Busbar

What it is (precise)

At the smallest scale, busbars — interconnect tabs or plates — bond individual cells in series and parallel inside a Module. Welded to the terminals, they continue the Current collector foils that carry current inside each Cell, and in many designs the whole set is integrated into a Cell Contacting System that merges busbars, voltage-sense leads and temperature sensors into one assembly.

At larger scales, copper or aluminium bars link modules within a Rack, tie racks together, and aggregate rack output onto a common DC bus running to the PCS. The same conductor family reappears in AC switchgear, but in stationary BESS the term usually means the DC current-collection path inside the enclosure.

Material choice is an engineering trade-off. Copper has higher conductivity — common aluminium busbar grades sit around 60-62% of copper on the IACS scale, so an aluminium bar needs roughly 1.6 times the cross-section to match a copper bar's ampacity — and it stays mechanically robust at bolted joints.

Aluminium is lighter and cheaper for a given ampacity but needs that larger cross-section plus careful joint treatment to manage surface oxidation and creep. Plated contact surfaces, usually tin or nickel over the base metal, and controlled bolt torque keep joint resistance low and stable across a 15-20 year project life.

Construction varies with position. Cell- and module-level bars are thin stamped or laser-welded plates, sometimes laminated flexible designs that tolerate cell swelling and vibration — the one place stationary practice borrows directly from EV pack engineering.

Rack and DC-bus conductors are heavier bolted bars, insulated with heat-shrink sleeving or epoxy coating and supported on standoff insulators rated for the full DC bus voltage. Every bolted interface is an engineered joint: plated mating surfaces, a specified torque, and hardware such as Belleville washers that hold clamping force through repeated thermal expansion and contraction.

Why it matters in a real grid-scale project

Busbars set the resistive (I²R) loss and thermal behaviour of the DC path. Because heating scales with the square of current, an undersized bar or a poorly made joint becomes a hot spot that wastes energy, erodes round-trip efficiency, and — worst case — can act as an ignition source beside adjacent Lithium Iron Phosphate cells. A loose or corroded joint is a classic field failure and a recurring finding in BESS commissioning and O&M inspections, precisely because it degrades silently: resistance creeps up over months, surfacing as quiet efficiency drift long before anything trips or alarms.

Commercially, busbar sizing trades capital cost against lifetime energy losses and warranty risk. Efficiency lost in the DC collection path flows straight through to the project's energy throughput and revenue measured at the point of interconnection. Joints are also a code and safety item: torqued to spec, often torque-marked, and inspected, because a high-resistance connection is both an efficiency penalty and a heat source that the enclosure's gas detection and explosion-protection systems were never meant to be the first defence against.

Typical values and standards

Rack-level continuous currents in modern 2-4 hour systems typically run ~70-200 A per rack: a rack storing roughly 350-450 kWh, discharging at 0.25-0.5C into a 1,200-1,500 V DC bus, lands in that band. Where racks aggregate onto a central bus feeding a 1-5 MVA PCS block, the main DC bars carry from 1,000 A up to several thousand amps.

A workable first-pass check for enclosed copper bars is roughly 1.2-2 A/mm² of cross-section, refined by thermal calculation for the actual enclosure, ambient and altitude derating. Utility DC buses are commonly 1,000-1,500 V class, with 2,000 V architectures emerging.

Temperature rise governs the final size: the bar must stay within the limits of its insulation, supports and joints at continuous rated current, and IEC 61439 ties the allowable temperature at a joint to its contact finish — bare copper joints are held to lower limits than tin- or silver-plated ones. Short-circuit withstand is the second gate.

During a DC fault the bar and its supports must survive both the thermal pulse and the electromagnetic force between parallel conductors, which scales with the square of peak current and inversely with spacing; battery-fed DC faults can reach tens of kiloamps, so bracing and support spacing are calculated, not assumed.

Design and verification lean on established references, each with a distinct role. IEC 61439 covers low-voltage assemblies — busbar ampacity, clearance and creepage, and short-circuit withstand within switchgear — while the IEEE 605 bus-design guide supplies ampacity and short-circuit-force methods engineers adapt for sizing.

At battery level, UL 1973 certifies packs and racks for stationary use, so module and rack busbars sit inside its scope, and IEC 62619 covers safety of industrial lithium cells. At system level, UL 9540 certifies the ESS as a product while UL 9540A is a fire-propagation test method characterising thermal runaway; its results feed NFPA 855, the US install standard, alongside NFPA 68/69 for deflagration venting.

How it shows up in specs, studies and contracts

On a datasheet the busbar hides inside a few lines: module and rack maximum continuous and peak current, terminal material and plating, and required bolt torque. Check the continuous rating against the worst real operating point, not the nominal one — at constant power, DC current is highest at end of discharge, when string voltage sags at low State of Charge, so a bar sized at nominal voltage can be 10-15% light at minimum voltage.

Ask the vendor three concrete things: does the continuous rating assume the container's real internal ambient, what bolt torque and re-torque interval apply, and how future augmentation racks will share the existing DC bus.

Busbar impedance and geometry then feed the project's DC short-circuit and arc-flash studies, which set protection settings and PPE boundaries inside the container. At commissioning, three checks are the ones inspectors sign off: torque verification with witness marks, micro-ohm (ductor) resistance measured across each bolted joint, and infrared thermography under load — a healthy joint reads no hotter than the adjacent bar.

Long-term service agreements commonly mandate periodic thermographic scans, and a hot joint found during a warranty claim is routinely argued as an excluded workmanship or maintenance defect rather than an equipment failure, so torque and scan records are worth keeping meticulously.

Common pitfalls

Dissimilar-metal joints are the classic trap: bolting aluminium straight to copper invites galvanic corrosion and creep-driven loosening, so copper-to-aluminium transitions need bimetallic plates or plated interfaces designed for the pair.

Daily cycling piles thermal cycling on top — every charge-discharge day expands and contracts the joint, and without conical spring washers and correct torque the clamping force walks down over thousands of cycles. Re-torquing is no cure-all either: over-tightening yields the metal and permanently reduces contact pressure, so torque values and hardware are specified, not guessed.

At 1,000-1,500 V DC, clearance and creepage distances deserve as much attention as ampacity, especially where condensation or dust raises the effective pollution degree inside the enclosure. And practice does not transfer blindly between the DC and AC sides: skin effect is irrelevant on the DC bus but reshapes current distribution in large AC bars, so a DC sizing habit applied to the PCS output bus, or the reverse, produces quietly wrong margins. When in doubt, size each side by its own standard calculation rather than pattern-matching the neighbouring bar.