Overview of the Drinking-Water Standards · water (except bottled water which must comply with the Food Act 1981). Drinking-Water Standards for New Zealand 2004 (DWSNZ 2004) list

Overview of the Drinking-Water

Standards

Published in October 2003 by the Ministry of Health

PO Box 5013, Wellington, New Zealand

ISBN 0-478-25860-7 (Internet)

This document is available on the Ministry of Health’s website:

http://www.moh.govt.nz

http://www.moh.govt.nz

Overview of the Drinking-Water Standards iii

Contents

List of Tables iv

1 Overview of the Drinking-Water Standards 1 1.1 Introduction 1 1.2 Scope of the Standards 3 1.3 Development of the Standards 4 1.4 Role of the Standards 6 1.5 Content of the Standards 7 1.6 Maximum acceptable values (MAV) 9 1.7 Components of a drinking-water supply 11

2 Determination of Compliance 14 2.1 Introduction 14 2.2 Compliance and transgression 15 2.3 Compliance with the Standards / MAVs 15

3 Microbiological Compliance 22 3.1 Rationale for microbiologocal MAVs 22 3.2 Microbiological compliance criteria 24 3.3 Microbiological monitoring requirements 58 3.4 Remedial action to be taken when transgression of a microbiological MAV occurs 83

iv Overview of the Drinking-Water Standards

List of Tables Table 1.1: Summary of the section numbers of the criteria, requirements for

sampling and testing, and remedial actions (some in process of changing) 13

Table 3.1: Log credits and conditions for control of protozoa 35 Table 13.1: C.t values for Cryptosporidium inactivation by chlorine dioxide 47 Table 13.1: C.t values for Cryptosporidium inactivation by ozone 48 Table 13.2: UV dose requirementsa for Cryptosporidium and virus inactivation

credits 50 Table 3.1: Minimum sampling frequency for E. coli in drinking-water leaving a

treatment plant for E. coli compliance criterion 1 63 Table 3.2a: Minimum sampling frequency for E. coli in a distribution system 68 Table 3.2b: E. coli sample distribution 70 Table 3.2c: Minimum sampling frequency for E. coli in a bulk distribution zone 71 Table 3.4: Minimum measurement frequency and reporting period for turbidity in

water leaving each filter, for protozoa compliance for all treatment plants using any form of filtration 71

Table 3.3: Minimum measurement frequency for protozoa compliance for bag and cartridge filters 72

Table 14.8: References to turbidity requirements in these Standards Note: the protozoa bits need redoing 88

Overview of the Drinking-Water Standards 1

1 Overview of the Drinking-Water Standards

Suggested changes

1.1 Introduction

Safe drinking-water, available to everyone, is a fundamental requirement for public health. The Drinking-Water Standards for New Zealand 2004 replace the Drinking-Water Standards for New Zealand 2000 with effect from 1 January 2005. They detail how to assess the quality and safety of drinking-water. The Standards define drinking-water: that is, water intended to be used for human consumption, food preparation, utensil washing, oral hygiene or personal hygiene. The Standards provide criteria applicable to all drinking-water (except bottled water which must comply with the Food Act 1981). Drinking-Water Standards for New Zealand 2004 (DWSNZ 2004) list the maximum concentrations of chemical, radiological and microbiological contaminants acceptable for public health in drinking-water. For community drinking-water supplies, the Standards also specify the sampling protocols that must be observed to demonstrate that the drinking-water complies with the Standards. Community drinking-water supplies are water supplies that serve more than 25 people for at least 60 days per year. All community drinking-water supplies known to the Ministry of Health are listed in the Register of Community Drinking-Water Supplies in New Zealand. Because of the wide variety of circumstances relating to individual household drinking-water supplies no general sampling recommendations are made for such supplies. If there is any concern about the quality of a household's drinking-water, advice on appropriate sampling programmes can be obtained either from the Environmental Health Officers of the local territorial authority or the Health Protection Officers at the public health service provider. Toxic chemical contaminants in drinking-water rarely lead to acute health problems except through massive accidental contamination of a supply. Before it presents a health risk the water usually becomes undrinkable due to unacceptable taste, odour or appearance.

Possibly 1 Jan 2005 These definitions may need to be amended as the changes in the Building Act and the Health Act regarding self-supplied community purpose buildings come into effect.


The problems associated with chemical contaminants of drinking-water arise primarily from their ability to cause adverse health effects after prolonged periods of exposure. Of particular concern are contaminants which have cumulative toxic properties, such as some heavy metals and substances which are carcinogenic. Because chemical contaminants of drinking-water do not usually give rise to acute effects they are placed in a lower priority category than microbiological contaminants, the effects of which are potentially acute and widespread. The control of risks arising from microbiological contamination is, therefore, given priority over the control of risks from chemical contaminants. The Drinking-Water Standards for New Zealand 2004 are intended to: • set out the requirements for compliance with the

Standards • facilitate consistency of application throughout

New Zealand • protect public health while minimising unnecessary

monitoring • be appropriate for large and small publicly and

privately owned drinking-water supplies. DWSNZ 2004 revise a number of small errors in the 2000 edition, update the analytical methods, and make a number of minor changes to improve the interpretation and robustness of the Standards. In addition, DWSNZ 2004 include the following significant changes. • The use of UV disinfection for inactivation of

bacteria, viruses and protozoa. • The sections relating to protozoa criteria have been

completely restructured. • The use of bags as a sole treatment process for

removal of protozoa is no longer considered to be adequate.

• Issues relating to tankered water supply systems have been incorporated.

4 new bullets – old ones gone Tankered supplies to come


1.2 Scope of the Standards DWSNZ 2004 are applicable to water intended for drinking, irrespective of its source, treatment, distribution system, whether it is from a public or private supply or where it is used. The exception is bottled water, which is subject to different standards set under the Food Regulations. MAF Standard D106.1 (1999) ‘Farm Dairy Water’ also covers water quality. The Standards specify maximum acceptable values (MAVs) for the microbiological, chemical and radiological determinands of public health significance in drinking-water and provide compliance criteria and procedures for verifying that the water supply is not exceeding these values. The companion publication, Guidelines for Drinking-Water Quality Management in New Zealand, provides additional information about determinands listed in the Standards, the management of drinking-water quality, the derivation of the concepts used in the Standards and references to the publications on which the Standards are based. Aesthetic considerations are not covered by the Standards. Guideline values for determinand concentrations that should avoid public complaints are given in Table 14.6 and are discussed in the Guidelines. These Standards are for the general protection of public health. For people with special medical conditions, or for uses of the water for purposes other than drinking, additional or other water quality criteria may apply such as the special requirements of the Animal Products Act, the Food Act, the Dairy Act, and the Meat Act. Water quality standards also appear in MAF Standard D106.1 (1999) ‘Farm Dairy Water’.

Added MAF reference


1.3 Development of the Standards The Standards were developed by the Ministry of Health with the assistance of an Expert Working Group. Extensive use was made of the World Health Organization's Guidelines for Drinking-Water Quality and addenda up to 1998. Reference was also made to the Drinking-Water Standards for New Zealand 1984, 1995 and 2000, and to the Australian Drinking Water Guidelines 1996. The Standards are based on the following principles:

1. The Standards define concentrations of chemicals of health significance which, based on current knowledge, constitute no significant risk to health to a person who consumes 2 litres of the water a day over their lifetime (taken as 70 years). It is usually not possible to define a concentration of contaminant (other than zero) at which there is zero risk because there is always some degree of uncertainty over the magnitude of the risk. Refer to the data sheets in the Guidelines for details of each determinand.

2. The Standards give top priority to health risks

arising from microbiological contaminants. Control of microbial contamination is of paramount importance and must not be compromised in an attempt to correct chemical problems, such as disinfection by-product formation.

3. The Standards set priorities to ensure that, while

public health is protected, scarce resources are not diverted to monitoring substances of relatively minor importance.

4. The Standards are set to protect public health

and apply to health significant determinands only. However, as the public generally assesses the quality of its water supply on aesthetic perceptions, guideline values for aesthetic determinands are also provided. Refer to the Guidelines for more details.

1998 still the latest? There is a 2003 edition – but not published yet – Task Force notes available Later version of Australian Guidelines promised in 2003 – still coming Added the word ‘chemicals’ in line 2, bullet 1 Added reference to data sheets


Where feasible, the sampling protocols are designed to give 95 percent confidence that the supply has complied with the Standards for at least 95 percent of the time. A minimum of 76 samples, none of which transgresses the MAV, is required before the Ministry can be 95 percent confident that the supply complies with the Standards for 95 percent of the time. To minimise costs to smaller supplies they are given the benefit of the doubt and it is assumed that if 38 successive samples are taken with no transgressions, then they are complying. However, if one transgression occurs the doubt no longer exists and to demonstrate compliance they must take as many samples as are necessary to comply with the 'allowable transgression' tables in Section 7.5.2 of the Guidelines. However, for those determinands monitored monthly it will take several years of results before this degree of confidence can be attained.

Rewritten

Use of the DWSNZ under the HDWAB & BA New section on interpretation needed.


1.4 Role of the Standards The Drinking-Water Standards for New Zealand 2004 contribute to the safety and quality of drinking-water by: • defining safety standards for drinking-water • detailing how compliance with these Standards is

to be demonstrated • facilitating the development of a consistent

approach to the evaluation of the quality of the country's drinking-water supplies.

Four barriers to disease are available in the provision of safe drinking-water.

1. Protection of the quality of the raw water source.

2. Removal of chemical and microbiological

determinands by physical means.

3. Inactivation of pathogenic micro-organisms by disinfection processes.

4. Prevention of contamination of treated water

whilst it is in the network reticulation. The Standards provide performance criteria for the second, third and fourth barriers to infection. The first barrier is discussed in the Guidelines.


1.5 Content of the Standards The DWSNZ 2004 set standards for drinking-water constituents or properties (determinands) and contain the information necessary to demonstrate whether a water supply complies with these Standards. Three types of compliance are included in these Standards: microbiological, chemical and radiological. The Standards define the Maximum Acceptable Value (MAV) for each determinand. For chemical deteminands this is usually the concentration at which the risk resulting from consumption of the contaminant over a lifetime is considered to be insignificant in the light of present knowledge. The Maximum Acceptable Values (MAVs) are discussed in Chapters 2, 3, 4 and 5. The determinands have been classified into four priority classes. These are discussed in Section 2.3.1. The monitoring and analytical requirements needed to demonstrate compliance for those determinands in Priorities 1 and 2 are given in Chapters 3, 4 and 5 for microbiological, chemical and radiological determinands respectively. MAVs for each of the individual health significant determinands are listed in Chapter 14. A summary of the location of the criteria, the requirements for sampling and testing, and remedial action is presented in Table 1.1. This ‘look-up’ table has been added to assist readers to ‘find their way around’. The Guidelines for Drinking-Water Quality Management for New Zealand (Guidelines) provide background and supporting information for the Standards and contain: • data sheets with background information about

each determinand including sources, environmental forms and fates, typical concentrations either in New Zealand or overseas drinking-water supplies, processes for removing the determinand from drinking-water, analytical methods, health considerations, derivation of the MAV and the guideline values for determinands of aesthetic interest

• chapters on microbiological, chemical and

radiological determinands providing background


information about each group of determinands • background information about chlorine and

alternative disinfection systems and their effect on drinking-water quality

• guidelines and risk management principles for community drinking-water supplies.


1.6 Maximum acceptable values (MAV) The Maximum Acceptable Value (MAV) of a determinand in a drinking-water represents the concentration of a determinand which, on the basis of present knowledge, is not considered to cause any significant risk to the health of the consumer over a lifetime of consumption of the water. Nearly all of the MAVs for the determinands covered in these Standards are based on the World Health Organization (WHO) publication Guidelines for Drinking-Water Quality 1998. The method of derivation depends upon the particular way in which the determinand presents a health risk. For some chemical determinands, adaptation of the method of derivation to suit New Zealand conditions has resulted in a minor difference between the guideline value recommended by WHO and the MAVs in these Standards. In addition, some chemical determinands not covered by the World Health Organization publication Guidelines for Drinking-Water Quality (editions and supplements up to 1998) have been added to these Standards because of their public health significance in New Zealand circumstances. MAVs of these determinands have been calculated using methods appropriate to the situation. In all cases the approach was conservative and considerable safety factors have been used. A general discussion on the methodology of the derivation of the MAVs is given in the Guidelines, together with specific information about the derivation of the MAV for each individual determinand. Note that:

1. the MAVs set in the Standards define water suitable for human consumption and hygiene. Water of higher quality may be required for special purposes, such as renal dialysis or certain industrial processes. The Standards do not address these issues.

2. short-term excursions above a chemical MAV do

not necessarily mean the water is unsuitable for consumption. Most MAVs have been derived on the basis of a lifetime exposure. The amount and the duration by which any MAV can be exceeded without affecting public health depends

1998? See above Ditto


on the characteristics of the determinand.

3. the chemical MAV values are set to be acceptable for lifelong consumption. The quality of drinking-water should not, however, be degraded to the MAV level. Ongoing effort should be made to maintain drinking-water quality at the highest possible level. Maximum Desirable Target Values (MDTVs) are given in the Guidelines to assist in treatment design.

For radioactive substances, screening values for total alpha and total beta activity are given, based on a reference level of dose.


1.7 Components of a drinking-water supply A community water supply comprises one or more of each of the following (see Figure 1.1): • the source or raw water • the treatment plant • the distribution system. Insert Figure 1.1: Schematic diagram of a drinking-water supply system

1.7.1 Source water A community water supply may abstract raw water from rainwater, surface water or groundwater sources. Surface water is frequently contaminated by micro-organisms. Shallow groundwater and some springs are microbiologically equivalent to surface water, along with rivers, streams, lakes and reservoirs. Secure groundwater, as defined in Chapter 7 and in Section 3.2.4, is usually free from microbiological contamination. A water supply may have more than one source of raw water. Secondary sources may be permanent or temporary.

1.7.2 The treatment plant A treatment plant is a facility that treats raw water to make it safe and palatable for drinking. For administrative purposes, the treatment plant is considered to be that part of the system where raw water becomes the drinking-water. This can range from a full-scale water treatment plant comprising chemical coagulation, sedimentation, filtration, pH adjustment, disinfection and fluoridation, to simply being the point in a pipeline where the water main changes from a raw water main to a drinking-water supply main. In a simple water supply, the water may be merely abstracted from a river, passed through a coarse screen and piped to town, that is, the water supply acts like a diverted stream. If raw water is chlorinated, however, the water will not be considered to become drinking-water until it has been exposed to chlorine for the design contact time. A treatment plant may receive raw water from more than one source.


1.7.3 The distribution system Once the water leaves the water treatment plant, it enters the distribution system (sometimes called the network reticulation) that consists of one or more distribution zones that serve the community. A distribution zone is defined as (Chapter 7): ‘...part of the water supply network within which all consumers receive drinking-water of identical quality, from the same or similar sources, with the same treatment and usually at the same pressure. It is part of the supply network that is clearly separated from other parts of the network, generally by location, but in some cases by the layout of the pipe network. For example, in a large city, the central city area may form one zone, with outlying suburbs forming separate zones, or in a small town, the system may be divided into two distinct areas. The main purpose of assigning zones is to separately grade parts of the system with distinctly different characteristics.’ A distribution zone may receive water from more than one treatment plant. The distribution system may comprise more than one distribution zone (see Figure 1.1). Distribution zones may be distinguished because they are either fed by a pumping station so that they are isolated from nearby zones by pressure, or because they are fed from a service reservoir which can markedly increase the retention time. Some distribution zones may vary seasonally due to supplementary sources being used at peak draw-off times while for other zones the boundaries may vary due to changes in pressure or draw-off. Others may vary due to the materials used in common sections of the distribution system. The distribution zones selected for public health grading of drinking-water supplies and for the Standards are based on water quality considerations and will not necessarily coincide with the distribution zones which the water suppliers identify for operational and management purposes. The Ministry of Health expects there would be more distribution zones based on hydraulics than there will be on water quality. Some community drinking-water supplies may comprise one distribution zone only. Some very small community water supplies may not have a network of water mains. For example, drinking-water supplies at factories, rural


schools and camping grounds may only have a communal tap. Some small drinking-water supplies may receive their water from another supply by tanker that pumps the water into a storage tank. Some water suppliers may receive their drinking-water from a water supply wholesaler via bulk mains.

Table 1.1: Summary of the section numbers of the criteria, requirements for sampling and testing, and remedial actions (some in process of changing)

Section numbers for:

Criterion Sampling Sites

Monitoring Frequency

Sampling Requirements

Test Requirements

Remedial Action

E. coli criterion 1 water leaving the treatment plant

3.2.2.1 3.3.1.1.1 3.3.2.1.1 - 4 3.3.3 3.3.4 3.4.1.1

E. coli criteria 2A and 2B water in the distribution system

3.2.2.2 3.3.1.1.2 3.3.2.1.5 - 6 3.3.3 3.3.4 3.4.1.2

E. coli criteria 3A amd 3B water in the bulk distribution zone

3.2.2.3 3.3.1.1.3 3.3.2.1.7 - 8 3.3.3 3.3.4 3.4.1.2

Protozoa criterion a water leaving each treatment plant filter

3.2.3 3.3.1.2.1 3.3.2.2.1 3.3.2.2.1 3.3.2.2.1 3.4.2.1

Protozoa criterion b water leaving each treatment plant filter

3.2.3 3.3.1.2.2 3.3.2.2.2 3.3.2.2.2 3.3.4.5 3.4.2.2

Protozoa criterion c water leaving the treatment plant

3.2.3 3.3.1.2.3 3.3.2.2.3 3.3.2.2.3 3.3.2.2.3 3.4.2.3

Protozoa criterion d secure water leaving the treatment plant

3.2.3 3.3.1.2.4 3.3.2.2.4 3.3.2.2.4 3.4.2.4

Secure groundwater criteria

3.2.4 Table 3.1 3.4.1.1

Priority 2 chemical criteria

4.2.1 4.3.1.1 - 2 4.3.2.1 - 2 4.3.3 4.3.4 4.4.1, Fig 4.1

Aggressive 4.2.3 4.3.1.2 4.3.2.3 4.3.3 4.3.4 4.2.3,


water criteria 4.4.2

2

2.1

Determination of Compliance

Introduction To comply with the Standards a determinand must be investigated according to the monitoring and analytical protocols given in Chapters 3, 4 and 5 for microbiological, chemical and radiological determinands respectively. Laboratories recognised for the purpose by the Ministry of Health shall be used for all analyses carried out to assess compliance with these Standards, except where special procedures are authorised for small remote drinking-water supplies or for analyses in the field or around the works. Analysis for the purpose of demonstrating compliance with the DWSNZ must be carried out in a laboratory that is recognised for the purpose by the Ministry of Health except where special procedures are authorised. These laboratories will be expected: • to hold laboratory accreditation to NZS/ISO/IEC

Guide 17025: 2000, or equivalent; • to use analytical methods that have been calibrated

against the referee methods; • to have quality assurance and control systems that

provide evidence of competency in testing; • to ensure that all samples used for compliance

testing are identified by the unique site identification code listed in the Register of Community Drinking-Water Supplies in New Zealand for the supply concerned. The site codes shall be provided by the water supplier generating the samples and shall accompany all samples sent for compliance testing.

