Integration of distributed generation into Europe’s electricity networks may be feasible, but the design of distribution grids in particular needs to evolve towards new models, in order to avoid medium-term problems. Following on from their first article (COSPP September–October 07), Angelo L’Abbate, Gianluca Fulli and Stathis D. Peteves consider the issues and options.
In response to the three European Union (EU) energy-related challenges – environmental sustainability, security of supply, and competitiveness – and within a context of growing electricity demand, distributed generation (DG) may further penetrate the European electric power sector. This trend could also be triggered by emerging technological solutions for more efficient, environmentally-friendly and small-size generating units, as well as by socio-economic and environmental constraints in building new, large generation and high-capacity infrastructures. It must also be noted that the recent new targets for Renewable Energy Sources (RES) deployment in the EU (globally 20% of energy consumption covered by RES by 2020) will foster a rising DG deployment in EU countries.
As stated by the European Commission in the European Strategic Energy Technology Plan (SET-Plan), one of the key EU technology challenges for the next 10 years is to enable an integrated, smart European electricity grid to accommodate a large penetration of RES and DG. Achieving a sustainable, interconnected European energy system requires extensive energy infrastructure changes, which represent one of the most important investments of the 21st century.
Impact of DG on distribution and transmission networks
Traditionally, distribution networks have mainly been designed and operated to distribute power passively from the upstream generation and transmission system to the final customers (see Figure 1). In this situation, with power flows mainly going mono-directionally from the substations to the consumers, the Distribution System Operators (DSOs) do not have the opportunity or the need to take active control of the power flows, unlike the TSOs (Transmission System Operators) for the transmission grids. It is for this reason that most distribution systems are designed as passive.
When increasing numbers of generators are connected to the distribution network, power can also be transferred reversely, from the distributed units to the distribution and the upstream transmission (see Figure 1).
Figure 1. Present and future architecture of electric power system
In this new situation, the distribution may be subject to change control properties and become more similar to the transmission, that is, have more ‘active’ control features.
Several aspects have to be considered regarding the impact of DG on distribution networks:
- network capacity and congestions
- short circuit currents
- protection selectivity
- network robustness
- voltage profile
- system stability
- system balancing and reserve
- power quality.
DG integration into electric power systems depends on the effects that DG installation may have, not only on the distribution networks, but also on the upstream transmission system. In this respect, different aspects have to be considered:
- steady-state effects
- contingency analysis effects
- protection effects
- dynamics effects.
In particular, DG connection may lead to the change or distortion of the profiles of voltage at transmission nodes. This may occur especially if the transmission network is generally ‘weak’ having not enough large generating capacity to control voltages.
Furthermore, the impact that DG may have on transmission losses is strongly affected by the DG location, but also depends on network topology as well as on DG size and type. The influence of DG on transmission congestions also depends on DG location. Strategically-located DG units may utilize the upstream transmission system less, if opportunely operated, and thereby help relieve overloaded branches in the transmission network.
DG issues at the transmission/distribution interface
DG integration in the Italian grid
The Italian Energy Regulator defines distributed generation as production from units with rated power lower than 10 MVA. This definition means that most generation qualifying as DG is hydro and fossil-fuelled plants connected to the low and medium voltage (< 30 kV) network and representing some 5% of the generation capacity (2005 data). If larger sized plants – up to 50 MVA – are considered as DG, the share of DG is estimated at some 15% of the total national generation capacity. This includes wind and biomass units, preferentially connected to higher voltage (60–150 kV) networks. The distribution grids are operated by 169 DSOs at present.
A considerable share of DG is installed in northern Italy, due to the higher level of industrialization and the abundance of hydro resources. Important shares of conventional thermal DG are also present in some central and southern regions. This situation may change in the next few years, due to the anticipated increase of DG – particularly renewables, like wind and photovoltaic (PV), and combined heat and power (CHP) – fostered by a favourable regulatory framework.
The growing deployment of DG technologies brings into play a number of issues related to DG connection and the co-ordinated development of the electricity system. Where DG grid integration is concerned, some specific criticalities affect the high voltage network (120–150 kV), which is one of the most targeted in Italy by the deployment of larger-sized DG resources. In fact, this network partially belongs to the TSO and partially to the largest DSO, notwithstanding the tight interconnection and physical infrastructure proximity.
The presence of different players is thought to hamper co-ordinated development of such intertwined transmission and distribution networks. In several cases – for example grid rationalization works, restructuring activities to mitigate the environmental impact, connection of RES spread on the territory – it is necessary to work on bordering inter-operating grids to build new substations or implement particular network reinforcements. The main reason for this rests with the need to relieve existing network constraints and manage operational issues. As the involved systems are strongly interconnected, it is often unclear how to share the investment costs and how to co-ordinate construction activities. The change in power flow patterns is not the only consequence of a mixed development of inter-operating networks; even the short circuit level increase is to be duly assessed so as to understand the need to upgrade electrical equipment. The solution now under study in Italy focuses on transferring the ownership of the high voltage distribution assets (19,000 km of lines) from the largest DSO to the TSO.
