Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Definition of Distributed Control for Smart Grids

Distributed control for smart grids is the use of distributed grid assets to achieve desired outcomes such as increased utilization of transmission assets, reduced cost of energy, and increased reliability. Distributed control is a key enabler to meet emerging challenges such as load growth and renewable generation mandates.

Introduction

The utility grid is a massive system with tens of thousands of generators, hundreds of thousands of miles of high-voltage transmission lines and millions of assets such as transformers and capacitors, all working with split-second precision to deliver reliable electrical energy to hundreds of millions of customers. The grid was designed at a time when electromechanical controls and generator excitation control were the only available control handles to keep the system operating. In addition, the inability to guide electricity flows and to store electricity inexpensively forced the adoption of rules that distorted market operation, and created inefficient use of assets. The decision of the US government to make electricity a universal right of the people and a regulated industry, further removed drivers that moved the industry toward effective and competitive operation. The result of these technology limitations and policy structures that were created, with the best of intentions, is that the electricity markets work ineffectively at best.

The electricity industry asset base in the USA is worth an estimated $2.2 trillion in terms of replacement value. Global and US electrical demands are expected to rise 80% and 24%, respectively over the next 20–25 years. In the USA, if grid-enabled vehicles (GEVs) are widely adopted, total electrical demand could be 45% higher in 2035 than 2008. Mandates for renewable energy are expected to require a doubling in annual transmission investment yet will lower utilization of the transmission system. A ubiquitous roll-out of smart grid functionality would provide better utilization of the transmission and distribution system, reduce the investment required to meet renewable mandates, and raise reliability. However, there is no economic basis for implementing such a far-reaching transformation. It is however possible, with correctly defined priorities, to implement an upgrade of selected portions of the grid, to achieve a large part of the important gains. Selective introduction of newer technologies that allow dynamic granular control (DGC) of voltage and power flows on the grid can enable a much more efficient use of grid assets, minimizing the need to build new transmission lines to accommodate variable generation resources such as wind and solar.

Such control capability will transform the electricity market, allowing transactions between willing generators and users, along pathways that have the capacity to handle more power. It will directly mitigate problems of cost allocation that have plagued transmission-line building. It will reduce the cost of absorbing more wind and solar energy, and of adopting GEVs to improve energy security and to reduce carbon emissions. This control cannot be implemented at the traditional generation excitation controls, but has to be implemented at various points on the grid, so that the utilization of diverse grid assets can be improved substantially. Such control will necessarily be distributed in nature, and points to a future vision of grid control.

Distributed control is a key enabler to meet emerging challenges in the electric power sector. This entry begins by surveying the emerging challenges. It then discusses general methods for two types of distributed control, namely VAR control and power flow control. Next, the technologies currently available to realize VAR control and power flow control are discussed. It then surveys emerging technologies and methods before concluding.

Emerging Challenges

Load Growth

Globally, electrical energy demand is expected to grow 2.5% per year from 2006 to 2030 with total demand rising from 15,665 TWh in 2006 to 28,141 TWh in 2030 [1]. Worldwide generation capacity is expected to grow 2.3% per year from 2006 to 2030, rising from 4,344 to 7,484 GW [1]. The DOE Energy Information Agency (EIA) expects US electrical load to increase 1% annually to 2035, a 24% increase in annual demand relative to 2010 demand [2]. These projections include negligible amounts of GEVs with electricity supplying less than 3% of global transport energy and 0.2% of US light-duty transit sector energy in 2030 and 2035, respectively [1, 3]. Historically, load growth has required investment in generation as well as the transmission and distribution systems. Utilities are facing pressure to meet this load growth at low cost to consumers, a challenge given high costs for fuel and materials.

Challenges Permitting Urban Generation and Transmission

The need for distributed control can be mitigated through the installation of generation in load centers or new transmission connecting distant generation with load centers. Siting generation in load centers is difficult due to noise, emissions, viewshed concerns, land acquisition cost, and the limited capacity of the urban fuel delivery system. In addition, generator cooling water availability is projected to be limited in many load centers [4]. A study in the USA shows that 19 of the 22 counties identified as most at risk for water shortage due to electricity generation are located in the 20 fastest growing metropolitan statistical areas.

Construction of new transmission to supply load centers is also difficult. Concerns include viewshed, land acquisition cost, and electromagnetic radiation. High-voltage transmission lines require more than a decade to build. Also, the allocation of cost is problematic, especially for lines that cross state and utility boundaries. Finally, the construction of new lines increases the MW-miles of line serving a given amount of load, decreasing rather than increasing utilization of the existing system, leading to cost increases and inefficiencies.

Policy Drivers for Renewable Generation

Policies are driving the adoption of renewable generation around the world. China aims to increase renewable energy from 7.1% of total electrical energy in 2005 to 20% in 2020 with the installed wind capacity rising from 1,260 MW in 2005 to 30,000 MW in 2020 [5]. Grid operators in China are required to accommodate renewable energy, with prices for wind decided in an auction in which only wind plants participate [5]. In the USA, 29 states plus the District of Columbia have enacted binding renewable portfolio standards (RPSs), and 7 states have voluntary goals [6]. RPSs require a certain percentage of annual electrical energy demand to be met with renewable generation by a specified year. In addition, the US federal government incents the development of renewable generation through the Production Tax Credit (PTC) and the US Treasury Section 1603 Investment Tax Credit.

The renewable mandates stress the transmission system and require additional investment. One metric of this stress is the ratio of peak generation capacity to average power demanded. In 2008, the US ratio was 2.18. In the EIA 2035 reference scenario, which sources 5% of national electricity from nonhydro renewables resources, the ratio improves to 2.03. All things being equal, this reduction shows increased utilization of the generation fleet in 2035 relative to 2008 and lower unit cost of energy. If a 20% RPS is mandated, the ratio rises to 2.29, a degradation resulting in lower utilization in 2035 relative to 2008. So, while the 2035 nonrenewable case shows a 7% increase in utilization, the RPS case shows a 5% degradation. To avoid curtailment of renewables, transmission must be built to accommodate the peak power output of the renewable fleet.

Globally, both wind and solar photovoltaic generation have grown exponentially over the last fifteen years, as seen in Fig. 6.1. As seen in Fig. 6.2, the highest quality wind and solar resources in the USA are remote from load centers. Siting generation at these high-quality resources maximizes the plant owner’s return on investment if the transmission network can accommodate the generated energy. A number of integration studies have been performed around the globe to assess the transmission and system operation impacts of wind and solar energy on the system [7–22]. Two studies covering large, multistate regions of the USA are the Eastern Wind Integration and Transmission Study (EWITS) and the Western Wind and Solar Integration Study (WWSIS). In EWITS, the design of the transmission build-out was iterated using a committee of stakeholders. Undiscounted transmission costs were $33–59 billion (2009$) higher than the reference case, which had minimal additional wind capacity. This amounted to $300–447 of added cost per kW of installed wind capacity. With the transmission build-out, wind curtailment averaged 7–11%, higher than current levels in the USA. In the WWSIS, Wyoming, Colorado, Montana, Arizona, and Nevada were supplied with up to 35% renewables. A simplified transmission build-out was developed to trade renewable energy among these states. The cost was $11B with a benefit-to-cost ratio of 0.9.

Fig. 6.1
figure 1

Logarithmic plot of global cumulative wind and solar PV capacity installed [23]

Full size image
Fig. 6.2
figure 2

Map of US wind and solar resources and major metropolitan areas. Solar resources are shown from light yellow to brown, with the highest resource sites in brown. Wind resources are shown from light blue to dark blue, with darker being more attractive. Only wind sites viable with current technology are shown. Major metropolitan areas are outlined in black. Gray lines in the background show the high-voltage transmission system

Full size image

For a scenario in which 20% of US demand is met with renewable energy, the estimated transmission cost is $360 billion over 25 years, or $14 billion per year. This assumes the average distance between renewable generation and load is 1,000 miles, a reasonable scenario if renewable generation is installed in high resource areas in the Midwest and Southwest. Currently, total transmission investment, including long-distance and local transmission, is $9 billion per year. If renewable energy transmission investment substitutes for 50% of the current investment, total annual investment would be $18.5 billion per year, a 105% increase over current investment levels. This increased level of investment would need to be sustained for 25 years to supply 20% of projected US demand with renewable energy.

Policy Drivers for GEVs

Grid-enabled vehicles (GEVs) are experiencing a renaissance following their last resurgence in the 1990s. Mainstream manufacturers and new entrants offer or plan to offer GEVs spanning a broad price range. In the USA, GEVs are currently supported through federal and state tax credits. Proposals have been developed to electrify 75% of the miles traveled by the light-duty fleet by 2040 [24]. In Europe, high fuel taxes and a requirement that fleet average emissions drop from 160 to 95 g CO2/km are likely to foster GEV development [24]. The Chinese government recognizes the ability of GEVs to improve energy security and has implemented a program to build charging infrastructure in the 13 largest cities. It also offers a tax incentive program for vehicles and buses on the order of $8,000 and $70,000 per vehicle, respectively [24]. The impact of GEVs on the power system is dependent on the level of GEV adoption, the charging model, and the level of coordination between the time of charging and the power system state. If 80% of the miles traveled by the US light-duty fleet are electrified by 2035, annual electrical demand is projected to increase 45% relative to 2008, compared to a 24% increase without GEVs. Uncoordinated charging will lead to overloading and accelerated aging of distribution assets [64]. The monitoring system for distribution assets is less developed than for transmission assets, making it difficult to proactively upgrade strained distribution assets. Without coordination of vehicle charging, significant new transmission and distribution investment will likely be required.

Differences Between the Electrical Network and Other Commodity Delivery Networks

In the USA, natural gas and petroleum pipelines have become less regulated over the last two decades. Before the 1990s, pipeline developers submitted plans for new pipelines to FERC. FERC selected which pipelines would be built based on its assessment of societal needs. It was assumed that FERC could not allow competition among pipelines. Following PUC CA v. CA 1990, FERC allowed more competition in the pipeline industry. Different rates could be charged to anchor customers, who supported pipeline construction at an early stage, and other customers who requested pipeline access after construction.

In petroleum pipelines, product differentiation is possible by separating shipments via a marker liquid called transmix. As of 2005, most shipment of oil was regulated by FERC, but refined products based on petroleum used market-based rates [25, 26]. However, market-based rates are the fastest growing rate type [25]. FERC permits the use of market rates on a case by case basis after ensuring that the pipeline does not have an unreasonable level of market power.

Due to deregulation, by 2000, the natural gas pipelines were no longer guaranteed a rate of return [27]. Users who reserved capacity were given firm transmission rights. Tariffs for firm capacity were regulated [28]. Remaining capacity was sold in bulletin boards at market rates. Owners of capacity can resell their capacity to realize arbitrage opportunities.

The electric network is considered a natural monopoly due to the inability of transmission line owners to control the flow of power through their lines. This lack of control means owners cannot take bids for the use of their lines, as is done in natural gas pipelines. The lack of control creates free-rider effects whereby entities benefit from investments made by others. Collectively, this creates a disincentive for transmission owners to upgrade their assets.

Reliability Challenges in Emerging Economies

A country-wide overview of reliability statistics was not identified for the emerging economies. Statistics for 1997–2008 were found for five utilities in Brazil, serving a combined electrified population of 389 million [29]. Statistics were also found for two power companies in India, the North Delhi Power Limited (NDPL) and the Bangalore Electricity Supply Company (BESCOM) [30]. Compared to the USA, the Brazilian utilities had eight times the annual outage duration and 14 times the annual outage frequency. The BESCOM system had 20 times the US outage duration. It is anticipated that continued economic growth will require more reliable supply of electricity, necessitating investment in emerging economies.

Distributed Control Techniques

Distributed control has been used to operate the electrical system since its inception. The initial subsection of this section will describe optimal power flow (OPF), which traces its history back to the earliest days of power system operation. Today’s power system relies primarily on OPF to realize control. As discussed in the previous section, numerous challenges are emerging for power system investment and operation. Accommodating renewable energy and GEVs will likely require a response time not possible with traditional control techniques such as OPF. This section will introduce VAR control and power flow control. Realized with appropriate technologies, VAR control and power flow control can provide fast-responding control, called dynamic granular control (DGC), to meet the emerging challenges.

The techniques described in this section can be applied to both the transmission and distribution systems. Controllability of the distribution system has lagged behind that of the transmission system. This is in part due to the large number of distribution assets, relative to transmission assets. The number of assets increases the complexity and cost of controlling the distribution system. Unlike the transmission system, the distribution system has traditionally been operated as a radial network rather than a meshed network. This simplifies control but leads to lower reliability. Smart grid activities to date have been largely aimed at improving control of the distribution system to improve efficiency, increase capacity without new asset construction, and accommodate distributed generation.

Optimal Power Flow

Optimal power flow (OPF) is used by system operators and planners to minimize the total cost of system operation. OPF is the primary control technique in today’s grid. The main control handles, or control actions, are adjustments of the real and reactive power injections of the generators. However, OPF can include other control handles, such as the settings of a phase-shifting transformer (PST). Although control decisions are typically made centrally, OPF is considered distributed control because the control actuators, such as generators and PSTs, are distributed throughout the system. OPF set points are not typically changed more frequently than every 5 min. OPF is configured to provide sufficient safety margins to accommodate unexpected events such as load forecast error, renewable generation forecast error, and contingencies. A contingency is when any one or more grid assets, such as a power line or transformer, suddenly go offline.

The realization of OPF is different in vertically integrated and market environments. In both, power flows are changed by directly changing the output power of the generators. In the absence of other power flow control technologies, OPF does not allow the control of flow in individual lines.

In a vertically integrated environment, a single utility is responsible for generation, transmission, and distribution in a given region. This utility owns all of the assets in the region and knows the costs and technical limitations of its assets. The inputs to the OPF are as follows:

  • Generator parameters – For each generator, the maximum power output, minimum power output, bus to which the generator is connected, and a cost curve are provided to the OPF. Each cost curve describes the variable cost of operating the unit at each potential operating point between minimum power and maximum power. Variable cost includes fuel and variable O&M costs.

  • Transmission line parameters – For each transmission line, the line impedance values, nominal voltage, current rating, and terminal locations are specified.

  • Transformer parameters – For each transformer, the transformer impedance values, nominal voltages, current rating, and terminal locations are specified.

  • Load data – The expected demand at each bus is forecasted for each time step over which the OPF will be run.

  • Contingency list – Normally, the OPF is run for a set of potential contingencies rather than all possible contingencies to reduce the solution time.

  • Stability limits – The OPF does not typically compute the stability of the network endogenously. Rather, rules are provided to the OPF in the form of a nomogram or lookup table to ensure that unstable solutions are avoided.

  • Reserve margin – The amount of surplus generation capacity that must be available to meet uncertainties in load and renewable forecasts as well as generator outages.

In a market area, an independent system operator (ISO) or regional transmission operator (RTO) runs the OPF to decide which generators will be used and at what level. The transmission network is typically owned by regulated entities, which may also own the distribution networks. Generators are independent of the owners of the transmission and distribution networks. Rather than cost curves, the generator owners submit bid curves to the ISO or RTO. The bid curves designate the minimum price generation owners are willing to accept to generate energy over the whole range of the generator output. In the simplest scheme, owners of the distribution network designate how much power their customers will require, and consumers are price takers. The other inputs to the OPF are the same as those used by the vertically integrated utility. Using the generator bids and forecasted demand, the ISO or RTO runs the OPF to compute the least cost dispatch of generators to serve the load and meet security requirements.

The economic impact of the lack of power flow control is visible in Fig. 6.3. In the figure, generators, each indicated by a circle with the letter “G” inside, are connected to bus 1 and bus 3. Load is connected at bus 3. Three transmission lines connect the three buses. Each line has a thermal rating of 1,000 MW. All transmission lines have the same impedance X. G1 is the cheapest generator. To minimize cost, G1 should serve the entire load at bus 3. However, due to transmission network constraints, G1 can only supply 1,500 MW. One thousand megawatts flows from bus 1 to bus 3 through line 1–3, and 500 MW flows from bus 1 to bus 3 via line 1–2 and line 2–3. The flow is less in line 1–2 and line 2–3 compared to line 1–3 because of the increased impedance. Because of this impedance difference, the flow through line 1–2 reaches the limit of 1,000 MW when the total power transfer is 1,500 MW. This sets the maximum amount of power G1 can supply to bus 3 even though 500 MW of unused capacity exists on line 1–2 and line 2–3.

Fig. 6.3
figure 3

Three-bus system showing power flows resulting from OPF control of generation

Full size image

OPF alone is limited to adjustments of generator output and is not able to resolve numerous challenges, including those described above. While the following sections will describe some power flow control capabilities beyond generation control that are currently implemented, the majority of power today is controlled, inefficiently, through OPF control of generator set points.

VAR Control

VAR control provides a number of benefits to the system. First, it can be used to control the voltage profile along a transmission line, thereby increasing the maximum amount of power which can be transmitted in the line. It can also be used to decrease the reactive power which must be sourced at the endpoints of a line and reduce line losses. VAR control can also be used to improve voltage stability if the end of a power line does not have sufficient generation. VAR control, if appropriately fast, can improve transient stability. Improvement of transient stability also increases the power which can be transferred on the line. Finally, VAR control can be used to damp power system oscillations, averting damage to rotating machinery, increasing power transfer capability, and potentially improving stability.

Power Flow Control

Under certain simplifying assumption, the power flow between two buses of an electrical network is determined by Eq. 6.1. The power flow can be modified by modifying any of the parameters of the equation:

  • V1 and V2 are the voltage magnitudes of the endpoints of the transmission line.

  • δ is the angle difference between the voltage phasors at the endpoints of the transmission line.

  • X is the reactance of the transmission line.

  • P is the power transmitted over the transmission line.

Voltage changes alone cannot impact power flow in a large way, since the voltage is typically maintained within ±5% of a nominal value to avoid damaging generators, the transmission network, the distribution network, or customer equipment.

The following sections will detail the methods used to change power flows:

Power flow between two AC buses

$$ {P} = \frac{{V_1V_2 sin\delta}}{X}. $$
(6.1)

Impedance Control

Impedance control of power flow aims to directly change the impedance, represented by the value X, as seen in Eq. 6.1. Technologies able to change impedance include the series mechanically switched capacitor (series MSC), the series mechanically switched reactor (series MSR), the thyristor-controlled series capacitor (TCSC), the static synchronous series compensator (SSSC), and the unified power flow controller (UPFC). These technologies will be discussed in more detail in section “Existing Distributed Control Technologies.”

The benefit of impedance control is shown in Fig. 6.4, a modification of the three-bus system previously shown in Fig. 6.3. Here, a power flow controller able to change line impedance has been installed on line 2–3. Through an appropriate modification of the impedance, the power flow in all lines can be equalized, allowing up to 2,000 MW to be transferred between bus 1 and bus 3, compared to the 1,500 MW using generator control alone. This allows the entire load at bus 3 to be served by the low-cost generator, resulting in a reduction in the cost of energy.

Fig. 6.4
figure 4

Three-bus system with power flow controller installed on line 2–3. The OPF determines the optimal settings for the generator and power flow controller

Full size image

Angle Control

Angle control aims to change power flows through the parameter δ as seen in Eq. 6.1. Technologies able to control the angle include the phase-shifting transformer (PST), also known as phase angle regulator (PAR), the variable frequency transformer™ (VFT), and the solid-state transformer. The power flow control of Fig. 6.4 can also be realized using angle control. The PST and VFT will be discussed in more detail in section “Existing Distributed Control Technologies.” The solid-state transformer will not be discussed as it is not commercially available and is expected to have high cost.

Voltage Control

Angle control aims to change power flows through the parameter V1 or V2 as seen in Eq. 6.1. Technologies able to control voltage include the static VAR compensator (SVC) and the static synchronous compensator (STATCOM). The power flow control of Fig. 6.4 cannot be realized via voltage control although voltage control can be used for limited power flow control in more complicated networks.

DC Control

It is also possible to control AC power flows by converting from AC to DC and then back to DC. Two technologies include the high-voltage DC transmission (HVDC) and the back-to-back converter (B2B). Both will be discussed in more detail in section “Existing Distributed Control Technologies.”

Existing Distributed Control Technologies

This section will discuss existing distributed control technologies for VAR control and power flow control. These technologies provide various levels of dynamic granular control (DGC). These technologies can be actuated through OPF for response times on the order of minutes or through separate control algorithms for response times on the order of microseconds.

VAR Control

Shunt Mechanically Switched Capacitor (shunt MSC)

A mechanically switched capacitor (MSC), when connected in shunt at a bus, can provide VAR control by delivering capacitive VARs. A typical installation is shown in Fig. 6.5. Switching causes deterioration of the mechanical switch. Thus, MSCs are typically not actuated frequently. MSCs are ubiquitous. The cost for an MSC was $8–10/kVAR in the mid-1990s [31].

Fig. 6.5
figure 5

Shunt mechanically switched capacitor (MSC) connected to bus 1

Full size image

Shunt Mechanically Switched Reactor (shunt MSR)

A mechanically switched reactor (MSR), when connected in shunt at a bus, can provide VAR control by delivering inductive VARs. A typical installation is shown in Fig. 6.6. A circuit breaker is typically used to connect and disconnect the MSR due to the inductive load. As with MSCs, frequent actuation deteriorates the mechanical switching element. MSRs are widely used.

Fig. 6.6
figure 6

Shunt mechanically switched reactor (MSR) connected to bus 1

Full size image

Thyristor-Switched Capacitor (TSC)

A thyristor-switched capacitor (TSC), shown in Fig. 6.7, is a shunt capacitor which is switched in and out using a thyristor pair. The use of a thyristor pair, rather than a mechanical switch as in the MSC, allows for longer life. Also, a TSC can respond faster than an MSC, allowing for an injection of VARs following a system transient.

Fig. 6.7
figure 7

Thyristor-switched capacitor (TSC) connected to bus 1

Full size image
Fig. 6.8
figure 8

Static VAR compensator (SVC) connected to bus 1

Full size image

Static VAR Compensator (SVC)

The static VAR compensator (SVC) is comprised of a combination of shunt MSCs, TSCs, and thyristor-controlled reactors (TCRs) to control grid parameters by changing the reactive admittance of the SVC. A typical SVC configuration is shown in Fig. 6.8. The maximum level of reactive current changes linearly with the bus voltage, and maximum VAR output varies as the square of the bus voltage. An SVC cannot increase VAR generation during a transient. However, since an SVC is typically implemented in a single-phase, line-to-ground manner, it can provide support during unbalanced faults. Mechanically switched capacitors are preferred for applications consistently requiring capacitive injection, as they have lower losses than TSCs. However, an MSC is limited to 2,000–5,000 switching cycles before the switch must be replaced, limiting the use of the MSC unless the required level of VAR compensation changes slowly [32]. In addition, an MSC has a slower response time than a TSC. The configuration in Fig. 6.8 combines an MSC for steady-state capacitive injection with a TCR for transient performance. The TCR generates harmonics which are removed with the shunt filters.

Worldwide, ABB has installed 499 SVCs totaling 73.3 GVAR, 54% of the global total [33]. Of these, 228 units, with a total capacity of 49 GVAR, are for utility customers. The remaining units are for industrial customers. Siemens has installed at least 45 utility SVCs representing almost 10 GVAR of capacity [34]. AMSC, formerly known as American Superconductor, has installed 130 SVCs and STATCOMs, although the breakdown between the two types is unknown [35]. An SVC cost $50/KVAR in the mid-1990s and is expected to cost roughly $80/KVAR now [31].

Static Synchronous Compensator (STATCOM)

A STATCOM uses solid-state gate turnoff devices to mimic the operation of a synchronous condenser, an electrical machine configured to only produce reactive power. The most common type of STATCOM, shown in Fig. 6.9, uses a single three-phase voltage source converter. The STATCOM maximum reactive output current is nearly constant over a wide voltage range, so reactive power output is nearly linear over a wide bus voltage range. Bus voltages are typically depressed during faults and portions of transients, allowing the STATCOM to provide additional VARs during periods with depressed bus voltages compared to an SVC of the same rating. The STATCOM response is an order of magnitude faster than an SVC, and its footprint is 30–40% smaller than an SVC [32].

Fig. 6.9
figure 9

One-line diagram of a static synchronous compensator installed at bus 1

Full size image

A STATCOM can also be configured to provide a temporary increase in reactive current during transients. Use of a STATCOM for three-phase fault and transient mitigation allows the use of a lower rating for the STATCOM than the SVC. However, the three-phase converter implementation limits the STATCOM’s ability to provide VARs during unbalanced faults unless the DC capacitor is significantly overrated relative to steady-state operation. This is a significant disadvantage compared to an SVC since an SVC can typically support unbalanced faults. However, AREVA has implemented a STATCOM able to support unbalanced faults through the use of separate converters for each phase.

A single, three-phase STATCOM was estimated to cost $55/kVAR in the 1990s [31] and $150/kVAR now. A system comprised of three single-phase STATCOMs is estimated to cost $200/kVAR.

Siemens has installed at least 14 STATCOMs [46], and ABB has installed an unspecified number of STATCOMs. AMSC has installed 130 SVCs and STATCOMs, although the breakdown between the two types is unknown [35].

Hybrid VAR Systems

Some installations use an SVC-STATCOM configuration, replacing the TCR of the SVC with a STATCOM. This allows for the use of an MSC, offering lower losses than a TSC, while retaining the quick response of the STATCOM. This also allows for smaller filters since TCRs generate harmonics [32]. The cost and performance characteristics lie between the SVC and STATCOM and depend on the relative rating of the components.

Comparison of Technologies

Table 6.1 provides a comparison of the VAR controllers discussed above.

Table 6.1 Comparison of salient elements of VAR controllers
Full size table

Power Flow Control

Series Mechanically Switched Reactor (series MSR)

A mechanically switched reactor (MSR) can modify line impedance if installed in series with the line as shown in Fig. 6.10 [36, 37]. Appropriately designed, a series MSR can be installed on an overloaded, low-impedance line within a meshed system to push power flow onto less utilized paths, to prevent frequent tripping of the line under system transients and contingencies, and to limit fault currents. It is possible to design the series MSR to be switched out, although that is not frequently done. Further, the speed at which the reactors are switched out is slow, designed mainly for a planned change in system/line impedances, and is not sufficient for the rapid response required to mitigate transients.

Fig. 6.10
figure 10

Series mechanically switched reactor (MSR) installed on a transmission line between bus 1 and bus 2

Full size image

A series MSR has a number of advantages and disadvantages relative to other methods to control or increase power flow. It is quicker to install than a new transmission line, with historical projects requiring 6–8 months for construction [38]. In addition, given that the MSR does not require new right of way, it requires significantly less time to plan and permit. However, it imposes an additional footprint requirement somewhere along the line, typically at one of the substations, which may require permitting and add cost. A 230 kV MSR solution in California required an additional 5,000 ft2 of substation area [37]. To reduce cost, manage mechanical stresses, and avoid core saturation during faults, air core reactors are typically used which leads to increased noise and magnetic fields. The magnetic fields are sufficiently high that care must be taken to mitigate unintentional heating of metallic equipment in the substation and hazards to workers in the substation. Finally, a series MSR is heavy and cannot be supported by the line. This requires that they be ground mounted with the requisite, and costly, isolation between ground and line potential. In the early 2000s, a series MSR cost $11–22/KVAR, although these figures are dated and the current cost is expected to be closer to $50/KVAR [36, 39].

Series Mechanically Switched Capacitor (series MSC)

A mechanically switched capacitor (MSC), when inserted in series with a transmission line, can control power flow by lowering the impedance of lightly loaded lines, thereby drawing power away from heavily loaded lines. A typical installation is shown in Fig. 6.11. Like a series MSR, a series MSC is not designed to be switched in or out rapidly and can thus not be used to mitigate transients. Since an MSC is not typically built to endure the high currents of a fault, protection methods are used to bypass the MSC during the fault and reengage it soon thereafter to increase system stability [40]. While early series capacitors were protected from fault currents using a maintenance intensive spark gap, more recent designs utilize an MOV or thyristor protection [41]. Like a series MSR, a series MSC imposes a footprint requirement along the line. As with a series MSR, a series MSC is sufficiently heavy that it cannot be supported by the line and requires an isolated, elevated platform. Finally, an MSC can lead to subsynchronous resonance with the rotating elements of a generator which can cause generator failure and power system instability [42]. This results in siting limitations. The cost for an MSC is $30–50/KVAR [43, 44].

Fig. 6.11
figure 11

Series mechanically switched capacitor (MSC) installed on a transmission line between bus 1 and bus 2 [45]

Full size image

Thyristor-Controlled Series Capacitor (TCSC)

A TCSC, shown in Fig. 6.12, is a series capacitor augmented with an additional shunt path comprised of an inductor and thyristor pair. A TCSC is able to inject a capacitance higher than the nameplate capacitance of the capacitor. However, like the series MSC, this solution is plagued by the need for an isolation platform, which increases cost and requires additional space at the substation. Unlike a series MSC, a TCSC can avert subsynchronous resonance challenges and is fast enough to improve transient stability.

Fig. 6.12
figure 12

Thyristor-controlled series capacitor (TCSC) installed on a transmission line between bus 1 and bus 2

Full size image

Multiple manufacturers produce TCSCs. Siemens alone has installed five TCSCs as of late 2009 [46], with the first installation conducted for WAPA in 1992. BPA’s 500 kV TCSC project was commissioned in 1993 [47]. Typically, the TCSC solution requires a construction period of 12–18 months. The typical cost for a TCSC is $150/kVAR [48].

Phase-Shifting Transformer (PST)

A phase-shifting transformer (PST), also known as phase angle regulator (PAR), controls power flow between two electrical buses by changing the δ term of Eq. 6.1. Numerous PST topologies exist. The simplest type has a fixed phase angle. In more complicated topologies, mechanical load tap-changers (LTCs) or solid-state switches are used to vary the phase angle. The most common type of PST uses a single core as seen in Fig. 6.13. While previous figures have been one-line diagrams, the PST figures show all three phases due to the cross-coupled nature of the PST. Winding 1 induces a voltage on windings 1′ and 1″, which are connected in series with phase A of the transmission line. Since winding 1 is connected between phase B and phase C, the voltages injected into phase A by windings 1′ and 1″ are in quadrature to the phase A voltage. The tap changers connected to windings 1′ and 1″ allow for the level of the injected quadrature voltage to be varied, regulating the phase shift. A symmetric structure allows a variable quadrature voltage to be injected into phases B and C as well.

Fig. 6.13
figure 13

Three-phase diagram of a single-core phase-shifting transformer (PST) installed on a transmission line between bus 1 and bus 2, inspired by [52]

Full size image
Fig. 6.14
figure 14

Three-phase diagram of a two-core phase-shifting transformer (PST) installed on a transmission line between bus 1 and bus 2, inspired by [52]

Full size image

This single-core design requires the full line current to pass through the LTCs, exacerbating fault current management and increasing the cost of the LTCs [49]. A more expensive PST relieves some of the drawbacks of a single-core PST through the use of shunt and series transformers, as shown in Fig. 6.14. This version couples a series connected transformer and a delta-connected exciting transformer. The voltages of windings Y′ and Z′ drive winding A′. The resulting winding A′ voltage is in quadrature to the phase A voltage. Winding A′ impresses a voltage upon phase A of the line, changing the angle. Like the single-core PST, LTCs are used to vary the level of quadrature injection. Another type of PST replaces the mechanical LTCs with pairs of thyristors. Finally, a fifth type of PST uses voltage source converters to synthesize the quadrature injection voltages. Although the PST is frequently used to alter power flows, only the versions using thyristors or voltage source converters are sufficiently fast to mitigate transients. Minimum cost for the single-core PST is estimated to be $150/kVA. Only a fraction of a PST’s rating will be controllable, so the actual cost per controlled kVA will be much higher.

PSTs have been identified with ratings up to 500 MW and 230 kV [50]. PSTs have been used in WECC for at least 35 years [51]. Most PSTs deployed in the USA can respond to operator commands or take automatic action to maintain preset power flows in a matter of minutes [50]. The Montana Alberta Tie Limited (MATL), believed to be the first AC merchant project in the USA, is slated to have a PST. In the USA, there are at least 20 PSTs in WECC, at least 36 PSTs in the Eastern Interconnection, and at least 3 in ERCOT.

High-Voltage DC Transmission (HVDC) and Back-to-Back Converter (B2B)

The high-voltage DC transmission system (HVDC) and the back-to-back converter (B2B) transform power from AC to DC and then back to AC. Both can be used to interconnect synchronous or asynchronous networks.

HVDC transports the power in DC form over a distance of up to thousands of miles between the two terminals of the line. HVDC provides full control of the line power. If HVDC is used to connect distant generation to a load pocket, the generation is assigned the same capacity benefits as if it was located in the load pocket [53]. Controllability and capacity benefits have incented the use of HVDC for merchant transmission projects in the USA. In addition, the HVDC line can be configured to limit maximum system fault current levels, an advantage over new AC lines which tend to increase fault current. The most popular type of HVDC system is the bipole, as shown in Fig. 6.15. In the figure, bus 1 and bus 2 are AC, and conversion from DC to AC occurs in the HVDC terminals. Each HVDC terminal includes a converter, a power factor correction system, and a harmonic filter. Power factor correction is required due to the reactive power requirements of the converter. Harmonic filtration is required due to the harmonics generated by the converter. A bipole uses two conductors and does not use the earth return under balanced operation. If one of the poles or conductors is offline, the system can continue to operate at a fraction of rated power. When a pole or conductor is offline, current is returned through the earth. A two-terminal, 3,000 MW bipole HVDC system costs $70/kW for each terminal and $1 M/km [54].

Fig. 6.15
figure 15

Bipole high-voltage DC transmission system (HVDC) installed between bus 1 and bus 2

Full size image

A B2B is an HVDC system with the two terminals directly connected, as shown in Fig. 6.16. The B2B consists of the same components of an HVDC, less the DC transmission line. Like a PST, the B2B allows for the control of the power flow. Unlike a PST, it is not limited by the phase shift and fault current limitations of the PST. Also, while a B2B can be used to interconnect asynchronous systems, a PST cannot. However, the B2B cost of $100–300/kW is higher than the PST [55]. Also, the availability of the B2B is around 97%, lower than the 99.9% availability of a transformer [56]. For additional cost, B2B can be used to support black start after a blackout.

Fig. 6.16
figure 16

Back-to-back converter (B2B) placed between two AC buses

Full size image

HVDC and B2B systems are realized using thyristors or gate turnoff devices. Gate turnoff devices are used for smaller power/voltage levels, but the power/voltage capability of HVDC and B2B systems using gate turnoff devices are increasing. Currently, thyristor-based systems dominate the HVDC and B2B markets.

In total, nearly 140 GW of HVDC and B2B capacity is installed or planned worldwide. There are numerous B2B systems installed globally. B2Bs are installed at the seams of the three US interconnections as well as interfaces between the US eastern interconnection and the Hydro-Quebec system. ABB alone has completed 30 HVDC projects and 13 B2B projects using thyristors [57] as well as 13 HVDC projects and 1 B2B project using gate turnoff devices. Siemens has installed 13 HVDC projects and 8 B2B projects using thyristor technology [58] as well as one HVDC system with gate turnoff devices [59].

Variable Frequency Transformer™ (VFT)

A variable frequency transformer™ (VFT), as shown in Fig. 6.17, provides functionality similar to a B2B without conversion to DC. The VFT is a large doubly fed induction motor developed by GE for power flow control applications. Power flow from the rotor to stator is varied by changing the torque on the rotor. Torque is applied using a drive motor and variable speed drive. The VFT is able to change power flow from full-rated output in one direction to full-rated output in the other direction in roughly half a second [60]. Each VFT is rated for steady state of 100 MW, but short-term overload power can exceed 150 MW [61]. GE specifies an efficiency of 99% [62] at full power. GE also claims that the area required for a 200 MW VFT station is 47% less than a 200 MW HVDC station [62]. The VFT cost is unknown.

Fig. 6.17
figure 17

Variable frequency transformer (VFT) connected between two AC buses

Full size image

GE has installed 5 VFTs at three locations. The first VFT was used to interconnect the Hydro-Quebec and New York networks. The next three were installed in parallel in Laredo, Texas. The most recent is the Linden VFT, between New York City and New Jersey.

Static Synchronous Series Compensator (SSSC)

The SSSC, shown in Fig. 6.18, improves upon the functionality of a combined series MSR and series MSC through the use of inverters using gate turnoff power electronic devices. Unlike the MSR/MSC combo, the SSSC can provide subcycle response times and actively mitigate transients. Using a voltage source inverter, a voltage is injected into the line in quadrature with the line voltage. Whether this voltage leads or lags the line current determines if the injected VARs are inductive or capacitive. The risk of subsynchronous resonance is mitigated. While no evidence of SSSC installations could be found, it seems the SSSC is designed to be ground mounted and connected to the line via a series transformer. This eliminates the need for the elevated platform used in the MSR, MSC, and TCSC. However, it still requires additional substation space. Also, the series transformer is expensive as it has to be designed to handle high basic insulation levels and high fault currents. The SSSC solution is also plagued by very high costs which are further exacerbated by series protection requirements and customizations. No evidence of global stand-alone SSSC installations was found. The Marcy convertible static compensator (CSC) can be configured to provide SSSC functionality on two lines simultaneously. Cost is unknown but expected to exceed $200/kVAR [63].

Fig. 6.18
figure 18

Static synchronous series compensator (SSSC) installed on a transmission line between bus 1 and bus 2

Full size image
Fig. 6.19
figure 19

Unified power flow controller (UPFC) with shunt element installed at bus 1 and series element installed between bus 1 and bus 2

Full size image

Unified Power Flow Controller (UPFC)

A UPFC, as shown in Fig. 6.19, combines a STATCOM and SSSC. The STATCOM and SSSC share a common DC bus, allowing for the injection of real and reactive power into the line. UPFC devices are able to regulate line voltage using shunt VAR control and real and reactive power flow control using series voltage injection. UPFC devices are able to meet the technical performance requirements for power flow control, but are expensive. Cost estimates range from $190/kVA up to $300/kVA [48].

There is at least one dedicated UPFC in the USA, located in Inez, Kentucky. It has a 40 MVA series converter and a 40 MVA shunt converter. There is also a UPFC in Kanjin, Korea. The Marcy convertible static compensator (CSC) can be configured to provide UPFC functionality.

Comparison of Technologies

Table 6.2 provides a brief comparison of the technologies. It is important to recognize that the cost of some technologies is priced in terms of $/kVAR and others in $/kVA. The PST is priced in terms of total throughput power, and the price per KVA controlled will be significantly higher than the value listed. The B2B is also priced in terms of throughput power, but the full-rated power of the B2B is controllable. The UPFC cost is given in terms of dollars per unit control effort.

Table 6.2 Comparison of salient elements of power flow controllers
Full size table

Emerging Technologies

The following section reviews three emerging technologies which may resolve the limitations of the power flow and VAR controllers described above. All three are thin AC converters (TACC).

Thin AC Converters

To increase the cost effectiveness and reliability of power electronic-based control technologies, the concept of a thin AC converter (TACCs) has been proposed. The TACC augments a passive grid asset, like a power line, shunt capacitor, or transformer, with a fractionally rated converter. In the process, the asset is transformed into a dynamically controllable asset that performs similarly to traditional FACTS devices. In addition, reliability is improved relative to traditional FACTS because the TACC can be bypassed if it fails, leaving the passive asset in service. This is called fail-normal operation. Three TACCs are presented below: smart wires, the controllable network transformer (CNT), and the dynamic capacitor (D-CAP). The technologies have been demonstrated in the lab and are currently being scaled to utility voltage and power levels.

Smart Wires (SW)

A family of distributed impedance control technologies has been proposed to control power flow. The simplest version of the technology, SW, is comprised of modules hung on the individual conductors of the transmission line as seen in Fig. 6.20. Each module monitors the line current. If line current crosses a preset threshold, the module autonomously takes action by injecting inductive impedance into the line. Other modules inject impedance at different current levels. The heart of each module, as shown in Fig. 6.21, is a single-turn transformer which when operated with the secondary open injects inductive impedance. The SW modules are self-powered using the line current and do not require communications among the modules or with a central control center. The modules operate at line potential and do not connect to the ground, eliminating the cost of isolation and substation footprint requirements. The target retail price for smart wires is $100/kVAR injected. Figure 6.22 shows the line utilization levels for a system with and without smart wires.

Fig. 6.20
figure 20

Overview of how smart wires modules will be installed in a transmission line

Full size image
Fig. 6.21
figure 21

A smart wire (SW) module, installed on a transmission line between bus 1 and bus 2, converts a traditional transmission line into a smart wire (SW)

Full size image
Fig. 6.22
figure 22

Line utilization of a test system with and without installation of smart wires

Full size image

Controllable Network Transformer (CNT)

The CNT provides simultaneous control of bus voltage magnitudes and phase angles by augmenting an existing multitap transformer with a fractionally rated converter, as shown in Fig. 6.23. Since power flow control typically requires small changes to system parameters, the converter can be fractionally rated with respect to the transformer and achieve meaningful, independent control of real and reactive power through the transformer. Any real and reactive power combination within the control range of the CNT, shown in Fig. 6.24, is possible. The CNT provides the functionality of both a PST and an LTC transformer. An LTC transformer is often used to adjust output voltage or VARs, but it cannot adjust phase angle. The CNT response is significantly faster than the typical LTC or PST, allowing it to help mitigate transients.

Fig. 6.23
figure 23

Controllable network transformer (CNT) installed on transmission line between bus 1 and bus 2

Full size image
Fig. 6.24
figure 24

CNT control range as a function of the transformer tap range. 0.10 means up to 10% of the transformer voltage is connected to the converter

Full size image

Dynamic Capacitor (D-CAP)

The D-CAP, shown in Fig. 6.25, combines a traditional shunt capacitor with a pair of AC switches and small filtering elements. The switches operate with a duty cycle to effectively change the total capacitance seen by at the bus. Like an SVC, the D-CAP is able to dynamically vary the amount of capacitance. However, unlike an SVC, the D-CAP can react within a fraction of a cycle. The D-CAP configured in a “boost” configuration as shown, also can maintain the VARs delivered to the line, even as the voltage sags under a grid fault. An SVC is unable to deliver this level of grid voltage support. In addition, the D-CAP is able to remove unwanted harmonics caused by other loads connected to the power system, as seen in Fig. 6.26. The ability to remove harmonics is in sharp contrast to the SVC, which creates harmonics. Costly and space-intensive filters must be used to remove the harmonic content generated by the SVC.

Fig. 6.25
figure 25

Dynamic capacitor (D-CAP) connected to bus 1

Full size image
Fig. 6.26
figure 26

Plot of fundamental and harmonic magnitudes in load current (top) and line current (bottom) with D-CAP installed. Without the D-CAP, the load current harmonics would appear in the line current

Full size image

Power Flow Control as an Enabler for Improved Energy Markets

Due to the lack of power flow control, contentious cost allocation methods have been developed to ensure transmission owners recover their costs. Because transmission owners have little incentive to invest in new lines, RTOs/ISOs mandate the construction of transmission lines deemed necessary to relieve pending reliability problems or severe economic congestion. The thresholds to judge proposed transmission projects to relieve economic congestion are conservative, resulting in less investment than would occur if the power flow was more controllable. This leads to higher cost of energy for consumers.

Every merchant transmission project built in the USA thus far has included some form of power flow control. Widespread power flow control would allow the types of market transactions seen in the petroleum and natural gas pipeline industries. Examples include bidding, bilateral transactions, and bulletin boards to match transmission owners with prospective transmission users. With control, a transmission user urgently needing to use a line to deliver to market, such as a wind generator which cannot cheaply store its energy, can pay a premium during windy periods to outbid other would-be users of the line. In this case, a user willing to defer generation, such as coal plant able to store energy by leaving coal unburned, would benefit from lower transmission costs when the renewable resource is unavailable. In contrast, under current transmission operation rules, transmission users of all types are often backed down pro rata based on the amount of long-term transmission capacity reserved. This leads to economic inefficiencies and higher costs.

The ability to inexpensively control power flows along specific paths can lead to significant benefits at the system and market level. The need for new transmission line build to accommodate renewable portfolio standards can be dramatically reduced. Loop flows, that reduce the capacity of the line available for use by the line owner, can be controlled. Improved interarea support during contingencies can improve the reliability and stability of the network. Costs associated with grid build-out and operation can be allocated more fairly. Owners of generation assets can sell energy to interested customers, finding uncongested pathways for power delivery. Overall, the smart and controllable grid is a critical element in realizing a greener and more economically efficient energy market.

Conclusions and Future Directions

This entry has discussed the emerging challenges which will place additional stress on the transmission system and some of the potential methods to relieve these stresses. A number of methods already exist to provide VAR control and power flow control. All of these methods are expensive and have disadvantages, which have limited their adoption. Emerging technologies that use fractionally rated control elements, and are inherently more distributed in nature, may overcome the significant disadvantages, allowing widespread adoption of VAR control and power flow control. These capabilities could transform the operation of the transmission system, enabling the delivery of low cost and renewable energy, and allowing efficient operation of a competitive energy market.