Liberalized markets bring benefits for consumers and market players alike, but by their nature are complex and challenging environments for network and system operators. What steps can be taken to ensure reliability?
Electricity is arguably the most important utility service provided today. Without it factories, offices and business in general could not function and life at home as we have come to expect it would be impossible. It is now so much a part of society that we take it for granted expecting to be able to use it as and when we need it. Without electricity we would also quickly lose most other key services such as telecommunications, water, gas and radio and television. It is the universal energy.
In autumn 2003 there were a number of high profile electricity failures worldwide. These have sown a seed of doubt that suddenly electricity supplies have become vulnerable. No longer can this essential product be taken for granted. Have electricity supplies become less reliable or were the failures merely coincidental?
The blackout in Canada and the northeastern USA in August 2003 was caused by tree flashovers and inadequate situational awareness
Investigations of these recent incidents have largely been completed and although there were some common features the reasons for these failures were far from simple, which is generally the case with major failures. Very few major disturbances have been attributable to a single event but rather they are due to a sequence of events often triggered by an extremely simple single incident. Weather, human error, equipment maloperation and combinations of these have historically been the most common causes. The following are the four most recent high profile events:
- 14 August 2003– north east USA and Canada affecting 50 million people and lasting for 30 hours – caused by tree flashovers, EMS failures and inadequate situational awareness leading to multiple trips and voltage collapse. Local disturbances, rather than being contained, were allowed to spread.
- 18 August 2003 – London affecting 0.5 million people and lasting for 35 minutes – caused by a wrongly rated protection relay.
- 23 September 2003 – Denmark and Sweden affecting 3.8 million people and lasting for over nine hours – caused by a nuclear power station trip, a double busbar fault and protection maloperations brought on by the low voltage and high transfers.
- 28 September 2003 – Italy 57 million people and lasting for over 18 hours – loss of two lines within 20 minutes first caused by a tree flashover in Switzerland and lack of awareness of the situation.
In most of these incidents the initial causes have been relatively straightforward. But causes like these are not a new phenomenon with similar events having taken place many times in the past. Although in these recent events the causes may have been simple and trivial the consequences were not. Here the electricity utility industry has been a victim of its own success. With failures so infrequent, customers appear to have forgotten that electricity can fail and have paid scant attention to contingency plans to cater for major failures.
However, is there an underlying problem? The electricity industry has moved rapidly from monopolistic, vertically integrated utilities and the procedures developed in this era seem unable to deal effectively with the liberalization, open access and cross-border trades of today. Already before the Italian failure the UCTE implemented a system for Day Ahead Congestion Forecasts (DACF) but there is still a need for improved on-line security assessment.
In the world today there is an insatiable need for information and instant answers and when electricity failures occur companies are bombarded with requests and are subject to speculative statements. What is true is that the public demands a supply of electricity that is 100 per cent reliable. In an increasing number of cases this impossible expectation is driven by statements from the media and politicians and also from industry experts and academics. Both the media and politicians are guilty of the soundbite approach. The reluctance of companies to speculate has to be balanced against the need to reassure the public with real answers. Although the industry is more exposed to public comment than in the past, with politicians and the public seeing things in black and white, comments by politicians are not new.
The advent of liberalization has allowed the freedom to carry out trading, and has in general been to the benefit of the customer. However the freedom that this has given has to be balanced against the requirements of the real world. System operators have had to operate their systems closer to the limit requiring more advanced tools providing better system visualization and closer monitoring.
Operating systems closer to their limits can lead to unexpected effects. For example on 27 September 2002 the heavy south to north flows on the Belgian network, caused by French exports and exchanges from Switzerland to Germany, reached a peak of 2600 MW and the Belgian system operator was forced to warn neighbouring system operators that the opening of the Belgian north border, as a last resort, was imminent – the system operator had the full power and authority to take this action. In the US outage, questions have been asked regarding the role and authority of the system operators.
System operators must clearly be given the authority to ensure that system security must always take priority over commercial considerations. Has the competitive instinct that has been generated made organizations less willing to share information? In a number of instances there has also been a rush to blame someone else whether or not this fits the facts.
In most cases liberalization and unbundling have gone hand in hand. Unbundling has, on the one hand, allowed network operators to focus on network reliability and security but it has required formal arrangements – e.g. the Grid Code, Connection Agreements etc. – to be put in place to ensure that organizations get all the data they need to operate the system. Unbundling has allowed the network operators to focus on network security and to operate it efficiently, securely and safely while at the same time facilitating market operation.
But the network operators are only responsible for the network and have no responsibility for the major issue of generation adequacy. The market must provide the appropriate signals for investment and the ability of electricity markets to function correctly in this respect is crucial. In none of the four failures was a lack of generation adequacy a factor. Unbundling has also included the separation, in vertically integrated companies, of the different parts of the business. It is worth considering if these arrangements are sufficient and that all organizations get all the information they need.
