Project Blue Water and Energy Worksheet -- by Russell Lowes, 5/25/2026 Draft

Introduction

Project Blue mega-data center supposedly changed its cooling methodology in September 2025 to not include wet/evaporative cooling technology on-site, but to switch to dry HVAC and other cooling technologies. In July 2025 they had promoted the concept of wet cooling.

This paper covers both cooling systems, wet and dry cooling, and shows an increased water usage from the dry cooling system. The move from wet to dry cooling pushes off water usage to the Tucson Electric Power electricity mix. For an apples to apples comparison, the July and September Project Blue versions are assumed at the same power service size, with the mix being provided by Tucson Electric Power (TEP).

The purpose of this analysis is to show how much water and energy usage that the first two mega-data centers known as Project Blue would use. The total water usage is being erroneously under-reported on the City website, at this webpage pdf: pdf: https://www.tucsonaz.gov/files/sharedassets/public/v/1/government/city-manager-office/documents/project-blue-updated-fact-sheet_250714.pdf, pg 3 See a copy of this pdf on "COT pdf" tab, directly copied from the pdf onto the next tab, titled "COT pdf" in July of 2025 (just in case the URL gets modified or deleted).

Under the wet cooling method, when the indirect offsite water usage is added to the direct on-site water usage, the total water usage goes up typically by approximately 3.4 to 5.9 times. This is a huge difference between the Project Blue-reported direct usage and the total of that and the indirect usage. Under the section, below, where the dry cooling method is considered, the total water usage increases due to increased inefficiencies of shifting evaporative cooling at the data center to evaporative cooling at the power plants.

Some might say that TEP could alter the energy source mix to be lower in water usage. However, they are now in a collaborative effort with the two largest Arizona utilities to potentially add nuclear plant capacity to their mix. Nuclear energy has the highest cooling needs of any major electric energy source in Arizona. This high water use is due to the non-supercritical temperatures of this thermal electric power source, with a thermal efficiency of just 32.5% (converting only 32.5% of the heat to electricity, versus about 60% thermal efficiency for combined cycle gas electric plants).

It is important to note that TEP said in an opinion editorial published in the Arizona Daily Star on 1/24/26, that Project Blue will be getting their energy from TEP at the "Large Power Service Time of Use High Voltage rate." However, the Arizona Attorney General says that the Energy Services Agreement passed by the Arizona Corporation Commisson allows TEP and Project Blue owner to set their own rate. The Energy Services Agreement, "Allows private parties to bypass the constitutional authority of the Arizona Corporation Commission, which is mandated to protect ratepayers from unfair price hikes. She warned that allowing a utility and a customer to "choose their own rate" is a recipe for massive cost shifts onto ordinary citizens." If the AGO is right, then there could be an even bigger cost shift to other ratepayers. Source: https://lakepowellchronicle.com/stories/arizonas-data-center-loophole-from-tucson-to-page-private-deals-spark-statewide-backlash,95747

I encourage members of the Tucson community to look this over, and send any questions, comments, and suggested improvements regarding this analysis to me, at [email protected].

Model inputs

Only two figures are adjustable (yellow fields): peak capacity and the dry-cooling multiplier. The appropriate tables below recalculate when you change either one of these inputs.

Energy data

Over 99% of all energy flowing into a data center ultimately becomes heat that must be removed by evaporative cooling, HVAC, or ambient air circulation. Sources: TEP 2024 Form 10-K (SEC filing, pp. 3 and 24); City of Tucson Project Blue fact sheet, p. 4 (capacity: primary 400–600 MW, secondary 500–700 MW, total 900–1,300 MW).

Tucson Water deliveries, 2024

Project Blue direct water use, wet cooling

The City webpage says full buildout "won't exceed 6% of the reclaimed water portfolio, or 1% of the entirety of Tucson Water's portfolio." Dividing 1,910 AFY by the 102,654 AFY Tucson Water portfolio gives the figure in the table instead, not 1%. Source: City fact sheet, pp. 3–4.

Reclaimed water recap

Energy and indirect water use, by cooling method

Indirect water = TWh × Σ(mix share × gal/kWh) for coal, gas and renewables at the 2024 TEP energy mix (source: TEP 10-K p. 24). Dry cooling uses the multiplier above for both energy and indirect water.

Water-use sources

Nuclear (PVNGS): NEI Magazine. Coal: U.S. Bureau of Reclamation, Four Corners Powerplant and Navajo Mine Final EIS. Gas: Union of Concerned Scientists. Solar: SEIA. For the sidebar on dry-cooling energy use rising 2–5× (averaging 3.5), see the "Evap vs HVAC" tab.

Additional indirect, off-site water use for wet cooling

Direct and indirect water use, wet cooling

Direct and indirect water use, dry cooling (revised September 2025 plan)

Only indirect water consumption is counted, multiplied by the dry-cooling factor to reflect less efficient heat rejection at thermal power plants.

Additional water use from switching wet to dry

Conclusion

In July of 2025, Project Blue mega-data center announced its planned to use 1910 acre-feet for two combined data centers, using wet/evaporative cooling.

In September of 2025 Project Blue announced a change in plans from an evaporative/wet cooling system to a dry/HVAC-C system. In doing so, they 1) have planned a shift in water use from on-site direct usage to off-site indirect water usage at power plants; and 2) have actually increased the total planned water usage in Arizona by likely between 42 and 66%. In the example below, I assume a 80% data center utilization factor, aka capacity factor.

