How Does Power Sharing Work in a Multi-Connector DC Charger?

Teams comparing multi-connector DC charger quotations can mistake plug count and headline kilowatts for simultaneously available power. The combined per-connector ratings may exceed what the cabinet and site can release, so the issue is specification clarity rather than necessarily defective hardware.

Summary: multi-connector DC charger power sharing allocates a limited pool among active vehicles; it does not guarantee the maximum connector rating at every outlet at the same time. Current U.S. rules for covered corridor-serving DCFC sites permit power sharing only while each charging port continues to meet an EV’s request up to 150 kW. Buyers should specify the cabinet pool, per-connector voltage/current envelope, allocation mode, module granularity, site cap, and a witnessed concurrent test before comparing bids.

Multi-connector DC charger power sharing allocates a limited available power pool among active vehicles; it does not automatically mean that every connector receives its maximum rating at the same time. The actual result depends on the charger’s internal architecture, allocation mode, connector limits, vehicle demand, and any site-level power limit. Buyers should therefore verify the named model’s simultaneous-use and allocation behavior before treating a connector count as delivered capacity.

Part 1. What does power sharing mean in a multi-connector DC charger?

Power sharing is the distribution of DC charging capacity among more than one active connector. A dynamic system can reallocate capacity as vehicles connect, taper, or disconnect. Kempower’s charging-power-management documentation describes a power-distribution module routing and re-routing available channels during a charging session.

The key phrase is “available power.” A charger may have a cabinet-level power pool, individual connector limits, and a site-level cap. Those three numbers can differ. A product label or a two-connector front panel alone does not establish how much power two vehicles can receive simultaneously.

For the NEVI rule’s regulatory scope, current 23 CFR 680.104 defines Power Sharing as “dynamically limiting the charging power output of individual charging ports at the same charging station to ensure that the sum total power output to all EVs concurrently charging remains below a maximum power threshold,” and says this is also called automated load management.

The current text of 23 CFR 680.106(d)(1) makes the cabinet-versus-port distinction concrete for covered U.S. corridor projects: each DCFC port must support 250 to 920 V DC, have a continuous rating of at least 150 kW, and meet an EV’s request up to 150 kW simultaneously. Power sharing is allowed only while every port continues to meet that request. This funding- and location-specific requirement is not a claim about every commercial charger.

Part 2. Which components set the available power pool?

DC charging equipment converts incoming electrical supply into controlled DC output. In some architectures, power modules are grouped in one cabinet and routed to one or more connectors or dispensers. In other designs, sharing is more limited by fixed internal paths or connector rules.

The EVBox PowerHub description is one example of a centralised architecture that routes power to multiple charging points. It also distinguishes static delivery from optional dynamic load sharing. That distinction matters: centralised architecture does not by itself prove that every deployment uses dynamic allocation.

A sound comparison separates the cabinet power pool and connector rating. At any moment, one connector is bounded by the vehicle’s requested voltage and current, its own cable and connector envelope, the modules or channels that can physically reach it, the remaining cabinet capacity, and the site’s active limit. The lowest applicable constraint governs; unused cabinet kilowatts cannot overcome a vehicle current limit or an unroutable module.

Module granularity turns architecture into an operating constraint. Ask for the allocation increment, minimum share, cross-connector routing restrictions, and behavior when a request falls between available increments; do not infer any of these from total cabinet power.

Part 3. Does power always split equally between active connectors?

No. Some systems use a fixed split, some use equal sharing, some prioritise one connector, and some dynamically assign available channels according to demand. The BTC Power explanation of dynamic sharing describes real-time allocation rather than a permanent equal division.

Real buyer questions reflect this difference: some units allow only one active vehicle, some use a 50/50 split, and some change allocation with finer granularity. That language is not a model specification; it reinforces why a quotation must name the allocation mode.

For procurement, dynamic power allocation is not a sufficient specification on its own. The RFQ should state whether released capacity moves automatically, the module or channel increment, priority rules, acceptable response time, and the fallback state if communication with a site controller is lost.

Important: Do not promise a vehicle a particular charging rate from cabinet power alone. Vehicle acceptance, connector limits, temperature, state of charge, and site controls can all reduce delivered power. Kempower’s charging-power-management documentation describes allocation as equipment-configuration dependent.

Part 4. What happens when one vehicle tapers or disconnects?

Vehicle charging demand commonly changes during a session. In a dynamic architecture, capacity that is no longer needed by one vehicle can become available to another active connector, subject to the charger’s allocation rules and limits. BTC Power uses this tapering behaviour to explain why dynamic allocation can use a shared pool more effectively than a fixed split.

The reallocation is not universal. It depends on the named charger’s routing design, minimum allocation increments, firmware/configuration, active connector limits, and the site’s permitted input. A commissioning test should confirm the observed response with representative vehicles or an approved test method, rather than relying only on a diagram.

