Advancements in real-time water monitoring technologies permit rapid detection of in-stream, in-pipe water quality, and alert of threats from waste loads. Legislation mandating the establishment of water resources monitoring, presence of hazardous contaminants in effluents, and perception of the vulnerability of the water distribution system to attacks, have spurred technical and economic interest. Alternatively to traditional analyzers, chemosensors operate according to physical principles, without sample collection (online), and are capable of supplying parameter values continuously and in real-time. This review paper contains a comprehensive survey of existing and expected online monitoring technologies for measurement/detection of pollutants in water. The state-of-the-art in online water monitoring and contaminant warning systems is presented. Application examples are reported. Monitoring costs will become a lesser part of a water utility budget due to the fact that automation and technological simplification will abate human cost factors, and reduce the complexity of laboratory procedures.

INTRODUCTION

Rapid and constant advancement of real-time water monitoring and sensing technologies will make these an ever more important tool for the evaluation of online, in-pipe water quality, and the assessment of related life and health risks. Established technologies are now permitting rapid detection of water quality changes, health and environmental threats induced by waste loads, and other impacts. Initially used in lieu of traditional monitoring mainly for reporting purposes, these instruments have recently been supplemented by specific software and interconnecting networks to become true online quality monitoring of distribution systems, also known under the names of contaminant warning systems (CWS), or water quality event detection systems (EDS). These consist of an integrated system of sensors, supervisory tools and data acquisition, to continuously monitor network conditions, warning of potential contamination events.

Water-quality monitoring programs represent a balance between several factors: analytical capacity, collection, processing, and maintenance of representative samples, and available resources, including technical, human and financial. While monitoring has been traditionally driven by the development of increasingly more sophisticated analytical equipment allowing lower detection limits and new constituent analyses, this increased capacity has often clashed against a number of limitations, including financial investment availability for purchase, and the capacity to collect uncontaminated and/or representative samples suitable for the new technologies. An alternative to these traditional chemical cabinet analyzers is the use of chemosensors, which operate according to physical principles (e.g. light measurement and others), without sample collection (directly in-stream), supplying (true or surrogate) parameter values in real time. The other alternative, traditional manual grab sampling followed by laboratory analysis, requires considerable manpower and only allows capture of small data sets, mostly unrepresentative of the variance at the source, and allows potentially important events to occur undetected (Copetti et al. 2014). In addition, long-term environmental databases often display significant data shifts that exceed natural variability, and which may be correlated with changes in sampling and laboratory analysis techniques and methods (Horowitz 2013). It has been shown, on the other hand, that remotely acquired, continuously in situ monitored data can provide important early warning information about water quality events (Glasgow et al. 2004). Comparative advantages of online monitoring sensor technology versus traditional cabinet analyzers and manual sampling are summarized in Table 1.

Table 1

Advantages of online monitoring sensor technology vs manual and fixed cabinet sampling

IssueOnline sensorsCabinet analyzersManual sampling
Installation costs One sensor can detect several parameters at once. Extremely simple in-stream/in-pipe installation (protection from vandalism/theft needed). Sensor costs are low compared to analyzers. Can run on battery for long times. Installation requires proper operator-accessible housing (protected from vandalism/ theft) with service lines. Instrumentation is expensive and needs automatic samplers. Very small installation costs. May require trucking a boat to the river or building sampling ports in pipe network. 
O&M costs Extremely low. Can work unassisted for long periods. Usually, visual inspection suggested bi-weekly or monthly. High. Frequent personnel control, calibration check, reagent costs. Highest. A team is engaged for each campaign. Cost of laboratory procedures and sample handling, and possible errors must be considered. 
Site accessibility Site must be accessed at installation and in case of maintenance/repositioning. No need to access if working properly. Minimal disturbance. Site must be accessed often. Medium–high disturbance. Requires team working on-site during campaigns. Maximum disturbance. Access can affect measurement. 
Sampling frequency Sample-less. Measurement is instantaneous, can be set from fractions of second on. Depends on technical times for analysis. Limited capacity (e.g. of automatic sampler). Need for sample handling. Even during ‘continuous sampling’ events the frequency is limited by the operators’ training and technology used. 
Data availability and uncertainty Instantaneous, can be transmitted wirelessly to a receiving station, and/or stored locally. Uncertainty due to missing data is highly unlikely. Systematic data error due to calibration deviation possible but retraceable. After analysis, same as online sensor, however, the delay due to the analytical process cannot be eliminated. Uncertainty of missing data due to failed procedure possible. Systematic data error due to miscalibration possible. Unless simple parameters are measured locally by hand-held devices, samples have to be transported and worked up in the laboratory. Uncertainty due to missing data depending on monitoring protocols, but quite possible. Systematic and random data errors due to manual handling of samples and human interference highly possible. 
Water quality dynamics Can fully capture water quality dynamics at the short- and long- term ranges, due to continuous, virtually unlimited data collection. Can capture some trends, depending on proper preliminary setting and sampler's capacity. Usually finite data collection capacity. Almost undetectable within a single campaign. 
Health protection Early online detection of contaminants may allow prompt response. Used on CWS/EDS systems. Detection of contaminants may not be quick enough for adequate intervention. Automatic determination of hazardous pollutants not always possible. Only for slow-moving contaminants in far-off locations (e.g. groundwater). 
IssueOnline sensorsCabinet analyzersManual sampling
Installation costs One sensor can detect several parameters at once. Extremely simple in-stream/in-pipe installation (protection from vandalism/theft needed). Sensor costs are low compared to analyzers. Can run on battery for long times. Installation requires proper operator-accessible housing (protected from vandalism/ theft) with service lines. Instrumentation is expensive and needs automatic samplers. Very small installation costs. May require trucking a boat to the river or building sampling ports in pipe network. 
O&M costs Extremely low. Can work unassisted for long periods. Usually, visual inspection suggested bi-weekly or monthly. High. Frequent personnel control, calibration check, reagent costs. Highest. A team is engaged for each campaign. Cost of laboratory procedures and sample handling, and possible errors must be considered. 
Site accessibility Site must be accessed at installation and in case of maintenance/repositioning. No need to access if working properly. Minimal disturbance. Site must be accessed often. Medium–high disturbance. Requires team working on-site during campaigns. Maximum disturbance. Access can affect measurement. 
Sampling frequency Sample-less. Measurement is instantaneous, can be set from fractions of second on. Depends on technical times for analysis. Limited capacity (e.g. of automatic sampler). Need for sample handling. Even during ‘continuous sampling’ events the frequency is limited by the operators’ training and technology used. 
Data availability and uncertainty Instantaneous, can be transmitted wirelessly to a receiving station, and/or stored locally. Uncertainty due to missing data is highly unlikely. Systematic data error due to calibration deviation possible but retraceable. After analysis, same as online sensor, however, the delay due to the analytical process cannot be eliminated. Uncertainty of missing data due to failed procedure possible. Systematic data error due to miscalibration possible. Unless simple parameters are measured locally by hand-held devices, samples have to be transported and worked up in the laboratory. Uncertainty due to missing data depending on monitoring protocols, but quite possible. Systematic and random data errors due to manual handling of samples and human interference highly possible. 
Water quality dynamics Can fully capture water quality dynamics at the short- and long- term ranges, due to continuous, virtually unlimited data collection. Can capture some trends, depending on proper preliminary setting and sampler's capacity. Usually finite data collection capacity. Almost undetectable within a single campaign. 
Health protection Early online detection of contaminants may allow prompt response. Used on CWS/EDS systems. Detection of contaminants may not be quick enough for adequate intervention. Automatic determination of hazardous pollutants not always possible. Only for slow-moving contaminants in far-off locations (e.g. groundwater). 

During the last two decades, several studies have revealed the presence of hazardous contaminants in waters due to ‘common’ anthropic activities, including pesticides (Öllers et al. 2001), natural and synthetic hormones (Kolpin et al. 2002), plasticizers, personal care products and pharmaceuticals (Daughton & Ternes 1999; Jones et al. 2002). Since these may end up in water supplies, there is a clear need to be able to rapidly detect instances of accidental (or deliberate) contamination in distribution systems, due to the potential consequences for human health. These data might not be measurable during routine offline monitoring at drinking water treatment plants, or in various distribution system locations. Since existing laboratory methods are too slow to develop operational responses, they cannot provide a sufficient level of public health protection in real time, therefore the need for better online monitoring of water systems is clear (Storey et al. 2011).

Water distribution systems are vital for the life and well-being of cities and nations, but unlike other similarly vital installations, they are potentially accessible (also in the ‘unauthorized’ meaning) to almost everyone willing to do so. In the current geopolitical climate, water distribution systems may thus become relatively easy targets of terrorist groups of any extraction, that could thus affect the fate of large numbers of people with limited effort. In 2002 the US FBI circulated a reserved warning to water industry managers indicating that al-Qaida operatives may have been studying American water-supply systems in preparation for attacks (IonLife 2002). No such attacks have been officially exposed (or disclosed) so far, however, accidental drinking water contamination events with variable (although usually nonlethal) consequences occur almost every day in many parts of the world. The safety of water distribution systems has thus become of primary concern to governments, and research related to water quality monitoring has increased significantly in recent years.

In view of these risks and the need for a safe and reliable water supply, traditional monitoring routines can no longer be considered satisfactory, especially since online, relatively cheap monitoring technologies, available for a larger number of parameters than previously thought possible, have become affordable (Capodaglio & Callegari 2009, 2015; O'Halloran et al. 2009; Capodaglio et al. 2016a).

Utilities around the world are now using some form of online monitoring as warning systems for drinking water contamination, in anticipation of yet-to-be-specified regulations. In the USA, turbidity is currently the only indicator bearing a regulatory requirement for continuous online monitoring (AWWA 2002); in Europe, current regulations (Council Directive 98/83/EC) do not specify the need for online measurements in drinking water systems, although good practice suggests that, at least in critical situations, some basic continuous monitoring (e.g., turbidity) should be implemented, given also the very affordable cost of last-generation sensors, today. Online CWS for water distribution networks (WDN) have been studied in the last few years, and are gradually being put into place. This paper contains a comprehensive review of existing and expected online monitoring technologies for measurement/detection of pollutants in water. The state-of-the-art in online water monitoring and CWS is also presented, with some application examples.

Online water quality monitoring

AWWA (2002) defines online water quality monitoring as the unattended sampling, analysis and reporting of parameters, producing data sequences at a greater frequency than that permitted by manual (grab) sampling, and allowing real-time feedback for process control, water quality characterization for operational or regulatory purposes, and alert/alarm.

Online monitoring of pollutants and dangerous substances is important for different purposes, including plant management, pollution control and reduction, and limiting the environmental impact of discharges. Online instrumentation must be placed at selected, representative locations in water system networks, and must be periodically maintained. Monitoring requirements can be defined according to monitored water type, considering its:

  • (a)

    variability, in space and time (in general very low for groundwater, low for lakes, high for rivers, very high for discharge channels and urban or industrial drainage, or in-plant, in-pipe monitoring);

  • (b)

    vulnerability, including type and location of possible contaminating activities, time-of-travel of contaminants to intake/point of use, natural/technological barriers’ effectiveness, control options after alarm.

The ‘ideal’ location for control of contaminants is as close to their potential source as possible. Source water with low vulnerability is characterized by few potential contaminant activities, transit times longer than those required for laboratory analysis, and the presence of multiple physical barriers between contaminating activities and point of intake. In a source with moderate vulnerability, online monitoring of surrogate parameters (such as total organic carbon (TOC), dissolved organic carbon (DOC), UV254, pH and conductivity) should be considered. In high-vulnerability water sources, online monitoring of chemical–physical–biological parameters (turbidity, pH, conductivity, redox, fish toxicity) and surrogate parameters, in addition to specific indicators (e.g. volatile organic compounds (VOCs), phenols and specific toxicity tests) may be preferred. In water/wastewater treatment applications monitoring must consider possible process optimization options, response times, significative sampling frequency, and allow adequate process-control lead time. For drinking water protection, multi-barrier approaches based on the concept that contaminants must be subject to as many points of control/treatment (barriers) as possible, prior to tap, are usually adopted (O'Halloran et al. 2009).

Table 2 summarizes monitoring requirements and objectives for various types of activities in the specific case of water distribution system applications.

Table 2

Online monitoring objectives and strategies for water distribution systems (modified from AWWA 2002)

ActivityMonitoring strategyObjectives
Contaminant source identification Surrogate parameters (TOC, DOC, UV254, pH, conductivity); Specific parameters (related to known sources of contamination); Biotests and toxicity tests Define potential contamination in relation to vulnerability of source water 
Monitoring of discharges into the source water Specific organic/inorganic contaminants Identify water pollution accidents 
Best management practices/protection of water source Hydrological parameters; Environmental parameters (solar radiation, O2, chloride) Prevent source deterioration; Environmental management 
Drinking water quality protection Specific organic/inorganic contaminants; Treatment-related parameters (flow, turbidity, pH, TOC, DOC, etc.); Biotests/toxicity Allow appropriate responses to contaminant presence (intake shut-up, additional treatment, treatment adjustment) 
Emergency response Specific organic/inorganic contaminants; Biotests/toxicity Drinking-water pollution control; Risk management; Treatment modification 
ActivityMonitoring strategyObjectives
Contaminant source identification Surrogate parameters (TOC, DOC, UV254, pH, conductivity); Specific parameters (related to known sources of contamination); Biotests and toxicity tests Define potential contamination in relation to vulnerability of source water 
Monitoring of discharges into the source water Specific organic/inorganic contaminants Identify water pollution accidents 
Best management practices/protection of water source Hydrological parameters; Environmental parameters (solar radiation, O2, chloride) Prevent source deterioration; Environmental management 
Drinking water quality protection Specific organic/inorganic contaminants; Treatment-related parameters (flow, turbidity, pH, TOC, DOC, etc.); Biotests/toxicity Allow appropriate responses to contaminant presence (intake shut-up, additional treatment, treatment adjustment) 
Emergency response Specific organic/inorganic contaminants; Biotests/toxicity Drinking-water pollution control; Risk management; Treatment modification 

Availability of real-time information is one of the key benefits of online monitoring: the information must be conveyed to the appropriate user by means of systems often referred to as supervisory control and data acquisition (SCADA). These consist of individual online instruments, connected to programmable logic controllers or remote telemetry units, that convert output signals to the desired units, compare them to criteria set by users, and generate signals for alarm or control to process equipment. A host computer is used to visualize, store, or to further utilize data for specific purposes (Figure 1). A few cities around the world have already adopted such systems, as illustrated by Allen et al. (2011).
Figure 1

SCADA system for environmental monitoring.

Figure 1

SCADA system for environmental monitoring.

Online monitoring technology overview

Table 3 summarizes the five main classes of online monitoring instrumentation. In this review, just the first four classes will be examined, with discussion of basic operating principles, state-of-the-art, and evaluation of technology for online applications in water and wastewater monitoring.

Table 3

Online monitoring instrumentation classes

Type of monitorsApplication examples
Physical Turbidity, particles, color, conductivity, total dissolved solids (TDS), streaming current, radioactivity, temperature, redox potential 
Inorganic pH, dissolved oxygen (DO), hardness, acidity, alkalinity, disinfectants such as chlorines and ozone, metals, fluoride, nutrients, cyanide 
Organic carbon (BOD, COD or TOC), hydrocarbons, UV adsorption, VOCs, pesticides, DBPs 
Biological nonspecific, algae, protozoa, pathogens 
Hydraulic flow, level and pressure 
Type of monitorsApplication examples
Physical Turbidity, particles, color, conductivity, total dissolved solids (TDS), streaming current, radioactivity, temperature, redox potential 
Inorganic pH, dissolved oxygen (DO), hardness, acidity, alkalinity, disinfectants such as chlorines and ozone, metals, fluoride, nutrients, cyanide 
Organic carbon (BOD, COD or TOC), hydrocarbons, UV adsorption, VOCs, pesticides, DBPs 
Biological nonspecific, algae, protozoa, pathogens 
Hydraulic flow, level and pressure 

Physical monitors

Well-established technologies used for monitoring physical parameters include: light scattering/blocking (turbidity, particles, suspended solids (SS)), light absorbance (color), electrochemical (conductivity, reduction–oxidation potential (Redox)), electrophoretic (streaming current), and other (radioactivity, temperature). Most of these have been commercially available as online instrumentation for some time (Table 4).

Table 4

Physical online monitor technology (modified from AWWA 2002)

ApplicationMost appropriate technologyOther technologies
Low turbidity raw water Single beam (tungsten or LED) turbidimeter Particle counters; Particle monitors 
Clarified water; Filter effluent Modulated four-beam turbidimeter  
High turbidity raw water Ratio turbidimeter; Modulated four-beam turbidimeter  
Filter backwash Transmittance turbidimeter; Surface scatter; Ratio turbidimeter; Modulated four-beam turbidimeter Laser light source (660 nm) and improved optics turbidimeters 
Color Online colorimeter; Spectrophotometer  
TDS Two-electrode conductivity probe; Electrode-less (toroidal) probes  
ApplicationMost appropriate technologyOther technologies
Low turbidity raw water Single beam (tungsten or LED) turbidimeter Particle counters; Particle monitors 
Clarified water; Filter effluent Modulated four-beam turbidimeter  
High turbidity raw water Ratio turbidimeter; Modulated four-beam turbidimeter  
Filter backwash Transmittance turbidimeter; Surface scatter; Ratio turbidimeter; Modulated four-beam turbidimeter Laser light source (660 nm) and improved optics turbidimeters 
Color Online colorimeter; Spectrophotometer  
TDS Two-electrode conductivity probe; Electrode-less (toroidal) probes  

Inorganic monitors

Inorganic monitors are used in online mode to detect influent and effluent water quality, and/or treatment process control; applicable technologies are listed in Table 5. Online monitoring of inorganic constituents, with the exception of chemical titration technology (alkalinity, acidity, hardness), is still in the early phases for many elements of interest to drinking water applications. For metals, typical available technologies (non-existent until very recently for many metals of interest, like As, Cd, Pb, Hg, Se, Zn) are adaptations to automatic mode of complex colorimetric methods originally developed for laboratory applications, and therefore expensive and/or complex to operate, and still not suitable for installation in remote or unmanned sites. Online instrumentation based on anodic or cathodic stripping voltammetry (ASV, CSV) methods was developed and launched on the market quite recently (Kissinger & Heineman 1996; Jothimuthu et al. 2011; Yue et al. 2012; Bullough et al. 2013; Nunes et al. 2015), with detection limits down to 0.5–10 μg/l (clean water), depending on sample type and actual analyte (ModernWater 2015).

Table 5

Online inorganic monitor technology (modified from AWWA 2002)

ParameterCurrently applied technologyOther technologies
DO, pH Ion-selective electrodes Fiber-optic chemical sensors (FOCS) 
Hardness EDTA tritration online; Ion-specific electrodes (ISE) (FOCSs or optodes) for pH, DO; Iodometric DO measurements 
Alkalinity Online alkalinity titrator ClO2: Iodometry, Amperometric meth. I, DPD, amaranth, chlorophenol red, LGB dye, ion chromatography 
Iron, manganese, metals X-ray fluorescence (complex), colorimetry CSV, ASV stripping voltammetry; Graphene-based EC-sensors 
Ammonia, nitrite Colorimetric, FOCS (ammonia)  
Nitrate Ion sensitive gas membrane electrodes, UV spectrometry  
Phosphorus, cyanide Colorimetric, FOCS (cyanide)  
ParameterCurrently applied technologyOther technologies
DO, pH Ion-selective electrodes Fiber-optic chemical sensors (FOCS) 
Hardness EDTA tritration online; Ion-specific electrodes (ISE) (FOCSs or optodes) for pH, DO; Iodometric DO measurements 
Alkalinity Online alkalinity titrator ClO2: Iodometry, Amperometric meth. I, DPD, amaranth, chlorophenol red, LGB dye, ion chromatography 
Iron, manganese, metals X-ray fluorescence (complex), colorimetry CSV, ASV stripping voltammetry; Graphene-based EC-sensors 
Ammonia, nitrite Colorimetric, FOCS (ammonia)  
Nitrate Ion sensitive gas membrane electrodes, UV spectrometry  
Phosphorus, cyanide Colorimetric, FOCS (cyanide)  

Developments in miniaturization technology and new materials, (i.e. carbon nanotubes) recently allowed design of fully automated, online metal monitors able to provide continuous monitoring in liquid streams (Hanrahan et al. 2004). Graphene has recently attracted strong scientific and technological interest, showing great promise in many diverse applications, from electronics, energy storage, and fuel cells, to biotechnologies because of its unique properties (Shao et al. 2010). Graphene-based electrochemical sensors have been developed for the detection of heavy metal ions. However, no commercial graphene-based products for environmental monitoring applications are available as of now.

Some promise for future applications comes from developments in optode technology, coupled with miniaturized spectrophotometry, due to their low-cost, low power requirements and long-term stability. Optodes (or optrodes) are optical sensors formed by a polymeric matrix coated onto the tip of an optical fiber, capable of (optically) measuring a specific substance, with the aid of a chemical transducer, applying various measurement methods, such as reflection, absorption, evanescent wave, chemiluminescence, surface plasmon resonance (SPR), and, by far the most popular, luminescence (fluorescence and phosphorescence). Optodes may provide viable alternatives to electrode-based sensors, or more complicated analytical instrumentation (Tengberg et al. 2006; Xie et al. 2014), although they often still do not have resolution comparable to the most recent cathodic microsensors.

Organic monitors

This class of monitors includes TOC analyzers, UV absorption and differential spectroscopes, chip-based micromachined devices and chromatographs. Although not all of these are suitable for online, on-site applications, this specific technology is much more developed than that for inorganics.

For this reason, in addition to its use in mandatory monitoring, and notwithstanding a lack of specific regulations, many water utilities already routinely use online organics monitoring to some degree. Table 6 shows different fractions measured by an organic carbon analyzer. Most organic compounds in water absorb UV radiation: their concentration can thus be estimated using spectrometry. Originally, a single UV source with wavelength of 254 nm was used for such measures. However, recently, instrumentation reading the entire UV–VIS (Ultraviolet–visible spectroscopy) adsorption spectrum (200–750 nm) was introduced (S-can 2015), and UV absorption is now a commonly used methodology. To quantify organic contamination, due to a multitude of substances, cumulative parameters such as chemical oxygen demand (COD), biochemical oxygen demand (BOD) or spectral absorption coefficient are often used. Evidence shows strong correlation between organic carbon measured with UV and that measured with standard methods (Figure 2). In addition, it was shown that several other parameters can be inferred by correlating their concentration values to UV full spectrum absorption (Figure 3). Furthermore, several common organic compounds have absorption spectra that make their identification quite easy with appropriate instrumentation (Figure 4).
Table 6

Carbon fractions measured by organic carbon analyzers (modified from AWWA 2002)

Carbon fractionAbbr.Definition
Total carbon TC Sum of organically and inorganically bound carbon (incl. elemental C) in water 
Total inorganic carbon TIC Sum of elemental carbon, CO2, CO, CN, CS, CCl4, etc. 
Total organic carbon TOC Organic carbon bound to particles <100 μm (TOC = TC–TIC) 
Dissolved organic carbon DOC Organic carbon in water bound to particles <45 μm 
Nonpurgeable organic carbon NPOC OC present after sample scrubbing to eliminate inorg. C and VOCsa 
Volatile organic carbon VOC TOC fraction removed from the sample by gas stripping 
Carbon fractionAbbr.Definition
Total carbon TC Sum of organically and inorganically bound carbon (incl. elemental C) in water 
Total inorganic carbon TIC Sum of elemental carbon, CO2, CO, CN, CS, CCl4, etc. 
Total organic carbon TOC Organic carbon bound to particles <100 μm (TOC = TC–TIC) 
Dissolved organic carbon DOC Organic carbon in water bound to particles <45 μm 
Nonpurgeable organic carbon NPOC OC present after sample scrubbing to eliminate inorg. C and VOCsa 
Volatile organic carbon VOC TOC fraction removed from the sample by gas stripping 

aMost commercial TOC analyzers actually measure NPOC.

Figure 2

Correlation between BOD5 and COD laboratory results and the results measured with a spectrometric probe (S::can website, 2015).

Figure 2

Correlation between BOD5 and COD laboratory results and the results measured with a spectrometric probe (S::can website, 2015).

Figure 3

Correspondence between spectral absorption areas and quality parameters (S::can website, 2015).

Figure 3

Correspondence between spectral absorption areas and quality parameters (S::can website, 2015).

Figure 4

Spectral absorption of benzene, with the typical five-peak shape (S::can website, 2015).

Figure 4

Spectral absorption of benzene, with the typical five-peak shape (S::can website, 2015).

Fluorescence spectroscopy has also been indicated recently as a promising tool for online monitoring of organic matter, although no commercial products exist, yet (Shutova et al. 2014).

In addition to organic matter, hydrocarbons are probably the main class of contaminants found in surface and groundwater. Methods for online detection include: fluorometry, reflectivity, light scattering and turbidity measurement, ultrasonic methods, electrical conductivity, spectroscopy, gas-phase detection (after volatilization), and resistance-based sensors; some methods, however, give merely an indication of the presence/absence of oil.

VOCs, including aromatic compounds, halogenates and trihalomethanes, evaporate when exposed to air, and can be of health concern when found in water (trihalomethanes are disinfection byproducts (DBPs) – possible precursors to formation of suspected carcinogens). Current monitoring technologies for VOCs include purge-and-trap gas chromatography (GC) with flame ionization (FID), electron capture (ECD) or photoionization detectors or mass spectrometry (MS). Most of these methods require skilled operators, purification and pre-concentration, sample injection and results analysis. Detection limits for different substances vary according to the detector method (Yongtao et al. 2000; Capodaglio & Callegari 2009).

Pesticides, including insecticides, fungicides and herbicides, comprise triazines and phenylurea compounds; they are monitored in surface waters in order to detect accidental pollution. Online monitoring of pesticides can be carried out using composite techniques, such as:

  • high-pressure liquid chromatography (HPLC)/diode array (DA) detection;

  • GC separation and mass spectrometer (MS) detection;

  • liquid chromatography/MS.

Each technique is capable of optimally detecting a group of compounds, for example, HPLC/DA can be used for atrazine, chlortoluron, cyanazine, desethylkatrazine, diuron, hexazinone, isoproruton, linuron, metazachlor, methabenzthiazuron, metobrorumon, metolachlor, metoxuron, monolinuron, sebutylazine, simazine and terbutylazine (AWWA 2002).

In theory, any analytical laboratory method can be adapted for online use, provided that requirements for consumables and manual intervention can be minimized: for this reason, current online systems are often a ‘robotized’ adaptation of offline procedures, however, this solution is not always the most efficient. Novel technologies, such as optochemical sensors, biosensors, and microbiological sensors, are being tested for organics and hydrocarbon analysis. Advances already in use include differential UV spectroscopy for DBP detection and microphase solid-phase extraction (SPE) for analysis of semivolatile organics (Yongtao et al. 2000).

A novel LED-based prototype instrument, detecting fluorescence peaks C and T (surrogate parameters for organic and microbial matter, respectively), was recently developed and tested. Although correlating well with regulatory organic surrogate measures, the device did not provide close correlations with regulatory microbial measures (Bridgeman et al. 2015).

LED UV fluorescence sensors adopting a special combination of UV and fluorescence measurements have been tested for online monitoring of dissolved organic matter and to predict DBP formation potential during water treatment. This application has demonstrated the potential applicability of LED UV/fluorescence sensors for online water monitoring (Li et al. 2016).

Biological monitors

Biosensors, defined as devices incorporating a biological, biologically derived, or biomimicking material, integrated within a physicochemical transducer, offer some advantages for environmental analysis, compared to conventional methods, since they are cheap and simple to use, and are frequently able to evaluate complex matrices with minimal sample preparation. Biosensors should be distinguished from bioassays or bioanalytical systems, which require additional sample processing (e.g. reagent addition). Advantages include miniaturization and portability possibilities, permitting their use as on-site devices. In addition to the identification of specific chemicals, some biosensors offer the possibility of measuring biological effects, such as toxicity, cytotoxicity, genotoxicity, or endocrine disrupting effects.

This information could be, in some cases, more relevant than specific chemical composition. Online biological monitors are an active area of R&D due to increasing regulatory and public demand. At this time, many biological monitors are relatively new and can still be considered experimental/unique laboratory-based applications, although commercial tests have started in water monitoring, for BOD, nitrate and pesticide assessment (Bahadır & Sezgintürk 2015).

Table 7 shows an overview of the most common types of online biological monitors. Table 8 summarizes the comparative features of biosensors versus current online LC–MS methods (Rodriguez-Mozaz et al. 2007).

Table 7

Common online biological monitors

TechnologyMeasurementComments
Fish tests Swimming pattern Low sensitivity 
Ventilation rate Sophisticated requirements 
Bioelectric field Requires exotic ‘electric fish’ species 
Avoidance patterns Interpretation complex 
Daphnid tests Swimming activity Good performance, no determination of causes 
Behavior 
Mussel tests Shell positions/opening Can concentrate pollutants to levels many times greater than found in the water. Long organism lifespan. Similar results over different species 
Algae tests Fluorescence (photosynthesis) Commercial monitors available 
Bacteria tests Luminescence Respiration of nitrifiers Commercially available, toxicity data for over 1,000 compounds 
Chlorophyll-a Fluorometry Interference with pigments, diss. organics, sensitive to environmental variables 
Chlorophyll-a and algal absorption Reflectance radiometry Commercial systems available 
Protozoan monitors Measurement requires preliminary concentration/centrifugation of sample By filtration on membrane cartridge 
Laser scanning cytometry W/modified blood cell separators, minimal operation time 
Particle characterization Analysis possible within 3 min, particles must be confirmed by trained operator 
UV spectroscopy Measure particle size/distribution, high number of false positive and negative results 
Multi-angle light scattering Online system, unlabeled parasites, differentiation problems 
Nucleic acid molecules and magnetized microbeads Successfully tested in laboratory 
Oocysts detected within 20 min, not fully automated 
TechnologyMeasurementComments
Fish tests Swimming pattern Low sensitivity 
Ventilation rate Sophisticated requirements 
Bioelectric field Requires exotic ‘electric fish’ species 
Avoidance patterns Interpretation complex 
Daphnid tests Swimming activity Good performance, no determination of causes 
Behavior 
Mussel tests Shell positions/opening Can concentrate pollutants to levels many times greater than found in the water. Long organism lifespan. Similar results over different species 
Algae tests Fluorescence (photosynthesis) Commercial monitors available 
Bacteria tests Luminescence Respiration of nitrifiers Commercially available, toxicity data for over 1,000 compounds 
Chlorophyll-a Fluorometry Interference with pigments, diss. organics, sensitive to environmental variables 
Chlorophyll-a and algal absorption Reflectance radiometry Commercial systems available 
Protozoan monitors Measurement requires preliminary concentration/centrifugation of sample By filtration on membrane cartridge 
Laser scanning cytometry W/modified blood cell separators, minimal operation time 
Particle characterization Analysis possible within 3 min, particles must be confirmed by trained operator 
UV spectroscopy Measure particle size/distribution, high number of false positive and negative results 
Multi-angle light scattering Online system, unlabeled parasites, differentiation problems 
Nucleic acid molecules and magnetized microbeads Successfully tested in laboratory 
Oocysts detected within 20 min, not fully automated 
Table 8

Comparative features of online SPE–LC–MS methods vs biosensors for environmental analysis (modified from Rodriguez-Mozaz et al. 2007)

Online SPE–LC–MSBiosensors
Comparatively higher sample volumes of water are necessary Small sample volumes are sufficient to obtain enough sensitivity 
Matrix effect; ionic suppression or enhancement in MS spectrometry Matrix effects. Variable depending on biorecognition principle and transduction element 
Preconcentration of the sample necessary (SPE) Direct analysis of the sample. Minimal sample preparation 
Multi-residue analysis Limited multi-analyte determination 
Automatization and minimal sampling handling Possible automatization of the system 
Direct and fast elution of the sample after preconcentration. Minimal degradation Direct analysis after sampling is possible. Minimal degradation 
No biological stability restrictions Low biological material stability 
Determination of chemical composition Determination of biological effect and of bioavailable pollutant content 
Compound selectivity by using specific sorbents (MIPs and immunosorbents) Compound selectivity by using specific biological recognition element 
Minimal consumption of organic solvents (elution with the LC mobile phase) Consumption of organic solvents avoided. Direct analysis of contaminant in water 
Generation of organic solvent waste Minimal and non-contaminating waste 
Short analysis time and high throughput Faster analysis. Real-time detection and high throughput 
Limited portability. Laboratory confined Availability of portable biosensor systems 
Applicability to early-warning and on-site monitoring Applicability to early-warning and on-site monitoring 
Qualified personnel required Qualified personnel not required. User friendly 
Expensive equipment Cost-effective equipment 
Online SPE–LC–MSBiosensors
Comparatively higher sample volumes of water are necessary Small sample volumes are sufficient to obtain enough sensitivity 
Matrix effect; ionic suppression or enhancement in MS spectrometry Matrix effects. Variable depending on biorecognition principle and transduction element 
Preconcentration of the sample necessary (SPE) Direct analysis of the sample. Minimal sample preparation 
Multi-residue analysis Limited multi-analyte determination 
Automatization and minimal sampling handling Possible automatization of the system 
Direct and fast elution of the sample after preconcentration. Minimal degradation Direct analysis after sampling is possible. Minimal degradation 
No biological stability restrictions Low biological material stability 
Determination of chemical composition Determination of biological effect and of bioavailable pollutant content 
Compound selectivity by using specific sorbents (MIPs and immunosorbents) Compound selectivity by using specific biological recognition element 
Minimal consumption of organic solvents (elution with the LC mobile phase) Consumption of organic solvents avoided. Direct analysis of contaminant in water 
Generation of organic solvent waste Minimal and non-contaminating waste 
Short analysis time and high throughput Faster analysis. Real-time detection and high throughput 
Limited portability. Laboratory confined Availability of portable biosensor systems 
Applicability to early-warning and on-site monitoring Applicability to early-warning and on-site monitoring 
Qualified personnel required Qualified personnel not required. User friendly 
Expensive equipment Cost-effective equipment 

At the moment, bacterial-based systems (Kim & Gu 2005) show poor sensitivity and low ease of operation. Developments will likely derive from improved fingerprinting of organisms, and cost reduction. Significant advances can be expected from protozoan monitor technology, with UV absorption/scattering analytical techniques that may soon allow automated detection of Cryptosporidium and Giardia. Molecular techniques initially applied to the recognition of organisms’ genomic sequence in clinical applications (Bej 2003) have shown great potential for detection of pathogens in water, and are producing interesting results that could soon lead to widespread online use. Very recently, a prototype automated biosensor for fast (8 hours) identification and quantification of Escherichia coli contaminations in ground, surface and drinking water was proposed and tested. The instrument is based on a three-electrode potentiostat using electrochemical assays to detect E. coli using their β-galactosidase activity (Ettenauer et al. 2015).

Microbial fuel cells (MFCs), biological systems capable of degrading organic matter with direct generation of electrical energy, intensively investigated as an alternative to traditional wastewater treatment processes (Capodaglio et al. 2013, 2016b), have recently been highlighted as a technology with potential for rapid and simple testing of water quality. The underlying principle is that the current generated by an MFC directly relates to the metabolic activity of the electroactive biofilm at the anode surface, thus any disturbances of their metabolic pathways are translated into a change in electricity production (Molognoni et al. 2016). Their application, supported by interpretation software, would not be limited to organic carbon, but also to water toxicity and specific compounds (Chouler & Di Lorenzo 2015; Yang et al. 2015). For interpretation of these data, the use of artificial neural networks, often adopted in wastewater-related modelling (Raduly et al. 2007), has been proposed.

Molecular methods for detection of microbial pathogens have in fact been established, however, most of these have important limitations, associated with the time necessary to isolate and/or identify the pathogen and detection accuracy. Research towards their improvement relies on methods of culture on selective media, immunological approaches, nucleic acid-based assays and DNA microarrays (Lemarchand et al. 2004). Molecular fingerprinting was recently demonstrated as an effective monitoring tool for detection of cyanobacteria in surface waters (Loza et al. 2013).

Fingerprinting

Fingerprinting methods describe the use of a unique chemical signature, isotopic ratio, mineral species, or pattern analysis to identify different chemicals. Optical fingerprinting by UV, VIS, and near-infrared (NIR) absorption spectroscopy can be effectively achieved by low-cost and compact devices that can be linked to an online diagnostic system, to directly identify some compounds (e.g. benzene, Figure 4) present in the water, or to indicate the possibility of their presence.

A fingerprint contains much more information about water quality than a single-wavelength instrument can provide, allowing more accurate and comprehensive assessments. In optical fingerprinting, a wide portion of the UV, VIS and NIR spectrum is monitored simultaneously at high measurement frequencies (minutes or fractions); Figure 5 shows, as an example, spectral fingerprinting of a municipal wastewater, with three spectral readings, in the wavelength range 230–730 nm, recorded at the same point within a short interval. Individual spectra show clearly different features, indicating a pronounced water-quality change occurring in the 18 minutes elapsed since the first reading. Although this indication alone, in general, will not individuate the compound(s) responsible, it can nevertheless trigger an alert to the operator, indicating deviation from routine conditions. Fingerprinting is used, in conjunction with sophisticated algorithms and statistical software, in CWS or EDS, described below.
Figure 5

Optical fingerprinting in a pipe, indicating rapid water-quality changes (S::can website, 2015).

Figure 5

Optical fingerprinting in a pipe, indicating rapid water-quality changes (S::can website, 2015).

Spectral photometric (spectrometric) methods are probably the most interesting, currently available mature technology to cover most online monitoring needs, and specifically fingerprinting. They are recommended by the US-EPA (EPA 2013) for online monitoring over traditional analytical techniques, having been tested for online drinking-water quality monitoring applications, instead of traditional reagent-based analyzers (EPA 2009).

The main features that have contributed to the wide acceptance of spectrometric methods, in comparison to photometric ones, are:

  • cost efficiency: the continuous UV/VIS spectrum enables simultaneous measurement of, e.g., organic carbon, nitrate and turbidity, for which only one spectrometer is required, instead of three photometers;

  • lower cross-sensitivity to turbidity, coloration, surface growth, etc.: potential interferences, not detectable by single/dual wavelength measurement, are nearly always compensated using spectral information;

  • greater precision, higher selectivity and reproducibility: since cross-sensitivity is substantially reduced, heterodyning of signals due to interference/noise is significantly less than with photometers; furthermore, individual substances and/or groups can be allocated to specific spectral features, resulting in very high reproducibility, without the absolute necessity of specific calibration;

  • qualitative evaluation: in addition to calibrated parameters, qualitative spectral information contained in the ‘fingerprint’ can be directly applied for alarm and control systems.

Continuous, consistent online data obtained by such instruments can be used to extract useful information about a water system not otherwise available. As shown in Figure 6, nitrate profiles measured continuously are compared to calculate travel times between monitoring sites, determine water age, and verify the network's hydraulic model (Thompson & Kadiyala 2013).
Figure 6

Comparison of continuous nitrate profiles in different sections of a distribution network (S::can website, 2015).

Figure 6

Comparison of continuous nitrate profiles in different sections of a distribution network (S::can website, 2015).

Integrated CWS

Following the ‘9–11’ events in the USA and the completion and review of a risk assessment procedure for public water systems serving populations greater than 3,300, as mandated by the US Bioterrorism Act of 2002, water distribution systems were identified as one of the most vulnerable areas of attack from potential terrorist or extremist groups. In consideration of deliberate hostile actions on water supplies, although none has been reported to date, in January 2002 the FBI circulated a reserved bulletin warning water industry managers that al-Qaida may have been studying American water-supply systems in preparation for attacks (IonLife 2002).

Homeland Security Presidential Directive 9 required the US Environmental Protection Agency (US-EPA) to develop a program for utilities to improve protection of their water distribution systems.

Online quality monitoring of distribution systems has been investigated extensively for some time (Grayman et al. 2001; Hasan et al. 2004; CH2MHill 2013), generating so-called CWS, or Water Quality EDS. These consist of integrated in situ sensors, SCADA systems, designed to continuously monitor network conditions and warn of any potential contamination events. In addition to security issue detection, their benefits may be categorized as operational enhancements, regulatory compliance, and contamination warning. Operational enhancements include continuous indication of water quality in the distribution system beyond that possible through routine regulatory sampling. Early indications of water quality problems may consist of unusually low residual chlorine, impending nitrification (elevated ammonia), turbidity excursions caused by mains breaks, and other unusual quality changes. Monitoring is achieved through measurement of parameters already familiar to utilities (e.g., chlorine residual), and/or other parameters relatively new to these applications (e.g., TOC). The US-EPA recommends monitoring four key parameters, namely: TOC, pH, conductivity, and chlorine (EPA 2009), while a later study, considering all instruments and technologies available, suggested the following parameters for online water-quality monitoring systems: conductivity, chlorine (combined), pH, oxidation/reduction potential (ORP), temperature, turbidity, and UV absorption (CH2MHill 2013).

Regulatory compliance benefits include the ability to maintain proper chlorine residuals and pH control in the network (to avoid Pb and Cu leaching from pipes). Warning of intentional or unintentional contamination in distribution systems is somewhat more complex. Specialized analyzers are available, including GCs that may detect specific contaminants and toxicity monitors that can provide general warnings. Due to the large number of potential contaminants, however, it is more practical to monitor for indications of contamination through changes in the same water quality parameters, or surrogates, often used for operational monitoring (Table 9).

Table 9

Parameters typically included in operational water quality monitoring systems

ParameterSignificance
TOC Total organic carbon. Elevated turbidity excursions can be associated with a breakthrough at the water treatment plant or scouring and release of biofilm within the distribution system. 
Residual chlorine A sudden loss in residual could promote biofilm growth and potential violation of the Total Coliform Rule. 
Conductivity Its measurement provides an easy method for identifying mixing or different water sources, which can have a significant impact on many industrial operations. 
pH Controlled for disinfection and corrosion control. The formation of some disinfection byproducts is pH dependent. 
Turbidity Provides warning of a system disruption created by a surge or reversal in flow that scours the pipeline. This could be caused by a pipeline break, hydrant knockover, or other problems that will impact chlorine residual and customer satisfaction. 
ParameterSignificance
TOC Total organic carbon. Elevated turbidity excursions can be associated with a breakthrough at the water treatment plant or scouring and release of biofilm within the distribution system. 
Residual chlorine A sudden loss in residual could promote biofilm growth and potential violation of the Total Coliform Rule. 
Conductivity Its measurement provides an easy method for identifying mixing or different water sources, which can have a significant impact on many industrial operations. 
pH Controlled for disinfection and corrosion control. The formation of some disinfection byproducts is pH dependent. 
Turbidity Provides warning of a system disruption created by a surge or reversal in flow that scours the pipeline. This could be caused by a pipeline break, hydrant knockover, or other problems that will impact chlorine residual and customer satisfaction. 

Note: Utilities that use chloramines for disinfection should also measure ammonia, nitrates, and DOC to provide early warning of nitrification in the distribution system. The first water quality indicator of nitrification will be the increase of ammonia, which will occur before nitrites and nitrates begin to increase.

Real-time monitoring strategies are the answer to detecting low probability/high impact events at an early stage, while chronic or long-term risks should still be monitored with traditional sampling. Considering the safety of drinking-water supplies, realistic detection limits of available online instrumentation must be taken into account (Figure 7). These could be enhanced by combining technologies, e.g. spectrometry with online toxicity tests (Weingartner 2013), but often their sensitivity will not suffice to warn about possible long-term contaminant effects.
Figure 7

An example of realistic detection limits of online sensors (redrawn from Weingartner 2013).

Figure 7

An example of realistic detection limits of online sensors (redrawn from Weingartner 2013).

An ideal EDS will react to most types of threat agents, at concentrations far below the LD50 lethal dose (in turn, much higher than drinking limits), provide distinct signals to each threat, not respond to harmless substances or operational fluctuations (i.e. ‘false alarms’), and have a ‘fast’ (real-time) response. The sensoristic component can consist of various sensing platforms, including contaminant-specific sensors, or quality sensors (e.g., pH, Cl, electrical conductivity, etc.) currently installed in many municipal water distribution systems to provide ‘surrogate’ data to the CWSs.

Table 10 lists in-pipe physical and chemical parameters that can be reliably measured along with current technologies. The most difficult issue is to distinguish actual contamination from natural fluctuations of the water matrix. Figure 8 summarizes capability, reliability and O&M requirements for existing physico-chemical in-pipe sensors.
Table 10

In-pipe measurable physical and chemical parameters with currently available technology

 Available technology
General parameters 
 Pressure On-chip 
 Temperature On-chip PPT 
 pH On-chip or electrolyte 
 ORP On-chip or electrolyte 
 Conductivity On-chip 
 Dissolved oxygen (on-chip) On-chip 
 Chlorine (free, total) On-chip amperometric; use of ORP 
 Chloramines: calculated from total free chlorine On-chip amperometric; use of ORP 
General optical parameters 
 Turbidity (optical) Optical 
 Color Optical 
 UV254–simple surrogate organics indicator Optical 
 Spectral TOC/DOC–broad organics detector Optical 
 UV spectral alarms Optical 
Spectral parameters for special purpose 
 NH4 (chloraminating systems): ISE ISE 
 NO2 (chloraminating systems): spectral hi-resolution UV-VIS Optical 
 NO3 (groundwater under agricultural influence): spectral UV-VIS Optical 
 Hydrocarbon alarm: UV-VIS or fluorescence Optical 
Other important parameters that no sensors exist for 
 Arsenic None 
 Endocrine disruptors None 
 Pesticides/Herbicides None 
 Available technology
General parameters 
 Pressure On-chip 
 Temperature On-chip PPT 
 pH On-chip or electrolyte 
 ORP On-chip or electrolyte 
 Conductivity On-chip 
 Dissolved oxygen (on-chip) On-chip 
 Chlorine (free, total) On-chip amperometric; use of ORP 
 Chloramines: calculated from total free chlorine On-chip amperometric; use of ORP 
General optical parameters 
 Turbidity (optical) Optical 
 Color Optical 
 UV254–simple surrogate organics indicator Optical 
 Spectral TOC/DOC–broad organics detector Optical 
 UV spectral alarms Optical 
Spectral parameters for special purpose 
 NH4 (chloraminating systems): ISE ISE 
 NO2 (chloraminating systems): spectral hi-resolution UV-VIS Optical 
 NO3 (groundwater under agricultural influence): spectral UV-VIS Optical 
 Hydrocarbon alarm: UV-VIS or fluorescence Optical 
Other important parameters that no sensors exist for 
 Arsenic None 
 Endocrine disruptors None 
 Pesticides/Herbicides None 
Figure 8

Synthetic assessment of capability, reliability and O&M requirements for existing physico-chemical in-pipe sensors: for the most common online quality parameters, a score representing an assessment of current technology in terms of selectivity, reliability and maintenance requirements is given. Higher scores generally indicate ‘better’ performance (except in the case of maintenance, where a lower score indicates lower requirements) (redrawn from Weingartner 2013).

Figure 8

Synthetic assessment of capability, reliability and O&M requirements for existing physico-chemical in-pipe sensors: for the most common online quality parameters, a score representing an assessment of current technology in terms of selectivity, reliability and maintenance requirements is given. Higher scores generally indicate ‘better’ performance (except in the case of maintenance, where a lower score indicates lower requirements) (redrawn from Weingartner 2013).

Laboratory experiments and pipe system tests have proved that a majority of potential contaminants will change the value of at least one surrogate parameter from normal background levels (Byer & Carlson 2005; Cook et al. 2005; Hall et al. 2007). Surrogate parameters can therefore provide valid information on the presence of contaminants within a distribution system. The challenge then is to analyze surrogate parameter signals to identify changes that are significantly beyond the range of natural ambient variability of the background water quality, or to establish a detection baseline. Detection of events by simply using upper and lower thresholds of parameter concentration is virtually impossible; hence complex pattern-recognition algorithms are indispensable. These can be implemented in advanced, user-friendly event-detection software, maintaining use of any already-installed sensors for event detection and water protection, resulting at the moment in the most economical and effective solution to distribution system security (Weingartner 2013).

EDS implementation examples

The CANARY EDS software (Hart et al. 2007) is an open-source software platform developed by the US-EPA. It gathers water quality inputs from SCADA systems and processes the data using event detection algorithms and statistical models to determine the probability of an anomalous event occurring within the distribution network (McKenna & Hart 2008).

CANARY was tested along with four other proprietary EDS software tools under real-life conditions by the US-EPA (EPA 2013). The results of this evaluation were encouraging, as the conclusions regarding EDS performance showed that event detection is possible, but the ability to detect anomalous conditions strongly depends on EDS configuration, baseline variability of the monitoring location, and the nature of the change, with overall positive response.

The US-EPA runs the Water Security Initiative (WSI), a nationwide project to support investigation and deployment of large water-security systems, based on distributed spectral (and other) sensors, centralized data collecting stations, and on several types of event detection software.

The first implementation of a working water security system in the USA was in Glendale (Arizona) with about ten monitoring stations fully integrated into a central database (Thompson 2008). In New York City, all existing manual sampling stations are being converted into monitoring stations with the use of online spectral monitors: over 30 have been already installed, and the proprietary software Moni::Tool was selected as the event detection software (EPA 2013). Philadelphia's water utility developed a comprehensive CWS for its drinking water system under an EPA-WSI grant, where selection of available instrumentation was made by field comparison between products from five different manufacturers (PWD & CH2MHill 2013). The City of Dallas (Texas) developed a CWS consisting of four spectral UV-VIS monitors based at the water treatment facilities, giving a ‘fingerprint’ of the water leaving each plant and 32 distribution monitors providing continuous water quality analysis at 16 checkpoints in the system. Monitoring is supported by an EDS constantly checking for anomalies in the background. Reported benefits of expanded water quality monitoring capabilities are a 24/7 view of water quality available to staff, including parameters such as nitrate, total chlorine turbidity, TOC, conductivity, UV spectral absorbance, DOC, pH, and free ammonia. All are web-accessible to the city's water operators and constitute valuable information for the detection of water quality changes from intentional or unintentional actions, natural phenomena and/or problems at treatment plants (Sanchez & Brashear 2011).

Bratislava Water Company (BVS) is responsible for the operation of the water and wastewater systems of the capital of Slovakia, supplying a population of over 600,000. Drinking water is produced in seven treatment facilities from more than 150 groundwater sources. Given the high quality of raw water the only treatment is chlorination, to prevent microbiological growth during distribution. To ensure that contamination of a source would not compromise overall high quality, BVS implemented a monitoring system that oversees all sources, coupled with an EDS sending an alarm in case of an unexpected event. Measured parameters include TSS, turbidity, NO3-N, COD, BOD, TOC, DOC, UV254, color, benzene, toluene, xylene (BTX), O3, H2S, assimilable organic carbon (AOC), temperature and pressure. Monitoring occurs by means of submersible online spectrometric probes combined with a centralized system, continuously analyzing four spectral alarm parameters from each site. An evaluation of the EDS showed that the spectral alarm system is able to detect contamination events down to 100 μg/L TOC, 25 μg/L carbendazim, 100 μg/L benzene and 50 μg/L saxitoxin (BVS 2013).

Under an EU-funded project, SMaRT-OnlineWDN, a group of European research and industrial partners is currently investigating the development of an online security management system for WDN based on sensor measurements of water quality and quantity, with planned applications ranging from the detection of deliberate contamination, to improved operation and control of a WDN under normal and stress conditions. An online running model, automatically calibrated to the measured sensor data, will give detailed information on contamination sources (localisation and intensity) (SMART-Online WDN 2016).

DISCUSSION AND CONCLUSIONS

In this paper, an overview of existing instrumentation applicable to water and wastewater online monitoring and forthcoming developments has been given, and a few state-of-the-art application examples examined. It is clear that technological development in this field is very rapid, and that astonishing advances are anticipated in several areas (fingerprinting, opto-chemical sensors, biosensors, molecular techniques). Some of the technologies mentioned, although promising, are not at commercial or at online, standalone application stage. Software applications, together with new-generation sensors, are also contributing to the identification of otherwise difficult-to-monitor parameters. In some cases, the presence of contaminants not directly observable online can be inferred by water property (e.g. absorbance) changes or by indirect indicators (with statistical analysis software), giving rise to the possibility of water quality ‘alerts’, pending more detailed analysis with traditional methods. Examples of CWS applications, arisen from perceived threats to the safety of water supply networks, have also been illustrated. CWS is perhaps the sector in which more rapid development is expected in the coming years, possibly creating a drive for further technological breakthroughs.

In spite of high technology instrumentation being developed, monitoring costs are bound to become a lesser and lesser part of water utility budgets due to the fact that automation and technological simplification will abate human cost factors (maintenance and other labor forms) and significantly reduce the complexity of procedures (sample preparation, reagent requirements, etc.).

Proper interpretation and use of the growing mass of water quality data that will become available through new monitoring and information technologies will allow better management of water resources, and of water/wastewater treatment facilities.

REFERENCES

REFERENCES
Allen
M.
Preis
A.
Iqbal
M.
Srirangarajan
S.
Lim
H. B.
Girod
L.
Whittle
A. J.
2011
Real-time in-network distribution system monitoring to improve operational efficiency
.
J. AWWA
103
(
7
),
63
75
.
AWWA
2002
Online monitoring for drinking water utilities
. In:
Cooperative Research Report
(
Hargesheimer
E.
Conio
O.
Popovicova
J.
, eds).
AWWA Research Foundation
,
Denver, CO, USA
.
Bridgeman
J.
Baker
A.
Brown
D.
Boxall
J. B.
2015
Portable LED fluorescence instrumentation for the rapid assessment of potable water quality
.
Sci. Total Environ.
524–525
,
338
346
.
Bullough
F.
Fenech
C.
Bridle
H.
2013
Advances in water quality monitoring of inorganics: current trends
.
Journal of Water Resource and Protection
5
,
40
48
.
BVS
2013
Early Warning System for Monitoring of Bratislava's Water Supply
.
Internal Report
,
BVS
,
Bratislava, Slovakia
.
Byer
D.
Carlson
K. H.
2005
Real-time detection of intentional chemical contamination in the distribution system
.
American Water Works Association Journal
97
(
7
),
130
133
.
Capodaglio
A. G.
Callegari
A.
2009
Online Monitoring Technologies for Drinking Water Systems Security
.
Risk Management of Water Supply and Sanitation Systems, NATO Science for Peace and Security Series
.
Springer
,
Dordrecht, The Netherlands
.
Capodaglio
A. G.
Callegari
A.
2015
Water supply systems security: novel technologies for the online monitoring of unforeseeable events
.
WIT Transactions on The Built Environment, Safety and Security Engineering VI
51
,
251
263
.
Capodaglio
A. G.
Molognoni
D.
Dallago
E.
Liberale
A.
Cella
R.
Longoni
P.
Pantaleoni
P.
2013
Microbial fuel cells for direct electrical energy recovery from urban wastewaters
.
The Scientific World Journal
2013
,
634738
.
Capodaglio
A. G.
Callegari
A.
Molognoni
D.
2016a
Online monitoring of priority and dangerous pollutants in natural and urban waters: a state-of-the-art review
.
Management of Environmental Quality: An International Journal
27
(
5
),
507
536
.
Capodaglio
A. G.
Molognoni
D.
Vilajeliu Pons
A.
2016b
A multi-perspective review of microbial fuel-cells for wastewater treatment: bio-electro-chemical, microbiologic and modeling aspects
.
AIP Conference Proceedings
1758
(
1
),
030032
.
CH2M HILL
2013
Contamination Warning System Demonstration Pilot Project: Implementation and Assessment
.
Internal Report
,
CH2MHill
,
Philadelphia, PA, USA
.
Cook
J.
Roehl
E.
Daamen
R.
Carlson
K.
Byer
D.
2005
Decision support system for water distribution system monitoring for homeland security
. In:
Proceedings of the AWWA 2005 Water Security Conference
,
Denver, CO, USA, WSC61381
.
Copetti
D.
Valsecchi
L.
Sanfilippo
L.
Capodaglio
A. G.
Moscetta
P.
Tartari
G.
2014
Real-time evaluation of nutrient concentrations in surface waters using a portable probe
. In:
Proceedings, IWA XIII Specialist Conference on Watershed and River Basin Management
,
9–14 September
,
San Francisco
,
USA
.
Daughton
C. G.
Ternes
T. A.
1999
Pharmaceuticals and personal care products in the environment: agents of subtle change?
Environmental Health Perspectives
107
(
6
),
907
915
.
EPA
2009
Distribution System Water Quality Monitoring: Sensor Technology Evaluation Methodology and Results
.
A Guide for Sensor Manufacturers and Water Utilities, EPA 600/R-09/076, US Environmental Protection Agency Office of Research and Development, Water Infrastructure Protection Division
,
National Homeland Security Research Center
,
Cincinnati, OH, USA
.
EPA
2013
Water Quality Event Detection System Challenge: Methodology and Findings
.
Office of Water (MC-140) EPA 817-R-13-002
,
Cincinnati, OH, USA
.
Ettenauer
J.
Zuser
K.
Kellner
K.
Posnicek
T.
Brandl
M.
2015
Development of an automated biosensor for rapid detection and quantification of E. coli in water
.
Procedia Engineering
120
,
376
379
.
Glasgow
H. B.
Burkholder
J. M.
Reed
R. E.
Lewitus
A. J.
Kleinman
J. E.
2004
Real-time remote monitoring of water quality: a review of current applications, and advancements in sensor, telemetry, and computing technologies
.
Journal of Experimental Marine Biology and Ecology
300
,
409
448
.
Grayman
W. M.
Deininger
R. A.
Males
R. M.
2001
Design of Early Warning and Predictive Source Water Monitoring Systems
.
AWWA Research Foundation
,
Denver, CO, USA
.
Hall
J.
Zaffiro
A. D.
Marx
R. B.
Kefauver
P. C.
Krishnan
E. R.
Haught
R. C.
Herrmann
J. C.
2007
On-line water quality parameters as indicators of distribution system contamination
.
American Water Works Association Journal
99
(
1
),
66
72
.
Hanrahan
H.
Patila
D. G.
Wang
J.
2004
Electrochemical sensors for environmental monitoring: design, development and applications
.
Journal of Environmental Monitoring
6
,
657
664
.
Hart
D. B.
McKenna
S. A.
Klise
K. A.
Cruz
V.
Wilson
M.
2007
CANARY: A water quality event detection algorithm development tool
. In:
Proceedings ASCE World Environmental and Water Resources Congress 2005
(K. C. Kabbes, ed.). ASCE, Reston, VA, USA
.
Hasan
J.
States
S.
Deininger
R.
2004
Safeguarding the security of public water supplies using early warning systems: a brief review
.
Journal of Contemporary Water Research and Education
129
,
27
33
.
IonLife
2002
How safe is our water? The threat of terrorism. Available at IonLife website http://www.ionlife.org (accessed July 2008)
.
Jones
O. A. H.
Voulvoulis
N.
Lester
J. N.
2002
Human pharmaceuticals in the aquatic environment
.
Environ. Technol.
22
(
12
),
1383
.
Jothimuthu
P.
Wilson
R. A.
Herren
J.
Haynes
E. N.
Heineman
W. R.
Papautsky
I.
2011
Lab-on-a-chip sensor for detection of highly electronegative heavy metals by anodic stripping voltammetry
.
Biomedical Microdevices
13
(
4
),
695
703
.
Kissinger
P. T.
Heineman
W. R.
(eds)
1996
Laboratory Techniques in Electroanalytical Chemistry 2
.
Marcel Dekker
,
New York
.
Kolpin
D. W.
Furlong
E. T.
Meyer
M. T.
Thurman
E. M.
Zaugg
S. D.
Barber
L. B.
Buxton
H.
2002
Pharmaceuticals, hormones, and other organic wastewater contaminants in U.S. streams, 1999–2000: a national reconnaissance
.
Environmental Science & Technology
36
,
1202
1211
.
Lemarchand
K.
Masson
L.
Brousseau
R.
2004
Molecular biology and DNA microarray technology for microbial quality monitoring of water
.
Critical Reviews in Microbiology
30
(
2
),
145
172
.
Loza
V.
Perona
E.
Mateo
P.
2013
Molecular fingerprinting of cyanobacteria from river biofilms as a water quality monitoring tool
.
Applied Environmental Microbiology
79
(
5
),
1459
1472
.
McKenna
S. A.
Hart
D.
2008
On-line identification of adverse water quality events from monitoring of surrogate data: CANARY software
. In:
Proceedings, Singapore International Water Week Conference
.
ModernWater
2015
.
Molognoni
D.
Puig
S.
Balaguer
M.
Capodaglio
A. G.
Callegari
A.
Colprim
J.
2016
Multiparametric control for enhanced biofilm selection in microbial fuel cells
.
J. Chem. Technol. Biotechnol.
91
(
6
),
1720
1727
.
Nunes
C. N.
Pauluk
L. E.
dos Anjos
V. E.
Chierici Lopes
M.
Quináia
S. P.
2015
New approach to the determination of contaminants of emerging concern in natural water: study of alprazolam employing adsorptive cathodic stripping voltammetry
.
Analytical and Bioanalytical Chemistry
5
,
1
9
.
O'Halloran
R.
Fogelman
S.
Zhao
H.
2009
Current Online Water Quality Monitoring Methods and Their Suitability for the Western Corridor Purified Recycled Water Scheme
.
Urban Water Security Research Alliance Technical Report No. 10
,
UWSRA
,
City East, Queensland
,
Australia
.
Öllers
S.
Singer
H. P.
Fässler
P.
Müller
S. R.
2001
Simultaneous quantification of neutral and acidic pharmaceuticals and pesticides at the low-ng/l level in surface waters
.
Journal Chromatography A
911
(
2
),
225
231
.
PWD & Ch2MHill
2013
Philadelphia Water Department Contamination Warning System Demonstration Pilot Project: Selection of Online Water Quality Monitoring Technologies and Station Design
.
White Paper submitted to EPA as part of the Water Security Initiative Grant awarded to Philadelphia Water Department
,
Philadelphia, PA
,
USA
.
Raduly
B.
Gernaey
K. V.
Capodaglio
A. G.
Mikkelsen
P. S.
Henze
M.
2007
Artificial neural networks for rapid WWTP performance evaluation: methodology and case study
.
Environ. Model. Software
22
,
1208
1216
.
Sanchez
C.
Brashear
K.
2011
City of Dallas Water Utilities Contamination Warning System. In City of Dallas Dept. of Public Works website http://dallascounty.org/department/pubworks/media/InfrastructureForum_2011 (accessed June 2015)
.
S-can
2015
S::can company website http://www.s-can.at (accessed May–July 2015)
.
Shao
Y.
Wang
J.
Wu
H.
Liu
J.
Aksay
I. A.
Lina
Y.
2010
Graphene based electrochemical sensors and biosensors: a review
.
Electroanalysis
22
(
10
),
1027
1036
.
SmartOnlineWDN
2016
Online Security Management and Reliability Toolkit for Water Distribution Networks. Project website
.
Accessible on http://www.smart-onlinewdn.eu/ (accessed 4 May 2016)
.
Storey
M. V.
van der Gaag
B.
Burns
B. P.
2011
Advances in on-line drinking water quality monitoring and early warning systems
.
Water Research
45
(
2
),
741
747
.
Tengberg
A.
Hovdenes
J.
Andersson
H.
Brocandel
O.
Diaz
R.
Hebert
D.
Arnerich
T.
Huber
C.
Kortzinger
A.
Khripounoff
A.
Rey
F.
Ronning
C.
Schimanski
J.
Sommer
S.
Stangelmayer
A.
2006
Evaluation of a lifetime-based optode to measure oxygen in aquatic systems
.
Limnology and Oceanography Methods
4
,
7
17
.
Thompson
K. A.
2008
Operational enhancements resulting from the development and implementation of a contaminant warning system for Glendale (AZ)
. In:
Proceedings 2008 AWWA Water Security Congress
.
Denver, CO, USA
.
Thompson
K.
Kadiyala
R.
2013
Protecting water quality and public health using a smart grid
.
Computational Water, Energy, and Environmental Engineering
2
,
73
80
.
Weingartner
A.
2013
Real-time technologies for early contamination warning in water security systems
.
Presentation given at ERNCIP Operator's Workshop, 11–12 September, Brussels, Belgium
.
Xie
X.
Zhai
J.
Bakker
E.
2014
pH independent nano-optode sensors based on exhaustive ion-selective nanospheres
.
Analytical Chemistry
8
6
(
6),
2853
2856
.
Yang
W.
Wei
X.
Choi
S.
2015
A two-channel bacteria-based biosensor for water quality monitoring
.
In: 2015 IEEE Sensors, pp. 1–4
.
Yongtao
L.
George
E. J.
Hansen
E. M.
Toma
J. J.
Werner
M.
Smith
P. D.
2000
SPE Twin-PAL. Online in situ monitoring of semivolatile organic compounds in water
. In:
Proceedings, AWWA Water Quality Technology Conference
,
Salt Lake City, UT, USA
.
Yue
W.
Bange
A.
Riehl
B. L.
Riehl
B. D.
Johnson
J. M.
Papautsky
I.
Heineman
W. R.
2012
Manganese detection with a metal catalyst free carbon nanotube electrode: anodic versus cathodic stripping voltammetry
.
Electroanalysis
24
(
10
),
10
18
.