$GLXY $CRWV $AEP Galaxy Digital announced completion of an ERCOT Large Load Interconnection Study (LLIS) and receipt of ERCOT approval for an additional 830 MW of computing demand at the Helios data center campus in West Texas, with a parallel execution of a service agreement with AEP Texas for the incremental capacity and Wind Energy Transmission Texas, LLC (WETT) identified as the transmission interconnection provider that facilitated the LLIS studies. The company stated that this brings total LLIS-complete, ERCOT-approved, utility-contracted power capacity at Helios to over 1.6 GW, effectively doubling approved campus power capacity and expanding the platform for multi-tenant development; the approval was described as following ERCOT review of required LLIS study elements, specifically steady-state and stability studies. Construction for the initial phase under a long-term lease agreement with CoreWeave was stated to be underway, with initial power delivery targeted to begin in early 2026.
The primary investment-relevant conclusion is that interconnection-risk de-risking at the 830 MW increment level has meaningful option value in an ERCOT environment where large-load requests are extensive but only a limited subset progresses to executed agreements, model inclusion, and energization. However, “LLIS-complete” and “ERCOT-approved” should not be treated as synonymous with unconditional authorization to energize or operate at full contracted demand; ERCOT’s large-load framework explicitly contemplates additional prerequisites for energization (including model/telemetry/stability assessments and ride-through evaluations) and the potential for curtailment under reliability-driven procedures, especially for large electronic loads. The magnitude of a potential single-site loss event is material at the system level; ERCOT has indicated that the maximum sustained system-wide load loss without exceeding the post-contingency frequency limit is approximately 2,600 MW, implying that a full 1.6 GW trip would represent approximately 61.5% of that threshold and would likely be treated as a non-trivial stability contingency in operational and planning studies.
EVENT DETAILS AND FACT PATTERN
The press release (Jan 15, 2026, 09:15 ET) describes a discrete milestone: completion of a LLIS and ERCOT approval for an additional 830 MW of computing demand at Helios, alongside a service agreement executed with AEP Texas for that additional capacity and identification of WETT as the transmission interconnection provider. Total “LLIS-complete, ERCOT-approved, and utility-contracted power capacity” at Helios was stated as over 1.6 GW after the approval, with the incremental approval framed as enabling multi-tenant partnerships and positioning Galaxy among large and fast-growing data center developers in North America.
The company stated that the incremental approval followed ERCOT review of LLIS study elements consisting of steady-state and stability studies. In ERCOT’s framework, these studies are intended to assess reliability impacts and identify needed transmission facilities or operational limits required to avoid operating-standard violations. The explicit mention of steady-state and stability studies suggests that the incremental 830 MW was evaluated not only for thermal/voltage loading but also for dynamic performance (frequency/voltage stability), which has become a focal point for large electronic loads.
INTERCONNECTION STATUS: WHAT “LLIS-COMPLETE, ERCOT-APPROVED, UTILITY-CONTRACTED” DOES AND DOES NOT MEAN
ERCOT’s large-load interconnection guidance defines a “Large Load” as aggregate peak demand of 75 MW or greater at a single site behind common points of interconnection or service delivery points. For loads meeting applicability requirements, the formalized process (associated with NPRR1234 and PGRR115) establishes the LLIS as the core reliability study mechanism for each new Large Load seeking to interconnect to the ERCOT system; the PGRR115 issue description also indicates that large loads evaluated via LLIS must be included in ERCOT’s quarterly stability assessment prior to initial energization and that additional requirements exist for inclusion in the Network Operations Model and for required interconnection equipment.
ERCOT’s Q&A emphasizes that completion and ERCOT approval of final interconnection studies is followed by execution of binding agreements (such as an interconnection agreement) between the interconnecting large-load customer and the applicable transmission service provider, and that energization is conditioned on completion of specified requirements (including agreements, notice-to-proceed and financial security for required interconnection facilities, model inclusion, operational telemetry, quarterly stability assessment, and commissioning-plan compliance). The press release’s use of “utility-contracted” and the explicit “service agreement with AEP Texas” for the incremental capacity are directionally consistent with that “post-study” phase of the process, where binding agreements and funding/security become gating items for construction and energization.
