In recent years, frequent urban drinking water contamination accidents have caused significant impacts on society in China. However, emergency measures are primarily realized as management policies lack systematic technologies especially for those within-network contamination accidents. The aim of this paper is to perform analysis on cases of contaminant accidents that occurred in water distribution networks and to propose a framework of emergency response technologies. The causes, impacts, and characteristic trends of urban water supply contamination accidents are analysed according to the data of eighteen typical contamination cases that have been publicly reported in China since 2003. The results show that the main problems for rapid response to those accidents are: (1) limited monitoring means; (2) lag in response, especially the lag in providing an explanation to the public; and (3) lag in recovery from the accident. To solve those problems, a framework of technological strategy for rapid response to contamination accidents including monitoring, identifying, evaluation, isolation and flushing is then proposed. Since drinking water contamination accidents are not a unique problem that only happens in China, the response policy and related technologies proposed in this paper are also applicable to other countries and regions that face the same problems.

Introduction

Water distribution network contamination accidents are an important social development issue faced by China. According to the data released by the Chinese Ministry of Environmental Protection in mid-March 2014, China has 250 million residents living near factories with major sewage emissions and major roads and 280 million residents using unsafe drinking water. Contamination accidents have a direct impact on the health and safety of residents and may cause severe social panic.

The Chinese government issued a series of documents regarding emergency management policies and response procedures in 2005–2006, including the ‘Decision of the State Council on the implementation of the scientific concept of development to strengthen environmental protection’ and the ‘National Environmental Emergency Plan’. Since issuing these documents, the overall increasing trend of environmental accidents has been reduced, to a certain extent. However, urban water supply contamination accidents still occur frequently. In the first three months of 2014 alone, ten accidents of drinking water that contained a peculiar odour were reported. Depending on the causes, emergency contaminations of water distribution networks can be divided into two categories: water source contamination and within-network contamination. Both types of contamination result in pollutants entering a water distribution network and spreading to users. Water source contamination is the more prominent cause of China's current drinking water contamination accidents (Han & Huang 2010; Tao & Xin, 2014). Many Chinese scholars have conducted studies on the causes of and management strategies for water source contamination accidents, such as the cyanobacteria contamination accident in Taihu Lake (Qu & Fan, 2010; Zhang et al., 2010, 2011). However, in terms of technical strategies for coping with within-network contamination accidents, systematic strategies and applications of distribution network emergency treatment plans are still lacking.

In this study, water contamination accidents that have happened in China since 2003 are analysed. Then a more detailed statistical analysis was performed on eighteen publicly reported cases of non-source contamination accidents in the water distribution network. Following the analysis, the main problems for rapid response to non-source contamination accidents are identified. To fight against those problems, we proposed a framework of technological strategies for quick response to a drinking water distribution network emergency, based on a comprehensive analysis of existing related technologies.

Statistics of contamination accidents

According to the statistics, there were 1,254 water contamination accidents in China from January 2003 to December 2013; of these, 257 accidents caused direct contamination to residential drinking water, accounting for 20.49% of all water contamination accidents (Figure 1).
Fig. 1.

Statistic of drinking water contamination accidents in China (2003–2013).

Fig. 1.

Statistic of drinking water contamination accidents in China (2003–2013).

The number of contamination accidents tended to increase between 2003 and 2006 (Figure 1) for two major reasons. (1) There were a greater number of environmental emergencies caused by various types of accidents and, thus, the overall situation was grim. (2) Ever since the major accidents of Songhua River pollution in Jilin province in 2005, which led to a water outage in Harbin, various types of urban pollution emergencies have received an unprecedented amount of attention and the number, depth and impact of related news reports have increased significantly, leading to an increase in available statistics (Li et al., 2007). From 2005 to 2006, China promulgated a series of documents that were designed to increase emergency response capabilities, urging the relevant units to strengthen their risk management plans. Hence, the number of contamination accidents has decreased, to a certain extent, since 2006. However, the frequency of occurrences still remains at a relatively high level. The main reasons are: (1) severe water pollution, as pollution problems that have accumulated over several years are more likely to be reported because there are more channels for acquiring and publishing information; (2) urban development, which has resulted in larger and more complex water distribution networks with insufficient network management; and (3) greater residential awareness of the quality of tap water, which has been associated with increased public concern over water safety.

