Behind-the-Meter vs. Utility-Scale Power: The Decision Framework

Every large power consumer — data centers, industrial facilities, electrolyzers, mining operations — faces a fundamental procurement decision: generate your own electricity on-site, or purchase it from the grid.

This is the behind-the-meter (BTM) vs. utility-scale (front-of-meter) question. It sounds simple. In practice, it is a multi-dimensional decision that depends on engineering constraints, regulatory frameworks, financial structure, timeline requirements, and risk tolerance — and the answer is increasingly shaped by policy changes that did not exist two years ago.

Neither option is inherently superior. Each allocates costs, risks, and obligations differently. The purpose of this article is to walk through the decision framework systematically so that the relevant factors are visible before any analysis begins.

The scale of the load is the first filter. Not every load size makes both options viable, and the shape of the load over time matters as much as its peak magnitude.

Small Loads (Under 10 MW)

For loads under 10 MW, behind-the-meter generation rarely makes economic sense unless there are specific reliability or power quality requirements that the grid cannot meet. The fixed costs of permitting, interconnection, and operations are spread across too few megawatt-hours to achieve competitive unit economics. Grid power from a retail provider is almost always simpler and cheaper. The exception is facilities in remote locations where grid extension itself would be prohibitively expensive — mining operations, remote data collection sites, or facilities in developing regions with unreliable grid infrastructure.

Mid-Range Loads (10–100 MW)

This is where the decision becomes genuinely complex. Both BTM and grid procurement are viable, and the marginal economics can tilt in either direction based on local conditions.

At 10–50 MW, a BTM natural gas installation might consist of several reciprocating engines or a single aeroderivative gas turbine. These are mature, commercially available technologies with well-understood costs. A 25 MW solar installation requires roughly 125–150 acres of land and produces power only during daylight hours. The economics depend heavily on the local retail rate structure, available BTM generation technology, land availability and cost, local gas pipeline access, and the cost and timeline of grid interconnection at that specific location.

At 50–100 MW, the infrastructure requirements increase substantially. A BTM gas facility at this scale is a small power plant — it requires natural gas pipeline connections capable of delivering fuel at industrial volumes, water supply for cooling (if using combined-cycle or steam injection), air quality permits that may be difficult to obtain in non-attainment areas, and a dedicated operations team. The capital cost of BTM generation at this scale typically ranges from $800–$1,500 per kW depending on technology, or $40–$150 million for the installation — a significant capital allocation that competes with other corporate investment priorities.

Large Loads (100+ MW)

At this scale, BTM generation becomes logistically challenging. A 200 MW natural gas combined-cycle plant is a major industrial facility in its own right — it requires fuel supply infrastructure capable of delivering approximately 40,000 MMBtu per day, emissions permits that place the facility among the significant sources in its airshed, a cooling water source, and a team of 20–40 personnel to operate and maintain. A 200 MW solar installation requires 1,000+ acres of land — roughly 1.5 square miles — which may not be available adjacent to the load site.

At this scale, most consumers are purchasing wholesale grid power, often through a combination of direct market participation and bilateral contracts (PPAs). The grid's ability to aggregate generation from hundreds of sources provides a level of diversification and reliability that is extremely expensive to replicate behind a single meter.

Even if you prefer grid power, the grid has to be able to deliver it. This is not a hypothetical constraint — it is the most common bottleneck in large load development.

The available capacity at the nearest substation or transmission bus determines whether grid service is feasible, how much it will cost, and how long it will take. This is determined through the interconnection study process — a formal engineering analysis conducted by the ISO/RTO or transmission owner.

How the Interconnection Study Works

When a large load applies for grid service, the utility or ISO conducts a series of studies — typically a feasibility study, a system impact study, and a facilities study. These evaluate three primary dimensions:

1. Thermal capacity: Will the additional load cause any transmission lines, transformers, or other equipment to exceed their thermal ratings? If a nearby 345 kV line is already operating at 85% of its thermal limit, adding 100 MW of load may push it over — requiring either a conductor upgrade, a new parallel line, or load management provisions.

2. Voltage stability: Will voltages at nearby buses remain within acceptable ranges (typically ±5% of nominal) under both normal operating conditions and contingency scenarios (loss of the largest nearby generator or transmission line)? Large loads draw reactive power, which can depress local voltage. Mitigation may require capacitor banks, STATCOMs, or synchronous condensers.

3. Short-circuit adequacy: Does the bus have sufficient fault current (a high enough Short-Circuit Ratio) to maintain voltage stability with the proposed load? Buses with low SCR — particularly those in areas with high inverter-based renewable generation — may require grid-strengthening measures.

