Issued by:

GES Coalition, Tepeyac Community Health Center, Green House Connection Center, Cultivando, Womxn from the Mountain, and 215 neighbors in Elyria-Swansea and Globeville.

Endorsed by:

The GrowHaus, GroundWork Denver, Communications Workers of America Local 7777, Denver Classroom Teachers Association (DCTA), Coloradans for the Common Good (CCG) Moms Clean Air Force Colorado

Sent to CoreSite on February 13, 2026

Written response requested by February 24, 2026

Primary contact: Robin Reichhardt, GES Coalition

Organizing lead: Alfonso Espino, GES Coalition

Why we’re writing 

CoreSite’s DE3 facility plans to operate in the Elyria-Swansea neighborhood for decades. At the proposed scale described, DE3 will function as round-the-clock industrial infrastructure: major electricity demand, backup generation, mechanical systems, potential major water use and water pollution by discharge, constant noise and heat rejection, and long-term operational and emergency risk.

DE3 is not “just one building.” It sits inside a durable industry shift: colocation is essentially a secure industrial warehouse for computing that sells “reliability & connectivity” as the product, and the AI era is pushing demand for high-density power and advanced cooling (often water/heat intensive).

CoreSite’s own sponsored industry research frames colocation as the backbone of “hybrid IT,” and reports that 98% of IT leaders have adopted or plan to adopt a hybrid IT model, meaning mega-load proposals will keep coming unless Denver and Colorado set clear, enforceable rules.

That’s why marketing claims can’t substitute for governance: no NDAs, full disclosure, and enforceable terms in writing must be the baseline.

But DE3 is not located in a blank industrial zone. DE3 will be across the street from more than 150 future units of senior housing, 150 units of mixed-income housing, and is located less than a block from Tepeyac Community Health Center, Johnson Recreation Center, and Elyria Park, and is located only a single block from the Elyria residential core. Besides its hyper-proximity to Elyria, it is also located less than 1.5 miles from the majority of households in the 80216 ZIP code (Globeville, Elyria–Swansea), widely documented as one of the most environmentally burdened ZIP codes in the country.

This cumulative burden is not accidental. It is the outcome of colonial dispossession and extraction, then decades of zoning, redlining, highway construction, and industrial siting that concentrated pollution next to working-class homes, alongside the legacy of the Vasquez Boulevard/I-70 Superfund site, a 4.5-square-mile smelting contamination footprint affecting multiple neighborhoods, including Elyria–Swansea. (CDPHE)

That is why “good neighbor” cannot mean PR, voluntary promises, or private conversations. In a frontline community like ours, being a good neighbor must be enforceable, measurable, publicly verifiable commitments, with real monitoring, public reporting, and real consequences.

Why this matters right now (HB26-1030 and the cumulative impacts of environmental racism)

Colorado lawmakers are currently debating HB26-1030, which would create a new data-center development authority and offer a 100% state sales and use tax exemption for 20 years (with potential extension) for certified data center projects, while also shaping how utilities plan and finance large-load infrastructure. (Colorado General Assembly)

In an already overburdened ZIP code like 80216, policy that accelerates data-center growth without ironclad, neighborhood-enforceable protections risks becoming the next chapter in a long pattern: public systems retooled for private industry, while frontline residents carry the health, noise, heat, traffic, and reliability risks.

This is the immediate terrain: Xcel is normalizing “grid expansion” as a core investment priority, with Denver alone framed as needing roughly ~$1.2B in distribution & transmission capacity investments (indicative, 2026–2035). The risk is predictable: data centers get framed as “growth,” while households absorb rising costs unless enforceable cost-causation rules are locked in.

At the same time, a national political opening is emerging: major national leaders and big tech companies are increasingly saying data centers should pay their own way and should not raise household electricity bills. That makes “no cost-shifting, no subsidies, enforceable accountability” politically mainstream, not “radical.”

There should be no giveaways and no backroom deals. If data centers want to build here, they must pay the full costs they trigger, and the public must be able to see and enforce the terms.

So we’re being direct: A good neighbor does not take public incentives without community protections and benefits that match the footprint and the risk.

And a good neighbor does not claim “innovation” branding, and does not receive expedited treatment at the cost of the neighborhoods living in the shadow of redlining and historic exclusion from public spending. No special exceptions should be pursued or accepted for DE3 unless enforceable protections and community-defined benefits are secured in writing, with real monitoring, public reporting, and consequences.