Several representative drinking water contamination accidents that have been publicly reported in recent years are summarized as Table 1.

Table 1.

Typical drinking water contamination accidents.

Series # Time Location Brief summary of the accident Population affected 
Apr. 2003 Jingjiang, Jiangsu A private access to water by a chemical factory allowed the water circulating in condensers to intrude into the water pipe, resulting in a benzene concentration as high as 168.65 mg/L (normal value ≤ 0.01 mg/L). The water distribution network for 4,000 people in the region was contaminated, and some residents exhibited symptoms typical of poisoning. 
Oct. 2003 Ruyang, Henan Deliberate pesticide poisoning of a highland reservoir. Poisoned 133 individuals, and affected more than 30,000 residential locations. 
Feb. 2004 Shapingba district, Chongqing Sewage entered the water distribution network after the repair of a burst pipe, causing regional drinking water contamination. 33 individuals became sick from drinking polluted water; 700 people were affected. 
Dec. 2006 Linxi county, Hebei The water tower of the village was contaminated by nitrides. 1 person died, and more than 40 individuals were hospitalized; 2,000 people were affected. 
May 2007 Wuxi, Jiangsu A sudden deterioration in the water quality of the water source in Nanquan caused a strong odour in the water. This water pollution caused problems for at least 1 million residents in Wuxi. 
July 2007 Shuyang, Jiangsu Water sources were contaminated by ammonia, resulting in water with an odour and the closing of the water plant. 200,000 people were affected. 
Jan. 2008 Fuxin, Liaoning Sewage entered the reservoir from the overflow pipe. The water source for nearly 10,000 people was contaminated, and 1,139 people received medical treatment. 
March 2008 Town of Zhongluotan, Guangzhou The well of a local business was contaminated by nitrite, and the contaminated water entered the drinking water distribution network. 41 villagers showed symptoms of poisoning, such as vomiting, chest tightness, black finger and cramps. 
Feb. 2009 Yancheng, Jiangsu Illegal waste water discharge from a company caused water sources to be contaminated by phenolic compounds, resulting in the closing of two water plants and water outages for a large portion of the city of Yancheng that lasted for nearly 67 hours. 200,000 people were affected. 
10 July 2009 Chifeng, Inner Mongolia Heavy rains caused sewage to enter water source wells, resulting in microbial contamination of drinking water sources. 4,200 individuals required medical treatment; 58,000 people were affected. 
11 March 2010 Ningjin, Heibei A mechanical processing company accidentally sucked water containing sodium nitrite back into the water distribution network. 5 villagers were poisoned; more than 400 people were affected. 
12 May 2010 Bayan, Heilongjiang Water distribution network construction caused a rupture and sewage contamination of drinking water. Nearly 200 students exhibited symptoms of vomiting and diarrhoea. 
13 Aug. 2011 Ruichang, Jiangxi The water distribution system inside a copper smelting company was corroded by raw materials and sewage. Thus, the water was contaminated when the water pressure was low at night, resulting in excessively high concentrations of copper and chlorine. 112 people were poisoned; more than 2,000 people were affected. 
14 Feb. 2012 Zhenjiang, Jiangsu Phenol leaked from a South Korean cargo ship, causing water contamination and an odour in the water of a large portion of the water distribution network. 1 million residents were affected, and a citywide panic over water pollution was triggered. 
15 Feb. 2012 Wuhan, Hubei In association with the construction of an upstream pump station, the water supplied by the Baisazhou water plant had an odour. More than 300 thousand residents were affected. 
16 May and Dec. 2013 Hangzhou, Zhejiang Tert-butylphenol emissions from upstream chemical companies caused a persistent odour in the water of the water distribution system. More than 2 million residents were affected, and a panic triggered consumption of bottled water. 
17 Feb. 2014 Chongming district, Shanghai Phenol contamination in the water source caused a persistent odour in the tap water. Affected four towns and a total of over 100,000 people. 
18 March 2014 Lanzhou, Gansu Artesian ditches of a water plant were contaminated by the wastewater of chemical plants, resulting in a benzene concentration as high as 20 times the standard. A panic triggered purchases of bottled water, and the water use of more than 240 million people was affected. 
