Drinking Water Utility-Level Understanding of Climate Change Effects to System Reliability

Climate change hazards, including increased temperatures, drought, sea level rise, extreme precipitation, wildfires, and changes in freeze–thaw cycles, are expected to degrade drinking water utility system infrastructure and decrease the reliability of water provision. To assess how drinking water utility manager perceptions of these risks affect utility planning, 60 semistructured interviews were conducted with utilities of various sizes, source water supplies, and United States geographical regions. This study analyzes these interviews (1) to evaluate which climate hazards are of primary concern to drinking water managers, (2) to develop a mental model framework for assessing utility-level understanding of climate change risks to system reliability, and (3) to examine the status of current water utility adaptation planning. The results show that concern and awareness of climate hazard risks vary geographically and are grounded in historical exposure; some participants do not believe climate change will influence their system’s overall reliability. When considering climate change risks, utility managers tend to focus on effects to water supply and infrastructure, as opposed to changes in operations and maintenance, water quality, or business functions. Most surveyed utilities do not have comprehensive climate adaptation plans despite federal and professional recommendations. The range of beliefs and actions concerning climate adaptation planning indicates that utilities need directed guidance, and policymakers should consider including climate hazards and projections as part of required utility risk and resilience assessments.

S4 affairs. 22 Climate-induced changes to pipe and pump station lifetimes will require water utilities to change maintenance and operations schedules. More broadly, changes in water quantity and quality will indirectly affect financial viability.
In addition to reducing the reliability of individual infrastructure components and indirectly affecting operations and financials, climate hazards directly affect water utility business functions. As an example, extreme heat affects a utility's ability to conduct maintenance and repairs and requires changes to public communication and external affairs, expanding the utility service mission from just delivery of clean water to communication about reliability. 23 Additionally, climate risk is analyzed when reviewing a water utility's operational risk management and financial management, and may begin to factor into credit and bond ratings. 24 The Water Utility Climate Alliance emphasizes the connection between direct effects to infrastructure and direct and indirect effects to operations and financials, arguing for links between climate risks and water utility business functions. 22 These links are needed to understand how climate change will affect the water utility as an institution, not just effects to individual infrastructure components.
The direct effects of five climate hazards (increased temperatures, sea level rise, extreme precipitation, wildfires, and changes in freeze-thaw cycles) to water infrastructure components were compiled to develop an initial model of climate risks to water utility reliability ( Figure S1). This proposed influence diagram illustrates the interconnected nature of water utility reliability and its links to climate hazards; this diagram shows how different climate hazards might affect reliability of different infrastructure components and water utility business functions. The interview begins with questions about the interviewee's background and experience at the utility (Section I). Though drinking water utilities are considered reliable, there is a lack of consistency in reliability of service standards. 25 Section II assess how the interviewee frames "reliability," from particular metrics to understandings in relation to business functions. Specific water utility business functions were compiled 22 and asked about in Question 8. The Environmental Protection Agency tracks water quality compliance violations but not level of service metrics, like length or frequency of outages. There are a range of indicators used to benchmark water utilities, including service outages, maintenance needs, and customer satisfaction indices, but most are self-reported to professional organizations 26 and few are regulated by state public utility commissions. 27 These various reliability indicators were specifically asked about in Question 12. Section III captures how climate risks might change utility reliability. The initial influence diagram (see SI Figure 1) was used to inform which specific events/hazards and reliability metrics were asked about in Question 19. Section IV finally asks specifically about climate change and the status of adaptation planning. Structure of interview questions was informed by work concerning electric utility resilience. 28

Interview Protocol
Let's first start with some basic information:

Section I: Background
1. Tell me about your water utility, such as system properties and customers served. 2. Tell me about your job. How long have you been working with this water utility and what are your roles and responsibilities? 3. Please walk me through the process your utility has when deciding on new projects.
These projects could be part of a capital improvement plan or maintenance/operations decisions. Now let's discuss the reliability of your drinking water system and water utility: Section II: Reliability Overview 4. How do you define reliability in relation to your work? 5. Do you think your definition of reliability is influenced by your background & field? 6. In general, are you more concerned about long-term or short-term reliability? 7. Do you think about component reliability (ability of one infrastructure component to function) differently than system reliability (ability of system to deliver clean water)? 8. Which of the following water and non-water business functions do you think influences system reliability?

