Electricity Transmission Grids
(Updated December 2013)
- National and regional grid systems connecting generators with wholesale customers are just as important as electrical power generation.
- Investment in these is often on a similar scale to generation capacity.
- New technology is enabling transmission at high voltages over long distances without great losses.
Countries with well-developed electricity infrastructure have established grids run by transmission system operators (TSO) to convey power to distribution systems where it is needed. Where generating plants can be located close to load centres, these are less important than where the plants are remote, as with many hydro-electric plants and wind farms. Lower voltage can be used. At higher voltages, eg 500kV and above, transmission losses over hundreds of kilometres are much reduced. At ultra-high voltages (UHV) eg 1000 kV AC or 800 kV DC, losses are further reduced (eg to 5% over 1000 km) but capital requirements are greater. In Germany consideration is being given to converting some existing AC lines to DC to increase their capacity.
Wholesale distribution companies ('discos') step the voltage down with transformers, eventually to domestic voltage, and market the electricity.
Transmission grids are normally alternating current (AC), which can readily be transformed to higher or lower voltages. Increasingly, direct current (DC) links are being used for particular projects, particularly undersea cables linking countries or linking offshore wind farms to onshore grids via converter stations.
In a synchronous grid such as western Europe, all generators are in phase, allowing AC power to be transmitted throughout the area, connecting a large number of electricity generators and consumers and potentially enabling more efficient electricity markets and redundant generation capacity. Sometimes AC grids are connected by high voltage DC (HVDC) links, using voltage-source converters (VSCs). HVDC allows connection of asynchronous AC systems. Over 300 GW of new HVDC transmission capacity is expected to be added to world grids by 2020, two-thirds of this in China to connect inland renewable (especially hydro) sources to coastal load centres.
One major issue for many countries setting out to add nuclear capacity to their infrastructure is the size of their grid system. Many nuclear power plants are larger than the fossil fuel plants they supplement or replace, and it does not make sense to have any generating unit more than about one tenth the capacity of the grid (maybe 15% if there is high reserve capacity). This is so that the plant can be taken offline for refueling or maintenance, or due to unforeseen events. The grid capacity and quality may also be considered regionally, as with Jordan for instance. In many situations, as much investment in the grid may be needed as in the power plant(s).
In Europe, the power transmission system operating body ENTSO-E, comprising 41 TSOs from 34 countries, has assessed the ability of Europe's grid networks to become a single internal energy market. This will require some $128 billion in new and upgraded power lines on order to meet the EU's renewables and energy market integration goals. It identified 100 power bottlenecks standing in the way, with 80% of them relating to the challenge of integrating renewable energy sources such as wind and solar power into national grids. One goal (set in 2002) is to have a level of interconnection for each country at lest equivalent to 10% of its generating capacity, to achieve trans-EU electricity infrastructure. This is far from being achieved in 2013, but the above investment will bring it about for all EU countries except Spain.
Germany is a prime example, having its traditional fossil fuel and nuclear power generation plants in the south, with lines spreading from there to the rest of the country, while wind power sources are along its northern Baltic coast. Hence its existing north-to-south to lines have become bottlenecks, incapable of transmitting sufficient wind-generated power from the north to replace closed capacity in the south.
Much of the European investment needs to be on refurbishment or construction of about 51,000 km of extra high voltage power lines and cables, to be clustered into 100 major investment projects dealing with the main bottlenecks. "The fast and massive development of renewable energy sources drives larger, more volatile, power flows over longer distances across Europe and is responsible for 80 out of 100 identified bottlenecks," according to ENTSO-E’s 2012 Ten-Year Network Development Plan.
The TSOs said their analysis showed that extending the grid by only 1.3 percent enables the addition of 3 percent generation capacity and the integration of 125 gigawatts of renewable energy sources -- all at a cost of 2 cents per kilowatt-hour for electricity consumers over a 10 year span. "Cumbersome permit-granting procedures and a lack of public acceptance for power lines are presently the most relevant obstacles" facing the efforts. Hence ENTSO-E proposes that each EU member state should designate a single competent authority responsible for the completion of the entire permit-granting process, which would not exceed 3 years.
Another goal of the EU's grid infrastructure efforts is reducing the "energy island" status of Italy, Iberian Peninsula, Ireland, Britain and Baltic states. This will be addressed by the upgrades, while reducing the total generation costs by about 5%.
