Operationalising water safety plans for Melbourne – a large city case study

Melbourne's water supply utilities have been recognised as pioneers in implementing water safety plans (WSP) since 1999, with the voluntary application of hazard analysis and critical control point (HACCP) to the drinking water supply. WSPs (Risk Management Plans (RMPs)) have been a mandatory requirement in the State of Victoria since the introduction of the Safe Drinking Water Act 2003 (SDWA 2003).

However, the operalisation of WSPs for Melbourne has not been without its challenges, with lessons continuing to be learnt some 24 years later. This technical paper seeks to share the Melbourne metropolitan utility's approach to implementing and optimising WSPs, identifying key factors underpinning their successful application in the Melbourne context.

This paper begins with an overview of Melbourne's diverse water supply systems, ranging from protected catchment unfiltered supplies to open catchment conventionally treated systems. It then sets out the regulatory context and describes the evolution of WSP operalisation for Melbourne.

Practical application methods and examples are shared in Methods: Applied to Operationalise WSPs section.

The ‘Discussion’ provides examples of the challenges encountered during WSP implementation, as well as improvement initiatives identified from lessons learnt through cited case studies.

The aim of this paper is to share practical experience and examples of ‘good industry practice’ which may be of benefit to practitioners and regulators, and particularly those new to WSP development and application.

Approximately half of Melbourne's water is sourced from forested, protected catchments. The catchment system consists of 11 water supply catchments and five water-holding storages. The catchments located within National Parks are co-managed with Parks Victoria, with management arrangements outlined in a National Parks Agreement. The catchments located within the State Forest are co-managed with the Department of Energy, Environment and Climate Action (DEECA). A Memorandum of Understanding details the arrangements to effectively manage human activity and land use for the purposes of protecting water resources in the State Forest.

The five water-holding storages are solely managed by MW. Most of Melbourne's water is supplied via Silvan Reservoir. Historically, Cardinia Reservoir has been supplied by the Silvan system, however, in recent years has received a significant share of its water from the Victorian Desalination Plant (VDP). Some of this desalinated water can then be used to supplement Silvan's demand. Greenvale Reservoir is supplied from the Silvan system.

Treated water from these multiple sources is supplied to the RWCs unfiltered, because of the high quality of water drawn from the protected catchments and large storages.

On average over the last 10 years, approximately 25% of Melbourne's drinking water has been sourced from open catchments that have mixed land uses including farming, rural properties and state forests that are open to activities such as camping and four-wheel driving. The balance of water not sourced from land-based open or closed catchments is provided by the VDP.

Whilst long retention times in storage reservoirs and primary disinfection plants help inactivate microorganisms such as pathogenic bacteria, protozoa and viruses in the untreated water, additional treatment barriers are required in some locations, depending on the risk level of the water.

Chlorination is the primary form of disinfection used to treat Melbourne's water supply, with chlorination plants located at all the major water treatment plants. In addition, MW operates six UV irradiation disinfection plants, which provide effective primary disinfection.

Water from unprotected catchments is also treated by filtration to ensure adequate pathogen removal. MW operates two large filtration plants with treatment processes including coagulation, clarification, media filtration and chemical addition for fluoridation, chlorination and pH correction. In addition, there are three membrane filtration plants associated with partially protected sources.

Melbourne's treated water supply is managed at two distinct levels:

•  – A bulk-treated water transfer system consisting typically of larger diameter water mains (up to 2,100 mm), pump stations, closed water storages and secondary disinfection plants managed by MW (the wholesaler).

•  – A distribution and reticulation network consisting of over 30,000 km of water mains (typically less than 800 mm diameter and a majority of 100–300 mm), 188 pump stations, 296 pressure-reducing stations, 148 closed water storages and 81 secondary disinfection plants managed by the RWCs (the retailers).

This network provides potable supply to a population of approximately 5 million residents in metropolitan Melbourne. As it was historically built and operated under a single water authority until 1995, there are hundreds of physical connection points between the bulk-treated water transfer system and the retailer networks.

MW and the RWCs have separate WSPs to manage risks within their parts of the supply system; however, it is essential these align to effectively manage risks from catchment through to the customer interface (which is typically located at the entry-to-premise plumbing at the customer property boundary).

