El Paso’s expanding data-center sector raises several distinct questions: what will be built, what benefits will reach local residents, how water and electricity will be supplied, and which obligations can be enforced over the life of a project. The available evidence supports substantial planned investment and construction activity. It also leaves important questions about long-term utility costs, public revenue, environmental performance and employment outcomes unresolved.
The central project is Meta’s northeast El Paso campus, developed through Wurldwide LLC. Nearby projects in southern New Mexico and a separate Army initiative at Fort Bliss add regional context, but have different contracts, utilities and approving authorities. Their investment and employment announcements should not be added together as though they describe a single El Paso project.
This briefing covers public information available through September 13, 2026, Mountain Time. Future dates are plans or scheduled events. Dollar amounts are nominal unless their source states otherwise. Corporate announcements establish what a company has announced; they do not independently establish completed construction, actual consumption or delivered community benefits.
Development and ownership
Meta publicly described an initial $1.5 billion El Paso development in October 2025. Its March 2026 expansion announcement increased the campus plan to approximately $10 billion and 1 gigawatt, with about 300 operational jobs and up to 4,000 construction workers at peak. These are development and employment expectations, not an audited count of permanent local employment. Construction jobs and ongoing operating jobs should always be reported separately. Meta’s October announcement, updated March 26; Meta’s March expansion update.
On July 28, Meta announced a joint venture with funds managed by BlackRock. The funds hold 80% and Meta 20%. Meta described approximately $14 billion in total development costs for buildings and long-lived infrastructure, supported in part by $12.5 billion of debt financing. These figures describe different financial categories and a later transaction; they cannot be treated as equivalent measures of money already spent locally. Planned capacity is expected to begin coming online in 2028. Meta’s venture announcement.
Meta’s initial leases last four years and include four renewal options, allowing occupancy for up to 20 years. The announcement also describes conditional residual-value guarantees, initially approximately $13 billion and declining over time, covering the first 16 years. The short initial lease therefore does not, by itself, establish that Meta could leave without substantial financial consequences. Nor do those guarantees establish uninterrupted operation or local employment for 20 years. Meta’s venture announcement.
A gigawatt is a measure of power, not annual energy consumption. The campus’s planned capacity should not be presented as its current electricity use. Comparisons with El Paso Electric’s system also require matching definitions: nameplate generation, dependable capacity, customer peak demand and annual energy are different measures.
Public decisions and incentives
The project’s relationship to Meta was public before the 2025 groundbreaking. The City’s December 4, 2023 presentation identified Meta by name. It also recorded a November 16 City Plan Commission recommendation and subsequent council steps concerning zoning, land sale and incentives. This does not resolve whether public notice was sufficient or participation effective. It does mean that an account saying Meta was unidentified throughout the 2023 approval process, or that no zoning process occurred, is inaccurate. City’s December 2023 presentation, pages 1–3 and 20.
The posted City agreements establish two connected incentive mechanisms. Chapter 312 provides an 80% property-tax abatement for ten years. Chapter 380 provides up to 15 annual grants equal to 80% of qualifying taxes after that abatement period, subject to the agreement’s terms. The documents therefore describe up to 25 eligible tax years for a phase, not 15 years including the initial ten. A separate 35-year agreement term constrains the timing of later phases. Chapter 312 agreement, section 2 and section 5.3; Chapter 380 agreement, “First Grant Year,” sections 3 and 5.1.
The City’s 2023 presentation modeled approximately $110 million in incentives for one phase. That was an estimate under stated development assumptions, not a cash payment already made. The original presentation itself described the 25-year incentive structure, making it a useful cross-check on the contracts. City’s December 2023 presentation, pages 11–18.
Fifty jobs is a condition for receiving full City incentives after the relevant deadline. The Chapter 380 agreement does not unconditionally require the company to achieve that number; it provides a cure period and reduced or unavailable annual incentives for specified shortfalls. That minimum is distinct from Meta’s larger operating-job forecast. The agreements’ employment definitions also matter when determining how many jobs are physically at the campus, held by local residents or supplied by contractors. Chapter 380 agreement, sections 4.3 and 6.3; Chapter 312 agreement, “Full-Time Job” definition.
Road funding involves separate commitments. The Chapter 380 document contains a $7.5 million reimbursement amount with a conditional overrun allowance. The posted road-development packet identifies a separate $5 million intersection allocation. This largely explains references to both $5 million and $12.5 million; it does not establish how much has been paid. Some material in the road packet is marked draft, so final reimbursement terms and spending require executed records. Chapter 380 agreement, section 4.6.1; City road-development packet.
State sales-tax treatment is separate from City property-tax incentives. Texas’s qualifying large-data-center program provides exemptions for specified purchases, with investment, employment, wage and power-contract conditions. It is not an exemption from all taxes. The Comptroller’s current listing associates the project with Wurldwide, Roadrunner Leap and Bluebonnet Crossing in different roles; entity names should be checked whenever linking tax benefits to ownership or operations. Texas Comptroller program; qualifying project list.
Public revenue and economic effects
The City’s public revenue figures require reconciliation. Its data-center information page lists approximately $83.4 million in net collections across five phases. An April 13, 2026 release projects more than $400 million for the City over 25 years and more than $1 billion for other taxing entities. These figures should not be combined, averaged or selectively substituted for one another without the underlying models. Their assumptions, covered taxes, development schedules and equipment replacement cycles are not sufficiently reconciled in the public explanations. City information page; City’s April 13 update.
That same April release states the City’s legal position that terminating without contractual cause could expose it to liability exceeding $1 billion. This is an attributed assessment, not a court judgment or a settled termination price. A public explanation should distinguish legal advice, negotiated obligations and adjudicated liability. City’s April 13 update.
School taxes need particular care. Texas school funding combines local revenue with state support under formulas that can change the net benefit of additional taxable value. It is misleading either to equate gross school-tax collections with an identical increase in classroom funding or to assert that every dollar is offset. A district-specific analysis must distinguish maintenance-and-operations revenue, debt-service revenue, state aid and other adjustments. Texas Education Agency’s school-funding overview; TEA’s 2025 funding-formula changes.
There are several plausible channels for local benefits: construction earnings, ongoing wages, purchases from regional businesses and tax collections. Their size depends on where employees live, where suppliers operate, the pace of development and public costs. Comparing a project only with previously undeveloped land does not establish the best alternative use of that land or of infrastructure spending. Conversely, an incentive’s face value does not establish that the same development and tax base would have occurred without it.
New supplier activity deserves attention without unsupported attribution. On September 10, Wiwynn and Amazon announced a Socorro manufacturing expansion associated with data-center equipment and expected nearly 1,000 additional manufacturing jobs by the end of 2027. This is a separate company forecast and a regional supply-chain development. It does not demonstrate that Meta’s campus caused those jobs. Wiwynn’s September 10 announcement.
Water supply, consumption and drought
Three water figures answer different questions. The posted 2023 agreement’s Tier III table lists 1.5 million gallons for an average day and a 2.5-million-gallon daily capacity allocation. EPWater’s newer public explanation estimates approximately 480,000 gallons per day of average use and describes 2.5 million as an extreme-heat maximum. Contractual capacity, modeled average use and actual metered consumption must remain separate. Water agreement, Exhibit B; EPWater’s Meta Q&A.
The agreement offers substantial service protections, but describing the project as categorically exempt from drought constraints overstates its text. Section 5(a) contains public-health and applicable-law qualifications; drought-related meter approval and consent to reservation reductions address different matters. EPWater separately says ordinary drought restrictions apply. A complete public explanation should show both the contract and the utility’s interpretation rather than imply that their relationship has been conclusively resolved. Water agreement, section 5(a); EPWater Q&A.
EPWater says the project pays the rates applicable to very large water users. A nominal consideration amount in a service agreement is not the price of the water delivered. The utility says Texas law prohibits it from publicly disclosing an individual customer’s use and that public disclosure would be at Meta’s discretion. Its statement that it had not received a conservation plan as of July 2026 should retain that date; it is not proof that no plan exists in September. EPWater Q&A.
