What a solar inverter actually does
The name "inverter" comes from its core job: inverting DC (direct current) from your panels into AC (alternating current) that your house and the grid use. But modern solar inverters are sophisticated power electronics devices — they do far more than simple voltage conversion.
Solar panels are photovoltaic (PV) devices — they convert light into direct current. A single panel typically produces 30–45 V and 8–12 A under full sun (240–500 W depending on panel size). Panels are wired in series to form strings: a string of 10 × 40 V panels produces 400 V DC at the inverter input.
The inverter's DC input range is designed to accept the full span of string voltage your panels can produce — from a cold, bright winter morning (highest voltage, no load) down to a hot, hazy summer afternoon (lower voltage, high current). A mismatch between panel string voltage and inverter input range is a common commissioning error.
The inverter outputs single-phase 230 V AC at 50 Hz — matching the UK grid and all standard household appliances. In larger commercial or three-phase domestic installations, the inverter outputs three-phase 400 V AC. The output connects into the household consumer unit, feeding power to circuits and allowing surplus to flow out to the grid.
In a hybrid system, the inverter is also the system controller. It reads data from the battery BMS, the CT clamp on the grid connection, and the solar input — then decides in real time where power flows: charge the battery, power the house, export to grid, or draw from grid. Every optimisation decision (charge from solar, avoid peak rate import, enforce export limit) runs through the inverter's control processor.
MPPT: how the inverter extracts maximum power from your panels
Solar panels are not fixed-output devices. The power they produce depends on the operating voltage — and the relationship between voltage and current changes constantly with sunlight intensity, temperature, and shading. MPPT is the algorithm that navigates this curve to find the peak.
Every solar panel has a characteristic current-voltage (I-V) curve. At one extreme (short circuit), current is maximum but voltage is zero — power is zero. At the other extreme (open circuit), voltage is maximum but current is zero — power is zero again. Somewhere in between is a peak: the Maximum Power Point (MPP), where voltage × current is at its highest.
The MPP shifts every time cloud cover, temperature, or shading changes. MPPT continuously repositions the operating point to stay at the peak.
The most common MPPT algorithm is Perturb and Observe (P&O): the inverter slightly increases the operating voltage, measures whether power rose or fell, then steps in whichever direction increased power. This repeats continuously — typically 10–100 times per second — keeping the inverter locked onto the peak even as clouds pass.
A single MPPT input handles one string (or parallel strings at the same voltage). If all your panels face the same direction and have no shading, one MPPT is fine. If panels are split across two roof orientations (south and east, for example), or if one section is shaded by a chimney while another is clear, you need two independent MPPT inputs — one for each string.
Panel voltage rises as temperature falls. On a cold, clear January morning, panel string voltage may be 15–20% higher than on a hot August afternoon. The inverter's MPPT voltage range must accommodate this spread. If the string voltage exceeds the inverter's maximum MPPT input voltage on a cold day, the inverter will operate below maximum power — or may not start at all. This is why string sizing calculations must account for cold-day peak voltage, not just standard test conditions.
How the inverter synchronises with the grid
Before an inverter can supply any power, its output must exactly match the grid: same voltage, same frequency, same phase. This synchronisation process happens automatically at startup, and the inverter maintains lock continuously during operation.
The inverter uses a phase-locked loop (PLL) circuit to track the grid's AC waveform in real time. The PLL samples the grid voltage thousands of times per second and adjusts the inverter's internal oscillator to produce an output that is phase-aligned — peaks and troughs match exactly. If the inverter's output were even slightly out of phase, a reactive current loop would develop between the inverter and grid, causing losses, tripping protection relays, and potentially damaging equipment.
At startup on a sunny morning, the inverter checks that the grid is present and stable before closing its output relay. This is why you see a 30–60 second delay between sunrise and the inverter beginning to produce power.
UK inverters are set to disconnect if grid voltage or frequency falls outside permitted bands. The standard settings under Engineering Recommendation G98/G99 are:
On networks with many solar installations, grid voltage can rise during peak generation periods — a phenomenon called voltage rise. If your inverter repeatedly trips on high-voltage grid faults during sunny periods, this is the most likely cause. See the grid overvoltage problem page for diagnosis steps.
Advanced inverters can also control reactive power — the component of AC power that oscillates between source and load without doing useful work. Network operators can require inverters to operate at a specific power factor (e.g. 0.95 lagging) to help stabilise local grid voltage. This is more common on larger G99-connected systems. For most domestic installations, the inverter operates at unity power factor (1.0) — exporting only real power.
Anti-islanding: why your solar stops in a power cut
When the grid loses power, your solar system also stops — even if the sun is shining and your panels could generate 5 kW. This feels counterintuitive but is a critical safety mechanism. Understanding it explains why a basic solar installation cannot provide backup power, and what is actually required to get backup capability.