When the Grid Has Capacity

If a nearby substation has sufficient thermal headroom, voltage stability, and short-circuit ratio to serve the proposed load, grid interconnection may require only minor upgrades — a new service transformer, a protection relay update, metering installation, or a short distribution line extension. Timeline: 6–18 months. Cost: typically $5–$30 million depending on the transformer size and site-specific civil work.

These situations exist but are increasingly uncommon for large loads. Most buses with abundant spare capacity are in areas far from existing infrastructure — which introduces other development challenges.

When the Grid Does Not Have Capacity

If the proposed load exceeds local grid capacity, the interconnection study will identify network upgrades — new transmission lines, substation expansions, transformer installations, protection system overhauls, or some combination. These upgrades are typically assigned to the interconnection customer under the ISO's tariff, meaning the customer pays for the infrastructure that benefits the broader grid.

Costs can range from tens of millions to hundreds of millions of dollars. A new 345 kV transmission line costs approximately $3–$6 million per mile depending on terrain and permitting complexity. A large power transformer (500 MVA class) costs $5–$15 million and has a 2–4 year lead time from order to delivery. Substation expansion with breakers, bus work, and protection systems can add $20–$50 million.

Timelines for network upgrade projects routinely extend to 5–8 years, driven by engineering design, environmental review, right-of-way acquisition, permitting, equipment procurement, and construction. In some cases, the required upgrades trigger additional cascading upgrades elsewhere on the system, extending timelines further.

In this scenario, BTM generation becomes attractive not because of superior economics in the abstract, but because of timeline. Building a gas turbine or solar installation on-site can be faster than waiting for grid upgrades — particularly when the interconnection study reveals multi-year upgrade requirements.

The timeline question is straightforward but consequential. The gap between "when power is needed" and "when power can be delivered" often determines the procurement strategy more than any economic calculation.

Grid Interconnection Timelines

Grid interconnection timelines are driven by the study process, network upgrade construction, and equipment procurement. The major phases:

• Queue entry and study completion: 12–36 months depending on the ISO and queue position. PJM's current queue backlog means new applications may wait 2+ years for study completion alone.
• Network upgrade design and permitting: 12–24 months for engineering design, environmental review, and permit acquisition.
• Equipment procurement: Large power transformers have 2–4 year lead times. High-voltage circuit breakers and other substation equipment have 12–18 month lead times.
• Construction: 12–36 months depending on the scope — a transformer replacement is faster than building a new 50-mile transmission line.

A straightforward interconnection with minimal upgrades can energize in 12–18 months. A complex interconnection with significant upgrades can take 5–8 years from application to energization. The uncertainty range is wide, and timelines frequently extend beyond initial estimates.

Behind-the-Meter Timelines

BTM generation timelines are driven by permitting, equipment procurement, and construction:

• Natural gas reciprocating engines: 18–30 months from order to operation. Engines are manufactured in modular configurations and can be shipped and installed relatively quickly. Gas pipeline connection is often the critical path item.
• Aeroderivative gas turbines: 18–36 months. Slightly longer than reciprocating engines due to more complex balance-of-plant requirements.
• Solar PV: 12–36 months depending on scale and permitting. A 20 MW installation on previously developed land can be built in 12–18 months. A 100+ MW installation requiring environmental review on greenfield land may take 24–36 months.
• Battery storage: 12–18 months. Battery modules are manufactured at scale and installation is relatively standardized. Procurement lead times for lithium-ion cells have shortened significantly since 2023.
• Combined-cycle gas turbines: 36–48 months. These are complex industrial facilities with extended construction timelines and more demanding permitting requirements.

For a facility that needs power in 2–3 years, BTM generation may be the only physically achievable option if grid upgrades are required. For a facility with a 5+ year horizon, grid interconnection may be preferable even if it requires upgrades — because the grid provides long-term operational simplicity and access to a diversified power supply.

The carbon dimension of the BTM vs. grid decision is less straightforward than it appears at first glance. The headline question — "is it clean?" — obscures significant complexity in accounting, timing, and verification.

BTM Natural Gas

Building a gas turbine or reciprocating engine behind the meter provides reliable, controllable power with high availability — but it is a direct source of CO2 emissions.