In circumstances where accreditation is not feasible, alternative evidence of competence may be accepted by the Ministry of Health. This will require compliance with the relevant clauses of the NZS/ISO/IEC Guide17025: 2000 to be demonstrated. The referee methods specified in Chapter 11.1 shall be regarded as the definitive methods for demonstrating compliance with these Standards. Alternative methods are acceptable but must have been calibrated against the referee methods. In the event of any dispute about differences in analytical results, results

Now bulleted Changed to NZS/ISO/IEC and deleted reference to the Aust Std (which is now 1999 anyway) in first bullet


obtained using the referee method shall be deemed to be correct. Compliance is determined by comparing the results of these monitoring programmes against the Standards' compliance criteria over 12 consecutive months. Records must be kept for at least ten years to enable trends to be detected and to establish the statistical significance of the results. MAVs are specified for determinands of all Priority classes in Chapter 14. The tables in Chapter 12 will assist in selecting the appropriate sampling and analytical methods.

2.2 Compliance and transgression The Drinking-Water Standards for New Zealand 2004 specify maximum acceptable values (MAVs) for the microbiological, chemical and radiological determinands of public health significance in drinking-water and provide compliance criteria and procedures for verifying that the water supply is not exceeding these values. The terms compliance and non-compliance apply to the supply. They are not applied to individual samples. Compliance is assessed on a running annual basis. In this way compliance can be assessed at any time during the reporting period using the previous period’s monitoring results. Unless otherwise stated, the reporting period is 12 months. The term transgression applies to a single sample. If every determinand in a sample is below its MAV, the sample meets the requirements of the Standards. A sample is said to transgress the Standards when it does not meet the requirements of the Standards, i.e. the MAV for one or more determinands is exceeded. Transgression of the Standards by a sample may not necessarily mean that the drinking-water supply itself is in non-compliance. This depends upon the verification requirements specified by these Standards for the determinand concerned. In the event of a sample transgressing any criteria, immediate action must be taken as set out in Sections 3.4 and 4.4.

2.3 Compliance with the Standards / MAVs

2.3.1 Priority classes for drinking-water determinands The determinands of public health significance have been divided into four priority classes to minimise


monitoring costs without compromising public health. To demonstrate compliance, only those relatively few determinands that fall into the classes with highest potential risk, Priorities 1 and 2, are required to be monitored. Monitoring of determinands in the lower potential risk, Priorities 3 and 4, is at the discretion of the supplier, unless required by the Medical Officer of Health for public health reasons. 2.3.1.1 Priority 1 determinands Priority 1 determinands are determinands whose presence can lead to rapid and major outbreaks of illness. Contamination of water supplies by pathogens usually arises from faecal material or wastes containing them. Humans, birds or animals may be the source. The determinands that currently are known to fall into this category include the pathogenic bacteria, viruses and protozoa. Escherichia coli (E. coli), a common gut bacterium living in warm blooded animals, is used as an indicator of the contamination of water by excrement and is a generally accepted indicator for the potential presence of pathogenic viruses and bacteria. However, E. coli is not a good indicator of the presence of the pathogenic protozoa Giardia and Cryptosporidium. For this reason the current Priority 1 determinands are: • Escherichia coli (E. coli) • Protozoa (Giardia and Cryptosporidium). Priority 1 determinands apply to all community drinking-water supplies in New Zealand and must be monitored in all supplies because they constitute a major public health risk. To comply with the Standards, Priority 1 determinands must be investigated according to the monitoring and analytical requirements given in Chapters 3,4 and 5 for microbiological, chemical and radiological determinands as relevant. Compliance is assessed and reported for each calendar year by comparing the results of these monitoring programmes over 12 consecutive months against the compliance criteria set out in Section 3.2. Note that there are some situations where a different reporting period has been specified


specified. Records must be kept for at least ten years. Giardia and Cryptosporidium are widespread in natural waters in New Zealand, and are not always removed reliably by conventional water treatment. Although there may be no correlation between the presence of E. coli and of pathogenic protozoa in raw water, increases in the turbidity of water which has been treated by flocculation and filtration have been linked with elevated protozoa counts. In view of the serious public health effects of contamination of a drinking-water supply by these protozoa, it is important that the likelihood of their presence in drinking-water is assessed. The most reliable methods currently available for direct determination of these organisms are still expensive and require highly skilled analysts, and the test requires at least one day. Also the organisms tend to appear sporadically, so that direct measurement techniques do not always give a representative assessment of the true extent of their presence in a drinking-water supply. If the water is subject to quiescent periods, the organisms may settle out so they are not present in the supernatant water. Disturbance of the sediments can resuspend the organisms, causing a sudden upsurge in their numbers. Because of these difficulties, direct determination of the presence of Giardia and Cryptosporidium is not used as a criterion of compliance with the Standards at present. Alternative ways of assessing the likelihood of the absence of these protozoa are therefore used. These are based on checking that the drinking-water has received a level of treatment which has a high probability of having removed the organisms. In these Standards, the criteria used are based on: • the effectiveness of particle removal to assess

treatment by filtration without coagulation • the use of turbidity to assess the effectiveness of

conventional coagulation/filtration treatment • C.t values by measurement of the chemical

di i f ’ id l h d f


disinfectant’s residual to assess the adequacy of disinfection, or specifying dosage and monitoring conditions for effective UV disinfection

• demonstration that the water has come from a secure groundwater source that will be free from these organisms.

The specific compliance criteria for each of these situations are given in Section 3.2.

2.3.1.2 Priority 2 determinands Priority 2 determinands are those that are present in a specific supply or the distribution zone, at concentrations that exceed 50 percent of the MAV. The Ministry of Health will carry out investigations on water supplies from time to time to identify the presence of P2 determinands, until this process is adequately covered by water supply risk assessment procedures carried out by the drinking-water suppliers. Determinands specified by the Ministry of Health to be Priority 2 determinands for the drinking-water supply under consideration are required to be monitored in order to establish compliance with the Standards. Priority 2 determinands are divided into three types: 2a, 2b and 2c. 2a Chemical and radiological determinands that could be introduced into the drinking-water supply by the treatment chemicals at levels potentially significant to public health (usually greater than 50 percent of the MAV). Priority 2a does not include disinfection by-products or determinands introduced into the drinking-water from piping or other materials. 2b Chemical and radiological determinands of health significance that have been demonstrated to be in the drinking-water supply at levels potentially significant to public health, (usually greater than 50 percent of the MAV). Priority 2b includes chemicals present in the raw water that may not be removed by the treatment process; any disinfection by-products; and determinands introduced into the drinking-water from piping or other materials that are

Now bulleted


present in the water when sampled under normal (flushed) protocols. Priority 2b does not include determinands introduced by the treatment chemicals or determinands introduced by the consumer’s plumbing. A separate category of ‘aggressive’ drinking-water is distinguished in which heavy metals are only found in the first flush of water collected from the tap but are not present at excessive levels in samples collected after flushing. These determinands are produced by corrosion of the consumer’s plumbing when water stands in contact with taps or other fittings, so that one or more of lead, antimony, cadmium, copper, nickel or zinc dissolve from the fitting. The presence of Priority 2a determinands will depend on the chemicals (and their impurities) used to treat the raw water or added to the water supply and, to some extent, the degree of management control over their use. The likelihood that a borderline determinand will be assigned to Priority 2a rather than Priority 3 will be much greater if the treatment process is operated in such a way that the concentration of the determinand varies greatly from time to time than if it is maintained at a relatively constant concentration. Some chemicals of health significance, for example copper sulphate for algal control, may be used only intermittently in the course of drinking-water treatment. In these situations the water supplier must advise the Medical Officer of Health and consider an appropriate monitoring programme. The Medical Officer of Health must also be advised of any long-term changes to the chemical treatment process so the Ministry’s drinking-water information system (WINZ), and the Register, can be revised (refer to the Guidelines). The frequency of monitoring of some Priority 2a determinands that can enter the drinking-water supply as impurities of water treatment chemicals may be diminished if water suppliers demonstrate to the Medical Officer of Health's satisfaction (for example from flow rates, dosing equipment and the use of treatment chemicals with verified specifications) that the determinand cannot be introduced into the drinking-water supply at concentrations greater than 50 percent of the MAV. 2c Micro-organisms of health significance that have been demonstrated to be present in the drinking-water supply. Micro-organisms listed in Table 14.1 may be listed as


Priority 2c determinands if there is reason to suspect that they are likely to be present in the drinking-water supply. This may occur, for example, when high numbers of these organisms are present in the raw water and E. coli is present in water leaving the treatment plant. The Medical Officer of Health may declare such organisms as Priority 2 if there are epidemiological grounds for suspecting the drinking-water supply. The designation of a Priority 2 determinand to a given supply will be based on monitoring and on knowledge of sources of health-significant determinands in the catchment, treatment processes and distribution system. The designation will be notified directly to the water supplier, after prior consultation, to enable review of any contrary evidence. Priority 2 determinands also will be listed in the Register of Community Drinking-Water Supplies in New Zealand published by the Ministry of Health. The requirement to monitor a Priority 2 determinand commences with the date of formal notification to the supplier of the designation of the determinand to Priority 2 by the Ministry of Health, not with the date of publication in the Register. A Priority 2 determinand may be relegated to Priority 3 or 4 with the consent of the Ministry of Health when monitoring has demonstrated that is should be assigned a lower priority. Refer to Section 4.2.3. Information about the compliance criteria and the sampling and analytical requirements for microbiological, chemical and radiological determinands are provided in Chapters 3, 4 and 5. Information on cyanobacteria and cyanotoxin monitoring protocols is included in the Guidelines for Drinking-Water Quality Management for New Zealand. 2.3.1.3 Priority 3 determinands 3a-3d

3a Chemical and radiological determinands of health significance arising from treatment processes in amounts known not to exceed 50 percent of the MAV. 3b Chemical and radiological determinands of health significance that are not known to occur in the drinking-water supply at greater than 50 percent of the MAV. The chemicals listed in Tables 14.2 to 14.5 are Priority 3a or


3b determinands unless they have been assigned to Priority 2 for a particular supply. 3c Micro-organisms of health significance that could be present in the drinking-water supply. Except for E. coli and the protozoa, the micro-organisms listed in Table 14.1 are Priority 3c determinands unless they have been assigned to Priority 2 for a particular supply. 3d Determinands of aesthetic significance known to occur in the drinking-water supply. Aesthetic determinands are classified as Priority 3 because they do not pose a direct threat to public health. People however, judge drinking-water mainly by the aesthetic characteristics of appearance, taste and smell, and an aesthetically unacceptable drinking-water supply may cause them to change to an alternative and potentially unsafe supply or treatment process. For this reason it is preferable that water supply authorities monitor these determinands, although this is not required in order to comply with the Standards.

2.3.1.4 Priority 4 determinands 4a-4c

4a Chemical and radiological determinands of health significance that are known not to be likely to occur in the drinking-water supply. 4b Micro-organisms of health significance that are known not to be likely to be present in the drinking-water supply. 4c Determinands of aesthetic significance not known to occur in the drinking-water supply. Priority 4 determinands for a specific supply will include those health significant or aesthetic determinands for which there is sufficient information to consider it unlikely they would be present in a particular supply. Some determinands, including some pesticides, will be Priority 4 for all New Zealand drinking-water because they are not used in this country at present. They are included in the tables to ensure that MAVs are available should the situation change. Priority 4 determinands may become Priority 2 if the


Ministry of Health considers this is warranted.

3 Microbiological Compliance

Suggested changes

3.1

3.1.1

Rationale for microbiological MAVs It is impracticable to monitor water supplies for all potential human pathogens. Surrogates have to be used to indicate possible contamination of the water supply with human and animal waste, the most frequent source of health-significant contamination of water supplies.

E. coli

The indicator organism chosen to indicate possible faecal contamination of drinking-water is E. coli. Thermotolerant coliforms (faecal coliforms) and total coliforms (which include both faecal and environmental coliform bacteria) may also be used to monitor water quality, but the results are harder to interpret than those from E. coli. If total coliforms or faecal coliforms are used for drinking-water monitoring to demonstrate compliance with the Standards instead of E. coli, a positive result shall be treated as though it were an E. coli result. E. coli should not be present in drinking-water in the distribution zones. However, unlike the drinking-water leaving the treatment plant, where microbiological quality is under the control of the treatment plant management, the quality of drinking-water in the distribution zones may be subjected to contamination from a variety of influences. Some of these may arise from poor design, or management practices such as faulty reservoir or mains construction and maintenance, or poor sanitary practices by water supply workers. Other contamination sources arise from the water users themselves, such as poor sanitation while making connections to the service, or inadequate backflow prevention. Therefore E. coli occasionally may be found in the reticulation. Their occurrence must always be followed up.


If more than 0.2 mg/L free available chlorine (FAC) is maintained in the drinking-water supply reticulation, coliform bacteria and E. coli are rarely found. For this reason it is permissible to substitute monitoring of free available chlorine for some (but not all) of the faecal coliform monitoring. A concentration of 0.2 mg/L chlorine dioxide expressed as ClO2 has a similar disinfecting power as 0.2 mg/L FAC.

3.1.2

Protozoa: Giardia and Cryptosporidium

The protozoa Giardia and Cryptosporidium occur in many New Zealand water sources. They are found in wild, farm and domestic animals. Surface waters, and non-secure groundwater, must be considered to be potentially contaminated. The risk associated with secure groundwater is much lower. Giardia and Cryptosporidium are pathogens which should be eliminated from drinking-water supplies. They are Priority 1 determinands because of their public health significance. The methods available for enumerating pathogenic protozoa and determining their viability are becoming less expensive and more reliable but they are not yet suitable for routine use. Until more rigorous procedures are available, criteria based on the probability that the treatment process used will have inactivated or removed any protozoa present will be used as criteria for compliance. There were four treatment categories in the 1995 and 2000 DWSNZ: (a) Filtration without coagulation; (b) Chemical coagulation plus filtration; (c) Disinfection without filtration; and (d) Secure groundwater. It is proposed to change the approach for DWSNZ 2004 to: • establish risk-based criteria by allowing for different

qualities of raw water • acknowledge the additive effect on protozoa removal

where more than one treatment process is used • make use of overseas studies that have assessed the

log removal efficacy of Cryptosporidium for a range of treatment processes

• include disinfection using UV light.

Same paragraph Mostly the same New paragraph New paragraph


3.1.3

Of the protozoa found in New Zealand waters, Crytosporidium is the most infective and is also the most difficult to remove or inactivate. Therefore it is assumed that if the treatment process successfully deals with Crytosporidium it will also deal with the other protozoa.

Turbidity

Reference to turbidity is made frequently throughout the Standards. It is measured: • as an aesthetic determinand • to demonstrate that the water is clean enough not

to impair the disinfection process • as a surrogate in the protozoa compliance

criteria. Having many applications, with different acceptable levels, the turbidity requirements can be confusing. To help clarify the situation, the various references to turbidity are summarised in Table 14.8.

New paragraph

A new section (3.1.3)

3.2 Microbiological compliance criteria

Separate criteria for compliance with the Standards are set for E. coli, and for the protozoa Giardia and Cryptosporidium. These are provided in Sections 3.2.2 - 3.2.3. In addition to these separate compliance criteria the following general criteria apply to all micro-organisms in Section 3.2:

3.2.1

General microbiological compliance criteria

Drinking-water complies with the microbiological compliance criteria if: • samples are taken at the required sites and

frequency for the determinand in question • the sampling, analytical and reporting procedures

comply with the requirements of the Standards • the remedial procedures specified in Section 3.4

Simplified


3.2.2

(action to be taken when transgression of the microbiological MAV occurs) are followed and the actions taken documented.

E. coli compliance criteria

E. coli compliance is assessed on the results of sampling for 12 consecutive months and requires that a drinking-water supply meets E. coli compliance criteria 1 and 2 below. If faecal, presumptive or total coliforms are measured, the results are to be treated as though they were E. coli. Note that a secure groundwater ‘leaves the treatment plant’ at the point where the water enters the distribution system. 3.2.2.1 E. coli compliance criterion 1: (for drinking-water leaving a treatment plant) All of a) – e) are complied with:

a) The water supply leaving the treatment plant is monitored for the presence of E. coli and turbidity.

b) The sampling and analytical techniques comply with the requirements of these Standards.

c) The frequency of sampling is equal to or greater than that specified in Table 3.1 (water leaving the treatment plant) for the population band and treatment type to which the water supply belongs.

d) Drinking-water leaving the treatment plant is sampled at any point at which the treatment process is fully complete including any required mixing or contact time, and before the first consumer. See Section 3.3.1.1.1.

e) The maximum number of 100 mL samples in which E. coli are found is equal to or less than:

0 in 76 samples

Merged criteria 1A and 1B Turbidity added in (a)


1 in 77 to 108 samples 2 in 109 to 138 samples 3 in 139 to 166 samples 4 in 167 to 193 samples 5 in 194 to 220 samples 6 in 221 to 246 samples 7 in 247 to 272 samples 8 in 273 to 298 samples 9 in 299 to 323 samples 10 in 324 to 348 samples

It should be noted that this table refers to the number of samples, irrespective of the frequency of sampling. Thus the number of transgressions in 250 samples is the same (7) whether all the samples are collected in one day or taken over the course of a year. For larger sets of samples, consult the Extended Table of Allowable Transgressions in the Statistical Considerations (Section 7.5.2) of the Guidelines. See also, Section 1.3 of these Standards for the concession re smaller supplies when testing fewer than 38 samples. The following additional requirements need to be satisfied in order to comply with E. coli criterion 1: a) Plants where chlorination is used: For the purpose of these Standards, chlorination is described as either continuously monitored, partly monitored, or undisinfected. Definitions follow: Continuously monitored chlorination:

Requires an on-line continuous FAC monitor, calibrated at least as frequently as recommended by the equipment suppliers, with an alarm system (FAC monitor or dosage monitor) that can prompt a site visit, without delay, to service the fault or condition. All plants with chlorination that supply a population greater than 10,000 must monitor FAC continuously. The FAC in the water leaving the plant should not fall below a concentration that is equivalent to a minimum of 0.2 mg/L FAC at pH 8.0 for more than 2 percent of the time in any one day. However, if the FAC concentration remains above 0.10 mg/L during this period, then the remedial action described in Section 3.4 need not be followed, but the event still needs to be reported.