An EC-sponsored Booz study in 2013 supported ENTSO-E’s plan to increase transmission by 40% by 2020, but said this rate must be maintained to 2030. “Around 90% of the benefits are achievable even if only half the desirable increment in transmission capacity is achieved, even without demand side reduction,” it said. The study said that integrating EU electricity markets more closely could yield up to €40 billion per year in benefits by 2030, and coordinating renewables investment could add €30 billion per year to that. Improving demand-side responses with smart grids could add up to €5 billion/year, and sharing balancing costs could add up to €0.5 billion/ year, the study says, bringing total potential benefits to €75.5 billion/year by 2030.
In a world perspective, France’s RTE estimates that $700 billion investment in the 16 largest grids handling 70% of world electricity is required over the ten years to 2022, partly on account of integrating renewable sources. The 16 grids have 2.2 million km of lines. RTE itself plans to invest $19 billion by 2020. In developed countries grid development is slow due to the approval process and public opposition.
The main challenges for grid management are frequency and voltage control, in the course of supplying demand which is constantly varying. This means that there must be dispatchable capacity available for TSOs to call upon. Traditionally this is dispatched in merit order, ie according to lowest marginal cost. However, with the establishment of preferential access for intermittent renewables, coupled with relatively high feed-in tariffs or other arrangements, this is compromised to some extent. Where there is large intermittent renewable capacity connected to the grid, supply from it may meet much of the demand (or even exceed it), and it means that low-marginal cost capacity is shut down. Since that kind of plant is often high-capital, low-running cost equipment, its economic viability is subverted.
Grid management authorities faced with the need to be able to dispatch power at short notice treat wind-generated power not as an available source of supply which can be called upon when needed but as an unpredictable drop in demand. In any case wind needs about 90% back-up, whereas the level of back-up for other forms of power generation which can be called upon on demand is around 25%, simply allowing for maintenance downtime. Some discussion on the integration costs of renewables is in the companion WNA paper on Renewable Energy and Electricity.
Voltage and frequency control
One of the basic functions of a TSO is to ensure that voltage in distribution networks and frequency do not depart significantly from the criteria set.
In Europe, there has been a proposal for TSOs to allow more variation in frequency, eg from 50 down to 47.5 Hz for extended periods, so that intermittent renewable sources can better be accommodated. Increased contribution from renewables is called for by some EU governments, regardless of reliability implications. ENTSO-E says that that the proposal for greater flexibility is to resolve "cross-border network issues and market integration issues" one of which requires "facilitating the targets for penetration of renewable generation.” Currently, variation of up to 1 Hz is allowed briefly. The Western European Nuclear Regulators Association (WENRA) has said this proposal has "the potential to negatively affect nuclear safety" because "the definition for the range of frequency and voltage is too large." In addition variability accelerates the ageing of some plant components, especially electric motors. ENTSO-E data shows that increased penetration of renewables is related to a sharp rise in the number and duration of frequency events.
Under technical and design specifications for nuclear safety the lowest frequency allowed for safety-related equipment is 48 Hz and a frequency below that means, for example, that a coolant pump might operate too slowly. Furthermore, nuclear legislation of several WENRA countries does not allow nuclear plants to participate in frequency control or load following, as proposed by ENTSOE-E.
In the EU the permanently permissible range of generator voltage variation is defined from 95% to 105% of rated voltage for up to 15 minutes. For a limited time, generators ought to be capable of operating in a voltage range from 92% to 108% of rated voltage in order to compensate for TSO issues, basically to ensure synchronous operation of the grid and support the system when local voltage problems occur (eg to avoid voltage collapse). At the transmission system connection point for distribution, the voltage is allowed to vary by 10%.
As the role of renewable sources increases, there has been more attention paid to system effects relating to the interaction of variable renewables with dispatchable technologies. System effects refer to the costs above plant-level costs to supply electricity at a given load and level of security of supply. A 2012 OECD Nuclear Energy Agency report focused on "grid-level system costs", the subset of system costs mediated by the electricity grid, which include a) the costs of extending and reinforcing transport and distribution grids as well as connecting new capacity, and b) the costs of increased short-term balancing and maintaining the long-term adequacy and security of electricity supply.
