Grid-Connected Converters
A grid-tied inverter controls current, not voltage — and anti-islanding, ride-through and synthetic inertia are all consequences of that one fact.
Skip to the animationA grid-tied inverter cannot set the voltage the grid has already fixed, so it controls its current instead — and anti-islanding, fault ride-through and synthetic inertia are all consequences of that role, as is the shift to grid-forming control.
Current source, not voltage source
An off-grid inverter defines the voltage and frequency because nothing else does. A grid-tied inverter faces a system that has already decided both, so it injects a current in phase with a voltage it cannot influence.
That requires knowing the grid's phase continuously. A phase-locked loop tracks it, and its accuracy directly sets the power factor — a one-degree phase error injects 1.7% reactive current.
Anti-islanding
If the grid disconnects while local generation matches local load, nothing about the voltage or frequency changes — and the converter would keep energising a section that linesmen believe is dead.
Active anti-islanding injects a small deliberate disturbance and watches whether the grid resists it. If it does, the grid is present; if the disturbance runs away, it is not. Detecting an absence is genuinely harder than detecting a presence.
The requirement that reversed
When solar and wind were a rounding error, disconnecting on any disturbance was safest. Once they became a large fraction of supply, every converter tripping on a voltage dip meant losing gigawatts at the worst possible moment.
Grid codes now require fault ride-through and reactive current injection during the dip — a complete reversal of the earlier rule, driven entirely by the technology's own success.
Inertia
A synchronous generator's rotating mass resists frequency change automatically — a service the grid received free for a century. An inverter has none, so a high-inverter system sees a steeper rate of change of frequency after a generation loss, leaving governors less time to act.
Synthetic inertia and fast frequency response are now specified explicitly in grid codes. It is a property nobody had to write down for a hundred years, because it came with the machine.
Grid-forming control
| Grid-following | Grid-forming | |
|---|---|---|
| Behaves as | Current source | Voltage source with virtual inertia |
| Needs an existing grid | Yes | No |
| Can black-start | No | Yes |
| Sets frequency | No — follows | Yes, with droop |
Grid-forming is largely a control-software distinction on the same hardware, and it is what makes a fully inverter-based grid possible — something has to define the reference.
What converters do that machines cannot
- STATCOM — reactive power in milliseconds, and independent of terminal voltage. A capacitor bank's output falls as V², so it supplies least exactly when a sagging voltage needs it most.
- Active filter — injects the inverse of the harmonic current, cancelling it at source.
- HVDC — links grids that are not synchronised, with exact control of the flow. Voltage-source HVDC can also start into a dead network and control P and Q independently.
HVDC wins over long distances, in submarine cable where AC charging current rules out anything past about 50 km, and for asynchronous interconnection. Offshore wind made submarine HVDC routine.
The numbers you will be asked for
- Injected power
P = V_grid · I_inverter · cos φ
- Phase error cost
1° of error → ≈ 1.7% reactive current
- Rate of change of frequency
df/dt = −ΔP / (2H · S_base)
- Droop, grid-forming
f = f₀ − m·P · V = V₀ − n·Q
- HVDC break-even
≈ 600 km overhead, ≈ 50 km submarine
Watch it work
Check yourself
question 1 / 4
One question at a time. Pick an answer to see why it is right or wrong, then move on — there is no score to keep and nothing is saved.