Resilient urban water supply: preparing for the slow-moving consequences of climate change

Disasters in the United States, such as the terrorist attacks on September 11, 2001, Hurricane Katrina in 2005, and Superstorm Sandy in 2012, have elevated the need for critical infrastructure preparedness and resilience to national priority level. Administrative policies and directives have been issued to facilitate the implementation of measures designed to respond and mitigate the consequences of risks (HSPD-5 2003; PPD-8 2011; EO 13653 2013; PPD-21 2013; State, Local, and Tribal Leaders Task Force on Climate Preparedness and Resilience (Task Force) 2014). A number of agencies and organizations have contributed to this effort (Table 1). To date, resilience planning has been principally focused on improving preparedness and the restoration of critical services in communities following extreme events, such as hurricanes, earthquakes or terrorism, and less so on the slow-moving consequences of climate change.

The term slow-moving consequence in this paper is used to differentiate between types of impacts caused by extreme events, and those caused by gradual, incremental changes in natural systems, such as drought and sea level rise (SLR) (California State Assembly 2014). Standard responses in the water sector to improve resilience to slow-moving consequences have often been reactive and have not fully considered long-term solutions. For instance, drought and extended low precipitation periods are generally associated with water supply shortages and/or excessive aquifer drawdown. Typical adaptation measures in these circumstances often focus on conservation, water supply demand forecasting, accumulating storage, supply diversification, and implementing drought and water shortage contingency plans (Brown & Skeens 2011; USEPA 2011, 2015a, 2015b, 2015c, 2015d). Similarly, water supply impacts related to SLR are often associated with salt water intrusion (SWI) in coastal aquifers. SWI is common in coastal locations and is often due to supply wells being over-pumped to meet the demands of growing development and population densities (Galloway et al. 2010). Remedies for mitigating SWI of groundwater resources vary from reducing abstraction to developing alternative sources.

Chloride contamination is not limited geographically to coastal locations or to drought areas. USEPA estimates that ‘two-thirds of the continental United States is underlain by saline waters that can intrude into fresh water supplies’ (USEPA 1999). Chloride contamination has been reported in nearly every state in the USA, and can usually be attributed to anthropogenic causes, such as over-abstraction, or stormwater runoff containing roadway deicers or fertilizers (Todd 1960; Anning & Flynn 2014). In drinking water, chloride can have aesthetic impacts when concentrations exceed 250 mg/L. Under the Safe Drinking Water Act (42 U.S.C. §§300f to 300j-26), chloride is managed as a secondary drinking water standard. Secondary standards are set to provide guidance on removing nuisance contaminants to levels that most consumers will find acceptable (USEPA 2013). Substantial action is often taken to maintain chloride concentrations below the secondary standard to ensure consumer satisfaction with finished water quality.

Health and aesthetics are not the only reasons that chloride is a threat to drinking water systems. It can also cause corrosion of drinking water infrastructure with negative effects on system integrity (Bonds et al. 2005; Muylwyk et al. 2014). Concerns with potential corrosion include the presence of toxic heavy metals, as well as iron and zinc, in drinking water, the latter rendering it aesthetically undesirable for consumption, and the deterioration of plumbing and distribution systems which frequently result in costly replacement (Kirmeyer & Logsdon 1983).

Given the risk of chloride contamination, coupled with the evolving risks from climate change impacts, challenges, adaptation responses and related operational strategies applied by water systems need to be examined and documented. Utility experiences provide a learning opportunity to enhance preparedness for the future. Integrating potential climate change consequences into decision-making is necessary to ensure long-term resilience in the water sector. Considering climatic impacts that could influence chloride concentrations, the aim of this research is to (1) identify indicators of disruption or reduced water service reliability, and (2) identify successful adaptation measures that may inform utilities evaluating options to manage the consequences of slow-moving climatic impacts.

The primary focus of the literature review was identifying indicators associated with the functional, operational and physical impacts that a water utility may experience. Water utilities across the USA have sought solutions to elevated chloride levels arising from many causes. Such actions were identified to glean information on useful practices and seek out potential survey respondents.