Series # Time Location Brief summary of the accident Population affected 
Apr. 2003 Jingjiang, Jiangsu A private access to water by a chemical factory allowed the water circulating in condensers to intrude into the water pipe, resulting in a benzene concentration as high as 168.65 mg/L (normal value ≤ 0.01 mg/L). The water distribution network for 4,000 people in the region was contaminated, and some residents exhibited symptoms typical of poisoning. 
Oct. 2003 Ruyang, Henan Deliberate pesticide poisoning of a highland reservoir. Poisoned 133 individuals, and affected more than 30,000 residential locations. 
Feb. 2004 Shapingba district, Chongqing Sewage entered the water distribution network after the repair of a burst pipe, causing regional drinking water contamination. 33 individuals became sick from drinking polluted water; 700 people were affected. 
Dec. 2006 Linxi county, Hebei The water tower of the village was contaminated by nitrides. 1 person died, and more than 40 individuals were hospitalized; 2,000 people were affected. 
May 2007 Wuxi, Jiangsu A sudden deterioration in the water quality of the water source in Nanquan caused a strong odour in the water. This water pollution caused problems for at least 1 million residents in Wuxi. 
July 2007 Shuyang, Jiangsu Water sources were contaminated by ammonia, resulting in water with an odour and the closing of the water plant. 200,000 people were affected. 
Jan. 2008 Fuxin, Liaoning Sewage entered the reservoir from the overflow pipe. The water source for nearly 10,000 people was contaminated, and 1,139 people received medical treatment. 
March 2008 Town of Zhongluotan, Guangzhou The well of a local business was contaminated by nitrite, and the contaminated water entered the drinking water distribution network. 41 villagers showed symptoms of poisoning, such as vomiting, chest tightness, black finger and cramps. 
Feb. 2009 Yancheng, Jiangsu Illegal waste water discharge from a company caused water sources to be contaminated by phenolic compounds, resulting in the closing of two water plants and water outages for a large portion of the city of Yancheng that lasted for nearly 67 hours. 200,000 people were affected. 
10 July 2009 Chifeng, Inner Mongolia Heavy rains caused sewage to enter water source wells, resulting in microbial contamination of drinking water sources. 4,200 individuals required medical treatment; 58,000 people were affected. 
11 March 2010 Ningjin, Heibei A mechanical processing company accidentally sucked water containing sodium nitrite back into the water distribution network. 5 villagers were poisoned; more than 400 people were affected. 
12 May 2010 Bayan, Heilongjiang Water distribution network construction caused a rupture and sewage contamination of drinking water. Nearly 200 students exhibited symptoms of vomiting and diarrhoea. 
13 Aug. 2011 Ruichang, Jiangxi The water distribution system inside a copper smelting company was corroded by raw materials and sewage. Thus, the water was contaminated when the water pressure was low at night, resulting in excessively high concentrations of copper and chlorine. 112 people were poisoned; more than 2,000 people were affected. 
14 Feb. 2012 Zhenjiang, Jiangsu Phenol leaked from a South Korean cargo ship, causing water contamination and an odour in the water of a large portion of the water distribution network. 1 million residents were affected, and a citywide panic over water pollution was triggered. 
15 Feb. 2012 Wuhan, Hubei In association with the construction of an upstream pump station, the water supplied by the Baisazhou water plant had an odour. More than 300 thousand residents were affected. 
16 May and Dec. 2013 Hangzhou, Zhejiang Tert-butylphenol emissions from upstream chemical companies caused a persistent odour in the water of the water distribution system. More than 2 million residents were affected, and a panic triggered consumption of bottled water. 
17 Feb. 2014 Chongming district, Shanghai Phenol contamination in the water source caused a persistent odour in the tap water. Affected four towns and a total of over 100,000 people. 
18 March 2014 Lanzhou, Gansu Artesian ditches of a water plant were contaminated by the wastewater of chemical plants, resulting in a benzene concentration as high as 20 times the standard. A panic triggered purchases of bottled water, and the water use of more than 240 million people was affected. 