Regulators have to introduce regulatory schemes that balance the requirements of cost and quality and deal with anti-competitive behaviour that works to the detriment of the customer. Regulators are however limited to their own country – or in the USA to a state level – and with electricity systems and flows recognising no borders there is a need for some form of overall regulation.
Weather is a factor in a large percentage of major electricity failures. In some cases it is virtually the only factor, as in Quebec in both 1989, the geomagnetic storm, and 1998, the ice storm, and in the very severe storm in France in 1999. In other cases it acted as a catalyst that set off a chain of events. The 1998 Quebec failure caused what was arguably the worst ever supply failure with some customers off supply for 33 days. Irrespective of whether or not we are going to see more severe weather patterns emerging, weather will continue to play a major part in electricity failures. Although it will never be possible to stop all weather related incidents, better designs may assist in preventing incidents and improved restoration methods will help to limit the effect.
With the spectre of climate change and greenhouse gases there is a major push to develop non-CO2 producing forms of generation such as wind. Large amounts of wind generation will bring new operational problems with large amounts of non-firm generation with a lot of it embedded within the distribution system.
Environmental and planning constraints have made the development of new facilities such as transmission lines and power stations much more difficult. The rise of the ‘Nimby’ effect and the green lobby has virtually put paid to nuclear power in large parts of the world and is now even making wind generation difficult. In 1998 the local power company in Auckland had faced major opposition to its attempts to upgrade the network to downtown Auckland. A total power failure ensued and reinforcement with an overhead line was authorised and installed in weeks rather than years.
Faster, larger more sophisticated SCADA/EMS systems are being deployed capable of providing ever-greater amounts of information. However control centres generally handle larger and larger areas with fewer staff. There are serious questions regarding the processing and presentation of large amounts of data that is ‘thrown’ at them in system emergencies. The staff can find it virtually impossible to assimilate and interpret all this information. The information should be appropriate, reliable and presented in a way that is easy to understand. Although tools have been used for post-fault evaluation, is it fully understood what is required to assist in the management of real-time events?
The systems providing the data must be secure and reliable and they must continue to function in a total blackout. The partial failure of SCADA/EMS appears to have played a major part in the US outage, and even the London failure alarm presentation seems to have played a part. “We have no clue. Our computer is giving us fits, too” (US control room operator, 14 August 2003). “We don’t even know the status of some of the stuff [power fluctuations] around us” (US control room operator, 14 August 2003).
Large amounts of wind generation will bring new operational problems with large amounts of non-firm generation embedded within the distribution system
With the advent of inter-state market trading, system operators must further improve communications with their colleagues controlling other systems. The electricity system is dynamic and flows cannot be directed down a particular path. Its security is affected not only by decisions taken by system operators on information regarding their own system but also by knowing what is happening in the connected systems. It is essential for a system operator to be aware of the situation within neighbouring systems and not simply at the boundaries. Initial indications would suggest that this was not the case in the 2003 failure in Italy where the Italian system operator was reliant on information passed verbally by his Swiss counterpart.
Terrorism and sabotage
Terrorism, like most other problems affecting a power system, is not new – the IRA were blowing up the interconnector between Northern Ireland and the Republic of Ireland 30 years ago. However, with recent events it has become very high profile and it is not only the power system primary assets but also the control centres and control equipment that are at risk. Sabotage is the main suspect for the initial cause in the recent total failure of the power system in Georgia.
It is impossible to accurately predict all possible situations beforehand but it is possible to understand and anticipate in general terms what situations can occur. There is a need for good planning, appropriate levels of investment in plant and equipment and good asset management backed up by an operating regime that is both reasonable and prudent. However there can never be guarantees that an event above and beyond what is deemed to be credible can occur. The system is particularly vulnerable at times of very high demand or when plant is out for maintenance.
It is therefore essential that the key risks are well understood and that contingency plans are available to assist in the management of all foreseen situations. The areas that need to be considered are as follows:
- Planning, design and asset management to limit the risk of failures
- Regulatory schemes that balance efficiency improvement objectives with the need to maintain quality of supply
- Defence mechanisms that will, in the event of system disturbances, automatically contain and limit failures so that a complete system collapse does not occur
- Black start plans to facilitate supply restoration quickly and safely in the event of a full or partial supply failure
- Training of staff to deal with both normal operation and abnormal situations
- Testing of black start and other contingency plans
- The above should be backed up with information systems that can handle large amounts of data and provide information in a clear and unambiguous way.
A key element of these plans is communication both with other electricity companies and with customers and the media. All these plans need to be tested and regularly updated.
A final thought
Of course blackouts can give rise to some unexpected and humorous tales. In the New York blackout of 1965 there was the story of the firemen opening the roofs of stuck, darkened, elevators in skyscrapers all over New York city. “Are there any pregnant women in there?” called the fireman. “Gimme a break. We’ve only been here ten minutes!” replied a New York wit.
However, although it was funny it was not so far from the truth. An intrepid New York Times reporter did a survey of maternity hospitals nine months after the blackout and discovered that there had been a great surge in the birth rate!
*The opinions expressed in this article are those of the authors and not necessarily their companies