Anees Jeddy is a transmission integration planning engineer for the New York Independent System Operator. This article reflects his personal views.
The electric industry is entering a planning cycle unlike any it has faced in decades. Utilities are being asked to serve hyperscale artificial intelligence data centers with load requirements ranging from hundreds of megawatts to well above a gigawatt, often on aggressive timelines and in places where the grid is already constrained.
At the same time, much of the new generation needed to serve that demand is renewable and remote from load centers. The result is a transmission planning challenge that is larger, faster and more stability-sensitive than the frameworks many utilities have relied on in the past.
That distinction matters because AI-driven load growth is exposing the limits of one-size-fits-all transmission planning. A utility serving a dense metropolitan data center cluster faces fundamentally different constraints than one moving wind generation across hundreds of miles. The challenge shifts again for offshore wind systems supplying coastal AI load while maintaining system strength at a weak onshore interconnection point. For example, a coastal grid integrating offshore wind and hyperscale data centers can face stability constraints that differ entirely from a long-distance bulk transfer problem.
In each case, the optimal technology choice is not predetermined — it depends on the underlying system physics and constraints.
High-voltage direct current technology continues to offer clear advantages. It remains highly effective for long-distance bulk transfer, submarine transmission and asynchronous interconnection. Modern voltage-source converter HVDC systems provide precise power flow control, black-start capability and strong performance in weak-grid conditions. For applications such as delivering remote renewable energy into major load centers over long corridors, HVDC will often remain the preferred backbone solution.
However, the AI-driven grid is not defined by distance alone. It is shaped by stringent power-quality requirements, minimal tolerance for interruption, clustered load growth, increasing penetration of converter-based resources and tighter stability margins under stressed conditions. These factors elevate the role of technologies that reinforce the alternating current system — providing dynamic voltage support, improving short-circuit strength and preserving reliability under severe contingencies — alongside or, in some cases, instead of purely transfer-focused solutions.
Where alternative transmission technologies fit
This is precisely where alternative transmission technologies deserve far more serious attention than they typically receive in conventional planning discussions.
Flexible AC Transmission Systems, or FACTS — especially static synchronous compensator, or STATCOM, and unified power flow controllers — are among the most practical near-term options for unlocking capacity on the existing network.
In many urban, suburban, and already-constrained corridors, the most valuable transmission asset is not a brand-new right-of-way but the one that already exists and is limited by voltage, stability or contingency performance. FACTS devices can rapidly improve power-transfer capability, deliver fast dynamic voltage support and mitigate the types of instability that become more probable as both generation and load shift toward power-electronics interfaces. For utilities facing hyperscale data center growth atop a stressed AC backbone, a FACTS-enhanced AC solution often delivers the optimal balance of speed, cost discipline and operational familiarity.
High-temperature superconducting, or HTS, cable also earns a prominent place in the modern utility toolkit — particularly in dense metropolitan environments where right-of-way, not distance, is the binding constraint. When hundreds of megawatts or gigawatts must be delivered through narrow underground or urban corridors to major load centers or campus clusters, HTS offers unmatched power density and superior power-quality performance. It is not a universal solution, and installed costs remain a hurdle, yet for short- to medium-distance, high-density applications, it has moved well beyond the experimental stage.
Multi-terminal DC architectures become increasingly relevant in the AI era. While traditional point-to-point HVDC fits certain long-haul needs, it falls short when multiple renewable injection points, multiple delivery nodes and clustered data-center loads demand greater routing flexibility and redundancy. A meshed or multi-terminal DC network can provide exactly that operational agility. The key caveats are the need for sophisticated controls, robust protection schemes and adequate AC system strength at every interface point with the broader grid.
Hybrid solutions may ultimately prove the most powerful category of all.
Utilities rarely face a pure “AC versus DC” choice. In practice, the strongest answer is often a tailored combination: HVDC for bulk long-distance transfer paired with targeted AC reinforcements to preserve system strength and contingency performance; offshore HVDC landing in a robust onshore hub supported by STATCOMs or storage; or embedded power-electronics solutions that enhance the flexibility of existing AC infrastructure without sacrificing the grid-support characteristics planners still require. In the AI-era planning environment, hybridization is not a compromise — it is the engineering approach best aligned with real world system needs, timelines and locational realities.
The real bottleneck may be instability. The defining grid risk of the coming decade is shifting from thermal congestion to dynamic instability and weak-grid performance.
Large AI facilities demand extremely high service quality, while renewable-heavy systems often operate with lower inertia and weaker fault characteristics than the grids utilities historically planned around. Converter interactions, low short-circuit ratios, limited frequency support and voltage recovery challenges are no longer edge cases. They are becoming central planning constraints.
A transmission solution that appears optimal on a cost-per-megawatt or loss basis can still fail if it weakens the system at the delivery point.
This shift requires a different approach to transmission planning. Technology selection should begin with a clear understanding of the constraint:
- Is the challenge long-distance bulk transfer, urban power-density delivery or reinforcement of an existing stressed network?
- Is the limiting factor right-of-way, system strength, voltage stability, fault duty or construction timeline?
- Does the load require a level of resilience and power quality beyond what the surrounding grid can provide?
- Can the solution perform under severe contingencies, not just under normal operating conditions?
When planning begins with these questions, the appropriate technology pathway becomes clearer. HVDC remains the leading option for long-distance transfer. FACTS-enhanced AC solutions are often optimal for reinforcing existing networks. HTS cable becomes viable in high-density urban corridors. Multi-terminal DC enables flexibility across distributed systems. Hybrid AC/DC solutions address the combined need for transfer efficiency and system strength.
The AI era doesn’t reduce the value of any transmission technology — it raises the cost of choosing the wrong one by selecting an option before fully understanding the system constraint. Utilities that plan the right way will be better positioned to integrate large loads without compromising system reliability.
