Drawing No. EH–CA–010 // Engineering Case Study
The 2003 Italy blackout was one of Europe’s most significant power-system disturbances. This interactive case study reconstructs how Swiss transmission-line flashovers, failed reclosing, delayed corrective action and cross-border separation led to frequency collapse across the Italian power grid. All 20 Italian administrative regions are displayed to show the geographic extent of the outage and restoration.
Select a stage or run the event.
A static summary of the same nine stages used by the interactive regional replay.
Italy imports approximately 6.65 GW across the northern border while the system remains N-1 secure. Alpine transmission corridors are heavily used.
The first 380 kV Swiss line trips after a flashover to vegetation. Automatic and manual reclosing attempts fail.
The parallel Swiss corridor becomes severely overloaded. Conductor heating and sag progressively reduce vegetation clearance.
A cross-border request reduces imports, but the action is not sufficient to return the remaining corridor to a secure state before its thermal window expires.
The Sils–Soazza / San Bernardino line flashes over and trips. Power swings immediately stress the remaining French, Austrian and Slovenian connections.
Almost simultaneous trips across the northern interface cut Italy from the wider synchronous grid. More than 6 GW of imported support disappears within seconds.
Frequency, voltage and angle stability deteriorate together. Generation and distributed resources disconnect while emergency controls cannot recover the balance.
The mainland Italian system and Sicily are de-energised. Sardinia remains shown separately because it was not part of the principal mainland-and-Sicily balance analysed by UCTE.
Operators rebuild the 380 kV backbone from neighbouring systems and black-start resources. Regional islands expand, synchronise and reconnect demand.
Interconnected, islanded and de-energised states.
The sequence in five linked mechanisms.
Official milestones with educational interpolation.
| Stage | Domestic generation (GW) | Northern imports (GW) |
|---|---|---|
| 1. Before 03:01 | 20.49 | 6.65 |
| 2. 03:01:42 | 20.50 | 6.55 |
| 3. 03:02–03:11 | 20.51 | 6.48 |
| 4. 03:11–03:21 | 20.58 | 6.30 |
| 5. 03:25:21 | 20.59 | 4.50 |
| 6. 03:25:34 | 20.10 | 0.05 |
| 7. 03:25:34–03:28 | 9.50 | 0.00 |
| 8. After approximately 03:28 | 0.00 | 0.00 |
| 9. 03:46–21:40 | 18.80 | 4.80 |
| Stage | Lukmanier | San Bernardino | Remaining Alpine interface |
|---|---|---|---|
| 1. Before 03:01 | 86 | 74 | 68 |
| 2. 03:01:42 | 0 | 118 | 78 |
| 3. 03:02–03:11 | 0 | 125 | 83 |
| 4. 03:11–03:21 | 0 | 116 | 80 |
| 5. 03:25:21 | 0 | 0 | 132 |
| 6. 03:25:34 | 0 | 0 | 0 |
| 7. 03:25:34–03:28 | 0 | 0 | 0 |
| 8. After approximately 03:28 | 0 | 0 | 0 |
| 9. 03:46–21:40 | 55 | 48 | 45 |
| Stage | Displayed frequency (Hz) | Replay state |
|---|---|---|
| 1. Before 03:01 | 50.00 | Near nominal |
| 2. 03:01:42 | 49.99 | Near nominal |
| 3. 03:02–03:11 | 49.98 | Near nominal |
| 4. 03:11–03:21 | 49.97 | Near nominal |
| 5. 03:25:21 | 49.80 | Near nominal |
| 6. 03:25:34 | 48.80 | Severe decline |
| 7. 03:25:34–03:28 | 47.70 | Severe decline |
| 8. After approximately 03:28 | 0.00 | Blackout indicator / de-energised |
| 9. 03:46–21:40 | 50.00 | Near nominal |
| Stage | Normal | Stressed | Critical | Offline | Restoring |
|---|---|---|---|---|---|
| 1. Before 03:01 | 20 | 0 | 0 | 0 | 0 |
| 2. 03:01:42 | 16 | 4 | 0 | 0 | 0 |
| 3. 03:02–03:11 | 12 | 5 | 3 | 0 | 0 |
| 4. 03:11–03:21 | 10 | 6 | 4 | 0 | 0 |
| 5. 03:25:21 | 6 | 6 | 8 | 0 | 0 |
| 6. 03:25:34 | 1 | 5 | 14 | 0 | 0 |
| 7. 03:25:34–03:28 | 1 | 5 | 6 | 8 | 0 |
| 8. After approximately 03:28 | 1 | 0 | 0 | 19 | 0 |
| 9. 03:46–21:40 | 1 | 0 | 0 | 0 | 19 |
The first tree flashover initiated the event but did not alone determine the outcome. The escalation depended on unsuccessful reclosing, sustained overload of the parallel corridor, delayed corrective action, a second flashover, cascading border trips and unstable island operation.
Official reports document the national transmission sequence and broad restoration milestones, not a second-by-second administrative-region chronology. This tool therefore uses Italy's 20 regions to make the geographic extent understandable while clearly labelling the regional progression as educational interpolation.
Sardinia was electrically separate from the principal mainland-and-Sicily system analysed in the UCTE capacity balance. The replay therefore keeps Sardinia normal during the principal blackout sequence rather than implying an unsupported identical progression.
Four concise conclusions.
A simplified barrier review.
Open each topic for the underlying principles.
Higher current heats the conductor and increases sag. Reduced clearance to vegetation raises flashover risk.
The phase-angle difference across the open line exceeded the closing criterion, making reconnection unsafe.
The island had to replace lost imports immediately while maintaining frequency, voltage and synchronism. It could not reach a stable operating point before additional generation and network elements disconnected.
Four transferable lessons.
Short answers to common questions.
No. The first flashover initiated a sequence that also required unsuccessful reclosing, sustained overload, delayed action, a second flashover, border trips and unstable island operation.
No. Region statuses are an educational reconstruction. The national chronology and key technical milestones follow the official investigation.
Sardinia was electrically separate from the principal mainland-and-Sicily system considered in the official capacity balance.
The event was primarily a system-security, coordination and stability failure rather than a simple single-relay malfunction.
Compare this event with other EngineerHub replays.
Official references and educational boundaries.
Primary investigation: UCTE Final Report on the 28 September 2003 Blackout in Italy.
Italian regional geometry: SimpleMaps Italy SVG map resources, used under the applicable free-map licence with attribution.
Neighbouring-country geometry: Natural Earth Admin 0 country boundaries, public-domain geographic data projected to the map frame.
Transmission-corridor geometry, intermediate frequency values, regional propagation and restoration percentages are educational approximations. Administrative and country contours use published geographic datasets. Official event times, line names, import and generation balances, the overload window, separation time and emergency endpoint follow the investigation.
Include the inputs and assumptions used.
EngineerHub.org