Reliability and resilience of the electricity system with high levels of renewables

The electricity supply system in Britain, which has provided reliable energy supplies for decades, is undergoing a radical transformation in the delivery of a zero carbon supply system. The headline changes are from a system based on large thermal power stations to one with a high penetration of intermittent renewables such as wind and solar. This transformation is changing the supply system characteristics from being highly controllable and highly stable with low technical risk, to a system with a large proportion of highly intermittent generation, less stable and with enhanced levels of technical risk.

Reducing carbon emissions from energy systems requires adoption of clean electricity generation and substitution of carbon fuel combustion with electricity in heating and transport. Renewable electricity from wind and solar currently play a significant and increasing part but they introduce increased technical risks which need to be managed. These increased risks come about from the major change in grid system characteristics which required radical changes in the way the network is planned and operated to ensure that the security and quality of supply standards are maintained. The increased exposure to extreme weather events adds to this changing background, resulting in an elevated risk of a major system failure.

The traditional generation capacity surplus [see glossary 1] and operating margins [see glossary 2] are significantly depleted although this has been compensated by the introduction of European interconnectors which currently supply about 10% of the GB electrical power.

The ability to recover from a grid failure or partial grid failure (a major feature of the resilience of the system) is under question. There is an urgent need for long term sustainable energy planning with less reliance on short term market solutions.

Reasons for the electricity supply to be interrupted

The causes of major system disturbances that may lead to system blackouts may be summarised as follows:

• Electricity supply is unable to meet high demand, resulting in progressive automatic disconnection of demand across the Grid System.
• Severe weather events that cause the network to be depleted beyond the level for which it is designed or intended to operate.
• Insufficient reactive power leading to voltage collapse.
• Lack of system inertia impairing the ability of the System Operator to match the system generation output to changing system demand and the ability of the network to recover automatically from transient faults.  
• Other exceptional events, e.g. solar storms, cyber security breaches, unforeseen technical anomalies, etc

Major supply interruptions are thankfully rare in GB. When they occur, they are often the result of a combination of the above issues. However, the increase in severe weather events from climate change and the significant change in characteristic behaviour of the grid system resulting from the large-scale integration of renewable generation, give rise to an overall increased probability of a major system failure. Given the societal reliance on the electricity supply and the profound impact of a widespread blackout, it is imperative that recovery from such an event should be fast as practically possible.

Issues Contributing to Major Network Failure

Background

The Grid System Operator (SO) and Transmission Owners have Licence obligations to plan and operate their networks to be compliant with the National Electricity Transmission System Security and Quality of Supply Standard (NETS). This provides a prescriptive minimum level of network capacity and defines the levels of network security under normal operation and under system maintenance conditions. There are no equivalent standards for the future planning or procurement of generation capacity, this is left to the electricity market.

The change in characteristics of the electricity supply system with large scale integration of renewable generation has required continuous changes to the Grid and Distribution Codes (the technical codes that govern the operation and planning of the Transmission and Distribution networks), to ensure that new plant technologies do not adversely affect the security and quality of supply.

Security of generation supply

The success of renewable generation currently results in approximately 25% of UK annual energy produced from wind, with an installed capacity of approaching 20GW. These figures are predicted to rise significant over the next decade with an offshore target of between 31 and 47GW of wind by 2030 (see NGC Future Energy Scenarios document) ]. The highest demands for electricity sometimes occur when there are very small quantities of renewable generation available. This may happen when a large high-pressure cold weather system sits over Europe, sometimes for a number of days, resulting in very little wind generation and reduced Interconnector import availability. Hence at times of reduced renewable output, generation security requires gas generation to be available and places a high reliance on gas imports from eastern Europe, Russia and Norway, with high price volatility.

The closure of coal fired power stations has resulted in severely depleted plant margins, [see glossary 2] and the closure of the nuclear fleet in coming years will further exacerbate this issue. An unintended consequence of the introduction of renewable generation is erosion of the economic case (loss of volume generation opportunity), for developers to build new conventional power stations that may back-up supplies when renewables are not available. Regulatory intervention to provide a ‘Capacity Market’, i.e.to pay generators to be ‘available’ without necessarily generating, has been ineffective in stimulating new generation projects.

To address the shortage of generation capacity, Ofgem have also promoted a number of mitigation initiatives including Demand Side Management (contractual demand reductions), increased HVDC Interconnectors [glossary 3] to Europe and Energy Storage initiatives (batteries).