New data This section now includes O3, ClO2 and UV New categories of FAC ‘monitoring’ New ‘concession’


If the chlorination process is ‘fully monitored’, then collecting samples for E. coli testing is optional (refer to Table 3.1).

Partly monitored chlorination:

Any continuously monitored supply that does not satisfy the above conditions, plus those supplies that do not have on-line FAC monitoring. For plants where the FAC is measured manually to be considered partly monitored, the FAC (equivalent at pH 8) must be at least 0.2 mg/L and be measured at least:

weekly at plants supplying up 500 people 3 times at week for plants supplying 501-5,000 people twice daily for plants supplying more than 5000.

Refer to Table 3.1 for E. coli sample collection requirements.

Undisinfected:

Any partly monitored supply that does not satisfy the above conditions, plus any supply that is not continuously chlorinated. Refer to Table 3.1 for E. coli sample collection requirements.

Free available chlorine (FAC) in the drinking-water leaving the treatment plant is to be monitored at a point after the contact time. Refer to Section 3.3.1.1.1 for comments on sampling sites. See Section 3.3.2.1.2 for a discussion on FAC monitoring. The contact time is greater than 30 minutes and the contact tank has been verified to be free of short-circuiting. The downtime of on-line monitoring equipment must be less than 1 hour in any week. Note that to gain maximum benefit in the distribution system section of the Ministry of Health’s Grading, the FAC concentration needs to be consistently above 0.2 mg/L. Because the FAC concentration falls as the water passes through the distribution system the optimum concentration leaving the treatment plant will need to be determined. If the pH is greater than 8.0 the equivalent FAC

Added note re 0.2 mg/L being too low if want an A grade for distribution system


concentration shown in Figure 3.1 shall be used. See Section 3.3.2.1.3 for a discussion on pH monitoring. The turbidity of the water leaving the treatment plant is less than 0.5 NTU for at least 95% of the time. Turbidity should be measured continuously or when a sample is collected for E. coli testing. See Section 3.3.2.1.4 for a discussion on turbidity monitoring. Note that for treatment plants serving fewer than 10,000 people, on-line process control measurements of FAC concentration made after only a short contact time may be used instead of readings from drinking-water leaving the plant, provided that:

a reliable correlation has been established, documented, and monitored, between the FAC concentration after the short contact time and the FAC concentration of drinking-water leaving the treatment plant

and the minimum value of the process control FAC concentration that has been established to be necessary to attain an FAC equivalent to a minimum of 0.2 mg/L FAC at pH 8.0 in drinking water leaving the treatment plant becomes the value used to demonstrate compliance.

b) Plants where chlorine dioxide is used: Before considering the use of chlorine dioxide, trials are required to show that the resultant chlorite concentration is not likely to exceed the MAV. Some free available chlorine (FAC) may appear in the final water so both FAC and chlorine dioxide (measured as ClO2) concentrations are monitored in the drinking-water leaving the treatment plant, at a point after the contact time. Refer to Section 3.3.1.1.1 for comments on sampling and Section 3.3.2.1.2 for monitoring. The contact time is greater than 30 minutes and the contact tank has been verified to be free of short-circuiting. If chlorine dioxide is being dosed to also inactivate protozoa (see Section 3.2.3), the contact time will likely be much greater than 30 minutes. The sum of the FAC and ClO2 concentrations in the water leaving the plant does not fall below 0.2 mg/L for more than 2 percent of the time in any one day. Chlorine dioxide reactions in drinking water are a lot less dependent on pH than is the case for FAC; studies have

Added the second sentence on turbidity, and, for at least 95% of the time

These process control measurements were meant to have been on-line ClO2 has been added


dependent on pH than is the case for FAC; studies have shown that ClO2 disinfection is better at a higher pH. The turbidity of the water leaving the treatment plant is less than 0.5 NTU. Turbidity should be measured continuously or when a sample is collected for E. coli testing. See Section 3.3.2.1.4 for a discussion on turbidity monitoring. All treatment plants disinfecting with chlorine dioxide and supplying a population of more than 10,000 should be ‘fully monitored’ ie, have continuous on-line FAC and ClO2 monitoring and (see b above for a definition of fully and partly monitored chlorination). While the plant is not manned there should be appropriate alarm arrangements to alert the operator to the need for a site visit to service the fault. The downtime of on-line monitoring equipment must be less than 1 hour in any week. If the sum of the FAC and ClO2 concentrations falls below 0.2 mg/L for less than 30 minutes in any one day, and stays above 0.10 mg/L, then the remedial action described in Section 3.4 need not be followed. For the purpose of determining E. coli sampling frequency in Table 3.1, fully monitored chlorine dioxide disinfection is equivalent to fully monitored chlorination, but if any one of the above conditions is not satisfied, the supply should be considered equivalent to an ‘undisinfected’ supply. The requirements for plants without continuous FAC and ClO2 monitoring that supply a population fewer than 10,000 are:

• a minimum of 2 samples should be tested for FAC and ClO2 per day, and

there should be an appropriate dosage alarm arrangement that indicates the need for a site visit to service the condition. c) Plants where ozonation is used: Before considering the use of ozone, trials are required to demonstrate that the resultant bromate concentration is not likely to exceed the MAV (0.025 mg/L). The conditions for acceptable ozone disinfection are described in Section 3.2.3, Protozoa Criterion c (i).

New Rewrite Section numbers later


These conditions should also satisfy the E. coli criterion. d) Plants where UV disinfection is used: The conditions for acceptable UV disinfection are described in Section 3.2.3, Protozoa Criterion c (ii). These conditions should also satisfy the E. coli criterion, except that for bacteria and virus inactivation, a minimum UV dose (fluence) of 40 mJ/cm2 (400 J/m2) is required.

New Rewrite Section number

3.2.2.2 E. coli compliance criteria 2A and 2B: for drinking-water in the distribution system E. coli compliance criterion 2A (criterion using E. coli monitoring only) All of a) – g) are complied with:

a) The water supply in the distribution system is monitored for the presence of E. coli.

b) Turbidity in water in the distribution system is

more variable than in water leaving the treatment plant. Therefore a turbidity sample should be collected at the same time as the E. coli sample.

c) The sampling and analytical techniques comply

with the requirements of these Standards.

d) Drinking-water is sampled at distribution system sampling sites as specified in Section 3.3.1.1.2.

e) The frequency of sampling for E. coli and

turbidity in the distribution system is equal to or greater than that specified in column 2 of Table 3.2a (water in a distribution system) for the population band to which the water supply belongs.

f) The maximum number of 100 mL samples in

which E. coli are found when sampled at the frequency specified in Table 3.2 is defined in note (e) in Section 3.2.2.1.

g) The median turbidity value shall be less than 1.0

NTU and no sample shall exceed 5.0 NTU. or

b) has been added g) has been added, to match Grading


E. coli compliance criterion 2B (criterion allowing partial substitution of E. coli monitoring in the distribution system by FAC or ClO2 monitoring). A water supply that maintains a residual of disinfectant throughout the distribution system has a reduced risk from microbial contamination. Therefore, provided that there is evidence of satisfactory disinfection residuals, there is a reduced need for E. coli testing. All of h) – n) are complied with:

h) The population is greater than 30,000, and the FAC in the water leaving the treatment plant does not fall below has a concentration that is equivalent to at least 0.2 mg/L FAC at pH 8.0 for more than 2% of the time.

Or, the population is greater than 30,000, and the sum of the chlorine dioxide (as ClO2) and FAC concentrations does not fall below 0.2 mg/L for more than 2% of the time.

i) The sampling and analytical techniques comply

with the requirements of these Standards.

j) The water leaving the treatment plant has a turbidity less than 0.5 NTU for at least 95 percent of the time.

k) The number of E. coli samples substituted by

FAC tests does not exceed 75 percent of the number of E. coli samples that are specified in Table 3.2a to be used when no substitution by FAC measurements occurs.

l) E. coli and FAC (or ClO2 and FAC) samples are

taken at the frequency and distribution specified in Tables 3.2a and 3.2b for the situation where substitution for E. coli samples by FAC (or ClO2 and FAC) measurements occurs.

m) The maximum number of 100 mL samples in

which E. coli are found when sampled at the frequency specified in Table 3.2 defined in note (e) in Section 3.2.2.1.

n) For partial substitution of E. coli testing by FAC

testing, all samples in the distribution system i FAC i l l 0 2

ClO2 added here – it is consistent with the residual in distribution system concept Now 98 percent of time, like Criterion 1 i) has been added in j) turbidity has been added


must contain FAC equivalent to at least 0.2 mg/L FAC at pH 8.0, except in areas of low flow where the FAC concentration may diminish to 0.1 mg/L. The median turbidity should be less than 1.0 NTU, with no sample exceeding 5.0 NTU. If these conditions are not met for any particular sample, E. coli is to be tested for.

Or, for partial substitution of E. coli testing by ClO2 plus FAC testing, the sum of the disinfectants in all samples must contain at least 0.2 mg/L, except in areas of low flow where the disinfectant concentration may diminish to 0.1 mg/L. The median turbidity should be less than 1.0 NTU, with no sample exceeding 5.0 NTU. If these conditions are not met for any particular sample, E. coli is to be tested for.

Changed from 0.5 NTU in n) to match the Grading values

3.2.2.2.1 FAC correction for pH greater than 8 At pH values greater than 8 the disinfection equivalent of the FAC decreases. If, for any reason, the pH temporarily exceeds 8.0, the chlorine concentration shall be increased accordingly. The FAC concentration necessary to provide the disinfection equivalent of 0.2 mg FAC/L at pH 8.0 is shown in Figure 3.1.

Previous first sentence deleted – it was incorrect

Figure 3.1: FAC concentrations at different pHs required to provide the disinfection equivalent of

0.2 mg FAC/L at pH 8.0

FACpH8eq

00.20.40.60.81

1.21.41.6

7.8 8 8.2 8.4 8.6 8.8 9 9.2

pH

FACmg/L


Suggested changes

3.2.2.3 E. coli compliance criteria 3A and 3B: for drinking-water in a bulk distribution zone

E. coli compliance criterion 3A (criterion using E. coli monitoring only): All of a) – f) are complied with:

a) The water supply at the transition point is monitored for the presence of E. coli and turbidity.

b) The sampling and analytical techniques comply with

the requirements of these Standards.

c) Drinking-water is sampled at bulk distribution zone sampling sites as specified in Section 3.3.1.1.2.

d) The frequency of sampling for E. coli at the

transition point is equal to or greater than that specified in column 2 of Table 3.2c (water in the bulk reticulation zone) for the population band that the bulk distribution zone services.

e) The maximum number of 100 mL samples in which

E. coli are found when sampled at the frequency specified in Table 3.2 shall be equal to or less than those prescribed in Section 3.2.2.1 note (e).

f) The turbidity values shall have a median less than

1.0 NTU, with no sample exceeding 2.0 NTU.

Bulk distribution zone a new concept Turbidity added in a) Added f) re turbidity. Allows a little for turbidity in bulk mains being intermediate between that leaving WTP and that in client’s distribution system

E. coli compliance criterion 3B (criterion for chlorinated fully monitored supplies) Both g) – h) are complied with:

g) The FAC (and if relevant, ClO2) content is monitored continuously.

h) The water always contains FAC equivalent to at least

0.2 mg/L FAC at pH 8.0, or the sum of the chlorine dioxide (as ClO2) plus FAC is at least 0.2 mg/L.

Monitored where? At the sample site (which is the transition point – ie, bulk meter)? And at how many? Cannot do it at them all – how about the one that’s furthest away?


3.2.3

Protozoa (Giardia and Cryptosporidium) compliance criteria Protozoa can be removed physically by filtration processes, or inactivated by disinfection processes. Chlorine is very effective in controlling bacteria and viruses, but not a practical process for protozoa. Recent studies have shown that UV light can inactivate protozoa but higher doses are required for inactivation of some bacteria and viruses. To produce a safe drinking-water, complying with both the E. coli and protozoa criteria, a combination of treatment processes may offer the most practical option. International studies have measured the efficiency of protozoa removal or inactivation of many treatment processes. The efficiency is measured as percent removal or converted to log removal rates. These are cumulative in multi-process treatment systems and can be called protozoa credits. The credits are earned on the condition that various criteria are met. The rationale is described in Section 3.1.2. The degree of treatment needed is related to the quality of the raw water. For the purposes of these Standards, raw waters are divided into two groups: ‘clean’, or ‘dirty’, based on the results of either an E. coli monitoring programme, or a sanitary survey. Clean raw waters require 3 log removals of Cryptosporidium, and dirty raw waters require 4 log removals. This applies to all waters except secure groundwaters. The credits that may apply to the various treatment processes are listed in Table 3.1. The protozoa compliance criteria that follow apply to water in, or leaving, a treatment plant at the sampling points specified in Section 3.3.1.2. If the protozoa criteria for water leaving the treatment plant are complied with, and the water in the distribution system complies with relevant E. coli criteria, then the water in the distribution system is considered to offer only a small risk due to protozoa. Although reliable direct enumeration of Giardia and Cryptosporidium strains can now be made, this is not used as a compliance criterion because of the high degree of uncertainty as to the interpretation of the results. The method is suitable for use in the investigation of drinking-water supplies, particularly in the case of disease outbreaks.

Completely rewritten Or do we need a third category for ‘very dirty raw waters’? Or would it be better to simplify that and ask for 4 log removal for all raw waters? A new Table 3.1, so others will be renumbered


Drinking-water complies with the protozoa criteria if the treatment it receives earns the required number of protozoa log removal credits, and satisfies the criteria for the treatment processes used. Alternatively, the source may be declared a secure groundwater (refer Section 3.2.4). Table 3.1 lists 14 treatment processes that can earn protozoa log removal credits. Unless stated otherwise, these are additive. The number of credits required is dependent on the quality of the raw water. Raw waters are divided into two groups: ‘clean’, or ‘dirty’, based on either an analysis of monthly E. coli testing over a 12 month period, or a sanitary survey. The boundary to be used for springs, groundwaters that are not secure, reservoirs and lakes shall be that 95 percent of samples shall contain fewer than 10/100 mL E. coli. The boundary to be used in flowing river and stream systems shall be that 95% of samples shall contain fewer than 50/100 mL. Clean raw waters require 3 log removals of Cryptosporidium, and dirty raw waters require 4 log removals. The raw water samples shall be collected according to a predetermined programme that covers different weather and flow patterns, and different days of the week. Alternatively, raw waters that will clearly fall into the ‘dirty’ category can be classified as such without the need for E. coli testing.

Table 3.1: Log credits and conditions for control of protozoa

Treatment Process Cryptosporidium log credit if complying with criteria (which follow)

Once the new raw water E. coli data has been assessed new boundaries may need to be established Note: some categories in the EPA version of the Table have been eliminated– not relevant in NZ. These were : watershed control pre-sed with coagulation lime softening + some for which no credits were on offer


Membranes Bag filters Cartridge filters Combined filter performance Individual filter performance Second stage filtration Chlorine dioxide Ozone UV

Credit equivalent to removal efficiency demonstrated in challenge test for device if supported by direct integrity testing. Potentially, 5 or more log removals. 1 log credit with demonstration of at least 2 log removal efficiency in challenge test. 2 log credit with demonstration of at least 3 log removal efficiency in challenge test. 0.5 log credit for combined filter effluent turbidity ≤0.15 NTU in 95% of samples each month. 1.0 log credit for demonstration of filtered water turbidity < 0.1 NTU in 95 percent of daily max values from individual filters (excluding 15 min period following backwashes) and no individual filter >0.3 NTU in two consecutive measurements taken 15 minutes apart. 0.5 log credit for second separate filtration stage; treatment train must include coagulation prior to first filter. No presumptive credit for roughing filters. Log credit based on demonstration of log inactivation using ClO2 C.t table. Log credit based on demonstration of log inactivation using ozone CT table. Log credit based on demonstration of inactivation with UV dose table; reactor testing required to establish validated operating conditions. If dose high enough to inactivate bacteria and viruses, can earn up to 3.0 protozoa log credits.

Bank filtration (infiltration gallery): All the water must be drawn from wells in an unconsolidated, predominately sandy aquifer. Core samples must contain at least 10 percent fine-grained material (less than 1.0 mm diameter) in at least 90 percent of their length. The turbidity of water leaving the wells must be <1.0 NTU in 95% of the samples, with none greater than 5.0 NTU. Supplies that do not monitor turbidity continuously must have documented evidence that the turbidity of the water leaving the well does not exceed 2.0 NTU for the week after a flood in the river.


To earn the 0.5 log credit for 7.5 m setback, or the 1.0 log credit for 15 m setback, water supplies using bank filtration must monitor turbidity at the position described in Section 3.3.1.2 and at the frequency described in Section 3.3.2.2. Flocculation, coagulation, sedimentation, filtration: To obtain 3 log credits: • All water passes through the full process, which must be

continuous. • At all times the chemical dosed must form a floc or

particles capable of removal by filtration. If a floc does not form in low turbidity raw water, it is possible for water to pass through the filters with little removal of particles. This condition may not be detected by an increase in turbidity. Protozoa may not be removed in these conditions.

• Sampling frequencies and reporting periods are discussed in Section 3.3.2.2 and summarised (for turbidity) in Table 3.4 (now Table 3.5).

• Measurement of the turbidity of the water leaving each filter satisfies the following requirements:

• (i) 95 percent of the turbidity measurements over the reporting period (at the frequency specified in Table 3.4) (now Table 3.5) do not exceed 0.2 NTU.

and

(ii) at any time during a filter run (excluding any period of filtering to waste) the turbidity does not exceed 0.5 NTU.

and (iii) for continuous monitoring, no increases of more than 0.2 NTU shall occur in any 10 minute period, except that up to two turbidity records per day may exceed 1.0 NTU to allow for spurious peaks.

As at 1 January 2005 all drinking-water suppliers servicing more than 10,000 people will be required to continuously monitor each filter for turbidity, and all filters must give an alarm to duty staff in the event of transgression. Coagulation, direct filtration: To obtain 2.5 log credits: • All water passes through the filters. If treatment operates

on/off, filtrate must be recycled or wasted for the length of time that has been demonstrated to be necessary to reach optimum performance.