This analysis uses a conservative approach, in that it gives the benefit of the doubt to Project Blue by only projecting a doubling of water usage for dry cooling, which translates to a doubling of energy and water usage for the more energy intensive dry cooling method.

If the evaporative cooling advantage range of 50-80% is used, for example, and put at the half way point between 50 and 80%, at 65%, the doubling would then become a multiplication of 2.86 instead of 2.0 times. This would increase the projected water consumption substantially. Instead of being, for example, 59% more water for dry systems than wet, it would increase to 127.4% over the wet cooling water usage levels. In terms of water usage compared to Tucson Water delivery, it would go from 14.4% to 20.6% of the water that Tucson Water delivers.

Clearly this is a major shift from on-site water usage to off-site water usage. It is a shift that the citizens of the city of Tucson and the state of Arizona need to consider, along with governments and businesses.

Glossary

AFY means acre feet per year. There are 325,851 U.S. gallons per acre-foot. All gallons here are U.S. gallons.

Direct water use is water that is used on the PB industrial site. Indirect water is water that is used by Tucson Electric Power for cooling at its power plants. It is evaporated and released to the atmosphere. This water loss may occur with water withdrawn in Pima County Active Management Area (AMA), or elsewhere in the state. So it might be a mix of water from our watershed and from our state's watersheds. Either way, it is depleting our state's water resources.

Energy units come in kilowatt-hours, megawatt-hours, Gigawatt-hours and Terrawatt-hours. Each one of these is 1000 times higher as they go from left to right. The abbreviations are kWh, MWh, GWh and TWh. Power Capacity: the units are similar, but do not have the element of time (h for hour): kW, MW, GW, TW. Power over time (or power times time) = energy.

Gals. is in U.S. gallons Gal./kWh is gallons of use per kilowatt-hour of energy. In this table Gal./kWh refers to water for power plants that evaporates into the atmosphere. In other words, it is lost to the environment.

Reclaimed water is water that has gone through primary, and perhaps secondary, treatment. When it goes through secondary treatment, it is often cleaner than the bulk of our water supply in Tucson. Hence, reclaimed water that has gone through secondary treatment can be recharged into the groundwater table for future use as potable water.

Project Blue is a project under Beale Infrastructure, which is owned by Blue Owl Real Assets. They are managing the data center construction for Amazon.

Tucson Water is a department of the City of Tucson. Tucson Electric Power, or TEP, is an investor-owned subsidiary of Fortis, a Canadian company.

Notes

a> Food for thought about efficiency of water cooling is in the next tab. Generally speaking, evaporative cooling in an arid climate like Tucson has an efficient reduction in electricity by 50-80%. Conversely, HVAC technology for this region is 2-5 time the water use for the same heat displacement. For futher explanation, see the EvapVsHVAC tab of this Excel workbook.

b> Water use of dry cooling would be expected to be 2-5 times the water use of wet cooling. Here the lower usage level is used. If a higher level is used, the water usage would go up linearly.

Evaporative vs HVAC cooling: AI query results

Pasted from a DeepSeek query (September 2025). Russell Lowes's comments are shown as shaded call-outs.

Formula: 1 ÷ (1 − savings).

This is an excellent question that gets to the heart of data center efficiency in arid climates. Here is a detailed comparison of electrical use for evaporative cooling versus traditional HVAC (vapor-compression) cooling for data centers in the Tucson, AZ area.

Executive Summary

For a data center in Tucson, evaporative cooling will use significantly less electricity than traditional HVAC cooling—often 50-80% less for the cooling system itself. The primary reason is that evaporative coolers use the natural process of water evaporation to cool the air, which requires only fans and small water pumps. HVAC systems use energy-intensive compressors to mechanically cool the air. However, this massive electrical savings comes with important trade-offs regarding water usage, humidity control, and maintenance.

Detailed Comparison

1. Evaporative Cooling (Direct or Indirect)

How it Works: Uses the evaporation of water to cool the air. Tucson's low humidity is

ideal for this process.

Direct Evap (Swamp Cooler): Air is pulled through water-saturated pads, cooling it

through evaporation and adding humidity. This is rarely used directly for IT equipment due to the humidity addition.

Indirect Evap: A heat exchanger separates the evaporatively-cooled air from the data

center air. This cools the data center air without adding moisture, making it the standard for modern data centers.

Two-Stage Evap: Often used as a supplement. The first stage uses indirect evap to pre-

cool air before it enters a more efficient mechanical cooling stage.

Electrical Consumption:

Very Low. The main electrical consumers are fans (to move large volumes of air)

and small water pumps.

It completely avoids the largest energy consumer in a traditional system: the

compressor.

Key Metric - Efficiency: The efficiency of cooling systems is often measured by their

Coefficient of Performance (COP). A high COP means more cooling per unit of energy.

Typical COP: 10 - 40+ (depending on outside conditions). This means for every 1 kW

of electricity used, it can remove 10-40 kW of heat.

Pros for Tucson:

Extremely low energy use, leading to massive operational cost savings.

Perfect for Tucson's long periods of hot, dry weather.

Cons for Tucson:

High Water Usage: This is the major trade-off. Tucson is in a desert, and water is a

precious resource. A large data center can use millions of gallons of water per year.

Effectiveness Diminishes with Humidity: During Tucson's brief "monsoon" season

(July-August), the rise in humidity reduces the cooling efficiency, often requiring a backup mechanical system.