The same ERCOT guidance also identifies time-bound compliance gates that can create schedule and cancellation risk if not met. Planning Guide Section 9.4/9.5 requirements are described as needing to be satisfied within 180 days in order to avoid potential cancellation pathways, and additional timing provisions can require reassessment if the project does not achieve initial energization within 365 days of the requested initial energization date used in the LLIS study. These procedural elements increase the importance of aligning engineering, procurement, construction, and commissioning schedules with the LLIS assumptions and the Load Commissioning Plan ramp schedule.
GRID PERFORMANCE AND RELIABILITY CONSIDERATIONS FOR LARGE ELECTRONIC LOADS
ERCOT has specifically highlighted voltage ride-through and stability risk for “large electronic loads” (LELs), including data centers and cryptomining loads, noting observed tendencies for some LELs to trip during voltage excursions, with potential system stability implications. ERCOT’s June 23, 2025 market notice requested voltage ride-through capability information from customers seeking interconnection of data center and cryptomining large loads 75 MW or greater and stated that sudden loss of load can contribute to local voltage and system frequency impacts that could trigger additional generator/load trips and potential cascading outages. In this context, large-load projects face not only traditional steady-state constraints but also an increasingly explicit dynamic-performance compliance layer.
A quantitative framing from ERCOT is particularly relevant for a single-campus 1.6 GW target: ERCOT stated that the maximum amount of load loss that can be sustained system-wide without exceeding the post-contingency frequency limit is approximately 2,600 MW. If the Helios campus were ultimately energized near its stated over-1.6 GW contracted/approved capacity, an unmitigated site-wide trip could represent approximately 61.5% of that system-wide “sustained load loss” envelope, implying that the project’s ride-through, protection settings, and staged commissioning controls are likely to be treated as high priority in ERCOT’s interim ride-through evaluation and in any future voltage/frequency ride-through standards.
ERCOT’s Q&A describes an “Interim Ride-Through Evaluation” as a condition for energization and indicates that ERCOT may allow energization subject to curtailment if modeled ride-through failures would result in violation of a System Operating Limit (SOL) or Interconnection Reliability Operating Limit (IROL), contingent on ERCOT implementing curtailment tools and procedures; even where no SOL/IROL violation is predicted, future curtailment exposure can exist as additional loads connect in the same area. This creates a non-trivial operational risk dimension for AI/HPC tenants whose economics and service-level commitments may assume high availability and limited curtailment, and therefore increases the value of technical mitigations that improve ride-through performance and reduce the likelihood that the campus is modeled as a trip-prone contingency.
Galaxy’s own technical positioning for Helios indicates design intent consistent with hyperscale AI/HPC requirements, including 100% uninterruptible power supply and battery backup and electrical topologies described as distributed redundant and block redundant systems, with targeted rack densities of 130–140 kW and claimed low-latency connectivity to Dallas. While these characteristics primarily address tenant uptime and compute density, they can also be relevant to the ride-through and protection philosophy at the facility level, depending on how UPS and switchgear coordination is implemented relative to grid-side voltage/frequency excursions. Actual grid-beneficial performance (ride-through vs trip) remains design- and settings-dependent and should be treated as a due-diligence focus rather than a guaranteed consequence of “UPS present.”
DEVELOPMENT TIMELINE AND DELIVERY READINESS
The company stated that construction to support the initial phase under the CoreWeave long-term lease is underway and that delivery of initial power remains on track for early 2026, with the company also scheduling 4th quarter and full year 2025 earnings for Feb 3, 2026. The near-dated “initial power” milestone is important because it will provide empirical evidence on commissioning execution, utility coordination, and any incremental conditions imposed by ERCOT for energization (telemetry, model updates, ride-through validation, and operational constraints).
ERCOT’s process design implies that energization and ramp are expected to follow a Load Commissioning Plan that can include demand limits tied to the final studies and that requires ERCOT approval before increasing demand after identified transmission upgrades become operational. For a campus scaling from initial energization toward 1.6 GW, the most likely path is multi-step energization across multiple commissioning increments rather than an instantaneous ramp to full demand; this is consistent with ERCOT’s stated need to monitor consumption limits and approve load increases as the system model and infrastructure upgrades evolve. Consequently, the “over 1.6 GW” figure is best interpreted as an upper bound of approved/contracted potential capacity rather than a near-term delivered load profile.