The role of DG in the European 2006 blackout
The increasing penetration of DG capacity in Europe has to be properly managed by the DSOs and TSOs. Special attention needs to be devoted to the intermittent nature and operational features of some DG sources (such as wind and, to a certain extent, CHP). The European blackout of 4 November 2006 gives a clear example of how DG operational behaviour may affect not only the distribution, but also the upstream transmission grid. On that occasion, the European power systems of the UCTE (Union for the Coordination of Transmission of Electricity) synchronous area were largely impacted by a serious system disturbance originating from the North German transmission grid. Many high voltage lines were tripped, leading to significant power imbalances and frequency deviations. The UCTE grid was then divided into three areas: west (in under-frequency), north east (in over-frequency) and south east (in slight under-frequency).
In case of frequency variations in the pre-defined range around 50 Hz, generators are normally required to remain connected to the grid and contribute to the system’s frequency regulation. During the disturbance on 4 November 2006, a large amount of DG, mainly wind and CHP units, was automatically disconnected in the under-frequency area, as well as automatically reconnected in the over-frequency area. This increased the imbalance between supply and demand, thus worsening the effects of the disturbance, e.g. in terms of number and duration of consumer disconnections. This was due to the national connection rules and technical characteristics of DG plants having less stringent operational standards than larger conventional power plants. An additional issue was that TSOs do not have control and real-time data about DG plants connected to distribution grids, neither at aggregated level for a grid area, nor for a DSO.
As highlighted by the European Energy Regulators, it is important that DG units contribute to network frequency and voltage control in the same way that larger conventional power plants do. Also, information provision by DG plants and DSOs to the TSOs, and procedures for automatic tripping and co-ordinated reconnection, need to be reformulated to enable TSOs to control the state of the system more effectively. These aspects can play a major role in the presence of an increasing penetration of DG, especially during disturbances and abnormal operation of the power system.
The evolution of the distribution systems
Today’s distribution grids can be suitable to accommodate increasing shares of DG in a short-term perspective, provided that all the requirements set by the DSOs are met. However, in the mid–long term, the distribution systems may be subject to the most profound changes in terms of system design, development, and network operation philosophy. Such evolution is expected to be gradual and uneven in its progress in the several European electricity distribution grids.
The more DG devices penetrate the distribution networks, the more these systems are expected to evolve towards transmission-like architectures. Unlike transmission grids, the distribution networks have generally not been designed to operate in the presence of power injections. Indeed, the distribution networks have been developed with radial structures (especially at low and medium voltage level) or with meshed structures and radial operation (mostly at medium and high voltage level).
Figure 2. Expected evolution of electricity grids
Figure 2 graphically summarizes the options in terms of network structure and operation. With an increasing DG penetration the situation is changing and resulting in possible bi-directional power flows (uni-directional power flows without DG).
Figure 3. Basic scheme of distribution grid evolution process
The transition process of distribution systems towards transmission-like schemes is generally gradual and may require several intermediate steps. Figure 3 schematically shows the possible stages, linked with the level of DG penetration increase. This process would then move from the traditional approach of simple DG connection (‘fit and forget’) to different changes/upgrades (grid protection systems, network reinforcement), up to the development of new control strategies.
Some possible developments of the distribution system in the near future as parts of the above described evolution process are discussed below. In particular, focus here is on active networks, microgrids, and virtual power plants.
These schemes may need the utilization of advanced solutions such as information and communication technology (ICT) and/or flexible controlling devices (FACTS).
Active networks are foreseen as the probable evolution of today’s distribution networks. The latter systems, which are passive, can evolve to be structured and operated in a similar way to transmission systems, which are active, managing bi-directional power flows. This change of the distribution design may be triggered by the connection of an increased number of small generating units.
This evolution shall be accompanied by an appropriate upgrade of the protection schemes, along with the introduction, at different stages, of new soft (ICT) and hard (power electronics-based devices) technologies for a more flexible system control.
Figure 4. Configuration of traditional and active distribution grids
Figure 4 depicts an example of the evolution of distribution architecture, from the traditional one towards the active distribution scheme. This first transformation is already ongoing in several European countries.
Within the power system, a microgrid is a portion of a distribution network containing DG sources, together with local storage devices and controllable loads. It can be regarded as a controlled entity which can be operated as a single aggregated load or generator, eventually providing network support and services. Microgrids generally have a total installed capacity between a few hundred kW and a few hundred MW.
The unique feature of microgrids is that, although they operate mostly connected to the distribution network, they can be automatically disconnected by intentional islanding in case of faults affecting the upstream network. Then, microgrids having sufficient generation and storage resources can ensure power supply to local customers. After a fault has been cleared and the upstream network operation restored, microgrids can be resynchronized to the rest of the system.
Additionally, the islanding procedure could also allow a microgrid to support ‘black start’ in case of a widespread system outage. In the case of large disruption, a number of dedicated DG units may be able to feed local loads and eventually to resynchronize this part of the grid with the main system. In this way, a microgrid would contribute to the restoration of normal operation conditions of the whole system.