Maintenance: Pads need regular replacement, and water mineral deposits (scale) must

be managed to maintain efficiency.

2. Traditional HVAC (Vapor-Compression Refrigeration)

How it Works: The standard "air conditioner." It uses a refrigerant cycle involving a

compressor, condenser, expansion valve, and evaporator coil to move heat from inside the building to the outside.

Electrical Consumption:

Very High. The compressor is the largest energy consumer. Second are the

condenser fans and internal fans.

Energy use increases exponentially as the outside temperature rises.

Key Metric - Efficiency: Measured by COP or EER (Energy Efficiency Ratio).

Typical COP: 3 - 6. This means for every 1 kW of electricity used, it can only remove

3-6 kW of heat. This is far less efficient than evaporative cooling in a dry climate.

Pros for Tucson:

Precise Control: Maintains exact temperature and humidity levels regardless of

outside conditions.

No Water Consumption: Uses only electricity (a "dry" system).

Cons for Tucson:

Extremely High Energy Costs: Electricity demand, especially during the summer

peak, will be enormous.

High Strain on Electrical Grid: Contributes to peak demand charges and requires more

robust electrical infrastructure.

Higher Carbon Footprint: If the grid electricity is generated from fossil fuels.

Head-to-Head Electrical Use Comparison Table

The Modern Hybrid Approach for Tucson

Most modern, efficient data centers in climates like Tucson use a hybrid or multi-mode system to get the best of both worlds and maximize overall efficiency (minimizing both energy and water).

Evaporative Cooling (Primary): The system runs on indirect evaporative cooling for

~85-90% of the year when the air is dry enough. This saves massive amounts of electricity.

Mechanical Cooling (Backup): During the humid monsoon season or on exceptionally

hot days, the system automatically switches to (or supplements with) a highly efficient mechanical vapor-compression system to maintain precise conditions.

This approach optimizes the PUE (Power Usage Effectiveness – a key data center metric where lower is better). A well-designed hybrid data center in Tucson can achieve an annualized PUE of 1.1 to 1.3, meaning almost all energy is going to the IT equipment, not the cooling. A data center relying solely on old HVAC might have a PUE of 1.5 or higher.

Conclusion

From a purely electrical perspective, evaporative cooling is the undisputed winner for a data center in Tucson. It will drastically reduce operating costs and strain on the electrical grid.

However, the decision is not purely about electricity. The critical trade-off is water. In a desert community like Tucson, the significant water consumption of an evaporative system is a major environmental and potentially reputational concern. Therefore, the optimal solution is a hybrid system that uses water-efficient indirect evaporative cooling as the primary method and only uses energy-intensive mechanical cooling as a necessary backup, thus minimizing both total electricity and total water consumption.

The previous answer synthesizes established engineering principles and best practices within the data center industry, particularly for arid climates. Here are the key research citations, reports, and resources that support the analysis, categorized by topic.

1. General Superiority of Evaporative and Air-Side Economization in Arid Climates

The fundamental principle that using outside air for cooling is far more efficient than mechanical compression is well-documented by leading industry authorities.

American Society of Heating, Refrigerating and Air-Conditioning Engineers (ASHRAE): ASHRAE's

Thermal Guidelines for Data Processing Environments have progressively widened the recommended and allowable environmental ranges for IT equipment. This has directly enabled the use of economizers (both air-side and water-side) and evaporative cooling by proving that IT hardware can reliably operate at higher temperatures and wider humidity ranges than traditionally believed.

Citation: ASHRAE Technical Committee (TC) 9.9. (2021). _Thermal Guidelines for Data

Processing Environments_ (5th ed.). ASHRAE.

Relevance: This guideline is the bedrock of modern, efficient data center cooling design. It

allows data centers in Tucson to use outside air for cooling for the vast majority of the year without risking equipment health.

U.S. Department of Energy - Lawrence Berkeley National Laboratory (LBNL): Researchers at LBNL

have published extensively on data center energy efficiency. Their work consistently shows that air-side economization (which includes evaporative cooling) is one of the most effective strategies for reducing cooling energy use.

Citation: Shehabi, A., Smith, S., Sartor, D., Brown, R., Herrlin, M., Koomey, J., Masanet, E.,

Horner, N., Azevedo, I., & Lintner, W. (2016). _United States Data Center Energy Usage Report_. LBNL- 1005775. Lawrence Berkeley National Laboratory.

Relevance: This report provides national-level analysis, noting that adoption of economizer

modes can reduce data center cooling energy use by over 70% in suitable climates, which aligns perfectly with the Tucson use case.

2. Efficiency Metrics (COP) and Performance Data

The cited COP values for the different cooling technologies come from engineering textbooks and manufacturer performance data.

Citation: Dvořák, P., & Vít, T. (2018). _Comparison of Cooling Systems for Data Centers_. In

Proceedings of the 6th International Conference on Systems and Control, University of Batna, Algeria.

Relevance: This paper provides a direct comparison, stating: "The indirect evaporative cooling

unit can achieve a COP of 15 to 25 while the compression chiller COP is typically between 3 and 6." This supports the specific COP ranges provided in the answer.

Citation: Intel Corporation. (2008). _Reducing Data Center Cost with an Air Economizer_.

Relevance: While an older study, it is a foundational real-world case study from a major tech

company. Intel implemented air-side economization (a precursor to full evaporative cooling) at a data center in New Mexico (a climate similar to Tucson) and demonstrated a 67% reduction in cooling energy consumption and a $143,000 annual energy savings. This directly supports the claim of 50-80% energy savings.