COMMERCIAL STRUCTURE, ANCHOR TENANT, AND FINANCING CONTEXT
Public Galaxy disclosures provide additional context on the revenue model and the relationship between “gross power” and “critical IT load.” In a May 2025 SEC-filed press release, Galaxy disclosed Phase I (133 MW critical IT) and Phase II (260 MW critical IT) commitments with CoreWeave (total 393 MW critical IT), with Phase I expected delivery in 1H26 and Phase II expected delivery in 2027; the same disclosure stated anticipated average annual revenue of approximately $900M for the combined phases and “expected EBITDA margins” of 90%, and noted approximate gross power requirements of 200 MW for Phase I and 400 MW for Phase II. These ratios imply that 393 MW of critical IT load maps to approximately 600 MW of gross power capacity (a gross-to-critical multiplier of approximately 1.53x), reflecting facility overhead (cooling, electrical losses, redundancy, ancillary systems). These are company expectations and should be treated as management estimates conditioned on build-out, utilization, and contract structure.
In an Aug 2025 announcement, Galaxy stated that it closed a $1.4B project financing facility with an 80% loan-to-cost structure and a 36-month term to fund the 1st phase of Helios AI data center development and that CoreWeave had committed to the full 800 MW of approved power capacity at Helios; the same announcement stated an expectation of average annual revenue of more than $1B over the 15-year term of the CoreWeave agreements based on committed terms, internal capex estimates, and full utilization. The financing terms indicate that the initial-phase build is being leveraged in a project-finance style structure and that the scale of power approvals is being treated as a bankable underpinning to tenant leasing economics.
SCALING IMPLICATIONS OF THE INCREMENTAL 830 MW APPROVAL
Using Galaxy’s disclosed gross-to-critical relationship (approximately 600 MW gross for 393 MW critical) as a directional benchmark, the incremental 830 MW of approved computing demand could correspond to approximately 544 MW of incremental critical IT load capacity, and total “over 1.6 GW” gross could correspond to approximately 1,048 MW of critical IT capacity, subject to meaningful uncertainty depending on final cooling/electrical design, redundancy tier, and rack-density-driven overhead. This framing is useful because hyperscale AI/HPC leasing demand is typically contracted and priced based on deliverable critical IT, while interconnection capacity is constrained by gross power at the point of interconnection. The incremental approval therefore has the potential to expand monetizable critical IT supply materially beyond the previously disclosed 800 MW approved-power envelope, but it simultaneously increases absolute execution and capex requirements for shells, electrical distribution, cooling, water strategy, and fiber.
The incremental 830 MW approval also increases the importance of load-shaping and risk-sharing mechanisms in tenant contracts. A high-load-factor AI/HPC campus that ramps toward >1.0 GW critical IT likely imposes a large and relatively flat demand profile on the local grid; absent long-term hedging or robust pass-through structures, exposure to ERCOT price volatility can be material. Galaxy’s history at the Helios site includes prior cryptomining operations, and the company has disclosed impairment and wind-down costs associated with mining at Helios; this background underscores that power price volatility and operational curtailment can materially affect economics when load is not fully contracted on a pass-through basis. Tenant lease terms are not fully disclosed in the cited releases, so conclusions on the degree of energy-price risk transfer should be treated as unresolved and central to underwriting.
ERCOT SYSTEM CONTEXT: DEMAND FORECASTS, QUEUE SCALE, AND ATTRITION RISK
ERCOT’s system planning materials highlight why a “study-complete and utility-contracted” milestone can be disproportionately valuable. ERCOT’s 2025 Report on Existing and Potential Electric System Constraints and Needs states that the 2025 TSP-provided load forecast projects demand of 218 GW for 2031 versus 150 GW in the 2024 forecast, with the increase attributed to future data center load growth attested to by TSP officers (“Officer Letter Loads”). The same report describes ERCOT’s use of an adjusted forecast methodology that reduces officer-letter loads for planning purposes, including a specific adjustment for data center loads (an additional 49.8% reduction on top of other reductions, with an example reducing a 100 MW data center load to 27.6 MW). This approach is effectively an institutional acknowledgement of high uncertainty and high attrition in “announced/attested” load pipelines.
In the same ERCOT report, ERCOT stated that it is tracking nearly 238.6 GW of large load interconnection requests and that 7,502 MW has been approved to energize since Jan 2022. The gap between requested volumes and approved-to-energize outcomes indicates that interconnection and energization gating items (studies, agreements, security, model requirements, and technical performance) are functioning as binding constraints. Within that context, an incremental 830 MW approval that advances the project into “utility-contracted” status can be interpreted as a substantive progression in a funnel where many projects stall or are de-rated in planning assumptions.