Water Utility Functions
Influences System Reliability? 13. Which of the identified metrics do you think is most important? 14. Does your utility have level of service expectations associated with this identified metric? Now let's discuss how the reliability of your water utility has changed over time: Section III: Changing Reliability 15. Is your drinking water system more or less reliable today than it was 10 years ago? 16. Which functions/departments have seen the biggest changes in reliability over time? 17. What do you think are some of the drivers behind why/how system reliability is changing? 18. In general, do you think people are talking about the same risks as they were 10 years ago? 19. Which of the following hazards/events do you believe could pose risks to your system reliability in the near future? Check all metrics you believe the event/hazard could affect.

Figure S3. Identified Hazard Risks to System Reliability (n = 60).
Hazards ordered in terms of overall concern by study participants; bar graphics and percentages illustrate same information. Northern Rockies and Plains region not included due to limited sample size (n = 3).

Gaps in Mental Model
The results indicate that water utility managers' understanding of climate-induced changes to system reliability is primarily focused on changes in water resources and supply. Participants demonstrated understanding of how climate change is expected to affect water resources, but they appear less certain about how other factors will be affected. Some participants shared robust understandings of climate projections in their regions; one stated, "What we see in our area for climate change is that storms in the wintertime will be sort of maybe more prominent, the average rainfall will be similar, but rainfall events and flooding events may be more severe. And with the lack of snowpack, which helps us to restore our aquifers, I do think about the long-term impacts to aquifers." Only four participants referred to the hazard-reliability matrix when asked about climate change and only two participants explicitly identified the hazards in that matrix as associated with climate change. This could suggest a disconnect between risk awareness from current climate hazards and how future changes to those hazards will affect utility risk. Most participants identified an individual hazard (typically drought) as their "relevant" climate hazard and discussed changes in system reliability in relation to that individual hazard, rather than discussing a range of relevant climate hazards.
Participants identified cascading, indirect consequences of climate change hazards to infrastructure (i.e., how heat-related changes in demand will strain pump stations and pipes).
Less discussed were the direct effects of climate hazards on infrastructure. Increased temperatures can degrade individual infrastructure components. Only 8 out of 60 participants identified concerns about electronic/motor overheating failures of pumping and lift stations, which have been linked to service outages during summer heat waves 2 and are expected to increase as ambient temperatures rise. 3 There were several mentions of increased temperatures S22 affecting pipe breaks, but no participants described temperature-dependent corrosion of metal pipes and degradation of thermoplastic pipes, which may shorten pipe lifespans. [4][5][6][7] Climate change may lead to differential soil settlements, as severe heat waves and long droughts lower ground water levels and consolidate soils. These changes can damage underground pipe infrastructure. 8,9 Further, the combination of longer, drier periods followed by periods of extreme rainfall can cause soil expansion and settlement; the resulting differential pressure can cause pipe cracking and failure. 10,11 Only one participant mentioned climate change affects to soil properties and buried infrastructure. Five participants noted that the amount of buried infrastructure in a water system provided resilience against hazards.
Extreme precipitation often causes flooding, and flooded treatment plants and pump stations can reduce the utility's ability to treat and distribute drinking water. 2 Of the four participants who mentioned climate-related flooding of facilities, three had experienced large flooding events recently. Coastal utilities mentioned concerns about sea level rise, which can also induce flooding, change soil pressure and settlement, and in turn, fracture pipes. 12,13 Saltwater intrusion degrades buried pipe mechanical properties 14 and poses risks to water quality throughout the water system; 15 five participants discussed this degradation and identified water quality issues with salt water intrusion into groundwater wells.
Another climate hazard, wildfires, can deteriorate water quality, which was mentioned by three participants (all from western regions where wildfire is common). Apart from affecting source water quality, wildfires can introduce contamination directly into distribution system water quality because of thermal degradation of PVC pipes or infrastructure components like hydrants valves and meters. 16,17 Only one participant identified this potential consequence.