3. The Hybrid Approach and PUE

The concept of the hybrid/multi-mode cooling system is the industry standard best practice, advocated by major design guides and consulting firms.

Citation: The Uptime Institute. Various White Papers and Tier Standard Guidelines.

Relevance: Uptime Institute publications consistently discuss the design and redundancy

implications of using economizer modes. They endorse the hybrid approach as a method to achieve high efficiency without compromising availability, which is critical for data centers.

Citation: Schneider Electric. (2013). _A Practical Guide to Efficiency Cooling for Data Centers_.

White Paper 228.

Relevance: This guide provides detailed schematics and control strategies for multi-mode

cooling systems. It explicitly outlines how to sequence between different cooling modes (e.g., free cooling, evaporative, mechanical) based on outside conditions to minimize PUE. It provides the engineering rationale for the hybrid system described.

4. The Water-Energy Trade-off in Arid Regions

This is a critical area of ongoing research, especially in the Southwestern U.S.

Citation: Masanet, E., Shehabi, A., & Koomey, J. (2017). _Characteristics of low-carbon data

centres_. Nature Climate Change, 7(9), 627–630.

Relevance: This paper discusses the environmental trade-offs of data center cooling strategies.

It highlights that while evaporative cooling saves significant energy (and thus carbon emissions in many grids), it can increase water consumption by 40-80% compared to mechanical cooling. This directly supports the identified key trade-off for Tucson.

Citation: Carter, K. (2020). _The Thirsty Data Center: Addressing Water Sustainability in Data

Center Design_. Uptime Institute Intelligence Report.

Relevance: This report focuses specifically on the growing concern over water usage. It

analyzes different cooling technologies and their water usage effectiveness (WUE), pushing the industry towards solutions that use less potable water (e.g., using reclaimed water, dry coolers, and more efficient hybrid designs). This contextualizes why the water usage in Tucson is a major consideration beyond just the electrical savings.

Summary

The provided answer is supported by a consensus from:

Industry Standards Bodies (ASHRAE) that enable the technology.

National Laboratory Research (LBNL) that quantifies the energy savings.

Academic and Industry Case Studies that provide real-world performance data (COP, PUE).

Engineering Design Guides that formalize the best-practice hybrid approach.

Contemporary Analysis on the critical water-energy nexus, which is particularly relevant to a

desert environment like Tucson.

City of Tucson Project Blue fact sheet

Copied July 2025 from tucsonaz.gov in case the URL changes. Page numbers from the original PDF appear as stray numerals in the text.

Project Blue: Developing Sustainable Digital Infrastructure in Tucson FACT SHEET Updated July 2025 Executive Summary: Project Blue is a proposed data center project under current development. It represents an opportunity for more than $3.6B in economic development for the City of Tucson. Project Blue includes a major reclaimed water infrastructure expansion that is funded, designed, and constructed by the Project in partnership with Tucson Water. Project Blue will be water positive and match 100% of all water consumed with additional water replenishment projects. Project Blue models how public-private partnerships can support responsible development of critical digital infrastructure by prioritizing community needs. Data Centers

A data center is a specialized facility designed to house computer systems, servers and related hardware that store,

process and manage digital information.

Data centers support technological advancement, driving economic growth and securing national leadership in

innovation. Through cloud computing, they support online education, strengthen community connectivity, and enhance the network resiliency of surrounding areas. Cloud infrastructure remains the primary application of data centers, and requires proximity to population centers to benefit customers.

Data centers are high quality infrastructure assets, generating significant recurring revenue for communities with

minimal impacts to local traffic, noise, and air quality relative to other industrial projects. Project Blue

Project Blue identified Tucson as a potential location for a data center due to its location near major markets such as

Phoenix, Los Angeles and Las Vegas which makes it a prime hub for interconnectivity between major population centers. The region also features low natural disaster risks, strong availability of long-haul transport, and a strong and diversified pool of skilled labor.

Project Blueʼs Primary Project site is located within the Pima County Southeast Employment and Logistics Center,

North of the Pima County Fairgrounds. The Initial Phase of the Primary Project could be operational as soon as 2027.

A Secondary Project is under exploration within the Tucson City limits and would also utilize the reclaimed water

infrastructure outlined below. Exact location is still being determined.

Feasibility studies are under way for a third site in the metro area, but outside of Tucson city limits. Only the Primary

and Secondary Projects are contemplated in this Fact Sheet and in the Development Agreement with the City of Tucson. Water Stewardship

Project Blue will not deplete any Tucson Water resources available to the City today, and will therefore not result in

deductions to Tucson Waterʼs current assured Water Supply portfolio. Since Project Blue is replenishing all water losses and will be providing water in excess of consumption, it will be a Water Positive customer. Water positivity will be achieved through two correlating mechanisms which in sum are Water Positive:

Tucson Water will charge a rate for all water used at the Project to include the Tucson Water cost to increase

Tucsonʼs water portfolio; and

Project Blue will fund local water projects within the region or bring new water supplies to the region to

augment and supplement Tucsonʼs replenishment efforts.

This commitment towards water positivity is a contractual obligation, outlined specifically in the Development

Agreement with the City. All potable water consumed will also be replenished via the Water Positivity requirements in the Development Agreement. The Development Agreement also includes punitive language that does not allow for long term use of potable water for industrial purposes.

