The American Water Works Association’s 2025 Utility Benchmarking Survey put a precise figure on something discussed in general terms for years: 21% of utility employees are eligible to retire in the next five years, against an average vacancy rate of 9%. According to EPA’s America’s Water Sector Workforce Initiative, 30% to 50% of water sector workers are expected to be eligible for retirement within the next 5 to 10 years. In West Virginia, a recent study found that more than half of water utility operators and managers expect to retire within the next decade, highlighting the workforce challenges facing the sector. Similar concerns have been reported by utilities across New England as experienced workers leave the profession and recruitment struggles continue. Those numbers sit alongside the pressures documented in the 2026 Black & Veatch Water Report, where the ageing workforce ranks among respondents’ top concerns, sharing the list with ageing infrastructure, PFAS compliance and a capital squeeze forcing utilities to defer work they know cannot wait.
The statistics flatten what is actually happening. According to EPA’s America’s Water Sector Workforce Initiative, the median age of water sector employees is 48, slightly above the national median across all occupations. According to Brookings Institution research, more than 30% of the broader water workforce was aged 55 or older. The veteran operator who retires on a Friday afternoon does not hand over a folder. She takes with her a working model of the system that was never written down: which pressure zone becomes unstable at peak summer demand, why a particular pump should never start before its neighbour has been running for ten minutes, what a faint conductivity spike at monitoring station seven means about an industrial discharge kilometres upstream. That knowledge was transmitted informally, over years of standing next to someone who already had it. When she leaves, the chain breaks.
21% of utility employees are eligible to retire in the next five years, against an average vacancy rate of 9%
Dan Rickard, Senior Operations Leader at Pennsylvania American Water, has spent four decades working through exactly this transition. “Probably a little bit of everything,” he says of what happens when a veteran retires. “We are able to transfer a lot of knowledge by providing job-shadowing, mentorship and other handover method procedures; we also document a lot of ‘best practices’ with SOPs, but we also lose some, which is just going to happen.” The candour matters. What gets lost is not incidental; it is often the most consequential knowledge the person held.
The 2026 Black & Veatch report is direct about the digital dimension of this gap. While 70% of surveyed utilities report collecting sufficient data, only 19% say they leverage it effectively. Staffing is the single largest barrier cited by utilities failing to meet their digital strategy objectives — 71% of that group name it. Losing experienced operators does not just create a headcount problem; it weakens the institutional capacity to make sense of data already gathered.
When the knowledge was never in the system to begin with
Roles where the risk is most acute tend to be operational and technical ones where experience takes years to develop. Treatment plant operators, distribution engineers and instrumentation specialists carry system knowledge that is rarely fully formalised. The 2026 AWWA State of the Water Industry report ranked workforce issues among the sector’s top concerns, though at a lower position than in earlier surveys when the issue placed among the top five. That movement obscures how the nature of the challenge has shifted. The retirements have not arrived all at once; utilities adapt to individual departures. The cumulative effect, however, is a sector moving from a headcount problem to a knowledge-continuity problem: not how to replace people who retire, but how to keep the knowledge of the utility moving from one generation to the next. Water is delivered by people, and workforce pressure touches nearly every other challenge on the sector’s priority list.
Roles where the risk is most acute tend to be operational and technical ones where experience takes years to develop
The knowledge that walks out is not primarily the knowledge recorded in manuals. SCADA systems log what the system is doing; they do not log why an operator made a particular decision, or what local context shaped it. An operator who knows that a pump should never start before its neighbour has been running for ten minutes has not written that down, because to her it is not a rule, it is just how the plant works. CMMS platforms capture maintenance history without the reasoning behind it. GIS systems map the network without reflecting decades of familiarity with how its buried components actually behave. Research on European water utilities has found that even the precise location of network assets is frequently held in individual memory or on paper records that may not survive a staff transition. The more experience a person accumulates, the harder it becomes to extract because the most valuable knowledge is the kind that has been so thoroughly absorbed it no longer feels like knowledge at all.