A modern combined-cycle gas turbine emits approximately 0.35–0.45 tonnes of CO2 per MWh, depending on efficiency and operating conditions. A simpler open-cycle turbine or reciprocating engine emits 0.45–0.60 tonnes per MWh. For a 100 MW facility running at 90% capacity factor, a combined-cycle plant produces roughly 275,000–350,000 tonnes of CO2 per year.

If the organization has net-zero commitments or science-based targets, BTM gas creates a direct Scope 1 emissions liability. Scope 1 emissions are the most visible and hardest to explain in sustainability disclosures. They cannot be addressed through REC purchases — they require either physical carbon capture (expensive and immature at this scale), offset credits (facing increasing scrutiny regarding additionality and permanence), or eventual retirement and replacement of the gas asset.

This creates a stranded asset risk: a BTM gas installation built today with a 25-year economic life may face regulatory or reputational pressure to retire well before the end of that life.

BTM Renewables

BTM solar or wind avoids direct emissions but introduces intermittency — the generation profile does not match a 24/7 load.

A solar installation in the Southwest U.S. achieves a capacity factor of roughly 25–30%. In the Northeast, 15–20%. This means a 100 MW solar installation produces the equivalent of 25–30 MW on average, and produces zero power for roughly 12 hours per day. Pairing BTM solar with battery storage improves dispatchability but adds $200–$400 per kWh of storage capacity, and even a 4-hour battery system only extends solar production into the early evening — it does not provide 24/7 coverage.

Most BTM renewable installations require a grid connection as backup for nights, cloudy periods, and winter months with shorter days. This means the facility still pays for grid interconnection capacity (demand charges), still relies on the grid for reliability, and the "behind-the-meter" label is somewhat misleading — the facility is partially BTM and partially grid-served.

Grid Power with PPAs

Purchasing grid power and contracting for renewable generation through a Power Purchase Agreement (PPA) is the most common approach for large consumers with sustainability targets. The structure:

1. Physical power delivery: The consumer takes grid power at their delivery point, paying the wholesale energy price (LMP) or a retail rate.
2. Financial PPA: The consumer enters a contract-for-differences with a renewable generator. When the generator produces, both parties settle against a reference price. The consumer receives Renewable Energy Certificates (RECs) representing the environmental attributes of the generation.
3. The gap: The generator and the consumer are typically at different locations on the grid. The power does not physically flow from the generator to the consumer. The REC is an accounting instrument that transfers the clean energy claim.

The accounting treatment of these arrangements is evolving. Historically, most corporate buyers used annual matching — purchasing enough RECs over the course of a year to match total annual consumption. This allows a consumer to claim "100% renewable" even if they consumed grid power from gas plants at 2 AM and received solar RECs from a generator in a different state that produced at noon.

The industry is moving toward more stringent standards:

• 24/7 Carbon-Free Energy (CFE): Pioneered by Google, this framework requires matching consumption with clean energy production on an hourly basis within the same grid region. A facility consuming 100 MW at 2 AM in Virginia must be matched with 100 MW of clean energy produced between 2 AM and 3 AM within the PJM footprint — not solar RECs from California generated at noon.
• EnergyTag: An international standard for time-stamped energy certificates that enables hourly matching verification.
• Granular certificates: The evolution of RECs into time-stamped, location-specific instruments that capture when and where the energy was generated.

Hourly matching is significantly more expensive than annual matching because it requires procurement of clean energy during hours when renewables are not producing — which means nuclear, geothermal, or battery-backed renewables. This affects both the cost and structure of PPA procurement and blurs the line between BTM and grid strategies.

BTM and grid procurement allocate risk differently across multiple dimensions. Understanding these risk profiles is essential because the "cheaper" option on a spreadsheet may carry risks that the organization is poorly positioned to manage.

Price Risk

BTM gas is exposed to natural gas commodity price volatility. Natural gas has ranged from under $2/MMBtu to over $9/MMBtu in the U.S. in recent years. For a 100 MW combined-cycle plant operating at 90% capacity factor with a heat rate of 6,500 BTU/kWh, fuel cost is approximately:

This exposure can be hedged through gas futures or fixed-price supply contracts, typically for 1–5 years. Longer-term hedges are less liquid and more expensive. The residual fuel price risk beyond the hedge period remains with the facility owner.

BTM solar has no fuel cost, but the risk is concentrated in the upfront capital investment. If grid electricity prices decline (due to cheap renewables flooding the market, for example), the BTM solar installation may produce power at a higher levelized cost than available grid power — an outcome that cannot be reversed because the capital is already deployed.

