Data Center Busway Market (2026-2036)

This report addresses busbar trunking systems supplied to data centre facilities, covering alternating current busway and the emerging direct current variants, from uninterruptible supply output distribution through to rack and power shelf feed. It is a vertical treatment of a product category addressed across all end-use industries in separate coverage of busbar systems, and the direct current portion is addressed in greater depth in dedicated coverage of 800 V direct current busways and busbar systems, so none of the three totals should be read as additive.

The growth of this market is driven by rack power levels that have moved well beyond what conduit and cable distribution serves economically, and by the operational advantage of a distribution system whose tap-off points can be added, moved or removed without de-energising the run. The second of those is the reason busway displaced cable in this sector and it remains the stronger argument, because a data centre changes its rack layout many times over the life of the building that houses it. The weaker argument is capacity, and the risk that argument conceals is addressed directly in this report. Busway is nonetheless the distribution method least able to absorb a change in the voltage it was designed to carry, which is the structural question this market now faces.

A busway is a prefabricated assembly of conductors in a rigid enclosure, installed as a continuous run with tap-off points at intervals. It displaced conduit and cable in data centre white space for a reason that has little to do with capacity. A tap-off box can be added, relocated or removed on an energised run, which means a row can be re-provisioned for a different rack layout without rewiring and without an outage, whereas cable distribution requires a new pull for every change. In a facility whose computing contents are replaced several times over the life of the building, that property is worth more than the marginal cost of the conductor.

Busbar trunking systems are governed within the low-voltage switchgear and controlgear assemblies series by IEC 61439-6, Busbar trunking systems, first published in 2012 with a second edition in development, and in North America by the Underwriters Laboratories standard for busways. That framework is stable and the product is mature, which has historically made this a competition on price, lead time, current rating and tap-off ecosystem rather than on technology. The circumstances that made it a stable category are now changing, and the change is not the one the industry usually emphasises.

The conventional account holds that rising rack power drives busway demand, which is true but incomplete. The more consequential development is that the distribution architecture itself is in transition. A white paper dated 25 September 2025, authored by Pietro Scalia, Yong Perry Li and Ashish Ekbote of Renesas Electronics and written in compliance with Open Compute Project tenets, records that artificial intelligence driven rack power growth is outpacing the conventional 48 V in-rack distribution scheme, which would otherwise cause large distribution losses or require heavy copper use. It states that 800 V, in either two-wire or three-wire form, is expected to be selected, with the two-wire arrangement offering higher copper utilisation and the three-wire arrangement lowering isolation requirements at the cost of more complex load balancing.

In the architecture described, the DC bus is generated in an external sidecar rack from a three-phase 480 Vrms alternating current source, using a Vienna-topology front end with bidirectional gallium nitride switches, and the information technology rack receives that DC voltage and houses the accelerators. A high-power-density isolated 16 to 1 DC to DC converter steps 800 V down to 48 V so that the existing 48 V ecosystem can be reused. The paper records that the 800 V approach reduces copper weight by almost 50 percent and that the current 48 V structure would require massive copper bus bars. Two details in that description matter more to this market than the headline voltage. The first is that the alternating current busway does not disappear, because the direct current bus is generated in a sidecar rack fed from a three-phase 480 Vrms source, so alternating current distribution continues and is pushed one stage upstream rather than eliminated. The second is the copper weight reduction of almost 50 percent, which is a direct statement that the conductor cross-section appropriate to one architecture is substantially larger than that appropriate to the next.

That is where the uncomfortable conclusion for this category sits. Busway is marketed as future-proof infrastructure, and in one respect it genuinely is, because tap-off flexibility makes it resilient to changes in rack layout. It is considerably less resilient to a change in distribution voltage. A busway run is a fixed building asset bolted to a ceiling or installed under a floor, sized and routed for a particular voltage and current, and it cannot be re-pulled the way a cable in containment can. Cable is replaceable in situ at modest cost. A busway run is not.

Set that against the pace of change on the load side. On rack power the paper records that several hundred kilowatts per information technology rack is already occurring, citing a forthcoming platform at more than twice the power of the current generation, with sidecar alternating to direct current building blocks trending towards approximately 20 kW or higher and a validated high-voltage direct current design of 12 kW built from two 6 kW units. It notes that accelerator product cycles have shortened to 12 to 15 months against approximately 30 months traditionally, and contrasts a historical semiconductor scaling rate of doubling every three years with artificial intelligence demand doubling every 12 to 18 months, concluding with reference to megawatt-scale demands. A busway specified today therefore sits between a building measured in decades and a computing architecture turning over in roughly a year, during a period in which the distribution voltage is explicitly described as expected to change and in which the copper required is expected to halve. The distinction worth holding is precise: busway is robust to layout change and fragile to voltage change, and the industry has been selling the first property as though it covered the second.

This does not imply that alternating current busway demand collapses, and a report suggesting so would be wrong. The sidecar architecture requires alternating current distribution to the sidecar, uninterruptible supply output distribution remains alternating current, mechanical plant distribution is unaffected, and the installed base requires maintenance, extension and tap-off supply for decades. What changes is where the alternating to direct current boundary sits and therefore how far the alternating current run extends into the hall, which shortens the downstream alternating current scope while leaving the upstream scope intact. The composition of this market shifts considerably faster than its total.

Conductor material is the second axis under pressure and for a reason unrelated to architecture. Copper remains the specified default on conductivity and connection reliability, but aluminium offers a materially lower cost per ampere and a lower weight for the same rating, and weight matters for a ceiling-suspended run in a building whose structural steel has been value-engineered. As conductor cross-sections rise with current rating, the absolute cost and weight difference between the two materials widens, which strengthens the aluminium case precisely in the high-current segment that is growing fastest.

A caution belongs in any assessment of this market. The direct current transition is well evidenced in architecture documents and in component development, and it is not yet widely evidenced in installed busway. The paper cited here describes an expected selection rather than a completed standardisation, the standards framework governing busbar trunking is a low-voltage alternating current framework whose extension to 800 V direct current is an open question, and a buyer specifying direct current busway today is ahead of the certification position. Suppliers should expect the growth of the direct current segment to be paced by standards and certification rather than by demand.