Depending on the causes, the representative contamination accidents (2003–2013) listed in the table can be divided in to four categories.

  • (1) Water source contamination. This type of accident has the greatest and widest impact, accounting for a large proportion of all the drinking water contamination accidents reported, and is thus the most concerning. Among the seventeen accidents listed in Table 1, water source contamination accidents accounted for 47% of the total number of accidents, not including the large-scale water outage accidents, such as the one caused by contamination of the Songhua River (no report indicated that the polluted raw water entered the drinking water distribution network during this accident). Contamination in water sources is relatively easy to identify by source water monitoring and analysis. Meanwhile, the major emergency response measure implemented is to close the water treatment plant, which is relatively simple.

  • (2) Wrong or illegal pipe connections. This type of accident is also an important cause of drinking water contamination. Wrong or illegal pipe connections tend to occur in areas with many township enterprises and problematic issues, such as the lax supervision of small businesses. Lax supervision leads to a lack of effective protection and separation between the water for production use and the water for residential use. Thus, once the water supply network was at very low pressure, contaminants can enter the network and pollute residential water through industrial water pipes, which cause serious contamination accidents.

  • (3) Poor management of a water supply network. For example, factors such as the failure to clean pipes before construction can result in the transport of contaminants into the distribution network. Sewage that flows back into a network through reservoir relief valves and pipes that are corroded by chemicals can also cause the contamination of drinking water in a network. Although this type of problem has occurred to varying degrees in various cities, once pathogenic microorganisms or toxic substances enter a network, major water safety concerns, such as those associated with the accident in 2014 in Lanzhou, can arise.

  • (4) Deliberate poisoning. Despite the fact that deliberate poisoning accidents represent unique cases among contamination accidents, these accidents can cause an enormous amount of damage and have a considerable impact. Contaminants such as chemicals, biological substances or radioactive substances can be introduced at any node of the network or at any water storage facility. Therefore, this type of contamination accident is more difficult to prevent and treat. In particular, China has entered a period associated with frequent terrorist attacks and, as lifelines, water supply systems have become a high-risk target of terrorist attacks and should be guarded vigilantly.

Issues analysis

The preceding analysis indicates that there are two major tasks in coping with urban water distribution system contamination accidents in China: (1) how to reduce or prevent the introduction of contaminants into a water supply system from the source water by implementing water pollution control and pipeline renovations; (2) how to rapidly and effectively stop the spread of a contaminant once it has entered the water supply system and then quickly restore the water supply.

The Chinese government has placed a great deal of effort into the following measures that aim to limit sources of contamination. In response to occurrences of water source contamination accidents that caused tremendous impacts, such as the contamination of the Songhua River caused by the explosion of a chemical plant in Jilin and the outbreak of cyanobacteria in Taihu Lake in Wuxi, the Chinese government has increased its levels of investment in drinking water source protection, emergency treatment processes and other measures. In addition, the urban water supply industry in China has also gained extensive experience in emergency management procedures and emergency treatment technologies for water source contamination accidents. These experiences have resulted in positive effects in terms of preventing and managing occurrences of water contamination accidents.

Compared with water source contamination, contamination accidents that occurred within the water distribution network are more covert and thus require more complex response measures. Therefore, in addition to stricter pollution control efforts, the Chinese government has invested in improvements in the condition of water supplies year after year, especially the water supplies in rural areas and towns. Centralized and decentralized water supply facilities have been constructed. According to ‘The plan of urban water supply facility transformation’ released by China's Housing and Urban Construction Department in 2013, China will invest 267.8 billion yuan and 200 million yuan for distribution network renovations and for emergency treatment measures, respectively, which should play a positive role in reducing the number of this type of drinking water accident.

However, given that the current rapid rate of urbanization in China and the associated increasing trend in environmental pollution will continue for a long period of time, especially considering the reality of the high risk of a terrorist attack, there is another core issue that affects the security of China's urban water supply system that must be solved: how to implement rapid and effective measures to prevent the spread of a contaminant once a pollutant enters the distribution network, and how to rapidly recover from the impacts of a contamination accident to return to a normal state. In this regard, although the Chinese government, at all levels, developed an emergency plan in 2006 after a series of environmental emergencies, the plan focuses primarily on management and lacks clear guidance on specific technical response strategies, especially technical measures executed at the network level. Therefore, there is an urgent need for a series of practical technical solutions to cope with drinking water accidents in China.

Currently in China, several issues concerning the response of urban water service units and related government departments to occurrences of sudden water distribution network contamination accidents need to be addressed.