When the DNO isolates a section of grid for maintenance or fault repair, engineers assume the cables are dead. If your inverter continues to generate power and export to the grid after isolation, it will energise cables that engineers believe are off — creating a lethal hazard for anyone working on those cables.
This scenario — where a generator continues to power a section of isolated grid — is called "islanding." Anti-islanding protection is required by law under Engineering Recommendation G98/G99 and must be tested during commissioning.
The inverter continuously monitors grid voltage and frequency. If the grid is lost, voltage collapses to zero within milliseconds. The inverter detects this, opens its internal relay within 200 ms (G98 requirement), and stops producing power. Before reconnecting, it waits a minimum hold-off period (typically 20–30 seconds) to confirm grid stability, then re-synchronises.
More subtle islanding scenarios (where household load perfectly matches generation and voltage appears stable) are detected by additional algorithms such as frequency drift detection (Frequency Shift), impedance measurement, or sanity-checking the grid waveform shape.
Hybrid inverters with EPS (Emergency Power Supply) mode solve this by physically disconnecting the household from the grid when it is lost — creating a true island, isolated from the DNO network. Within this isolated section, the inverter can safely generate power from solar and battery without any risk to grid engineers. EPS mode powers a dedicated essential loads circuit (pre-wired at installation) and is not affected by anti-islanding requirements because it is fully isolated from the public grid. See our dedicated guide on how backup power works for the full technical detail on EPS, UPS, and island mode.
String, hybrid, and microinverters: what is the difference?
The type of inverter determines what your system can do, how it handles shading, and whether backup power is possible. Most UK domestic installations use one of three architectures.
One central inverter for all panels. Panels wired in series strings. Simple, cost-effective, and the most proven technology.
String inverter plus battery management in one unit. DC battery connection. Full system control in one device.
One small inverter per panel. DC-to-AC conversion at panel level. All panels operate independently.
If you already have a string inverter and want to add battery storage, one option is AC coupling. An AC-coupled battery has its own separate inverter (sometimes called a hybrid inverter or battery inverter), which connects to the household AC circuits rather than to the DC side of the existing solar inverter.
AC coupling is slightly less efficient than DC coupling (the solar energy converts DC→AC→DC→AC before being used), but it can be added to virtually any existing solar system without replacing the solar inverter. Brands like GivEnergy and Tesla Powerwall frequently operate in AC-coupled configuration on retrofit projects. See the battery storage guide for a full comparison of AC and DC coupling.
Inverter operating modes: what your system is doing and when
A hybrid inverter switches between several modes depending on solar availability, battery state, house load, and the time of day. Understanding these modes explains why your system sometimes imports from grid even when the sun is shining, or why the battery charges overnight.
Solar generation powers the house directly. Surplus goes to battery. If battery is full and generation exceeds consumption, surplus exports to grid (up to any export limit). This is the normal daytime operating mode — the goal is to use as much self-generated power as possible before touching the grid.
When solar exceeds house load and battery is full, the inverter exports surplus to the grid. The rate is capped at any export limit set by the DNO — typically 3.68 kW for single-phase G98 connections. Exported units may earn Smart Export Guarantee payments from your energy supplier, measured by your smart meter.
Hybrid inverters can be configured to charge the battery from the grid during off-peak rate windows — typically between 00:30 and 05:00 on Octopus Go, Intelligent Octopus, or Economy 7. The inverter draws power from the grid at controlled rate to fill the battery before the peak-rate morning period. If this mode is not working, the most common causes are incorrect timer settings, a firmware reset, or a reversed CT clamp preventing the inverter from seeing grid power correctly.
When export limit is set to 0 W, the inverter continuously curtails generation to prevent any power reaching the grid. All solar must be consumed on-site or stored in battery. When battery is full and house load is low, generation is actively reduced by the MPPT control loop. A system in zero export mode with a full battery and low house load will appear to generate significantly less than its panel rating — this is not a fault.
When grid power is lost, a hybrid inverter with EPS mode disconnects from the grid and switches to island mode — powering an essential loads circuit from battery and/or solar. EPS response time varies by brand: GivEnergy hybrid inverters typically switch within 10–20 ms; some older models take 50–100 ms. EPS only powers the dedicated essential loads output — not the whole house, unless specifically wired that way at installation. See the no backup during power cut problem page if EPS is not activating.
Related guides and problem pages
LFP chemistry, BMS, SoC management, AC vs DC coupling, and round-trip efficiency.
How CT clamps give the inverter real-time grid data, and what happens when they are reversed.
Why DNOs set export limits, how the inverter enforces them, and whether yours can be changed.
Diagnosis for inverters that show no display, no output, or fail to start at sunrise.
What a red fault light means, which codes to check first, and when a remote diagnostic is needed.
Zero generation despite sunshine — inverter faults, DC wiring issues, and isolation failures.