Re: second bullet: so should we monitor for aluminium (or iron if used)? Should we delete requirement (iii) now that effluent from each filter shall be <0.2 NTU for at least 95 percent of the time? And continuous monitoring for towns with population >5000 by 2008?


• At all times the chemical dosed must form a floc or particles capable of removal by filtration. If a floc does not form in low turbidity raw water, it is possible for water to pass through the filters with little removal of particles. This condition may not be detected by an increase in turbidity. Protozoa may not be removed in these conditions.



• (i) 95 percent of the turbidity measurements over the reporting period (at the frequency specified in Table 3.4) do not exceed 0.5 NTU. If the turbidity of the raw water is <0.5 NTU, the turbidity of the filtrate must always be less than that in the raw water. (Raw water monitoring of turbidity is already being carried out in order to determine the coagulant dose rate).

and (ii) at any time during a filter run (excluding any period of filtering to waste) the turbidity does not exceed 1.0 NTU.

and (iii) for continuous monitoring, no increases of more than 0.2 NTU shall occur in any 10 minute period, except that up to two turbidity records per day may exceed 1.0 NTU to allow for spurious peaks. As at 1 January 2005 all drinking-water suppliers servicing more than 10,000 people will be required to continuously monitor each filter for turbidity, and all filters must give an alarm to duty staff in the event of transgression.

Diatomaceous earth filtration: To obtain 2.5 log credits: • All water passes through the filters. If treatment operates

on/off, filtrate must be recycled or wasted for the length of time that has been demonstrated to be necessary to reach optimum performance.

• Trials are required to determine the minimum diatomaceous earth precoat thickness that will reliably remove protozoa in different raw water conditions, before switching from the recycle to production mode.


NOT ditto – as above because filter effluent can be more turbid for this treatment continuous monitoring for towns with population >5000 by 2008?



• (i) 95 percent of the turbidity measurements over the

reporting period (at the frequency specified in Table 3.4 now Table 3.5) do not exceed 0.5 NTU. If the turbidity of the raw water is <0.5 NTU, the turbidity of the filtrate must always be less than that in the raw water. (Raw water monitoring of turbidity is already being carried out in order to determine the diatomaceous earth dose rate).


and (iii) for continuous monitoring, no increases of more than 0.2 NTU shall occur in any 10 minute period, except that up to two turbidity records per day may exceed 1.0 NTU to allow for spurious peaks. As at 1 January 2005 all drinking-water suppliers servicing more than 10,000 people will be required to continuously monitor each filter for turbidity, and all filters must give an alarm to duty staff in the event of transgression.

Slow sand filtration: To obtain 2.5 log credits for a slow sand filter operating as the primary treatment process: • All water passes through the filters. If treatment operates

on/off, filtrate must be recycled or wasted for the length of time that has been demonstrated to be necessary to reach optimum performance. The filter must not be allowed to dry out.

• Disinfecting chemicals must not be dosed upstream of the filter beds.

• Slow sand filters should operate at a constant flow rate, which should be less than 0.35 m/h.

• After sand scraping or replacement, the water must be recycled until the treatment process has been demonstrated to have returned to full efficiency.

• The efficiency of slow sand filtration may be reduced appreciably at water temperatures below 6°C so additional microbiological monitoring may be required.


• Measurement of the turbidity of the water leaving each

NOT ditto – as above because filter effluent can be more turbid Continuous monitoring for towns with population >5000 by 2008?


filter satisfies the following requirements: • (i) 95 percent of the turbidity measurements over the

reporting period (at the frequency specified in Table 3.4) do not exceed 0.5 NTU. If the turbidity of the raw water is <0.5 NTU, the turbidity of the filtrate must always be less than that in the raw water. This will require regular turbidity testing of the raw water.


Membranes: To receive removal credits, the membrane filtration process must:

1. meet the basic definition of a membrane filtration proces

and 2. have its removal efficiency established by challenge

testing and verified by direct integrity testing and

3. undergo periodic direct integrity testing and continuous indirect integrity monitoring during use.

All water must pass through the filters. If treatment operates on/off, filtrate must be recycled or wasted for the length of time that has been demonstrated to be necessary to reach optimum performance. Most direct integrity test methods are applied periodically because the membrane unit must be taken out of service to conduct the test. In order to assess process performance between direct integrity testing events, continuous indirect integrity testing is required. This involves monitoring some aspect of filtrate quality that is indicative of the removal of particulate matter. The maximum removal credit that a membrane filtration process is eligible to receive is equal to the lower value of either: • The removal efficiency demonstrated during challenge

testing or • The maximum log removal value that can be verified

through the direct integrity test (i.e. integrity test sensitivity) used to monitor the membrane filtration process.

Cross-ref to E. coli criterion? (re 6° C)


Potentially, 5 or more log removal credits are available. Membrane material configured into a cartridge filtration device that meets the definition of membrane filtration, and that can be direct integrity tested according to the criteria specified in this section, is eligible for the same removal credit as a membrane filtration process. Challenge testing is a product-specific rather than a site-specific requirement. Water suppliers using membrane filtration may adopt the equipment or appliance supplier’s certification, provided it meets the requirements of the Membrane Filtration Guidance Manual (USEPA 2003e), or a standard formally recognised by the Ministry of Health as being equivalent, and that the testing was performed by an appropriately accredited inspection body. The equipment to be installed must be identical (or validated as equivalent) to the equipment that was tested during the certification process. A direct integrity test is defined as a physical test applied to a membrane unit in order to identify and isolate integrity breaches. An integrity breach is defined as one or more leaks that could result in contamination of the filtrate. The direct integrity test method must be applied to the physical elements of the entire membrane unit including membranes, seals, potting material, associated valving and piping, and all other components that under compromised conditions could result in contamination of the filtrate. The direct integrity tests used at present include those that use an applied pressure or vacuum (such as the pressure decay test and diffusive airflow test), and those that measure the rejection of a particulate or molecular marker (such as spiked particle monitoring). The direct integrity test must meet performance criteria for resolution, sensitivity, and frequency. The direct integrity test used at any installation is defined by the manufacturer/supplier for use with their equipment. Provided it meets the requirements of the Membrane Filtration Guidance Manual (USPA 2003e), or a standard formally recognised by the Ministry of Health, the technique will meet the requirements of these Standards. A control limit is defined as an integrity test response which, if exceeded, indicates a potential problem with the system and triggers a response. A control limit for a direct integrity test must be established that is indicative of an integral membrane unit capable of meeting the Cryptosporidium removal credit awarded. If the control limit for the direct integrity test is exceeded the membrane unit must be taken off line for

Will put definitions in Glossary and description of membrane filtration in the Guidelines Put detailed description of challenge testing in Guidelines


exceeded, the membrane unit must be taken off-line for diagnostic testing and repair. The membrane unit can only be returned to service after the repair has been completed and confirmed through the application of a direct integrity test. Most direct integrity tests available at present are applied periodically and must be conducted on each membrane unit at a frequency of not less than once every 24 hours while the unit is in operation. If continuous direct integrity test methods become available that also meet the sensitivity and resolution criteria described earlier, they may be used in lieu of periodic testing. The reporting period summarising all direct integrity test results above the control limit associated with the Cryptosporidium removal credit awarded to the process and the corrective action that was taken in each case shall be monthly. Continuous indirect integrity monitoring is defined as monitoring some aspect of filtrate water quality that is indicative of the removal of particulate matter. If a continuous direct integrity test is implemented that meets the resolution and sensitivity criteria described previously, continuous indirect integrity monitoring is not required. Continuous indirect integrity monitoring must be conducted according to the following criteria: • continuous filtrate turbidity monitoring on each membrane

unit • if the filtrate turbidity readings are above 0.15 NTU for a

period greater than 15 minutes, direct integrity testing must be performed on the associated membrane units

• at no time should the filtrate turbidity exceed the turbidity of the feed water. This will require regular turbidity testing of the raw water.

• if indirect integrity monitoring includes an approved alternative test (such as particle counting) and if the alternative test exceeds the approved control limit for a period greater than 15 minutes, direct integrity testing must be performed on the associated membrane units

• a monthly report must be submitted summarising all indirect integrity monitoring results triggering direct integrity testing and the corrective action that was taken in each case.

Sampling frequencies and reporting periods are discussed in Section 3.3.2.2, and in Table 3.4 (now 3.5) for turbidity.

Discussion on resolution, sensitivity and control limit in Guidelines


Bag filters: To receive removal credits, the process must meet the basic definition of a bag filter, and establish its removal efficiency through challenge testing. To receive 1 log credit, a bag filter must have a demonstrated efficiency of 2 log or greater for Cryptosporidium removal. The 1 log factor of safety is applied to the removal credit because: • the removal efficiency of bag filters has been observed to

vary by more than 1 log over the course of operation • bag filters are not routinely direct integrity tested during

operation in the field, so there is no means of verifying the removal efficiency of filtration units during routine use.

Challenge testing is a product-specific rather than a site-specific requirement. Water suppliers using bag filtration may adopt the equipment or appliance supplier’s certification, provided it meets the requirements of the Bag and Cartridge Guidance Manual (USEPA 2003e), or a standard formally recognised by the Ministry of Health as being equivalent, and that the testing was performed by an appropriately accredited inspection body. Testing must be on entire units, including filtration media, seals, filter housing, and other components integral to the process. The equipment to be installed must be identical (or validated as equivalent) to the equipment that was tested during the certification process. All water must pass through the filters. A slow opening/closing valve is required ahead of the filter to minimise flow surges, and a flow restrictor is required to maintain flows below the certificated maximum operating rate of the filters. The flow through each filter shall be recorded. If treatment operates on/off, filtrate must be recycled or wasted for the length of time that has been demonstrated to be necessary to reach optimum performance; a minimum of 5 minutes is recommended. Filtrate should pass to a storage facility (eg, pressure tank or service reservoir). If sized correctly, this should eliminate surges through the filter. At a minimum, pressure gauges should be located before and after each filter. These should read to within 1.0 kPa accuracy. The differential pressure between the inlet and outlet header should be monitored to determine when the filter needs replacement, or attention, as per the conditions in the manufacturer’s certification. An alarm should be linked to the pressure gauge system that ensures the operator is alerted


when the system is out of specification. Sampling frequencies and reporting periods are discussed in Section 3.3.2.2 and summarised (for turbidity) in Table 3.4 (now Table 3.5). Measurement of the turbidity of the water leaving each filter satisfies the following requirements:

(i) 95 percent of the turbidity measurements over the reporting period (at the frequency specified in Table 3.4) do not exceed 1.0 NTU. If the turbidity of the raw water is <1.0 NTU, the turbidity of the filtrate must always be less than that in the raw water. This will require regular turbidity testing of the raw water.

and (ii) at any time during a filter run (excluding any

period of filtering to waste) the turbidity does not exceed 2.0 NTU.

Cartridge filters: To receive removal credits, the process must meet the basic definition of a cartridge, and establish its removal efficiency through challenge testing. To receive 2 log credits, a cartridge must have a demonstrated efficiency of 3 log or greater for Cryptosporidium removal. Testing must be on entire units, including filtration media, seals, filter housing, and other components integral to the process. The 1 log factor of safety is applied to the removal credit because: • the removal efficiency of some cartridge filters has been

observed to vary by more than 1 log over the course of operation

• cartridge filters are not routinely direct integrity tested during operation in the field; hence, there is no means of verifying the removal efficiency of filtration units during routine use.

Challenge testing is a product-specific rather than a site-specific requirement. Water suppliers using cartridge filtration may adopt the equipment or appliance supplier’s certification, provided it meets the requirements of the Bag and Cartridge Guidance Manual (USEPA 2003e), or a standard formally recognised by the Ministry of Health as being equivalent, and that the testing was performed by an appropriately accredited inspection body. Testing must be on entire units, including filtration media, seals, filter housing, and other components integral to the process. The equipment

Pressure gauges, with suitable alarm telemetry – but for what size communities? And what does a pressure gauge tell us? Differential too high – that means it’s pretty much blocked. But if it’s short-circuiting, there may be no differential. Or as it blocks it may start to discharge particles with the result that the differential stays much the same. Ditto re cartridges. And surge control? A lot of good ideas or practices can only really be Guidelines, not sound enough to be criteria for DWSNZ? Definition of bag filters and their challenge testing protocol to go in Guidelines Integrity testing: the non-destructive tests applied to membrane filtration processes cannot be applied to most bag


to be installed must be identical (or validated as equivalent) to the equipment that was tested during the certification process. For further discussion, refer to the bag filtration section of the Guidelines. Cartridge filters and housing shall be certified to the cyst/oocyst reduction requirements of NSF/ANSI 53-2002 (plus addenda 1 and 2) or a standard formally recognised by the Ministry of Health as being equivalent. The cartridge filter (or the packaging containing up to 50 individual cartridges) shall be labelled in accordance with clause 7.3 of NSF/ANSI 53-2002 (plus addenda 1 and 2). The cartridge filter housing shall be labelled in accordance with clause 7.2 of NSF/ANSI 53-2002 (plus addenda 1 and 2). All water must pass through the filters. A slow opening/closing valve is required ahead of the filter to minimise flow surges, and a flow restrictor is required to maintain flows below the certificated maximum operating rate of the filters. The flow through each filter unit or module or shall be recorded. If treatment operates on/off, filtrate must be recycled or wasted for the length of time that has been demonstrated to be necessary to reach optimum performance; a minimum of 5 minutes is recommended. Filtrate should pass to a storage facility (eg, pressure tank or service reservoir). If sized correctly, this should eliminate surges through the filter. At a minimum, pressure gauges should be located before and after the filter unit or module. These should read to within 1.0 kPa accuracy. The differential pressure between the inlet and outlet header should be monitored to determine when the filter needs replacement, or attention, as per the conditions in the manufacturer’s certification. An alarm should be linked to the pressure gauge system that ensures the operator is alerted when the system is out of specification. If the module comprises multiple cartridges, there must be a process that indicates a flow or pressure problem with an individual cartridge. Sampling frequencies and reporting periods are discussed in Section 3.3.2.2 and summarised (for turbidity) in Table 3.4 (now Table 3.5). Measurement of the turbidity of the water leaving each filter satisfies the following requirements:

filters. At present (EPA says) there are no alternative direct non-destructive tests that can be used reliably with these devices. Typical process monitoring for bag filters includes turbidity and pressure drop to determine when filters must be replaced or faults attended to. However, the applicability of either of these process monitoring parameters as tools for verifying removal of Cryptosporidium has not been demonstrated with confidence. But I have included turbidity anyway! Many raw waters (with a turbidity of say 0.4 NTU) that are passed through a cartridge or bag filter (or even DE filter) will ultimately block it. Despite all those particles being removed, it is not uncommon to find that the turbidity of the effluent is not significantly different from the influent. That is because most of the particles contributing to the turbidity are very small and pass through the filter. But the filter is probably still removing protozoa effectively – so how can we prove it? If we cannot do it by measuring turbidity, then we can hardly ask water suppliers to monitor for turbidity. In these raw water conditions, it looks increasingly like particle counting is going to be needed. But a lot of towns treating ‘cleanish’ raw waters with cartridges are rather small and probably do not want to buy a particle monitor. So what can we do? The smaller the system, the worse it gets! The alternative for very small supplies (<500?) is (for Ministry of Health?) to contract out (subsidise?) somebody (HPO surveillance?) with a particle counter to do a regular circuit of suitably sized regions. Or: Some researchers (eg, Rice et al


(i) 95 percent of the turbidity measurements over the reporting period (at the frequency specified in Table 3.4) do not exceed 1.0 NTU. If the turbidity of the raw water is <1.0 NTU, the turbidity of the filtrate must always be less than that in the raw water. This will require regular turbidity testing of the raw water.

and (ii) at any time during a filter run (excluding any

period of filtering to waste) the turbidity does not exceed 2.0 NTU.

Integrity testing: any membrane cartridge filtration device that can be direct integrity tested is eligible for removal credits as a membrane, subject to the criteria specified in that section. This section applies to cartridge filters that cannot be direct integrity tested. Typical process monitoring for cartridge filters includes turbidity and pressure drop to determine when filters must be replaced or faults attended to. However, the applicability of either of these process monitoring parameters as tools for verifying removal of Cryptosporidium has not been demonstrated with confidence. Combined filter performance: A 0.5 log credit is available if the turbidity of the combined filter effluents is less than 0.15 NTU in 95% of samples tested in each reporting period, and no sample should exceed 0.5 NTU. This requires continuous turbidity monitoring of the combined filtrates, or a system that calculates the mean turbidity from the readings from on-line turbidimeters on each filter. This credit is only available to plants that include a coagulation process. Sampling frequencies and reporting periods are discussed in Section 3.3.2.2. Individual filter performance: A 1.0 log credit is available by demonstrating that:

(1) the filtered water turbidity is less than 0.10 NTU in at least 95% of the maximum daily values recorded at each filter in each month, excluding the 15 minute period following backwashes,

and (2) no individual filter has a turbidity greater than

0.3 NTU in two consecutive measurements taken 15 minutes apart.

JAWWA Sept 1996 122-129) have reported that aerobic spore removal rates are very similar to particle removal rates. I like the idea of doing plate counts of raw and treated water, ie, using the natural bacteria. Mostly they’d be aerobic bacteria, which are generally smaller than the protozoa. I would use the m/f method so large volumes can be filtered. Trials would be needed for each supply to see how much water needs to be filtered. With such low turbidity water it should be possible to filter 5 L (for example) to ensure a reasonable number of bacteria grow. Also, the optimum incubation temperature and time will need to be determined. Some waters may be OK (if fairly large bacterial population) at 35° C for 48 h; others may be better at 22° C (or even ‘room temperature’ – ie, no incubator needed!) for 72 h. An example: Say 25 mL of raw water was filtered through the membrane and incubated, and 170 colonies were counted = 6800 colonies per L. Say 5 L of filtered water was filtered through the membrane and incubated, and 50 colonies were counted = 10 colonies per L. The removal calculates out at 99.7%, which is 2.5 log removal. Would that give confidence that the filter was removing the protozoa effectively? If so, it looks like being much cheaper than having to buy a particle counter. Or am I being a bit over-confident about their bacterial removal efficiency? Roughly speaking, I’d guess protozoa are about double the size of these bacteria – maybe 50% removal of bacteria is more like


This credit is only available to plants that include a coagulation process. Systems that receive the 1 log additional treatment credit for individual filter performance cannot also receive the additional 0.5 log credit for combined filter performance. Sampling frequencies and reporting periods are discussed in Section 3.3.2.2. Second stage filtration: A 0.5 log credit is available for a second, separate filtration stage. The treatment train must include coagulation prior to the first filter, and both stages must treat all of the flow. To be eligible for this credit, the secondary filtration must consist of rapid sand, dual media, granular activated carbon (GAC), or other fine grain media in a separate stage following rapid sand or dual media filtration. A cap, such as GAC, on a single stage of filtration will not qualify for this credit. The turbidity of the water leaving each filter unit which together comprise the second filter must not exceed 0.15 NTU in 95% of samples, and no sample shall exceed 0.5 NTU. Sampling frequencies and reporting periods are discussed in Section 3.3.2.2. Chlorine dioxide: The credits available are based on the demonstration of log inactivation as stated in the ClO2 C.t table (Table 13.1). Refer to the Guidelines for Drinking-Water Quality Management for New Zealand for guidance on determining contact times.