Limited monitoring means

The majority of drinking water contamination accidents listed in Table 1 were detected through health department reports or user complaints. There were only three cases in which the contamination was detected by monitoring (Figure 2). Among these, two accidents were detected by water source monitoring, and the third case was not detected at the earliest monitoring sampling time because the contaminant indicator was not tested daily. This limitation is a problem that cannot be solved in the near future.
Fig. 2.

Detection of eighteen contamination accident cases.

Fig. 2.

Detection of eighteen contamination accident cases.

Lag in response, especially the lag in providing an explanation to the public

As is reported, the sensing, emergency response, and compliance process can prolong the expected impact time of a water contamination event by a period of time measured in days, not in hours or minutes (Bristow & Brumbelow, 2006). According to the investigations of the accidents listed in Table 1, there are only four cases in which the response to the contamination accidents was initiated in less than 12 hours after the accident was reported (Figure 3).
Fig. 3.

Response delay of eighteen contamination accident cases.

Fig. 3.

Response delay of eighteen contamination accident cases.

Lag in recovery from the accident

On average, 3–5 days were required to completely recover from the occurrence of a contamination accident. In a few instances, more than a week passed before recovery, severely impacting the residents (Figure 4). This lag in recovery time is attributed to the lack of technology and means to effectively curb the spread of a contaminant or to rapidly clear polluted water from a pipeline.
Fig. 4.

Duration of water contamination accident in China.

Fig. 4.

Duration of water contamination accident in China.

Emergency response policy

Based on an analysis of technical studies and engineering measures for emergency responses to water distribution system contamination accidents in China and abroad, a flowchart of water distribution network contamination emergency response measures is proposed in this study. The proposed flowchart (Figure 5) focuses on the formulation of specific measures for the emergency response to water distribution network contamination accidents.
Fig. 5.

Flowchart of emergency response to water distribution system contamination accident.

Fig. 5.

Flowchart of emergency response to water distribution system contamination accident.

The proposed measures to cope with a contamination emergency focus on the following aspects: emergency-oriented deployment of water quality monitoring sites, locating and identifying the source of contamination, evaluation of the scope and impact of the contamination accident, control of contamination spread, and flushing of the distribution network and restoration of the water supply.

Emergency-oriented deployment of water quality monitoring sites

Water quality monitoring is still the first barrier to the prevention of water contamination in a distribution network, there has been a great deal of research on optimal sensor placement in case of contamination accidents (Hart & Murray, 2014). However, two problems exist: first, there are numerous types of contaminants, but common means and equipment for monitoring all types of contaminants are lacking; there are also difficulties such as low accuracy in monitoring measurements and inaccurate data. Second, water supply companies are unable or unwilling to pay for the high costs associated with the installation, operation, and maintenance of monitoring equipment. A more practical approach is to scientifically and rationally develop a technical monitoring plan that is based on conventional online monitoring data, such as residual chlorine and turbidity, in combination with network hydraulic modelling (Yang et al., 2009; Liu et al., 2014); thus, the scope and spread of a contamination accident can be rapidly identified to facilitate the identification of the source of contamination. However, it should be noticed that for non-reactive contaminants, more types of monitoring data such as specific conductivity needed to be involved for detection of the accidents.

Locating and identifying the source of contamination

Contamination source identification is critical for controlling the development of a water distribution network contamination accident. Compared to traditional contamination source identification methods that have high requirements for network monitoring equipment and data, the user complaints based technology can effectively resolve issues of insufficient and low-accuracy network monitoring equipment and data (Ostfeld et al., 2008; Tao et al., 2012). Although there are fewer studies of user complaint-based contamination source tracking methods, the application costs are low. Therefore, within the context of an increased risk of terrorist attacks and, especially, of the currently insufficient means for network monitoring and inadequate monitoring capabilities, it is important to develop complaint-based contamination source identification technology to enable the rapid deployment of measures to control the contamination accident and minimize human health risks.

