The EU is therefore facing questions about what needs to be done to facilitate the addition of more wind power on to the power system networks.
Countries new to wind can look to the experience of several European countries in how to manage high wind penetration into the network. Meanwhile, these established nations have greater ambitions for wind, bringing fresh challenges for themselves.
The issues presented when trying to integrate more wind are largely technical ones relating to the power systems networks, but they also affect policy makers. As wind power is subsidised in order to encourage generation and therefore achieve lower emissions, it is important to ensure that these emissions reductions will not be offset by an increase in emissions when other power sources are used to balance the system, and that hidden costs from integration are minimised.
With Europe hoping to generate more than 13% of its electricity from wind by 2020, ongoing studies into wind's integration are vital. The International Energy Association's wind agreement (IEAWind), whose Task 25 is co-ordinating international efforts to integrate large amounts of wind, is currently preparing a report entitled Recommended methodologies for wind integration studies. These methodologies will be presented at the European Wind Energy Association conference in Copenhagen this month, with the full study to be published later in the year.
Defining integration costs
Before integrating any new power plant, the adequacy of the grid and the implications of balancing the system need to be assessed. While many new power plants require grid reinforcements, not all are able to take part in balancing the system, thus adding to the burden of those plants that can. All power plants incur balancing needs in the rare event of failures.
Integration costs are those incurred to ensure that the needs of the customers are reliably met, so that when power is required by industry and individuals, the voltage and freqency is available. They do not include the costs of installing and connecting a new power plant. In principle, it should be possible to calculate the integration cost for any power plant. But in practice it is not that simple. Both the grid and balancing actions are used by the system as a whole, and the system services are there for all loads and generators. This makes it much more challenging, if not impossible, to separate out and apply every different system cost to specific generators.
Sharing the load
The beauty of a large power system is that the balance between the consumption of many electricity users and the generation of all the power producers can be made across the whole system. Every hour there are many consumers and generators with individual imbalances but, the larger the system, the more they cancel each other out. However, this does make it challenging to identify which individual players are causing the imbalances. And it would not make much sense to balance any individual generator or load, as balancing the system's net imbalance will yield the desired result with much less effort.
Energy storage is sometimes seen as a specific solution for wind power. However, it is necessary to keep in mind that it is less useful when paired with any specific plant or technology than when viewed as a general resource for the whole grid network. This also means that it will have to compete with other options for flexibility, such as the conventional power plant.
Grid reinforcements, similarly, benefit the wider system, not just a new power plant. System operators do not usually earmark investments in the grid for a certain reason but calculate the needs and benefits for the whole system. The simplest way to apply a portion of costs specifically to wind power is to allocate the costs associated with the parts of the grid that are mostly needed for wind. Doing this, however, risks overlooking the overall benefit to the users of that grid.
Variable, not intermittent
Wind (and solar) are variable, "scheduled by nature", and their output changes over time in a way that can be predicted. These predictions are more accurate for a couple of hours ahead and for a larger, dispersed fleet of wind farms, and less predictable when forecast a day ahead for a single project. Wind generation is often referred to as "intermittent", which is actually misleading, as large-scale wind power changes in quite a smooth way over several hours and days. Production from any single power plant (including nuclear and coal) can be intermittent, as they can trip offline unexpectedly in seconds.
The variable nature of wind power does affect a power system's operational security, reliability and efficiency. Integrating wind requires changes in that system, such as incorporating wind forecasts and monitoring the current level of generation in the control rooms. At higher wind penetration levels, the methods and tools used for planning and operation, such as allocation of reserves, need to be adapted as transmission capacity and balancing resources need to increase with high amounts of wind power.
Current situation in Europe
There is already high installed wind capacity in Germany and Spain, and several other countries have a high share of wind on their power systems. In 2011, wind accounted for more than 5% of the EU's annual consumption. The percentage of wind power achieved per country in the EU are shown on the map.
Denmark has been the pioneer, with more than 25% of its annual electrical energy coming from wind power, much of it in the western part of the country. This means that there are occasions during the year when wind-power production is higher than the consumption in the area. Denmark has managed this oversupply by using its strong connections to neighbouring countries. The hydropower-dominated Nordic market has enabled Denmark to use short-term forecasting and market bidding to deal cost-effectively with wind variability and forecast errors.
Spain and Portugal offer another interesting example since these two countries are part of a large power system but there is very limited connection from the Iberian peninsula to France. In addition to short-term forecasting, the system operators here are provided with online information and they are able to control generation levels through centres for distributed generation. This has enabled coping with periods with as much as 50-70 % instant penetration levels of wind power.
The island of Ireland forms a small synchronous system, with the same electrical frequency, and has about a 10% share of wind power on a yearly basis, with several situations when the wind provides a 50% share or more. Such large penetration levels in a whole system that operates at a synchronised level is one current challenge, and research is under way to establish how to best manage the networks.
The National Renewable Energy Action Plans (NREAPs) submitted by EU member states in 2010 pointed to a threefold increase in the share of wind power by 2020. Thirteen countries are aiming for at least 10% of electricity from wind and solar, and eight for more than 20%. Ireland and Denmark have targets in excess of 30%. The map on page 16 also illustrates the 2020 targets of EU countries for wind power's share of electricity generation.
With such high targets comes the need to research the operation and design of power systems to ensure that the power supplies remains reliable and secure. More transmission — connection to neighbouring areas and a strong grid within countries — is required by all. Interconnection between Ireland and the UK, and Spain and France is planned. Even well-connected Denmark plans to increase interconnection capacity to Norway so that it can double its current wind generation.
Europe-wide studies and co-ordinated transmission planning of system operators are also taking place, overseen by the European Network of Transmission System Operators for Electricity
Another prerequisite is flexibility in the power systems. Spain and Portugal are increasing their pumped hydropower capacities, and Denmark is looking at all options for demand-side flexibility, such as channelling excess power towards a system for electric vehicles. Further studies are also taking place to look at the best way to calculate reserve power requirements.
Ireland has been studying the changes needed to enable secure operation with up to 75% penetration of wind power. Given that a synchronous system relies on the physical properties of synchronous generators when disturbances deviate the frequency from the nominal, the country is testing the technical limits of its system to find out how much asynchronous (wind) generation there can be in a synchronous system.
Wind turbine capabilities are continuously evolving, making them easier to control according to power system needs. Regulating with wind power does mean production losses but, in critical high wind penetration situations, it can reduce the need to curtail wind in order to manage the system with other power plants online.
Building on previous research, wind integration study methodology has evolved. And with international collaboration under the IEA's Task 25, best practice is emerging. A complete integration study will include several parts, and usually means an iterative process, as illustrated by the flow chart further down.
Calculating the cost
Analysing the results of a study to calculate the cost of integrating wind on the network depends on the assumptions made and the set-up of the study. Assumptions of a larger wind share in the power system usually refer to 10-30 years ahead, so researchers must consider other investments that would be made during that time period. As system costs are difficult to allocate to any single plant or technology, the studies must aim to quantify incremental increases in costs. Similarly, as most grid upgrades benefit non-wind power generation plants too, it is also difficult to allocate grid reinforcement costs specifically to wind. Another issue is operational costs, which will actually be reduced by the use of wind power because the bulk of operating costs are fuel-related.
The integration cost,then, can potentially be produced by a cost-benefit analysis: the difference between the credit for cutting operating bills and the cost of the increased variability introduced by wind. Gas or coal power plants will have to adjust for this variability but the consequent lower efficiency and higher wear and tear are quite modest compared to the total operating costs. The studies published so far have reported reasonable integration costs for wind power that will not offset the benefits of emission-free power production.
Increasing international co-operation between research institutes and Europe-wide power system operators will simplify the search for ways to achieve smooth integration into the European network of the increased proportion of wind power set out in the NREAPs.
ANATOMY OF A WIND STUDY — WHAT IS INVOLVED
A wind integration study starts with data from a plant location and output, configuration of the rest of the power system and the load level for the year(s) under consideration.
From this penetration level (black boxes in the diagram above), the scope of the system to be studied should be determined — the whole synchronous system or part of it?
The portfolio development step defines the details of the system to be studied: the present or a future system; assumed generation fleet; demand; flexibility options available; and interconnection options to neighbouring areas. The basic setup assumptions will have a crucial impact on the results achieved by the study. How is the wind power added — will it replace something else or will the existing generation stay the same? What is the aim of the research: to study a policy limit or economic consequences, or assess reliability implications?
Changes in system management may need to be made from the start to accommodate large amounts of wind power. This involves checking the options for flexibility through both operational measures and the transmission grid. Transmission planning entails contingency analysis and stability studies as well as profitability analysis of the investment options. Allocation, procurement and use of reserves in a cost-effective manner may also have to be reviewed.
There will likely be investigations of transmission adequacy, simulations of the operation of power plants in the system (when to start and stop, hourly generation levels, including wind plant output forecast error) and calculations of the adequacy of capacity to meet peak load situations.
Dynamic simulations and flexibility assessment are necessary when studying higher penetration levels. Reliability constraints from transmission, capacity or reserve margins will require an iteration on the results to change the installed capacity of the remaining power plants, the transmission grid, the operational methods or the reserves.
Hannele Holttinen is co-ordinating the International Energy Association Wind Task 25 research collaboration