The NGC Future Energy Scenarios (linked above) predict that the peak demand for electricity will increase from about 58GW today to around 69GW in 2030, rising to between 92 and 113GW by 2050, under their four future energy scenarios. Electricity demand increase to supply electric vehicle charging and electric heating (including the replacement of gas central heating boilers with heat pumps), is a key deliverable for the government decarbonising policy. It is however highly questionable where this additional generation capacity may be realised without a radical change to the way future generation needs are planned, procured and delivered. It is appropriate to point out that the 20GW of gas fossil generation that plays a vital role in providing generation security today, is targeted be removed by 2050.

The future generation requirements need careful planning to ensure there is a coordinated, sustainable, efficient and secure plant mix including renewables, large and small nuclear and large-scale energy storage etc.

Reactive Power

All parts of the network require a satisfactory balance of reactive power generation, whether this is produced by generators, static reactive power compensation devices, or the power system overhead lines and cables themselves. A shortage of reactive power gives rise to low voltage and a surplus gives rise to high voltage. This is a non-linear characteristic and imbalances can give rise to very high voltages or voltage collapse, in a short period of time. Reactive power and hence system voltage, is controlled by a mixture of automatic and manual control procedures.

System inertia and synchronous generation

The system inertia, or flywheel effect, is the energy stored in rotating machines connected to the grid system, that helps to keep the system running during short periods of imbalance between generation and demand and also helps the system ride through transient network faults.

With the reduction in the number of heavy fossil fuel generators, the inertia (i.e. the rotational mass) of the grid system has significantly reduced. Wind turbines (and solar generators) make no contribution to the inertia of the grid system because they are electrically de-coupled through their DC/AC power inverter control systems. This affects the grid system performance in a number of ways.

With a lower inertia system, the rate of change of frequency is greater when there is an imbalance between generation and demand, so generators with governor response [see glossary 4] need to respond more quickly to regulate frequency within operational limits. This is not a problem for gas turbine generators, or where HVDC interconnectors are providing frequency response, but the output from wind farms (or solar panels) cannot be increased, only decreased under emergency conditions. Furthermore, the older designs of nuclear generators are not capable of providing a significant response level because their rate of change of output is too slow.

When the output from wind and solar generators is high and demand is low, the proportion of generation capacity with sufficient response will make it more difficult for the SO to contain the frequency within operational limits. This would be particularly difficult if the level of generation loss approaches or exceeds the infrequent generator loss limit [see glossary 5] and may result in low frequency demand shedding. This was a factor with the load shedding event in August 2019 when a lightning strike caused the Hornsea windfarm and Little Barford gas turbines to trip. Notably there was also a loss of about 500MW of embedded generation [see glossary 6] that tripped because of the low frequency deviation, highlighting the vulnerability of embedded generation providing additional operational uncertainty for the SO.

When a fault occurs on the transmission system, the transient response of generators must not cause them to lose synchronism [see glossary 7]. With significantly reduced system inertia, the rate of change of frequency (or acceleration) during the fault will be higher and stability limits (assessed by computer simulation) may impose significant power flow constraints on the operation. This issue has required the SO to implement an additional ancillary services market for System Inertia and ScottishPower is seeking to install Synchronous Compensators [see glossary 8] on their transmission system, to maintain the Anglo-Scottish power transfer capability. The need for such devices poses a legitimate question as to who should pay for restoring system inertia caused by the closure of coal and nuclear stations. Should this fall on the operators of the remaining rotating machines, should it be the operators of the new, low-inertia generation or should it be their customers?

Network faults

In simplified terms, credible faults, e.g. caused by lightning strikes to overhead lines, are assumed to affect both circuits on a tower line, or two separate circuits on two tower lines. Under more frequent extreme weather events from climate change, this conservative level of system depletion may be exceeded several times and may increasingly result in significant loss of supply.

Under post fault outage conditions the SO is required to re-secure the network by re-scheduling generation, e.g. to reduce the flow over stressed parts of the network. However, as the amount of controllable generation has been significantly reduced, the opportunities of the SO to re-secure the network post fault have also reduced. The output from renewable generation cannot be increased beyond its normal output, only reduced under emergency conditions.

System Design Uncertainty

Windfarm generators and a significant proportion of the equipment installed on the transmission and distribution systems, have complex power electronic control systems. The modelling required to securely integrate these diverse control systems is technically challenging and may not always represent the behaviour of the plant in the real world. It is not unusual for unexpected adverse control system interactions between equipments on the network to take place, particularly under unusual or extreme operating conditions. This could lead to significant disruption to electricity supplies.

Resilience

A measure of the resilience of a system is its ability and time to return to normal after a breakdown (which may be local, regional or national in its field of disruption).

