What a three-phase hybrid inverter adds for C&I sites
A three-phase hybrid inverter is a three-phase on-grid inverter with bidirectional battery control added on top. That addition is what enables peak shaving, tariff arbitrage, and EPS backup with fine, per-phase export regulation. This guide focuses on where that combination fits in commercial and industrial sites, not on three-phase power fundamentals.
A standard three-phase on-grid inverter only converts PV DC to synchronized three-phase AC and follows the grid. A three-phase hybrid inverter goes further, coordinating PV, grid, and battery together.
| Capability | Standard three-phase on-grid inverter | Three-phase hybrid inverter |
|---|---|---|
| Primary function | PV to three-phase AC, grid synchronized | PV, grid, and battery coordination, bidirectional |
| Energy paths | PV to grid or site only | PV to site or battery, battery to site or grid, grid to battery under TOU or EMS logic |
| Storage integration | External AC-coupled only, limited control | Native DC-coupled and AC-coupled options, SOC and DOD governed by the inverter’s BMS interface |
| Export regulation | Coarse limiter or relay-based curtailment | Continuous, setpoint-based export control per phase or total site |
| Demand charge management | Indirect, relies on EMS curtailment | Direct, schedules battery discharge for peak shaving |
| Backup capability | None, shuts down on outage | EPS feeds priority loads during grid loss |
| Phase balancing | Follows grid phases passively | Applies active phase balancing and buffers rapid swings with storage |
Both topologies synthesize three sinusoidal phases from a shared DC bus. Only the hybrid version buffers transients and enforces site-wide power targets on top of that. That’s what improves phase balancing under variable loads, and what makes unbalanced load handling possible without nuisance trips.
When a C&I site should specify three-phase
Three-phase makes sense in a handful of situations, and each comes with its own design consideration.
- Three-phase utility service with three-phase loads. Native three-phase generation simplifies phase balancing and feeder loading, avoiding phase-by-phase PV fragmentation and mixed EPS coverage.
- Motor or compressor loads (VFDs, chillers, pumps). Higher starting currents and torque transients get shared across phases; size for kVA and allowable voltage sag at motor start.
- High peaks or demand charges. Storage pairs with three-phase for peak shaving across all phases, and a hybrid topology keeps DC-coupled response fast under TOU peaks.
- Long feeders to subpanels or pumps. Lower conductor losses and better voltage drop management than single-phase, though cable runs, transformer impedance, and tap settings still need evaluating.
- Harmonic limits and reactive power targets. Three-phase inverters manage current symmetry and power factor more predictably; VAR schedules and flicker limits still need coordinating with the interconnect.
- Export-limited or reverse-power-protected interconnects. Tighter control of export limits across all phases, though CT placement and per-phase setpoints must satisfy the meter.
- Planned or future batteries. A unified battery bus serves every phase for site-wide EPS; reserve space, cooling, and BMS integration for growth.
- Transformer-rated metering and utility telemetry. Cleaner compliance with C&I metering requirements, confirmed against CT ratios, polarity, and SCADA points with the utility.
Phase balancing and unbalanced load handling
A three-phase hybrid inverter manages phase balance in real time. It regulates per-phase current and voltage, then routes PV and battery power toward whichever leg is most loaded. Modern control loops sample current on all three phases every cycle. That’s what lets them react to an imbalance almost immediately, rather than applying a slow, symmetric cap that trips under stress.
Three things matter most in that control loop. Per-phase current has to stay within rating without losing three-phase sync. Get this wrong and the failure mode is breaker trips or thermal derating. Neutral current needs minimizing through phase sharing and harmonic suppression. Get this wrong and the neutral overheats, or the neutral-to-ground voltage rises. PV and battery power need dispatching to the most loaded phase quickly. Get this wrong and the site ends up exporting on light legs while a deficit remains on the heavy one.
In practice, this plays out as two patterns. When one phase runs heavy, the inverter boosts that leg from the DC link. It restricts export on the lighter legs and holds each phase within its rated limit. When the imbalance is only mild, it shares power symmetrically instead, using reactive power control to minimize harmonics.
Smart energy management in a three-phase context
Grid export, zero export, and EPS work the same way conceptually on a three-phase system as on any hybrid inverter. The underlying mechanics are covered in separate guides on export limitation and outage backup. What’s different on three-phase is that every one of them has to be enforced per leg, not just at the site total.
