Two Battery Racks on One 51.2V Inverter: Rack-to-Bus Design, Protection, and Current Sharing

Introduction

A commercial or larger residential ESS begins with one rack containing:

4 × 51.2V battery modules.

The project later expands to a second rack:

Rack A: 4 modules
Rack B: 4 modules

All eight batteries need to support one common inverter system.

A simple installer might connect:

Rack B → Rack A → inverter.

The system powers up.

But several months later:

Rack A carries more current
Rack A batteries show more cycles
Rack A runs warmer
Rack B SOC remains higher
One rack reaches protection before the other

The problem is not necessarily the battery modules.

The system may have created a preferred rack because the two racks do not have equal electrical paths to the inverter.

Once battery systems expand across multiple physical racks, the design should move from:

“parallel batteries”

to:

rack-level DC distribution architecture.

1. Treat Each Rack as a Sub-Bank

A useful design concept is:

Rack A

Multiple matched batteries connected to an internal or local bus.

Rack B

Another matched group.

Both racks then connect to a:

common main DC bus

which supplies the inverter.

This makes each rack a defined sub-bank rather than extending one long daisy chain.

2. Avoid Rack A as the Connection Point for Rack B

Poor architecture:

Rack B
→ Rack A bus
→ inverter

Rack A is physically closer to the inverter connection.

Its batteries may have a lower effective resistance path.

Better architecture:

Rack A
→ main positive/negative bus

Rack B
→ same main positive/negative bus

Inverter
→ same main DC distribution

Each rack now has a clearer and more symmetrical route.

3. Why Rack-Level Resistance Matters

Current sharing depends on total path resistance.

For an individual rack, this can include:

Internal module branches
Internal rack busbar
Rack breaker
Rack-to-main-bus cable
Main busbar joint

Even if every individual battery cable is matched inside each rack, unequal rack-to-bus cables can still create rack-level imbalance.

4. Example

Rack A-to-main bus: 0.5m cable.

Rack B: 3m cable.

Same conductor cross-section.

Rack A has much lower rack-level cable resistance.

At a 300A total load:

Rack A may supply: 190A.

Rack B: 110A.

Inside each rack, batteries may share current reasonably.

But the racks themselves are imbalanced.

5. Individual Battery Balance Does Not Guarantee Rack Balance

Suppose Rack A contains four batteries each supplying: 47–48A.

Rack B batteries: 27–28A.

Within each rack: excellent balance.

Across racks: poor balance.

If the installer only checks battery-to-battery current inside Rack A, the larger system problem can be missed.

Always compare:

Rack A total current vs Rack B total current.

6. Rack-Level Current Measurement

Install or provide measurement access for each rack feeder.

Example:

Rack A feeder

165A

Rack B feeder

160A

Reasonable.

Another case:

Rack A

230A

Rack B

95A

This deserves investigation even if no individual BMS alarm exists.

7. Rack-to-Bus Cable Design

For two identical racks, aim for appropriately engineered rack feeders with:

Same conductor material
Same cross-section
Similar effective length
Same connection hardware
Similar protection devices

The objective is similar rack-level resistance.

Do not intentionally undersize the “stronger” rack cable to force balance.

8. Main Busbar Placement Matters

If the main busbar sits directly beside Rack A but far from Rack B, cable paths naturally differ.

Possible solutions include:

Central DC distribution location
Appropriately engineered larger feeder for the farther rack
Symmetrical cabinet arrangement

The best solution depends on the physical installation.

Electrical layout should be planned together with rack placement.

9. Rack-Level Protection

Each rack may require its own:

DC fuse
Breaker
Disconnect

before joining the common main bus.

This provides:

Fault isolation
Maintenance isolation
Protection of rack feeder cable

The exact protection design must coordinate with:

Battery branch protection
Main protection
Available fault current

10. Module Protection and Rack Protection Are Different Layers

A system may have:

Individual battery branch fuse

protecting each module.

Rack breaker

protecting rack feeder.

Main DC breaker

protecting common inverter feed.

Each has a different role.

Adding a second rack often requires reviewing the protection hierarchy rather than simply copying the first rack.

11. Fault Current Increases With a Second Rack

If one Rack A feeder faults, Rack B may also feed that fault through the common bus.

