a study in Ann Arbor, Michigan, Skadsen et al (2008) found that a total delay time greater than about 8 hours between the entry of the contaminant into the distribution system and the cessation of water use by customers (associated with the initial detection, initial actions, characterization and mitigative response steps in Figure 4) would significantly reduce the effectiveness of the CWS. This threshold may vary considerably in other systems. Discussions with water utility personnel suggest that in order to meet a delay time goal of less than eight hours would require a very robust on-line monitoring network and a very efficient emergency response plan.

Over the past decade, there has been extensive research on the optimal placement of a set of sensors in a distribution system (Murray and Hart 2010; Grayman 2010). Berry et al. (2005) provide a succinct description of the sensor placement problem for CWS design.

“The goal of a sensor placement optimizationformulation is simple: to place a limited number of sensors in a water distribution networksuch that the impact of an accidental or intentional injection of contaminant to publichealth is minimized. However, no specific, concrete formulation for sensor placementhas emerged that is widely accepted by the water community. There are a wide rangeof alternative objectives that are also important when considering sensor placements,such as minimizing the cost of installing and maintaining the sensors, minimizing theresponse time to a contamination event, and minimizing the extent of contamination(which impacts the recovery costs). Additionally, it is difficult to quantify the healthimpact of a contamination event because human water usage is often poorly characterized,both in terms of water consumption patterns as well as how the water consumptionimpacts health effects. Consequently, surrogate measures like the total volume of waterconsumed at all sites have been used to model health impacts; this measure assumes thathuman water consumption is proportional to water consumption at all junctions withinthe network.”

Many different algorithms have been proposed and tested for designing optimal sensor networks (Ostfeld 2008). Most of these methods utilize integrated search techniques in conjunction with distribution system hydraulic/water quality models that assess the benefits associated with alternative sensor network designs. The most widely used software in this category is the Threat Ensemble Vulnerability Assessment –Sensor Placement Optimization Toolbox (TEVA-SPOT) distributed by the Environmental Protection Agency (Berry et al 2010).

Design of Water Systems to Reduce Vulnerability

Contamination warning systems and othermethods for detecting and mitigating contamination events require active intervention in order to minimize impacts. An alternative or supplemental“passive” mechanism for reducing impacts involves the re-design of distribution systems aimed specifically at improving security (Baker 2008). In the United States and many other countries, distribution systems are designed as looped systems composed of transmission lines andinterconnected local delivery pipes used to deliver water to customers. Looped systems generally result in multiple paths that water can follow from thetreatment plant to customers and provide redundancy in case of outages. However, the loops also result in multiple pathways for contaminants and greater difficulties in isolating contaminants.

Baker (2008) suggested the greater use of “zoned” distribution systems based in part on the District Meter Area (DMA) concept that is widely used in the United Kingdom and other areas. A DMA consists of an isolated water distributionsystem zone for which there can be both control and calculation of flowamounts. The Baker approach, shown schematically in Figure 5, extends this approach further to minimize interchange between zones and “hardening” and reduction of access to the transmission system to discourage entry of contaminants into the transmission backbone.

Figure 5. Schematic representation of a zoned distribution system

Murray et al (2010) tested this zoned approach using mathematical modeling and simulation of several distribution systems and found that such an approach may significantly reduce the spread of contaminants with little negative impact on water age (a surrogate for water quality) and reliability. Converting a system from a looped to a zoned system would require the addition of control valves, check valves and some piping. The level of effort and costs associated with such a conversion is likely to vary considerably depending on the existing distribution system configuration.

Remediation, Decontamination and Recovery

When a contamination event occurs, the immediate concern is minimizing the impacts of the event on the customers of the water system. Once the immediate concern is taken care of, attention turns to the goal of remediating and decontaminatingthe system and then returning it to normal operation (USEPA 2008). There are many challenges during this period when the system is being returned to normal operation including, for example:

  • Providing interim water supply while the water system is not available
  • Treatment of water contaminated through accidental or deliberate actions. This water needs to be treated before being released into treatment systems or the environment.
  • Decontaminating pipes. Many contaminants that have been identified as likely threats to drinking water and distribution facilities can adhere to, or become embedded in, rusty or corroded pipes or biologically active layers (biofilm).
  • Mitigating the “fear factor” associated with the contamination event and regaining the confidence of customers that their water supply is safe.