In practice, MW and the RWCs consider impacts beyond their respective WSP boundaries, to ensure (so far as reasonably practicable) that supplied water is ‘fit for purpose’.

Historically (pre-1999), the approach of Melbourne's water utilities (and others around Australia) was to measure the quality of their drinking water supply by comparing monitoring results against the guideline values in the Australian Drinking Water Guidelines (ADWG). There was not a structured preventative risk management approach for drinking water. Water safety was reliant on end-point Escherichia coli testing, which was deficient due to lengthy microbial sampling processing times and small sampled volume relative to supplied water (not necessarily being representative).

Within Australia, the Melbourne metropolitan water utilities developed systematic drinking water quality management systems in the 1990s, which were informed by the then-current guidance in AS/NZS 4360-1995 Risk Management (Standards Australia 1995; Stevens et al. 1995).

Building on that experience, and spurred on by a sectoral response to the 1998 boil water advisories in Sydney, during the late 1990s, several Australian water utilities developed drinking water quality management systems (Deere & Davison 1998; Deere et al. 2000) based on the HACCP (FAO/WHO 1996) system. This included Brisbane Water (Gray & Morain 2000), Gold Coast Water (Queensland) (Smith 2004), YVW (Melbourne) (Chapman et al. 2003), and South East Water (Melbourne) (Deere et al. 2001; Mullenger et al. 2002).

South East Water was the first water authority in Australia to obtain HACCP accreditation for the supply of drinking water, in November 1999 (Mullenger et al. 2002), immediately followed by YVW in December 1999. The other MW utilities followed, with Quality Assurance (HACCP) certification in subsequent years: MW in 2005 (Wise 2005) and City West Water (now merged with Western Water to form GWW) gained HACCP certification in 2002 and Western Water who gained HACCP certification in 2012.

Third-party certification helped build consumer and regulator confidence in water safety.

Internationally, the HACCP system had previously been proposed as a concept that could be applied to assuring drinking water safety (Havelaar 1994), and a HACCP system had been implemented and third-party audited by the water utility in Reykjavík, Iceland, in 1997 (Gissurarson 2004). The approach was similarly considered in the United States (Barry et al. 1998; Deere & Davison 2006; Martel et al. 2006).

Following the operationalisation of this approach in Australia, consideration was given as to whether the food sector's risk-based HACCP approach was the best one to adopt for drinking water supplies, or whether a modified system, tailored to drinking water supplies, would be preferable (Deere & Davison 1998; Davison & Deere 1999; Davison et al. 1999; Deere 1999). It was concluded that a broader approach was required (Cunliffe 2001; Deere et al. 2001; Stevens et al. 2003). Australia ultimately adopted a tailored approach that was based on the then-current guidance in AS/NZS 4360-1995 Risk Management (Standards Australia 1995), as first proposed by Stevens et al. (1995), and tailored for use in water supply, whilst wholly incorporating all the principles of HACCP and its Supporting Programmes (Deere et al. 2001).

Building on these operational experiences, the ADWG ‘Framework for the Management of Drinking Water Quality’ was developed (Cunliffe 2001). Water utilities were central to the development of that approach, with the Water Corporation (Western Australia), Sydney Water, MW, South East Water (Melbourne) and Power Water (Northern Territory), all undertaking full-scale pilots. Some New Zealand utilities adopted similar Public Health Risk Management Plan approaches (Nokes 2001) during the same period. WHO directly drew upon this experience to propose a similar risk-based ‘Water Safety Plan’ approach (Deere et al. 2001) that was ultimately embedded within the Third Edition of the Guidelines for Drinking water Quality in 2004 (WHO 2004).

Within Australia, the State of Victoria pioneered the formal implementation of a statutory approach to risk management for drinking water in line with this approach through its Safe Drinking Water Act 2003 (SDWA 2003; Labza 2004). Water utilities were required to implement water quality RMPs. Other Australian states and territories introduced similar requirements over time.

WSPs and HACCP implementation have since driven a more proactive reduction in water quality risk, numerous improvements, and facilitated early identification and more rapid response to water quality events and incidents.