Surveys and interviews were conducted to provide as near term as possible review of current utility activities. The questions elicited indicators or action trigger points related to salinity that could hinder functionality, and the measures used to overcome them. The survey and interviews also helped establish a time horizon for adaptation actions. Some supplementary reports and publications were also obtained for better definition of response actions.

Elevated chloride concentrations have many causes. Those discussed in this paper are shown in Figure 2.

SWI occurs when saline water moves into a body of fresh water, raising the chloride concentration. Increased salinity, affects the aesthetic quality of drinking water negatively. Salt water can intrude laterally or vertically (upward or downward) from ground- or surface- water sources, most often in coastal regions (Todd 1974; Barlow 2003; Johnson 2007; Barlow & Reichard 2010).

Increased chloride levels in coastal aquifers are primarily caused by over-abstraction, which lowers the water table, enabling seawater infiltration and contamination. Since seawater is denser than freshwater, a relatively tall freshwater column above sea level generally inhibits saltwater intrusion into fresh aquifers. Over-abstraction has already caused ‘permanent’ contamination of several aquifers in urbanized coastal locations (Barlow & Reichard 2010). The Upper Floridan Aquifer, the drinking water source for approximately 10 million people in Florida, Georgia and South Carolina, is just one location where this has occurred. Distance from the coast, however, does not equate automatically to immunity from SWI. The Monterey/Salinas Valley in California has also suffered SWI because of over-abstraction, and seawater has intruded up to 5–8 km (3–5 miles) inland (Nico Martin 2014). Freshwater aquifers that are below sea level can also be susceptible to chloride contamination, as in the Carpinteria Groundwater Basin, California and the Biscayne Aquifer, southern Florida (Trimble et al. 1998; Deyle et al. 2007).

While snowfall has diminished in many parts of the United States, there is more winter rain (USEPA 2016). Winter rainfall often leads to icy roads that require treatment and salt is a common deicing agent. The runoff from melting snow and ice transports dissolved salts and other contaminants into urban waters, significantly increasing chloride concentrations. Dissolved salts in lakes and streams can permeate through soils, and eventually into groundwater (Hammann & Mantes 1966; Terry 1974; Godwin et al. 2003). There are also concerns about these salts corroding water distribution networks (Bonds et al. 2005; Muylwyk et al. 2014).

The investigation of water utility strategies to address various impacts (see Figure 2) yielded useful results for dissemination about preparing for future impacts. Detailed information from over thirty utilities in the USA was collected and is summarized in Table 2.

Adaptation measures implemented in response to SWI by water utilities are often similar for both ground- and surface- waters. Protection from saltwater migration has required physical barriers. Solutions for lateral SWI have included bentonite slurry walls (due to over-pumping) and tide gates (in tidal canals) to protect freshwater sources in California and Florida (Deyle et al. 2007; Galloway et al. 2010; GPI Southeast Inc. 2012). In Louisiana, the Army Corps of Engineers has built and rebuilt a sill several times for a 14 m (45 foot) section of channel in the Mississippi River. It was rebuilt in 1999 and 2012, due to erosion caused by flow increases following drought, and is expected to require restoration approximately every 5 years. The estimated cost of the original sill – 1988 – was $800,000 (Soileau et al. 1990; USACE 2015). Physical barriers have also been installed on the Sacramento and San Joaquin rivers as emergency measures, when surface waters have receded during drought, potentially enabling seawater encroachment (Croyle 2015). Other types of physical barrier include freshwater injection into purpose-built wells along the coast, as a barrier between sea- and ground- water – e.g., the groundwater replenishment system built by the Orange County Water District (California, USA) (Chalmers et al. 2010). Basins on the central and west coast of California installed similar SWI barrier wells in the 1950s and 60s as a mitigation measure. They consisted of 290 injection wells up to about 210 m (700 ft) deep along 27 km (17 miles) of the coast. Both potable and reclaimed water, the latter extensively treated, were injected. This cost approximately $14 million in 2007/2008, with $5 million in annual maintenance (USEPA 1999; Johnson 2007).