The authors here presented a Bayesian based model (Xin et al., 2014) that use user complaints data to fast identify the location of pollution source as Equations (1)–(4). 
formula
1
where, is probability of user complaint on node i; is probability of user complaint related to water consumption on node i; is probability of user complaint related to occurring time of contamination; is probability of user complaint related to time delay after contamination occurs; , , , are weights for ,, and .
,, and can be calculated separately by the following equations: 
formula
2
where, is factor which maps to [0, 1]; is water demand at node i; is total demand in water network; 
formula
3
where, is factor which maps to [0, 1]; P2(t) is investigated water intake probability relate to time t. 
formula
4
where, Δt is time delay after contamination occurs; is standard deviation; is factor which maps to [0, 1].
Figure 6 shows a real case of contamination accident in ZJ water distribution network in Southern-Eastern China. The pipe network serves about 870,000 people and has two water treatment plants and three booster pump stations. This accident was confirmed to be due to chemical mass intake nearby one of the two plants (node JX-2 in Figure 6) according to the result of the investigation that lasted for more than two weeks. During the first 24 hours after the accident happened, forty-eight user complaints were received as shown in Figure 6. Based on the model described as Equations (1)–(4), once a third complaint is received, the scope of a contamination source can be limited to six nodes with the help of a hydraulic network simulation (Figure 7), and the candidate source nodes can be further narrowed to the real contaminant node using the above method, which greatly improving the efficiency of the emergency response to the contamination accident. However, this methodology could be improved by better consideration of the effect of time uncertainty in user complaints data.
Fig. 6.

User complaints distribution during a water distribution system contamination accident.

Fig. 6.

User complaints distribution during a water distribution system contamination accident.

Fig. 7.

Number of candidate contaminant source nodes versus number of received user complaints.

Fig. 7.

Number of candidate contaminant source nodes versus number of received user complaints.

Evaluation of the scope and impact of the contamination accident

Once a contamination accident occurs, an investigation of the accident-caused damage to the water supply system can provide the scientific basis for the adopted emergency measures. This evaluation primarily considers the water source and water intake structures as well as the emergency treatment facilities and pipelines. The content of this evaluation includes the causes of damage, their range of impact, the extent of damage, and the repair time. The spread of a contaminant throughout a network and its impact can be assessed using monitoring techniques such as online water quality monitoring, emergency water sampling and measurement procedures as well as a hydraulic simulation of the water distribution network. Due to the time variant nature of demands, use of hydraulic water quality models combined with the supervisory control and data acquisition-based method is more reasonable for contamination spread prediction (Davidson et al., 2005). As sensors are getting cheaper and used more widely in water supply networks, along with the rapid popularity of smartphones, there is the opportunity to use big-data technologies combined with traditional ones including simulation and optimization to achieve goals of evaluating and controlling contamination accidents (Mounce et al., 2015).

Control of contamination spread

The shutdown of the water distribution network is the most direct and effective measure to prevent a contaminant from spreading to users. However, the initiation of a blind, large-scale water outage without any clear knowledge of the contamination source may cause severe social panic. Therefore, after a preliminary assessment of the contaminant toxicity has been conducted and, under certain conditions, factors such as fire safety and social stability have been considered, the action of supplying water at a lower pressure can be considered to meet non-drinking daily water use demands before a government order to shut off the water supply. Supplying water at a lower pressure can reduce water use by users and can also effectively reduce the rate at which a contaminant spreads. When supplying water at a reduced pressure, the operational status of the water distribution network differs from that during the typical water supply conditions. Thus, the effect of pressure on water consumption should be fully considered in water quality and hydraulic simulations. Furthermore, pressure-dependent demands should be considered in water quality and hydraulic simulations (Rasekh & Brumbelow, 2014; He et al., 2016). In this regard, there is still a large amount of research to be done to evaluate the water distribution network system in Chinese cities.

An isolation valve is another key water system component that can be used to control the spread of an accident. The spread of a contaminant through the network requires time; thus, appropriate valves can be closed during this period of time to prevent a contaminant from spreading to downstream users. However, a water distribution network is a pressured system, and the operation of valves in large diameter pipes often requires considerable time and manpower. Therefore, the operation of a pollutant isolation-oriented valve should be accomplished with the help of technologies such as simulations of contaminant spreading and distribution network geographic information system spatial analyses (Baranowski & LeBoeuf, 2008; Eliades & Polycarpou, 2012). Researchers have also proved that water networks' sectorization and partitioning by closing gate valves or redesign of water networks as district metering areas are efficient ways to protect networks from intentional contamination, and reduce the spread of contamination (Grayman et al., 2009; Di Nardo et al., 2015). Although some scholars have proposed technical methods for implementing valve-closing plans (Alvisi et al., 2012; Di Nardo et al., 2013), as described for several cases listed in Table 1, in most cases, a rapid and effective isolation of contamination via valve closure was not achieved, except during contamination accidents that occurred in small-scale tree-like distribution networks.