Selected projects need to be located within watersheds that connect to Tucson Waterʼs supply portfolio, must

demonstrate measurable benefits to water quality, quantity, or availability and contribute to Tucson Waterʼs Assured Water Supply as regulated by ADWR.

Potential projects may include bringing additional water supply and long-term storage credits into the basin,

supporting potable reuse through funding advanced treatment technologies or public education; implementing PFAS removal treatment at wellheads or treatment plants; restoring rivers and wetlands to enhance infiltration; developing green stormwater capture infrastructure; or advancing water conservation through initiatives like Advanced Metering Infrastructure for early leak detection or funding efficient toilet replacement programs.

Project Blue has aligned its approach to water stewardship with Guiding Principles from Tucson Waterʼs One Water

2100 Plan, including delivering reliability, reinforcing resiliency, enhancing the communityʼs quality of life and achieving accessibility. Reclaimed Water Infrastructure

Project Blue includes a $100M investment to expand Tucson Waterʼs reclaimed water system, extending it from

Mission Manor to the southeast Tucson economic development areas. This represents the largest public infrastructure project funded by a private developer in Tucsonʼs history. All reclaimed water infrastructure described will be funded, designed, and constructed by Project Blue, in partnership with Tucson Water, then delivered to Tucson Water to own and operate.

Throughout project development, Project Blue has partnered closely with Tucson Water, an industry leader in water

stewardship for 125 years and reclaimed water for over 40 years. Tucson Water has guided the water approach for the project to be in close alignment with the Tucson Mayor and Councilʼs water policies and the adopted One Water 2100 Plan.

Project Blue will ultimately utilize 100% renewable water sources for industrial purposes and has committed to

replenishing all consumptive losses. None of the sites under exploration would use potable water for industrial purposes on a permanent basis.

Renewable water sources, as defined by Tucson Water, includes reclaimed water from the Agua Nueva Water

Reclamation Facility (discharged effluent from the wastewater treatment process), the Tucson Airport Remediation Project TARP, and previously recharged reclaimed water from the new aquifer storage and recovery facility. All sources are currently commingled in the Tucson Water Reclaimed Water System.

Project Blue enables significant potable water conservation and community benefit by supporting existing and future

users in replacing potable water with reclaimed water:

The reclaimed water line extension was purposefully routed to abut two potential users of reclaimed water.

The future reclaimed water system will provide an opportunity to convert large scale potable water consumption to reclaimed water at the Los Reales Sustainability Campus, which is currently projected to use up to 306 acre feet per year AFY of potable water, and the future Los Reales Park (a Master Planned multi-purpose sports park).

The route will also facilitate future industrial and commercial customers in the southeastern Tucson region

and SELC to utilize reclaimed water.

The new reclaimed water infrastructure funded by Project Blue includes two main components:

An 18-mile reclaimed water line extension that will be over-sized in capacity, enabling Tucson Water to

facilitate sustainable growth in southeast Tucson and convert existing potable water customers to non-potable sources.

A new 30-acre aquifer recharge facility that will improve system-wide supply reliability and offer recreation

benefits similar to the Shirley C. Scott SHARP facility. Over the duration of the project, more water is expected to be recharged than withdrawn by Project Blue, resulting in a net increase in groundwater over time; this contribution is above and beyond the additional water positive projects. 2 Water Allocation Water use is closely tied to climate and weather conditions, with cooling systems designed to meet demand during the hottest possible days—ensuring reliable performance and safe operating temperatures even under extreme weather events. Estimated average annual allocation volumes are shown in the table below, although actual usage may vary higher or lower on any given year. The allocation volumes shown are based on the projected average water usage due to the climate conditions of many years. Metric Primary Project Secondary Project Initial Phase Full Build Full Build Average Annual Reclaimed Water Allocation 440 AFY 870 AFY 1040 AFY AFY Acre-Feet per Year

Water supplied to the Primary Project (full build) would be comparable to the annual water use of two typical Pima

County 18-hole golf courses. If the Secondary Project moves forward, the water supplied would be equivalent to slightly more than two additional golf courses.

All of the wastewater from the cooling systems is sent to the Pima County sewer system, which eventually supplies

the reclaimed system. No pollutants or toxic chemicals are added to the water. Project Blue will hold an industrial discharge permit from the Pima County Regional Wastewater Reclamation Department and must comply with strict wastewater limitations in order to operate.

At full buildout of both the Primary and Secondary Projects water use wonʼt exceed 6% of the reclaimed water

portfolio, or 1% of the entirety of Tucson Waterʼs portfolio. Per Tucson Water, the reclaimed water supply for this project has been allocated for economic development projects, like Project Blue, and will not adversely affect other existing reclaimed water customers and uses, including the water currently supplying the Santa Cruz River Heritage Project and SHARP.

In 2026, the City will have access to about 32,000 AFY of supply to the reclaimed water system for city uses,

made up of 22,000 AFY of treated effluent and 10,000 AFY from TARP.

The city also wheels effluent owned by other entities to their sites through the Reclaimed Water System,

which is not derived from the cityʼs water portfolio.

Currently, the Reclaimed Water System delivers approximately 11,000 to 14,000 AFY to city customers, using

about 40% of the available supply. The remaining supply volume is either recharged at city facilities such as SHARP or discharged to the Santa Cruz River, where the city does receive some recharge credits, but also loses a significant volume from the region annually.