Table 13.1: C.t values for Cryptosporidium inactivation by chlorine dioxide

Water temperature, °C1 Log credit 1 2 3 5 7 10 15 20 25 0.5 1.0 1.5 2.0 2.5 3.0

305 610 915 1220 1525 1830

279 558 838 1117 1396 1675

256 511 767 1023 1278 1534

214 429 643 858 1072 1079

180 360 539 719 899 1286

138 277 415 553 691 830

89 179 268 357 447 536

58 116 174 232 289 347

38 75 113 150 188 226

1 C.t values between the indicated temperatures may be determined by interpolation.

99.7% removal of protozoa? Definition of cartridge filters goes in Guidelines See above for a discussion re what we can do about that, ie, nothing obvious to monitor Is it a good idea to have a category worth an extra 0.5 log credit for dropping from 0.2 NTU to just 0.15? 95% of daily values in a month means one daily maximum value per month is allowed to exceed 0.10 NTU. It is assumed that every filter can exceed 0.10 NTU once a month – so long as it is <0.30. The USEPA wonders whether this is a very practical credit. They say a turbidity of 0.14 NTU can be rounded to 0.1 NTU so it’s hardly any different from the combined filter effluent credit. That’s pretty much all in the area of measurement uncertainty anyway! However, they do say a large plant with 1 dud filter could still have a combined filter effluent <0.15 NTU.


All water treated with chlorine dioxide shall have a turbidity of less than 0.5 NTU. Refer to Section 3.2.2.1 (c) for information about chlorine dioxide with respect to E. coli compliance criteria. Sections 3.3.1.1.1 and 3.3.2.1.2 discuss sampling and testing for compliance with E. coli criteria. The high dose required for effective inactivation of protozoa by chlorine dioxide will almost invariably cause the concentration of chlorite to exceed its MAV of 0.3 mg/L. Based on a 50-70 percent conversion from ClO2 to chlorite (ClO2

-) the maximum dose is limited to about 0.5 - 0.6 mg/L. Therefore very long contact times are required.

The C.t values in this table are not the same as in DWSNZ 2000. Replace, with units! Take out the 2, 3, 4, and 7° C columns. Question – re last paragraph - why do we have 0.3 MAV when WHO provisional Guideline is 0.7 and USEPA MCL is 1.0 mg/L? (WHO was 0.2 in 1993)

Ozone: The credits available are based on the demonstration of log inactivation as stated in the ozone C.t table (Table 13.1).

Table 13.1: C.t values for Cryptosporidium inactivation by ozone

Water TemperatuLog credit 1 2 3 5 7 0.5 1.0 1.5 2.0 2.5 3.0

12 23 35 46 58 69

10 21 31 42 52 63

9.5 19 29 38 48 57

7.9 16 24 32 40 47

6.5 13 20 26 33 39

1 C.t values between the indicated temperatures may be determined by interpolation.

The dissolved ozone concentration shall be measured continuously in the contactor. The monitor shall include an alarm that indicates to an operator when the ozone

This is new C.t data ex LT2ESWTR – so becomes Table 13.2, with units added! Delete the 2, 3, 4, and 7° C columns? Alarms for small schemes? But do small schemes use ozone?


concentration is too low. All the water passes through the ozone reactor, and for at least 95 percent of the time shall have a turbidity of less than 1.0 NTU, and no sample shall exceed 2 NTU. C.t values are established to provide a conservative characterisation of the dose of ozone necessary to achieve a specified inactivation of Cryptosporidium. C.t is defined as the product of the disinfectant concentration and disinfectant contact time:

C.t = disinfectant (mg/L) x contact time (in minutes)

where t is the time it takes the water to move from the point where the initial disinfectant residual concentration is measured to the point where the final disinfectant residual concentration is measured in a specified disinfectant segment, and C is the measured concentration of dissolved ozone in mg/L

There are three common methods for calculating C.t.

1. T10 2. Continuous stirred tank reactor (CSTR) 3. Extended-CSTR

These methods are described in the Guidelines. The C.t is calculated daily, using the peak hourly flow. The T10 method calculates C.t through a contactor assuming hydraulic conditions are similar to plug flow and can be used without tracer study data. T10 is the time it takes for 90 percent of the water to pass through the contactor. Even in well-baffled contactors, the T10 is most often less than 65 percent of the average hydraulic detention time through the contactor, and generally under-estimates the true C.t achieved. Ozone inactivation kinetics are not affected by pH in the range of pH 6-10. However, ozone becomes less cost-effective as the pH increases. The C.t data in Table 13.1 (or 13.2) are valid for ozone concentrations in the range 0.2-5.0 mg/L. Sampling frequencies and reporting periods are discussed in Section 3.3.2.2. UV:


The credits available are based on the demonstration of inactivation as stated in the UV dose table (Table 13.2 or 13.3). If the UV dose is high enough to inactivate bacteria and viruses, the process can earn up to 3.0 protozoa log credits. The doses in Table 13.2 or 13.3 are based on UV light at 254 nm as delivered by a low pressure mercury lamp. The equivalent doses for other lamps (such as medium pressure) can be determined through reactor validation testing. Irradiance is the power per unit area incident from all upward directions on an infinitesimally small element of surface area dA, divided by dA, whereas fluence rate (intensity) is the power incident from all directions on to an infinitesimally small sphere of cross-section dA, divided by dA. Both have the SI unit of W/m2. The fluence (UV dose) and radiant exposure (both J/m2 or mJ/cm2 or mW.s/cm2) are the counterparts of irradiance and fluence rate respectively, where power is replaced by energy. UV dose is the product of the average fluence rate acting on a micro-organism from all directions and the exposure time. Micro-organsisms are inactivated by the total fluence incident on them, that is, UV light from all directions. UV doses of around 40 mJ/cm2 (or 400 J/m2) are required to inactivate many bacteria and viruses. If chlorine is used as a second disinfectant, lower UV doses can be employed for inactivation of protozoa.

Table 13.2: UV dose requirementsa for Cryptosporidium and virus inactivation credits

Log Credit Cryptosporidium UV Dose (mJ/cm2)

Virusb UV Dose (mJ/cm2)

0.5 2 c 39 1.0 (90% removal) 3 58 1.5 4 79 2.0 (99% removal) 6 100 2.5 9 121 3.0 (99.9% removal) 12 143 3.5 NA 163 4.0 (99.99% removal) NA 186

a these doses are based on UV light at 254 nm as delivered by a low pressure mercury lamp. The doses can be applied to other lamps such as medium pressure through reactor validation testing.

b based on adenovirus studies (one of the most resistant micro-organisms)

Move table to back of Standards, and because ClO2 and O3 will have a table each this time, this will become Table 13.3 12 mJ/cm2 gives 3 logs for Crypto. But 38 or 40 needed for bacteria and viruses. So can use lower UV dose if using chlorine too


c rounded to whole numbers To be awarded the protozoa credits, a water supplier must demonstrate the effectiveness of the UV dose:

1. by using the results of a UV reactor validation test (from the manufacturer or supplier),

and 2. with on-line process monitoring of UV irradiance and

flow in each reactor or appliance, and

3. with on-line monitoring of turbidity of the raw water. For supplies serving a population of fewer than 5000, the on-line monitoring of irradiance and turbidity is optional. Refer to Section 3.3.2.2 for monitoring frequencies. The reactor validation test establishes the operating conditions under which a reactor can deliver a required UV dose. On-line monitoring of irradiance is used to demonstrate that the system maintains these validated operating conditions during routine use. Turbidity is measured because organisms can be shielded from the UV light by particulate matter. For any operational flow rate, the irradiance measured (W/m2) must be greater than that established by an independent bioassay process to correspond to a germicidal effectiveness, at that flow, equivalent to that of at least 400 J/m2 (40 mJ/cm2) at 253.7 nm. Each sensor used to monitor the UV irradiance level in the appliance must be calibrated annually to a traceable international standard. For any operational flow rate, the absorbance of the water shall be lower than that for which the unit has been certified, by an independent bioassay process, to be capable of producing, at that flow, a germicidal effectiveness equivalent to that of at least 400 J/m2 at 253.7 nm. Note that “the spectral attenuation (absorbance) of the water shall be lower” is synonymous with ‘the transmission (UVT) of the water shall be higher’. Absorbance = - log10(transmission), or A = -logT. At present, three strategies are commonly used for monitoring UV dose delivery:

1. UV intensity setpoint approach 2. UV intensity plus UVT setpoint approach 3. Calculated dose approach.

1. UV Intensity Setpoint Approach. Measurements made by

Always ‘must be greater than’ - or for 98% of the time (= 1/2 h/day)? And what if it only falls say to 200 J/m2 – that’s not as bad as being completely off? OK to not have to follow remedial action? So to go from 400 to nil J/m2 for 30 min is a transgression. But if it falls to 200 J/m2 is a transgression when it lasts for 30 min or 60 min? When drops to 200 J/m2 alarm should have rung already and operator alerted – fixed before public health problem?


the UV intensity sensor are used to control the UV reactor. The sensor is located in a position that allows it to properly respond to both changes in the output of the lamps and also UVT of the water. The UV intensity sensor output and the flow rate are used to monitor dose delivery. The setpoint for UV intensity over a range of flow rates is determined during validation. 2. UV Intensity plus UVT Setpoint Approach. The UV sensor is placed close to the lamp so that it only responds to changes in UV lamp output. UVT is monitored separately. For a specific flow rate, the UV intensity and UVT measurements are used to monitor dose delivery. The setpoints for UV intensity and UVT over a range of flow rates are determined during validation. 3. Calculated UV Dose Approach. The UV intensity sensor is placed close to the lamp. Flow rate, UVT, and UV intensity are all monitored, and the outputs are used to calculate UV dose via a validated computational algorithm developed by the UV reactor manufacturer. The transmission or absorbance of a sample (not filtered, not pH adjusted) is measured when using approach 2 or 3 because natural organic matter, iron and manganese can absorb UV light to such an extent that disinfection is impaired. Measurements of transmission or absorbance shall be made in a silica cell with a path length of at least 40 mm. Ideally the water being treated in the disinfection appliance should have a transmittance of at least 94% per cm at 253.7 nm. This is equivalent to a path-length dependent transmittance at 253.7 nm of:

94% measured in a 10mm cell 76% measured in a 40mm cell 71% measured in a 50mm cell 50% measured in a 100mm cell.

All the water passes through the UV disinfection unit, and for at least 95 percent of the time shall have a turbidity of less than 1.0 NTU, and no sample shall exceed 2 NTU. If treatment operates on/off, UV disinfected water must be recycled or wasted for the length of time that has been demonstrated to be necessary to reach optimum performance. Accurate flow measurement is essential to ensure that the UV reactors are operating within their validated conditions. For plants that rely on a common flowmeter (e.g. raw water flowmeter), the functionality of the flowmeter should be verified in conjunction with its intended use with the UV installation The accuracy and operating range of the

(Why 40 mm cell? Isn’t a 10 mm cell is OK with modern instruments?)


installation. The accuracy and operating range of the flowmeter should be verified and the availability of the necessary output signals from the meter should be confirmed. If pressure gauges are used to monitor the flow split between UV reactors, the calibration and installation of the pressure gauges should be verified as well. Alarm systems must be present that will warn of the failure of any component or factor that will affect the performance of the appliance. The radiation spectrum emitted by the lamps must be such that it does not cause the nitrite concentration to exceed 50 percent of the MAV. All water suppliers must keep a log book that records all relevant data that can be used for trouble-shooting, such as dates the lamps are installed, in and out of service, hours run, manually cleaned, flows through the reactors, power usage/UV dose. The equipment must be certified to supply the required minimum germicidal effectiveness by one of German DVGW Technical Standard W294, Austrian Standard oNORM M5873-1 or Austrian Standard oNORM M5873-2, or a standard formally recognised by the Ministry of Health as being equivalent. An appropriately accredited inspection body must have conducted the certification. When specifying the equipment being purchased, allowance should be made for when the water supply that will be treated has a lower temperature than that used in the validation tests. The equipment (including the sensor) to be installed must be identical (or validated as equivalent) to the equipment that was tested during the certification process, in terms of components that influence the relationship between dose delivery and measured irradiance. The installation of the equipment must be such that the certification is not invalidated. For example the stability of the flow should be at least as well developed as during the certification process. As a minimum, the following literature must be supplied with the equipment, written in English. • Fully detailed and illustrated manuals covering all aspects

of the technical specification and the operation and

Alarm with telemetry for plants > what population? Earlier version suggested zero increase in nitrite – why? Etc etc? - what else to be logged?


maintenance of the equipment. • Full details of the biodosimetric performance testing of the

equipment. • Full certification to one of the standards outlined above of

the ability of the equipment to deliver a germicidal effectiveness equivalent to that of at least 400 J/m2 at 253.7 nm.

• The uncertainty related to all measurement devices incorporated in the system.

• The spectral sensitivity of the sensor and the spectral emission of the lamp.

Some small supplies may not have automatic lamp cleaning facilities, in which case, UV disinfection will probably be unsuitable if the raw water contains more than 0.3 mg/L iron plus manganese, 120 mg/L total hardness, or 5 mg/L suspended solids. Without a sensor for monitoring irradiance, small supplies will be unsure of when their lamps are in need of attention, manual cleaning, or replacement.

Leave in DWSNZ (as criterion?) or move to Guidelines? So what do they do about that?

3.2.4 Criteria for demonstrating the security of

groundwater Groundwater is considered to be secure when it can be demonstrated not to be likely to be contaminated by pathogenic organisms because it is: • not directly affected by surface or climate influences, as is

demonstrated by compliance with criteria (a), (b) and (c) below

• abstracted via a secure well-head or similarly proven structure.

These are tougher now (just like DWSNZ 2000 were tougher than 1995). People have been asking that if they ‘won’ secure status with older criteria – what do they do now? Remain secure, or have to do all the new bits?


(a) E. coli is absent from the groundwater (prior to any disinfection or filtration):

Initially at least 38 samples shall be taken at approximately equal intervals over a period of at least 12 months at a point prior to any treatment, and tested for E. coli. If no E. coli are found in any samples (these 38 or any additional samples) taken for the purposes of demonstrating security, and conditions b and c below are met, the groundwater is classed as secure and monitoring can then revert to that required for secure groundwater in Table 3.1. Until all these criteria are met, the groundwater is not secure and monitoring should be undertaken as per the requirements for non-secure groundwater in Table 3.1.

And (b) the well head has been demonstrated to be secure by a qualified person.

Apart from a visual inspection of the surface, this should include such matters as confirming the depth and condition of the bore and casing.

And (c) the groundwater is not directly affected by surface or climate influences:

The lack of surface or climate influences can be demonstrated by the residence time in the aquifer or by the lack of significant and rapid shifts in determinands that are linked to surface effects as shown by:

(i) less than 0.005 percent of the water shall

have been present in the aquifer for less than one year (demonstrated by the tritium, CFC or sulphur hexafluoride (SF6) methods);

and

(ii) Variations in the concentrations of all of the

following determinands shall not exceed a coefficient of variation of more than:

3.0 percent in conductivity 4.0 percent in chloride concentration

(a) rewritten Take a secure supply, testing E coli quarterly Find a positive (so immediately non-secure?) If +ve was P/A, then enumerate E coli Or only non-secure if this repeat sample +ve too? Do daily E. coli and boil water until they get consec 3 clear days After 3 clear days – stop boiling water To be secure again: need 38 clear samples in 12 months If that’s OK, do monthly E. coli after 12 more months If they are OK, revert to quarterly – ie, 2 years and 50 E. colis after the failed test New sentence added in (b). SF6 added Coefficient of variation mentioned here, and


2.5 percent in nitrate concentration (standardised variance)

when measured:

at least monthly for one year, or once every two months for two years, or three monthly for three years.

Should the concentration of any one of these determinands be near the limit of detection, so that the coefficient of variation cannot be determined reliably, the results for that determinand may be disregarded at the discretion of the Medical Officer of Health. Once the well has been verified as secure, these determinands should be tested annually to check that the results remain within the range of concentrations found originally.

(d) An acceptable alternative to (c):

A verified hydrogeological model demonstrating that the bore is extracting from a secure aquifer will be acceptable. Verification must include demonstration that the parameters of the model (diffusivity, water age, etc.) are mutually consistent. If there are multiple bores and aquifers, from each aquifer the bore that the verified model indicates is the least likely to be secure must be subjected to criterion (a) above. Provided that no positive E. coli results are obtained, the security of the other bores from that aquifer will be presumed, but must be verified by collecting three samples for E. coli testing from each bore, collected at intervals of 1 month. This verification must be carried out for each individual aquifer.

Standardised variance mentioned here – they are not the same thing. Complicated? Will people understand this? Should we include a worked example – YES - still to be done. Better though: we could delete std’ised variance by limiting use of NO3 to when concs >0.1 N mg/L? Should we say the result of these calcs will be dependent on the quality of the data? So need very competent lab – some Min. of Health recognised labs may not be good enough! The point is: say 1 lab does 12 chlorides in 1 batch – should get 15 +/- 1 (say) so SD very low. BUT: if 12 monthly samples – done in 12 batches, maybe different analysis, different technician, even different lab! Now the answer is may be 15 +/- 5 so SD now large. To see how large – look at IANZ Watertest results! Or this could go in Guidelines as discussion (d) is a new section

3.2.4.1 Age determination

1) The bore shall have been pumped at full capacity for at least one hour prior to collection of the sample. This can only be done after the bore has been pump-

Section 3.2.4.1 is new


tested and developed.

2) The zero point used for age determination of the water shall be the time at which the water leaves the surface.