Flushing of the distribution network and restoration of the water supply

In a contamination accident, contaminants often need to be removed from the distribution network as soon as possible to reduce the associated hazards. The most commonly employed method is to open the fire hydrants to flush the pipelines (Ostfeld et al., 2008; Haxton & Walski, 2009). However, this method is not sufficient for coping with contamination accidents. Dedicated outlet ports that can be used to drain the contaminated water as soon as possible should be established within an urban water supply network, and valves should be closed to reduce contamination to the downstream distribution network. The locations and quantities of outlet ports within the distribution network to be operated during a contamination accident should be determined using optimization technology by simulating contamination spread through the network and the impact of emergency flushing. For example, a network outlet layout and plan for pipeline flushing have been developed for the city of Zhenjiang in China, focusing on the scenario of a source-induced drinking water contamination accident in the network (as shown in Figure 6). An optimized outlet layout can reduce the time required to flush the contaminated water from the distribution network from 20 hours to less than 10 hours (Liu, 2013).

Conclusions

While experiencing rapid urbanization, China is also facing risks to drinking water quality and safety. Because the current measures for preventing water pollution are still not sufficient, water source pollution is still the main cause of current urban drinking water contamination accidents. In addition, as the process of urbanization continues, water supply networks continue to expand and their structures become more complex. The majority of emergency response plans and treatment procedures for contamination accidents have been established at the level of management and policy, but lack systematic and reliable technological strategies. As China continues to face an increasingly serious threat of terrorism, the role of urban water supply systems as lifelines to cities also presents significant safety risks. In addition, compared with water source contamination, deliberate within-network poisoning is more covert and destructive. Thus, given that the current monitoring measures are inadequate to fully address these potential threats, the development and deployment of practical and reliable technologies for locating contamination sources, closing isolation valves, and flushing pipelines to achieve accurate source identification and rapid recovery from contamination (to reduce human health risks as much as possible when an accident does occur) is an urgent problem that China's water industry must address in the coming years.

Therefore, technological research and development in this area should be a focus for developing emergency response plans for China's water supply system in the near future. In addition to an increased effort to protect centralized water sources and control water pollution, further efforts should also focus on strengthening the renovation, supervision, and inspection of the distribution networks of mid/small towns that are under industrial development. Technologies that support the rapid response to contamination accidents and measures that adapted to the level of available technology should be developed. Especially, accompanied by the rapid development of intelligent terminal device, user feedback on water quality is becoming more and more frequent and timely, which is as important as those data collected by sensors. Therefore, big-data related technologies which can combine all of this valuable information together, should be incorporated into water supply network management for fast evaluation and control of contamination accidents.

In addition, drinking water contamination accidents is not a unique problem that only happens in China. Countries all over the world, including many developed countries, also face the risk of sudden pollution of drinking water. Due to the complexity of water distribution systems which are normally constructed underground over a long period of time, those countries are also faced with a similar predicament in dealing with these events. So the response policy and related technologies proposed in this paper also apply to other countries and regions.

Acknowledgements

This project was supported by the National Natural Science Foundation of China (Funding No. 51478326, No. 51378374) and Fundamental Research Funds for the Central Universities of China (No. 0400219309, No. 0400219331).

