Comment: Lessons from the Iberian grid outage

There have been several articles speculating on the cause of the major power outage in Spain a couple of weeks ago, as we would expect. Much attention has been given to the impact of large volumes of renewables connected to the grid, particularly solar photovoltaics (PV) panels, and whether these have contributed to the outage. The European Network of Transmission System Operators for Electricity (ENTSO-e) announced last week that it will investigate the incident and issue two reports. The first will reconstruct the events of the outage and determine its cause; the second will propose recommendations to help prevent similar incidents. Both reports are likely to attract considerable interest, particularly amongst other grid operators and planners.

MORE FROM OPINION

Although major grid events are fortunately very few, removing the risk of all such events is not possible. The risk can be considerably reduced through engineering design and operation, but there will always be some risk of failure. Attention needs to be given to how such failures can be managed to minimise their effect and to enable the grid to return to normal operation as safely and as quickly as possible.

The fundamental operation of power systems relies on the balance of generation with demand. Mismatches cause the grid frequency to change, and generation will need to compensate accordingly. Initially this is through an exchange of energy from the inertia of the rotating plant, but it is then followed by generator governor action to vary the primary power input to adjust for the change in demand.

For much larger mismatches (such as loss of generation), more radical action is needed. When this occurs, there will be a rapid fall in frequency, perhaps within seconds, and although grid inertia will help to cushion the fall, an increase in power input is needed from governor action. However, this takes much longer to operate and is unlikely to be very helpful in such circumstances. Without any further action the outcome will be a complete collapse of the grid with possible serious damage to plant and equipment. The only alternative is to disconnect demand. This is accomplished automatically by under-frequency load shedding protection relays, which detect the fall in frequency and then disconnect load to arrest the fall to save as much of the grid as possible. They are normally located at transmission substations, and the protection scheme will have multiple stages which trigger demand disconnection at different substations and declining frequency levels.

The design of the underfrequency load shedding (UFLS) scheme needs to consider the characteristics of the grid system. Over the last 30 years there have been substantial changes in the British grid and elsewhere, with less rotating power plant and more inverter connected plant including interconnectors, solar PV, and wind. The reduction in rotating power plant means less mechanical grid inertia. However, inverter connected plant can still provide electronic grid inertia, often referred to as synthetic inertia. This can provide fast frequency response which mimics the stored energy characteristic of rotating plant by using electronic sensors to increase the power output of the inverter connected plant. Hence although there has been some speculation that renewables may have adversely contributed to the grid outage due to their lack of inertia, this may not be so, and will need to be investigated by ENTSO-e along with the design of the UFLS (or any other) scheme.

Restoring the Spanish grid would have been a major challenge. Initial estimates indicated that it would takes several days to complete, but this was accomplished within 23 hours. This is remarkable, considering the extent of the outage. The challenge for the grid operator is to increment power production cautiously with demand. This requires careful coordination with other network operators such as the regional distribution networks, interconnectors, as well as power plant operators, with safety always paramount.

There is no doubt that there will be important lessons to be learned from this event. However, as much as the cause is important, its subsequent management both in terms of minimising impact as well as grid restoration is also important and hopefully will be thoroughly investigated. Consequently, we should await the outcome of ENTSO-e investigation and avoid speculation. Until then, grid operators would be wise to ensure they are well positioned if such an event were to occur to them.

Dr Robert Sansom is a member of the IET Sustainability and Net Zero Policy Centre