Most recently, Australian water utilities have also been rolling out a health-based target (HBT) approach to managing drinking water risks, embedding HBTs within WSPs. The standardised approach provided by the peak Australian industry body, Water Services Association of Australia (WSAA) in 2015, and more recently the ADWG (2022), has assisted water quality specialists in enhancing source water protection measures, treatment augmentation and importantly, guided prioritisation and timing of capital investment programmes to enable best practice water quality management.

The Melbourne utility's WSPs are supported by their organisation-wide integrated management systems. The WSPs satisfy the requirements of relevant ISO standards and HACCP Codex, and broadly align with the 12 Element Framework in the ADWG, which in turn is aligned with the 10 Module approach in the WHO – Second Edition of the WSP Manual (WHO 2023).

Effective regulation of drinking water supply underpins the successful application of WSPs by water utilities in Victoria (Australia).

The Essential Services Commission (ESC) exercises its responsibilities under the Essential Services Commission Act (2001), which include oversight of markets, making price determinations (every 5 years), and assessing and approving utility applications for water pricing and spending (Water Plans, ESC 2023).

The Victorian Department of Health (DH) ensures drinking water is delivered to Victorians by water businesses in accordance with the requirements of the Safe Drinking Water Act 2003 and the Safe Drinking Water Regulations 2015 (DH 2023). Key features of the Act require water suppliers and water storage managers to:

•  – Prepare and implement plans (WSPs) to manage risks to drinking water.

•  – Ensure the drinking water they supply meets the quality standards specified by the regulations (variation is only provided, after community consultation, of water quality standards that relate to aesthetic parameters).

•  – Report any known or suspected contamination of the drinking water to DH.

•  – Publish an annual Drinking Water Quality Report. The report is reviewed by DH and must disclose to the public, information concerning the quality of drinking water.

•  – Engage an independent, trained auditor from a DH-approved panel to audit the WSP every 2 years.

• The benefits of establishing robust regulation specific to WSPs cannot be overstated and include:

•  – Clear direction and guidance on WSP requirements and clear accountabilities related to public health and drinking water supply.

•  – Clear reporting of incidents related to known or suspected contamination.

•  – Public disclosure of water quality data and incidents.

•  – Well-defined requirements for risk-based water quality monitoring programmes.

•  – Well-defined requirements for installation of water sampling points.

•  – Public disclosure of water quality data and incidents.

•  – Drives continuous improvements which assists with prioritising resources and obtaining funding from the economic regulator (ESC).

•  – Well-defined requirements for risk-based water quality monitoring programmes.

•  – Well-defined requirements for the installation of water sampling points.

WSPs have been implemented in over 90 countries (WHO 2023). At least 64 countries have policies or regulations in place that promote or require WSPs or their equivalents. A global survey on WSP policies and practice found that almost two-thirds of countries reporting policies in place or under development did not have complementary requirements for auditing (WHO/World Bank/UNICEF 2022).

WSP's success is heavily dependent on the implementation and operationalisation of all components. Significant improvements in these areas have been made by MW and the RWCs since 1999 to ensure the continual supply of safe and clean drinking water. This section presents some practical examples of operationalising WSPs.

In early development, as MW and the RWCs created their respective WSPs and HACCP processes and documentation, collaboration enabled the utilities to problem solve and ‘learn together’ to implement this ‘then new’ approach. With time, however, experience and confidence (as well as competing priorities), inadvertently led to ongoing renewal of WSP documentation and assessments with somewhat less collaboration between the Melbourne utilities.

In recognition of this, and also in response to feedback following a water quality event in 2020 (refer Results and Discussion section), there has been a renewed effort to strengthen consultation during the updating or preparation of new WSP documents. Examples of this include:

•  – MW and the RWCs jointly participate in the assessment of drinking water quality risks from catchment to tap, rather than risk registers being updated internally by one utility and then shared with the other.

•  – MW and the RWCs jointly attend and assess ground conditions for catchment sanitary surveys used in the application of HBTs.

Various collaborative activities and integrated procedures have also been developed as outlined in the next section, on managing transferrable risks.

