Road transportation equipped with conventional combustion engines is responsible for a significant share of carbon-dioxide emissions and polluting cities with fine dust. On the contrary, electric vehicles are environmentally cleaner and more efficient than conventional cars, even when all the energy losses from the power plant to the electric vehicle are taken into consideration.
The depletion of fossil fuel reserves is stimulating the use of renewable energy resources (for example, wind power and solar energy), but the intermittent output characteristics limit the optimal use of these resources and complicate their integration into the power system. The flexible load of electric vehicles can use the electricity produced when available, thereby supporting a high penetration of intermittent renewable energy generators. Furthermore, electric vehicles need to be regularly recharged, offering storage capacity that creates the opportunity to transfer more energy through the distribution network by using the surplus load-transfer potential of the existing network.
When managed intelligently, it is possible for the network to both supply the additional energy demand for electric vehicles and cope with intermittent energy production from renewable energy resources. To achieve this optimal approach, the distribution network must have sufficient capacity for the additional load and be equipped with the secondary systems for communication and control.
Enexis, a distribution network operator in the Netherlands, has analyzed the available capacity on a section of the existing distribution network to determine how this can be used to support charging electric vehicles through the introduction of the mobile smart grid, which also will have to include secondary systems to control the flexible loads.
Topology and Operation of Medium-Voltage Networks
The design of medium-voltage (MV) networks in the Netherlands comprises a (regional) distribution network supplied by high-voltage (HV) to MV substations. Typical primary voltages of HV/MV transformers are 220 kV, 150 kV, 110 kV or 50 kV; secondary voltages include 25 kV, 20 kV and 10 kV.
The design of the MV network is normally based on the use of single transformer feeders, but more complex variations frequently occur in which, for instance, a MV substation is connected to several other MV substations. Also, many MV installations at HV/MV substations feed both MV transmission networks and MV distribution feeders, designed to operate as open rings.
MV transmission networks normally satisfy the n-1 criteria, which establishes system firm capacity whereby no load is lost when one of the cables in a parallel circuit is isolated because of a system fault or for maintenance work.
Capacities of Medium-Voltage Cables
At Enexis, the maximum current rating of an existing MV cable under n-1 operational conditions is based on the nominal rating, with de-rating factors assigned to take into account the proximity of parallel circuits, circuit loading characteristics, soil temperature and type of cable insulation.
Medium-Voltage Networks in the Province of Limburg
The capacity of the distribution networks installed in the Province of Limburg in the south of the Netherlands has been analyzed. This region comprises one-fifth of the total distribution networks operated by Enexis and includes 29 HV/MV substations. No transformers are installed on the MV networks; all transmission and distribution cables in these networks operate at 10 kV.
The average load profile of the 69-kV transmission cables in the Province of Limburg in 2007 on the day of peak system demand, which was Dec. 12, and the average capacity of these cables were determined. The capacity of the cables was based on the nominal cable loading with allowances made for the differences between cross-linked polyethylene-insulated and paper-insulated lead-covered cables, and aluminum and copper conductors. To determine the cable capacity, it was assumed the cables were continuously loaded, a conservative assumption.
The results showed that 64% of the total energy transfer capacity of the MV transmission cables was not used in normal operation. This is mainly because of the n-1 criteria applied in the design (when a fault occurs, a part of the capacity is needed to meet this criteria) and a conservative estimation of the simultaneity of peak loads. Moreover, the networks are designed and installed to supply foreseeable future loading.
For the distribution cables, an average load profile and capacity of 147 MV distribution cables supplied by MV transmission cables was evaluated and, as continuous loading conditions were assumed, correction factors were applied. The evaluation confirmed that an even higher percentage of the capacity (76%), compared to the capacity of the MV transmission cables, is not used in normal operation. This high percentage is mostly caused by the fact that these distribution networks are designed to be operated as open rings. When a fault occurs, one part of the ring can be fed through the other side of the ring by closing a net opening. Therefore, extra capacity margin is needed.
The results confirm there is cable load-transfer capacity available to supply additional energy in MV transmission and distribution cables, provided no faults occur. Further studies were undertaken on one of the Limburg distribution networks to determine the available surplus load-transfer capacity available in this distribution network and the MV/low-voltage (LV) transformers.
Three loading situations were simulated for a MV network supplied by a HV/MV substation that comprised 228 cables and 187 MV/LV transformers. The three situations were as follows:
Situation 1. Under normal operation it is presumed domestic customers have a dynamic load profile, using, on average, 70% of the capacity needed for peak demand. For industry, a more continuous load profile is applied.
Situation 2. Maximum continuous loading of cables and transformers is the same maximum load used in Situation 1. However, in general, more energy can be transported while the load is continuous throughout the day.
Situation 3. Only 50% of the extra available capacity of cables and transformers is used (for example, by electric vehicles). It is assumed the additional load is as equally spread over the network as the load in the first situation.
The results presented that in the first situation, the peak load current on a few cables and transformers exceeded the nominal, continuously allowable current. Because of the relatively long thermal time constants in the cables and transformers, this is currently not a problem.
In the third situation, the peak load current in 14 of the 228 cables and in 88 of the 187 transformers exceeded the nominal, continuously allowable current. Therefore, to use the surplus grid capacity, some of the MV cables and MV/LV transformers would need to be upgraded.