The report showed that while all technologies generate system costs, those of dispatchable generators are at least an order of magnitude lower than those of variable renewables. If the system costs of variable renewables were included at the level of the electricity grid, the total costs of electricity supply increased by up to one-third, depending on country, technology and penetration levels. While grid-level system costs for dispatchable technologies are lower than US$ 3 /MWh, they can reach up to $40 /MWh for onshore wind, up to $45 /MWh for offshore wind and up to $80 /MWh for solar. In addition, the greater the penetration of intermittent renewables, the higher the system costs.
Currently, such grid-level costs are simply absorbed by electricity consumers through higher network charges, and by the producers of dispatchable electricity in the form of reduced margins and lower load factors. Failing to account for system costs means adding implicit subsidies to already sizeable explicit subsidies for variable renewables. As long as this situation continues, dispatchable technologies will increasingly not be replaced as they reach the end of their operating lifetimes, thereby seriously diminishing security of supply. However, maintaining high levels of security of electricity supply in decarbonising electricity systems with significant shares of variable renewables will require incentives to internalise the system costs, as well as market designs that adequately cover the cots of all dispatchable power production, including low-carbon nuclear energy.
The NEA report concluded that nuclear power will fare relatively better than coal or gas in the short run due to its low variable costs. In the long run, however, when new investment decisions need to be made, reduced load factors will disproportionately penalise technologies with high fixed costs such as nuclear. In systems that currently use nuclear energy, the introduction of variable renewables is therefore likely to lead to an increase in overall carbon emissions due to the use of higher carbon-emitting technologies such as gas as back-up.
The existence of high system costs implies that significant changes will be needed to enable an economically viable coexistence of nuclear energy and renewables in increasingly decarbonised electricity systems. Such changes may include more widespread use of carbon pricing, long-term power supply contracts, and capacity payment mechanisms in order to provide adequate incentives for new investment.
China is developing a very sophisticated grid system, since its main coal deposits are in the north, its main wind potential in the far west and its nuclear plants are on the coast – close to load centres. The grid system run by the State Grid Corporation of China (SGCC) and China Southern Power Grid Co (CSG) is growing rapidly, utilising ultra high voltage (1000 kV AC and 800 kV DC) transmission. By 2015 SGCC is investing CNY 500 billion ($75.5 billion) to extend the UHV grid to 40,000 km. By 2020, the capacity of the UHV network is expected to be some 300-400 GW, which will function as the backbone of the whole system, having 400 GWe of clean energy sources connected, of which hydropower will account for 78 GWe, and wind power from the north a further significant portion. Wind capacity by 2020 is planned to be 100 GWe. At the end of 2009, China had budgeted to spend $600 billion upgrading its grid. By 2020 operational transmission losses are expected to be 5.7%, down from 6.6% in 2010.
The northern part of India suffered two massive grid failures in July 2012 leaving first 390 million people without power, and a day later, some 680 million in 22 states, highlighting the country’s infrastructure challenges. The Northern grid was the first affected, then this plus parts of the Eastern and North-eastern grids, after low voltage in one place tripped a link and this led to cascade tripping. Most of the under-frequency relays (UFRs) in the northern region did not work, and load dispatch centres did not respond to the problem. Power to some essential services was restored after a few hours each time, but others were out for more than a day. All five grids are controlled by the Power Grid Corporation, which operates 95,000 km of transmission lines. The country has 33 state load dispatch centres (SLDCs), five Regional load dispatch centres (RLDCs), and a national load dispatch centre.
The USA has a patchwork of grids which are often barely interconnected. In 2012 a report from the American Society of Civil Engineers said that ageing equipment and lack of capacity was leading to intermittent failures, and said that an extra $107 billion investment was needed by 2020. This may be conservative. In September 2011 a simple error led to a cascading and uncontrolled failure which affected southern California and was the most extensive in the state’s history. It rivalled the 2003 failure which left much of the northeast and 50 million people without power. Among the four main causes for the Northeast blackout that investigators listed six months later: The primary utility “did not recognize or understand the deteriorating condition of its system.” Over 1965 to 2009, there were 57 major grid failures in the USA and Canada, according to a study by the Institute of Electrical and Electronics Engineers, 41 of those in the USA and two of them shared.
A 2011 MIT report said that the US grid faced "a number of serious challenges over the next two decades, while new technologies also present valuable opportunities for meeting these challenges." Incorporating more renewable generation is one challenge, increased penetration of electric vehicles is another. But "the diversity of ownership and regulatory structures within the U.S. grid complicates policy-making, and a number of institutional, regulatory, and technical impediments remain that require action." It made recommendations accordingly.
In April 2013 a new 530 km, 1000 MWe capacity HVDC line, the Champlain Hudson Power Express (CHPE) was approved and will be installed underground and underwater, originating at the Canadian border in Quebec and running the length of Lake Champlain and through parts of the Hudson, Harlem, and East Rivers to New York City. Cost is put at $2 billion.
Another HVDC project, the 1000 MWe Lake Erie CleanPower Connector, will bring Ontario power 105 km to the PJM grid supplying 13 states in NE USA at a cost of about $1 billion.
In November 2013 the 1500 MWe Gateway project was approved by the US government, a 1600 km HVAC link from southern Wyoming to southern Idaho to strengthen the western grid and transmit wind power west to a more populous area.
In Germany the existing north-to-south to lines are overloaded and incapable of transmitting sufficient wind-generated power from the north to replace closed capacity in the south. In May 2011 the German federal network agency and grid authority, Bundesnetzagentur (BNA), reported that reserve capacity was fully utilised, no buffer was left, and "For this reason the risk of non-controllable network disturbances is increasing distinctly." Much scheduled service and maintenance work on the grid had to be postponed. BNA warned of very uncertain supply situations likely over winter 2011-12, especially in southern Germany, along with increased costs. In mid 2012 the four German TSOs proposed building 2100 km of DC lines and 1700 km of AC ones to connect North Sea and Baltic Sea wind farms with southern load centres. The cost was put at EUR 20 billion, and getting approval through planning procedures is a major hurdle.
In December 2012 a report from the German Energy Agency (Deutsche Energie-Agentur GmbH, DENA) showed that investment of EUR 27.5 billion to EUR 42.5 billion is required by 2030 to expand and upgrade electricity distribution to cope with increased renewables share of supply. The DENA Distribution Study said that grid and distribution expansion of 135,000 km to 193,000 km was needed. It called for reforms to the regulatory framework to help network operators realise the returns needed as incentives for the necessary investments. DENA is 50%-owned by federal government ministries and 50% by German financial institutions. The Distribution Study was supported by German generation and grid companies including EnBW, EOn and Vattenfall.
Later in December 2012 cabinet agreed to accelerate the program proposed by the four TSOs for building new high-voltage transmission and achieve it in four years rather than the ten years originally envisaged. As well as 2800 km of new lines, about 1500 km of the present grid will be upgraded. It is expected to cost about EUR 10 billion. The planned law aims to speed up the grid construction by placing limits on what legal options opponents can take. State governments have agreed to let the Bundesnetzagentur coordinate plans, rather than asserting regional interests. One project under construction in 2013 is the 380 kV so-called Thuringer Bridge transmission link connecting Saxony-Anhalt in eastern Germany with Bavaria and due to be completed in 2015. Bundesnetzagentur has warned that if it is late, grid security will be compromised due to the closure of E.On's 1345 MWe Grafenrheinfeld nuclear plant then. An increase in north-south transmission capacity is planned by converting the 400 km North Rhine-Westphalia to Baden-Württemberg line to DC, scheduled for 2019 when EnBW’s 1392 MWe Phillipsburg 2 nuclear plant is due to close.
With the permanent shutdown of the eight nuclear reactors affected by the moratorium, Germany could no longer support security of supply in the European interconnected grid to the extent it has done so far. Hence the German situation has implications much more widely in Europe, as the country becomes a net importer of power rather than major exporter. Grid stability is the other major concern, along with generation and transmission capacity. As of mid 2012, a total of 15 out of the 24 most urgent grid expansion projects identified in 2009 were delayed by between one and five years and only 214 km of 1,834 km of transmission lines had been completed.
Plans for a 1400 MWe HVDC link with Norway promise to help Germany's renewable energy objectives, as the interconnector between Denmark and Norway has been doing for Denmark's wind energy for some years. (Denmark's wind turbines depend heavily for their effective utilization on 29 GWe of hydro capacity in Norway, over 1 GWe of which can be dispatched promptly when wind power is unavailable in West Denmark. This will rise by 700 MW in 2014. Hence, there is a natural interdependence between West Denmark's wind and Norway's hydro. With good winds, power can be exported back to Norway and there conserve hydro potential.)
The Czech Republic is one of the adjacent countries affected by Germany's grid problems. Since mid-2012 the 2 GWe Temelin plant has operated about 100 MWe below capacity as instructed by grid operator CEPS because of grid security issues caused by power surges due to renewable power production in Germany.
Austria is another country experiencing difficulties due to input from subsidized wind and solar PV. The Austrian Power Grid (APG) is having increasing difficulty in balancing unpredictable supply to demand requirements. This has raised the need for adequate sourcing of balancing power, requiring dependable sources such as gas-fired generating units to be available. In Austria most of these are now out of service, unable to compete economically, and hence the country has high reliance on uncertain German supply. APG proposes capacity payments to keep fossil fuel capacity in standby mode, especially as further wind capacity comes on line with limited grid access.
France's grid operator RTE plans to invest EUR 15 billion ($19 billion) on the nation's grid by 2020, and a further EUR 20 billion by 2030 with the present energy mix. However, it says that EUR 50 billion would be required by 2030 if the government decides to reduce nuclear share from 75% to 50% of supply and replace this with renewables. Either way, the grid investments are needed to improve security of supply and accommodate rising renewable power capacity. RTE grid transport costs account for about 10% of consumers' bills.
Russia’s Federal Grid Company is 80% owned by the government and controls 125,000 km of transmission lines over 13.6 million sq km. Its customers are regional distribution companies ('discos'), electricity suppliers and large industrial enterprises.
Japan’s grid is very unusual in that on the main island, Honshu, the northeastern half including Tokyo is 50 Hz, served by Tepco (and Tohoku), the southwestern half including Nagoya, Kyoto and Osaka is 60 Hz, served by Chubu (with Kansai & Hokuriku), and there is only 1 GWe of frequency converters connecting them. This arises from original equipment coming from Germany and USA respectively. The interconnection is being increased to 2.1 GWe, funded by the utilities. Early in 2013 it was announced that METI will establish a new body to balance electricity supply and demand in wide areas across Japan, as early as 2015. The new body will manage grid and transmission facilities, which are currently owned and managed by utility companies.
Major regional grid projects
Baltic Energy Market Interconnector Plan (Baltic grid map pdf)
The planned Visaginas nuclear plant is envisaged as the cornerstone of the Baltic Energy Market Interconnector Plan (BEMIP) linking to Poland, Finland and Sweden. A high-voltage (400 kV) 1000 MW DC southwest interconnection – PowerBridge or LitPol Link – costing €250-300 million, to improve transmission capacity between Lithuania and Poland is to be built, with 500 MW by 2015 and another 500 MW by 2020. Much of the funding is from the European Union (EU), and work is ahead of schedule.
This follows inauguration of an interconnector between Estonia and Finland to the north – Estlink-1, a 150 kV, 350 MW HVDC cable costing €110 million and also supported by EU funding. The 170 km 450 kV HVDC Estlink-2 further east and now under construction will provide a further 650 MW early in 2014. The project is budgeted to cost around €320 million, which will be divided between TSOs Finngrid and Elering (Estonia), with €100 million to be provided by the EU as part of the EU’s extensive economy recovery package. Both will be operated by the two TSOs.
Another major transmission link westward under the Baltic Sea, the 700 MWe NordBalt 300 or 400 kV HVDC project, is planned between Klaipeda in Lithuania and Nybro in Sweden (400 km), by Svenska Kraftnat and LitGrid. The €550 million project is expected to be completed by 2016. (The Baltic states and Belarus have good interconnection of grids from the Soviet era, but this did not extend to Poland, let alone to Germany. Kaliningrad gets all of its electricity from Russia, via the Lithuanian grid.)
Lithuania's revised energy policy in 2012 involves rebuilding the grid to be independent of the Russian system and to work in with the European Network of Transmission System Operators (ENTSO) synchronous system, as well as strengthening interconnection among the three Baltic states.