In addition to these formal arrangements, managing the transfer of risk is also enabled by informal activities such as:

•  – Joint planning for water quality investigations, options assessments and capital projects. Representatives from all water authorities ensure capital projects align with WSP scope, promoting collaborative risk management.

•  – Fortnightly Melbourne metro water utility water quality meetings: multi-stakeholder discussions addressing water quality issues and sharing best practices for a unified WSP approach.

•  – Knowledge sharing and support via quarterly Victoria-wide water utility meetings.

•  – Participation in a Victoria water quality network working group hosted by the WSAA.

•  – Jointly funded research via industry bodies such as Water Research Australia (WaterRA) and WSAA.

•  – Dedicated, jointly funded resources to drive the delivery of cross-organisational projects, most recently involving the appointment of a water quality-focused resource.

Key improvements from rolling reviews and detailed risk assessments include:

•  – Development of a tank risk assessment tool to prioritise tanks for inspection and maintenance (ongoing programme).

•  – Use of drone inspections for storage tank roof ingress points.

•  – Risk-based water mains cleaning programme.

•  – Backflow prevention enhancements and risk-based retrofit programme.

•  – Improved staff water quality training including more accessible, interactive, online modules.

•  – Development of disinfection controls for tools and equipment (for Contract partners).

•  – Improved data collection associated with flushing after planned or unplanned works to verify water quality performance and manage risks associated with depressurisation of water mains.

MW manages the drinking water supply catchments (watersheds) in accordance with the relevant Acts and Regulations, and takes heed of the following principle set out in the ADWG:

Drinking water quality risks managed at the treatment plant include those introduced within the upstream catchment and which persist through to the raw water abstraction point (treatment plant inlet), as well as new risks introduced during treatment.

The drinking water quality risk assessment evaluates these risks through the treatment process train. The HACCP Codex decision tree was used to identify CCP and quality control points (QCP) for significant items of the risk assessment.

The CCPs and QCPs shown in Table 1 control health-related and aesthetic water quality issues, respectively. They cover water treatment appropriate for the level of catchment protection, reservoir storage and measured water quality characteristics (including disinfection for all supplies, filtration for less protected sources, fluoridation, pH correction as necessary and appropriate levels of chlorine to manage biofilm growth bacteria in the distribution system).

An example of a CCP management and corrective actions is provided in Table 2.

In the event that a critical exceedance of a CCP cannot be effectively managed to assure water supply safety, then the utility must notify DH under the WSAA 2015; Safe Drinking Water Act (2003).

All instances of known or suspected water contamination or issues with the potential to impact water supply, events resulting in widespread customer complaints, or fluoride notifications are published annually in a water utility's Water Quality Annual Report (GWW 2022; MW 2022; SEW 2022; YVW 2022).

Table 3 provides an example of a process control plan to monitor and manage a QCP (customer complaints) and Table 4 how an exceedance related to customer complaints was resolved.

This example demonstrates the operationalisation of the management of a significant risk.

The Melbourne supply system is unique in that protected source water (catchments) enables a significant proportion of supply (over 50%) to be unfiltered. Application of HBTs has identified the variety in the types of source waters being managed by MW, with all four source water categories being identified (ADWG Categories 1–4). The WSP, in turn, must be adaptable to the breadth of risk types and risk levels presented across these catchment types. This is achieved via the drinking water quality risk register.

Typically, the focus of risk evaluation for unfiltered supplies is not surprisingly, on catchment protection, primary disinfection control and protecting against re-contamination during reticulation (closed, pressurised system downstream of primary disinfection). Melbourne's unfiltered supplies are also largely gravity-driven with no post-treatment water storage and some direct offtakes to reticulation networks (i.e. with limited buffering storage and hence inability to ‘halt’ supply as a control).

Examples of tailored risk reduction initiatives implemented for unfiltered supplies include:

•  – Emergency ‘backup’ primary disinfection barriers to reduce the likelihood of primary disinfection failure.

•  – Power reliability improvements to increase treatment reliability.

•  – Robust catchment security (e.g. CCTV, security fences, patrols) to exclude human pathogen sources.

•  – Hydrodynamic modelling of storage to understand pathogen log reduction capability.

