Chapter 2 Transmission Grid Voltage Control 2

Chapter 2
Transmission Grid Voltage Control
2.1 Introduction
Load frequency control and automatic voltage control in electrical power grids have always been considered the main regulating functionalities 1–12. Frequency control is tied with active power control, worked out and settled first and foremost because it relates more to the power system energy trade, the physical running speed of generators and the cost of energy to consumers.
Voltage control problems were dedicated mainly to system operators ever since the time of the first power systems, but the efforts to find a solution based this problem not been sufficient, and so a reliable and standard solution has not yet been achieved. A full understanding of voltage problems and what are thought to be their reliable solutions varies widely among cultures and countries. In addition, there are differences in the practical ways voltage is controlled in the field by individual utilities, methods that are generally inadequate and ineffective for satisfying real voltage needs. There are many reasons for this deficiency:
• Based on a theoretical view, static analyzing of grid voltage isn’t capable enough to the task of linking study results to real grid performance. In fact, only recently have engineers come to a general consensus on what constitutes a proper dynamic analysis of voltage instability. Up until now, such uncertainty has made it difficult for system operators to trust theoretical studies on system or static simulation result;
• Lack of software package tools in the past to enable dynamic analysis with high reliability, despite their more recent availability for large power systems, with tested dynamic models including operating on-field controls and protections. Nowadays, modern simulation tools based many software applications enables a system operator to study and rebuild the links between voltage and reactive power of power grid;
• The difficulty of the voltage control, which requires, the regulation of all power system buses compared to the simplicity of frequency regulation which is considered a single variable;
• Complexity from the operators side to detect the values of voltage magnitude of all bus bars due to lack of real time information in addition to unavailavlity to track the power system dynamics to reach an optimal solution or even accepted one. Also the complexity to take the decision of tap changing of the tranformers due to uncertainty for the effect of doing so on the power system stability. Also a lack of knowledge of a real time reactive power reserve and consumption.
• lack of proposals to improve voltage stability by manufacturers who consider problems only, at a given bus. Such proposals often include expensive equipment controls. Conversely, voltage problems often need a management from all the stakeholders and a plan to achieve optimal solution based on the all the voltage buses in a grid;
• The unavailability at control centers of continuous system voltage regulation, which should replace the less effective manual dispatching, always insufficient to face on-time real voltage problems. In fact, only large voltage variations can be recognized and managed in practice by manual recovery controls. From this perspective the distinction between continuous regulating control and stepping protection fades. In fact, often the system operator’s understanding of “voltage control” is so confused that he assumes it to be only voltage protection control;
• Previous absence of sufficient information technology (IT), Wide area measurement systems (WAMS) and SCADA/EMS control systems, which nowadays highly support most TSO control centers, therefore providing in real time most of the information sent from the site, including bus voltages and available reactive power reserves;
• Absence of an existing SCADA/EMS system with available control functionality for real-time continuous, reliable, optimal and fast control of grid voltages.
The above points didn’t cover all the possible reasons voltage control problems are not adequately met today. The main reason is surely linked to the old way system operators follow of manual control; for this reason, they are often working away from getting the importance of innovative automatic solutions and promoting them in practice, despite their high concern about voltage problems. Often, their manual intervention comes too late, that is, in extreme voltage conditions, when uncoordinated control risks failing. Despite that, application of grid voltage control improvement through technology innovation is a big research topic.
The problem of effective and automatic voltage and reactive power control in large and complex grids has been seriously considered since 1980; its solution demands realizing and defining of sophisticated control strategy able to reach optimal, reliable and secure system. Utility and transmission system operators (TSO) are, on the one hand, certainly interested in enlarging the reliability, security and quality of supply with an effective solution that has a minimum impact on investment costs. However, due to the novelty of this area, utility companies and TSOs are monitoring one another to gain an advantage, each learning from its competitors’ on-field experiences before making its own investment. Unfortunately, this approach is too prudent and often stalls decisions and delays the application of advanced and currently available wide area control solutions.
Voltage-reactive power control is indispensable in power systems that operate under normal or emergency conditions. During normal operation power/voltage control ensures the transmission of electrical energy at the required voltage quality and in conditions that are most convenient for suppliers and users. In emergencies the role of voltage control is to increase system security by enlarging the margin with respect to system voltage instability limits, thus ensuring continuity in system operation and proper operating conditions for the largest number of consumers.
Voltage regulation and reactive power compensation problems generally require a different approach whether we consider transmission or distribution level. At transmission level, the high voltage (HV) network can benefit from voltage-reactive power support that is provided by the largest alternators, through which the whole grid is fully controlled. At the distribution level, voltage control generally concerns independent, individual distribution areas: each area represents a small, separate part of the overall distribution system.

2.2 Structure of Voltage Control Hierarchy
Traditional power systems based on HV grid is partitioned into regions and on automatic managing of each region’s reactive power generating element aiming to control region bus voltages have been established first in Europe mainly Italy and France since 1980. These systems are called either coordinated voltage control, to focus on the required managment among region control sources, or they are also called secondary SecVC and TerVC, to emphasize the full image of the regulation levels. experience implementation, research and applications applied in Italy and France, followed by Belgium, Spain and more recently by United States, Brazil, Taiwan, South Korea, Romania and South Africa. An international CIGRE task force studied to investigate the idea and after that published an extensive report. After 1990, based on experimental applications in voltage wide-area control systems, certain European countries (mainly Italy and France) concluded general applications based wide-area voltage control systems as their actual grids. those projects implemented and studied for many years, for several considerations: its novelty; SCADA linked updates at dispatch control centers; Different formation of production and transmission companies; and the great interest at the beginning of 2000 of energy market rules despite the revolution of voltage and frequency regulation. With changing utilities organizing structure, and under the competition of the energy market abilities, TerVc, SecVC and PriVC are becoming more importany and highly required, mainly in the countries which already include them, but elsewhere, too, as research and capacity building regarding them increases. Global Power system engineers, in fact, knew that both SecVC and TerVC enable simplifying of automation complexity of transmission grid voltages overall by improving overall grid efficiency and security and the distinction of the contributions of different participants to voltage ancillary service in correct, simplified ways.
The focus on the improvement and application of transmission grid voltage control became highly required, at the beginning of 2000’s and under the concern of global smarting the electrical Grids dream, an important step towards this dream is directed mostly from “manual control” to innovative “automatic control”, through simple, effective closed-loop regulating systems, managed and supervised directly by electrical grid dispatching centers. Cost/benefit analyses strongly encouraged this trend. Moreover, because regulating voltage is mainly a local problem, any proposed ideas must consider automation in terms of managing local reactive power resources, mainly those of generators, static condensers, static inductors and different types offacts, the objectives of presenting optimal voltage service during different operating conditions can be accomplished through a decentralized voltage regulating strategy which manages resources at each grid zone. Such coordination requires exchanging data between regional control center and power stations or substations. The best policy was understood: “The more exchange there is of real-time electrical data in accordance with power system dynamics, the more improvement there will be in a voltage control system’s performance and effectiveness”.
The benefit of grid voltage control is high grid efficiency which becomes strongly related to system with different regions managment rather than source local control, requiring effective exchanging data among regional (either geographical or virtual) centers and the national system control center. In addition to exchanging measurements with surrounding grids (e.g., node-bus voltage magnitudes and connected lines or cables reactive power consumpation) and coordination of mutual control signals is also very important to minimize the power losses. Within the strategy of power sector liberalization and high quality service market competition, on-line and wide area monitoring of the performance of actual HV control strategy also represents a challenge to properly and optimally detect generating unit’s sharing to reach high voltage stability 26. Definite improvements that result from coordinated “automatic” real-time voltage regulation can then be summarised as follows:
• High quality-based power grid operation, in terms of reduced variations around the defined voltages profile across the overall grid;
• High security-based power system operation, in terms of larger reactive power reserves kept available by generating units for facing emergency conditions;
• Transfer capability of the power system is improved, in terms of increased active power levels transmissible, with reduced risk of voltage instability and collapse29;
• Efficient power network operation, in terms of active loss minimization, minimization of reactive flows and better management of reactive resources;
• Simplification of controllability and measurability-based voltage quality service, in terms of power system performance measurement criteria
Voltage and reactive power control of a network requires geographical and temporal coordination of many on-field components and control functions achievable by a hierarchical control structure. A real-time and automatic voltage control system can, in fact, be basically structured in three hierarchical levels: primary (component control), secondary (area control) and tertiary (power system control and optimization) levels.
Figure 2.1 gives a preliminary spatial view of the three overlapping hierarchical levels of a voltage -reactive control system. It also shows the interaction of the tertiary level with the not-real-time and off-line forecasting level based on state estimation and optimal power flow (OPF). This scheme distinguishes real-time levels with automatic closed-loop voltage and reactive power controls from day-before or short term optimal forecasting computation (necessarily delayed with respect to real-time power system operating conditions). In doing so, it offers clarity that helps us recognize relevant differences between real-time and forecasting levels.
It often happens that the tertiary voltage (closed-loop) control is confused with the static optimization problem of voltage-reactive power, which must be considered (due to its long delay with respect to a system’s operating conditions) as open-loop control or off-line forecasting related to system operation scheduling.
The most commonly employed OPF objective function is power system loss minimization, which forecasts, by the use of not-real-time data, the generators’ reactive power scheduling in order to maintain appropriate voltage levels within a power system’s normal operating limits. Obviously, higher performances are obtained with an automatic closed-loop voltage control that minimizes losses in real time (such as by TerVC) in comparison with a system operation based simply on a forecast computation linked to past working points (e.g., by OPF). In fact, a present grid operating condition could be, at times, very different from a forecasted one, mainly during critical operating conditions or in case of a great delay in the computation of a reliable state estimation.

Fig. 2.1 Voltage control Hierarchy implementation
The non-online OPF voltage-reactive power issue is, in any event, a useful input reference for TerVC computing, as shown in Fig. 2.1 and later fully described. In Fig. 2.1, the SecVC level is subdivided into two parts: the decentralized voltage control in the system regions (SecVC) overlaps the power station layer (secondary reactive power regulation), which controls the rotating generators, SVC, STATCOM and UPFC reactive powers. A dispatcher can interfere with the main control levels, mostly with not-real-time levels. He can also switch off the TerVC and manually define SecVC pilot node voltage set-point values, but in this case, it renounces on-line real time system optimization as well as the stability benefits deriving from TerVC. On the contrary, dispatcher manual control inside the SecVC level is to be avoided and is very dangerous for system security due to the criticality of the manual reactive power control at the high control speed provided by SecVC. In other words, SecVc should be fully automatic, while a manual TerVC can be managed by the dispatcher’s operator who, so doing, renounces the high reliability and efficiency that automatic TerVC provides.
2.3 Primary Voltage Control (AVR or PriVC)
Primary regulation aims controlling the terminal voltage of synchronous generators and compensators with the major objective of allowing correct and secure operation of the equipment. Obviously, primary voltage control impacts the transmission network, mainly at the MV buses to which generators and compensators are connected, by sustaining the local medium voltages during normal and perturbed operating conditions. Control actions are based on station measurements and aimed to bring out the voltage at the reference desired value automatically, with a dynamic performance characterized by a dominant time constant value within a milliseconds up to one second: this compensating control must be considered a high-speed local voltage regulation.
An automatic voltage regulator (AVR) realizes the primary voltage control of a generator (in Fig. 2.2 the unit controller). The AVR regulates voltage at the generator’s terminal by controlling field excitation voltage Vf (Fig. 4.3). Synchronous machine operation at one of the limits, over excitation (OEL) or under excitation (UEL), recloses the correspondent Vlim, OEL input compares actual excitation current If with its maximum value Iflim. Under steady-state operation, If is lower than Iflim, and the integral controller doesn’t participate in the primary voltage control. When If is greater than Iflim, the integral regulator generates a Vlim negative value that reduces the excitation voltage of the synchronous generator. A certain transient level of the excitation windings’ overload can be allowed based on the slow dynamics of the thermal phenomena. Accordingly, the stator current limiting value is transiently increased. In this way, excitation current If can reach very high values for a short time, being limited by the generator thermal limits (rotor and winding heating, etc.).
• UEL input compares the reactive power Q with a reference Qvref. Under SteadyState operation Q is greater than Qvref and the integral regulator doesn’t participate in the primary voltage control. When Q is lower than Qvref, the integral regulator generates a Vlim- negative value that increases the excitation voltage of the synchronous generator.
Control actions are based on local measures and aimed at preventing operating points from overcoming generator thermal limits, with a dynamic performance characterized by a dominant time constant value from within a few seconds to some twenty-plus seconds.
In a hierarchical automatic voltage control system, the role played by OEL and UEL limits is very important, and the shape of their curve and loop dynamics must be carefully reconstructed and considered by the power station regulator who controls the generators’ reactive power. In fact, the generator operating point must be maintained inside the operating limits during normal and perturbed operating conditions, thus avoiding any generator thermal stress and wasted control effort due to possible differences between the real and the not well-reconstructed AVR limits.

Fig. 2.2 Primary voltage control considering reactive power limits
2.4 Tertiary and Secondary Voltage Control Levels
A. Secondary Voltage Control
Secondary voltage regulation has, as its first objective, the automatic voltage control at a power system’s main transmission buses (i.e., the most important load buses) by controlling the largest available reactive power resources on site. Therefore, primary and secondary voltage controls have different and sometimes opposite aims.
Secondary voltage control plays an important role both during normal operating conditions and in front of contingencies:
• In normal grid operation, it ensures:
? Maintenance of network voltages at a specified value and reduction in their changes.
? improve optimization control efficiency;
? optimal distributed online controls of reactive power generating units;
? Dynamic performance of first-order type to HV voltage transients, with a dominant time constant of about 50 s.
• Under disturbed conditions, secondary voltage regulation:
? Offers real time controls of injected/consumed reactive powers in the disturbance area;
? Speedily recovers the disturbance area voltage level;
? Imposes a first-order dynamic response to voltage transients in accordance with PID (PI) control mechanism, with a proposed time constant of about 50 s as well as fast recovery of most of the peak variations resulted from large disturbance
When the pilot node is considered to be the controlled HV bus in figure. The basic principle of SecVC is voltage control of a wide HV grid through controlling of a small number of buses—the most important ones each of them able to determine voltage in surrounding buses, so each defining its area of influence. SecVC therefore requires dividing the transmission network into regions, within which the voltage is controlled in the main bus, called the region “pilot node”. A regional controller (which controls the pilot nodes and therefore the region) separately coordinates the generators of a given area by automatically adjusting their reactive powers to control the voltage of the region pilot bus. Similar to high side voltage control (HSecV), pilot control voltage consists of closed-loop control of the pilot bus voltage through a PID control action, which defines an area reactive power level “q”, the reactive powers of all the control power plants in the area. The secondary voltage regulator inputs the instantaneous voltage measure of the region pilot bus and compares it with the pilot node voltage set-point, determining instant by instant the reactive power level to be sent to the control power plants in the area. The reactive power level “q” therefore determines the alignment of each area’s generating units, contributing in proportion to their capabilities to total area reactive power.
The automatic voltage and reactive power control of a transmission network considers the hierarchical structure shown in Fig. 2.3, where the control apparatuses are now apparent:
• In this control structure, the first control level (the primary level) consists of conventional generator voltage regulators (AVRs). These make it possible to take fast-control action in the face of local perturbations (for instance, short circuits near a generator) and thereby collectively determine the “primary” voltage regulation of the network.
• The second hierarchical level consists of power station CC regulators, which achieve the reactive power required by the CAC or the RVR at a higher hierarchical level, by operating on the primary voltage control set-points.
• The third hierarchical level consists of a slower CAC (or a few RVRs if the grid is subdivided into more than one region: for example, the case of a national dispatcher operating on-field through regional dispatchers), which regulates in an integral way the voltage of the pilot nodes by controlling the reactive power of participating power stations to the second hierarchical level.
The switching of compensating equipment such as capacitor banks and shunt-reactors or the blocking of OLTC tap-changers is part of SecVC control action. It operates at each region on the local switching resources only when needed, according to the region control margin value, given by the difference of the real-time value of region reactive power level (control signal) “q” with respect to its + 1 or ? 1 p.u. limits. Proper thresholds of the “q” value habilitate area on/off switching according to pre-defined sequences.
B. Tertiary Voltage Control
The basic idea of TerVC derives from the need for a system’s operating security and efficiency to increase through central real-time coordination of the distributed SecVC structure:
• Pilot node voltage set-points must be adequately updated and coordinated online and in real time, with dynamics slower than SecVC, by consideration of the real operating condition of the overall grid and by avoiding pointless and conflicting SecVC inter-area control efforts.
• To this end, pilot node voltage set-points can be computed and updated in real time simply by use of the SecVC control system operating conditions that give reliable, synthetic, timely information on what is going on at the overall system: “SecVC controls that are active on the physical process and the pilot node measurement feedback provide, at any instant, an undoubtable figure of the most important essential happenings in the real process”.
• Therefore, pilot node voltage set-points can be optimized in real time to effectively minimize grid losses while still preserving the control margin by simply referring to the “grid equivalent” real-time system model, based on few but very reliable and significant data on control variables the SecV is able to provide to TerVC.

The TerVC control level is therefore aimed at optimizing nationwide voltages by a suboptimal real-time control. This involves determining moment by moment the pilot node voltage set-point values by minimizing the differences of the measured pilot node voltages with respect to their historical references or off-line forecasted values, always maintaining the control margin in each area. Having a proper selection of SecVC areas, this simplified TerVc optimisation is able to achieve a safe and efficient closed-loop system control by a slower than SecVC dynamic performance. Therefore, the tertiary loop represents the continuous computing of a wide-area, real-time, updated, optimal voltage plan, applied to the grid through the global coordination of automatic control actions achieved by SecVC. The main TerVC objectives are these: (i) the management, at a low speed, of the reactive power flow between the power system areas, accomplished by minimizing power system losses; (ii) the increasing of the power system’s controllability and stability.

Fig. 2.3 Three levels of voltage control.
Fig. 2.4 Secondary Voltage Control Circuit
C. Tertiary and secondary voltage control modeling
• Unit Cluster Control (CC), is shown in Fig. 2.3 The AVR voltage reference VGS could be varied between Vmin and Vmax which obtains the unit reactive power production QG corresponding to its reference value Qref as illustrated in (1).

(1)
Where KG is the regulator integral gain and it is equal to (XTG + Xeq) / (TG) and tuned in such a way that the closed loop has a dominant time constant (TG) of about 5s, XTG represents the transformer generator reactance while Xeq represents the equivalent reactance of the line connecting the supporter generator with the pilot bus. The AVR dynamic response should be fast enough to deal with the local network disturbances to avoid being affected with the reactive power loop. The reference value Qref is obtained from the product of the reactive power level q by the unit capability limit QGL as shown in (2). QGL is computed on-line based on the operation conditions of the reactive power source (Capability curve of the generator).

(2)
• The reactive power level q may be provided by the pilot Central Area control (CAC). In closed-loop and real-time, the reactive level q in the interval between its minimum qmin = -100% and its maximum qmax = 100% which achieves load bus voltage magnitude Vp corresponding to its reference value Vpref. 4.

(3)

(4)
.
where KP, KI, KD are the parameters of the coordinated proportional integral derivative controller.
• The Vpref is given originally by optimal load flow calculations to achieve a certain objective function (as minimum load shedding or minimum power losses or even optimal reactive power reserve) at a certain condition with restriction to some constraints as active and reactive power limits and also bus voltage limits.

2.5 Previous experiences of TerVC and SecVC
A. French model27
The original French secondary voltage regulation system regulates voltage profile in each selected zone by distributing reactive power from the various regulating generators in the zone as illustrated in Fig. 2.5.

Fig 2.5 TerVC and SecVC model in France.
This original SecVC control scheme, which has been in operation since the early 1980s, has the following sample of limitations.
1. The control level N of each pilot node is calculated by a dedicated microprocessors located in the zone’s regional dispatching center. Therefore, each pilot node has a dedicated RVR controller that becomes a regional controller. This does not allow considering possible dynamic interaction among the pilot node controls of the critical regions, unless it is of a sophisticated data exchange.
2. Very slow dynamic performance affects application, the dominant time constant turns out to be greater than 30 min. This is very slow with respect to the declared 3 min and not comparable to the 50s faster.
3. Abnormal transients due to SQR initialization and a standby period of 5 min to allow reactive power alignment among generator sets.
4. Need for an operator to take corrective action on the control level due to transients induced in primary voltage control systems.
To summarize the situation very briefly, France’s SecVC would be more efficient if its controller design had been more accurate and had a less critical subdivision of the grid into zones.
B. Italian model
The hierarchical voltage control system presented its mission to operate successfully in the Italian power system since 1985, contributing to simplifying and improving of grid voltage operation. Secondary and tertiary voltage controllers allow grid operators to achieve a full operation of transmission grid transfer capabilities, as required by today’s energy markets. In the framework of the ancillary services market, data made available by the proposed control system also allow simple, correct recognition of the real contribution of each generator to the voltage service 14.
The Italian experience started with experimental applications in the Florence Area (1985) and in Sicily (1989), which recorded significant benefits. The control system grew step by step, with plants first operating by reactive power high side voltage control, then with all SecVC pilot node control. The very promising results at that time leads TERNA (the Italian TSO) to promote widespread application of SecVC and the development of TerVC-LMC 5–10. Further developments in control apparatus technology and disputable decisions made on SecVC and TerVC integration into the TSO’s SCADA/EMS control systems caused interruptions and delays in the overall control system’s activation and operation. Moreover, forcing SecVC and TerVC into the SCADA\EMS architecture, which was not designed to properly integrate such real-time automatic voltage regulation, runs a highly probable risk of impoverished implementation of SecVC and TerVC regulating loop dynamics.

Fig 2.6 TerVC and SecVC model in Italy.

C. Brazilian model
The control scheme in Fig. 2.7, even if it achieves good quality in the grid but it presents a reason for some critical remarks concerning the system in practice; a few were previously noted in the French model solution:
• No reactive power control loop, but direct control of the primary set-points: In practice, this choice could create dynamic interaction and reactive power recirculation among generators.
• The pilot node control law is 100% integral: therefore, it is too slow in the first part of the transient in front of a large contingency.
• The pilot node voltage control is consisting of a very large number of parallel integrators (Fig. 2.7), as many as the number of reactive power resources controlling that voltage: This determines a high risk of improper operation in practice because of generator limits that are not properly considered and also because of the different dynamics and offsets the considered PriVCs can have, with consequent strong interaction among them and with overlapped integrators.
Looking at 12, the results involving the use of SecVC show the benefits gained improving voltage stability and security. The results also make evident the importance of proper selection of the pilot buses and reactive power resources participating in the SecVC mission.

Fig. 2.7 SecVC model in Brazil

2.6 Technical and Economic Benefits of SecVC+TerVC
The benefits coming from the general application of secondary and tertiary voltage control to a wide area power system are now considered.
a) The technical benefits include:
• Loss reduction turned to be not less than 5 % with continuity;
• Rise in of reactive power reserves available during transients following large disturbances.
• Widening voltage stability margins: voltage collapse delayed;
• Improving Angle stability: SecVC high speed voltage control significantly contributes
to damping electromechanical oscillations too;
• Active power transfer capability increases and power flow congestion reduction;
• Increase in lifetime of equipment (OLTC, capacitor banks, shunt reactors) by reduction of switching maneuvers: 30–70 % reduction in OLTC stepping maneuvers;
• Enhancement in Voltage support: flat voltages at HV and EHV levels against normal load change and transient amplitude reduction in front of contingencies.
• Better voltages with reduced reactive power control effort.

b) The economic benefits include:
• Production and transmission cost recovery due to reducing the losses;
• Investment reduction for Facts equipment (SVC, STATCOM, capacitor banks, shunt reactors);
• Improved management of more economical production under security constraints due to increase in line transfer capacity and reduction in bus congestion;
• Reduction of time during which contractual quality of voltage at customer end is not guaranteed;
• Reduction of not-fed loads for security reasons;
• Reduction of heavy costs and negative social impact due to power system blackouts linked to voltage collapse

2.7 Summary
The chapter presented the main idea of tertiary and secondary voltage control. The reason, mathematical modeling and overview of its implementation and model limitations in some countries were presented. The chapter also presented the benefits of applying SecVC + TerVC from the technical and economical point of views.