Renewable energy resources, which are becoming integrated into electric power systems around the world, connect to existing transmission grids at a range of voltage levels. The changes brought about by these new power sources are certain to have a significant impact on system performance and efficiency and to necessitate advances in the planning and operation of electric grids.
This article focuses on the impact of the penetration of renewable resources on the electric grid in terms of system performance and the technical challenges and opportunities for achieving higher levels of reliability and efficiency in grid performance. The specific focus is on wind and solar energy, the renewable resources with the most potential for significant penetration in the near term.
The Current Grid
The electricity delivery infrastructure can be broadly divided into two subsystems—the transmission subsystem and the distribution subsystem—which are predominantly distinguished by different voltage levels (Figure 1). The transmission subsystem, also called the bulk power system, primarily delivers electricity generated at central stations to locations close to load centers. In North America, the transmission system usually operates at voltage levels of 69 kilovolts (kV) to 765 kV, is highly meshed, and has significant levels of automation and control.
The distribution subsystem, which delivers electricity from load centers to customers, operates at voltage levels ranging from 26 kV to 120 V, is predominantly radial in structure, and does not have the same level of automation as the transmission subsystem. The network infrastructure at 69 kV, which serves as the interface point between the transmission and distribution subsystems, is usually referred to as the sub-transmission system.
Early versions of wind turbine generators consisted of fixed-speed wind turbines with conventional induction generators. This class of machines was rugged but was limited to operation in a narrow wind-speed range. In addition, the conventional induction generator, which was directly connected to the electrical grid, required that reactive power support be provided locally to achieve the desired voltage level.
Advances in power electronics have revolutionized wind turbine technology and led to the development of the doubly fed induction generator (DFIG) (Figure 2). The stator of the DFIG is directly connected to the grid, and the rotor winding is connected via slip rings to a converter, which only has to handle a fraction (20 to 30 percent) of the total power. The highly efficient, variable speed DFIG is designed to extract maximum energy from the wind, and it puts out electricity at a constant frequency no matter what the wind speed.
Most modern wind farms have DFIGs and are available in ratings that range from 1.5 megawatts (MW) to 4.5 MW. Newer generations of wind generators, which have permanent magnet synchronous generators and fully rated converters, have a range of control over both real power and reactive power for varying wind speeds.
The conversion of solar energy to electricity is currently accomplished mostly in two ways—by direct conversion using photovoltaics (PVs) or by solar thermal conversion. These are briefly described below.
In the direct-conversion method, PVs generate a direct current (DC) output that is converted to alternating current (AC). This conversion is achieved by a power electronic device called an inverter. Most PVs are rooftop units, and PV-based solar energy primarily has limited distribution and capacity. However, some large commercial PV-based solar facilities of up to 60 MW have been built recently.
The principal problems with the large-scale integration of PVs into the grid include limited capacity, high cost, low energy-conversion efficiency, and deteriorating performance as PV cells age.
Solar Thermal Conversion
In solar thermal conversion, the sun’s rays are directed by mirrors to heat a thermal exchange agent (e.g., mineral oil) to a sufficiently high temperature. This agent then exchanges the heat generated via a conventional steam cycle and runs a steam turbine that drives a synchronous generator.
The solar thermal method also has the capability of storing energy using a thermal phase-transition approach. This is commonly achieved by using molten salt to store heat for up to six hours; the stored heat is used to run a conventional steam cycle when energy from the sun is not available. Although solar thermal facilities have plant capacities in the range of several hundred MWs, they also require significant quantities of water for cooling and steam generation. Unfortunately, water resources are limited in many parts of the United States where solar insolation is plentiful.
Grid Interface with Renewable Resources
Wind farms are typically located in areas where wind resources are plentiful and can satisfy certain requirements (for details, see http://www.nrel.gov/gis/wind.html). Most onshore wind farms are located in rural areas where the transmission system voltages are typically in the range of 69 kV to 161 kV. The nominal terminal voltages at the wind turbines range in value from 575 V to 4,160 V, depending on the turbine ratings (Miller et al., 2005). The unit transformer at each wind turbine steps up the voltage and feeds power into a collector system that operates at voltages ranging from 12.5 kV to 34.5 kV. The high side node of the collector system is then connected to the main substation transformer for the wind farm, which again steps up the voltage to the desired level and connects the wind farm to the transmission system in the geographical vicinity.