•  – In situ water quality (VPS) monitoring in storage to guide selective harvesting through reservoir outlet towers with gates at different depths.

•  – Increased functionality for water transfer to enable individual source waters to be isolated if required, and contingency supplies to be transferred from elsewhere (i.e. interconnectivity of large transfer mains and pump stations).

•  – Use of CCTV for monitoring bird counts on reservoir outlet towers.

•  – Application of a quantitative microbial risk assessment (QMRA) tool to inform contingency plans for treatment failure scenarios.

•  – Use of Microsoft PowerBI dashboards to present data in real-time, to assist in rapid decision-making (BI – business intelligence is a powerful platform for data analytics).

•  – Preparation of business cases (capital planning) for treatment augmentation (e.g. UV, filtration) to reduce lead times for their implementation, if required in the future.

•  – Information sharing with other utilities from other jurisdictions with similar (unfiltered) supplies to guide thinking.

•  – A long established and comprehensive protozoa scat monitoring programme to inform periodic re-evaluation of protozoa risk for protected catchments (Haydon et al. 2022).

•  – Chlorination trials to understand the impact of turbidity from bushfire ash and/or eroded catchment sediment on primary disinfection efficacy, to assist in decision-making during incidents impacting protected catchment/unfiltered supplies (Hellier & Stevens 2009; Hobbs et al. 2010).

For filtered supplies, limited utility control of catchment protection measures results in a shift of focus into building resilience in treatment plants, with robust process control, asset maintenance, provision of critical spares, monitoring and treatment upgrade (additional barriers) where application of HBTs has highlighted a vulnerability. Examples of tailored risk reduction initiatives include:

•  – Planned filter upgrade to convert mono media sand filters to dual media filters and improve backwashing effectiveness.

•  – Study to improve treatment performance and tighten controls for a clarification process at a conventional filtration plant.

•  – UV treatment addition to a 620 ML/d conventional filtration plant.

•  – Augmentation of a UV reactor to enable higher log removals in response to anticipated change in risk profile for an open catchment.

•  – Introduction of Granular Activated Carbon (GAC) as a new treatment process for a supply that experiences taste and odour compounds (MIB and geosmin) in source water.

When water quality events affect two or more of the Melbourne metropolitan water utilities, a coordinated, industry-level approach is required. To achieve this, a comprehensive emergency management framework has been set up. Some of the lessons learnt from the management of complex incidents are described herein.

As a result of a severe storm in August 2020, power outages caused disruption to MW's Silvan water treatment plant power supply. The onsite backup generator was able to maintain power to the site before failing late that night, in turn causing intermittent disinfection failure over a period of approximately 7 h. This resulted in approximately 100 mL of non-disinfected water entering the supply network impacting YVW and SEW customers across 98 suburbs in Melbourne's outer East.

Separate incident management teams were established at MW, YVW, SEW and GWW (then City West Water), each working closely with each other, DEECA (then DELWP) and DH on a coordinated response to the incident. Corrective actions taken to mitigate water quality risks included:

•  – Restoration of emergency power and disinfection processes by approximately 6:30 am on Friday, 28 August.

•  – Submission of a report to the DH under Section 22 of the Safe Drinking Water Act 2003.

•  – Issuance of precautionary Boil Water Advisories (BWA) notices supported by extensive customer communication through multiple digital and media channels by YVW and SEW for impacted areas.

•  – Development and implementation of a plan to return the network to normal operational service.

•  – Targeted chlorine dosing and extensive water quality testing to ensure the water was safe to drink before lifting BWAs.

An extensive internal post-incident investigation was completed by MW to ensure the root cause was established and countermeasures implemented to prevent a recurrence of the issue including:

•  – Asset upgrades and improvements including provision of immediate additional generator capacity at Silvan Treatment Plant and installation of permanent, automatic emergency dosing units for each main supplied by the plant.

•  – Improvement to generator and chlorine dosing preventative maintenance regimes.

•  – Improvements to contingency plans including review of triggers for escalation and criteria for issue of BWA notices.

•  – An ongoing comprehensive review of disinfection control effectiveness at all Water Treatment Plants in MW's system to identify and correct any potential systemic issues.