A handful of commissioning details make or break accurate control in the field, and most of them are specific to having three legs instead of one or two. Meter polarity and the import/export sign convention need checking on each phase individually. Phase rotation and mapping matter too, confirming L1, L2, and L3 run correctly from meter to inverter rather than being swapped. CT ratio, arrow orientation, and burden all need verifying on all three CTs, not just one. Time sync and TOU schedule alignment round out the list.
EPS behavior when the grid fails
Anti-islanding opens the grid relays first. The inverter then forms its own three-phase reference at compliant voltage and frequency for motors and IT loads. EPS enforces per-phase current limits during this island. It tolerates a realistic amount of imbalance, and can derate or shed a phase to protect the DC bus if needed. When the grid returns, the controller resynchronizes phase angle, frequency, and voltage before reclosing. It then restores whatever export policy was active before the outage.
The battery covers the transient kVA during this whole sequence. PV supports continuous kW, subject to irradiance and temperature. Tiered load shedding, combined with correct neutral bonding, is what keeps the island stable throughout.
Sizing a three-phase hybrid inverter for C&I
Sizing works through five stages. These are demand and phase split, AC stage sizing, DC/AC and MPPT matching, battery power and energy, and grid/export settings.
| Stage | What to calculate | Key checks |
|---|---|---|
| Demand and phases | 12-month max kW and the worst phase split | Per-phase limits, motor inrush, transformer reverse power |
| AC stage sizing | Continuous kW, short-term kVA headroom, temperature and altitude derate | Unbalance tolerance, protection, compliance |
| DC/AC and MPPT | kWp, strings per MPPT, Voc at minimum temperature, Vmp at maximum | MPPT window, clipping risk by irradiance and temperature |
| Battery power and energy | Shave kW, event duration in hours, usable kWh by DOD and SOC | BMS current, C-rate, cabling losses |
| Grid and export | CT placement, export setpoints per feeder or service | Anti-islanding, per-phase export limits |
A worked example shows why the phase split matters, not just the site total.
- Total site peak: 480 kW
- Naive assumption (evenly split): 480 ÷ 3 = 160 kW per phase
- Actual worst-phase reading from interval data: one leg carries 40% of the total = 192 kW on that single phase
- Extra headroom needed on that leg alone: 192 − 160 = 32 kW, about 20% more than the balanced assumption would suggest
Sizing off the site-wide average alone would undersize that one leg by 32 kW. The inverter’s per-phase current rating, not just its total kW rating, has to cover the worst phase specifically. This is the core reason single-phase sizing math, like the 10kW Hybrid Inverter Sizing Guide uses, doesn’t translate directly to a three-phase site without checking the phase split first.
Conclusion
A commercial three-phase hybrid inverter earns its place under a few conditions. The site already carries balanced three-phase feeders, sees frequent motor starts, or faces strict export caps that benefit from integrated storage. Control quality decides the outcome more than the headline spec. Phase balancing and genuine unbalanced load handling keep feeders within limits. Grid export, zero export, and EPS modes match the site to its actual tariff and compliance strategy.
For interconnect approvals and commissioning steps, that’s covered in more depth in the guide on regulatory compliance for hybrid inverters. For the broader system architecture this guide sits within, see the hybrid and energy storage inverter overview.
FAQ
Do I need zero export on a commercial three-phase system?
Only when the utility forbids feed-in or caps it. Zero export uses a meter or CT at the point of common coupling to trim inverter output so no energy crosses the boundary. Where export is paid, a normal grid export mode with a TOU schedule usually returns more value.
How fast does EPS output start after a grid outage?
Most three-phase hybrid inverters restore EPS output within 10 to 20 milliseconds. That’s short enough for lighting, refrigeration, and standard IT loads, but sensitive process controllers may still reboot. Sites with zero-tolerance equipment should add a small online UPS in front of that specific panel.
Can a three-phase hybrid inverter run on an unbalanced site?
Yes. It monitors each phase separately and dispatches battery power per phase, so a heavy single-phase load doesn’t starve the other two. Check the datasheet for the allowed per-phase imbalance ratio, since that limit decides whether one large motor can safely sit on a single leg.
Can I add batteries later to an existing three-phase on-grid PV system?
Yes, through either an AC-coupled or DC-coupled retrofit path. Choosing a hybrid-ready inverter from the start, even before batteries are installed, generally makes that later upgrade simpler and cheaper than retrofitting a pure grid-tie unit.
Which grid codes and certifications should I verify before specifying a C&I inverter?
Regional grid codes, anti-islanding certification, protection class ratings, and the utility’s specific interconnection requirements all need confirming before specifying equipment. These vary by jurisdiction and change periodically.