Distributed PV resources with inverters produce AC output at the desired voltage. In residential neighborhoods, these would connect directly with the utility supply point to the residence. Utilities around the country have established standards for these connections to minimize the significant safety risks of the bi-directional flow in electricity in existing residential supply circuits if the customer sells power back to the utility (e.g., http://www.srpnet.com/electric/pdfx/gen_guidelines.pdf). Commercial PV units, which have similar interconnection requirements, would most likely interface with the distribution system at slightly higher voltage levels than residential PV units, depending on their ratings.
Central solar thermal resources have significantly higher ratings and would connect to the transmission grid at high voltage levels ranging from 230 kV to 345 kV.
Power System Planning
The increasing penetration of renewable resources will have a significant impact on the performance and reliability of the electricity grid. This is largely because of the variability of renewable resources and the lack of large-scale economical storage capability. This impact will be discussed with respect to planning and operation, primary functions related to grid performance and reliability.
Traditional planning for a power system and for expanding transmission functions has been undertaken in response to the needs of the transmission system based mainly on past and projected loading levels, which have traditionally been estimates of future demand. In the deregulated market, and in the present case of using different renewables (i.e., different in source and in temporal characteristics, as well as in geographic location), transmission planners must respond to the needs of power generators. In other words, planning to expand transmission may now be driven by the location and type of generation, rather than by the needs of the transmission system. To compare, traditional transmission planning processes are driven by loads and have a “bottom up” structure, whereas current transmission planning is driven more by generation needs.
The term “integrated system planning” (see Figure 3) refers to the inclusion of the temporal, stochastic, and voltage-level characteristics of generation sources in plans for system expansion. In addition, because renewable resources have characteristics that favor large-scale energy storage, the storage components must be included in the integrated system plan. In the real-world environment, federal and state projections and long-term plans and portfolios may also strongly influence the expansion of the transmission system. The factors that must be considered in planning for increased renewable energy transmission are briefly described below.
Scalability of Network Topology
Because a future power transmission and distribution network with a high percentage of renewables may have more generation sources than existing networks, scalability will be a significant factor. Planners will have to determine (1) the network topology best suited for this new scenario and (2) the effects on system performance and reliability of having a large number of spatially distributed generation sources.
Network topology will significantly impact total transmission losses, as well as performance of the overall network when subjected to disturbances. If the network has a very large number of power sources, the range of possible power-flow configurations will be enormous (Hecker et al., 2009). Although this will make the performance and reliability problems much more challenging, it will also provide opportunities for designing networks that can out-perform traditional networks.
The main objective for legacy transmission systems was to transmit power from relatively local (e.g., within a radius of about 500 kilometers [km]) generation sources to load centers. Under deregulation, this objective has migrated to much longer distances and much higher operating power levels (e.g., many hundreds of megawatts, perhaps > 1,000 MW, for 1,200 km or more). With the assumed renewable energy portfolio and the degree of variability from wind and solar sources, it will be critical for planners to take advantage of the geographical diversity among renewable resources. To facilitate a balance, a large high-voltage backbone may be necessary. This backbone network could consist of an interconnected transmission grid at voltage levels of 765 kV or greater that would provide the capability of moving a large amount of power from where it is generated to the locations where it will be used.
Because distributed resources may be widely dispersed and have diverse temporal characteristics, their operating level and transmission paths are a combination of high and low MW levels and long and short distances. At the 50 percent penetration level, one might expect transmission paths and transmission loading well into the hundreds of kilometers and hundreds of MW.
These changes in transmission topology to account for distributed resources can be facilitated, at least in part, in the following ways (Osborn and Zhou, 2008):
- Voltage-level and power-level upgrades would be made to existing high-voltage DC (HVDC) systems and/or new parallel HVDC systems to convert existing 12-pulse bipolar designs into 24-pulse bipolar designs. These changes could dramatically increase operational power levels and reliability and concomitantly decrease the impact of HVDC converters on power quality.
- Transmission routes would be determined after taking into account the intermittency of resources, load patterns, and available rights-of-way.
- The performance of planned routes under varying conditions would be evaluated, including the analysis of adequacy and reliability. The stochastic nature of the renewable resources would also have to be accounted for.
