Solar Power Factor Correction: Inverter kVA Limits

A commercial site can add solar and still retain a power-factor-related charge. It can also see its grid-meter power factor fall even though the PV inverter reports unity power factor. Before proposing a setting change or extra equipment, establish what the meter measures and what the tariff actually charges.

Last updated: November 2026.

Start at the facility connection point

Separate three questions:

• What is the load’s active and reactive demand?
• What active and reactive power does the inverter exchange?
• What net flows does the utility meter record at the connection point?

The readings need not match. Site loads, transformer effects, existing compensation equipment and other generators can sit between the inverter and the billing meter. A fixed inverter power-factor setting cannot by itself establish a facility-meter target.

For terminology, see the power factor definition and the explainer on active and reactive power. This article focuses on the site assessment, control constraints and evidence needed for a commercial decision.

Distinguish true power factor from displacement power factor

Power factor is the ratio of active power P, in kW, to apparent power S, in kVA. State the meter convention, including how it treats import and export direction.

Under sinusoidal conditions, displacement power factor is cos φ, and the familiar power triangle gives:

S = sqrt(P² + Q²)
PF = P / S

Here Q is reactive power in kvar. These equations are used in the simplified examples below. With distorted waveforms, true power factor and displacement power factor can differ. Do not interpret cos φ or a fundamental-frequency Q reading as a complete harmonic assessment.

Fluke’s power-quality application note explains the distinction and the interactions between harmonics and correction capacitors. It is a reason to obtain suitable power-quality measurements, not to assume every low-PF problem can be solved with additional kvar.

Why unity-PF solar can lower grid power factor

Consider hypothetical balanced, sinusoidal measurements at one operating instant. Assume constant site demand, no storage and no other compensation changes. Use a convention where positive Q means the site’s inductive reactive demand supplied from the grid.

In the second row, active imports decrease, but reactive demand is unchanged. The meter’s PF falls. Apparent import also falls, so a lower PF does not automatically mean higher current or a larger charge than before solar. The tariff determines the financial effect.

The third row assumes an available local source offsets 45 kvar in the stated direction. It illustrates the site balance, not an instruction to set an inverter to a particular leading or lagging value. Manufacturers use different sign and excitation conventions; the approved design must identify them.

A target requires a site balance, not an inverter label

If the engineer evaluates a hypothetical 0.95 displacement-PF target at the 30 kW grid-import operating point, the permissible reactive magnitude in this simplified model is:

Q_grid_target = 30 × tan(arccos(0.95)) = 9.86 kvar
Required local offset = 60 − 9.86 = 50.14 kvar

That calculation does not authorize the target, prove the control can achieve it or establish a tariff saving. The target must come from the actual project requirements. Near zero active import or during export, the meter’s quadrant and averaging rules become especially important.

Read the inverter’s real capability envelope

A kW active-power rating and a kVA apparent-power rating describe different limits. In the ideal sinusoidal envelope, a rated apparent power S rated gives:

|Q| ≤ sqrt(S_rated² − P²)

For example, 110 kVA at 100 kW gives 45.8 kvar of geometric headroom. At a hypothetical 110 kW active output, that same envelope has no reactive headroom. This is a screening calculation, not a guaranteed manufacturer operating point.

Fronius’s country-setup manual documents model-specific reactive limits, maximum apparent power/current, country-setting constraints and active/reactive priority. Its different listed ranges show why one claimed “0.8 PF capability” cannot represent every inverter family or operating point.

For the earlier site’s 70 kW PV output and 50.14 kvar illustrative offset, the ideal combined apparent power is about 86.1 kVA. Check the actual model, temperature, voltage and current limits before interpreting that as achievable. Do not add a universal sizing margin or assume a mathematical circle overrides the manufacturer’s curve.

Active-power priority changes the result

When a requested operating point reaches a device limit, active-power priority and reactive-power priority can lead to different outputs. A design that reserves kvar may reduce active output at some times; a design that prioritizes active output may leave the reactive target unmet.

This does not mean every kvar supplied causes an equal annual generation loss. Evaluate the chronological operating points, device limits and approved priority. Extra losses and thermal conditions also need model-specific evidence. A nameplate PF setting alone cannot establish an annual yield percentage.

Evaluate control at the right point

SMA’s Data Manager guidance discusses feedback from the point of interconnection and warns that PV inverters may be unable to meet connection-point setpoints where self-consumption and reactive demand are high. The measurement architecture and available equipment capability both matter.

A facility-correction objective and a network-voltage objective can call for different reactive behavior. Do not assume modes can be enabled together without a documented priority scheme. Existing capacitor banks, plant controllers and other reactive sources also need coordination.

