3.1 Inverter Types and Topologies

3.1 Inverter Types and Topologies

Quick Answer: Modern photovoltaic installations utilize four primary inverter topologies: central inverters for utility-scale MW blocks, string inverters for multi-kilowatt commercial and residential arrays, microinverters for distributed AC conversion at the module frame, and DC power optimizers paired with string inverters. Transformerless (non-isolated) inverters dominate contemporary design due to higher efficiency and lighter weight, operating with neither dc conductor solidly grounded and relying on sensitive residual current detection and ground-fault protection (NEC 690.41).

In a photovoltaic (PV) power system, the inverter serves as the foundational energy conversion engine. Beyond performing direct current (DC) to alternating current (AC) power conversion, the inverter executes maximum power point tracking (MPPT), enforces system safety disconnects, and actively regulates grid stability. Selecting an inverter topology dictates array wiring methods, balance-of-system (BOS) component selection, fire-safety compliance under National Electrical Code (NEC) Article 690, and overall system energy yield under real-world shading conditions.

---

Central Inverters in Commercial and Utility-Scale Systems

Central inverters represent the traditional utility-scale and large commercial topology, concentrating the power conversion of hundreds or thousands of series-parallel PV module strings into a single monolithic conversion unit.

Electrical Architecture and Power Ratings

• Capacity Range: Typically spans 500 kW to 4.5+ MW per individual enclosure.
• DC Operating Voltage: Modern utility systems operate at up to 1500Vdc; under NEC 690.7, listed dc PV equipment rated 1500 V or less that is not on or in a building is exempt from Parts II and III of Article 490. Higher voltage minimizes resistive ($I^2R$) conduction losses across extensive ground-mount arrays compared to legacy 600Vdc or 1000Vdc designs.
• Combiner Infrastructure: PV strings connect in parallel within array junction boxes, which feed intermediate DC combiner boxes or master recombiners before terminating at the central inverter DC busbars.
• Medium-Voltage Integration: Central inverters generally output three-phase AC at 480Vac, 600Vac, or 690Vac, which immediately couples to a dedicated pad-mounted step-up transformer converting voltage to distribution or transmission collection levels (e.g., 12.47 kV to 34.5 kV).

Engineering Advantages and Vulnerabilities

Central inverters offer the lowest initial equipment capital expense on a per-watt basis (frequently $0.04 to $0.08/W for the power conversion equipment). Maintenance is centralized at accessible equipment pads rather than dispersed across vast racking fields. However, central inverters present distinct operational drawbacks:

1. Single Point of Failure: An internal fault in a 2.5 MW central inverter drops the entire generation block offline until certified field technicians perform repairs.
2. Array-Level MPPT Limitations: A single MPPT channel oversees thousands of modules. Non-uniform soiling, vegetative shading, or localized cloud cover forces the MPPT algorithm to track a compromised average I-V curve, creating substantial mismatch power losses across the array field.
3. Heavy Installation Logistics: Multi-ton enclosures require reinforced concrete foundations, crane rigging, and specialized heavy civil logistics.

---

String Inverters: Architecture and Multi-MPPT Flexibility

A string inverter aggregates a series connection of PV modules (a "string") into a compact, wall-mounted or rack-mounted conversion unit. Standard residential units range from 3 kW to 11.4 kW single-phase (120/240V split-phase or 208V), while commercial three-phase string inverters range from 20 kW to 250+ kW (208Vac or 480Vac wye).

[Module 1]--(+)--[Module 2]--(+)-- ... --(+)--[Module n] ---> String Inverter DC Input (MPPT 1)
[Module 1]--(+)--[Module 2]--(+)-- ... --(+)--[Module n] ---> String Inverter DC Input (MPPT 2)

Series String Physics and Mismatch Vulnerability

In a series circuit, current ($I$) remains identical through every component while voltages sum ($V_{\text{string}} = \sum V_{\text{module}}$). The primary constraint of a simple string inverter is current bottlenecking: if one module in a 12-module series string experiences 40% surface shading from a plumbing vent or tree branch, its internal photogenerated current drops. Although bypass diodes activate to route current around shaded cell groups, the entire string's operating current ($I_{mp}$) can be restricted, depressing the overall energy production of unshaded modules on the same circuit.

Multi-Channel MPPT Design Flexibility

Modern string inverters address array variability by integrating multiple independent MPPT inputs (frequently 2 to 12 distinct MPPT channels on commercial units). Multi-MPPT tracking provides crucial architectural flexibility:

• Multiple Orientations: Allows strings on an east-facing roof plane and a west-facing roof plane to terminate at the same inverter without cross-string current fighting.
• Unequal String Lengths: Accommodates roof geometries requiring strings of different module counts (e.g., one 10-module string on MPPT 1 and one 13-module string on MPPT 2), provided each string voltage remains within the inverter's rated MPPT voltage window.

---

Module-Level Power Electronics (MLPE)

Module-Level Power Electronics distribute power conversion circuitry directly to each individual solar module. The two primary MLPE implementations are microinverters and DC power optimizers.

