Today, most discussions around powering data center campuses focus on generation and grid interconnection. But inside the fence line and along the last mile(s) of distribution, two other critical constraints get less attention: the copper that carries the power and the skilled labor it takes to install it.

Neither constraint is new. What has changed is that copper now carries greater price risk, while skilled electrical labor brings more schedule uncertainty. Skilled electrical trades account for 30-40 percent of construction hours on a data center project, and some reporting projects an electrician shortfall beginning in 2027. Electrical topology decisions increasingly determine a facility's exposure to both risks.

Copper: Unpredictable in the near term, limited in the long term

Near-term copper forecasts have proven unreliable. The International Copper Study Group (ICSG) projected a 150,000-tonne refined copper deficit for 2026 in October 2025, then revised it six months later to a 96,000-tonne surplus, with a 377,000-tonne surplus projected for 2027.

The long-term picture is more consistent. S&P Global's January 2026 study, Copper in the Age of AI, projects global demand rising about 50 percent, from 28.3 million tonnes in 2025 to 42.4 million tonnes in 2040. It expects mined output to peak around 2030 and decline afterward, with recycled copper covering roughly 31 percent of demand by 2040. The study attributes only about ten percent of projected demand growth to AI and data centers, while core economic growth, the energy transition and defense account for most of the rest.

Supply responds slowly. On average, a new copper mine takes roughly 17 years to move from discovery to production. It’s not that copper will run out.

Rather, the marginal tonne is likely to become more expensive and less predictable to secure, with data center developers competing alongside other industries that need copper for their own power distribution infrastructure: charging stations for electric vehicle (EV) fleet depots; building electrification (heat and cooling systems); and advanced manufacturing facilities (semiconductor fabs). A design that requires fewer tonnes per megawatt carries less of that exposure.

How labor and campus layout affect deployment schedules

For a data center, deployment cost is partially informed by the construction timeline. Each month between construction completion and energization is a month of capital deployed without revenue. Lead times for long-lead equipment are largely outside a project's control. But field installation time is more directly influenced by the amount of material that needs to be installed and the availability of skilled electrical labor.

Layout matters too. At 480 volts, copper’s cost and bulk keep transformers and generators close to the building, which concentrates electrical work in a single corridor. Industry productivity factors suggest losses of about 25 percent when crews are crowded to 200–250 square feet per person.

Design metrics worth tracking

If copper exposure and schedule exposure are both shaped by design, they can be measured early on in the planning process. Two metrics are useful alongside cost per megawatt:

1. Conductor tonnage per megawatt delivered, which captures exposure to copper price and availability.
2. Field labor hours per megawatt, which captures schedule and labor exposure.

When comparing conductor options for a given route and voltage, cost per kiloampere-meter ($/kA-m) provides a common basis: the cost to carry 1,000 amperes over one meter. It can cover conductor cost alone or full installed cost, including terminations and labor, so the scope should be stated when options are compared. Either way, it measures cost, not the tonnage or field labor hours a design depends on, so it is most useful when considered alongside the two metrics above.

Several approaches move one or both metrics, each with tradeoffs:

• Higher distribution voltages (e.g., 800VDC versus 415 or 480VAC) reduce current for the same power and with it, conductor cross-section. The tradeoff is ecosystem immaturity, arc-flash safety hazards, and workforce retraining requirements.
• Aluminum conductors substitute for copper at lower material cost but need a substantially larger cross-section for comparable ampacity and careful termination practice.
• Prefabricated electrical skids and modules shift labor from the site to the factory. They shorten field schedules but add transport limits and earlier design freezes.

No single approach is best for every project. But the point is that conductor tonnage and installation time are design outputs, and they can be explicitly evaluated.

Superconducting power distribution as an alternative

Superconducting power distribution addresses both design metrics more directly. Superconductors carry much higher current density than copper, so the same power can move through fewer, smaller cables and narrower routes, using far less conductor material.

The benefit depends on where the superconducting cable is applied. One application where superconducting cables have a clear advantage is longer, high-current feeder runs where the near-zero resistive losses have a greater impact on system efficiency. Shorter, lower-ampacity runs are typically left in copper, where conventional cable remains the practical choice.

A 72-megawatt (MW), 480 volts reference design illustrates the effect when power equipment sits about 105 meters from the data hall. At that distance, the conventional design needs 564 tonnes of copper, compared with 332 tonnes when the same equipment sits about 15 meters away.

Replacing the long, high-current feeders with superconducting cable cuts copper tonnage by 68 percent (564 to 183 tonnes) and installation labor by 55 percent (63,700 to 28,400 hours). These figures compare the complete superconducting-based design against the complete conventional copper design, including the feeders that remain copper in both.

These tradeoffs are valid. Superconducting systems require cooling infrastructure and careful system integration, which should be weighed against the material and installation savings.

Copper forecasts will likely keep swinging, and the electrician shortfall is projected to begin in 2027. Developers cannot wait for either to resolve. Conductor tonnage and field labor hours, however, can be influenced by design decisions. Measuring both at the design stage, alongside cost, shows pre-construction teams which options, superconducting among them, can reduce installation complexity and shorten project timelines.

Visit veir.com to learn more about superconducting power delivery solutions.