References

References
Alvisi
S.
Franchini
M.
Gavanelli
M.
Nonato
M.
(
2012
).
Near-optimal scheduling of device activation in water distribution systems to reduce the impact of a contamination event
.
Journal of Hydroinformatics
14
(
2
),
345
365
.
Baranowski
T. M.
LeBoeuf
E. J.
(
2008
).
Optimal Hydraulic Response to a Water Distribution System Contamination Event
.
American Society of Civil Engineers
,
Honolulu, HI
,
USA
.
Bristow
E. C.
Brumbelow
K.
(
2006
).
Delay between sensing and response in water contamination events
.
Journal of Infrastructure Systems
12
(
2
),
87
95
.
Davidson
J.
Bouchart
F.
Cavill
S.
Jowitt
P.
(
2005
).
Real-time connectivity modeling of water distribution networks to predict contamination spread
.
Journal of Computing in Civil Engineering
19
(
4
),
377
386
.
Di Nardo
A.
Di Natale
M.
Guida
M.
Musmarra
D.
(
2013
).
Water network protection from intentional contamination by sectorization
.
Water Resources Management
27
(
6
),
1837
1850
.
Di Nardo
A.
Di Natale
M.
Musmarra
D.
Santonastaso
G. F.
Tzatchkov
V.
(
2015
).
Dual-use value of network partitioning for water system management and protection from malicious contamination
.
Journal of Hydroinformatics
17
(
3
),
361
376
.
Eliades
D. G.
Polycarpou
M. M.
(
2012
).
Water contamination impact evaluation and source-area isolation using decision trees
.
Journal of Water Resources Planning and Management
138
(
5
),
562
570
.
Grayman
W. M.
Murray
R.
Savic
D. A.
(
2009
).
Effects of Redesign of Water Systems for Security and Water Quality Factors
.
American Society of Civil Engineers
,
Kansas City, MO
,
USA
.
Han
X. G.
Huang
T-.L.
(
2010
).
Statistical analysis of sudden water pollution accidents
.
Water Resources Protection
26
(
1
),
84
90
(in Chinese)
.
Hart
W. E.
Murray
R.
(
2014
).
Review of sensor placement strategies for contamination warning systems in drinking water distribution systems
.
Journal of Water Resources Planning and Management
136
(
6
),
611
619
.
Haxton
T. B.
Walski
T. M.
(
2009
).
Modeling a Hydraulic Response to a Contamination Event
.
American Society of Civil Engineers
,
Kansas City, MO
,
USA
.
He
P.
Tao
T.
Xin
K. L.
Li
S. P.
Yan
H. X.
(
2016
).
Modelling water distribution systems with deficient pressure: an improved iterative methodology
.
Water Resources Management
30
(
2
),
593
606
.
Li
L.
Liu
L. L. C.
Zhang
L.
Jiang
D.
Li
J.
(
2007
).
Analysis and strategy of drinking water pollution accidents in recent 20 years in China
.
Acta Geographica Sinica
62
(
9
),
917
924
(in Chinese)
.
Liu
L.
(
2013
).
Optimal Layout of Draining Outlets in Urban Water Supply Network, in College of Environmental Science and Engineering
.
Tongji University
,
Shanghai
,
China
(in Chinese)
.
Liu
S.
Che
H.
Smith
K.
Chen
L.
(
2014
).
Contamination event detection using multiple types of conventional water quality sensors in source water
.
Environmental Science-Processes & Impacts
16
(
8
),
2028
2038
.
Mounce
S.
Gaffney
J.
Boult
S.
Boxall
J.
(
2015
).
Automated data-driven approaches to evaluating and interpreting water quality time series data from water distribution systems
.
Journal of Water Resources Planning and Management
141
(
11
),
1
11
.
Ostfeld
A.
Uber
J. G.
Salomons
E.
Berry
J. W.
Hart
W. E.
Phillips
C. A.
Watson
J.-P.
Dorini
G.
Jonkergouw
P.
Kapelan
Z.
Guestrin
C.
VanBriesen
J.
Small
M.
Fischbeck
P.
Preis
A.
Propato
M.
Piller
O.
Trachtman
G. B.
Wu
Z. Y.
Walski
T.
(
2008
).
The Battle of the Water Sensor Networks (BWSN): a design challenge for engineers and algorithms
.
Journal of Water Resources Planning and Management
134
(
6
),
556
568
.
Qu
J. H.
Fan
M. H.
(
2010
).
The current state of water quality and technology development for water pollution control in China
.
Critical Reviews in Environmental Science and Technology
40
(
6
),
519
560
.
Xin
K. L.
Xiang
N.
Tao
T.
Yin
Z. L.
(
2014
).
Contamination source identification in water distribution network based on consumer complaints
.
Journal of Tongji University
42
(
2
),
283
286
(in Chinese)
.
Zhang
X. J.
Chen
C.
Ding
J. Q.
Hou
A.
Li
Y.
(
2010
).
The 2007 water crisis in Wuxi, China: analysis of the origin
.
Journal of Hazardous Materials
182
(
1–3
),
130
135
.