The Community Floor: GES DE3 Data Center Platform 

What follows is our platform that can guarantee a positive impact on the health and well-being of the neighborhood it calls home. The Good Neighbor Agreement must implement these as binding terms, campus-wide and end-state (not phase-by-phase promises).

1) No expansion / no “phasing” loopholes

No increase in DE3’s footprint, capacity, power demand (MW), water demand, diesel inventory/runtime, emissions, noise, heat discharge, or truck traffic beyond what the public has evaluated and consented to. No stacked approvals that quietly add up to an end-state nobody agreed to. The Agreement governs the full end-state campus.

2) Transparency is the baseline: disclosure by default (no NDAs)

No NDAs and no secrecy on public-interest impacts: grid, water, diesel generators, emissions, noise, heat, emergency ops, and traffic/logistics.
CoreSite must maintain a public, continuously updated DE3 Disclosure Packet (campus-wide/end-state) with real numbers (not percent-only claims), including:

• MW ramp timeline & end-state capacity

• interconnection status & upgrade needs

• cooling type & water-use ranges

• withdrawal sources & discharge pathways/permits

• generator inventory,  hour meters, fuel & testing/usage logs

• mitigation measures & compliance logs

• emergency operations plan

• Health Equity Analysis (and updates): methods, assumptions, raw monitoring data, and mitigation/penalty triggers

3) Health Equity Analysis before permits or energization (no “study later”)

Before CoreSite files for, seeks acceptance of, or relies on any approvals—land use/building permits, air/generator permits, water taps/discharge approvals, or utility interconnection/energization—CoreSite must complete an independent Health Equity Analysis (HEA) that is campus-wide and end-state.

Minimum HEA requirements:

• Independent reviewer paid by CoreSite, selected with the resident governance body (not developer-controlled).

• Cumulative burden & incremental impact: quantify DE3’s added impacts on NOx/PM, noise/sleep, heat, traffic, water withdrawals, and blowdown/discharge, including routine ops and worst-case/emergency scenarios.

• Who carries the burden: children, elders, medically vulnerable residents, renters, and households closest to the site.

• Alternatives analysis: cooling design, potable minimization, generator configuration/runtime, load shape, and noise/heat controls—show the least-harm approach.

• Binding outcomes: measurable caps/limits, continuous monitoring, and automatic corrective actions and penalties tied to exceedances.

• Update rule: any scope change (MW, water, generators, discharge pathway, noise/heat profile, truck traffic) triggers an HEA update before any new approvals.

No permits-by-default: no approvals first, analysis later.

4) No cost-shifting: ratepayers pay $0 (build what’s needed up front)

100% cost causation: CoreSite/tenants must pay for all electric system upgrades triggered by DE3—make-ready, feeders, transformers, protection equipment, and reinforcements—via an enforceable payment mechanism with public accounting.

If a new substation (or major facility) is needed, CoreSite must fund and build it up front before additional load is energized.

No back-door pass-throughs. No secret discounts. Colorado grid-modernization cost recovery pathways (including those enabled through recent state law, such as SB 24-218) must not become a subsidy for DE3. DE3 must be locked into full cost causation, no stranded-cost risk shifting, and full disclosure of every upgrade it triggers.

5) Clean air: reject Tier 2 diesel; hard limits & public monitoring

Full campus generator disclosure (make/model/year, kW, EPA tier, permits, locations, hour meters, testing plan). Minimum: Tier 4 Final (or better) for all engines, no “temporary/ rental/ mobile” loopholes; Tier 4 standards are designed to reduce key pollutants like NOx and PM dramatically compared to earlier tiers. (DieselNet)

Emergency-only means emergency-only (no peak shaving/grid services/profit runtime). Binding runtime caps & independent fenceline monitoring (NOx/PM) with a public dashboard, automatic exceedance alerts, and automatic penalties.

Quantification: a ~2,000 kW diesel generator can burn on the order of ~129–167 gallons/hour, depending on load; 50 hours/year of testing can mean ~6,450–8,350 gallons per engine per year, and diesel combustion emits about ~22.5 lbs CO₂ per gallon (plus co-pollutants). (Clifford Power Systems, Inc.)