A limited potable water allocation at the Primary Project will be used for industrial purposes only on an interim basis

until the reclaimed water infrastructure is operational. The switch to reclaimed water is required to be complete within two years of operation. The Secondary Project will not use any potable water for industrial purposes. Energy Supply and Renewable Energy

Project Blue was drawn to Tucson by Southern Arizonaʼs unique access to clean energy generation resources – such

as solar, wind, and energy storage – coupled with the regionʼs existing electric grid infrastructure.

Project Blue is supportive of TEPʼs 2050 Net Zero Emissions goal and planned coal retirements, as well as the City of

Tucsonʼs Climate Action and Adaptation Plan. As one of TEPʼs largest future customers and a major catalyst for regional energy transformation, Project Blue will advocate for aggressive renewable deployment while maintaining affordable rates and reliable service.

The Initial Phase of Project Blue will be supplied under a long-term electric utility agreement with TEP and supported

by local clean energy generation and storage resources, subject to ongoing discussions. These resources are already operating or under advanced development because of TEPʼs all-source request for proposal ASRFP procurements 3 conducted between 2022 and 2024, which successfully identified a large pipeline of new solar, wind and energy storage resources.

The values shown below are estimates of the potential power capacity that may be required to support Project Blueʼs

potential developments. Like the Initial Phase, Project Blue envisions that competitive ASRFP processes will once again enable TEP to maximize the identification and deployment of additional clean energy resources at the lowest cost. Energy Infrastructure & Service

Project Blue will help reduce TEPʼs cost of service by diluting TEPʼs fixed costs over a larger volume of electricity sold.

Project Blue will fund all grid upgrades completed for its direct benefit, ensuring that these costs are not paid by other

customers. The rates paid by Project Blue will cover all of TEPʼs associated costs of service.

Project Blue will support economic growth while improving local and overall electric system reliability. TEP has

conducted rigorous studies of its system and the overall electric grid will benefit from the addition of Project Blueʼs electricity use. Energy Capacity Metric Primary Project Secondary Project Initial Phase Full Build Full Build Potential Power Capacity 250 - 350 MW 400 - 600 MW 500 - 700 MW MW Megawatt

The capacity values shown are estimates of potential power capacity that may be required to support future Project

Blue developments.

Power capacity is one of several enablers to a successful project. Capacity is typically secured through a power

contract with a local utility, separate from land, development, and other agreements. Project Blue anticipates contracting for capacity with TEP when needed. Economic Impact & Job Creation

Data centers generate substantial and reliable property tax revenue for local governments while requiring minimal

local government services and have little impact on traffic congestion in surrounding neighborhoods. This local tax revenue from data centers would otherwise need to be supplied by other revenue streams, such as increases in residential property taxes.

The Chamber of Southern Arizona, in collaboration with Pima County and the City of Tucson, retained Applied

Economics, an independent 3rd party consultant, to perform an Economic Impact Study EIS which calculated the following potential direct economic benefits for the Initial Phase of the Primary Project:

$3.6 billion total capital investment planned over the multi-year construction period.

$250 million in total tax revenues, with $97 million to the City of Tucson, $60 million to Pima County, and $93

million to the State of Arizona over a 10-year period.

180 new full-time, permanent jobs created by the third year of operations in 2029. These positions are on-site

and include engineers, technicians, operations, and security, with an average salary of $64,000 per year.

3,000 direct construction jobs created during the multi-year construction period anticipated to take place

between 2025 - 2028 that will require skilled union and trade labor that is intended to be sourced locally in Tucson. 4

The Primary and Secondary Projects of Project Blue, if fully built, are projected to create over four times the number of

full-time permanent jobs as those estimated for the Initial Phase. The number of construction jobs and the economic impact would also exceed those generated by the Initial Phase by more than threefold. This is an unprecedented investment in the city of Tucson, and an economic development opportunity that will bolster city and county budgets without an increase in taxes on residents.

Project Blue is located in the Vail School District - one of the fastest growing districts in Arizona and the only growing

district in Pima County. Alongside direct tax revenue allocations, Project Blue will unlock a much needed increase in bonding capacity for all Districts in Pima County - creating new potential for capital improvements and long-term investment in local education.

Sources of tax revenue include, but are not limited to, construction sales tax, real estate property taxes, personal

property taxes and transaction privilege taxes on lease rental income. Data centers also upgrade capital equipment over time, generating further substantial tax revenue for local governments. Community Investment & Workforce Development

Project Blue will generate a significant number of new local construction and permanent jobs, creating opportunities

for skilled Tucsonan workers to find employment close to home, rather than having to travel or commute long distances.

Project Blue is committed to building meaningful relationships with the local Tucson community and will partner with

local experts to identify and deliver programs that meet unique local needs, such as:

Apprenticeships & Internships

STEM Education Classes

Skills Certifications

Educator Workshops

These programs benefit Project Blue by upskilling people who can work on the project while also providing valuable

professional skills that participants can use throughout their career. Sustainable Design Features

Project Blue will incorporate many sustainable design features of interest to the community.

Tree Planting: The project will meet and exceed tree planting requirements for the Native Plant Protection

Ordinance NPPO. Where feasible, the project will cooperate with the City in implementing components of the Tucsonʼs Million Trees Program.

Public Green Space: The planned aquifer recharge facility will include a public recreation area with native

landscaping, ramadas and walking trails, similar to SHARP.

Native Vegetation: The site plan includes native plant mitigation to enhance flood control.