Fronius limits parameter adjustment to authorized specialists and approved or requested utility requirements. Obtain the approved control design and connection conditions; this guide does not provide commissioning passwords, curve settings or DIY wiring instructions.

Verify nighttime capability instead of assuming a battery is required

Some solar-only models can remain grid-connected for reactive operation outside PV feed-in hours. SMA’s documented Q on Demand 24/7 function uses grid power for internal assemblies and describes conditions affecting overnight operation and restart.

That is evidence for the documented model/function, not every SMA inverter or every PV inverter. Check the exact hardware, firmware, permitted mode, auxiliary consumption and availability after outages. Do not claim a universal “battery kWh per kvar” requirement or treat nighttime support as free energy.

Determine the actual billing mechanism

Obtain the current tariff and account classification before estimating value. Possible mechanisms include apparent-power demand charges, reactive-energy charges or a power-factor adjustment. They require different measurements and calculations; a generic national penalty table is not sufficient.

Record:

• Whether the meter reports true or displacement PF and its averaging method.
• Which demand interval, billing period and import/export quadrant apply.
• Whether billed demand includes a minimum, lookback or ratchet.
• How reactive energy is counted and whether leading and lagging quantities are treated differently.
• The actual rates, exceptions and eligibility conditions for that account.

Do not divide an independently observed monthly kW maximum by a monthly PF average and call the result a measured kVA peak. Those quantities may come from different intervals. Likewise, a PF average cannot identify how much compensation is required at every operating point.

A review package for a facility engineer

Collect the connection agreement, current tariff, bills and compatible interval P/Q/S measurements. Include voltage, true/displacement PF and power-quality findings where appropriate. Cover representative production, operating, closed and nighttime periods; identify missing records and unusual events.

Use load profile analysis to describe the time basis. For harmonics and power quality, retain the measured evidence rather than assuming inverter reactive support addresses distortion.

The engineer’s comparison should show the base case, available inverter capability, proposed approved control objective, unmet periods and expected energy/charge changes. Include any additional equipment, auxiliary consumption, active-output restrictions and maintenance costs. Compare equivalent billing periods and verify the result against meter records after an approved implementation.

A software evaluation can start with the generation and financial workflow and a demonstration using the project’s records. Confirm which calculations and reports are available; this article does not claim SurgePV automatically performs reactive dispatch or removes PF charges.

Methodology and limitations

SurgePV publishes this guide as a solar software provider. The worked examples are independent balanced-sinusoidal calculations, not measured customer projects or verified savings cases. Manufacturer documents support bounded capability/control observations; the exact connection agreement and tariff remain necessary for a facility decision. No universal penalty, country-code curve, annual yield loss, battery requirement or payback is claimed. Sources last reviewed September 30, 2026.

Frequently Asked Questions

What is power factor correction with a solar inverter?

It uses available inverter reactive-power capability as part of a site or grid-control strategy. Whether it improves power factor at the facility meter depends on the site load, measurement point, sign convention, approved control mode and equipment limits.

Why can solar make grid-meter power factor worse?

Unity-power-factor PV can reduce the active power imported from the grid without reducing the site’s reactive demand. The grid supplies a different ratio of active to apparent power, so the measured grid power factor can fall even though the inverter itself operates at unity power factor.

What power factor threshold triggers utility penalties?

There is no universal threshold or penalty formula for every site. Read the current tariff and meter definitions for the account. Power-factor adjustments, apparent-power demand charges and reactive-energy charges are different billing mechanisms.

How much reactive power can a solar inverter provide?

A simplified sinusoidal model gives an apparent-power envelope of Q squared plus P squared no greater than S rated squared. The actual operating range can be smaller because of model-specific reactive limits, current, voltage, temperature, firmware, control priority and connection requirements.

Does reactive-power support reduce solar generation?

It can reduce active output when the requested operating point exceeds the available current or apparent-power envelope and reactive power has priority. A fixed percentage of annual yield loss cannot be inferred from the power-factor setting alone.

Can a solar-only inverter provide reactive power at night?

Some documented models support nighttime reactive operation using grid power for their internal losses. This is not universal, and a battery is not automatically required. Verify the exact model’s functions, operating conditions, connection agreement and measured consumption.

What is the difference between kW, kVA and kvar?

kW is active power, kvar is reactive power and kVA is apparent power, the combination of the two. Power factor is active power divided by apparent power, so a 100 kW load at 0.8 power factor draws 125 kVA.

Sources

• Fluke power-quality application note
• Fronius country-setup manual
• SMA Q on Demand 24/7 documentation