Microbiological Sensors for the Drinking Water Industry

Bo Højris, Torben Lund Skovhus

• Publisher: IWA Publishing
• Year: 2018
• eBook ISBN: 9781780409559
• Topic: Biological Sciences
• Subtopic: Applied Sciences
• Index: Biological Sciences

This most up-to date book addresses the interdisciplinary area of drinking water quality monitoring by microbiological sensors. It is edited and written by leading water professionals and experts, and binds together interests and competences within sensing technology, system behavior, business needs, legislation, education, data handling, and intelligent response algorithms. The book contains chapters on: - the history of water monitoring and early sensors - the current landscape of microbiological sensors and different measuring concepts - needs from the water industry - detailed description of several state-of-the-art sensor technologies - water industry case stories with operator experience - examples of on-line data collection and data handling - microbiological sensors applied in education and industry workshops - regulator aspects of introducing microbiological sensors It is the hope that the book will be widely used by water utility managers, technical staff working with drinking water safety, educators in the field of water quality and regulators in this field worldwide. Furthermore, that it will help bridge the gaps between these diverse and otherwise differently oriented water professionals.

© IWA Publishing 2018. Torben Lund Skovhus and Bo Højris. Microbiological Sensors for the Drinking Water Industry DOI: 10.2166/9781780408699_0001

Section 1

Background and perspectives

Chapter 1

History and perspective – the challenge of the gap between legislation and scientific achievements: The new era of enlightenment

Hans-Jørgen Albrechtsen

1.1 INTRODUCTION

Controlling microbial water quality is crucial to ensure safe drinking water – and this subsequently calls for proper monitoring and documentation of the occurrence of microorganisms in drinking water.

Since the end of the 1880s, monitoring has been achieved by various culture-based methods, in which the microorganisms must proliferate to be detected. Today these principles still form the basis for almost all legally binding guidelines and standards relating to microbial drinking water quality (e.g. EU, 1998; US EPA, 2017; US SDWA, 1974; WHO, 2017; EN ISO 6222:2000; EN ISO 9308-1:2014; EN ISO 9308-2:2012).

Recently, a range of different sensors has emerged with abilities to measure microbial quality aspects of drinking water. Furthermore, we are entering the age of “Industry 4.0” (Lu, 2017; Sedlak, 2015), and automation and digitalization is playing an increasingly important role. Also for this reason, the use of sensors becomes essential. However, one major challenge for the implementation of sensors for microbial water quality is that they often are based on measuring principles that are quite different from the classical culture-based methods.

Therefore, one of the major challenges in the implementation of sensors in modern management of microbial drinking water quality is to overcome the clash between signals from the sensors and the standards based on classical measuring methods.

1.2 THE HISTORICAL DEVELOPMENT OF CULTURE-BASED METHODS

Already in 1881, Robert Koch had demonstrated the use of culturing bacteria on solid media, i.e. the application of agar or gelatin to solidify the bacterial culture medium containing growth substrates such as ammonium salts, yeast residues and a variety of sugars. Within a few years Percy Frankland had – in 1885 – applied this method to measure the number of bacteria in water, e.g. for evaluation of the efficiency of filtration. However, this raised a discussion on which medium is best? How many bacteria are acceptable? How good is this measure for understanding the occurrence of pathogens? (Payment et al., 2003).

A few years later, in the 1890s, Theobald Smith proposed Escherichia coli as a drinking water indicator since it is part of the normal bacterial flora of mammals and warm-blooded animals (Prescott & Winslow, 1915). In the absence of a specific test for E. coli, surrogate tests for E. coli were developed, e.g. the ‘fecal coliform test’ proposed by Eijkman in 1904, based on the principle that E. coli is more thermotolerant than other members in its family. This method was further modified to detect the ‘total coliform’, which in addition to E. coli included a range of bacteria for which growth is not restricted to the mammalian colon. In 1914 the US Treasury Department proposed this test to be a standard for drinking water safety (Edberg et al., 2000) and these principles have been the foundation for monitoring microbial drinking water quality since.

Several methods have been implemented to detect and quantify the number of E. coli. A simple version of the Most Probable Number or MPN procedure was implemented in 1914 (Ashbolt et al., 2001). This is a statistical-based assay where a sample is diluted until extinction, and the serial dilutions are inoculated into broth media. The number of bacteria in the sample is estimated based on the distribution of the positive and negative responses (expressed as growth or biochemical reactions such as gas production during fermentation) among the dilutions.

E. coli can also be detected as colony formation in solidified selective media either on the top of agar (spread plate) or by mixing the water sample with melted agar (pour plate method). During the 1950s membrane filtration became an alternative to the MPN approach (Ashbolt et al., 2001). This method is based on filtering the water sample through a membrane which retains the bacteria; after filtering the filter is incubated on top on a solid agar plate where the bacteria on the filter form colonies. The advantage with this approach was the increased sensitivity of the method, since a larger volume (100 mL or more) can be concentrated on the filter. This is still an accepted method for enumeration of E. coli (EN ISO 9308-1:2014).

In the early 1990s the MPN methods were further developed into a miniaturized setup, where detection is based on an enzymatic reaction in the so-called Colilert® technique (IDEXX), which is the state-of-the-art method (EN ISO 9308-2:2012).

However, all these methods require growth of the investigated cells, either to provide enough cells to make a visible colony or to produce enough enzyme to catalyze transformation of the substrate to produce a detectable signal. Furthermore, these methods require incubation time and subsequent relatively long response times, to ensure that the numbers are sufficiently low.

1.3 ADVANTAGES AND LIMITATIONS OF THE CLASSICAL CULTURE-BASED METHODS

All culture-based methods have this strong limitation – the incubation time – since it takes time for the microorganisms to grow to such a quantity that they can be detected, whether that is as a colony or a biochemical reaction. Although the methods have improved substantially, as exemplified by the state-of-the-art and standard method for E. coli and coliforms (Colilert® from Idexx), they still need 18 hours of incubation to provide certainty whether a given water sample is safe – although a contaminated sample may give a positive signal earlier. When monitoring drinking water distribution systems, the water might have been used or consumed before the result finally is reported.

On the other hand culture-based methods all have one clear advantage – the cells which are detected are obviously viable, otherwise no colony would have been formed. The use of E. coli provides an extremely sensitive analytical approach. Since the concentration of E. coli in wastewater is on the order of 106–107/100 mL (Lucas et al., 2013; Andersen, 2015), and the level of detection for the standard method for use in drinking water is <1/100 mL, this method is able to detect traces of wastewater even after 1–10 million times of dilution.

However, the reliability of the viability on the other hand is also a challenge for this measuring principle, since e.g. E. coli is more sensitive to UV radiation and disinfection chemicals such as chlorine than many other pathogens (Chang et al., 1985). In this way the absence of the indicator microorganism may be a false negative result, because the given water sample may still be infectious.

Despite these aspects, E. coli is still seen by many as the best biological drinking water indicator for public health protection (Edberg et al., 2000).

1.4 MEASURING GENERAL MICROBIAL POPULATIONS OR MICROBIAL INDICATORS

The methods described above can be used in two quite different principles. The first principle – to count the number of unspecific bacteria in the water – is referred to as heterotrophic plate counts (HPC). This can be done in many different ways, but a common method for drinking water is described by EN ISO 6222:2000.

The second principle is to use indicator microorganisms (E. coli, Clostridium perfringens, intestinal enterococci, etc.), which are indicators of the risk of infection, i.e. the presence of pathogens. Since the majority of waterborne diseases are related to fecal contamination of the water, the philosophy is based on the selection of an indicator microorganism which always is present in the digestive system of mammals. So, if this microorganism is detected in a given ...

Table of contents

• Cover
• Half Title
• Title Page
• Copyright Page
• Contents
• Editorial preface
• Foreword by Eberhard Morgenroth
• Foreword by Koen Huysman
• About the Editors
• List of Contributors
• Section 1: Background and perspectives
• Section 2: Industry needs
• Section 3: Sensor technologies
• Section 4: Data collection and interpretation
• Section 5: Water safety, education and legislation
• Index