Rainwater Harvesting: The project will meet the Cityʼs Rainwater Harvesting ordinance and the Countyʼs

requirements for “first flush.ˮ

EV Charging: Project Blueʼs design voluntarily includes EV charging stations for employees onsite.

Solar: Project Blue will evaluate an installation of solar panels above the parking areas on-site.

Open Space Preservation: The Project is preserving current open space and complying with local ordinances

related to flood control. 5

Capacity (utilization) factor for mega-data centers

80% is used as the typical utilization factor at Digital Realty. Other utilization factors, including a 90% forecast for future data centers, are reported by the Nicholas Institute.

Excellent request. I will provide you with verifiable documentation, citing specific sources and studies, to support the statistics and claims made in the previous explanations.

Core Documentation & Sources

Here is a categorized list of the most authoritative sources, including direct quotes and data points.

1. Google's Direct Disclosure on Power Utilization (Primary Source)

This is the single most direct piece of evidence for hyperscale utilization factors.

Source: Google, *2024 Environmental Report*

Relevant Section: "Data center efficiency" (p. 27 of the 2024 PDF).

Key Quote:

"In 2023, our data center fleet’s average annualized power utilization was 79%—meaning that on average, we used nearly four-fifths of the maximum possible power across our data centers... We target a utilization rate of around 80%, leaving headroom for growth, failures, and maintenance."

Why it's critical: This is a first-party, explicit confirmation from a leading hyperscaler of the ~80% utilization factor target and achievement. It directly validates the "65-85%" range and the rationale for the safety buffer.

Link: [https://sustainability.google/reports/google-2024-environmental-report/](https://sustainability.google/reports/google-2024-environmental-report/)

2. Uptime Institute Global Surveys (Industry Benchmark)

Uptime Institute's annual surveys are the industry standard for data center infrastructure and management trends.

Source: Uptime Institute, *Annual Global Data Center Survey 2023*

Key Data Points:

On average, operators report running at 82.5% of their maximum critical IT capacity (p. 19 of the 2023 report summary).

The report notes a long-term trend: "Most data centers operate at between 70% and 90% of their critical IT capacity—a figure that has changed little in recent years."

It explicitly states that operators maintain headroom for "failures, future growth and daily/seasonal fluctuations."

Why it's critical: This provides broad, third-party validation of the utilization range across the entire industry (including hyperscale and enterprise).

Link: [https://uptimeinstitute.com/resources/asset/2023-data-center-industry-survey-results](https://uptimeinstitute.com/resources/asset/2023-data-center-industry-survey-results) *(Note: Full report requires membership, but summaries and key findings are publicly released.)*

3. Lawrence Berkeley National Laboratory (LBNL) Technical Analysis

LBNL produces deeply technical, peer-reviewed analyses of data center energy use that are foundational to U.S. government policy.

Source: Lawrence Berkeley National Laboratory, *United States Data Center Energy Usage Report* (LBNL-1005775, 2016) and subsequent updates.

Key Concepts: While this report is famous for forecasting overall energy use, its methodology hinges on understanding server utilization and facility load factors. It distinguishes between "peak stock" (installed capacity) and "energy draw" (actual use), highlighting the significant gap.

It cites typical server utilization in legacy enterprise data centers at 12-18% (a key source for that specific figure).

It notes that hyperscale facilities achieve significantly higher IT equipment utilization through advanced software and workload consolidation.

Why it's critical: This is the academic/technical bedrock for the discussion on IT vs. facility utilization and the efficiency gains of hyperscale.

Link: [https://eta.lbl.gov/publications/united-states-data-center-energy](https://eta.lbl.gov/publications/united-states-data-center-energy)

4. Schneider Electric White Papers (Engineering & Design Perspective)

As a major vendor of data center physical infrastructure, Schneider's white papers explain the engineering rationale behind design margins.

Source: Schneider Electric, White Paper 272, *How to Define Data Center Availability: Tier Standards and Topology Choices*

Relevant Section: Discussions on N, N+1, 2N, and 2(N+1) redundancy.

Key Concept: The paper explains that to achieve Tier III (concurrently maintainable) or Tier IV (fault tolerant) availability, capacity cannot be fully utilized. For example, in an N+1 cooling system, if you run at 100% load, the failure of the "+1" unit means the remaining N units are at 100% capacity with zero margin, creating a high-risk scenario. Good practice dictates operating at a maximum of ~80% of system capacity to maintain true redundancy.

Why it's critical: It provides the engineering design principles that *mandate* a utilization factor of less than 100%.

Link: [https://www.se.com/ww/en/download/document/SPD_WSTR272/](https://www.se.com/ww/en/download/document/SPD_WSTR272/)

5. Utility & Grid Operator Reports (Load Factor Evidence)

Reports from power companies confirm the high, stable load profile of data centers.

Source: Dominion Energy (Virginia), *Integrated Resource Plan (IRP) Filings*; Electric Reliability Council of Texas (ERCOT), *Load Forecast Reports*.

Key Data: These filings often note that large data center customers have annual load factors exceeding 90%, meaning their average demand is very close to their peak demand. This "flat" load is distinct from the facility's internal utilization factor but confirms that, from the grid's perspective, these are massive, constant loads.

Why it's critical: It corroborates the scale and consistency of power draw from the utility side, supporting the idea that hyperscale facilities are not cycling on and off but operating at a high, steady state.

Example Search Terms: "Dominion Energy data center load factor," "AWS ERCOT load profile."

