ThinkNL urges government to include AI and data centers in Churchill Falls review
Nov 16, 2025 - Churchill Falls. ThinkNL has asked the provincial government to expand its review of the Churchill Falls MOU to include AI computing and data center opportunities. The group delivered a brief to Premier Tony Wakeham calling for an evaluation of "value-added opportunities that clean power can unlock."
Charlie Oliver of ThinkNL says AI and data centers are growing fast, need clean and reliable electricity, and that Labrador has the right "personality": hydro, abundant water, a cold climate, and fiber-"in spades." The ask is simple: add these industries to the province's scenario analysis and convene a cross-sector roundtable to "test assumptions and surface options."
Why this matters for government
AI workloads and data centers are now a strategic load class. They bring long-term contracts, stable revenue, and the chance to use off-peak and surplus energy efficiently. Cold climate and hydro-based grids reduce cooling needs and emissions intensity, which improves the province's competitive position.
International analysis shows data centers are a material and growing electricity load, with clear implications for grid planning and policy. See the International Energy Agency's briefing on data centers and networks for context.
What ThinkNL is proposing
- Include AI and data centers as explicit demand scenarios in the Churchill Falls review.
- Establish a cross-sector roundtable to validate assumptions, assess trade-offs, and surface options across energy, telecom, water, land, workforce, and permitting.
What to evaluate in the scenario analysis
- Grid capacity and reliability: Firm vs. interruptible load options, seasonal water flows, reserve margins, and impacts on ratepayers and existing industry.
- Generation and transmission: Current hydro commitments, potential upgrades, new substations, and fiber redundancy to key sites.
- Water and cooling: Withdrawal permits, discharge standards, air-side cooling potential in a cold climate, and on-site heat reuse.
- Siting and land: Proximity to transmission, fiber routes, and water; site-readiness; environmental constraints; cumulative effects.
- Indigenous and community partnerships: Early engagement, equity pathways, benefit agreements, and local procurement.
- Revenue models: Long-term PPAs, capacity reservations, connection queues, and tariff design for large flexible loads.
- Regulatory process: Clear permitting timelines, one-window coordination, and environmental assessment requirements.
- Workforce: Skilled trades, network engineers, facility operators, and training pipelines tied to real project timelines.
- Resilience and security: Redundant fiber, black-start and backup strategies, and cybersecurity expectations for critical facilities.
Risks to manage
- Over-commitment of firm energy that could raise rates or constrain existing users.
- Single-sector exposure if market demand slows; avoid stranded assets by phasing infrastructure.
- Water use, fish habitat, and thermal discharge-require strict monitoring and adaptive management.
- Community impacts: housing, services, and local infrastructure capacity near potential sites.
Roundtable: who should be at the table
- Energy: NL Hydro, system planners, independent experts.
- Telecom: fiber owners, carriers, and network architects.
- Water and environment: regulators, scientists, and local authorities.
- Indigenous governments and organizations with rights and interests in Labrador.
- Municipal leaders near candidate sites.
- Economic development, labor, and training partners.
- Prospective AI/data center operators and investors.
Practical next steps for public servants
- Issue a public addendum to the MOU review scope that adds AI/data center scenarios.
- Publish a baseline: available firm energy, seasonal variability, transmission constraints, and water availability by region.
- Open an expression-of-interest (EOI) to gauge site needs, load profiles, and timing from credible proponents.
- Set principles for affordability and reliability so residential and industrial customers are protected.
- Define siting criteria, permitted load flexibility, and standard interconnection timelines.
- Stand up the cross-sector roundtable with a 90-day mandate to "test assumptions and surface options," then report publicly.
Skills and capacity
Government will need policy, technical, and procurement skills to evaluate proposals and structure deals. If your team is ramping up on AI concepts, mission profiles, and vendor due diligence, a focused course catalog can speed up onboarding.
The opportunity is straightforward: clean, steady electricity, cold climate, water, and fiber make Labrador a credible home for compute-heavy industries. A structured review-guided by data and the public interest-will reveal where AI and data centers fit, and on what terms.
Your membership also unlocks:
