Liquid-cooled busbar and DC power distribution

What the technology is

A liquid-cooled busbar is a high-current electrical conductor — usually copper or aluminum, often laminated or encapsulated — with an integrated liquid channel, cold plate, or adjacent liquid-cooled heat exchanger. The liquid removes Joule heat (P = I²R) and heat conducted from adjacent power semiconductors, connectors, or fuses, allowing a given conductor cross-section to carry more current, or the same current at lower steady-state temperature.

In DC power distribution, the conductors often carry continuous current rather than cyclic AC current, so I²R heating is sustained and skin/proximity effects are usually lower than in AC busbars at the same RMS current. DC distribution appears in:

• data-center rack power and 400–800 V DC distribution;
• electrolyzers, arc furnaces, and smelters;
• battery formation/test and EV fast-charge systems;
• traction, marine, and renewable-energy collection.

Source-tier classification used here

• Tier 1: IEEE/ISO/IEC standards; DOI-verified peer-reviewed paper.
• Tier 2: company technical/product pages and SEC filing pages; named press releases.
• Prohibited: SEO farms, anonymous benchmarks, AI-generated summaries, unretrieved patent/paper proxies.

Thermal and electrical fundamentals

For a busbar of length L, cross-sectional area A, and resistivity ρ, the electrical resistance is:

R = ρL / A

The Joule heat generated at current I is:

P = I²R = I²ρL / A

This is the core reason liquid cooling becomes interesting. As current rises, the required copper or aluminum area rises with I² if cooling is fixed; liquid cooling attacks the temperature constraint directly.

Standard constraints

Two standards bodies govern permissible conductor temperature rise, though exact numeric limits must be checked against the paid standards text:

• IEC 60943:2001 — “Guidance concerning the permissible temperature rise for parts of electrical equipment, in particular for terminals.” This is relevant because connector and joint temperature limits often constrain busbar ampacity more than the bar itself. Exact limit values should be treated as needs primary source validation; the standard is real and is sold by IEC.
• URL: [webstore.iec.ch
• IEEE C37.23-2015 — “IEEE Standard for Metal-Enclosed Bus.” It defines ratings, temperature rise, and test requirements for metal-enclosed bus. It supports feasibility claims at the standards level.
• URL: [standards.ieee.org

Both are Tier 1 sources. They do not by themselves prove that a particular liquid-cooled busbar product performs as claimed; they define test and rating frameworks.

Data-center liquid-cooling context

The peer-reviewed paper cited here:

• Ebrahimi, K., Jones, G. F., and Fleischer, A. S., “A review of data center cooling technology, operating conditions and the corresponding low-grade waste heat recovery opportunities,” Renewable and Sustainable Energy Reviews, Volume 71, 2017, pages 832–847.
• DOI: 10.1016/j.rser.2016.10.041
• Publication date: 2017-05
• Access date: 2026-04-29
• Tier 1 peer-reviewed.

This paper is not busbar-specific, but it documents the data-center liquid-cooling operating conditions and waste-heat tradeoffs that motivate liquid cooling of high-current DC distribution. It is the one DOI-verified paper used in this note. Load-bearing product-level claims still need industry-specific validation.

Why liquid-cooled busbars matter for DC distribution

In DC distribution, sustained high currents create several practical limits:

1. Conductor temperature rise at bolted joints and insulation interfaces.
2. Resistance increase with temperature for copper/aluminum, which raises I²R losses further — a feedback loop.
3. Insulation and enclosure thermal class — hotter busbars force higher-rated insulation or derating.
4. Footprint — increasing current usually requires more conductor cross-section, which increases weight and mechanical support cost.
5. System reliability — oxidation, joint loosening, and thermal cycling accelerate if temperatures are high.

Liquid cooling attacks the temperature constraint. If the conductor can be held at a lower steady-state temperature, designers can either:

• keep the same cross-section and carry more current; or
• reduce the cross-section, weight, and footprint for the same current; or
• reduce temperature-dependent resistance losses.

Company product material from Mersen states, in self-reported marketing language, that liquid-cooled busbars can reduce size/weight or improve ampacity for power electronics and heavy-current applications:

• Mersen, “Liquid cooled busbar,” product page.
• URL: [mersen.com
• Access date: 2026-04-29
• Tier 2 company technical/product page.

However, the quantitative ampacity improvement or size reduction is not independently confirmed in the company source I could inspect and should be marked needs primary source validation. This is a marketing-vs-filed-disclosure discrepancy: Mersen’s product page is technical marketing, while SEC filings of system integrators discuss thermal management demand in general and risk terms, not segment-level busbar ampacity.

Patent and paper landscape

Peer-reviewed paper — available

One Tier 1 peer-reviewed paper is cited above:

• DOI: 10.1016/j.rser.2016.10.041
• Title: A review of data center cooling technology, operating conditions and the corresponding low-grade waste heat recovery opportunities
• Authors: Ebrahimi, K.; Jones, G. F.; Fleischer, A. S.
• Journal: Renewable and Sustainable Energy Reviews, Vol. 71, pp. 832–847
• Publication date: 2017-05
• Access date: 2026-04-29

This paper establishes the data-center liquid-cooling context. It is not specific to busbar design, so direct busbar-performance claims should not be attributed to it.

Granted patent — not cited because unverified

I did not include a granted patent number because I could not verify one resolves to a real document in this session. Inventing a number would violate the source-integrity rule. The correct next-step is:

• search USPTO/EPO/WIPO for "liquid cooled busbar", "cooled busbar", "busbar cooling", and "bus bar liquid cooling";
• verify assignee and grant date;
• cross-reference against company product literature and SEC filings.

A patent is evidence of recorded claims, not proof of performance or commercial viability. This distinction applies to any patent that would be added in a second pass.

Worked calculation: 3,000 A DC busbar

Assumptions

Resistance per meter

R = ρL / A

R = (1.68e-8 Ω·m × 1 m) / 0.001 m² = 1.68e-5 Ω/m = 16.8 μΩ/m

Joule heat per meter

P = I²R

P = (3,000 A)² × 1.68e-5 Ω = 9,000,000 × 1.68e-5 = 151.2 W/m

Required coolant flow at ΔT = 5 K

Q = m_dot × c_p × ΔT

m_dot = Q / (c_p × ΔT)

m_dot = 151.2 W / (4,180 J/(kg·K) × 5 K) = 0.00724 kg/s

Volume flow for water density ≈ 1 kg/L:

V_dot ≈ 0.00724 L/s = 0.434 L/min per meter

For a 10 m busbar, the flow would be approximately 4.34 L/min under these assumptions.

Sensitivity

This calculation is a first-order thermal-electrical model. It does not include connector/joint resistance, coolant pumping power, thermal interface resistance, or ambient-to-coolant thermal resistance. Those inputs are product-specific and needs primary source validation.

Parameter table with source classification

Cross-reference: marketing claims vs. filed disclosures

Marketing/technical claim. Mersen’s liquid-cooled busbar page describes liquid-cooled busbars as suitable for high-power converter and power-electronics applications and implies performance benefits such as reduced size/weight or improved current capacity. This is self-reported Tier 2 material.

Filed-disclosure context. Vertiv Holdings Co (VRT), a US-listed supplier of thermal and power management for data centers, discusses liquid-cooling demand and product development in its SEC filings, but its public filing page does not let me verify a specific busbar-level ampacity or efficiency number from the retrieved filing. The SEC filing page:

• Vertiv SEC filings, [investors.vertiv.com
• Access date: 2026-04-29
• Tier 2.

Discrepancy. There is a gap between supplier marketing (“liquid cooling improves busbar performance”) and investor-grade disclosure (which reports segment-level demand and risk, not busbar-specific ampacity or efficiency numbers). This is a classic claims-vs-filed-evidence gap. Until a product datasheet or independent test report is retrieved, quantitative ampacity/size/efficiency claims remain needs primary source validation.