Figure 5. Configuration of a microgrid and a virtual power plant
Figure 5 shows the structure of a microgrid.
Virtual power plants
The virtual power plant (VPP) is a decentralized energy management system tasked to aggregate different small generators for the purpose of energy trading and/or to provide system support services. The VPP concept is not itself a new technology, but a scheme to combine DG and storage and exploit the technical and economic synergies between systems’ components. This aggregation is not necessarily pursued by physically connecting the plants but by interlinking them via soft technologies (ICT). For this reason the result is a virtual power plant, which may then be a multi-fuel, multi-location and multi-owned power station.
A VPP balances required and available power in identified areas, based on offline schedules for DG, storage, demand-side management capabilities and contractual power exchanges. For a grid operator or energy trader, buying energy or ancillary services from a VPP is equivalent to purchasing from a conventional station. VPP using DG, RES and energy storage can potentially and gradually replace conventional power stations. Figure 5 illustrates the concept of VPP.
The integration of DG into today’s distribution grids in Europe is ongoing and can be technically feasible, provided that the DG plants fully meet the requirements set out in the system operators’ grid codes.
However, in the case of increasing DG penetration, planning and developing new distribution grid architectures (like the active network, the microgrid and the virtual power plant schemes) represent crucial measures to avoid serious mid-term problems that may occur in the power system.
If the growing DG output is not handled appropriately (see, for example, the Danish case on page 24), portions of the distribution systems may be exposed to risks of weakened reliability.
In addition, as seen in the Italian case and the European blackout, the possible effects on the transmission system, caused by the growth of DG, cannot be neglected. One reason for that is the lack of clear technical or legal definitions of the borders between electricity transmission and distribution. Furthermore, without the use of properly co-ordinated system interfaces, flexible controlling devices and smoother communication tools between DSOs and TSOs, the consequences of a power disruption at distribution level may be suffered (if not amplified) at transmission level.
A closer interaction between TSOs and DSOs is then essential in the operation and planning of distribution systems with DG.
Angelo L’Abbate, Gianluca Fulli and Stathis D. Peteves are with the Institute for Energy, Petten, The Netherlands, which is part of the European Commission’s DG Joint Research Centre.
The views expressed in this article are the sole responsibility of the authors and do not necessarily reflect the views of the European Commission
The Danish experience with DG integration
Denmark currently has the highest DG share in Europe, with over 30% of installed power generation coming from DG. At distribution level, there are currently 107 DSOs (mainly small companies) operating the Danish distribution networks. The DG present in the Danish power system mostly comprises small and medium CHP and wind power plants. This is particularly a feature of the system in western Denmark, where DG capacity amounts to over 50% of the total. This DG capacity is installed throughout the distribution systems at voltage levels of 60 kV and below.
Due to this large DG penetration, problems may arise when predicting and controlling the total power generation, as CHP units mostly operate on the basis of heat demand, while wind plants produce according to wind availability. When the DG production exceeds local consumption, the transmission lines to the interconnected neighbouring countries are used to sell the surplus production.
However, different problems in terms of security and reliability (e.g. contingency analysis criteria frequently not met, missing information on DG power production, protection systems often ineffective or not selective in their intervention) have to be tackled by the system operators with the increasing DG penetration.
To face the above issues with DG in the western Denmark system, the Danish TSO is developing advanced technical solutions like the ‘cell architecture’. The aim of this system architecture is to streamline and centralize the control of a large number of DG units widespread throughout the distribution systems, while also exploiting DG benefits and counterbalancing the DG impact on grid operation.
A cell can be characterized as a portion of the western Denmark distribution system down to a 150/60 kV substation, containing DG capacity and local loads typically up to 100 MW aggregated. A cell is structurally like a microgrid.
By this project the TSO expects to interact with the DSO whose distribution network contains the cells. The TSO will then be able to control the cells centrally as conventional power plants, that is in terms of: power output; ancillary services and network support request; disconnection from the transmission network (intentional islanding) in case of upstream emergencies; black-start capability request for a post-fault restoration operation.
In normal situations, a cell relies on the upstream transmission system, exporting the excess production from DG or importing the needed power when DG cannot cover all the local load consumption. When implementing intentional islanding, communication assumes a crucial role. In fact, a command from the SCADA-based transmission control centre (operated by the TSO) communicates to the cell controller (operated by the DSO) that the local generation and demand must be quickly balanced. Also, in the islanded mode, voltage and frequency control must then be carried out at cell level by using the DG resources. Other key issues to be addressed concern the dynamic seamless transition of protection, control, and other network schemes from grid-connected mode to islanding and vice versa.
Communication between the DSO and the TSO is also essential for requesting the cell to provide black-start capability support and restore the service after a fault.
It is evident that by this new distribution system architecture it will be possible for the DSO to control real and reactive power exchanges between the cell and the upstream transmission, as well as to monitor the DG components and the flows within the cell.