Integrated solutions beyond monitoring are often required to ensure resilience. Since the economy in Southern California depends on groundwater irrigation of high-value, salt-sensitive crops,a groundwater management plan was implemented. Although the water table elevation remained relatively consistent, chloride concentrations in the Monterey and Salinas Valley kept rising (Nico Martin 2014) because seawater was intruding upward from below. An integrated adaptation measure was adopted which supplies reclaimed water for irrigation to offset potable water use (Feldsher & Wu 2010; MCWRA 2016). The Salinas Valley Water Diversion Project augments this by diverting excess surface water flows during wetter periods. The project included a rubber dam designed to retain water released from two reservoirs. When activated, it creates a detention pond and mixes diverted water with reclaimed water for irrigation when needed. In combination, these projects provide multiple benefits: retarding SWI into aquifers, reducing groundwater demand, ensuring reliable alternative irrigation sources, and minimizing wastewater discharges into the Monterey Bay National Marine Sanctuary.

Coastal aquifer SWI due to over-abstraction can yield insights into salinity that may be caused by SLR. The Upper Floridan Aquifer, a primary water source in Florida, Georgia and South Carolina, was first pumped in the 1800s and has been used extensively ever since. High demand has led to over-abstraction causing SWI in Brunswick, Georgia, and elsewhere. High chloride concentrations, sometimes exceeding 2,000 mg/L, have increased source sensitivity awareness. Further development has been limited since then, with real time water level and quality monitoring equipment installed as part of an early warning system. The United States Geological Survey (USGS) monitors the data and maintains the equipment (USGS 2015). Industrial consumption has also been reduced by between 25 and 50% as part of the regional Coastal Georgia Water & Wastewater Permitting Plan for Managing Salt Water Intrusion (GAEPD 2006). This included mandatory audits by Georgia Environmental Protection Division (GAEPD) when industrial permit holders' groundwater abstraction permits needed renewal and, potentially, revised permit/use conditions, as well as new abstraction limits. Abstraction was also moved to shallower sources, e.g., local surficial aquifer systems, where SWI is less likely to be a problem. The plan's primary focus was to stabilize and reverse SWI by managing the urban water cycle, involving water conservation and wastewater reuse, as well as surface- and ground- water abstraction permitting. The plan includes policies and actions for managing wastewater discharges into sensitive ecosystems in coastal Georgia.

Wichita (Kansas) monitors a chloride plume caused by a variety of activities, as well as natural salt deposits and sporadic drought. The USGS simulated chloride transport over 18 years and modeled wellfield management scenarios to control groundwater levels and chloride movement. Scenarios were designed to determine the aggravation causes (groundwater abstraction in different locations) and the best response (e.g., amount of artificial groundwater recharge). Similar work has helped not only define and understand the water quantity and quality needed to ensure supplies, but also provide information on the effectiveness of aquifer storage and recovery (ASR) in other states (Lavista 2014).

Saline water can migrate up rivers, increasing water salinity. The Atlantic coast of South Florida has seen significant development, which can be the main driver for elevated chloride concentrations in some areas. In some cases, land drainage canals, for development, have allowed seawater to migrate and infiltrate freshwater aquifers. Current remedies in South Florida include salinity control dams on various canals, abandoning wellfields contaminated by saltwater, and drilling new water supply wellfields to the west (away from the coast). Coastal canal water levels will probably need to be raised eventually, however, to recharge the Biscayne Aquifer and protect it against further SWI (Trimble et al. 1998; Bloetscher et al. 2010).

Integrated solutions generally provide the best flexibility towards climate and demand fluctuations affecting long-term water supplies. El Paso Water Utilities (Texas) had used reclaimed water for non-potable reuse and recharged the aquifer for indirect potable reuse. The utility then implemented strong conservation measures and built a very large inland desalination plant to treat brackish groundwater. An extended drought and limited surface- and ground- water resources led it to pioneer direct potable reuse. In this case, it was realized that, given its arid climate, population growth and drought, this solution diversifies its water resource portfolio and bolster drought mitigation efforts, with a controlled and reliable year-round and drought tolerant supply. The treatment was designed to maximize public health protection using advanced processes, rigorous critical process monitoring, risk management, and public outreach (Maseeh et al. 2015; Megdal & Forrest 2015).

Typical drought responses are focused on meeting demand. Lessons from the multi-year drought in California, which began in 2011, show the challenges that might be faced by water utilities. The current drought is expected to last long past 2016, and various actions have been taken to improve water conservation and minimize water loss. Bulk water transfers are another emergency measure that can be used (Hodges et al. 2014). Mediterranean coastal cities, such as Barcelona, Spain, received water by barge from France to overcome source supply deficits (Sauri & Domene 2006).

It is clear from this study that emergency measures implemented now and in the past are strongly linked to water resource protection from chloride (NRC 1987; Roehler et al. 2013). The issues that dictate response type are salinity levels and available funds. Acceptable chloride concentrations vary, and seasonality, tidal influence and time-scales all affect the type of response implemented. Cost also has a major influence on community.

In general, actions were reactive, rarely including infrastructure retrofits or modifications to operations, which have not been considered as potential responses. Peace River Manasota Regional Water Supply Authority (Lakewood Ranch, FL), however, modeled groundwater conductivity data relative to tide-levels, which enabled proactive planning to improve water quality. Potential salinity problems are at their most serious during the dry season (April–June), when river flows are low, at high tide and when the wind is onshore from the Gulf of Mexico. To guard against these risks, greater reliance is placed on off-stream reservoirs, and ASR. The utility continues to track future developments because of the dynamic water quality situation (Morris et al. 2015).

Chloride contamination is a real threat to water source quality. There is potential for saline intrusion into fresh water sources across the USA and climate change is likely to make it increasingly common. The impacts of urbanization – over-pumping to meet high demand – SLR and drought, can all be linked to increased chloride levels that threaten drinking water supply quality (Figure 2).

In this study, the challenges and adaptive actions taken in various places to protect water supply sources and manage water quality resilience were observed. If insufficient consideration is given to the threat of SWI, then service disruptions, reduced reliability and potential long-term consequences of source water contamination are likely to occur. The effects on individual water supplies will vary and a common recommendation for protective measures is not suitable. Strategies must be developed to fit the needs of individual communities. Key observations from this study are:

• Be proactive, not reactive. Emergency measures can be effective, but plans for investing in capital infrastructure projects that consider both current and future threats may be a wiser course of action.

• Monitor and be aware of local threats. Establish a source water quality awareness and monitoring program. Monitoring water levels, flows and key parameters such as chloride levels may serve to alert of potential/imminent water supply issues.

• Develop a portfolio of alternative water sources. Invest in alternative water infrastructure – e.g., for water recovery and reuse – to reduce raw water consumption. Other options may include using seawater in sanitation and/or industry. Higher-level treatment may be required in some cases – e.g., reverse osmosis or desalination plants.

• Conduct modeling studies. Salinity issues are complex but mathematical models can be used to clarify major issues, while pilot- or lab- scale tests can help define water quality boundaries, etc.

Future research should include investigation of best practice, and strategies and operational techniques for dealing with salinity issues, including the effects of corrosion. As communities evaluate the potential consequences of climate change on water supplies, solutions will be needed to strengthen long-term water supply resilience.

The authors would like to thank many individuals for their help: Curt Baronowski and Brian C. Pickard (USEPA), Frederick Bloetscher (Denia Beach, Florida), Gregory S. Cherry (USGS), Alan Cohn (NYC Environmental Protection), David Dinkins (University of Florida/IFAS), Tricia H. Kilgore (City of Savannah, Georgia), Michael Low (Boynton Beach, Florida), Barry Stewart (St. Johns County Utility Department, Florida), Kevin Morris (Peace River Manasota Regional Water Supply Authority, Florida), Michael A. Moschenik (City of Dunedin, Florida), Diana Koontz and Nancy T. Norton (SWFWMD, Florida), and Jayantha Obeysekera (SFWMD, Florida).

Pacia Díaz's internship with AWWA was supported with funding from the Water Industry Technical Action Fund (WITAF).