3) A full description of the procedure used to determine

the age of the water shall be provided by the consultant and/or analyst.

4) Age of groundwater reports should include full

details of the interpretation.

5) A confirmatory dating shall be carried out if specified to be necessary by the analyst.

In case of conflict of conclusions from chemical variability and water age data, the Medical Officer of Health will adjudicate. 3.2.4.2 Multiple bores serving a drinking-water supply Raw water for a drinking-water supply may come from a number of bores; if more than three this can lead to the requirement for a large number of E. coli samples to be collected if each bore is to be monitored separately. To demonstrate groundwater security in these circumstances, a hydrogeological model of the aquifer should be developed to establish: • whether the bores are drawing from a single aquifer or

more than one aquifer • the extent of the aquitard and whether it is continuous

throughout the whole bore field • the extent of variability of the chemical determinands

throughout the bore field. If it is established that the aquifer is single and homogeneous the number of bores to be sampled may be reduced, provided the sampling programme can be shown to give a representative picture. In designing the sampling programme, consideration should be given to the inclusion of bores at which drawdown, under certain circumstances, may introduce shallower, potentially contaminated water into the aquifer. It is not possible to give a general guide as to what the

#4 is new New section


requirements will be. This will need to be established by an independent investigator. The need to establish the integrity of each well head will remain.

3.3 Microbiological monitoring requirements In all records and documents used for demonstrating compliance with DWSNZ 2004 the sampling site identification details shall include the source, treatment plant or zone code listed in the Register of Community Drinking-Water Supplies in New Zealand. Priority 1 microbiological determinands E. coli and protozoa must be monitored regularly and frequently at the treatment plant. E. coli shall also be monitored in the network reticulation. The sampling frequencies specified are based on the population served, the type of water source and the treatment employed. Sampling frequencies do not apply when a supply has been shut down, for example, during holiday periods or for part-time supplies such as at ski fields. In community drinking-water supplies serving seasonal populations, additional samples must be collected during periods of peak population. The collection of these samples should commence at least two weeks before the seasonal increase commences so there is time to rectify any problems. Both the base and peak population are required for the Register of Community Drinking-Water Supplies in New Zealand. Procedures for sampling, sample preservation, storage and sample transport shall be agreed beforehand with the Ministry of Health recognised laboratory carrying out the analysis, except where special procedures are authorised for small remote drinking-water supplies or for analyses in the field.

New comment This (penult) paragraph has been expanded

3.3.1

Microbiological sampling sites

3.3.1.1 E. coli sampling site specifications

3.3.1.1.1 For E. coli compliance criterion 1 (water leaving


a treatment plant, refer Section 3.2.2.1)

• Samples for E. coli shall be taken from drinking-water

leaving the treatment plant, which may be at any point before the first consumer, but after the prescribed contact time.

• The FAC (and if relevant, the ClO2) sampling site shall be located at a point where the adequacy of the residual, the 30 minute minimum disinfection contact time and the pH can be demonstrated clearly (refer Guidelines), but before the first consumer. Note: if ClO2 is being used for protozoa inactivation, the contact time may be greater than 30 minutes.

• During maintenance such as tank cleaning, use can be made of the contact time in the mains prior to the first consumer. Peak flow must be used in the calculation.

• The UV irradiance in the appliance must be measured continuously by a sensor situated at the point where the irradiance was measured during the bioassay that was used to validate the performance of the appliance.

• The ozone sampling site shall be at a point in the contact tank that has been demonstrated to satisfy the C.t requirements. See also Section 3.3.3.5.

• For untreated water, sampling shall be at the point of entry into the distribution system.

3.3.1.1.2 For E. coli compliance criterion 2, (water in the network reticulation, refer Section 3.2.2.2)

Samples shall be taken from sites that represent all of the various conditions that exist in the distribution zones and provide a complete geographical coverage. To ensure that the network reticulation is evaluated adequately, sampling locations shall be planned carefully, based on the characteristics of each zone and what is known about age and state of repair of the various components, number of households served by sections of the distribution system etc. The sampling plan, shall be approved by the Medical Officer of Health or an accredited Ministry of Health designated officer, before the sampling programme commences and shall include: • fixed points such as pumping stations and reservoirs • random locations throughout the distribution zone • extremities of the distribution zone • taps representing the mains water being supplied to

houses


• extra sampling in the event of mains construction and maintenance.

Ministry of Health approval of the water supplier’s risk management plan for distribution zones satisfies this condition.

3.3.1.1.3 For E. coli compliance criterion 3, (water in the bulk water zone refer Section 3.2.2.3) Sampling shall be carried out at the transition point with the client distribution zone.

Does this mean bulk meters? See earlier note re how many and where

3.3.1.2 Protozoa (Giardia and Cryptosporidium) sampling site specifications Bank filtration (infiltration gallery): Turbidity samples shall be taken from water leaving the well or pump discharge from each well. Flocculation, coagulation, sedimentation, filtration: Turbidity samples shall be taken from water leaving each filter. Coagulation, direct filtration: Turbidity samples shall be taken from water leaving each filter. Diatomaceous earth filtration: Turbidity samples shall be taken from water leaving each filter. Slow sand filtration: Turbidity samples shall be taken from water leaving each filter. Membranes: The term module is used to refer to the various commercial membrane configurations and is defined as the smallest component of a membrane unit in which a specific membrane surface area is housed in a device with a filtrate outlet structure. A membrane unit is defined as a group of membrane modules that share common valving that allows the unit to be isolated from the rest of the system for the purpose of integrity testing or other maintenance.

Nearly all new details


Direct integrity tests are performed on each membrane unit. Continuous indirect integrity monitoring is usually performed on the effluent from each unit. Smaller plants may be able to sample individual modules. Bag filters: Challenge testing must be conducted on a full-scale filter element identical in material and construction to the filter elements proposed for use in full-scale treatment facilities. Compliance monitoring testing shall be performed on the effluent from each bag filter. A means of measuring the differential pressure between the feed water and filter effluent is required. Cartridge filters: Challenge testing shall be carried out on entire units, including filtration media, seals, filter housing, and other components integral to the process. Compliance monitoring testing shall be performed on the effluent from each cartridge. At large installations, a module may be sampled instead, but there must be some means of determining when individual cartridges in a module require attention. A means of measuring the differential pressure between the feed water and filter effluent is required. Combined filter performance: Continuous turbidity monitoring shall be carried out on the combined filtrates. Alternatively, a system can be used that calculates the mean turbidity from the readings of on-line turbidimeters on each filter. Individual filter performance: Continuous turbidity monitoring shall be carried out on the effluent from each filter. Second stage filtration: The turbidity of the water leaving each filter unit which together comprise the second filter shall be measured continuously. Chlorine dioxide: The ClO2 sampling site shall be located at a point where the adequacy of the residual and the minimum disinfection


contact time can be demonstrated clearly (refer Guidelines), but before the first consumer. The water temperature can be read at the same point, or in the raw water. The contact time is the average time, at peak flow, for the water to flow from the ClO2 dose point to the sampling point. Refer also to Section 3.3.1.1 (E. coli sampling sites) and Section 3.3.3.2 (Sampling requirements). Ozone: The site for sampling ozone shall be at a point in the contactor that has been demonstrated to satisfy the C.t requirements. See also Section 3.3.3.1 (E. coli sampling sites) and Section 3.3.3.5 (Sampling requirements).

The site may be established by tracer studies or by computational fluid dynamics, verified by direct measurement.

The on-line analyser must be placed at the site so defined. Grab sampling downstream of the contactor is inappropriate because by the time the water has emerged there will not normally be any free ozone residual. Grab samples for calibration purposes shall be taken from as close to the on-line analyser as possible.

UV: Water suppliers must monitor the flow rate, lamp outage, UV irradiance as measured by a UV intensity sensor, and turbidity, in or at each reactor or appliance. The UV sensor must be situated at the point where the irradiance was measured during the bioassay that was used to validate the performance of the appliance. If using the UV Intensity Setpoint Approach the transmission or absorbance can be measured in the raw water or in the water leaving the appliance. The test may be carried out either on-line or manually. If using the UV Intensity plus UVT Setpoint Approach, or the Calculated UV Dose Approach, the UVT must be measured on-line, usually as a component of the appliance. Secure groundwater: Provided the continued security of the groundwater shall be demonstrated, no further sampling is required.

3.3.2

Microbiological monitoring frequencies

3.3.2.1 E. coli monitoring frequencies

Table 3.1 has been completely


3.3.2.1.1 Sampling for E. coli in drinking-water leaving a treatment plant

redesigned

Table 3.1: Minimum sampling frequency for E. coli in drinking-water leaving a treatment plant for E. coli compliance criterion 1

Supply Type

Population served

Minimum samples per

calendar quarter

Maximum days

between samples

Minimum days of the week used

Undisinfected or non-secure 0 -- 500 13 13 5 Undisinfected or non-secure 501 -- 10,000 26 5 6 Undisinfected or non-secure more than 10,000 92 (ie, daily) 1 7 Chlorinated: partly monitored*

0 -- 500 3 40 3

Chlorinated: partly monitored 501 -- 10,000 13 13 5 Chlorinated: partly monitored more than 10,000 26 5 6 Ozonated: fully monitored 0 -- 500 3 40 3 Ozonated: fully monitored 501 -- 10,000 13 13 5 Ozonated: fully monitored more than 10,000 26 5 6 UV disinfection 0 -- 500 3 40 3 UV disinfection 501 -- 10,000 13 13 5 UV disinfection more than 10,000 26 5 6 Secure groundwater supplies regardless of

population 3 (or 1: see

notes) 40 3

Chlorinated*: fully monitored regardless of

population 0 - -

* and chlorine dioxide Notes to Table 3.1:

1. Refer to Section 3.3.2.1.2 for definitions of fully and partly monitored chlorination and undisinfected, and to Section 3.2.2.1 for information on E. coli criterion 1. 2. When the population increases, additional sampling shall be performed so that the sampling frequency is that specified in the Table for the population actually present.


3. Sampling must not always be on the same day of the week. It should be carried out on at least the number of different days shown (e.g. if 3 shown, then Monday, Wednesday, and Friday could be the days chosen). 4. For the monitoring frequencies designated for secure groundwater to apply, the security of the groundwater must have been demonstrated as specified in Section 3.2.4. 5. Monitoring requirements for secure groundwater supplies may be reduced to 1 sample per quarter after no E. coli have been detected in 12 consecutive months of sampling after the initial compliance tests are complete. 6. If E. coli is detected in a groundwater supply that has been classified as secure, that supply shall be monitored as a non-secure supply and must be fully reassessed to verify security. 7. The frequency of E. coli monitoring of drinking-water derived from secure groundwater shall be increased to at least weekly for four weeks when the source water quality may have changed, for example at peak abstraction during dry spells, following flooding of the recharge area, after an earthquake, or when chemical or physical water quality changes have occurred. 8. Where a treatment plant is supplied by both a secure groundwater and a non-secure source the supply shall be considered to be non-secure. 9. For supplies serving fewer than 500 people, samples prescribed to be taken from water leaving the treatment plant may be taken from the distribution zone instead if this is more convenient. This is on condition that the ‘treatment plant’ samples are taken from the first available tap after the treatment plant; sampling is done at the frequency specified in Table 3.1; and no E. coli are found. These samples are additional to those required for monitoring the distribution zone (Table 3.2) which are to be collected from points closer to the extremities of the distribution zone. 10. The samples prescribed to be taken from drinking-water leaving the treatment plant may be omitted for supplies to a single building (or a complex of not more than three buildings) that serve a population of fewer than 100 people and where, because of the short length of the reticulation system, contamination is unlikely to occur in the reticulation.

Re 6) see comment elsewhere – this is tougher response than other types of raw water/treatment process. That’s because being secure attracts large savings in monitoring. So increased sampling is in relation to demonstrated risk. 7) is clarified 10) is clarified – this is not the same as note 9 which just allows flexibility in the sample site for small water supplies.


11. In other than secure groundwater supplies, at least 10 samples per quarter need to be collected in order to satisfy the Ministry’s 95percent/95percent target. Water supplies serving a population of fewer than 500 and that do not practise fully monitored chlorination fall into this category. 12. Chlorine dioxide is not used presently in New Zealand but if it is in the future, its efficacy shall be considered equivalent to chlorination.

11) is clarified Added #12

3.3.2.1.2 Disinfection monitoring frequency in water leaving a treatment

plant

For other than continuously monitored systems, the frequency of monitoring shall be increased if there is a flood, emergency operation, interruption to the supply system, or when other circumstances may give rise to an increased risk of faecal contamination. (a) Chlorination and disinfection with chlorine dioxide: Refer to Section 3.2.2.1 for definitions of continuously monitored chlorination, partly monitored chlorination, and undisinfected, and for other details of E. coli compliance criterion 1. Supplies serving more than 10,000 people shall be monitored continuously. E. coli sampling is optional. FAC in partly monitored supplies must be measured at least:

weekly at plants supplying up 500 people 3 times at week for plants supplying 501-5,000 people twice daily for plants supplying more than 5000.

Table 3.1 describes the different E. coli sampling frequencies for partly monitored chlorination and undisinfected supplies. Plants disinfecting with chlorine dioxide but that do not have continuous monitoring are considered to be undisinfected for the purposes of E. coli sampling. (b) Ozone and UV disinfection: The measurement of chemical residual or UV irradiance shall be made on-line. If the ozone or UV disinfection process is not continuous, or does not have alarms to indicate faults, or is not continuously monitored, the water leaving the treatment plant shall be sampled for E. coli at the undisinfected (Table 3.1) rate.

Section expanded ClO2 included Ozone and UV added

3.3.2.1.3 pH monitoring frequency in water leaving a treatment plant Testing regime expanded


In all plants that chlorinate and serve more than 10,000 people, the pH of water leaving the contact tank shall be measured continuously. Monthly manual pH testing will be acceptable if it has been demonstrated that results found in weekly sampling for 12 months fall within a 0.3 pH range. For smaller plants that chlorinate but do not have a continuous pH monitor, a sample shall be tested for pH at the same time FAC samples are tested. At plants that do not chlorinate, pH shall be measured at the same time as the E. coli sample is collected.

3.3.2.1.4 Turbidity monitoring frequency in water leaving a treatment plant

In all plants that chlorinate and serve more than 10,000 people, the turbidity of the water leaving the contact tank shall be measured continuously; monthly manual turbidity testing will be acceptable if it has been demonstrated that results found in weekly sampling for 12 months do not vary by more than 5% from the mean. The mean shall be <0.5 NTU. Daily sampling will be required after heavy rain or earthquake.

At plants that do not have continuous on-line turbidity measurement, for compliance with E. coli Criterion 1, turbidity shall be measured at the same time as the E. coli sample is collected. At plants that do not chlorinate, turbidity shall be measured at the same time as the E. coli sample is collected.

Testing regime expanded

3.3.2.1.5 Sampling frequency for E. coli compliance criterion 2A, in a distribution zone

The minimum sampling frequencies for E. coli in drinking-water in the distribution zones, when E. coli monitoring is not partially substituted by FAC monitoring, are given in Table 3.2a. Testing shall be carried out on different days throughout the week as shown in Table 3.2b. While for small communities the minimum sampling frequency is set at three per calendar quarter, all drinking-water supplies should ideally be monitored at least ten times a quarter to provide improved statistical confidence in the results.

3.3.2.1.6 Sampling frequency for E. coli compliance criterion 2B in a distribution zone (E.coli monitoring partly substituted by FAC)


In cases where the FAC leaving a treatment plant does not fall below the equivalent of 0.2 mg/L at pH 8.0 for more than 2 percent of the time in any one day, and the turbidity is less than 0.5 NTU for 95% of the time, FAC monitoring in the distribution system may be substituted for up to 75 percent of the E. coli tests required provided that: or, in cases where the sum of the chlorine dioxide (as ClO2) plus FAC leaving a treatment plant does not fall below 0.2 mg/L for more than 2 percent of the time in any one day, and the turbidity is less than 0.5 NTU for 95% of the time, disinfectant monitoring in the distribution system may be substituted for up to 75 percent of the E. coli tests required provided that: • at least four FAC (and if relevant, ClO2 plus FAC) tests are performed

for each E. coli test omitted • pH is measured at the time of FAC (and if relevant, ClO2 plus FAC)

sampling • chlorine monitoring (and if relevant, ClO2 plus FAC) is carried out daily. FAC results at pH greater than 8.0 are to be calculated as their FAC equivalents at pH 8.0. The number of FAC (and if relevant, ClO2 plus FAC) and E. coli tests needed is calculated from the formulae:

Added ClO2

Number of FAC tests = percent / 100 of E. coli tests replaced x (number of E. coli tests required

by Table 3.2a column 2 if no substitution with FAC testing is done) x 4 The number of E. coli tests

= percent / 100) of E. coli tests replaced x (number of E. coli tests required by Table 3.2a column 2 if no substitution with FAC testing is done) (this is the number shown in Table 3.2a column 3).

EXAMPLE (for 75 percent

FAC = 0.75 x (number of E. coli tests required by Table 3.2a if no substitution with FAC testing is done) x 4

replacement) E. coli = (1-0.75 E. coli tests replaced) x (number of E. coli tests required by Table3.2 if no substitution with FAC testing is done).

The sampling protocols resulting from the application of these formulae are summarised in Table

3.2a. The results of calculations for a range of distribution zone populations are rounded-up to the nearest whole number. Any interpolations to the table using the formulae are to be rounded up on a similar basis. E. coli monitoring shall be carried out regularly throughout the month on different days throughout the week, including weekends. The FAC tests are to be conducted daily. The maximum number of days between collection of E. coli samples and the minimum number of days of the week on which samples shall be taken is shown in Table 3.2b.