S23
Finally, freeze-thaw cycles have known correlations with pipe breaks. 19,20 Climate change will likely decrease total snowfall in the winter but increase freeze-thaw cycles, though the extent to which that will happen varies by region. In some locations with seasonal cold weather, pipe failure rates might decrease as a result of climate change due to fewer freeze-thaw cycles. 21 Five participants from the Upper Midwest and Ohio Valley regions noted that climate change could affect cold temperatures, in turn changing pipe break rates.
In addition to gaps in understanding the potential effects of specific climate hazards on infrastructure, the aggregated mental model also illustrates gaps related to direct effects of increased temperatures, sea level rise, extreme precipitation, wildfires, and changes in freezethaw cycles on water utility business functions. As an example, extreme heat affects a utility's ability to conduct maintenance and repairs and requires changes to public communication and external affairs, expanding the utility service mission from delivery of clean water to communication about reliability. 23 Only two participants identified increased heat exposure and stress as potentially affecting personnel. In the participants' responses to the hazard-reliability matrix changes in maintenance need was identified as the most common consequence of the climate hazards, but when asked about effects of climate change, few participants identified changing maintenance needs. On the management side, climate risk is analyzed when reviewing a water utility's operational risk management and financial management, and may begin to factor into credit and bond ratings, 24 but no participants mentioned credit or bond risks from exposure to climate change. Finally, half of the participants identified challenges with communicating with customers, with some expressing that customer perceptions and expectations of the drinking water sector have soured in recent years (five out of 60 participants).

Discussion of Institutional and Technical Barriers
Participants described both technical and institutional limitations for why adaptation planning wasn't taking place, though institutional barriers were mentioned more frequently.
Technical limitations (Table S5) describe barriers to engineering project designs and execution; participants highlighted the uncertainty associated with climate projections, the potential for magnifier events or tipping points, and the expense as reasons for why the utility has not integrated climate adaptation into planning efforts. These technical limitations mirror those identified a decade ago, 29 indicating progress is still needed in providing utilities with usable climate projections they can easily integrate into existing planning processes. Ten out of sixty participants (17%) said their utility used climate change projections in planning efforts; streamflow and rainfall intensity projections were most commonly mentioned, followed by projections of sea level rise. Three participants mentioned using climate projections in the utility's future demand models to assess how climate change affects water consumption and usage. Only one interviewed participant described their utility's use of climate scenario planning to determine what future climate projections to design for; their utility was making decisions about the level of water supply reliability based on different future climate scenarios. Only one participant mentioned any of the EPA's CRWU tools; he identified the CREAT tool as one that his utility hasn't made use of yet but hopes to in the future. Eleven out of the 60 participants (18%) discussed tracking changes to source water, infrastructure assets, and water quality as a way to assess when climate change effects begin to be observed. aware of the general vulnerabilities to climate change a water utility might face and accurately understood that there are regional differences in changing climate hazards. When asked about the effects of climate change on his utility's drinking water system, one participant said, "I think in different parts of the country, you're gonna get different answers. In the part of the country where the utility is water rich, you'll get one answer. If you're out in the arid west, or in S26 California, you're gonna get a different answer. Those folks are experiencing 20-year droughts and they're gonna be many times more concerned about climate change. And I would say, the more water rich the utility, the less concerned they're gonna be with climate change." While climate risk to water resources and supply does vary depending on geographical region, 30 all regions of the United States are exposed to some climate hazards. 31 Climate hazards can pose risks to utility infrastructure systems, operations and maintenance, water quality, and business functions, separate from affecting water supply. Every participant that perceived the issue as one for "others" to contend with did not see climate change as posing a risk to utility business functions. Another institutional limitation is the focus on historical precedent to guide concern for climate hazards. One participant said, "locally here, I don't think a lot of utilities have a lot of plans in place, just because of how fortunate we have been historically." When discussing whether or not a climate hazard could pose a risk to their system's reliability, participants relied on recent experience of that hazard.