The EPA has taken practical steps to address the problem, developing a Knowledge Retention Tool Spreadsheet to help small water systems consolidate operational information ahead of staffing changes. The tool is a useful starting point. But it also illustrates the boundary the sector keeps running into: a spreadsheet can record what a system does and what procedures exist. It cannot record why an experienced operator makes the calls she makes, or what accumulated local context sits behind them.
Tom Walski, Senior Water Resources Engineer at Bentley Systems and one of the most widely cited engineers in the water sector, frames the preparation window in unambiguous terms. The principle, he argues, is simple: before an individual leaves an organisation, someone must already be trained and ready to fill the position. “Do not wait until someone leaves to advertise the role,” he says. “On-the-job training is probably the best form, but it tends to be narrow. Licensing requirements encourage staff to understand systems beyond ‘which button do I push?’ and that broader understanding is exactly what cannot be improvised once the veteran has gone.”
The 2026 American Water Works Association State of the Water Industry report ranked workforce issues among the sector’s top concerns
Culture, Rickard argues, matters as much as any formal process. “When young employees work in a culture where they are not afraid to ask questions or provide input, they are empowered to want to learn. That leads to a strong knowledge base and a leadership skill-set that stays with the organisation.” At Pennsylvania American Water, the approach combines electronic SOPs stored on shared platforms with GIS, hydraulic models and digital twins, and equally, mandatory annual talent profiles and development goal plans for every employee, including milestones tied to union pay scales. The underlying logic is that knowledge must be embedded simultaneously in systems and culture, and that one without the other is insufficient. A well-stocked SOP library does not help the new operator who does not yet know which questions to ask; a strong mentoring culture without documentation produces knowledge that remains as personal and perishable as the veteran who carries it.
Why technology narrows the gap but cannot close it
The digital water industry’s response has been substantial. Simulation environments, cloud-based asset platforms and digital twins have all advanced materially. Modelling software now allows utilities to run failure-scenario drills for events so rare or catastrophic they could never be tested on a live system, creating the experience before the emergency arrives. Cloud-based environments make the decision history of a capital programme continuously accessible to incoming staff, replacing paper plans that became obsolete overnight with a project record any team member can access from anywhere on day one.
Water is delivered by people, and workforce pressure touches nearly every other challenge on the sector’s priority list
Walski poses the central challenge of AI-assisted knowledge transfer with precision: “How do you train machine learning for something that has never happened, but will someday? With modelling software, it is possible to run simulation drills for rare or catastrophic events that could never be tested on a real system. It is better to fail in a drill than during an actual crisis.” The U.S. EPA’s ‘A Day Without SCADA’ national cyber drill reflects the same logic, rehearsing manual operation across utilities precisely because the embodied knowledge that veterans carry needs to become a designed-in institutional capability, not an inherited one.
New digital tools create opportunities for automation and innovation, but they also demand a workforce skilled enough to apply them
The limits of technology are equally important to understand. AI systems learn from data that exists. If operational knowledge was never captured in structured form, there is nothing for an algorithm to learn from. There is a subtler risk too: as SCADA becomes the default interface, a generation of operators is arriving that has never run a pump station without a screen. Walski is direct about what that means in practice. “I’ve heard of systems where the control room operator only knows of a pump station as a symbol on a SCADA screen. The operator has no idea where this station is physically located and would not know how to operate a pump manually if necessary.” That is a fragility, not a capability. New digital tools create opportunities for automation and innovation, but they also demand a workforce skilled enough to apply them, and the sector’s own data suggests it is not yet there. The 2026 Black & Veatch report rates utility expertise in digital twins at 1.69 out of 5 and in data science and AI at 1.78, both among the lowest scores recorded across all digital capability categories surveyed.
The window to transfer what veterans still carry, through mentoring overlaps, simulation development and digital documentation, is open but narrowing. The SOPs memorialise best practice; the mentorship culture ensures new employees feel entitled to ask the questions that no SOP anticipates. Neither works as well without the other, and neither works at all if the veteran has already left before the relationship was built. The utilities that invest in that bridge now will be in a fundamentally different position from those that treat knowledge transfer as a post-retirement problem.