This EU integration was an important factor leading to Russia suspending work on its new Baltic nuclear power plant in its exclave of Kaliningrad. It was designed for the EU grid and about 20% built. Despite endeavours to bring in west European equity and secure sales of power to the EU through proposed transmission links, the 1200 MWe plant is isolated, with no immediate prospect of it fulfilling its intended purpose. Kaliningrad has a limited transmission link to Lithuania, and none to Poland, its other neighbour. Both those countries declined to buy output from the new Baltic plant. Lithuania does not wish to upgrade its Kaliningrad grid connection to allow Baltic NPP power to be sent through its territory and Belarus to Russia. As well as upgrading the Lithuania link, Russian grid operator InterRAO had plans to build a 600-1000 MWe link across the Kaliningrad border to Poland and a 1000 MWe HVDC undersea link to Germany, but with no customers these plans are not proceeding. In March 2013 Rosatom said that Russia had applied for Kaliningrad to join the EU grid system (ENTSO-E), evidently without response.
North Sea grid
In a move towards meeting the EU’s goal of reaching a 20% share of energy from renewable sources by 2020, nine European countries have agreed to build a power grid of high-voltage cables under the North Sea. It would be the first multinational grid designed to address the fluctuating nature of ‘green’ power generation. The North Sea Grid Initiative includes Germany, Denmark, Norway, Sweden, Belgium, France, Netherlands, Luxembourg, and the United Kingdom.
The project aims to link up some 100 GW of offshore wind power which is currently planned by European power companies. The UK has launched a £100 billion programme to boost its offshore wind farms; already the world’s biggest at around 1 GW, to as much as 40 GW by 2020. The project will have an estimated cost of about $40 billion, and is expected to be operational by 2023, balancing supply and loads among regions and from large wind and solar farms.
East African grid
The World Bank is funding a new Eastern Electricity Highway project to connect Ethiopia with Kenya and eventually beyond. It is the first phase of a $1.3 billion East Africa power integration program, with $243 million for Ethiopia and $441 milion for Kenya from the Bank, which says that the “project will change the fundamentals of the power sector in East Africa”.
“Smart grid” refers to a class of technology for electricity delivery which uses computer-based controls to monitor and match supply with real-time end-user demand, varying prices accordingly. It involves two-way communication between distributor and customers’ meters and switchboards, with the management of that information to optimise efficiency. A key feature of the full smart grid is automation technology that allows the utility to adjust and control each individual device or millions of devices from a central location. Smart grids allow optimal integration of domestic-scale renewables into the grid, and also integration of electric vehicles into the system.
Smart grids have major implications at the distribution level, but little at the TSO level. Some 80% of smart grid investment is at DSO level, very little at TSO. Despite talk of electricity highways, HVDC etc, most non-hydro renewable sources are connected to low-voltage distribution networks rather than the high-voltage grids.
Impediments to improvement
The high cost of transmission projects is one disincentive to investment in new capacity.
Acquiring and managing the right of way for transmission assets is a complicated and cumbersome process in many countries, with reliability and customer opinion at stake. Electric utilities and TSOs must manage numerous and often competing interests as they negotiate easements for transmission projects. These will be driven by reliability and capacity objectives, but landowners and government officials have other priorities and interests.
In France, opponents of the 163-km Cotentin-Maine project connecting the new Flamanville reactor to the main grid argued that uncertainty over the safety of living near high-voltage power lines, including the risk of causing leukemia in children, meant that the project should not proceed. Opponents included environmental groups and local public associations. The country’s top administrative court rejected the appeal, saying it was a public interest project and that enough safety assessments had been carried out.
International Atomic Energy Agency, Nuclear Energy Series No. NG-T-3.8, Electric Grid Reliability and Interface with Nuclear Power Plants (2012)
International Atomic Energy Agency, Technical Report Series No. 224, Interaction of Grid Characteristics with Design and Performance of Nuclear Power Plants (1983)
Energy News 24/8/12
Energize Weekly 22/8/12
OECD Nuclear Energy Agency (November 2012), Nuclear Energy and Renewables: System Effects in Low-carbon Electricity Systems, ISBN 9789264188518
OECD/NEA, 2013, CSNI Technical Opinion Papers No. 16: Defence in Depth of Electrical Systems. NEA #7070.