Conclusions

In the United States, following the events of September 11, 2001, the concerns over possible terrorist acts affected almost all aspects of life including water supply systems. Previous concerns over accidental contamination of water sources or intrusion into distribution systems took a backseat to the much greater concerns over intentional contamination of the water supply. This fostered a significant upturn in research and regulations affecting water supplies. The goal was certainly a water system that was much less vulnerable to intentional and accidental events, a robust and ubiquitous monitoring system that would immediately detect a wide range of contaminants at any point from source to tap, and a response plan that would protect customers from exposure to unwanted agents. In the years since 2001, some progress has been made in reaching for the goals of a safer water system. However, since there have been no major intentional contamination events in water systems in the U.S., one can postulate that: (1)it is difficult to fully assess the degree of protection that has resulted and; (2) the rate of research and development (and progress towards the lofty goals of 2001) has slowed down.

There are still many issues that need to be confronted in order to ensure a safe water supply. Some issues and questions that are specifically related to potential contamination of water distribution systems are outlined below:

  • Given the current state of technology, can a contamination warning system be effective in protecting customers from contamination events?
  • Can we afford the relatively high capital and operational costs associated with a contamination warning system to defend against what is assumed to be very rare events?
  • Can we broaden the usage of the monitoring equipment associated with contamination warning systems in order to address more routine water quality issues and thus, broaden the cost basis?
  • What are the prospects for future breakthroughs in monitoring capabilities that will lower costs and improve capabilities?
  • Should we consider more radical systems such as point-of-use treatment or dual water delivery systems as a mechanism for improving the security of the water delivery process?

References

Baker Corporation. (2008), Municipal Water Distribution System Security Study: Recommendations for Science and Technology Investments, Final Report to the U.S. Department of HomelandSecurity, Washington, DC.

Berry, J. W., Hart, W. E., Phillips, C. A., Uber, J. G., and Watson, J.-P. (2005). Validation and assessment of integer programming sensor placement models. In Proceedings of the World Water & Environmental Resources Congress, EWRI, ASCE, Reston, VA.

Berry, J. et al. (2010). User’s Manual TEVA-SPOT Toolkit. Version 2.4. U.S. Environmental Protection Agency. EPA 600/R-08/041B.

Grayman, W.M. (2010). Designing an Optimum Water Monitoring System. Wiley Handbook of Science and Technology for Homeland Security, Ed., J.G. Voeller. John Wiley & Sons, Inc.

Grayman, W.M. (2006). Use of Distribution System Water Quality Models In Support Of Water Security. In Security of Water Supply Systems: From Source to Tap, Edited by J. Pollert and B. Dedus. Springer Press, Netherlands.

Murray, R.E., Grayman, W.M., Savic, D.A., and Farmani. (2010).Effects of DMA redesign on water distribution system performance. In Integrating Water Systems – Edited by Boxall & Maksimovi´c. Taylor & Francis Group, London.

Murray, R. and Hart, W.E. (2010). Review of Sensor Placement Strategies for Contamination Warning Systems. Journal of Water Resources Planning and Management. American Society of Civil Engineers (ASCE), Reston, VA, 136(6):611-619.

Ostfeld, A., et al. (2008). The battle of the water sensor networks (BWSN): A design challenge for engineers and algorithms.J. Water Resources Planning and Management, American Society of Civil Engineers (ASCE), Reston, VA. 134(6): 556–568.

Rossman, L. A. (2000). EPANET 2 Users Manual. U.S. Environmental Protection Agency.

Skadsen, J., Janke, R., Grayman, W., Samuels, W., TenBroek, M., Steglitz, B., and Bahl, S. (2008). Distribution System Monitoring For Detecting Contamination Events And Water Quality, J.AWWA. 100(7): 81-94.

U.S. EPA. (2007). “Water security initiative: Interim guidance on planning for contamination warning system deployment.” EPA-817-R-07-002, Office of Water, Office of Ground Water and Drinking Water, U.S. EPA, Washington, D.C.

U.S. EPA. (2008). “Decontamination and recovery planning: Water and wastewater utility case study”.” EPA-817-F-08-004, Office of Water, U.S. EPA, Washington, D.C.

1