Cooling terminology can obscure water demand. A closed internal loop can transfer heat to equipment that evaporates water outside that loop. Other designs reject heat primarily to air and use little or no water for cooling, potentially with different energy requirements. Water treatment, sanitation and electricity generation can add other demands. The relevant questions are the actual cooling configuration, seasonal consumption, water source and efficiency measured under El Paso conditions. U.S. Department of Energy’s cooling-water guidance.
Meta’s August 2026 explanation of closed-loop cooling describes its technology broadly. Microsoft has also announced a design avoiding water consumption for cooling in new facilities. Neither establishes the equipment installed or annual water performance at El Paso. Project-specific plans and measured results are needed before making technology comparisons. Meta’s cooling explanation; Microsoft’s design announcement.
Water replenishment also differs from reducing withdrawals at a campus. Restoration or conservation projects may provide benefits, but the accounting requires location, timing, additionality and verification. A company’s aggregate replenishment goal does not automatically establish a gallon-for-gallon offset in the aquifer or season affected by a particular site. Meta’s water program.
El Paso’s wider water strategy includes conservation, groundwater management, desalination and reuse. EPWater currently describes its Pure Water Center as a roughly $290 million project planned to begin operation in 2029 with capacity for 10 million gallons per day. Those are utility plans, not present supply, and the facility should not be portrayed as available today or dedicated to Meta. USGS groundwater research provides historical aquifer evidence but does not, by itself, forecast this campus’s marginal effects. EPWater’s Pure Water Center; USGS Hueco Bolson investigation.
Electricity supply and customer costs
A large electricity customer can pay substantial utility revenue while requiring new generation and transmission. Whether other customers benefit or face additional costs depends on construction spending, operating costs, minimum payments, contract duration, guarantees and regulatory decisions. Neither the project’s size nor a company’s general promise answers the cost-allocation question alone.
The proposed McCloud facility is central to this debate. In June 2026, the City described a 366-megawatt natural-gas proposal with an estimated cost of $551.8 million and recommended that regulators deny the application. It questioned procurement, alternatives, costs and long-term need. Those are the City’s positions in a contested proceeding, not findings by the Public Utility Commission of Texas. ERock separately describes the facility as bridge power with a possible later grid-connected role. City’s June 18 statement; ERock’s June fact sheet.
The City’s July 7 statement announced a July hearing and anticipated an October proposal for decision and December commission vote. That announcement does not establish the latest procedural status. Precise household bill impacts, the final service agreement and current tariff terms require the underlying filings and orders; the available evidence here does not establish an approved or predicted household increase. City’s July 7 statement.
Separate proceedings also need separate explanations. Council’s April 27 minutes identify docket 59611 as the High Load Factor Large Power Service tariff application and authorize intervention. A July 20 announcement concerns proposed changes to the existing Economic Development Rate. The two should not be merged simply because both concern large customers. April 27 minutes, page 3; City’s July 20 rate statement.
El Paso Electric participates in the Western Interconnection. Several prominent Texas large-load measures expressly concern ERCOT: provisions of Senate Bill 6 specify the ERCOT power region, and the Governor’s August audit announcement addressed ERCOT’s interconnection process. That geographic distinction matters, but does not mean El Paso has no utility regulation or ratepayer protections. EPE 2027–2036 system plan, section 3; SB 6 enacted text; Governor’s August 3 announcement.
Meta’s August 10 statement broadly commits to paying for energy, water and needed grid infrastructure. Its wording does not limit the commitment to ERCOT. The outstanding El Paso question is how such commitments appear in enforceable local agreements and regulatory orders. Meta’s August 10 statement.
Air quality, noise and heat
Power generation and cooling can affect air emissions, noise and nearby temperatures, but their impacts depend on equipment, operating hours, controls, distance and weather. Permit limits are not the same as actual emissions. Countywide air-quality classifications are pollutant-specific and should not substitute for a local exposure assessment. TCEQ’s El Paso air-quality planning page.
A May 2026 study of four Phoenix-area data centers reported localized warming during a small set of field traverses, including average downwind–upwind temperature differences around 0.7–0.9°C and peaks around 2.2°C. This is useful evidence that waste heat can warrant investigation. It is not an El Paso measurement or a basis for drawing an identical impact radius around a proposed campus; the accessible institutional abstract does not permit a full independent review of its methods. Sailor, Abolhassani and Martin, institutional research record.