The probability of a complete or partial failure of the British grid system is very small but, for the reasons given above, it is increasing. The societal, financial and political consequences of a major blackout are substantial, so the SO has a Licence obligation to put in place a viable Black Start procedure. The procedure requires all generators with a Black Start capability to start independently and supply island loads [see glossary 9], before the SO reconnects the islands to restore the complete system. The process of synchronising the islands may take considerable time.

A significant part of the generation portfolio with Black Start capability used to reside in the coal fired fleet but these facilities are now all being decommissioned. Since Longannet power station closed, the ability for supplies to be restored to the central belt of Scotland following a Black Start has been severely compromised. The delay to re-establish supplies from England could now take up to 5 days or more.

Recognising this unacceptable recovery time, a funded innovation project, ‘Distributed ReStart`, promoted by ScottishPower Energy Networks, with partners National Grid SO and consultants TNEI, will investigate a bottom-up approach to Black Start recovery. The project seeks to use the large volume of distributed generation now connected to the grid, to play a role in Black Start recovery. However, there are formidable technical, organisational and procurement challenges to overcome. In addition, the majority of this distributed connected generation is renewable energy sourced and might not be available without wind or sunlight at the time of need.

Conclusion

The radical changes in the electricity network in the transition to net zero carbon emissions unfortunately lead to great cost and technical risk. Our reliance on electricity in the modern world is not appreciated until there is a loss of supply for a significant period of time. The partial grid failure in August 2019 provided a stark reminder of the effect of an electricity blackout, although this event was minor when compared to other well documented, widespread blackouts that have been experienced in many countries in the developed world.

The current measures taken to manage the worryingly low plant margin are, in the main, short term and significant. Predictable firm generation capacity is required if the increasing demand from electric vehicles and electric heating is to be met. This requirement needs to reflect conditions where there is minimal availability of European wide renewable resource, i.e. reduced interconnection availability, to ensure continuity of supply in the UK.

It is imperative that the grid system issues caused by reducing inertia and increasing intermittency are sustainably managed and resolved. The current industry regulatory framework promotes uncoordinated renewable generation without adequate consideration of the operational risk and costs. The System Operator is required to respond reactively with expensive, inefficient ancillary services, when a proactive holistic planning approach to renewables integration would be more effective and efficient in the long term.

The role of the industry Regulator should be changed, or a new body created, to address the optimum long-term energy needs in terms of plant mix, including diversity, controllability, sustainability and technical compliance etc.

There needs to be increased political awareness of the high impact of a major blackout and the inadequate measures currently available to restore the network to all parts of Great Britain in acceptable timescales, even though this is a low probability event. There needs to be a civil contingency plan to support public services and manage societal breakdown when basic services, e.g. water, sewage, food, fuel and communications are lost for a prolonged period of time.

If the innovative distributed generation approach cited above is to provide a significant benefit, there will need to be massive investment in control communications infrastructure, to enable the connected demand to be managed during a Black Start procedure.

The adequacy of the Security and Quality of Supply Standards (Operation and Planning) should be reviewed in the light of the increased likelihood of extreme weather events and their impact on the network.

Glossary

1. Generation capacity surplus: all the generation available to run, but whose operating costs are such that they are never called to operate in the energy market. Traditionally this would be over 35% more than system demand but in recent years most of this surplus has gone.
2. Operating plant margins: the difference between available generation and the system demand.Traditionally there would be 15% more generation despatched to cover for plant breakdown or shortfalls and demand forecasting errors etc. Plant margins now in the order of 5% and require sustained imports from European interconnectors and demand reduction contracts.
3. HVDC interconnector: A High Voltage Direct Current connection that uses AC to DC converter stations at each end of the line. Enables very long high-capacity cables to be used, not possible with AC.
4. Governor: a control system regulating the output of a generator in response to grid frequency. For the UK the mains frequency is required to be maintained between 49.5Hz and 50.5Hz (50 cycles per second ±1%)
5. Infrequent generator loss limit: The system is operated to withstand the loss of this amount of generation without infringing statutory frequency limits. The limit is currently 1800MW.
6. Embedded generation: generation, usually less than 100MW, connected to the distribution system at 132, 33 and 11kV
7. Synchronous system: System wide generators that run at the same frequency (50 cycles per second) tied together in synchronism. When a synchronous generator loses synchronism, it is automatically disconnected to preserve the security of the network.
8. Synchronous compensator: Similar to a generator but with no turbine drive to produce active power. It provides a contribution to system inertia in transient fault timescales and provides system voltage control.
9. Islands: sub-sections of the National Grid which contain generators and consumers and can operate in isolation from the rest of the grid on a temporary emergency basis

Author: This article was written by Colin Bayfield, MSc, CEng, FIET.  Retired Industry professional.