Write that commissioning check as controlled cases rather than one convenient session: two high-demand vehicles, unequal requests, a reduced request or taper, a disconnect, a lower site cap, and the agreed communications-loss fallback. The contract must set the instruments, stabilization time, tolerances, and pass/fail values; this article does not invent them.

Part 5. Which limits can reduce the power each vehicle receives?

The power visible to a driver is the result of several constraints working together. A multi-connector charger can have unused cabinet capacity while a particular vehicle is limited by its acceptance curve. Conversely, two vehicles can request more than the cabinet or site is allowed to deliver.

The commercial EV charging load-profiling guide should be completed before selecting a shared-power configuration. It identifies the site demand and concurrency assumptions that product ratings cannot answer on their own.

The site-control schedule should also state the normal cap, governing meter, update interval, ramp behavior, priorities for other building loads, and safe fallback if the meter or communications path fails. Retain the configured limit and synchronized meter/controller trace as acceptance evidence.

Which standards or rules belong in the specification?

The Open Charge Alliance’s OCPP overview identifies OCPP 1.6, 2.0.1, and 2.1 as available versions and defines the protocol’s role between charging stations and charging management systems. Ask for evidence against the contracted version and functions, while keeping electrical output and routing acceptance as a separate test.

For total-cost comparison, separate charger price from upstream capacity, switchgear, installation, networking, testing, and the operational effect of queuing or missed departure targets. An illustrative model may compare installed cost with energy delivered inside required dwell windows, but its inputs must come from the site’s load profile and contracted performance—not an assumed gain from “dynamic” sharing.

Part 6. What should buyers verify before specifying a shared-power charger?

An RFQ needs explicit sharing questions. “Two connectors” is not enough to compare offers.

How should the concurrent test be evidenced?

A useful concurrent charging acceptance test places at least two connectors under demand at the same time with representative vehicles or approved EV emulators. Retain synchronized records of each request, connector voltage/current/power, cabinet total, configured site cap, allocation state, firmware/configuration, alarms, and timestamps. A diagram explains intent; the trace shows whether the purchased configuration met the contract.

Part 7. When does a multi-connector DC architecture fit the site?

Shared-power DC architecture can fit commercial sites, fleets, or public locations where concurrent charging is expected and a measured demand model supports the chosen pool. It may be less suitable where each vehicle requires a guaranteed dedicated rate, where the site capacity is not established, or where the model documentation does not clearly state allocation behaviour.

XYDF publishes DC fast charger and commercial solution routes, including a split DC station product route. These are appropriate starting points for a product discussion. They do not, without model-specific evidence, establish a sharing algorithm, module granularity, simultaneous-use behaviour, or delivered-power outcome.

For bid comparison, ask each supplier to return a model-specific compliance matrix, power-path diagram, installed-module schedule, voltage-current curves, configuration record, OCPP evidence where required, and proposed site acceptance procedure. Unsupported “compliant” or “dynamic” labels can leave offers incomparable or cause acceptance disputes.

For an informed discussion, send XYDF the required connector count, vehicle mix, dwell and departure pattern, site power limit, target throughput, preferred allocation policy, and expansion plan.

FAQ

Can a dual-connector DC charger charge two vehicles at once?

It can only do so if the named model supports simultaneous operation. Verify this in the model documentation and quotation.

Does power always split 50/50?

No. Some systems use a fixed split, while others use equal, priority-based, or dynamic allocation.

What happens when one EV tapers?

In a dynamic system, unused capacity may be reassigned to another active connector, subject to the charger’s rules and limits.

Is connector power the same as cabinet power?

Not necessarily. Cabinet capacity is the shared pool; connector and cable limits can cap the output available at an individual outlet.

What limits shared charging power?

Site caps, cabinet capacity, routing design, connector limits, vehicle acceptance, allocation rules, and operating conditions can all limit it.

Can site load management reduce charger output?

Yes, when the site control design applies a cap to charging. Verify the integration, settings, and fallback behaviour for the specific project.

What should a buyer verify in an RFQ?

Verify simultaneous use, cabinet and connector limits, allocation mode, routing granularity, site-control interface, and commissioning test method.

References

• Electronic Code of Federal Regulations: 23 CFR 680.104 Definitions
• Electronic Code of Federal Regulations: 23 CFR 680.106
• Federal Register: National Electric Vehicle Infrastructure Standards and Requirements
• Open Charge Alliance: Open Charge Point Protocol
• Kempower Charging Power Management Documentation
• BTC Power: Dynamic Power Sharing
• EVBox PowerHub Centralised Architecture
• EV Connect: EV Charger Load Management

The durable rule is simple: size the shared pool from concurrent demand, then make the allocation promise measurable.

When the load profile and acceptance criteria are ready, contact XYDF to review the charger configuration and RFQ evidence.