Grid (merchant exposure) means purchasing power at the prevailing wholesale price, which can be extremely volatile. In ERCOT, the system-wide offer cap is $5,000/MWh, and prices regularly reach that ceiling during summer peak hours and scarcity events. A 100 MW load exposed to a $5,000/MWh price spike for even 4 hours faces a $2 million cost event. In PJM, price spikes are less extreme but sustained high prices during winter cold snaps (Winter Storm Elliott in 2022 produced multi-day average prices above $200/MWh) can generate significant unexpected costs.

Grid with PPA provides price certainty for the contracted volume at the contracted node, but introduces basis risk — the difference between the PPA settlement price at the generator's node and the actual price the consumer pays at their delivery node. During congestion events, this basis can be large and persistent.

Operational Risk

BTM generation places operational responsibility on the load owner. Equipment failures, maintenance scheduling, fuel supply disruptions, and staffing are all the owner's problem. A forced outage on a BTM gas plant means the facility either curtails load or switches to grid power (if a grid connection exists) — and grid power during the conditions that cause plant outages (extreme heat, extreme cold) is typically at its most expensive.

Gas turbines have planned maintenance intervals (typically every 8,000–12,000 operating hours for hot gas path inspections, 25,000–50,000 hours for major overhauls) during which the unit is unavailable for days to weeks. This must be planned around load requirements or backed by redundant capacity.

Grid procurement distributes operational risk across the entire generation fleet managed by the ISO. A single plant outage does not affect service under normal conditions — the grid operator re-dispatches other generators to maintain supply. The consumer's exposure is to system-level events (widespread outages, extreme weather, fuel supply disruptions to the entire region) rather than single-asset failures.

Regulatory and Political Risk

BTM generation is subject to local and state permitting, air quality regulations, zoning, and the evolving co-location rules discussed in the next section. Permits can be challenged. Regulations can change. A facility permitted under today's rules may face different obligations in 5 years.

Grid procurement is subject to wholesale market rules set by FERC and the relevant ISO, retail rate structures set by state regulators, and transmission tariffs that are periodically updated. All of these can change through regulatory proceedings — but changes typically apply broadly to all market participants rather than targeting individual facilities.

The BTM calculus has changed materially in the last two years due to regulatory actions in multiple jurisdictions. This section explains why — and what it means for facilities in development today.

The Core Issue: Free-Riding on Grid Reliability

The grid is a shared resource. Its reliability — the ability to keep the lights on under all conditions — depends on adequate generation capacity, sufficient transmission infrastructure, and financial mechanisms (capacity markets, ancillary service markets) that pay generators to be available when needed.

These mechanisms are funded by grid users. Transmission costs are allocated to loads based on their usage. Capacity costs are allocated based on peak demand. Ancillary service costs are socialized across the market.

When a large load co-locates with behind-the-meter generation, it effectively exits the grid — or appears to. The load reduces its metered grid consumption, which reduces its share of transmission costs, capacity costs, and ancillary service payments. The costs that load was paying are shifted to remaining grid customers.

But here is the critical nuance: the load has not actually left the grid. When the BTM generator trips offline, the load will attempt to draw power from the grid instantly and at full capacity. The grid must be capable of serving that demand — which means the grid must have been planned, built, and maintained as if the BTM load were a full grid customer. But the BTM load has not been paying for that planning, building, and maintaining.

This is the free-rider problem. And it scales: if one 200 MW data center goes behind the meter, the impact is modest. If ten do it in the same region, the grid planning, cost allocation, and reliability frameworks begin to break down.

ERCOT: Senate Bill 6 (SB6)

Texas Senate Bill 6, signed into law in 2023 and effective in stages through 2024–2025, imposes new obligations on co-located generation and load in ERCOT.

Background: ERCOT operates as an energy-only market — there is no capacity market that pays generators simply to be available. Reliability is maintained through scarcity pricing (prices rise sharply when reserves are low, incentivizing generation availability) and ancillary services (generators are paid for frequency response, spinning reserves, and other reliability products). This structure depends on accurate load forecasting and transparent participation by all significant loads.

Co-located facilities that operate entirely behind the meter are invisible to ERCOT's market operations. They do not appear in load forecasts, they do not contribute to ancillary service procurement, and they do not respond to scarcity price signals. If a large co-located facility's BTM generation fails during a grid emergency, the sudden appearance of a 200+ MW load that was not in the forecast can precipitate a reliability event.