Traffic equivalency: A ~2,000 kW diesel generator burning ~129–167 gallons/hour emits roughly ~2,900–3,750 lbs of CO₂ per hour of testing. Using the EPA’s average passenger-vehicle tailpipe rate (~400 g CO₂/mile), that’s about the same CO₂ as ~55–70 typical cars driving on I-70 for one hour at freeway speed (~60 mph). (US EPA) Over 50 hours/year of testing (~6,450–8,350 gallons/year per engine), that’s roughly ~66–85 metric tons CO₂/year. (EIA) On the Central 70 stretch (I-25 → Chambers Road ~200,000 vehicles/day over ~10 miles, so one engine’s annual testing CO₂ is roughly equivalent to ~16,000–21,000 cars making a single 10-mile trip, or about ~2–2.5 hours of all Central-70 daily traffic. (Colorado Department of Transportation) And that’s only CO₂, it excludes the localized co-pollutants (NOx/PM) that are the frontline health threat in 80216. 

Generator operations are increasingly being treated as a public health & regulatory issue, not a private facilities decision, and “backup power” claims can mask routine or expanded on-site generation unless rules are explicit. This is why DE3 must have hard runtime caps, transparent logs, and enforceable penalties.

There is already national precedent for enforcement against tech companies’ on-site generation, reinforcing that generators aren’t a “trust us” issue; they are a compliance and public-health issue.

6) Noise & light: enforceable standards that protect sleep

Set enforceable dB(A) limits at the property line and nearest sensitive receptors, with stronger nighttime standards. Require:

• continuous, public noise monitoring

• engineering controls

• automatic corrective actions/penalties for exceedances
  No light spill: shielding, glare controls, enforceable response timelines, and a public issue log.

7) Water & heat accountability as public-health infrastructure (including discharge)

Require campus-wide water caps with automatic penalties; reuse-first cooling with potable minimization; and drought-stage operating rules that protect household reliability and affordability.

Require independently verifiable metering and a public dashboard showing daily totals, monthly peaks, annual totals, and source mix (potable vs reuse/other).

Discharge transparency is required: the Disclosure Packet must include:

• withdrawal sources (potable/reuse/other)

• blowdown volumes (daily/monthly/annual)

• discharge destination & permits/approvals

• chemicals/additives & expected concentrations

• thermal impacts & mitigation

Heat is part of the cooling footprint: waste heat must deliver a measurable neighborhood benefit with milestones and verification, or provide an enforceable equivalent (e.g., neighborhood electrification/efficiency & heat-mitigation fund). No voluntary language.

8) Construction rules that prevent disruption and harm

Enforceable rules for: truck routes, delivery windows, peak volumes, dust suppression, anti-idling enforcement, safe routing, and no nighttime construction except true emergencies. Require a hotline, response standards, posted schedules/compliance logs, a public issue log, and penalties for repeat violations.

9) No contracts that fuel surveillance and over-policing

DE3 must not support mass surveillance, dragnet policing, deportation operations, or war/occupation decision systems. Require tenant/use transparency, civil-liberties guardrails, and enforceable off-ramps for violations.

This is not abstract: Palantir’s work has been widely reported in connection with ICE and deportation infrastructure, including recent reporting on an “ImmigrationOS” initiative. (American Immigration Council)

And concerns about surveillance tools like ALPR networks are active and real, including reporting and civil liberties analysis about Flock-style systems and data access/sharing. (San Francisco Chronicle) 

In Denver, audit logs obtained from the Denver Police Department show Flock ALPR camera data was used in immigration-related “national searches” more than 1,400 times (June 2024–April 2025) (Colorado Newsline)

CoreSite itself markets colocation services to the public sector (including government). (CoreSite)

So the Good Neighbor Agreement must include use/tenant transparency and enforceable prohibitions, not trust-me language.

10) Binding agreement, resident governance & real investment (with enforcement capacity)

A binding Good Neighbor Agreement with timelines, monitoring, transparency, and consequences (penalties/clawbacks/third-party enforceability). Establish resident-governed oversight with data access and authority, and fund independent technical capacity selected by the community. Provide sustained annual community investment tied to measured impacts.

Include a Reparative Health Equity & Enforcement Fund that pays for:

• independent air/noise/water monitoring & analysis

• resident-selected technical experts

• legal enforcement capacity (so neighbors can enforce if agencies fail)

• home filtration/health protections & heat mitigation

• independent Health Equity Analysis (baseline and updates) &  community-controlled review

What we’re asking CoreSite to put in writing (Implementation terms)

A) No NDAs; no gag clauses; no forced complicity

We will not accept any agreement that restricts or hides decisions that impact the public good.