Table 3.2a: Minimum sampling frequency for E. coli in a distribution system

Population serviced 1 Minimum number of E. coli samples per quarter where no FAC substitution 2, 4

Minimum number of samples per quarter where FAC testing substitutes 75 percent of E. coli testing 3

E. coli FAC4 up to 500 3 Not Applicable Not Applicable 501 - 5,000 13 Not Applicable Not Applicable 5,001 - 10,000 16 Not Applicable Not Applicable 10,001 - 15,000 19 Not Applicable Not Applicable 15,001 - 20,000 22 Not Applicable Not Applicable 20,001 - 25,000 25 Not Applicable Not Applicable 25,001 - 30,000 28 Not Applicable Not Applicable 30,001 - 35,000 31 8 93 35,001 - 40,000 34 9 102 40,001 - 45,000 37 10 111 45,001 - 50,000 40 10 120 50,001 - 55,000 43 11 129 55,001 - 60,000 46 12 138 60,001 - 65,000 49 13 147 65,001 - 70,000 52 13 156 70,001 - 75,000 55 14 165 75,001 - 80,000 58 15 174 80,001 - 85,000 61 16 183 85,001 - 90,000 64 16 192 90,001 - 95,000 67 17 201 95,001 - 100,000 70 18 210 100,001 - 110,000 73 19 219 110,001 - 120,000 76 19 228 120,001 - 130,000 79 20 237 130,001 - 140,000 82 21 246 140,001 - 150,000 85 22 255 150,001 - 160,000 88 22 264 160,001 - 170,000 91 23 273 170,001 - 180,000 94 24 282 180,001 - 190,000 97 25 291 190,001 - 200,000 100 25 300 etc Notes to Table 3.2a:

1 When the population increases, additional sampling will be performed so that the sampling frequency is that specified in the Table for the population actually present. 2 Results of the calculations rounded up to nearest whole number. 3 Testing is to be distributed evenly throughout the quarter, carried out on different days of the week and shall give a representative geographical coverage of the distribution system (see microbiological site specification in Section 3.3.1.1.2 for details). Calendar quarters are to be used (Jan - Mar, Apr-Jun etc.).


4 A turbidity test is required when collecting a sample for E. coli or FAC testing, and a pH test is required when testing for FAC. For water supplies using chlorine dioxide, read chlorine dioxide as ClO2 plus FAC in place of FAC. Note also that monitoring additional to that required for compliance monitoring shall be carried out after installation of new mains or following connection or repair in the network reticulation. This monitoring is to be carried out within 12 hours of completion of the connection, repair, or restoration of flow.

Turbidity added in 4 Added ClO2 Last paragraph added


Table 3.2b: E. coli sample distribution

Number of E. coli samples per quarter

Maximum interval between E. coli samples (days) 1

Minimum number of days of the week used

1 -- 2 90 1

3 -- 7 45 3

8 -- 12 17 4

13 -- 18 11 5

19 -- 21 8 6

22 -- 30 6 7

31 -- 36 5 7

37 -- 45 4 7

46 -- 60 3 7

61 -- 92 2 7

over 92 1 7

Note to Table 3.2b:

The maximum interval between samples is determined by the number of E. coli samples, not by the size of the population. See the example below. Example: If the zone population is 68,155 and 75 percent replacement is used: without replacement 52 E. coli samples are required per quarter (Table 3.2a) with 75 percent replacement of E. coli by FAC this requires: 13 E. coli per calendar quarter (i.e. 52 x 25 percent, rounded up if necessary) and 156 FAC per quarter (i.e. 52 x 75 percent x 4) which, in accordance with Table 3.2b, have a maximum sampling interval of 11 days and are sampled over five different days of the week. For water supplies using chlorine dioxide, read chlorine dioxide as ClO2 plus FAC in place of FAC.


3.3.2.1.7 Sampling frequency for E. coli compliance criterion 3A in a bulk distribution zone Table 3.2c specifies the minimum number of samples to be tested per calendar quarter from each transition point from a bulk distribution zone depending on the population served by the client supply serviced by that transition point.

Table 3.2c: Minimum sampling frequency for E. coli in a bulk distribution zone

Nominal Population

Minimum samples per calendar

quarter

Maximum days between samples

Minimum days of the week used

0 - 500 13 13 5 501 – 10,000 26 5 6 > 10,000 92 1 7

Table 3.4: Minimum measurement frequency and reporting period for turbidity in water leaving each filter, for protozoa compliance for all treatment plants using any form of filtration

Notes to Table 3.4: 1. Reporting period is the length of time over which 95 percent of measurements shall meet

the compliance criteria. Reporting periods are sequential. Continuous data records shall not be more than one minute apart. These should be compressed using a procedure that preserves accuracy of raw data, and reported as percent of time value exceeded on a reporting period basis.

2. Continuous monitoring is highly desirable, but manual monitoring is acceptable for small plants.

Population Served Minimum Measurement Frequency Reporting Period 1 Continuous Manual 2 (per

Calendar Quarter) Continuous Manual

More than 10,000 on each filter N/A one day N/A 5,001 - 10,000 2 at least one

turbidimeter to every 2 filters 3

184 (twice a day)

one day 2 weeks

501 - 5,000 at least one turbidimeter to every 4 filters 3

92 (daily)

one day 4 weeks

101 - 500 at least one turbidimeter to every 4 filters3

26 (twice weekly)

one day 3 months

100 or fewer N/A 13 (weekly)

N/A 6 months


3. Each filter shall be sampled sequentially (no blending) for five minutes (two filters per turbidimeter) or for two minutes (four filters per turbidimeter) before switching to the next filter. Sample lines should be short and sample flows high enough to prevent adsorption or precipitation.

Table 3.3: Minimum measurement frequency for protozoa compliance for bag and cartridge filters

Population Served Minimum Measurement Frequency per Calendar Quarter

Minimum Measurement Frequency per Calendar Quarter *

On-line particle counting MPA Over 10,000 Continuous and Weekly 5,001 – 10,000 Continuous or 92 (i.e. daily) 501 – 5,000 Continuous or 46 500 or less Continuous or 13 * Notes to Table 3.3: Additional testing shall be carried out immediately after any adjustment to a filter (eg, change

of cartridge) until the effluent particle size distribution has stabilised. Trials are required to determine the minimum diatomaceous earth precoat thickness that will reliably remove protozoa before switching from recycle to production.

3.3.2.1.8 Sampling frequency for E. coli compliance criterion 3B in a bulk distribution zone FAC (and if relevant, ClO2) must be monitored continuously and turbidity at the same frequency as would be required for E. coli in Table 3.2.

This means that they are continuously monitoring FAC in the bulk mains, so what about pH too? If so, where (at bulk meter?), and at how many of them?

3.3.2.2 Monitoring frequency for protozoa (Giardia

and Cryptosporidium) Bank filtration (infiltration gallery): • supplies that serve a population greater than 10,000

must monitor turbidity continuously • supplies that serve a population 5000-10,000 must

measure turbidity once a day • supplies that serve a population fewer than 5000 must

measure turbidity weekly. Flocculation, coagulation, sedimentation, filtration: Turbidity is used as a measure of the efficacy of the


coagulation plus filtration process. For protozoa compliance monitoring of water leaving the treatment plant, turbidity shall be measured at each filter at the frequencies specified in Table 3.4 as a minimum and shall be measured continuously where possible. Continuous monitoring is required for water supplies servicing populations greater than 10,000; the monitor must give an alarm to duty staff in the event of a transgression. Coagulation, direct filtration: As for flocculation, coagulation, sedimentation, filtration. Diatomaceous earth filtration: As for flocculation, coagulation, sedimentation, filtration. Additional testing shall be carried out determine the minimum diatomaceous earth precoat thickness that will reliably remove protozoa in different raw water conditions, before switching from the recycle to production mode. Slow sand filtration: As for flocculation, coagulation, sedimentation, filtration. The water must be recycled after sand scraping or replacement until additional testing has demonstrated that the treatment process has returned to full efficiency. Additional monitoring (including for E. coli) will also be required if the water temperature falls to 6°C so temperature should be monitored daily (during the relevant seasons) in areas where this could occur. Membranes: Challenge tests are performed on the complete module by the manufacturer or supplier before the water supplier purchases the plant. They may also be required by the Medical Officer of Health to check that a treatment problem has been rectified. Direct integrity tests are performed on each membrane unit at least daily. Continuous indirect integrity monitoring (turbidity) is usually performed on the effluent from each unit. Smaller plants may be able to sample individual modules. Bag filters: Challenge tests are performed by the manufacturer or supplier before the water supplier purchases the plant. They may also be required by the Medical Officer of Health to check that a treatment problem has been rectified

Alarm added to second paragraph Added a row to Table 3.4 for plants <100 – and increased rate for 100-500 - OK? 2/day for 2 wks = 28 tests; daily for 4 wks = 28; 1 per wk for 6 mths = 26. 5% of 26 = 1.3, so one transgression is OK, 2 is non-complying So after first transgression – wait for second in same reporting period – if it comes – they have failed to comply (for that whole period), even if the two events were up to 26 weeks apart. Is this a problem? Obviously yes – so probably need to report every transgression – if not report, then at least respond as though it were one – discuss. Note 1:Continuous monitoring will be defined better soon Continuous monitoring by 2008 for >5000


rectified. The differential pressure between the feed water and filtrate shall be monitored (monthly reporting period): • continuously at treatment plants servicing more than

100 people • at least daily for smaller units. The flow through each filter or filter module shall be monitored (monthly reporting period): • continuously at treatment plants servicing more than

500 people • by cumulative recording turbine meter (read daily) for

smaller plants. Compliance monitoring testing for turbidity shall be performed on the effluent from each bag filter at the frequency that appears in Table 3.4 (now Table 3.5 I think). Additional testing shall be carried out immediately after cartridge replacement to confirm that it is performing satisfactorily. Cartridge filters: As for bag filtration. Combined filter performance: Turbidity monitoring of the combined filter effluents shall be continuous. Alternatively, a system can be used that calculates the mean turbidity from the readings of on-line turbidimeters on each filter. Individual filter performance: Turbidity monitoring shall be carried out on the effluent from each filter continuously. Second stage filtration: The turbidity of the water leaving each filter unit which together comprise the second filter shall be measured continuously. Chlorine dioxide: All water supplies using chlorine dioxide shall monitor the residual chlorine dioxide continuously. The monitor must give an alarm to duty staff in the event of a transgression. The peak daily flow or demand (whichever is the more appropriate) shall be used for calculations.

Table 3.3 is no longer relevant or acceptable – so Table numbering changes AGAIN! Wait though to see if particle counting finds its way back into DWSNZ 2004.


Temperature shall be measured daily. Turbidity is monitored as per the E. coli compliance criterion 1. Ozone: All water supplies using ozone shall monitor the residual ozone continuously. The monitor must give an alarm to duty staff in the event of a transgression. The peak hourly flow shall be used for C.t calculations which shall be made on a daily basis. If the water temperature has been shown to vary by less than 2° C in 24 hours over a month in the summer, it shall be measured daily; if it is more variable, it shall be measured at least every 4 hours. The turbidity of the raw water shall be measured continuously. But if 95 percent of daily samples over a year are less than 1.0 NTU it shall be measured as per Table 3.4: Minimum measurement frequency and reporting period for turbidity in water leaving each filter, for protozoa compliance for all treatment plants using any form of filtration. UV: All water supplies using UV disinfection shall monitor the irradiance in each reactor continuously. (This is what is said in Criteria section: For supplies serving a population of fewer than 5000, the on-line monitoring of irradiance and turbidity is optional.) The turbidity of the raw water shall be measured continuously. But if 95 percent of daily samples over a year are less than 1.0 NTU it shall be measured as per Table 3.4: Minimum measurement frequency and reporting period for turbidity in water leaving each filter, for protozoa compliance for all treatment plants using any form of filtration. The flow rate shall be monitored continuously through each reactor. If using the UV Intensity plus UVT Setpoint Approach, or the Calculated UV Dose Approach, the UVT must be measured continuously by an on-line instrument.

First paragraph: small supplies probably will not have continuous irradiance monitors – so what do they do instead to demonstrate that the UV system is still working? Massively overdose? Or be non-complying? Fourth paragraph - but not for small supplies – how do they do it? Estimate flow based on pressure


If using the UV Intensity Setpoint Approach the transmission or absorbance can be measured manually. If 95 percent of weekly samples for a year are not less than 94% transmission (or do not exceed an absorbance of 0.027) based on a 10 mm cell, then samples can be tested monthly. Secure groundwater: The water shall comply with the specification of secure groundwater (Section 3.2.4). No further monitoring is required to demonstrate compliance with the protozoa criterion.

gauges for example – some may record, some may be read only? Should supplies without autoclean have to monitor something – what, and how often? Like TH and Fe, or NTU more often – and if so, what trigger points, or what = a transgression?

3.3.2.3 Direct measurement of protozoa

Direct measurement of protozoa is not used as a compliance criterion at present. Protozoa enumeration and viability tests may be used to confirm suspected presence of protozoa but failure to find protozoa does not demonstrate that drinking-water is free from protozoa. The frequency of testing is at the water supplier's discretion, unless otherwise directed by the Medical Officer of Health.

3.3.2.4 Monitoring frequencies for Priority 2c micro-organisms

Micro-organisms designated as Priority 2c shall be monitored at a frequency specified by the Medical Officer of Health.

3.3.3 Microbiological sampling requirements

3.3.3.1 E. coli

Procedures for sampling, sample preservation, storage and sample transport shall be agreed beforehand with the Ministry of Health recognised laboratory carrying out the analysis, except where special procedures are authorised for small remote drinking-water supplies or for analyses in the field (see Section 3.3.4). In cases where special procedures are authorised, samples shall be collected aseptically, in sterile bottles, using thiosulphate to dechlorinate the sample if necessary. Ideally, testing should commence within six hours of sample collection and it must never be

Std Meths says 30 hrs – leave it at 24? USEPA says 24 hr holding time OK


six hours of sample collection and it must never be delayed more than 24 hours after collection. Samples shall be transferred to the laboratory in a cool, dark container. If delivery time exceeds one hour, samples shall be maintained at no more than 10° C but shall not be frozen. If samples cannot be processed immediately upon arrival in the laboratory they should be stored in a refrigerator.

Changed this to 10° C to align with Std Meths – and USEPA

3.3.3.2 Free available chlorine (FAC) and chlorine dioxide

Measurement of FAC and the residual chlorine dioxide must be made in the field.

3.3.3.3 pH

The analysis of pH should be carried out as soon as possible after sampling, with a maximum delay of 4 hours unless there is documented evidence that the pH of the water will not change during the interval between sampling and analysis. This latter remark applies particularly to groundwater samples.

Groundwater remark added

3.3.3.4 Turbidity and particle counting

Measurement of turbidity and particle counts preferably should be made on-site using a continuous monitor. Otherwise a portable instrument or a laboratory remote from the water treatment plant may be used. In the latter case, samples shall be transported to the laboratory as soon as possible, at the latest within 36 hours.

3.3.3.5 Ozone

The ozone residual in water decays rapidly. The half-life ranges from less than 1 minute to more than 20 minutes. Ozone contactors are sealed vessels with sample lines that penetrate the walls or roof structure of the contactor. The detention times in the sample lines should be as short as possible in order to minimize ozone residual decay in the sample lines. Preferably less than 10 percent of the ozone should decay in the sample lines.

Rewritten

3.3.3.6 UV

The operating UV irradiance level shall be monitored continuously by a certified sensor sited at the point at which the irradiance was measured during the

New


performance validation of the appliance, which shall be at the remotest point in the appliance from the UV source. Manual UV absorbance or transmission tests shall be conducted on samples that have not been filtered and not been pH adjusted.

3.3.4

Microbiological analytical requirements

Maybe add a brief reference to Direct Integrity Tests (mf)? And maybe a bit about Challenge tests for MF, Bag/Cartridge? Delete particle counting if it doesn’t play a role in DWSNZ.

3.3.4.1 Introduction

Laboratories recognised by the Ministry of Health shall be used for all analyses carried out for the purpose of assessing compliance with these Standards, except where special procedures are authorised for small, remote drinking-water supplies or for analyses in the field. The competence of field analysts shall be verified by an assessor accredited for the purpose. The limit of detection, precision and uncertainty of test methods should be included on all analytical reports. Refer to Part 1000 of Standard Methods for the specification for IANZ accreditation. More detailed information about analytical requirements for E. coli, free available chlorine, pH, turbidity and particle size is given in the Guidelines for Drinking-Water Quality Management in New Zealand.

This new paragraph added Will it be called the 2004 edition?

3.3.4.2 E.coli

For convenience, Presence/Absence tests or other rapid methods for E.coli that are acceptable to the Ministry of Health for the purpose may be used for routine monitoring. However, should a positive result be obtained, a second sample must be collected within 12 hours (wherever possible) from the same sample site and E. coli enumerated by a laboratory recognised by the Ministry of Health. The Ministry of Health maintains a register of laboratories recognised for this purpose (see Section 2.1). The remedial procedures following a transgression are specified in Section 3.4 apply.


3.3.4.3 Free available chlorine

Measurement of FAC must be made in the field or on-line as soon as the sample has been collected. Suitable FAC field methods are DPD tablets or powder in foil. Amperometric techniques may also be used. All methods should be calibrated against the referee method, APHA 4500 Cl F DPD, ferrous ammonium sulphate titrimetric method, at least once every six months. When field tests involve colorimetric interpretation, all persons who have undertaken FAC tests shall undertake the calibration exercise against the referee method. The identity of the person performing each field test shall be recorded. The analyst making the measurement should be familiar with both the referee and field methods and possible causes of inaccuracy. The reading from a continuously monitoring chlorine or chlorine dioxide analyser shall be calibrated against a grab sample from the water (tested by field method) at least once a week at the point where it is continuously monitored. The reading on the continuous monitor at the time of taking the grab sample shall be noted. Once every six months the grab sample is to be split and an interlaboratory calibration with a Ministry of Health recognised laboratory carried out, on-site. The instrument shall be recalibrated if the output as measured from a grab sample and that logged for compliance purposes varies by more than 0.1 mg/L. This reading may be different from the reading observed on the instrument display. The precautions specified in the Guidelines should be followed.

Clarified That means: 0.23 and 0.32 mg/L (varies by 0.09) are OK But 0.23 and 0.34 (varies by 0.11) - require action That OK?

3.3.4.4 pH

The pH electrode should be calibrated before each set of measurements is made, and manufacturer’s instructions should be followed for the storage of the electrode when not in use. Calibration solutions used should be prepared by an analytical laboratory using the formulations given in the above instructions, or purchased as a certified solution from a chemical manufacturing company. Readings from continuous pH monitors should be


calibrated on-site against a calibrated hand-held pH electrode once every four weeks. Readings should agree to within 0.10 pH units. The pH meter should be calibrated every day/time with pH 7 and 9 (or 10) buffers (most samples will fall in this range). Also, pH 4 buffer should be included once a month to check linearity over a wider range. For potable waters, many of which in New Zealand are unbuffered, the time taken for the pH to return from measuring the pH 9 buffer to reading the pH of a water supply sample shall be recorded every four weeks. If this has become slow, then the electrode needs attention or is unsuitable. Meters being used for potable water require special thin glass electrodes to work properly on unbuffered waters. "Robust" electrodes are not suitable.