Key SB6 provisions:

• Registration requirements: Large loads (above 75 MW) co-located with generation must register with ERCOT and provide visibility into their operating characteristics — including the circumstances under which they would draw from the grid.
• Emergency Response Service (ERS): Co-located loads may be required to participate in ERCOT's ERS program, meaning they must be willing to curtail demand during grid emergencies in exchange for payments. This gives the grid operator a tool to manage the reliability risk of BTM generation failures during system stress.
• Market participation: Co-located generation that previously operated entirely outside the ERCOT market may now be required to offer capacity into the market during scarcity conditions — effectively requiring BTM generators to act as grid resources when the grid needs them most.
• Reliability demonstration: New co-located facilities above certain size thresholds must demonstrate through study that they do not degrade grid reliability — similar to the interconnection study process for grid-connected resources.

Practical impact: SB6 does not prohibit BTM generation in ERCOT, but it significantly reduces the regulatory advantages. A co-located facility that was previously invisible to the market must now register, may face curtailment obligations during emergencies, and may need to offer its generation into the market during scarcity — which erodes the core value proposition of co-location (avoiding market participation and grid obligations).

PJM: FERC Scrutiny of Co-Located Arrangements

In PJM — the largest U.S. ISO by load, covering 13 states and D.C. with approximately 180 GW of peak demand — the Federal Energy Regulatory Commission (FERC) is actively examining co-located load-generation arrangements, particularly for data centers in Northern Virginia (the largest data center market in the world).

The context: Northern Virginia's "Data Center Alley" along the Loudoun County corridor has seen extraordinary load growth. Dominion Energy, the transmission owner, has struggled to build transmission infrastructure fast enough to keep pace. Some data center developers have explored co-locating gas generation or nuclear to bypass the transmission bottleneck.

Several high-profile proposals — including arrangements to co-locate data center load directly at existing nuclear plant sites — brought the co-location question to FERC's attention.

FERC's concerns:

• Transmission cost allocation: PJM's transmission costs are allocated to loads based on their Network Service Peak Load (NSPL). When a load co-locates with generation and reduces its NSPL, the transmission costs are shifted to other customers. FERC is examining whether this cost-shifting is just and reasonable — the standard under the Federal Power Act.
• Capacity market implications: PJM's capacity market (Reliability Pricing Model, or RPM) procures generation to meet a reliability standard based on forecasted peak load. If large loads exit the grid through co-location, the forecasted peak decreases — reducing capacity procurement. But those loads may still rely on the grid during BTM outages. FERC is examining whether co-located loads should face capacity obligations that reflect their potential grid impact during BTM outages.
• Interconnection service type: Grid-connected loads in PJM receive Network Integration Transmission Service (NITS), which entitles them to reliable delivery from anywhere on the network. Co-located loads that primarily self-supply but occasionally draw from the grid occupy a gray area — they want the reliability of NITS during BTM outages without paying the full NITS cost at all times. FERC is examining whether a new service category is needed.
• Market power: If a large generator co-locates with a large load and serves that load directly, the generation is effectively removed from the competitive wholesale market. During scarcity, this reduction in available supply could increase prices for remaining grid customers. FERC is evaluating the market power implications, particularly when the co-located generation is a large baseload resource like a nuclear plant.

Status: Multiple FERC proceedings and PJM stakeholder initiatives are actively developing new rules. FERC issued an Advance Notice of Proposed Rulemaking (ANOPR) in 2025 seeking comment on co-located configurations. PJM has proposed tariff modifications. The outcome is uncertain, but the direction is clear: co-location will face more scrutiny and more obligations than it has historically.

CAISO (California)

California has long imposed stringent requirements on behind-the-meter generation. The South Coast Air Quality Management District (SCAQMD) and other air districts impose emissions limits that effectively prohibit new BTM gas generation in most urbanized areas. Even in areas where permits are obtainable, the emissions offset requirements — purchasing credits from facilities that reduce emissions elsewhere — add substantial cost.

BTM solar is common in California but has been affected by successive rounds of Net Energy Metering (NEM) tariff changes. NEM 3.0 (effective 2023) significantly reduced the compensation rate for exported BTM solar, making solar-plus-storage the economically preferred configuration. For large industrial loads, the economics of BTM solar depend heavily on the specific utility rate schedule, time-of-use rate design, and demand charge structure.

MISO and SPP

Co-location rules in MISO and SPP are less developed than in ERCOT and PJM, but are drawing increasing attention as data center development expands into these regions. MISO's central corridor (Iowa, Illinois, Indiana) has seen significant data center interest driven by available land, relatively low power costs, and proximity to fiber routes. As large co-located proposals emerge, MISO's tariff and interconnection procedures will face the same questions PJM is addressing.