•  – Approval of a new Mt Evelyn Water Treatment Plant. This new primary disinfection plant will be able to treat water supplied from Silvan Reservoir, downstream of the existing Silvan water treatment plant to enable continuity of primary disinfection for water supplied from Silvan Reservoir during future planned and unplanned outages of the existing disinfection plant.

Internal debriefs were also undertaken by each of the RWCs, along with a joint debriefing process involving MW and the RWCs. The debrief identified joint learnings and collective actions to improve joint emergency response plans, communication protocols and water distribution management.

As part of a suite of systematic improvements in response to the Silvan BWA incident described above, a Joint Action Plan was formed between MW, the RWCs, DH and DEECA (then DELWP). The Joint Action Plan was co-designed by all parties to implement further improvements using the inputs of debriefs and investigative reviews of the event (DH 2022).

The JAP assigned actions to MW, the RWCs and DH, with many assigned collectively to more than one organisation. This necessitated collaboration to deliver the actions. Initiatives used to facilitate collaboration included:

•  – Recruitment of two joint resources (new roles – one planning-focused, the other emergency management-focused) to work equally across the four utilities.

•  – Provision of a dedicated DH point of contact to liaise with the joint resource as a utility point of contact on Joint Action Plan projects.

•  – Fortnightly meetings are attended by the Joint Resource and BWSA Operating Representatives.

•  – Use of shared online spaces for document collaboration and project tracking.

•  – Use of online workshops early in project timelines to seek stakeholder input during development stages.

•  – Numerous improvement projects including network modelling, sensor technology trials and opportunistic pathogen research.

•  – Development and implementation of a strategy to enhance online monitoring of chlorine residuals across strategic interface points.

•  – Development of a joint chlorine residual management strategy focusing on improving chlorine residuals in the distribution and reticulation network.

MW's operational area was impacted by heavy rainfall and very strong winds over several consecutive days beginning 9 June 2021.

Strong winds disrupted the main power supply to 10 water treatment sites and three pump stations for periods of up to 14 days. Access to these sites was also difficult and hazardous due to falling trees. Safety of staff, generator fuel supply and monitoring of water supply quantity and quality were critical components of the first stage of the incident.

Catchment runoff following heavy rain impacted on raw water quality, most significantly in the Upper Yarra Reservoir where high turbidity restricted its use for drinking water for a period of time due to the inability to guarantee treatment efficacy and aesthetic concerns. MW worked closely with YVW and the DH to establish a contingent supply of water to the Yarra Valley townships from the Silvan Reservoir via reverse flow pumping arrangements up the Yarra Valley Conduit, rather than the normal gravity-fed supply from Upper Yarra Reservoir. This reverse flow arrangement commenced on 18 June 2021. However, this alternative supply did not reach the first potable water supply customers until 5 July 2021, as evidenced by fluoride monitoring at the Lusatia Park treatment plant. MW worked closely with YVW on a joint communications plan.

Ultimately MW, working closely with the State Incident Command, YVW, DEECA (then DELWP) and the DH were able to successfully prevent the deterioration of water quality from reaching customers.

In October 2022, 1 orgs/100 mL of E. coli was detected in a routine sample taken from the suburb of Richmond which forms part of GWW's service area. The Richmond distribution and reticulation system is managed by GWW but is supplied from YVW's Kew reservoir. The Kew Reservoir is subsequently supplied from MW's Surrey Hills Reservoir.

GWW immediately notified MW, YVW and DH and incident teams were established at both GWW and YVW. Rapid risk assessments were undertaken by both incident teams with each contributing different sections to the other's assessment. The risk assessments included:

•  – Review of the upstream system design, operating conditions, asset condition, works on the system and assessment of the effectiveness of barriers (YVW and MW).

•  – Review of the local system design, operating conditions, asset condition, and works in the area and assessment of the effectiveness of local secondary chlorination (GWW).

•  – Review of sample results and reactive sampling was undertaken (all).

•  – Assessment of the risk to public health (all).

Key information was shared through regular liaison between the incident teams, briefings and the sharing of situation reports and risk assessments.

The risk assessment ultimately found the E. coli detection to be unrepresentative of the water supplied to customers and no corrective actions were undertaken beyond the detailed investigation.