REGULATORY AND MARKET-RULE TRAJECTORY IN ERCOT FOR LARGE LOADS
ERCOT’s formalization of large-load requirements has accelerated. The PGRR115 issue documentation indicates that the revision creates a new process for studying reliability impacts of large loads, requires LLIS for each new large load meeting applicability, requires inclusion of LLIS-evaluated large loads in ERCOT’s quarterly stability assessment prior to initial energization, and adds requirements for inclusion in the Network Operations Model and required interconnection equipment. PGRR115 is shown as approved by the Public Utility Commission of Texas (PUCT) on May 15, 2025 with documented effective dating for at least certain sections later in 2025, which aligns with the broader narrative of 2025 as a transition year from interim process toward a codified regime.
ERCOT’s Q&A also flags prospective rule changes that can affect projects delivering in 2026–2027, including PUCT rulemaking activity (Project No. 58481) to establish new interconnection standards for large loads, and ERCOT’s proposed voltage and frequency ride-through standards (referenced as NOGRR 282) that could apply to large electronic loads as a condition for interconnection and continued operation, with an indicated limited exemption framework tied to an approval/milestone date of Nov 14, 2025. For a project delivering initial power in early 2026, these evolving standards introduce a non-trivial design and compliance risk: a requirement to modify protection settings, control systems, or facility-level ride-through performance could impose incremental capex, commissioning complexity, or energization delay.
TRANSMISSION/UTILITY COUNTERPARTIES AND COST RECOVERY CONSIDERATIONS
The press release identifies AEP Texas as the utility counterparty for the additional capacity via a service agreement and WETT as the transmission interconnection provider. ERCOT describes TDSPs as regulated by the PUCT with rates set by the PUCT and required to provide non-discriminatory access to the grid; therefore, interconnection of a large load can have both a regulated-utility interface (tariffed services and cost recovery) and a project-specific contract layer (service agreements, security, construction funding for interconnection facilities). WETT describes itself as a transmission service provider developing and operating high-voltage transmission lines and substations across West Texas (over 400 miles of transmission lines and 12 substations), with capabilities including interconnection and substation development. Together, these counterparties imply that a non-trivial portion of the required infrastructure is embedded in regulated transmission/distribution frameworks, even though ERCOT’s wholesale market is deregulated on the generation/retail side. (ERCOT)
The ERCOT constraints-and-needs report also highlights broader grid-strength and transmission investment needs in areas with rapidly changing resource mixes and large load additions. The report describes system changes driven by inverter-based resource penetration and references ongoing system-strength initiatives (including synchronous condenser deployments) to address reliability needs; large electronic load additions in West Texas can interact with these dynamics through their effect on voltage stability, fault response, and local congestion patterns. While the incremental 830 MW approval indicates that the LLIS studies did not identify unresolvable violations at the study stage, the magnitude and concentration of load increase the probability that additional localized upgrades, operational limits, or phased energization conditions will be imposed as the broader interconnection landscape evolves.
MARKET IMPACT IN ERCOT: PRICES, CONGESTION, AND NEWBUILD INCENTIVES
From a power-market perspective, incremental high-load-factor demand at scale can tighten energy balances in ERCOT’s energy-only market, increasing scarcity pricing frequency and supporting merchant generation margins, particularly if incremental supply additions lag realized load growth. However, the locational impact depends heavily on the campus’s nodal placement and associated transmission constraints: a large data center in West Texas can reduce renewable curtailment and negative pricing in constrained zones by absorbing local surplus generation, but it can also create new congestion patterns and reactive power/stability requirements depending on intra-zone topology and coincident renewable output. Given that the incremental 830 MW approval explicitly references steady-state and stability studies, the project’s integration has likely been assessed under modeled contingencies; nevertheless, real-world outcomes depend on the evolving mix of nearby inverter-based generation, transmission outage conditions, and the timing and geography of other large loads energizing in the region.
At the U.S. macro level, EIA has explicitly linked forecast growth in electricity demand through 2027 to demand from large computing facilities, projecting U.S. electricity use growth of 1% in 2026 and 3% in 2027. While ERCOT-specific drivers are idiosyncratic, the national framing supports the plausibility of sustained hyperscale demand and provides a backdrop in which ERCOT’s large-load pipeline (and the need for transmission and generation investment) is being evaluated by policymakers and market participants as a structural rather than cyclical theme.