Synthesis of Documentation

Putting these sources together provides a complete, documented picture:

Google's 79% is a real, published figure for hyperscale power utilization.

Uptime's 82.5% average confirms this is an industry-wide phenomenon.

LBNL's analysis explains the historical context of low IT utilization and the hyperscale efficiency leap.

Schneider's engineering standards document the *why*: redundancy and safety margins prevent 100% utilization.

Utility reports show the external consequence: these are massive, steady loads on the grid.

How to Cite This Information

For a formal report or presentation, you can structure your citations as follows:

For Utilization Factor (65-85%):

Industry benchmarks and first-party reports indicate hyperscale data centers target an annual average power utilization factor of approximately 80%. Google reported a 79% utilization rate in 2023, while Uptime Institute's global survey found an average of 82.5% of critical capacity in use [1, 2].

For the Rationale (Headroom & Redundancy):

This sub-100% utilization is a deliberate design and operational requirement. It maintains necessary headroom for equipment failures, maintenance, and future growth, ensuring compliance with N+1 redundancy standards and avoiding operational risk [2, 4].

For Historical IT Utilization (12-18%):

In contrast to modern hyperscale operations, legacy enterprise data centers have historically suffered from low server utilization, often in the range of 12-18%, contributing to poor overall efficiency [3].

References (Formatted):

Google. (2024). *2024 Environmental Report*. Retrieved from https://sustainability.google/reports/

Uptime Institute. (2023). *Annual Global Data Center Survey 2023*.

Shehabi, A., et al. (2016). *United States Data Center Energy Usage Report* (LBNL-1005775). Lawrence Berkeley National Laboratory.

Schneider Electric. (2023). *White Paper 272: How to Define Data Center Availability*.

This documentation provides a robust, multi-source foundation for the claims regarding mega-data center capacity and utilization factors.

Rough estimate of heat Project Blue would emit at full build

Sherrill Thomson, a contact on Facebook posed on 1/27/26, "closed water systems," in relation to a discussion on cooling data centers. This was my response that day. Closed water systems are there. They are the "primary" cooling system for many data centers. However, this "primary" term is a little ambiguous -- you would think that it might refer to the main way in which thermal energy/heat is transferred into the atmosphere/removed from the data center computers. "Primary" really refers to the first system, not the system that disburses the heat to the atmosphere. And it is the "secondary" cooling system that does the heavy lifting of removing the heat. Just like at a nuclear plant, the primary cooling system moves the hot liquid to the heat exchangers that then allow the cooling towers to transfer the heat to the atmosphere.

There is a way to make the secondary system closed, but that involves a huge capital investment, and a much bigger repair and maintenance budget. It usually includes radiator tech, like the radiator in a car. I am not an expert on thermal heat removal, but am following the dots on this fairly simple equation. It is all simple math. When you are transferring that much heat, in the case of Project Blue, the heat will be something like this:

So when you are moving that much heat away per hour, the cost of a closed secondary heat system would be enormous. That is why in the arid southwest, evaporative cooling towers with a COP of 10-40, vs a 4.5 COP of heat exchangers (aka HVAC tech, aka heat pump) are the preferred way. They actually use less water than the heat exchanger tech.

U.S. electricity in 2023 vs projected data center use

Sources: EIA for 2023 capacity and generation; Deloitte for data center projections.

Why data centers in the Southwest make little sense

Data centers getting supplied energy by Tucson Electric Power, do not make any sense, here in the arid and water-overdrawn Southwest. The so-called "dry" cooling method even uses more water than the wet method. We need to count the direct onsite water use of wet PLUS the indirect water use of offsite power plant cooling. Tucson Electric Power is 86% fossil fuel-based. The amount that the Project Blue data center was going to use with wet cooling was 1910 acre feet per year. TEP and PB never mentioned the indirect water needs for cooling the data center generated by TEP power plants. I have done the math on that and it comes out to an additional 7409 acre feet per year (AFY), totaling about 9300 AFY. Direct message me with your email address and I will send my spreadsheet-based model to you.

The City of Tucson would not support this water diversion from their watershed (1910 directly plus whatever power plants in the Tucson watershed that provided power to Project Blue. So, PB announced that they are planning to go with "dry" cooling. That is, they plan to use water completely indirectly, and that will actually increase the total water use. It does so because cooling towers (mechanical evaporative cooling) is much more efficient. HVAC is only at a 4.5 COP/coefficient of performance. That means, for every kilowatt-hour you put into an HVAC, aka heat pump, you only get 4.5 kilowatt-hours of thermal energy moved from the data center to the atmosphere. Evaporative cooling towers remove on average, for the mostly dry Tucson area, between 10 and 40 kWht/kilowatt-hours thermal for every kWhe/kilowatt-hour of electricity.

Hence, the water consumption for cooling with so-called dry cooling (HVAC) is at least 14800 AFY. Some say they TEP could reduce their water consumption by going to solar, but that claim is dubious for two reasons: 1) they have been meaningful fighting solar expansion by reducing the amount they pay for rooftop solar about 10% per year. They actually went down in solar % to total energy produced from 2022 to 2023 from 13.8% to 13.2%, but then went up to 14.0% in 2024, for a net increase of 0.2% from 2022 to 2024. 2) TEP is "studying" nuclear energy with the two biggest utilities in Arizona. Nuclear energy uses MORE water than fossil fuels.

Data centers make no sense in the arid SW, with the TEP mix of fossil fuel and with the potential plan to go with high-water-using nuclear energy.