Infrastructure and Technology Responses

After examining the structural causes of water shortages in the western United States, the second discussion of the 2026 Marshall & Stevens Water Forum focused on responses that could increase supply, improve storage and reduce demand. The panel considered groundwater recharge, the use of data to identify suitable recharge locations, safeguards for water quality, investment in recharge and water-banking infrastructure, advances in wastewater treatment and reuse, lower-water cooling technologies for data centers, and the operating practices needed to maintain reliable, cost-effective water treatment and desalination plants.

The Forum was moderated by Jennifer Simmonson, MAI, Director in the Real Estate practice at Marshall & Stevens. She was joined by Megan Nicholas Harper, Director of Water Resources at Nuveen Natural Capital; Rosemary Knight, Professor in Stanford University’s Geophysics Department and Founding Director of the Center for Groundwater Evaluation and Management; Nati Louzon, VP of Global O&M at IDE Technologies; and Bobby Majumder, Partner at FBT Gibbons and Co-Head of the firm’s Energy Group. Together, the panel brought perspectives spanning groundwater science and management, institutional agricultural investment, water technology and operations, and energy and industrial development.

Groundwater Recharge Turns Wet Years Into Stored Supply

Rosemary Knight described groundwater recharge as a way to address the imbalance between the water withdrawn from aquifers and the water returned to them. Although low groundwater levels are a serious concern in parts of California, she noted that the depleted space below ground also provides substantial capacity for storage. Knight said the available underground storage space is equivalent to the volume of water that could be held in 30 Lake Shastas.

Recharge occurs naturally when precipitation, Sierra snowmelt, water in rivers and lakes, and excess irrigation water infiltrate the soil and move into an aquifer. Knight also discussed restoring land so that natural recharge processes can function more effectively. Floodplains historically contributed to recharge when rivers overflowed, but she said levees have separated many rivers from those floodplains.

Managed aquifer recharge adds an intentional operational component. Water can be spread on the ground or injected underground. Knight discussed Flood-MAR, which uses excess floodwater for managed aquifer recharge by moving it to locations where it can infiltrate and replenish groundwater. In a state that moves between flooding and drought, the approach can store water from wet periods for recovery from, or preparation for, the next drought.

“Let our engineered infrastructure tap into our natural infrastructure.” – Rosemary Knight

The Location of a Recharge Project Determines Whether It Works

The availability of water and underground storage capacity does not make every location suitable for recharge. Knight said a recharge pond can fail to deliver the intended benefit if water remains at the surface and evaporates rather than moving efficiently into an aquifer. For growers who make land available for recharge, prolonged ponding can also destabilize orchard roots and introduce disease.

Knight described the use of geophysical measurements to examine what lies below the surface and determine how readily water placed at a site will reach the groundwater system. Her team uses those data to generate maps that identify areas as suitable or unsuitable for recharge. In the Tulare Irrigation District, she said, Stanford collaborated with the district to acquire a towed electromagnetic system that can move between rows of almond trees and collect subsurface data.

The objective is to convert measurements into information that growers and water managers can use when deciding where to direct water. Knight described interest among growers who want to participate in recharge but also want evidence that flooding their land will produce the intended result.

“They don’t want to just flood their land and have a problem. They want to flood their land and have a solution.” – Rosemary Knight

Recharge Requires Both Local Participation and Public Oversight

Asked about the respective roles of water users and government, Knight emphasized grassroots participation supported by government action. She said California has removed permitting obstacles that previously could delay approval until the floodwater available for recharge had already passed. At the same time, she cautioned against removing safeguards needed to prevent unintended consequences.

Water quality was the central risk she identified. Moving water downward can potentially carry residual pesticides near the surface into drinking-water supplies. Knight therefore called for adaptive management and active monitoring so an effort to increase groundwater quantity does not degrade groundwater quality.

Her description assigned related but distinct roles: landowners can make land available and participate in recharge, while government can facilitate projects and help ensure monitoring remains in place. The combination allows recharge to move forward while maintaining attention to its effects.

“When you recharge, monitor.” – Rosemary Knight

Water Reuse Is Becoming a More Practical Source

Nati Louzon described advances in treatment technology and accumulated operating experience as factors making recycled water a more reliable and practical source. He pointed to improvements in equipment, membranes, treatment processes and recovery rates. According to Louzon, wastewater recovery that had been approximately 60 to 70 percent can now reach 80 to 85 percent or more in some applications, increasing the amount of usable water while lowering its cost.

Cost matters because users can compare recycled water directly with tap water. Louzon said lowering the cost of treatment can increase the role of recycling. As an example, he described a planned wastewater-recycling project in Israel that would combine technologies to treat wastewater for irrigation and industrial use rather than discharge it to a river or another location.

Louzon also emphasized that these projects depend on both construction and operation. Modern recycling systems can combine membrane bioreactors with nanofiltration or reverse osmosis, along with other technologies intended to increase recovery. He also referred to zero-discharge approaches that have become more common in industrial applications as the technology has improved.

“The main key here is successfully to build and operate those plants.” – Nati Louzon

Recharge and Water Banking Need Long Term Capital

Megan Nicholas Harper agreed that recharge and water banking can contribute to California’s water response, but she emphasized the resources required to implement them at scale. These projects need infrastructure, operating expertise and long-term capital rather than land and water alone.

Nicholas Harper said basic earthwork for a surface recharge basin can cost approximately $8,000 to $10,000 per acre, in addition to ongoing cultural costs. Surface basins can also experience evaporation losses. She described subsurface recharge, which can use buried perforated piping to increase infiltration while keeping productive land in use, as an alternative that can significantly increase the project cost.

The full system can include conveyance facilities, basins, land, taxes, operations, maintenance, monitoring, accounting and a governance framework. Nicholas Harper pointed to the Tule Subbasin, where water districts and landowners have invested in groundwater recharge and banking programs for more than 40 years and have banked hundreds of thousands of acre-feet of water.

She also explained why the physical structure of an aquifer affects both project benefits and investment decisions. Recharge water may benefit an upper aquifer while a confined lower aquifer does not receive the same benefit. When the parties making the investment cannot capture the full value created by recharge, individual landowners and water managers can find it difficult to justify substantial upfront costs. Nicholas Harper identified governance, accounting and policy as important foundations for further investment.

“Investors can provide long-term funding needs for projects that generate water security over decades rather than years.” – Megan Nicholas Harper

Some Storage Approaches Carry Higher Treatment and Infrastructure Requirements

Looking ahead, Nicholas Harper identified aquifer storage and recovery and deep-well injection programs as significant areas of both opportunity and challenge. These approaches can expand groundwater storage, but she described them as capital intensive.

Water intended for injection into confined systems must receive substantial treatment, approaching drinking-water quality, to avoid contamination. That adds treatment and infrastructure costs to the storage project. In Nicholas Harper’s assessment, institutional investors can contribute by funding systems designed to capture wet-year supplies and store them for future use.

Data Center Cooling Is Changing the Water and Cost Equation

Bobby Majumder turned the discussion from expanding and storing water supplies to reducing water use by industrial facilities. He said data center developers have an incentive to improve water efficiency because doing so increases the number of sites at which facilities may be located.

Majumder described total cost of ownership as incorporating initial capital expenditures for a facility and associated infrastructure upgrades, as well as annual operating expenditures. Against that framework, he outlined four cooling technologies that had moved to the forefront: rear-door heat exchangers, direct-to-chip cooling, single-phase immersion and two-phase immersion.

A rear-door heat exchanger uses a liquid-cooled panel installed at the back of a server rack to remove heat. Direct-to-chip systems place liquid cooling at the server blade and chips and transfer heat away more efficiently than traditional facility-level evaporative cooling. In single-phase immersion, server hardware is submerged in a thermally conductive fluid. Two-phase immersion uses an electrically insulating liquid that boils at a low temperature to transfer heat.

Majumder described these technologies as a progression in initial capital cost, with rear-door heat exchangers at the lower end and two-phase immersion at the higher end. He said rear-door heat exchangers can produce total-cost-of-ownership savings of 25 to 35 percent over traditional air cooling across a five-year period. His larger point was that water requirements and the cost of alternative cooling systems now belong in the initial business case for a data center, alongside land and power.

“Now you have to have water at the outset, and investors are asking, what is your water plan?” – Bobby Majumder

Long Term Performance Depends on How Plants Are Operated

Louzon closed the second topic by discussing the operating practices and performance measures needed for water treatment and desalination plants to deliver high-quality water while controlling long-term costs. He placed reliability, efficiency and proactive operation at the center of that effort.

Rather than treating plant performance as fixed once construction is complete, Louzon described a continuous cycle of identifying an area for improvement, making the improvement and checking performance again. He said stopping that process leaves a plant behind both operationally and financially.

Louzon also described the use of artificial intelligence with supervisory control and data acquisition systems and computerized maintenance management systems. The objective is to reduce reaction time, increase production availability and lower cost. Other performance measures he identified include chemical consumption, energy consumption and water recovery.

Electricity tariffs can also affect operations. Louzon referred to improvements at a California plant that allow it to respond to lower and higher tariffs during the summer. Across desalination and other water-treatment facilities, he said continued improvement in operations and maintenance is necessary to reduce the cost of water production.

Technology Works Within a Larger Delivery System

The second Forum discussion presented several ways to increase usable water supplies, store water for future need and reduce demand. Knight described natural and managed aquifer recharge supported by site-specific subsurface data and water-quality monitoring. Louzon explained how advances in treatment and plant operations can increase wastewater recovery and improve long-term performance. Nicholas Harper detailed the infrastructure, governance and patient capital required to carry recharge and water-banking projects forward. Majumder outlined cooling technologies that can reduce water use while changing the total cost of data-center ownership.

Across those responses, the panelists repeatedly connected the technology itself to the conditions required for implementation: suitable locations, reliable measurements, protective monitoring, capable operators, governance, infrastructure and capital. Their comments framed infrastructure and technology as operating systems that must be planned, financed, monitored and improved over time, rather than as stand-alone solutions.

Explore the other articles in this four-part series for additional perspectives and insights from the 2026 Water Forum.