Some rewritten

3.3.4.5 Turbidity

On-line turbidimeters and particle counters should be installed in such a manner that air entrainment, temperature changes, long sample lines, algal growth, low velocities or changing velocities, are minimised. To operate a turbidimeter with confidence at around the 0.10 NTU level requires an instrument with a limit of detection better than 0.02 NTU and requires the use of sophisticated calibration techniques. Continuous monitors must be calibrated at least as frequently as recommended by the manufacturer. Continuous turbidimeters that do not use the nephelometric technique need to be calibrated in NTU units. The reading from the turbidimeter shall be calibrated against a grab sample from the water leaving the filter at the time of the turbidimeter reading at least once a week. Calibration readings should agree to within 0.05 NTU for test samples with turbidity up to 0.50 NTU, and to within 0.10 NTU for test samples above 0.50 NTU. Field and laboratory turbidimeters should be checked against a secondary standard that is traceable to a primary standard at least once every day they are used. The test technique shall be checked and calibrated

Tolerance added. if turbidity should be <0.10 NTU, when has a transgression occurred? Can’t be at 0.11 NTU. If using the 0.05 ‘allowance’, then transgression will occur at 0.16 NTU. Answer to this is probably best in Criteria Section?


against a primary standard at least six-monthly, preferably by a Ministry of Health recognised laboratory. See APHA 2130. Note that different manufacturers’ instruments claiming to comply with the referee method may not give the same results with the same water, even when each instrument has been calibrated in accordance with the manufacturer’s manual. Refer to the Guidelines for Drinking-Water Quality Management for New Zealand.

3.3.4.6 Particle counting

The reading from the instrument shall be calibrated against a grab sample from the water leaving the filter at the time of the instrument reading at least once a week. Once every six months the grab sample is to be split and an interlaboratory calibration with a Ministry of Health recognised laboratory carried out.

So far, particle counting not used in DWSNZ 2004. What tolerance is OK?

3.3.4.7 Ozone

On-line ozone analysers shall be calibrated at least once a week against a grab sample from the water at the point where it is continuously monitored. Generally, probe-type monitor readings tend to drift downward over time due to weakening of the electrolyte solution. Checking by a Ministry of Health recognised laboratory is preferred, but if the analyser is checked using a field test method, the field test method should be calibrated against the referee method (Indigo Method, Standard Methods 4500-Ozone, 20th Edition) at least once every six months by a Ministry of Health recognised laboratory. Normally the indigo method is used for calibration. Because ozone is so reactive it is necessary to cross-check multiple samples, preferably five, using the following procedure:

1. Obtain an analyser reading while the grab sample is being collected, with an appropriate time delay to allow for flow time in the sample line.

2. Promptly measure the ozone residual concentration in the grab sample, using the indigo method.

3. Calculate the average grab sample ozone residual value and the average analyzer

Should this re testing go in Guidelines instead?


residual value and the average analyzer ozone residual value.

4. Compare the average of the on-line and grab-sample results. The average of the on-line analyser should not deviate more than 10% or 0.05 mg/L (whichever is larger) from the grab sample average. If it is more than this, adjust the meter reading as per the manufacturer’s instructions. It is important that the on-line analyser not record more than 10% or 0.05 mg/L greater than the grab samples. A negative deviation bias, while not affecting public health, may also be useful as an indication of a malfunctioning unit.

5. Allow the analyser to stabilize for a period of 30 minutes after adjusting the meter reading and repeat steps 1 to 4 until the difference calculated in step 4 is less than 10% of the grab sample average or less than 0.05 mg/L.

The indigo method assumes that high-purity reagents are used. Since the publication of the 20th edition of Standard Methods, several reports have been published discussing a potential biasing where reported results could be significantly low. The potential biasing involves the value of the sensitivity factor used in the calculation. This should be addressed in the next edition of Standard Methods; if not, an alternative analytical technique will appear in the Guidelines.

Tolerance addressed

3.3.4.8 Chlorine dioxide

Continuously monitoring ClO2 analysers shall be calibrated against a grab sample of the water at the point where it is continuously monitored, at least once a week. Checking by a Ministry of Health recognised laboratory is preferred, but if the analyser is checked using a field test method, the field test method should be calibrated against the referee method at least once every six months by a Ministry of Health recognised laboratory. The referee method can be either 4500-ClO2 C. Amperometric method 1, or 4500-ClO2 D. DPD method. If the on-line analyser and manual test differ by more than 10% or 0.10 mg/L (whichever is larger), confirm with another check. If they still differ by more than

What tolerance is OK?


10% or 0.10 mg/L (whichever is larger), adjust the on-line instrument as per the manufacturer’s instructions, allow the analyser to stabilize for a period of 30 minutes, and recheck.

3.3.4.9 UV disinfection

Most UV installations provide a reference and a duty sensor. These sensors must have been calibrated to a traceable international standard. The operating UV irradiance level shall be monitored continuously by the duty sensor, and must be calibrated against the reference sensor at least monthly. The reference sensor must be checked, and recalibrated if necessary, by a certifier accredited for the purpose to NZS/ISO/IEC 17025 Standards (or equivalent) at least annually. If the reference sensor is found to be out of calibration, the calibration interval shall be shortened. On-line UV transmission meters require calibration at least as frequently as recommended by the manufacturer, using a buffer solution of known absorbance, or compared with a bench top instrument. Once every six months a grab sample is to be split and an interlaboratory calibration with a Ministry of Health recognised laboratory carried out.

Tolerance? Abs (or T) normally quite accurate. Should be +/- 1% T

3.4 Remedial action to be taken when transgression of a microbiological MAV occurs

When a transgression of the microbiological Standards occurs there must be an immediate response. The action to be taken in the different circumstances is summarised below. This should be documented in each case and the remedial action record provided to the Medical Officer of Health. Samples collected as a result of a transgression are not counted as part of the routine compliance monitoring programme; this resumes once the cause of the transgression has been remedied.

Added

3.4.1

Transgressions involving E. coli When a positive result has been obtained using a


Presence/Absence, faecal or total coliform test carried out by field or works staff, a second sample must be collected within 12 hours (wherever possible) and E. coli enumerated by a Ministry of Health recognised laboratory. If the compliance monitoring test directly enumerates E. coli, and a transgression occurs, begin the remedial work at the first “Action Box” in Figure 3.2.

Added

3.4.1.1 Response to a transgression involving E. coli criterion 1 for drinking-water leaving a treatment plant Because contaminated drinking-water leaving the treatment plant can affect the whole community, immediate action is required if a positive E. coli test or a failure in the requirements of FAC monitoring, occurs. Refer to Section 3.2.2.1 and Figure 3.2. Corrective action to be taken (which should be documented in the Risk Management Plan) includes inspecting the plant and equipment; increasing the disinfection if necessary; if no fault is found at the plant, inspecting the water supply catchment; increasing sampling; and advising the Medical Officer of Health. Any remedial action indicated shall be applied promptly, and be reported fully. If repeat samples continue to be positive, the Medical Officer of Health shall be consulted and remedial action, such as the issue of a ‘Boil Water’ notice, shall be carried out if considered necessary. Corrective action shall be intensified. Corrective and remedial action shall be continued until samples have tested free of E. coli for three successive days. If one of the three further days sampling called for coincides with a scheduled compliance sample, only one sample need be taken on that day (paragraph 2 of Section 3.4). If E. coli is detected in a groundwater supply that has been classified as secure, that supply shall be monitored as a non-secure supply and must be fully reassessed to verify security. Figure 3.2: Response to E. coli contamination of drinking-water leaving a treatment plant. Insert Figure 3.2 here – it needs modifying

The period of a transgression shall not exceed 3 days [treatment plant] or 5 days [distribution zone] from the date that the positive sample was collected. After this period it becomes a new transgression. This is really more to do with Grading – include in DWSNZ?


3.4.1.2 Response to transgression involving E.

coli in drinking-water in a distribution system or bulk distribution zone 3.4.1.2.1 E. coli present in a distribution system or bulk distribution zone Figure 3.3 summarises the degree of response depending on the number of E. coli found. The response outlined in Figure 3.3 also applies to total coliforms or faecal coliforms if these are used in place of E. coli in the monitoring programme. When a positive result has been obtained by field or works staff using a Presence/Absence or equivalent test, a second sample must be collected within 12 hours (wherever possible) for enumeration of E. coli by a Ministry of Health recognised laboratory. Figure 3.3: Response to E. coli contamination of a drinking-water supply distribution zone Insert Figure 3.3 here– it needs modifying Figure 3.3 summarises the successive stages of response, reflecting the fact that a distribution zone sample containing ten or more E. coli per 100 mL, or more than one sample testing positive, must be considered to be seriously contaminated. The response (which should be documented in the Risk Management Plan) includes resampling to confirm whether E. coli is still present, ascertaining the source of the contamination and increasing the disinfectant dose if necessary. For a heavy contamination (more than 10 E. coli per 100 mL) or a second positive result, the Medical Officer of Health must be advised. Daily sampling should be continued until three successive clear days have occurred.

Included bulk distribution zone Reference to Risk Management Plan added, here and below.

3.4.1.2.2 Free available chlorine too low in a distribution system or bulk distribution zone If the FAC (and if relevant, ClO2) is less than the equivalent of 0.2 mg/L at pH 8.0 more often than is prescribed in E. coli compliance criterion 2B, full monitoring of E. coli according to Table 3.2a should be carried out in addition to the FAC monitoring schedule. The FAC monitoring regime may be re-instated when the level of FAC has continuously met the requirements of the Standards for one week.

Added ClO2


3.4.2 Giardia and Cryptosporidium transgressions

Sections 3.4.2.1 – 4 outline the responses to when four different types of water supply fail a protozoa criterion. Waters that are not covered by these ‘treatment’ processes, eg, non-secure groundwaters and surface waters that are not treated, do not have a corresponding protozoa compliance criterion, and therefore are non-complying.

Section 3.4.2 needs to be rewritten. We will not have protozoa criteria a, b, c and d anymore either What degree of transgression would lead to a boil water notice for all types of treatment? Can we assume E. coli present? Last sentence is a new comment

3.4.2.1 Response to particles in drinking-water leaving each filter: (protozoa criterion a for filtration normally without coagulation) In the event of a transgression, e.g. the particle counter showing a particle distribution and/or total count that indicates that the filter performance is not satisfactory (distribution range counts and/or total count changes by more than 50 percent), the filter array shall be taken off-line and the transgression procedures implemented immediately. Compliance criterion a is discussed in Section 3.2.3. The reason for the increase in numbers of particles should be investigated. Possible responses should be documented in the Risk Management Plan. Filter performance tests shall be carried out, and if unsatisfactory, then the fault diagnosed and repaired. The problem could be due to a failure of the filter material or precoat; a failure of the sealing of the filter membrane, or cartridge unit into its housing; or an overloading of the filter with particulates (ie, raw water turbidity exceeds filter rating). Once the fault has been remedied the filter or the array shall be re-tested and if satisfactory the array brought back on line. For non-continuous monitoring, the frequency of testing the drinking-water leaving each filter should be increased until the particle compliance criterion is met. Where this cannot be met, the Medical Officer of Health should be advised.

Unless particle counting is included in DWSNZ 2004, this section will be deleted Some has been deleted from here

3.4.2.2 Response to turbidity in drinking-water leaving each filter: (protozoa criterion b for coagulation plus filtration) The reason for sudden increases in turbidity or transgressions should be investigated immediately.

This section should now cover all treatment processes monitored by turbidity, so will need a rewrite. Check Fig still OK too


Compliance criterion b is discussed in Section 3.2.3. Investigations (which should be documented in the Risk Management Plan) should consider the effect of possible occurrences in the catchment that affect the raw water quality, and the need for any remedial action involving coagulant dose adjustment, floc carryover, filter operation, individual filter condition. Refer to the Guidelines for Drinking-Water Quality Management in New Zealand for further guidance. For non-continuous monitoring, the frequency of testing the drinking-water leaving each filter should be increased until the turbidity compliance criterion has been met. Figure 3.4 shows the steps to be followed in response to turbidity transgression for water leaving a filter. Figure 3.4: Response to turbidity transgression for drinking-water leaving a filter

Fig Needs modifying

3.4.2.3 Response to incorrect disinfection C.t of drinking-water leaving a treatment plant (protozoa criterion c) Compliance criterion c is discussed in Section 3.2.3. (i) Chemical disinfection. If the disinfection C.t value specified for the temperature of the water has not been achieved, investigations (which should be documented in the Risk Management Plan) should commence. Either the disinfectant dose rate, or the contact time, or both, should be increased. Where this does not achieve the C.t value required by the Standards, the Medical Officer of Health should be advised. Refer to the Guidelines for Drinking-Water Quality Management in New Zealand for further guidance. Figure 3.5: Response to disinfectant C.t transgression for drinking-water leaving a treatment plant Insert Figure 3.5 – needs to be modified (ii) UV disinfection. If the operating UV irradiance level falls below 400 J/m2 at 253.7 nm at the sensor, investigations (which should be documented in the Risk Management Plan) should commence. This may include inspection of the power supply, all relevant lamps, possibly cleaning or replacing them. If there has been a change in the quality of the raw water, the

(ii) UV added We want a new Fig (3.6, or modify


turbidity and UV transmission or absorbance should be checked. Refer to the Guidelines for Drinking-Water Quality Management in New Zealand for further guidance.

Fig 3.5?) for this new sub-section on UV. Even if only because of the large number of plants using UV.

3.4.2.4 Response to when a secure

groundwater becomes non-secure If a secure groundwater fails any of the conditions described in Section 3.2.4, the drinking-water no longer complies with protozoa criterion d. Section 3.2.4 includes the procedures required to verify that a groundwater is secure.

New section. Is that enough for here? If it takes 12 months of monitoring to show it’s secure again, does that mean it fails the protozoa compliance for the 12 months? It may not be failing E coli criteria during that time – just increased its monitoring

Table 14.8: References to turbidity requirements in these Standards Note: the protozoa

bits need redoing

Compliance Criterion Application Turbidity, NTU Frequency of turbidity test

E. coli Criterion 1: (all waters, ex WTP) Section 3.2.2.1

(a) If chlorinating (b) If ozonating (c) If disinfecting with ClO2 (d) UV disinfection Treated, no disinfection Untreated

<0.5 <1.0, none >2 <0.5 <1.0, none >2 1.0 median, none >2 1.0 median, none >2

Preferably continuous Preferably continuous Preferably continuous When collecting E. coli sample When collecting E. coli sample When collecting E. coli sample

E. coli Criterion 2A: (water in distribution system) Section 3.2.2.2

When testing for E. coli option

1.0 median, none >5

When collecting E. coli sample

E. coli Criterion 2B: (water in distribution system) Section 3.2.2.2

If FAC substitution for E. coli testing

1.0 median, none >5

When measuring FAC, or if relevant, ClO2

E. coli Criterion 3A: (water in bulk main) Section 3.2.2.3

When testing for E. coli option

1.0 median, none >2

(See comment below)


E. coli Criterion 3B: (water in bulk main) Section 3.2.2.3

If FAC substitution for E. coli testing

1.0 median, none >2

(See comment below)

Continuous (See comment below)

Protozoa Criterion a: (Applies to ex filters only) Section 3.2.3

Plants that filter 1.0 median, none >2

As per E.coli Criterion 1, but sampled from each filter

Protozoa Criterion b: Plants that coagulate/filter <0.5 (or <0.1 Preferably continuous,


(Applies to ex filters only) Section 3.2.3

later?) each filter

Protozoa Criterion c: (Applies to ex WTP only) Section 3.2.3

(a) O3 C.t disinfection (b) ClO2 C.t disinfection (c) UV disinfection

(a) 1.0 median, none >2 ?? (b) <0.5 (c) <1.0

(a) As per E.coli Criterion 1 (b) As per E.coli Criterion 1 (b) As per E.coli Criterion 1

Protozoa Criterion d: (Applies to ex WTP only) Section 3.2.3

Secure groundwater 1.0 median, none >2


Guideline Value General process control and measure of water quality

2.5 maximum None specified

Comments: ‘compulsory’ turbidity testing important because turbidity can give rise to 12 demerits in distribution system Grading. Hard to assess without data!

1. Water leaving the treatment plant

1.1 What do you think about the idea of not >2.0 NTU for water ex WTP? This is based on turbidity being potentially higher in distribution systems where not >5.0 NTU applies. Or should this be 2.5 NTU instead of 5, ie – the same as the GV? In which case, does turbidity stop being a GV and become a MAV? Not if we say it is always tested as a ‘micro support test’.

1.2 Assuming we add a turbidity requirement for O3, that means all water leaving a treatment plant will now have a turbidity condition on it. For treatments a-d, the turbidity test is to show disinfection is probably OK, or as a protozoa surrogate, ie, part of the relevant criteria. For the undisinfected and untreated waters, it’s just a general measure of water quality – ie, a GV that now has to be tested regularly? How can we word/sell that?

2. Water in the distribution system

Previously, unless FAC tests were being substituted for E.coli tests, turbidity did not have to be measured, unless related to Guideline Value work, such as complaints. In NZDWS 2000 the turbidity was to be not >0.5 NTU (for this one criterion). Now all water in the distribution system is given a condition as part of the criterion 2A/2B. And this condition is now the same as appears in the Grading. So assuming no deterioration in quality after leaving the treatment plant, all water in the distribution system ideally (ie, GV) should have a turbidity <2.0 NTU, while many should still be <0.5 NTU (hangover from relevant Criterion).

3. General

Now, unless turbidity is being measured continuously by an on-line instrument, it is being tested (leaving the WTP and in the distribution system) every time a sample is collected for E. coli testing, or whenever a sample is tested for FAC as an E. coli substitution.

4. Water in bulk main


4.1 All plants serving >10,000 should have continuous FAC etc monitoring so maybe there should not be a Criterion 3A? But continuous FAC, pH and turbidity monitoring at every sample site is a bit over the top! Maybe there should be one such site (representing each WTP), chosen to represent water that’s normally been in their system the longest, and monitored at a bulk meter. The other sample sites on bulk mains should use the E. coli option.

4.2 If all plants serving >10,000 have to have continuous FAC, then the turbidity of the water leaving the WTP has to be <0.5 NTU. So why do we say the turbidity in the bulk main can be as high as a median 1.0 NTU? That could be construed as a sudden deterioration in quality. Maybe it should be intermediate between the expected quality ex WTP and the quality allowed in their customers’ mains – how about: median of 0.5 and never >1.0?

Documents

Overview of the Drinking-Water Standards · water (except bottled water which must comply with the Food Act 1981). Drinking-Water Standards for New Zealand 2004 (DWSNZ 2004) list