Useful local evidence would include permit applications and final conditions, operating-hour assumptions, baseline sound measurements, nearby sensitive locations and repeated monitoring under different weather conditions. Until those data are obtained and assessed, a categorical claim that impacts are negligible or severe would go beyond the evidence presented here.
Southern New Mexico and Fort Bliss
Project Jupiter in Doña Ana County has changed materially since its original announcement. Oracle’s June 2026 account describes a revised power plan using fuel cells in place of previously proposed turbines and diesel equipment. Its September 8 update announces a request for proposals for 2 gigawatts of renewable energy, with delivery contemplated over 2027–2031 and subject to required approvals. Fuel cells using natural gas are distinct from renewable electricity; a procurement request is not evidence of operating renewable generation. Oracle’s June power-plan update; Oracle’s September 8 update.
Oracle’s current project page says customer delivery is expected to begin in the first half of 2027. Its reported construction workforce, tax payments and community commitments should be identified as company statements and reconciled with county records where possible. Earlier completion dates and workforce projections should remain in a dated history rather than silently becoming current facts. Oracle’s data-center information; July 28 Jupiter update.
A congressional statement dated August 24 describes a New Mexico Supreme Court stay that temporarily halted Project Jupiter’s air-permit proceeding, including the hearing previously set for September 14. The Department of Justice’s September 3 statement supports challenges to the permitting process; it does not establish that the stay was lifted. A new hearing date requires a later official order. These procedural developments do not decide whether the project ultimately may operate. August 24 congressional statement; New Mexico DOJ’s September 3 statement.
The Army announced a conditional agreement for hyperscale data-center development on March 26, 2026, including approximately 1,384 acres at Fort Bliss. Conditional selection does not establish final construction authorization, operating capacity or delivered jobs. Precise multi-gigawatt and employment figures circulated elsewhere should remain out of factual summaries until their project-specific documentation is available. Army’s March 26 announcement.
Broader research and its limits
Historical experience offers questions to investigate rather than a ready-made verdict for El Paso. Virginia’s legislative research agency examined data-center economic effects, electricity demand, water and noise in 2024. Its work is valuable because it examines benefits and infrastructure pressures together, but Virginia’s industry scale, power market and tax arrangements differ from El Paso’s. Virginia JLARC’s 2024 report.
New 2026 working papers examine local employment, economic activity, housing and electricity prices around data-center development. Their reported findings include economic gains alongside possible price pressures. Working papers and accessible abstracts should not be treated as settled consensus, and national average estimates should not be converted into a forecast for a particular El Paso household. Yue and Zeng working-paper record; Alvarez, Argente, Chow and Van Patten, NBER working paper.
An August 2026 Lawrence Berkeley National Laboratory report inventories 55 large-load tariffs, including proposed and approved designs, using an inventory principally from March and April. It provides concrete examples of minimum payments, security and other mechanisms used to address utility costs. It does not establish which terms apply in El Paso’s Texas proceedings; its EPE Rate 50 entry concerns New Mexico. Berkeley Lab’s tariff report.
Local policy and outstanding decisions
On July 20, the City adopted a data-center policy framework. Its announcement describes further work on ordinance changes and other implementation. Adoption of a framework should not be represented as proof that every proposed zoning or environmental provision is already enforceable. Participation figures from public outreach also describe participation, not a representative opinion poll. City’s July 20 framework announcement.
Advance material for the September 14 council work session describes continuing community-agreement discussions, possible workforce and neighborhood initiatives, and planned district meetings. It expressly indicates that discussions had not identified a specific El Paso allocation or funding commitment from Meta’s national community fund. These are proposed or developing arrangements, not an adopted agreement or money awarded to El Paso. Advance council presentation, pages 2–3 and 7–10.
The most consequential next evidence is specific: updated tax models, incentive compliance reports, executed utility agreements and amendments, final regulatory orders, project-specific water plans, environmental permits and measured operating results. For community benefits, the relevant record is a signed commitment with amounts, deadlines, beneficiaries and reporting requirements. Publishing these documents alongside understandable explanations would allow residents to assess both performance and unanswered questions as development proceeds.