Microinverters: Distributed Module-Level AC Conversion

Microinverters mount directly to the module frame or racking beneath each panel, converting the low-voltage DC output of a single module (or a dual-module pair) directly into grid-synchronized AC electricity right at the roof level.

• AC Trunk Architecture: Microinverters connect in parallel along a pre-molded AC trunk cable. Multiple units form an AC branch circuit that lands directly on a dedicated overcurrent protection device (OCPD) in an AC load center.
• Branch Circuit Limitations: Manufacturers specify maximum continuous AC output currents that dictate branch circuit limits. For example, a 20A 240V branch circuit can carry up to 16A of continuous inverter output current, because the overcurrent device must be at least 125% of the inverter output circuit current (2017 NEC 690.8(A)(3) and 690.9(B)); such a branch typically supports 11 to 16 microinverters depending on individual VA ratings.
• Rapid Shutdown Compliance: Because DC wiring is restricted to the short module leads directly beneath the panel (always operating below 60V–80Vdc), microinverter systems comply inherently with NEC 690.12 rapid shutdown boundaries without requiring external DC shutdown transmitters.
• Reliability Trade-Off: Moving semiconductors, capacitors, and microprocessors onto the roof exposes components to extreme thermal cycling and moisture ingress. A hardware failure requires roof access to unclamp the module and replace the unit.

DC Power Optimizers: Distributed Buck/Boost Regulation

DC power optimizers are module-level DC-to-DC converters paired with a specialized central string inverter. Instead of converting DC to AC on the roof, the optimizer performs MPPT at each panel and regulates its DC output.

1. Buck/Boost Voltage Adjustment: The optimizer dynamically steps its output voltage up or down while adjusting current ($P_{\text{in}} \approx P_{\text{out}}$), maintaining an optimal, fixed string DC voltage (e.g., 380Vdc to 400Vdc single-phase, or 480Vdc three-phase) delivered to the inverter.
2. Rapid Shutdown Integration: When the inverter is powered off or grid AC is removed, the optimizers automatically switch to a safe 1V output per module. A string of 18 modules drops to 18Vdc, satisfying the NEC 690.12 limits of $\le 30\text{V}$ outside the array boundary and $\le 80\text{V}$ inside it within 30 seconds.
3. Telemetry and Granularity: Optimizers communicate with the inverter over power-line carrier (PLC) communication, transmitting module-level current, voltage, and power metrics to monitoring platforms.

---

Galvanic Isolation vs. Transformerless (Non-Isolated) Inverters

Historically, solar inverters incorporated internal 60 Hz iron-core or high-frequency toroidal transformers to provide galvanic isolation—a physical magnetic barrier preventing direct electrical conduction between the DC array wiring and the AC utility grid.

Transformerless (Non-Isolated) Inverters

Contemporary installations almost universally utilize non-isolated, transformerless inverters (designated as "TL"). By substituting solid-state H-bridge silicon switching topologies and advanced filtering networks for heavy copper transformers, transformerless inverters achieve significant operational advantages:

• Higher Peak Efficiency: Elimination of transformer core and winding losses raises peak efficiency to 98.0%–99.2%.
• Reduced Weight and Footprint: A 10 kW transformerless residential inverter weighs under 35 lbs (16 kg), enabling single-technician wall mounting, compared to legacy transformer-based units exceeding 120 lbs.

NEC Grounding and Safety Requirements for Transformerless Arrays

Because transformerless inverters share a direct conductive path between the DC array and AC grid, bonding either the positive or negative DC conductor to earth creates a catastrophic short circuit through the internal switching bridge during AC polarity reversals. Therefore, transformerless arrays operate with neither dc conductor solidly grounded. The 2017 NEC 690.41(A) lists this as a PV array "not isolated from the grounded inverter output circuit" (a functionally grounded configuration); the 2014 NEC and field slang call it an ungrounded or floating array:

• Floating Potentials: Both positive and negative conductors operate at a potential relative to ground (e.g., $+200\text{V}$ and $-200\text{V}$ relative to earth on a 400V string).
• Disconnecting Both Conductors: Because neither current-carrying conductor is solidly grounded, equipment disconnecting means must simultaneously open both the positive and the negative conductor (2017 NEC 690.15(D)). Overcurrent protection is different: 690.9(C) permits a single overcurrent device per circuit, provided all such devices in the PV system are placed in the same polarity.
• Insulation Resistance Monitoring: Prior to grid synchronization, the inverter performs an automated insulation resistance check by applying an internal test potential to measure leakage resistance to ground. If insulation resistance falls below the inverter's threshold (set by its listing; IEC 62109-2 ties the minimum to the maximum array voltage, so thresholds from tens of kilohms up to about $1\text{ M}\Omega$ are common), the inverter locks out to prevent hazardous fault energization.
• Residual Current Detection (RCD): An internal differential current sensor continuously monitors the vector sum of currents in the positive and negative conductors ($I_+ + I_-$). Any differential exceeding milliamperes indicates leakage to ground, triggering sub-second disconnect.

---

Inverter Architecture Comparison Matrix

The following matrix summarizes the design trade-offs across all four primary PV conversion topologies: