4
AGC in Restructured Power Systems
As the world moves toward competitive markets in electric power systems, the shift of ownership and operational control of generation from the vertically integrated utilities to independent, for-profit generation owners has raised a number of fundamental questions regarding AGC systems. Key questions relate to the new AGC designs that are more appropriate to the new operational objectives of a restructured power network, including the revising of traditional control schemes and the AGC model by taking into account bilateral transactions.
After deregulation of the electricity sector, all reliability entities in the world, such as the North American Electric Reliability Council (NERC) and Union for the Coordination of Transmission of Electricity (UCTE), updated the control performance standards for AGC. The crucial role of AGC systems will continue in restructured power systems, with some modifications to account for some issues, such as bilateral contracts and deregulation policy among the control areas. In a real-time power market, AGC as an ancillary service provides an essential role for ensuring reliable operation by adjusting generation to minimize frequency deviations and regulate tie-line flows.
This chapter reviews the main structures, configurations, and characteristics of AGC systems in a deregulated environment. Section 4.1 addresses the control area concept in restructured power systems. Modern AGC structures and topologies are described in Section 4.2. A brief description of AGC markets is addressed in Section 4.3. Some concepts of the AGC market and market operator, the needs for intelligent AGC markets in the future, and also an updated conventional frequency response model concerning the bilateral transactions are explained in Section 4.4. Finally, the chapter is concluded in Section 4.5.
4.1 Control Area in New Environment
Most deregulated utilities have chosen to control the frequency and tie-line power to the same quality as before deregulation. The AGC schemes and control strategies have also mostly remained similar to before deregulation, except that some definitions are changed and services provided by participants are now classified as ancillary. However, the introduction of electricity markets has added to the pressures to redefine some concepts and to update the way that frequency/real power is controlled.
In an open energy market, generation companies (Gencos) as independent power utilities may or may not participate in the AGC task. On the other hand, distribution companies (Discos) may contract individually with Gencos, renewable energy plants, or independent power producers (IPPs) for power in different areas. Therefore, in the new environment, control is highly decentralized. Each load matching contract requires a separate control process, yet this process must cooperatively interact to reestablish system frequency and tie-line power interchange.1 In real-time markets, new organizations, market operators, and supervisors, such as independent system operators (ISOs), are responsible for maintaining the real-time balance of generation and load for minimizing frequency deviations and regulating tie-line flows, which would facilitate bilateral contracts spanning over various control areas.
In the new structure, there are no constant boundaries for control areas. The definition of a control area is somewhat determined by pooling arrangements and contract agreements of AGC participating utilities. The boundary of the control area encloses the Gencos, transmission company (Transco), and Discos associated with the performed contracts. In order to supply the load, the Discos can get power from Gencos directly or through Transco. Such a configuration is conceptually shown in Figure 4.1. The control areas are interconnected to each other, through either the Transco or Gencos.2
FIGURE 4.1
A (virtual) control area in a deregulated environment.
In a modern power system, the AGC system should track moment-to-moment fluctuations in the system load to meet the specified control area performance criteria, such as those criteria provided by NERC and UCTE. These criteria require the AGC system to maintain the area control error (ACE) within tight limits. All control areas in a multiarea power system are required to follow the determined control performance standards. The assigned control area performance criteria are measurable and in use for normal functions of each control area’s energy management system (EMS).
Currently, in many countries, electric systems are restructured; new market concepts were adopted to achieve the goal of better performance and efficiency. Operating the power system in a new environment will certainly be more complex than in the past, due to restructuring and a considerable degree of technical and economical interconnections. In addition to various market policies, numerous generator units in distribution areas and a growing number of independent players and renewable energy sources (RESs) are likely to impact on the operation and control of the power system.
The classical AGC scheme may not be as straightforward to use in a deregulated power system, which includes separate Gencos, RESs, IPPs, Discos, and Transcos in a competitive environment, as for vertically integrated utility structures. In response to the new challenges, novel modeling and control approaches are required to maintain reliability, to follow AGC tasks, and to get a new trade-off between efficiency and robustness.1 As the electric power industry moves toward full competition, various industry consensus on definitions, requirements, obligations, and management for AGC ancillary service is being developed by many entities across the world, such as NERC, UCTE, the Federal Energy Regulatory Commission (FERC), Oak Ridge National Laboratory (ORNL), etc.
4.2 AGC Configurations and Frameworks
4.2.1 AGC Configurations
The MW-frequency regulation issue in a multiarea power system is mainly referred to as frequency control, load following, and scheduling. The main difference between frequency control and load following issues is in the timescale over which these fluctuations occur. Frequency control responds to rapid load fluctuations (on the order of a few seconds to a minute), and load following responds to slower changes (on the order of a few minutes). While frequency regulation matches the generation with a seconds-to-minutes load change, load following uses the generation to meet minutes-to-hour and daily variations of load.3 These issues are addressed by governor systems, AGC, and economic dispatch mechanisms.4
In practice, AGC configurations could differ according to their timing, the amount of information individual suppliers and loads provide to the market operator, and the role of the market operator in facilitating or directing this ancillary service. A general configuration for the AGC system in a deregulated environment is shown in Figure 4.2. The Gencos send the bid regulating reserves to the AGC center through a secure network service. These bids are sorted by a prespecified time period and price. Then, the sorted regulating reserves with the demanded load from Discos, the tie-line data from Transco, and the area frequency are used to provide control commands to track the area load changes. The bids are checked and re-sorted according to the received congestion information (from Transco) and screening of available capacity (collected from Gencos). The control signal is transmitted to the Gencos once every one to few seconds, while the results of computing participation factors and load generation scheduling by the market operator (economic dispatch unit in the AGC center) are executed daily or every few hours.
FIGURE 4.2
General AGC configuration in a deregulated environment.
A general scheme for AGC participants in restructured power systems can be considered as shown in Figure 4.3. The Gencos (and many distributed power producers) would interact with the market operator by providing bids for the supporting AGC service. In fact, the market operator is responsible for trading the regulation power. Transcos also post their information regarding the availability and capability of transmission lines via a secure communication system. Discos submit their demand bids to the market operator to be matched with Genco’s bids while satisfying the performed regulation standards provided by reliability entities.
FIGURE 4.3
AGC participants in a deregulated environment.
In many market models, such as Poolco,5,6 since the AGC ancillary service auction is operated by the market operator, the market operator is the single-buyer party to meet the reliability obligations. The main market objective is to minimize AGC payments to Gencos while encouraging Gencos to provide sufficient regulation power. Gencos would anticipate submitting a bid that would maximize their profits as allocations are made. The AGC bids should include financial information for capacity reservation and energy, as well as operational information, such as location, ramp rate, and quantity blocks. Based on the central operator requirements for the AGC issue and participants’ bids submitted to the market operator, the price and quantity of regulation power are determined, and payments are calculated by the market operator.
The above-mentioned mechanism mostly deals with a centralized AGC market (usually called pool market), which is cleared by a unique market operator that collects the offers to sell and the bids to buy. On the other hand, in a decentralized AGC market (usually called exchanges market), sellers and buyers can enter directly into contracts to buy and sell. In this case, the transaction is of a bilateral type.
As mentioned, with a centralized structure, the market is cleared by a unique entity. In pool-energy-only markets, which are popular across the United States, participants provide bids and offers. Then, the system operator directly commits and dispatches the producers. Therefore, this approach favors links between AGC services and other products, such as energy. However, markets with a centralized structure are deemed to be opaque because the clearing process is quite complex. In addition, bidders have to provide a lot of information, and this system hardly takes into account all the variables of the system. In a real-time power market, the energy component of an ancillary services bid is used for energy balancing and ex-post pricing systems. Resources available in the energy balancing system include regulation, spinning, nonspinning, and replacement reserves, as well as resources for submitted supplemental bids for real-time imbalances that are pooled in the energy balancing system and arranged in merit order based on their energy bid prices.7 In the decentralized structure, which is popular across Europe, participants propose bids and select offers directly in the market. Therefore, a global co-optimization is difficult since participants buy and sell independently from each other. Instead, each participant does its own co-optimization with its assets.8
In defining an AGC market, a key factor is to attract enough regulation power producers to make the market competitive, while maintaining an acceptable level of security and reliability. If too much energy is traded close to real time, then the market operator must contract more reserves to ensure that the predicted demand can be met. One of the principles recommended in the FERC standard market design is that the design should ensure there are few incentives for a participant not to be in balance prior to real time.9 The AGC performance can also be significantly influenced by the deregulation policy,5,7,10 communication structure/facilities,11,12 and most importantly, the reserve levels.13
The market operator usually provides a priority list sorting by regulating price as described in Bevrani.1 The capacity of one regulating object in the table should be fully utilized before calling on the next, which is more expensive. This applies to both cost and the market environment. In essence, this would mean that at any given time the cheapest solution should be in place. However, this is often far from the reality, and due to many reasons, the economic solution is not what it should be. Reserve levels also need to be considered, as a cheap generator might have its output reduced to ensure sufficient reserve levels. The AGC algorithm needs to be set up so that an expensive generator decreases and a cheap generator increases its regulation power, simultaneously.14
4.2.2 AGC Frameworks
In the real world, different AGC frameworks/schemes are available to perform supplementary control among different countries/regions. The AGC scheme that has been implemented in some countries differs from the design adopted in most other parts of the world. Considerable differences exist between the AGC characteristics, reserve service, topology, and related standards defined in various jurisdictions. This diversity is the source of some confusion because the diversity is extended not only to the specification of existing AGC systems, but also to the terms used to describe them. Below, the general AGC framework provided by UCTE is briefly explained. In the UCTE terminology, instead of AGC, the secondary control or load-frequency control (LFC) term is used.
According to the UCTE definitions,15 a control area is the smallest portion of a power system equipped with an autonomous AGC system. A control block may be formed by one or more control areas working together to satisfy predefined AGC performance requirements with respect to the neighboring control blocks, within a synchronous area connected to the UCTE network.
Within the synchronous area, the control actions and reserves are organized in a hierarchical structure with control areas, and control blocks with a coordination center. The AGC, the technical reserves, and the corresponding control performances are essential to allow transmission system operators (TSOs) or block coordinators to perform daily operational business.
As shown in Figure 4.4, the synchronous area consists of multiple interconnected control areas/blocks, each of them with a centralized supplementary control loop. Each control area/block may divide up into subcontrol areas that operate their own underlying AGC, as long as this does not jeopardize the interconnected operation. Figure 4.4 shows the hierarchy of an AGC that consists of the synchronous area, with control blocks and (optionally) included control areas.
FIGURE 4.4
A UCTE synchronous area.
If a control block has internal control areas, the control block organizes the internal frequency regulation according to one of the centralized, pluralistic, and hierarchical schemes.15 In a centralized scheme, the AGC for the control block is performed centrally by a single controller (the control block including only one control area); the operator of the control block has the same responsibilities as the operator of a control area. This scheme is currently in use in some UCTE countries, such as Italy, Austria, Belgium, and the Netherlands.
In a pluralistic scheme, the AGC is performed in a decentralized way with more than one control area; a single TSO (block coordinator) controls the whole block toward its neighbors, with its own supplementary loop and regulating capacity, while all the other TSOs of the block regulate their own control areas in a decentralized way by their own. This scheme exists in France, Spain, and Portugal, with France operating as the block coordinator.
Finally, in a hierarchical scheme the AGC is performed in a decentralized way with more than one control area. The AGC is carried out by several separate supplementary control loops, one for each control area within the hierarchical control block. They are separately controlling their cross-border exchanges. But at a higher control level, a single TSO (block coordinator) operates the superposed block controller that directly influences the subordinate supplementary loops of all control areas of the control block; the block coordinator may or may not have regulating capacity on its own. Switzerland represents an application of this AGC scheme.
In particular, most European power systems in UCTE use one of the aforementioned control schemes. However, there are important differences among them in details, and some AGC schemes exhibit some important differences with respect to the standard structures.16–20
4.3 AGC Markets
After the advent of deregulation, there was much effort to form competitive markets for ancillary services. Currently in many countries, similar to available competitive markets, some markets exist for ancillary services, such as the AGC market, with different structures and even titles, depending on the rules and regulations of the area. However, since a Genco can make the choice to allocate 1 MW of production capacity as energy or an AGC service, markets for AGC services and energy are closely linked. Furthermore, AGC markets as well as energy markets are highly influenced by other markets and commodities, such as fuel and environment markets.
Systems across the world have adopted different methods to calculate the needs for regulation services, which leads to different types of AGC markets. In North America, regulation reserve markets for AGC with fully dispatchable regulation power capacity within 10 min are available. In some regions, such as England (without an official AGC system), AGC market is summarized to a part of the spinning reserve as a power exchange system. It provides a system with a 30 min short-term market for balancing, operating 1 h ahead of real time. However, in many countries, such as Japan, Australia, China, the Nordic countries, and continental Europe, the AGC markets and frequency performance standards are strongly influenced by grid rules and regional policies.21
Flat rate, price based, and response based are three famous AGC markets.22 Flat rate, because of simplicity, is the most common type of AGC market (specifically in North America and North Africa). It provides a 10 min regulation market with a uniform price payment at the rate of the market clearing price (MCP), without considering the ramp rates and response quality of the participant generating units. A price-based AGC market provides a 5 or 10 min regulation market, and the participant generating units are paid based on their ramp rate performance. Finally, the price-based AGC market provides two separate markets for fast ramp regulation (5 min market) and slow ramp regulation (10 min market), and the AGC participant generating units are paid at the rate of MCP as the maximum possible payment available in each auction market.
It has been demonstrated that in comparison to the flat-rate AGC market, the price-based and ramp-rate-based AGC markets increase competition, encourage participant generators with incentive, and explore more control options to optimize AGC performance. The separation of fast and slow ramp generators in the response-based AGC market makes it possible for this market to call upon the appropriate service, depending on the magnitude of the disturbance that is suitable for contingencies and related power reserves. However, in this case (separate markets for fast ramp and slow ramp regulation), making a decision is more difficult. To procure a certain amount of regulation from such a market, the market operator has to decide how much of the fast and slow regulation powers are to be bought. Then there are multiple options available as to how to use them in time of need.22
All the generators participating in the AGC markets mentioned above are required to meet specified technical and operating requirements, and also, they should determine regulation capacity, price, and operational ramp rate (MW/min) in their bids. The AGC markets are usually cleared for every dispatch interval during the trading interval ahead of real time.23
In the AGC markets, the structures of bids are related to the scoring and clearing processes, while the structures of payments are related to the settlement process. In the literature, most of the discussions on structures of offers and payment of AGC services are concentrated on capacity, utilization, and opportunity cost components.24–27 The more common structures for offers and payments are known as (1) a fixed allowance, (2) an availability price, (3) a price for kinetic energy, (4) a utilization payment, (5) a utilization frequency payment, and (6) a payment for the opportunity cost.8 A fixed allowance is paid to the provider in every instance. An availability price is paid only when the unit is in a ready-to-provide state. A price for kinetic energy remunerates the quantity of kinetic energy made available to the system. It recognizes the machines with high kinetic energy, and thus high inertia to shape the rate of frequency change following a contingency. A utilization payment remunerates the actual delivery of the service. A utilization frequency payment is based on the number of calls to provide a service over a given period of time. It thus reflects the extra costs that may be incurred each time the service is called upon. The payment of the opportunity cost has been identified for a long time by the community as an important allowance.
Coordination of the AGC market with other ancillary and energy markets is also an important problem. Since a generating unit may provide several ancillary services (including AGC), and thus contribute to several markets, coordination between different markets, in both quantity and price issues, is required. For example,8 if a generator provides reserves for an AGC system, it cannot sell all its capacity on the only energy market. On the other hand, if a generator is committed through energy dispatch, it is then able to provide reactive power support for voltage control service. A direct consequence of this feature is that the prices of ancillary services and electrical energy will interact.
4.4 AGC Response and an Updated Model
4.4.1 AGC System and Market Operator
As mentioned, the main objectives of an AGC system are to maintain the frequency within control areas close to the nominal value, as well as to control tie-line flows at scheduled values defined by utilities’ contracts. Similar to the conventional AGC system, the balance between generation and load can be achieved by detecting frequency and tie-line flow deviations via ACE signal through an integral feedback control mechanism. If supply and demand do not match in the long run, as well as in the short run, the market will fail. The supply of AGC services is mostly ensured by conventional generating units. Marginally, other participants also provide regulation services, such as storage devices that smooth either consumption or generation, consumers that can modulate their consumption upon request or automatically, and to some extent, RESs. The demand for AGC services is defined by the market operator and depends on the power system structure.
As explained in Chapter 2, the generating units could respond to fast load fluctuations, on a timescale of 1 to 3 s, depending on the droop characteristics of governors in the primary frequency control loop. The generating units could respond to slower disturbance dynamics in the range of a few seconds, measuring the ACE signal via a supplementary frequency control loop in the AGC system. The longer-term load changes on a timescale of 10 s to several minutes could respond based on economic dispatch plans and special control actions that would utilize the economics of the AGC system to minimize operating costs.
The deviations in load and power could be procured by the market operator on purpose, because of planned line and unit outages. This kind of deviation may be produced by the market operator as a control plan in response to energy imbalances following unpredicted disturbances. These deviations are basically different than unpredicted frequency/tie-line deviations that usually occur by variations of load and generation from scheduled levels following a fault, such as unplanned line and unit outages. The AGC participant generating units in an AGC market could respond to unpredicted frequency/tie-line deviations proportional to the assigned participation factors from their schedules within a few seconds. The market operator will change the set points of AGC units, which have submitted energy price/quantity bids for the real-time energy imbalances, by means of a new control plan, as shown in Figure 4.5.
Determining AGC participation factors by market operator is an important issue in deregulated environments. For this purpose, several factors, such as regulation price, ramp rate, and bid capacity provided by the candidate generating units, should be considered. The impacts of these factors on AGC performance and system frequency response characteristics, including maximum frequency deviation, the time taken to bring the frequency back within safe limits, and the time taken by ACE to cross zero for the first time following the disturbance, are studied in PSERC.22 It has been shown that the frequency response is better when the participation factors are proportional to the units’ ramp rate.
FIGURE 4.5
The AGC–market operator loop.
The market operator may procure the required power regulation from various existing reserves, such as normal AGC regulation, spinning reserves that are usually available within 10 min, nonspinning reserves, and replacement reserves. In this process, the participant Gencos would be allowed to rebid their uncommitted resources and regulation powers at new prices. The market operator, which is responsible for AGC procurement, can use various methods to obtain AGC services. Some methods are known as compulsory provision, bilateral contracts, tendering, self-procurement, and spot market. These methods are defined in Rebours.8 Various factors, such as market concentration, mode of energy and transmission trading, risk aversion, the costs recovery method, and centralized or decentralized AGC control, influence the choice of one of these methods over the others.
In addition to the quality and quantity of AGC services, the location is also important. Although frequency control ancillary services act on global frequency, their physical locations should be considered while procuring AGC services for some reason, mostly with the security and reliability issues.8
Congestion of transmission lines is an important reason that can affect the reliable provision of AGC services. If enough transmission capacity is not available, the affected zone has to secure enough ancillary services from within its perimeter. Therefore, a part of the transmission capacity has to be allocated for AGC services. This congestion of transmission lines and overloading is more important in the presence of contingencies and emergency conditions, and it should be carefully considered for performing emergency control actions.28,29
Regarding the high cost of reserving transmission capacity, the contributions to the AGC are likely to be distributed across the whole power system network to reduce unplanned power transits following a large generation outage.8 A distributed framework for AGC services can also be useful following the islanding issue. Islands cannot stay stable without any frequency control system service.