The network load analyses confirm that the existing MV networks have surplus capacity that could be made available to supply flexible loads. Electric cars are flexible, not time-critical, loads that can be disconnected from the network or adjusted to a lower power when required. However, to optimally use this capacity, these loads should be coupled to the grid and operated in an intelligent way.
Charging Electric Vehicles
If flexible loads, such as electric vehicles, could be controlled in such a way the load is optimized throughout the day, the capacity of the existing networks can be used more efficiently and more energy can be transported. Electric cars can be charged during off-peak load periods when the demand is low, and in this way, the increased electricity demand can be supplied with only a limited need for further investment in grid capacity. This is in contrast with the design of the existing grid, which traditionally is based on peak load forecasts. Flexible loads offer the opportunity to use the extra capacity that is already available in the existing grid but is currently not used.
The Mobile Smart Grid Concept
To be able to use the surplus capacity, or hidden potential, of existing grids for electric vehicles and intermittent renewable energy sources, a control strategy is needed. Enexis is introducing the mobile smart grid concept. This concept includes data collection from the flexible loads, on the basis of which a loading schedule can be determined, taking into account customer preferences, local grid capacity, and the actual and forecast availability of electrical energy generators.
An adequate communication structure, such as the Internet, must support the flow of information needed for intelligent charging of electric cars (that is, adjusting the loads to the fluctuating in-feed at the distribution network (decentralized generation) without a car owner experiencing any inconvenience).
The loading schedules of the electric vehicles can be adjusted when extra electrical energy is available or when service interruptions occur. It should also be possible to slow down the charging speed of the electric vehicles in case of emergencies on the MV transmission network to prevent interruptions. In this way, the n-1 principle of the MV transmission network is still guaranteed and the reliability of supply to all other customers is not adversely affected by the mobile smart grid approach.
The in-depth examination of the existing capacity on Enexis' grid showed that 64% of the load-transfer capacity of the MV transmission cables is available to transport extra energy and an even higher percentage (76%) is available in the MV distribution cables fed by these MV transmission cables. However, these two percentages are reduced when the surplus capacity is corrected by factors that consider soil resistivity, soil temperature and the thermal influence of cables laid in close proximity.
Provision must be made for surplus cable capacity that is needed to supply future increases in load. Nevertheless, a part of the existing capacity can be used for flexible loads without the need for further investment.
Based on the assumption that 50% of this available existing capacity is used to transport energy, load calculations for a MV distribution network show that some cables and a large percentage of the MV/LV transformers installed will need to be upgraded. Furthermore, to use the available surplus capacity for flexible loads such as electric cars, the flexible loads will need to be controlled. This can be done by the mobile smart grid concept, which will include the intelligence systems to slow down the charging of electric vehicles, when necessary, to secure the n-1 principle.
The mobile smart grid concept will be developed further to make optimal use of the existing capacity and when additional work involving a complete load flow survey of the Enexis grid is completed.
The colleagues of the regional asset management department of Enexis are acknowledged for their cooperation and availability of the data provided.
Else Veldman (email@example.com) received a MSME degree from the University of Twente in the Netherlands and now works in the innovation department of Enexis B.V., a large Dutch distribution network operator. Veldman is also with the electrical energy systems group at the Eindhoven University of Technology in the Netherlands, where she is pursuing a Ph.D. on the future function, planning and operation of electricity distribution networks.
André Postma (firstname.lastname@example.org) graduated in 1975 in communication and high-voltage power systems, and from 1975 to 1985, he worked for the Royal Dutch Air Force as a special officer for communication and air base power systems. In 1985, he joined PNEM, a Dutch energy company, later re-named Essent, and he now works for Enexis B.V., an independent distribution network operator. For the past few years, Postma has focused on the investigation and introduction of electric vehicles, RES and superconducting techniques.
Han Slootweg (email@example.com) received a MSEE degree in 1998 and a Ph.D. in 2003, both from Delft University of Technology in the Netherlands. Slootweg, who also has a MBA degree, is currently manager of the innovation department at Enexis B.V. Slootweg has a professorship in smart grids with the electrical energy systems group at the Eindhoven University of Technology in the Netherlands.
Madeleine Gibescu (firstname.lastname@example.org) received a diploma in power engineering from the University Politehnica, Bucharest, Romania, in 1993, and a MSEE degree and Ph.D. from the University of Washington in the United States. Gibescu has worked as a research engineer for ClearSight Systems and as a power systems engineer for AREVA T&D Corp., but currently she is an assistant professor with the electrical power systems group at the Delft University of Technology in the Netherlands.
Wil L. Kling (w.l.kling @tue.nl) received a MSEE degree from the Eindhoven University of Technology in the Netherlands and, from 1978 to 1983, he worked for KEMA before joining TenneT, the Dutch transmission system operator, until 2008. In December 2008, Kling was appointed a full-time professor and chair of the electrical energy systems group at the Eindhoven University of Technology. Kling has been a part-time professor at the Delft University of Technology (since 1993) and at the Eindhoven University of Technology (since 2000).
Company mentioned: Enexis www.enexis.nl