For optimal use, many renewable resources require energy storage. The following factors must be considered in designing optimal storage systems:
- type of storage (batteries; flywheels; superconducting magnetic energy storage systems; pumped hydro storage systems; compressed-air, molten-salt, fuel cells + hydrolyser; and other active and passive innovative systems)
- voltage level and power level at the point of interconnection
- energy-storage rating
- time duration and time profile of the charge/discharge cycle
- physical location of the storage device in the system (e.g., proximity to loads, sources)
- control objectives
- ownership and operator of the storage elements
- maintenance of the storage elements
- cost and efficiency (energy recovered/energy stored)
- availability and commercialization—including availability at a specific time in the future
Flexible Alternating Current Transmission System (FACTS)
A critical prerequisite for 50 percent penetration of renewable generation is dynamic control of power flow along optimal corridors in the transmission and networked distribution systems. This can be achieved with different types of high-power electronic controllers:
- centralized, large FACTS devices (Hingorani, 1993), especially unified power-flow controllers (UPFCs), which can control power flow in high-voltage AC transmission systems
- distributed electronic power-flow controllers, which are similar to FACTS devices but are highly distributed in the network and have high-frequency operation, lower power rating, and extensive real-time communication capability
- power-conversion devices for interfacing renewable resources (e.g., DFIGs for wind energy) that are supplementary to power-flow controls
Communications and Monitoring
The electric power grid is becoming an increasingly automated network and is expected to have increased functionality, higher efficiency, more programmability, and more flexibility. A variety of communication networks are interconnected to the electric grid for sensing, monitoring, and control. These communication networks are closely associated with the supervisory control and data acquisition (SCADA) systems in the network.
The data provided by the SCADA systems are used in the energy-management systems (EMS) for a wide range of systems-operation functions and real-time control of the power grid. The SCADA network and EMS are the main factors in the operation of the system under normal and emergency conditions. Any disturbance or dislocation in the network is sensed primarily by observations and analysis of the behavior of the system based on data obtained by the SCADA network.
The present method of securing the electric grid is real-time monitoring of the electrical behavior and performance of transmission lines. Wide area monitoring system (WAMS) technologies are key to increasing access to available maximum capacity of the transmission lines. WAMS provide real-time monitoring, which enables grid operators to determine precisely the operating margins of transmission lines while maintaining stability limits.
One WAMS technology that is increasingly being used is phasor measurement units (PMUs) (Phadke and Thorp, 2008), which are GPS-enabled sensors that take accurate measurements of grid conditions at strategic points in fine-grain time intervals (e.g., microseconds). GPS—time-stamped measurements (e.g., voltage and phase angle)—from multiple PMUs are gathered through a real-time communication network and used to conduct online security assessments of the grid. The PMU-based WAMS technologies, which effectively monitor the dynamic state of the grid, including voltage and angular stability and thermal limits, provide early warnings to network operators of imminent failures, stress, or potential instability, thus enabling them to take preventive action.
Interfaces between the Grid and Renewables
In light of the very wide range of capacity ratings for the renewable mix and the well diversified technologies used to integrate them into power grids, renewable sources can be categorized into concentrated energy resources (CERs) and distributed energy resources (DERs).
Concentrated Energy Resources. Among CERs, geothermal, biomass, and concentrated solar systems have conventional synchronous generators and steam prime movers. As a result, integration of these CERs is expected to be less challenging than for DERs. However, large-scale wind farms and large-scale PV systems present a spectrum of technical challenges that will require thorough investigation. The technical challenges arise mostly from the expanding application of electronic devices at high power ratings.
For instance, large wind turbines with power ratings of more than 1 MW nowadays commonly have DFIGs. A wind generation system with DFIGs requires an AC-DC-AC power converter rated at about 30 percent of the full power rating of the generator to achieve variable operation frequency within the range of ±30 percent of the nominal frequency of 60 Hz.
In addition, emerging direct-drive wind generation systems that use permanent magnet synchronous generators (PMSGs) are expected to prevail at the power rating of 3 to 5 MW, which is suitable for offshore wind farms. Systems with PMSGs require power converters that can handle the full power rating of the generator. For CERs with large-scale PV, the power electronic interface is indispensable because of the necessity of converting DC voltage generated by PV into the 60 Hz AC voltage of the grid.
Distributed Energy Resources. For DERs, namely distributed wind and PV generation systems, large numbers of small-scale generation sources are dispersed at the distribution level. Facilitating the integration of DERs will require microgrid and power management systems that transparently provide control and regulation.