In fact, trading AGC services in a distributed framework across systems allows more efficient use of flexible resources, reduces the potential exercise of market power, diminishes imbalance exposure, and makes better use of interconnection capabilities.8,30 A necessary condition to perform such a useful framework is using a distributed generation scheme across the whole interconnected network.
In a competitive environment with a decentralized market structure, a Disco has the freedom to contract with any Genco in its own area or sign bilateral contracts with a Genco in another area that would be cleared by the market operator. If a bilateral contract exists between Discos in one control area and Gencos in other control areas, the scheduled flow on a tie-line between two control areas must exactly match the net sum of the contracts that exist between market participants on opposite sides of the tie-line. If the bilateral contract is adjusted, the scheduled tie-line flow must be adjusted accordingly. In general, using bilateral contracts, Discos would correspond demands to Gencos, which would introduce new signals that did not exist in the vertically integrated environment. These signals would give information as to which Genco ought to follow which Disco. Moreover, these signals would provide information on scheduled tie-line flow adjustments and ACEs for control areas.7
In a competitive electricity environment, Poolco and bilateral transactions may take place simultaneously. As already mentioned, in Poolco-based transactions, the power generating units and consumers submit their bids to the market operator, and market players quote a price and quantity for upward and downward adjustment. For each time period of operation, generators’ bids are selected based on the principle of the cheapest bid first for upward regulation and the most expensive bid first for the downward regulation. In addition, during the low-frequency conditions, consumers having interruptible loads may also be selected based on the cheapest bid first. The resultant price/quantity list is used for achieving the balance between consumption and production.
The AGC in a deregulated electricity market should be designed to consider different types of possible transactions,5,31,32 such as Poolco-based transactions, bilateral transactions, and their combination. Over the last years, some published works have addressed the updating of the traditional AGC model and the redesign of conventional control schemes to accommodate bilateral transactions.31–34 An AGC scheme required for Poolco-based transactions, utilizing an integral controller, has been suggested in the literature.5,31,32,35
4.4.2 AGC Model and Bilateral Contracts
Using the idea presented in Donde et al.31 the well-known AGC frequency response model (Figure 2.11) can be updated for a given control area in a deregulated environment with bilateral transactions. The result is shown in Figure 4.6. This model uses all the information required in a vertically operated utility industry plus the contract data information.
The overall power system structure can be considered as a collection of Discos or control areas interconnected through high-voltage transmission lines or tie-lines. Each control area has its own AGC and is responsible for tracking its own load and honoring tie-line power exchange contracts with its neighbors. There can be various combinations of contracts between each Disco and available Gencos. On the other hand, each Genco can contract with various Discos. Therefore, a Disco in any of the areas and Gencos in the same or in a different area may also negotiate bilateral contracts. These players of the electricity market are responsible for having a communication path to exchange contract data, as well as measurements to perform the load following function. In such contracts, a Genco changes its power output to follow the predicted load as long as it does not exceed the contracted value.
The generation participation matrix (GPM) concept is defined to express these bilateral contracts in the generalized model.1 GPM shows the participation factor of each Genco in the considered control areas, and each control area is determined by a Disco. The rows of a GPM correspond to Gencos, and columns to control areas that contract power. For example, for a large-scale power system with m control areas (Discos) and n Gencos, the GPM will have the following structure, where gpfij refers to the generation participation factor and shows the participation factor of Genco i in the load following of area j (based on a specified bilateral contract):
FIGURE 4.6
An updated AGC response model for deregulated environments.
New information signals due to possible various contracts between Disco i and other Discos and Gencos are shown as wide arrows. Here, we can write:
where
where ΔPdi (in Figure 4.6) is the area load disturbance, ΔPLoc-i is the contracted load demand (contracted and uncontracted) in area i, and ΔPtie-i, actual is the actual tie-line power in area i. Using Equation 4.2, the scheduled tie-line power (ΔPtie-i, scheduled) can be calculated as follows:
Interested readers can find more details and simulation results on the above generalized AGC scheme for restructured power systems in the literature.1,32,36
4.4.3 Need for Intelligent AGC Markets
The AGC markets of tomorrow, which should handle complex multiobjective regulation optimization problems characterized by a high degree of diversification in policies, control strategies, and wide distribution in demand and supply sources, must be intelligent. The core of such an intelligent system should be based on flexible intelligent algorithms, advanced information technology (IT), and fast communication devices.
The intelligent AGC market interacting with ancillary services and energy markets will be able to contribute to upcoming challenges of future power systems control and operation. This issue will be performed by intelligent meters and data analyzers using advanced computational methods and hardware technologies in both load and generation sides. The future AGC requires increased intelligence and flexibility to ensure that they are capable of maintaining a supply-load balance following serious disturbances.
4.5 Summary
During the last two decades, energy regulatory policies all around the world have been characterized by the introduction of competition in many electric power systems. The AGC issue as an ancillary service represents an important role to maintain an acceptable level of efficiency, quality, and reliability in a deregulated power system environment. To this aim, researchers and responsible organizations have started to analyze possible new AGC schemes and regulation solutions, with paradigms suited for the energy market scenarios. These new solutions can rely on recent advances in IT, artificial intelligent methodologies, and innovations in control system theory.
This chapter has emphasized that the new challenges will require some adaptations of the current AGC strategies to satisfy the general needs of the different market organizations and the specific characteristics of each power system. The existing market-based AGC configurations and new concepts were briefly discussed, and an updated frequency response model for decentralized AGC markets was introduced.
References
1. H. Bevrani. 2009. Robust power system frequency control. New York: Springer.
2. H. Bevrani. 2004. Decentralized robust load-frequency control synthesis in restructured power systems. PhD dissertation, Osaka University.
3. E. Hirst, B. Kirby. 1999. Separating and measuring the regulation and load-following ancillary services. Utilities Policy 8(2):75–81.
4. Power Systems Engineering Research Center (PSERC). 2009. Impact of increased DFIG wind penetration on power systems and markets. Final project report. PSERC, Arizona State University, Phoeniz, AZ, USA.
5. J. Kumar, N. G. K. Hoe, G. B. Sheble. 1997. AGC simulator for price-based operation. Part I. A model. IEEE Trans. Power Syst. 2(12):527–32.
6. J. Kumar, N. G. K. Hoe, G. B. Sheble. 1997. AGC simulator for price-based operation. Part II. Case study results. IEEE Trans. Power Syst. 2(12):533–38.
7. M. Shahidehpour, H. Yamin, Z. Li. 2002. Market operations in electric power systems: Forecasting, scheduling, and risk management. New York: John Wiley & Sons.
8. Y. Rebours. 2008. A comprehensive assessment of markets for frequency and voltage control ancillary services. PhD dissertation, University of Manchester.
9. W. Hogan. 2002. Electricity market design and structure: Working paper on standardized transmission services and wholesale market design. Washington, DC. http://www.ferc.fed.us.
10. R. D. Chritie, A. Bose. 1996. Load frequency control issues in power system operation after deregulation. IEEE Trans. Power Syst. 11(3):1191–200.
11. S. Bhowmik, K. Tomsovic, A. Bose. 2004. Communication models for third party load frequency control. IEEE Trans. Power Syst. 19(1):543–48.
12. H. Bindner, O. Gehrke. 2009. System control and communication. Risø Energy Report 8, 39–42. http://130.226.56.153/rispubl/reports/ris-r-1695.pdf.
13. Y. Rebours, D. Kirschen. 2005. A survey of definitions and specifications of reserve services. Technical report, University of Manchester. http://www.eee.manchester.ac.uk/research/groups/eeps/publications/reportstheses/aoe/rebours%20et%20al_tech%20rep_2005B.pdf.
14. G. A. Chown, B. Wigdorowitz. 2004. A methodology for the redesign of frequency control for AC networks. IEEE Trans. Power Syst. 19(3):1546–54.
15. UCTE. 2009. UCTE operation handbook. http://www.ucte.org.
16. B. Delfino, F. Fornari, S. Massucco. 2002. Load-frequency control and inadvertent interchange evaluation in restructured power systems. IEE Proc. Gener. Transm. Distrib. 149(5):607–14.
17. G. Dellolio, M. Sforna, C. Bruno, M. Pozzi. 2005. A pluralistic LFC scheme for online resolution of power congestions between market zones. IEEE Trans. Power Syst. 20(4):2070–77.
18. I. Egido, F. Fernandez-Bernal, L. Rouco. 2009. The Spanish AGC system: Description and analysis. IEEE Trans. Power Syst. 24(1):271–78.
19. L. Olmos, J. I. Fuente, J. L. Z. Macho, R. R. Pecharroman, A. M. Calmarza, J. Moreno. 2004. New design for the Spanish AGC scheme using an adaptive gain controller. IEEE Trans. Power Syst. 19(3):1528–37.
20. N. Maruejouls, T. Margotin, M. Trotignon, P. L. Dupuis, J. M. Tesseron. 2000. Measurement of the load frequency control system service: Comparison between American and European indicators. IEEE Trans. Power Syst. 15(4):1382–87.
21. I. Arnott, G. Chown, K. Lindstrom, M. Power, A. Bose, O. Gjerde, R. Morfill, N. Singh. 2003. Frequency control practices in market environments. In Quality and Security of Electric Power Delivery Systems 2003, CIGRE/IEEE PES International Symposium, Montreal, QC, ON, 143–48.
22. PSERC. 2008. Agent modelling for integrated power systems. Project report. http://www.pserc.org.
23. K. Bhattacharya, M. H. J. Bollen, J. E. Daalder. 2001. Operation of restructured power systems. Boston: Kluwer Academic Publishers.
24. H. Singh. 1999. Auctions for ancillary services. Decision Support Systems 24(3–4):183–91.
25. H.P. Chao, R. Wilson. 2002. Multi-dimensional procurement auctions for power reserves: Robust incentive-compatible scoring and settlement rules. J. Regulatory Econ. 22(2):161–83.
26. G. Chicco, G. Gross. 2004. Competitive acquisition of prioritizable capacity-based ancillary services. IEEE Trans. Power Syst. 19(1):569–76.
27. F. D. Galiana, F. Bouffard, J. M. Arroyo, J. F. Restrepo. 2005. Scheduling and pricing of coupled energy and primary, secondary, and tertiary reserves. Proc. IEEE 93(11):1970–83.
28. J. J. Ford, H. Bevrani, G. Ledwich. 2009. Adaptive load shedding and regional protection. Int. J. Elect. Power Energy Syst. 31:611–18.
29. H. Bevrani, G. Ledwich, J. J. Ford, Z. Y. Dong. 2008. On power system frequency control in emergency conditions. J. Elect. Eng. Technol. 3(4):499–508.
30. Frontier Economics and Consentec. 2005. Benefits and practical steps towards the integration of intraday electricity markets and balancing mechanisms. London: Frontier Economics Ltd. http://europa.eu.int/comm/energy/electricity/publications/doc/frontier_consentec_balancing_dec_2005.pdf.
31. V. Donde, M. A. Pai, I. A. Hiskens. 2001. Simulation and optimization in a AGC system after deregulation. IEEE Trans. Power Syst. 16(3):481–89.
32. H. Bevrani, Y. Mitani, K. Tsuji. 2004. Robust AGC: Traditional structure versus restructured scheme. IEEJ Trans. Power Energy 124-B(5):751–61.
33. H. Bevrani, Y. Mitani, K. Tsuji, H. Bevrani. 2005. Bilateral-based robust load-frequency control. Energy Conversion Management 46:1129–46.
34. PSERC. 2005. New system control methodologies: Adapting AGC and other generator controls to the restructured environment. Project report. http://www.pserc.org.
35. J. M. Arroyo, A. J. Conejo. 2002. Optimal response of a power generator to energy, AGC, and reserve pool-based markets. IEEE Trans. Power Syst. 17(2):404–10.
36. H. Bevrani, Y. Mitani, K. Tsuji. 2004. Robust decentralized AGC in a restructured power system. Energy Conversion Management 45:2297–312.