Audits play an important role in verifying that ‘we are doing the things we say we are doing’. There are several layers of WSP audits:

•  – Annual Internal audits and Continuous Process audits

•  – External HACCP certification audits

•  – Two yearly DH Regulatory audits

Internal audits of the WSP and HACCP system are conducted by audit-trained utility staff or third-party certified auditors to verify that actions comply with the WSP and to identify non-conformities and areas for improvement. The internal audit programmes are endorsed by the Senior Management and the Boards of the corporations.

In the State of Victoria, external auditing of a water utility's WSP is conducted by a DH-approved external auditor, approximately every 2 years. The external audit is a requirement of the Victoria Safe Drinking Water Regulations (SDWR 2015), to verify the WSP complies with the obligations of the Safe Drinking Water Act (SDWA 2003). Opportunities for Improvement (OFIs) and/or non-conformances may be identified that drive improvement prior to the next audit. DH has a comprehensive Guidance document detailing the requirements, audit finding types and actions in case of minor or major non-conformances.

Other external audits take place to re-certify for ISO-certified systems and HACCP certification.

Actions arising from internal and external audits are tracked to completion. Non-conformances arising from external audits are assigned to the relevant manager. Observations and non-conformances arising from internal audits are tracked and reviewed by an Audit team before closure. Audit findings and follow-up actions are reported to the Senior Management.

One of the most important audits are performed by trained and qualified utility Internal Compliance Officers. They are out in the field every day auditing construction and commissioning of new assets as well as planned and emergency works such as valve insertions and water main break repairs, water quality sampling, meter installations, tank maintenance etc. to ensure the developers and contractors comply with WSAA specifications and internal utility specifications. The outcomes are captured electronically in the field in dedicated software. Contractor performance reports and audit actions are tracked and reported using BI reports. In the past four calendar years (2019–2022), of the total of 3029 field audits performed by YVW including topics such as occupation health and safety, environment, plant and equipment, traffic management and customer service, 654 have been specifically for WSP related topics averaging 13 monthly audits.

This paper has described practical aspects of WSP implementation for effective drinking water quality risk management, drawing on recent examples of challenging scenarios with operational constraints, lessons learnt and achievement outcomes. Although MW, YVW, SEW and GWW each manage their own WSPs, collaboration is essential to ensure effective catchment to tap risk management. This is achieved through a combination of both formal and informal processes and underpinned by good communication and strong working relationships. The key contributing factors for a successful WSP implementation are:

•  – Ongoing risk reduction strategies

• One of the key contributors to improved ‘control measures’ for managing risks assessed in the WSP is to focus on process improvements and innovative technologies to manage risks. The examples cited include increased catchment security, hydrodynamic modelling of reservoirs to assist validation of pathogen risk reduction, emergency primary disinfection backup systems, application of QMRA and data analytics to support verification and process improvements.

•  – Continuous improvement

• WSP implementation requires an ongoing review of risks including after incidents and events. Practical challenges and implementation of comprehensive short, medium and long term improvements were demonstrated with a variety of incidents in the ‘Discussion’ above. In particular, leveraging learning from incidents and events has been valuable for identifying process improvements and refinement of communication with stakeholders.

•  – Audits

• This review has also discussed the demonstrable importance of a balanced approach to undertaking WSP audits, including annual to two yearly system audits, as well as the criticality of ongoing field activities and processes performed by the appropriately skilled utility compliance officers.

In the past 20 years, the WSP application has evolved for the Melbourne supply system, from the implementation of HACCP in the early 2000s to the recent integration of HBTs. Shared experiences, learning from each other, collaborating on industry research, benchmarking against others and keeping abreast of advances in industry ‘best practice’ have complimented and supported WSP's evolution for the Melbourne metropolitan utilities.

Strong drinking water regulations, underpinned by robust internal and external auditing processes, have also driven improvement in WSP implementation.

Finally, there have also been opportunities to learn from incidents and to continuously improve, which in turn has built greater resilience into Melbourne's water supply system, better preparing it for a future with potentially more frequent extreme events.

All relevant data are included in the paper or its Supplementary Information.

The authors declare there is no conflict.