Optimizing place-based data infrastructure siting: balancing energy, environment, and communities

Introduction

The rapid expansion of hyperscale data centers has recently drawn significant concern and pushback from host communities, with several counties/municipalities in the United States enacting development moratoria or outright bans.^1–3 Hyperscale data centers require large quantities of reliable electricity, local cooling resources, appropriate land, and associated workforce availability to be considered in a holistic manner (Figure 1).

Figure 1:

Holistic models for siting data centers must incorporate water access, power needs, workforce availability, use-appropriate land, and community environmental health effects to optimize facility performance and improve community-based outcomes.

Local Community Impacts

Residents near data centers have cited a range of concerns related to data center impacts on their communities, including noise from cooling equipment, air and noise pollution from backup power generation, reduced groundwater availability, increases in local electric utility rates, impacts on property values, preservation of rural landscapes, and minimal local economic benefit relative to land-use burdens.^1–3 These tensions highlight a critical knowledge gap: current facility siting analyses do not consider community acceptance or forecast how specific design choices translate into community impacts. Additionally, the community benefits associated with data centers are often overlooked in land use planning and zoning discussions. Data centers can provide well-paying jobs that do not require four-year college degrees and new tax income for local school districts. Proactively addressing public concerns, apprehensions, and misconceptions while engaging with the community could significantly improve the efficiency and effectiveness of data center placement and implementation timelines.

Community health impacts could be modeled before construction through the adoption of Health Impact and Environmental Impact Assessments, which are currently not mandatory. Noise and air pollution mitigation based on innovative technical solutions could be planned with communities in advance, and community benefit agreements could help limit Not-In-My-Backyard (NIMBY) or build absolutely nothing anywhere near anything (BANANA) sentiments. Ensuring local investment and health protection, as part of place-based data center siting tools and methodologies, could help emphasize community benefits.

By integrating community and facility parameters (e.g., proximity to residential settlements, projected water draw, sound power levels, grid load profiles, and local resources such as aquifers and mining tunnels) with demographic, socioeconomic, and regulatory datasets and tax incentives, a siting model can be developed to generate a “community environmental concern index” for candidate sites. For example, Kentucky has a wealth of low-cost land that does not compete with agricultural uses; however, only five of the state’s 120 counties are currently viewed as having feasible locations, with one recently removed due to community pushback.^1,4 Such a siting model could allow planners to test alternative design or mitigation strategies and quantify how each option shifts predicted community sentiment, thereby enabling data-driven engagement strategies and socially resilient facility designs.

Technical Challenges of Data Center Design

In 2022, data centers globally consumed an estimated 460 terawatt-hours (TWh) of electricity.⁵ There is a growing need to reduce the energy demand in data centers through the optimized use of software techniques, such as data deduplication, storage tiering, improved load distribution, and server optimization. Similar to how driver behavior influences the fuel efficiency of a vehicle, software design and execution profoundly affect system-level energy use. However, existing tools and methodologies for assessing and guiding software energy behavior are limited. Powering data centers with backup batteries, load management systems, nearby renewable sources, and energy scheduling, could make it a better alternative for currently used diesel generators.^6,7 Additionally, local contexts such as solar radiation, wind availability, access to fuel, utility prices, emissions intensity, and grid reliability may reveal the policy trade-offs and limitations in different areas.⁸

Computer systems must be cooled to maintain hardware performance, and this cooling demand can account for 25–35% of a data center’s total energy consumption.⁹ Data centers can also consume significant amounts of water to meet cooling demand based on current approaches. Challenges persist in integrating advanced approaches, such as liquid cooling, immersion, and direct-to-chip designs, which face barriers of cost, infrastructure compatibility, and standardization.¹⁰ More experimental strategies, such as subterranean siting or groundwater cooling, are constrained by insufficient performance data and unresolved environmental risks. The absence of universally adopted metrics prevents meaningful comparisons across cooling technologies and accurate assessments of environmental impacts. Empirical data from large-scale deployments are scarce, and predictive tools linking workloads to real-time cooling demands remain immature. Waste heat recovery is promising¹¹ but remains technically complex to scale, with current implementations being limited and dependent on further engineering advances. Without progress in addressing these challenges, cooling will remain a bottleneck in achieving energy-efficient and environmentally responsible data centers.

Finally, technical skills must be developed within the workforce. Community and technical colleges have an important role to play, as not all data center positions require a four-year degree. To prevent labor mismatches, such as local skill gaps or equitable labor outcomes, strong partnerships with industry and nonprofit organizations and communication with local government are needed in a parallel timeframe of siting tools.

Future Research and Policy Direction

Local governments are increasingly offering incentives to attract data center development; however, current siting approaches lack integrated tools that account for technical, environmental, and social dimensions. To address this gap, future research should focus on developing GIS-based multi-criteria decision analysis models that combine facility parameters (e.g., power demand, cooling requirements, and water availability) with community indicators, such as demographic profiles, health risk factors, and land-use priorities. These models should enable planners to simulate alternative design and mitigation strategies, quantify predicted community sentiment, and optimize site selection for resilience and equity.

Policy frameworks must evolve to incorporate mandatory Environmental Impact Assessments and formal Health Impact Assessments early in the planning process, ensuring that community well-being is explicitly considered alongside technical performance. Pilot projects will be critical for refining methodologies and building stakeholder trust. Additionally, collaboration among the industry, academic researchers, and local governments can accelerate the creation of standardized metrics for cooling technologies, carbon footprint benchmarks, and workforce development programs that are aligned with regional needs.

Conclusion

Data center placement is currently both a design and social challenge. A primary issue with siting hyperscale data centers is related to balancing the demands for power, water, and cooling for the center with the health and benefits of the communities. The demand for new facilities is rapidly increasing. Ultimately, a research-to-policy pipeline that integrates technical innovation with community engagement and health protection will reduce opposition, shorten implementation timelines, and ensure that data center growth aligns with sustainability and social well-being.

Biographies

Rochelle H. Holm is an Associate Professor at the Center for Healthy Air Water and Soil at the Christina Lee Brown Envirome Institute, University of Louisville.

Ted Smith is a Professor and the Director of the Center for Healthy Air Water and Soil at the Christina Lee Brown Envirome Institute, University of Louisville.