Unit economics / business model

Who monetizes this technology

The value chain has several monetization layers:

1. Material and conductor suppliers — copper and aluminum producers, laminated busbar manufacturers, and high-conductivity alloy suppliers. They sell busbar material, laminated busbars, and thermal interface materials by volume or by engineered assembly.
2. Busbar/connector component makers — companies such as Mersen, nVent, Rogers Corporation, and others sell engineered busbar assemblies. This is a project-based, engineered-component business. Margins are usually higher for engineered assemblies than for commodity copper bar.
3. Power distribution and equipment makers — Eaton, Schneider Electric, ABB, Vertiv, and others sell switchgear, busway, rack power distribution units, and DC distribution systems. They monetize integrated systems rather than the busbar alone.
4. System integrators and data-center operators — they monetize the total cost of ownership improvement, but the economic benefit is captured as lower energy loss, higher density, or longer asset life, not as a direct busbar subscription.

How money is made

• Product sale: busbar assemblies, PDUs, DC busway, liquid-cooled enclosures.
• Service and project engineering: for custom high-current industrial installations.
• Aftermarket/field service: connectors, thermal interface, coolant loop maintenance.
• Potential subscription/energy-performance contracts: not established for busbars specifically; the business model is predominantly equipment sale.

Margin and value capture — citation-bound

No specific public segment margin for liquid-cooled busbars was retrievable in this session. Therefore, exact margin numbers are needs primary source validation. The only safe statement is that the value shifts toward the companies that integrate busbar cooling with system-level DC distribution and controls, because they can monetize density and reliability rather than selling commodity copper.

Disconfirming Evidence

The market and technical thesis for liquid-cooled busbars would be contradicted if:

1. Standards and safety codes do not permit liquid-cooled conductors in occupied data-center spaces without substantially higher installation and inspection cost; that would stall adoption even if thermal performance is superior.
2. Leak failure rates or coolant compatibility problems are materially worse than air-cooled equivalents; a single leak into a power distribution busbar could cause arc-flash or corrosion failure.
3. Pump energy and maintenance costs exceed the electrical-loss savings; because the busbar is an efficiency component rather than a revenue-generating chip, total-cost-of-ownership must remain positive.
4. Data-center DC distribution does not standardize on high-current liquid-cooled busbars; if the OCP ORv3 48 V busbar standard is replaced or fails to scale, the high-volume market shrinks.
5. Component suppliers cannot demonstrate validated ampacity improvements; if the only evidence is self-reported marketing with no independent test reports or issued patents with viable claims, the moat is weak and prices commoditize.

Original scope and definitions

Routes power and cooling fluid within data center racks; enables 400-800V DC liquid-cooled architectures

Why it matters to the thesis

Enables AI rack cooling (120kW+); busbar content scales with liquid cooling penetration from 5% to 30%+ by 2028

Operating constraints

• Bottleneck type: Specification-in / Qualification / Volume supply
• Technical constraint: Busbar must integrate both power routing and coolant distribution; custom engineering per platform (NVIDIA NVL72, Rubin)
• Economic constraint: Specification-in during DC design phase creates switching costs; long lead times on specialized SKUs

Adoption

• Driver: GB200/NVL72/Rubin AI server deployment; liquid cooling adoption curve acceleration
• Blocker: Air cooling sufficiency perception; customer insourcing; commodity busbar competition from Eaton/ABB

Product categories

Liquid-cooled busbar, high-density rack PDUs, thermal management enclosures, cable management

Validation signals

NVT Data Solutions 25-35% YoY growth; VRT $16.5B backlog; ThermoKey acquisition signals sustained thermal demand

Invalidation signals

Liquid cooling penetration plateaus <10%; hyperscaler design wins shift to alternative cooling (immersion, spray); margin compression from commodity busbar competition

Who buys it

Named customer classes appear in the Mersen product page and industry context:

• Power electronics OEMs — high-power converters, UPS, drives.
• Data-center operators/hyperscalers — via rack busbar, 48 V DC distribution, and 400/800 V DC infrastructure.
• Electrolysis and smelting operators — continuous high-current DC busbars for aluminum, chlor-alkali, green hydrogen.
• Rail and traction systems — traction converters and DC switchgear.
• Battery manufacturing and test equipment — formation/test circuits with sustained DC current.
• EV fast-charging and battery integration — high-current busbars inside packs or charge posts.

These are inferred from industry context and the Mersen product page; no individual named buyer was retrieved.