The rack protection device must be capable of safely interrupting the available DC fault current.

Large parallel racks should be treated as high-energy DC systems.

12. Main Busbar Must Support Both Racks

Suppose:

Rack A potential operating current: 250A

Rack B: 250A

Combined maximum system current: 500A.

The main distribution bus must be engineered for the actual system maximum.

A busbar originally designed for one 250A rack may not remain adequate after adding the second rack.

13. Main Inverter Cable Must Also Be Reviewed

The two racks may supply more current than one rack.

But the inverter cable is still the single combined path.

If the inverter power also increased, review:

Main positive cable
Main negative cable
Main breaker
Inverter terminal capacity

Do not upgrade only the battery side.

14. What If the Inverter Power Does Not Change?

Example: One 8kW inverter.

Original: Rack A only.

Expansion: Rack A + Rack B.

If inverter remains 8kW:

main inverter current is broadly unchanged.

The second rack mainly provides:

More runtime
Lower current per module
More redundancy

The existing main bus may already be adequate if originally correctly sized.

But this must be verified, not assumed.

15. Communication Architecture Across Two Racks

Power wiring is only half the system.

The eight BMS modules may need to operate as one communication network.

Possible architecture:

Rack A modules
→ Rack A communication chain

Rack B modules
→ Rack B chain

Then a designated:

Master battery
Master BMS
Battery management unit

communicates with the inverter.

The exact design is manufacturer-specific.

16. Do Not Create Two Independent Masters Without a Defined Architecture

Suppose Rack A has Master A.

Rack B has Master B.

Both attempt to communicate with one inverter CAN port.

Possible outcomes:

CAN collision
Incorrect SOC
One rack ignored
Wrong current limit

Unless the battery system explicitly supports two rack masters with a central controller, one coordinated communication architecture should be used.

17. Rack Quantity Must Be Reflected in Total Capacity

Four 100Ah modules per rack:

Rack A: 400Ah.

Rack B: 400Ah.

Combined: 800Ah nominal

at the same system voltage.

After expansion, verify the inverter/master reports the correct:

Battery count
Ah
kWh
Charge-current limit
Discharge-current limit

If the inverter still sees only Rack A, SOC and current management may be incorrect.

18. Rack SOC Can Diverge

Even if all batteries started matched:

Rack A may gradually reach: 40% SOC

while Rack B remains: 55%.

Possible reasons:

Rack current imbalance
Different temperature
Communication differences
Different battery age

Monitor average rack SOC as well as individual module SOC.

19. One Rack Can Age Faster

If Rack A consistently carries:

60% of system throughput

and Rack B: 40%,

Rack A batteries may accumulate more:

Ah throughput
Cycle count
Heat exposure

After several years, the resistance difference between racks can increase further.

Early rack imbalance can therefore become long-term unequal aging.

20. Different Rack Ages Complicate Expansion

Original Rack A: 3 years old.

New Rack B: brand new.

The new rack may have:

Higher capacity
Lower internal resistance

It may carry more current after expansion.

Or the shorter original Rack A cable may dominate and keep Rack A heavily loaded.

Actual current must be measured rather than predicted only from age.

21. Commissioning Test 1 — Rack A Only

Before combining racks, verify Rack A independently where appropriate:

Module communication
Voltage
Branch current
Alarms

22. Test 2 — Rack B Only

Repeat for Rack B.

This confirms the new rack itself is functional before integrating the two.

23. Test 3 — Both Racks at Moderate Load

Apply a stable load.

Record:

Rack A feeder current
Rack B feeder current
Individual module current
Rack voltage
Temperature

Example:

Total:

200A

Rack A: 103A
Rack B: 97A

Reasonable.

24. Test 4 — Higher Approved Load

Now increase load.

Sometimes rack imbalance appears only at high current.

At low load:

A: 50A
B: 48A

At high load:

A: 190A
B: 110A

Now investigate:

Rack feeder resistance
Breaker drop
Cable length
Rack internal busbar

25. Test 5 — Charging Current

Rack sharing should also be checked during charging.

Example:

Total charge: 240A.

Rack A: 160A.

Rack B: 80A.