B) Reject the “jobs” argument as justification for harm; commit to reparative reinvestment instead

Data centers are highly capital-intensive and typically create limited permanent operations employment compared to the scale of infrastructure and public-system stress; most jobs are short-term construction. (hamminstitute.org)

So CoreSite must not use jobs claims to justify burden shifting onto an already overburdened community. The right frame here is reparative reinvestment: community health protections, monitoring, and long-term neighborhood wealth-building that residents control.

C) Water: disclose BOTH use and discharge, with binding caps and drought-stage protections

Include the full withdrawal & blowdown/discharge accounting described above, with monthly public reporting and enforceable caps, acknowledging that cooling tower strategies can drive significant water demand and wastewater (blowdown). (Congress.gov)

Because water is a high-leverage public-health infrastructure issue, disclosure and enforceable standards must be the baseline, not voluntary corporate promises.

D) Diesel: no Tier 2; emergency-only defined narrowly; pathway to eliminate combustion

No Tier 2 engines. Minimum Tier 4 Final (or better) plus enforceable runtime caps, fenceline monitoring, and automatic penalties. (DieselNet)

CoreSite must also present a time-bound plan to reduce and ultimately eliminate routine on-site combustion risk(battery-based resilience / alternative architectures), rather than normalizing diesel as “the cost of doing business” in 80216.

And because generator enforcement is already a national flashpoint, DE3 must treat generator operations as a public compliance issue with hard caps, public logs, and real penalties, not a private “backup” preference.

E) Substation and grid upgrades: who pays, what’s required, and how long it will really take

CoreSite must provide written answers (with Xcel documentation) to:

• Is a new substation required for end-state DE3 load? If yes, what voltage, where, and by when?

• What is the critical path timeline (studies, design, permitting, procurement, construction, commissioning)?

• What load will be energized before the substation is in service, and what are the reliability and cost-shifting risks of that sequencing?

• What is the binding mechanism that guarantees 100% cost causation and prevents later “surprise” cost recovery from ratepayers?

Given Colorado’s Grid Modernization Adjustment Clause (GMAC), pathways for cost recovery through rates, DE3 must not be allowed to trigger “Type 1/Type 2” distribution investments that later become socialized, this is why the binding mechanism must explicitly lock 100% cost causation & no stranded-cost risk shifting.

F) Community welfare assurances (measurable, audited, enforceable)

The agreement must include community welfare protections that are real and not symbolic:

• baseline & ongoing health-risk indicators (air/noise/heat)

• home mitigation (filtration/ventilation support prioritized for vulnerable residents)

• emergency notification standards (multi-language, rapid, documented)

• a resident-controlled grievance process with response deadlines and remedies

• annual third-party compliance audits with publicly downloadable datasets

G) Carbon and climate: reduce emissions in reality, not offsets

CoreSite must commit to:

- real, additional clean energy procurement tied to hourly load growth (not just annual RECs)

- load management commitments that reduce peak stress and avoid fossil peaker reliance

- transparent reporting that ties MW ramp, grid upgrades, and clean energy sourcing together

A practical path forward (proposed timeline)

We propose that by February  24th

• CoreSite responds to this proposal in writing and at the scheduled public meeting.

• CoreSite identifies authorized decision-makers for water (use & discharge), diesel/backup architecture, monitoring, and agreement authority.

Within 15 days

• CoreSite publishes the DE3 Disclosure Packet (campus-wide, end-state).

• CoreSite provides draft enforceable terms for diesel limits, monitoring & penalties.

• CoreSite provides cooling design, water withdrawal & blowdown/discharge accounting, and draft enforceable cap terms.

Within 30 days

CoreSite meets with a neighborhood delegation and decision-makers to finalize:

• monitoring and operating limits

• public reporting cadence

• resident governance/oversight structure

• the draft binding Good Neighbor Agreement (with enforcement)

Closing

We’re offering this proposal in the spirit of being a good neighbor: we want a stable relationship that reduces conflict, creates clarity, and protects the people who live next to this facility every day. But stability requires truth and enforceability, especially in 80216, where cumulative pollution burdens have been documented for decades. If CoreSite truly wants to be a good neighbor, this is a clear way to show it, with commitments that are understandable, enforceable, and durable.