An Open Letter to the Power and Digital Infrastructure Industries

Accelerating Time-to-Power for Data Centers and Large Loads Through a Flexibility Framework

The global demand for digital infrastructure is accelerating at an unprecedented pace. Artificial intelligence, cloud computing, advanced manufacturing, and electrification are driving the rapid deployment of data centers and other large loads at a scale and speed that the power system has not historically experienced.

In many regions, a common constraint has emerged: time to power. Large loads often face lengthy and uncertain timelines before they can be served, and this now represents one of the most significant barriers to deploying critical digital and industrial infrastructure, even in locations where capacity exists or could be unlocked with modest changes in how loads and grids interact. 

At the same time, a powerful but underutilized opportunity is available today. Many large loads—including some modern data centers—can provide meaningful, reliable flexibility to the power system. However, the absence of a shared language, common definitions, evaluation tools and transparent performance expectations has limited the ability of utilities, system operators, regulators, and load developers to use that flexibility to accelerate deployment.

To address this gap, the EPRI DCFlex program is developing a flexibility framework for large loads, designed explicitly to meet the immediate need to accelerate time to power while maintaining system reliability and long-term efficiency.

This letter is a call to the industry to collaborate around this framework and move from fragmented, bespoke approaches toward replicable and scalable solutions.

The Challenge We Face

Today’s interconnection and planning processes rely heavily on worst case assumptions about load behavior. While prudent from a reliability perspective, these assumptions could change when planning with flexibility in mind.

As a result:

  • Utilities and system operators lack confidence and transparency into what flexibility a large load can credibly provide.
  • Load developers lack clear signals on which flexibility capabilities matter most and how they could influence interconnection outcomes.
  • Each project individually or in a cluster is modeled as a custom case with unique characteristics, increasing study time, uncertainty, and cost for all parties.

This dynamic slows deployment precisely when speed matters most.

A Practical, Near Term Solution: Load Flexibility Framework

The DCFlex flexibility framework introduces a technology‑neutral way to describe and evaluate large load flexibility, based on power system requirements—such as congestion relief, peak reduction, balancing, and frequency response.

Rather than focusing on specific technologies or assets, the framework defines flexibility in terms of performance characteristics, including:

  • Notification time
  • Duration
  • Frequency of use
  • Depth of load adjustment
  • Ramp behavior 
  • Availability

These characteristics are grouped into a small number of flexibility classes, each representing a clear, consistent set of capabilities that can be understood and used by utilities, operators, and regulators across regions.

This approach enables a shared understanding of:

  • What flexibility is available
  • What flexibility is needed
  • How that flexibility can be relied upon in interconnection, planning, and operations

Why This Matters for Speed to Power—Now

The immediate value of this framework is not theoretical. It is practical and actionable.

By guiding and informing how flexibility is defined and communicated:

  • Interconnection studies can be streamlined, reducing the need for repeated custom modeling.
  • Available capacity and siting decisions can improve, aligning large loads with locations where flexibility unlocks grid headroom.
  • Utilities can more confidently consider non-firm or alternative interconnection arrangements, accelerating initial energization while preserving paths to firm service over time.
  • Load developers can design facilities intentionally, making informed trade‑offs that reduce deployment risk and delay.

In short, framework-based flexibility enables faster, more predictable decisions—while maintaining system reliability.

A Framework Built for Collaboration

The DCFlex flexibility framework is being developed through a collaborative, multi‑stakeholder process involving utilities, system operators, data center operators, generation owners, technology providers, and industry organizations. Its core principles are:

  • Nondiscriminatory: Technology neutral and vendor agnostic
  • Complete and specific: Detailed enough to be operationally meaningful
  • Structured: Replicable across geographies and time
  • Practical and ready-to-use: Framework is to be accompanied by easy to adopt tools, processes, and a guidebook
  • Transparent: Clear expectations for both loads and grids

Recognizing the diversity of stakeholders and respective local conditions that require consideration in adoption, the framework is intentionally designed to be adaptable—supporting pilots, voluntary programs, and regional implementation while maintaining a consistent foundation.

A Call to Action

We believe accelerating the deployment of data centers and other large loads is a shared imperative—and one that requires shared solutions.

We call on:

  • Utilities and system operators to engage with shared flexibility framework as a tool to unlock capacity and reduce interconnection friction.
  • Load developers and operators to design for flexibility and adopt common performance definitions.
  • Regulators and policymakers to enable pathways that recognize flexibility as a legitimate, reliability enhancing resource.
  • Industry consortia and standards bodies to align efforts and avoid fragmentation.

The challenge of time to power cannot be solved by any one stakeholder alone. But with a common framework, a shared language, and a commitment to collaboration, it can be solved—starting now.

The DCFlex flexibility framework represents a concrete step toward that future. We invite the industry to join us in shaping, piloting, and incorporating the learnings into meaningful solutions.

 "Antitrust Disclaimer: This letter is informational only and describes a non‑binding framework for discussing large‑load flexibility. EPRI is not a standards‑setting organization, and nothing herein constitutes an agreement, standard, or coordination affecting competition."

Signed,
Industry participants committed to accelerating reliable, sustainable power for data centers and large loads

Dark blue industrial landscape with Flex MOSAIC Logo and a play button

Flex MOSAIC™ at a Glance

Watch this overview video to get a quick, high-level introduction to the newly launched Flex MOSAIC™ framework—and why large-load flexibility can help accelerate time to power

Four People sitting at a conference table with blue saturation, the Flex MOSAIC logo and a document icon in the center

Inside Flex MOSAIC™: The Technical Framework Explained

This two‑page brief unpacks the Flex MOSAIC™ framework in detail, outlining the core concepts, structure, and logic behind how flexibility is defined, evaluated, and applied to support faster, more reliable large‑load integration

Aerial view of power lines among trees with Flex MOSAIC logo and a hand pointer icon

Experience Flex MOSAIC™

Jump into the interactive journey, pick your persona (DC Designer, Program Developer, Transmission Planner), and explore how Flex MOSAIC™ converts flexibility from a concept into an actionable pathway for accelerating time to power.