ISO-NE and NYISO

The dense urban geography and limited available land in the Northeast make large-scale BTM generation impractical in most cases. Grid procurement is the default for large loads. The regional focus is on offshore wind, battery storage, and transmission expansion to serve growing electrification demand — not on co-located facilities.

In practice, many large facilities adopt a hybrid strategy that combines elements of both BTM and grid procurement. This is not a compromise — it is often the optimal configuration when evaluated against all five decision factors.

Common Hybrid Configurations

Configuration 1: Grid-primary with BTM peak shaving

• Grid connection sized for full load capacity
• BTM battery storage (1–4 hours) to reduce peak demand charges and provide short-duration backup
• Net effect: lower demand charges, some energy arbitrage, minimal regulatory complexity

Configuration 2: Grid-primary with BTM renewables

• Grid connection for full capacity
• BTM solar (sized to offset 20–40% of daytime consumption) with battery storage
• Wholesale PPA for additional renewable procurement
• Net effect: reduced grid purchases, partial sustainability progress, maintains full grid backup

Configuration 3: BTM-primary with grid backup

• BTM gas turbines sized for base load
• Grid connection sized for N-1 redundancy (one BTM unit out) plus peak demand
• BTM solar or battery for supplemental clean energy
• Net effect: operational control, shorter timeline, but still subject to co-location regulations

Configuration 4: Phased development

• Phase 1: BTM generation to meet initial load while grid upgrades are constructed
• Phase 2: Transition to grid as primary supply once interconnection is complete
• BTM generation retained for peaking, backup, or sold/repurposed
• Net effect: accelerated initial timeline, long-term grid integration

The hybrid approach diversifies risk, provides multiple pathways to sustainability targets, and avoids full dependence on either the grid or on-site generation. It also requires significantly more sophisticated energy management — coordinating dispatch across multiple sources, managing overlapping contracts, ensuring regulatory compliance for each component, and maintaining engineering staff or service contracts for BTM assets.

The BTM vs. grid decision depends on the specific intersection of these factors for each facility. No single factor dominates — the answer emerges from the combination.

| Factor | Favors BTM | Favors Grid | |--------|-----------|-------------| | Load size | 10–100 MW sweet spot | Under 10 MW or over 200 MW | | Grid capacity | Insufficient local capacity, long upgrade timeline | Available capacity, minimal upgrades | | Timeline | Need power in 1–3 years | 5+ year horizon acceptable | | Carbon target | BTM renewables + storage | Grid + PPA + hourly matching | | Risk appetite | Prefer operational control, tolerate commodity risk | Prefer distributed risk, tolerate market/regulatory exposure | | Regulatory environment | Jurisdiction with stable BTM rules, minimal co-location scrutiny | Jurisdiction with active co-location proceedings (ERCOT SB6, PJM/FERC) | | Capital availability | Willing to deploy $50M–$150M+ in generation assets | Prefer operational expenditure model | | Organizational capability | In-house power plant operations and energy management | No energy operations capability |

There is no universal answer. The optimal strategy depends on the specific site, the specific grid conditions at the point of interconnection, the specific regulatory jurisdiction, the organization's financial position, and the organization's operational capabilities.

The behind-the-meter vs. utility-scale decision is not a binary choice but a spectrum of configurations shaped by engineering constraints, economic analysis, regulatory frameworks, and organizational strategy. The traditional assumption — that large consumers can simply build their own power and bypass the grid — is being challenged by regulators in ERCOT, PJM, and other jurisdictions who are concerned about reliability implications, cost-shifting to remaining grid customers, and wholesale market integrity.

The decision framework requires evaluating five factors in combination: load size and profile, local grid capacity and interconnection cost, development timeline, carbon and sustainability requirements, and risk tolerance across price, operational, and regulatory dimensions. The answer will be different for a 50 MW facility in West Texas with access to cheap gas and abundant land than for a 300 MW campus in Northern Virginia facing transmission constraints and FERC scrutiny, even if the underlying business is identical.

The regulatory trajectory is clear: co-location is receiving more scrutiny, not less. Facilities planned today should model scenarios in which BTM benefits are reduced by future regulatory changes. The safest strategies are those that remain economically viable regardless of how the co-location rules evolve — which often points toward hybrid configurations that maintain grid access while deploying BTM assets for specific operational or sustainability objectives.

What has not changed: understanding the grid at your specific point of interconnection is the prerequisite for any power procurement strategy, whether the electrons come from your own generator or from the transmission system.