Power System Operation
Power systems operate in a range of time frames from nearly real time to “operational real time” (i.e., a few seconds). Economic dispatch, that is, determining the most economical distribution of the committed generation outputs to meet a given pattern of load demand while accounting for system losses (Wood and Wollenberg, 1996), is performed in operational real time.
Figure 4 shows the main tools and controls for power system operation. Note that in the figure, several time frames are called out:
- SCADA: 2 to 4 seconds
- economic dispatch and automatic generation control (AGC): 2 to 10 seconds
- security and contingency analysis: 1 to 2 minutes
- state estimation: 1 to 5 minutes
Unit commitment, which has a time frame of one week or longer, is not shown in the figure because it is not usually considered an operational tool. The traditional unit commitment is the procedure by which the entire ensemble of generating units is examined to produce a subset of generators that satisfy the load, minimize operating costs (including start-up and shutdown costs), and satisfy a range of constraints, such as environmental impact mitigation, contractual limits, expected market power generation, and manpower limitations). The operational functions most impacted by the high penetration of renewable resources are unit commitment and economic dispatch.
The incorporation of renewable resources would significantly alter the traditional approach to unit commitment. The variability of renewable resources would require measures to accommodate fast generation (e.g., a few seconds) changes. The inclusion of storage devices would also alter unit commitment. Both features of renewable resources (i.e., variability and storage) would also alter economic dispatch. No fuel costs would be associated with the renewable energy resources, but increased operational and maintenance costs would be incurred and must be accounted for.
However, the most critical element would be the variability of renewable resources and accounting for sufficient commitment and dispatch of reserve generation to guarantee the reliability of the system in the event that the renewable resource suddenly becomes unavailable. For example, the wind might suddenly stop blowing, or the weather might become cloudy.
Variability is also closely tied in with automatic generation control to maintain system frequency. In power systems, electricity has to be produced to match the load on the system, and load patterns are highly variable. For example, a customer may switch on the TV and air conditioner and blend a smoothie almost simultaneously. The sudden increase in demand, however small, must be met by a concomitant increase in generation. If it is not, the system frequency will change, which will have an adverse effect on expensive power system components, as well as on customer-owned appliances. Hence, the system frequency has to be carefully controlled within tight tolerances.
Frequency control is achieved by providing control mechanisms that adjust the generation output to match the load. With the high degree of variability of renewable resources, either sufficient conventional generation would have to be maintained on active spinning reserve (i.e., be readily available) or sufficient energy storage would have to be provided to guarantee that load and generation remain in balance. Overall, the increasing penetration of variable renewable resources will require a re-examination of the economic dispatch/automatic generation control formulation and could require reevaluation of the limits of frequency variation in the system.
This article has highlighted the potential impact of increased penetration of renewable resources on the planning and operation of the bulk power system. Increased penetration of renewable resources has the potential to introduce major technological challenges that would have to be met to satisfy existing planning and reliability standards.
Hecker, L., Z. Zhou, D. Osborn, and J. Lawhorn. 2009. Value Based Transmission Planning Process for Joint Coordinated System Plan. Presented at the Power Systems Conference and Exposition, 2009. PES ‘09. IEEE/PES March 15–18, 2009, Page(s):1–5 Digital Object Identifier 10.1109/PSCE.2009.4840181.
Hingorani, N.G. 1993. Flexible AC transmission. IEEE Spectrum 30(4): 40–45. DOI 10.1109/6.206621.
Miller, N.W., W.W. Price, and J.J. Sanchez-Gasca. 2005. Modeling of GE Wind Turbine-Generators for Grid Studies, Version3.4b. Atlanta, Ga.: GE Energy.
Osborn, D., and Z. Zhou. 2008. Transmission Plan Based on Economic Studies. Transmission and Distribution Conference and Exposition, 2008. T&D. IEEE/PES, April 21–24, 2008. Page(s):1–4. DOI 10.1109/TDC.2008.4517276.
Phadke, A.G., and J.S. Thorp. 2008. Synchronized Phasor Measurements and Their Applications. New York: Springer Science+Business Media, LLC. DOI: 10.1007/978-0-387-76537-2_5.
Wood, A.J., and B.F. Wollenberg 1996. Power Generation Operation and Control, 2nd ed. New York: John Wiley and Sons, Inc.