If both racks have similar SOC, this may indicate a resistance or BMS issue.

If Rack B is almost full, the difference may be normal.

26. Test 6 — One Rack Isolated

If the system is designed for rack-level redundancy, test an approved N-1 condition.

At critical load:

Isolate Rack B
Confirm Rack A current remains within limits

Then test the opposite condition if the architecture allows.

This verifies that both racks can actually support the intended fallback load.

27. Rack A Works Alone but Rack B Does Not

Possible causes:

Rack B main breaker
Communication master configuration
Feeder cable
Pre-charge
Battery count
BMS current limit

A redundant architecture should not assume both racks are independently usable until tested.

28. Do Not Stack Both Rack Feeders on an Undersized Inverter Terminal

If the inverter has only one intended battery connection, use a properly engineered DC distribution system rather than forcing:

Rack A cable
Rack B cable

onto one terminal bolt.

Mechanical and electrical connection quality matters.

29. Thermal Inspection at Rack Level

Under sustained load compare:

Rack A breaker
Rack B breaker
Rack feeders
Busbar joints

Example:

Rack A breaker: 34°C.

Rack B breaker: 61°C.

Even if current is similar, Rack B protection may have high contact resistance.

30. Rack-to-Rack Troubleshooting Sequence

If one rack carries much less current:

Compare rack SOC.
Compare rack feeder current.
Compare rack voltage under load.
Measure voltage drop across rack breaker.
Measure rack feeder drop.
Compare internal module currents.
Check communication.
Inspect temperature.
Review rack age and capacity.

This narrows the problem from system level to rack level before examining individual cells.

31. Two Racks Should Not Be Treated as One Giant Cable Chain

A clean architecture makes every hierarchy clear:

Battery module

→ module branch

→ rack bus

→ rack feeder/protection

→ main DC bus

→ inverter branch

This is easier to:

Maintain
Expand
Troubleshoot
Document

than repeatedly adding modules to an informal daisy chain.

32. Future Third-Rack Expansion

If a third rack may be added later, plan:

Main busbar capacity
Spare rack protection position
Communication address range
Floor/rack space

during the two-rack design.

Avoid rebuilding the complete DC combiner every time capacity increases.

33. Two-Rack Design Checklist

Item	Check
Rack module compatibility	Confirmed
Rack A/B nominal voltage	Matched
Rack feeder cable	Engineered
Rack feeder resistance	Comparable
Rack protection	Installed
Main bus capacity	Verified
Main inverter cable	Verified
Communication architecture	Defined
Master BMS	Defined
Total battery count	Correct
Rack current sharing	Tested
Charge sharing	Tested
N-1 operation if required	Tested
Thermal inspection	Completed

Frequently Asked Questions

Can two LiFePO4 battery racks connect to one inverter?

Yes, when the batteries, DC distribution, protection and BMS communication architecture are designed for common operation.

Should Rack B connect through Rack A?

A common main DC bus usually provides a cleaner and more controllable architecture than daisy-chaining one rack through another.

Do both rack feeder cables need the same length?

The goal is similar effective rack-level resistance. Cable cross-section and complete electrical path also matter.

Should each rack have its own breaker?

Rack-level isolation/protection is often useful in larger systems, but the exact protection design is project-specific.

Can one rack be new and the other several years old?

Possibly, but capacity, resistance, firmware and current-sharing differences should be evaluated.

How do I know whether the racks are balanced?

Measure total feeder current from each rack during both charge and discharge under controlled conditions.

Conclusion

Once a LiFePO4 battery system expands into multiple physical racks, installers should stop thinking only in terms of individual parallel batteries.

Each rack becomes a sub-bank with its own:

Internal current sharing
Feeder resistance
Protection
Communication
Thermal behaviour

A good two-rack design connects both racks to a properly engineered common DC distribution system rather than making one rack the electrical path for the other.

For B2B ESS projects, rack-level architecture also makes future:

Expansion
Maintenance
N+1 redundancy
Fault isolation

much easier to manage.

HIZN Lithium supplies modular 48V/51.2V LiFePO4 battery systems for residential solar, telecom, UPS and commercial energy-storage projects with scalable rack configurations, CAN/RS485 communication and OEM support.