20
Integration of distributed renewable energy systems into the smart grid
Ghanim Putrus
Edward Bentley Electrical Power Engineering at Northumbria University Newcastle upon Tyne, United Kingdom
Abstract
About 30% of all primary energy resources worldwide are used to generate electrical energy. Since the invention of the electric incandescent light bulb in 1879, the growth of electric power systems progressed at an exponential rate, particularly after the development of AC (alternating current) power generation and the transformer. Transforming AC power from one voltage level to much higher levels meant that losses and voltage drops in the supply lines could be kept at acceptable values. The contribution of different renewable energy sources to the electricity generation mix varies from one country to another, but generally this is currently a small proportion of the total installed capacity. The increase in renewable energy generation involves significant challenges in establishing cost-effective and reliable renewable energy systems in addition to solving the technical problems associated with their connection to the grid. In order to understand the impact of increased connection of RES on the grid and the need for smart grid solutions, it is important first to understand how electricity is currently generated, the characteristics of generation from RES, and some aspects of grid control.
Keywords
renewable energy systems
wind energy conversion systems
voltage fluctuation
harmonics
smart grid
islanded operation
microgrids
point of common coupling
20.1. Introduction
About 30% of all primary energy resources worldwide are used to generate electrical energy. Since the invention of the electric incandescent light bulb in 1879, the growth of electric power systems progressed at an exponential rate, particularly after the development of AC power generation and the transformer. Transforming AC power from one voltage level to much higher levels meant that losses and voltage drops in the supply lines could be kept at acceptable values.
The contribution of different renewable energy sources to the electricity generation mix varies from one country to another, but generally this is currently a small proportion of the total installed capacity. However, energy policy targets for 2050 are very ambitious, setting the commitment of some countries to 80% reduction of greenhouse gas emissions below 1990 levels, by 2050 [1]. Renewable energy obligations such as the European 20/20/20 targets are driving a fast growth of renewable energy installations. For example, in the United Kingdom, renewable electricity installed capacity at the third quarter of 2014 was 23.1 GW, which is over 25% of the total installed capacity (around 85 GW). The contribution of renewable energy to electricity generation in this period was about 17.8% of total generation [2,3]. This represents an increase of 24% compared to the same quarter in 2013. The growth of renewable energy generation capacity in the United Kingdom during 2009–2013 is shown in Figure 20.1. As can be noticed, growth of renewable energy in the last few years is steadily increasing.
Figure 20.1 Growth of renewable energy generation capacity in the United Kingdom during 2009–2013.
The increase in renewable energy generation involves significant challenges in establishing cost-effective and reliable renewable energy systems (RES) in addition to solving the technical problems associated with their connection to the grid. In order to understand the impact of increased connection of RES on the grid and the need for smart grid solutions, it is important to first understand how electricity is currently generated, the characteristics of generation from RES, and some aspects of grid control.
20.2. Conventional power generation
Conventional power plant is the general term applied to the production of electrical energy from coal, oil, or natural gas using the intermediary of steam. The generator is usually a synchronous machine having a small number of poles (two or four) and running at high speeds (1500–3600 rpm). The overall efficiency of energy conversion from fuel to electrical is greatly influenced by the poor efficiency of the turbine and condenser. Typical overall efficiency ranges from 30% to 40%. The main features of these conventional plants are their low capital cost per kilowatt installed as compared to other plants and virtually no limit on their size.
The combined-cycle power plant is relatively more efficient and environmentally friendly. It operates in two stages with an overall efficiency of up to 55%. The first stage includes a gas turbine that drives the first alternator and the second stage uses hot exhaust gases from the gas turbine to produce steam through a heat exchanger, which operates a steam turbine coupled to a second alternator. Another way of increasing the overall efficiency of a conventional power plant is to utilize the remaining energy in the steam, after leaving the turbine for industrial processing or central heating (e.g., large areas of towns). In this arrangement, usually referred to as combined heat and power, the plant overall efficiency is increased to around 90%.
These plants are usually centrally controlled at national level, taking into consideration their capacity and dynamic characteristics, in order to match the overall demand, which continually varies but follows a fairly predictable pattern. This match is necessary in order to keep the frequency of supply constant. To meet a particular demand, the national grid control center selects a number of generating units with sufficient capacities to meet the demand and some spare capacity (generation dispatch). Units selected (committed) must be at appropriate locations in order to reduce transmission losses. To minimize generation costs, units committed must have the lowest production costs. Once a unit has been committed, it must be allowed to deliver at least the minimum power it can handle. The levels of units loading (generation schedule) is determined and thus controlled by the turbines’ governors. Also, minimizing the cost of starting up and shutting down units is an additional constraint. Therefore, generating stations are normally listed in the order of economic merit, which is used to determine when each unit is to be committed and its output power at any loading conditions (generation scheduling). The incremental fuel cost, defined as the additional cost to increase the output of a generating plant by 1 MWh, is usually used as the benchmark for unit loading. For optimum economical operation, any demand for extra power should be met by a unit with the lowest incremental fuel cost, until this exceeds that of another unit [4].
A power generation plant may be assigned as a PQ plant, where the plant is assigned a fixed real and reactive power generation, or as a PV plant, where the plant is assigned a fixed real power generation and reactive power is varied to keep the voltage at the plant constant. A plant may be assigned to provide a balancing service to control frequency and voltage, where the active and reactive power outputs are variable.
20.3. Electricity generation from renewable energy resources
Renewable resources such as water, wind, biomass, tidal, wave, and solar energy are available at zero cost, are pollution free, and inexhaustible. At present, the cost of generation per unit of energy (MWh) from these resources (with the exception of hydroelectric and wind power plants) is relatively high compared to fossil-fuel sources. However, advances in technology and economy of scale are bringing down the cost of generation from RES.
Hydroelectric power plants have been used for a long time. However, the cost of construction of these plants is very high, in particular the civil engineering works. The pumped-storage plant is similar to a normal hydro plant except that two (upper and lower) reservoirs are used and the turbine–generator set is designed to also operate as a motor–pump set. When electrical energy is most needed, during peak load, the plant operates in generating mode where water flows from the upper reservoir to the lower one and electric power is generated and fed to the system network. Water stored in the lower reservoir is pumped back to the upper reservoir during off-peak periods when electricity is cheap and the plant is made ready for the next peak load. The overall efficiency of the pumped-storage plant (defined as electrical energy output/electrical energy input) is about 67%.
In further sections, main characteristics and grid interface controllers for RES are described, with emphasis on photovoltaic (PV) and wind energy conversion systems (WECSs). Although the characteristics of other types of RES (such as biomass, tidal, and wave) may be different, the way they generate electricity, and their connection and impact on the grid are similar.
20.3.1. Photovoltaic systems
The performance of a PV module is usually determined by its I–V characteristics, shown in Figure 20.2. These characteristics define the power output at any load point, giving the power curve shown in this figure. As can be seen, maximum power is only generated at a certain voltage level. This voltage varies with incident radiation, shading, temperature, and module soiling conditions. Therefore, it is important that a PV module is operated at (or around) this value in order to maximize energy capture.
Figure 20.2 I–V characteristics and output power of a PV module.
(20.1)where,
ISC is the short circuit current,
VOC is the open circuit voltage,
VMPP and IMPP are the cell voltage and current at maximum power point (MPP), which may be found from manufacturer data sheets.
A maximum power point tracker (MPPT) controller is usually used to determine the voltage, at which a PV module is able to produce maximum power and shifts the operating point to that voltage, regardless of load voltage. The MPPT is a high efficiency DC-to-DC converter that presents an optimal electrical load to a solar module (or array) and produces a voltage suitable for the load, irrespective of voltage variation on the PV module side (due to variations in irradiance, temperature, etc.). The MPPT decouples the PV module voltage from the load voltage, that is, it allows the PV to operate at MPP voltage (at the controller input) while delivering the required load voltage at the output of the converter. In addition to providing MPPT and “matching” functions, the converter is usually of boost type, which steps up the PV module voltage to allow operation at higher voltage levels, thus reducing power losses. The converter also provides flexibility in selecting rated voltages of the modules and batteries.
There are different approaches to control the DC-to-DC converter to track the MPP. They vary from a simple current controlled system where the PV or load voltage is assumed constant and the MPPT tries to maximize the current to a more sophisticated system, where the output power generated is optimized, such as incremental conductance, perturbation, and observation (P and O) and hill climbing MPPT controllers, which may employ digital signal processing techniques [6]. The principle for controlling the DC-to-DC converter is to change the duty ratio (mark to space ratio) of the converter such that the PV module voltage is controlled to occur at a value that produces maximum output power.
In standalone PV systems, the operating point is largely determined by the battery voltage, which is usually close to the MPP. As the battery voltage varies (depending on its charging state), the operating point also varies and may move away from the MPP. In practice, the battery is not allowed to fully discharge, hence the voltage is kept reasonably constant and close to the MPP. Therefore, for standalone systems, MPPT is desirable (to maximize the PV energy output), but not essential, as the PV module operates most of the time around the MPP.
20.3.2. Wind energy conversion systems
An alternative source of electrical energy that is becoming increasingly popular is the wind power plant. Due to their commercial viability, WECSs are currently the fastest growing renewable energy resources. A single wind-powered generator produces a limited amount of electrical power and hence a large group of generators is normally deployed and is called a “wind farm.” As with many other renewable energy resources, wind energy has the disadvantage that degree and period of its availability are uncertain. Also, the balance between the capital costs of a wind turbine (or a farm) and the revenue from the electricity generated over the lifetime of the turbine must be carefully looked at.
In WECSs, the wind pressure rotates turbine blades attached to a shaft, which is coupled to the rotor of a generator. Large wind turbines usually employ a mechanical gearbox to increase the speed of the generator (usually induction type). Smaller wind turbines (below 10 kW) usually use permanent magnet generators.
The mechanical power output of a wind turbine (and hence electrical power output) is proportional to the cube of wind speed and may be expressed as in Ref. [7].
(20.2)where Pa is captured power by the wind rotor, R is the radius of the rotor in m, ρ is the air density in kg/m3, v is the speed of the incident wind in m/s, Cp is the power coefficient, which for a given wind rotor depends on the pitch angle of the wind rotor blades and on the tip speed ratio (λ) defined as,
(20.3)where ω is the rotational speed of the rotor in rad/s.
Equations (20.2) and (20.3) show that the output power of a wind rotor is a function of the wind speed (v) and rotational speed (ω) of the rotor and this relationship may be represented, as shown in Figure 20.3 [8]. Therefore, in order to maximize energy capture from the wind, it is necessary to regulate the wind turbine speed to match the optimal speed (ω1 or ω2) for any particular wind speed (v1 or v2), as shown in Figure 20.3.
Figure 20.3 Wind turbine characteristics.
There are two main types of wind turbines, namely, fixed speed and variable speed. Fixed-speed wind turbines employ a complex variable-pitch control mechanism of the blades and gearbox to optimize the energy captured. Consequently they have lower efficiency and require regular maintenance, as compared to variable-speed wind turbines, which employ power electronics converters to optimize the output power performance (without the need for aerodynamic controllers). In fixed-pitch variable-speed wind turbines, the electrical torque of the generator is usually controlled to maintain optimum wind rotor speed at a particular wind speed for maximum power output. A power electronic MPPT controller, similar to that used for PV systems, is usually used for this purpose.
In general, variable-speed wind turbines are characterized as having higher efficiency and better reliability. Hence, they are becoming more popular, particularly for small-scale applications.
The rotor in large wind turbines is usually a horizontal-axis, which provides better wind energy capturing capability. However, in turbulent wind environments, control of a horizontal-axis rotor to continuously follow the wind direction and operate at the MPP becomes difficult due to the fluctuation of wind speed. Hence, in such areas it is better to employ vertical-axis turbines, as shown in Figure 20.4.
Figure 20.4 Horizontal and vertical axes small wind turbines.
20.3.3. Other renewable energy resources
Other renewable energy resources such as water, biomass, tidal, and wave are also available. Hydroelectric, pumped-storage, and biomass plants have been in use for a long time. These are usually large plants and centrally controlled, similar to conventional power plants. Tidal energy is generated by the rotation of the earth within the gravitational attractions of the Moon and the Sun, which causes the sea level to change. Power generation involves extracting the kinetic energy from moving water, set up by tidal flows. The output power is periodic and reasonably predictable. Waves are produced by the action of wind on the surface of the seas. Power is generated when the main mechanical structure of the turbine moves with the waves, causing oscillations or driving hydraulic motors, which drives an asynchronous generator. Wave energy can be considerable, up to 70 MW/km of shoreline coastal processes [9]. However, one major problem is that wave energy is not consistent, and therefore the output power is variable and largely unpredictable.
20.4. Grid connection of distributed RES
PV panels generate DC output and hence they need inverters to connect to the grid. Generators used with wind turbines can be asynchronous or synchronous. An asynchronous (induction) generator may be connected directly to the grid (usually through a soft starter to reduce transient currents) while a synchronous generator is connected through power electronics converters. Converters are also used for power generation from wave and tide. Generally, these converters generate harmonics, which can be harmful to the grid (e.g., extra losses, overheating, etc.) and therefore they have to meet applicable harmonic emission standards, as explained in Section 20.7.
20.4.1. Grid interface of PV systems
There are two types of PV systems, namely standalone- and grid-connected systems. In the former, PV modules are used to supply individual loads in isolation from other sources of electricity (the grid). In the latter, PV system is connected to and operated as a part of the main grid. As a PV system generates DC voltage, a DC/AC inverter is required if the system is to supply AC loads or is to be connected to the grid.
The main components of a PV system, which is suitable for both standalone and grid-connected operation are, PV modules, rechargeable batteries, a control unit, and an inverter (if AC output is required), as shown in Figure 20.5 [6].
Figure 20.5 Main components of a PV system.
An inverter is the interfacing device between the PV modules and the grid. Line-commutated inverters usually employ thyristors as the power electronic switching devices. These devices are not self-commutated and rely on the grid voltage to operate (commutate). Therefore, line-commutated inverters are not suitable for standalone PV systems. They draw reactive power from the grid (for thyristors commutation), hence they increase losses and reduce overall efficiency.
Self-commutated inverters employ self-commutated power electronic switching devices (usually MOSFETs, for low power applications and IGBTs, for medium to high power applications), as shown in Figure 20.6. These devices are switched on or off by a gate signal, thus the grid voltage is not necessary. Therefore, these inverters are suitable for both grid-connected and standalone PV systems. They do not need reactive power to operate, hence fewer power losses and better overall efficiency. In fact, the inverter may be used to generate reactive power, if required.
Figure 20.6 Typical self-commutated PV inverter with low frequency transformer and (PWM).
Voltage source inverters (VSI) are usually used in PV systems and these may be voltage controlled, where the output voltage is controlled such that it follows the desired reference signal while the current is load dependent. Alternatively, the VSI can be current controlled, where the output current is controlled to follow the desired reference signal. The current controlled VSI allows control of the power factor (i.e., unity or leading power-factor operation) and fault current (i.e., low fault currents) and therefore it is widely used for grid-connected systems. In standalone systems, the output current is determined by the load, so a voltage controlled VSI is essential.
20.4.2. Grid interface of WECS
20.4.2.1. Fixed-speed wind turbine
In fixed-speed wind turbines, a variable-pitch fixed-speed wind turbine drives (through a gearbox) an asynchronous (induction) generator, which is directly connected to the grid, as shown in Figure 20.7. The gearbox is needed to increase the speed and together with the variable pitch they control the speed and optimize the energy captured from the wind. The gearbox reduces the overall efficiency of the WECS and requires maintenance. A soft starter is usually used to energize the generator and reduces transient currents while starting. Also, switched capacitors are needed to supply some of the reactive power demand of the induction generator.
Figure 20.7 Fixed-speed wind turbine.
When using an induction generator, the machine generates AC power by running at a small slip speed above synchronous speed. Hence, the turbine speed variations are also small, since any increase in wind speed increases the torque. Typical generator voltage is ∼690 V, so a three-phase transformer is needed to step up the voltage. This is a simple (electrical) and relatively cheap system. This system tends to overspeed during faults in the grid and network operators expect the wind turbine to meet their “fault ride-through” requirements (i.e., the turbine system continues to work for a short time during faults on the grid) [10].
20.4.2.2. Variable-speed wind turbines
An alternative scheme employs power electronics converters to connect the generator to the grid. The converter decouples the generator from the grid, maintaining synchronization while allowing the turbine speed to vary depending on wind conditions. Thus, there is no need for pitch control or it can be very simple. Compared to fixed-speed systems, variable-speed systems have improved system efficiency, energy capture, and reduced mechanical stresses. Hence, they are becoming more popular. Two versions of this system are common, and are briefly described as follows.
1. Wide-range variable-speed wind turbines with synchronous generator: in this system, a multipole-synchronous generator, designed for low-speed operation, is usually used in order to eliminate the need for the gearbox. Connection to the grid is through a power electronics converter, as shown in Figure 20.8, which completely decouples the generator speed from grid frequency [11]. The rating of a power converter corresponds to the rated power of the generator plus losses. The generator can be electrically excited through a field winding or a permanent magnet machine. Small-scale wind turbines usually employ permanent magnet generators. The generator produces variable frequency power, which is first rectified (AC-to-DC) and then converted to 50 Hz by using a self-commutated inverter (DC-to-AC).
2. Doubly fed induction generator (DFIG): in this case, a wound rotor induction (asynchronous) generator is used and variable-speed operation is obtained by injecting controllable voltage into the rotor at slip frequency [11]. The rotor winding current is fed through slip rings from a variable frequency power electronics converter, which is supplied from the grid, as shown in Figure 20.9. The power converter decouples the grid frequency from the rotor mechanical frequency, enabling variable-speed operation of the wind turbine. The power rating of this converter is determined by the rotor power, which in turn is determined by the operating slip of the generator, which is usually very small, resulting in a converter rating, which is a fraction of the full power (usually about one-third of the generator’s rated power).
Figure 20.8 Wide-range variable-speed WECS.
Figure 20.9 DFIG.
20.4.3. Grid interface of other RES
Similar to conventional power plants, synchronous generators are used in hydroelectric, pumped-storage, and biomass plants. The plants are usually centrally controlled and “dispatchable.” Therefore, no special grid connection issues will normally arise unless the power plant, for logistical reasons, is located in a weak part of the grid.
Power generation from tidal and wave energy is not steady and therefore not centrally “dispatchable.” However, it is worth noting that tidal power is periodic and reasonably predictable. Power is usually extracted from wave energy by using linear or asynchronous generators operating at variable speeds. Similar to WECSs, power electronics converters are usually needed to connect the generators (driven by tidal and wave turbines) to the grid. Therefore, grid connection considerations are similar to those applicable to WECS.
20.5. Distributed renewable energy sources
Electric power networks have evolved over the years to have large centrally controlled generators connected to the high voltage side of the network. Consequently, power flow is from the high voltage side (where generators are connected) to the low voltage side of the network (where medium and small size loads are connected).
Renewable energy sources connected to the grid can be either relatively very large plants (e.g., hydro and wind farms) with hundreds of MW power capacity or small plants less than 50 MW. The former is usually connected to the transmission system and is centrally controlled. The latter is relatively small, compared to central generation and is neither centrally planned, nor dispatched. Distributed RESs are relatively small and connected to the distribution network, typically below 33 kV (close to the point of use). This usually relates to noncentralized generators, ranging in size from around 1 kW to around 50 MW, and are referred to as distributed, embedded, or dispersed generation. Examples of distributed RESs are large wind turbines (or farms) up to 50 MW and small-scale wind and PV systems (or microgenerators) [12].
20.5.1. Benefits of distributed RES
Potential benefits of distributed RESs may be summarized as follows.
1. Energy efficient, reduced CO2 emissions, and environment friendly.
2. Reduced network (infrastructure) and central generation capacities.
3. Generation closer to loads, thus reduced power losses in the network.
4. Diversity in terms of energy supply and location, thus a more secure and reliable supply.
To encourage the uptake of low carbon technologies, several countries have introduced incentives to support generation from distributed RES up to 5 MW. This is usually referred to as the feed-in tariff (FIT), which is a payment made to customers for both generation and export of produced renewable energy.
20.5.2. Impact of distributed RES on the grid
Currently, distributed RES amounts to only a small proportion of the total network generating capacity. Hence, the impact on network performance is not significant. However, with the increased number of new and renewable energy resources being connected to the grid, these will start to have an impact and create potential problems in existing power networks (which were not designed to accommodate RES).
Existing power networks rely on central generation and a national transmission grid that enables central control of the system at a national level by means of accurate statistical prediction of overall demand (allowing for diversity). Loads vary with respect to size and time but are fairly predictable. To ensure continuity of supply, each supply utility maintains its own generation margin, typically about 10% of peak demand or the size of the largest generator. Large centrally controlled generators are connected to the transmission system, which is highly automated and centrally controlled. Loads are connected at the distribution network, which has limited automation, for example, transformer tap changers and autoreclosing of circuit breakers. This arrangement results in power flow from generating plants via the transmission and distribution networks to medium and smaller sized loads, that is, directional power flow, as shown in Figure 20.10. As power flow is directional, the quality of power supply (frequency and voltage control) is maintained by a central control, and utilities have developed the necessary skills and procedures to do this.
Figure 20.10 Existing power distribution networks.
Significant connection of RES at the distribution level will change the structure and power flow in future power networks. Figure 20.11 shows a possible structure of future power networks. Studies [13–16] have shown that significant deployment of distributed RES could lead to a bidirectional power flow in distribution networks and, unless adequately controlled, this may have significant impacts on the quality of power supply. Understanding these effects is very important, as they will influence the way future power networks will be designed and operated.
Figure 20.11 Future power distribution networks.
Potential problems for distribution networks from deployment of distributed RES have been identified as follows [13–16].
1. Change in voltage level and violation of statutory limits: these limits are set up according to the distribution code in each country, for example, in Europe these are –6% to +10% for the low voltage (400 V) level and ±5% for the higher voltage levels.
2. Difficulties in power flow control and potential reverse power flow: distributed RES will influence power flow in the network and this may lead to overloads on network feeders and equipment, depending on demand and generation profiles. Transformer tap changers normally have “line drop compensation” and this will malfunction with reverse power flow.
3. Increase in short circuit contribution and fault levels: network equipment has a defined fault handling capacity, usually for a short time (until protection operates and isolates the faulty part, which usually takes 0.2–1.0 s). Increase of fault level depends on the type of RES. When using rotating machines, for example, wind turbines, the increase in fault level can be significant (depending on the penetration level). When using power electronics converters to interface RES to the grid, the contribution to fault current can be made negligible, as the converter can be switched off almost immediately as soon as a fault is detected.
4. Issues with protection management: distributed RES will cause additional “uncontrolled” power flow in the network during faults, and this will affect existing relay settings and may also jeopardize network protection. It will also make it difficult to set relays, as fault current seen by the relay keeps changing depending on the state of generation of the distributed generator (DG) (whether on or off). Another issue to consider is that of islanding operation. As described in Section 20.5.3, according to current legislations, distributed RES can only operate when connected to the grid, that is, it must shut down if connection to the grid is lost (likely due to faults on the grid). Therefore, the DG must have anti-islanding protection or a loss-of-mains protection installed.
5. Potential voltage imbalance, which is specific to single-phase small-scale distributed generation.
The anticipated effects of large deployments of distributed RES need to be properly analyzed and appropriate measures should be taken in order to avoid potential problems in future power networks. Factors to consider include the following:
1. Type and capacity of the generator and voltage level for connection to the grid (point of common coupling, (PCC)).
2. Technology and type of grid interface.
3. Availability of power generated and reliability of plant.
4. Variation of power generated with time and predictability of variation.
5. Regulations and grid code for connection.
20.5.3. Islanded operation and microgrids
The distributed generation owner needs to ensure that the system is safe for installers, operators, and users (e.g., earthing, compliance with standards, labeling, isolation, etc.). Current regulations for power systems operation require that distributed generation be disconnected in the event of a fault on the grid or if the grid voltage drops [12,17]. This is usually done if any of the following exceeds a preset level: frequency, rate of change of frequency (ROCOF), or voltage level. The requirement for disconnection is mainly due to issues related to safety and control of the grid and the island. Therefore, it is not permitted to operate an isolated local power area (an “island”) electrically separated from the grid, even though this would permit DGs to carry on supplying local demand. Under IEEE Standard 1547 [16], DGs rated below 10 MVA must be disconnected in the event of a major fault occurring on the grid. There is no provision for setting up a power “island” in this situation, however desirable this might be from the point of view of continuity of supply. The reasons for this are related to safety and technical issues. If an isolated area continued to be powered from distributed RES internal to the area, repair workers will be at risk from unexpected dangerous voltages and there is also the risk of damage to distribution equipment during reconnection. The other safety issue is that the grid operator cannot guarantee adequate earthing in the island, if the grid connection is lost. The technical issue is with regard to balancing supply and demand to maintain constant voltage and frequency in the island.
The opportunity of setting up “islands” was not thought important enough when IEEE 1547 was drafted to be worth special provision, since distributed generation was in its infancy. IEEE Standard 1547.4 [18], written more recently, considers the deliberate retention of distributed generation within an “island” when a system fault is present elsewhere. Sources of distributed generation if connected to the grid in conformance to the standard must be capable of operating in “island mode,” thus benefiting adjacent consumers. Clearly for “islanding” to work, distributed generation within the island must be capable of supplying the load requirements, including those for reactive power. The standard calls for provision of distributed control and monitoring apparatus to ensure that power supply and demand are kept in equilibrium, and that frequency and voltage limits are observed. The cost of the additional equipment is considerable, restricting the degree to which “islanding” has so far been adopted. Autonomous control of voltage and frequency is only permitted during operation as an “island”, and is not needed when the area is reconnected to the remainder of the grid.
Microgrids are power “islands” that are designed to operate as separate entities separated from the grid for extended lengths of time. Typically a microgrid would contain its own sources of generation plus power storage and may be uninterruptible power supplies [19,20]. Where consumers require very high levels of supply reliability such a scheme may be attractive. Possible users may include hospitals and defense establishments. Microgrids operating as “islands” would ensure continuity of supply even if there were failures in the grid [21]. At present, developments in microgrids are in their early stages. As of 2011 there were only 160 ongoing microgrid projects worldwide, involving some 1.2 GW [22]. Microgrids suffer from the same cost drawbacks as “islanding” systems conformable with IEEE 1547.4 [18]. The benefits, in terms of reliability of supply, may be obtained using conventional backup generators, and this approach tends to be simpler and cheaper.
20.6. Voltage control in power networks
As explained in Section 20.5.2, the flow of power in existing power distribution networks is from higher to lower voltage levels. Therefore, utilities estimate the voltage drop in the feeders based on the network configuration and assume that minimum and maximum loading on the feeders will remain relatively constant. Accordingly, a transformer’s off-load tap position and on-load tap changer’s (OLTC) initial position are set to raise the voltage to a level that will compensate for the voltage drops.
20.6.1. Voltage drop in feeders
It is usual practice in power systems to represent a part of a network by its Thévenin equivalent circuit. A Thévenin circuit comprises of a voltage source (e.g., sending-end voltage of a feeder) and a series impedance (e.g., impedance of the feeder), as shown in Figure 20.12, where a line represented by series impedance, R + jX, is supplying a load drawing a power, P + jQ.
Figure 20.12 Equivalent circuit of a power network.
The phasor diagram of the circuit is shown in Figure 20.13 and the line voltage drop may be derived as follows.
Figure 20.13 Phasor diagram.
From the phasor diagram,
(20.4)
(20.5)Since P = VRI cos φ and Q = VRI sin φ,
(20.6)Normally, 
(20.7)Hence, magnitude of the line voltage drop is approximately given by,
(20.8)and the angular shift is determined by,
(20.9)where ∆V is the voltage drop, R and X are the resistance and reactance of the feeder, P and Q are the active and reactive power flow, and V is the grid voltage.
The effect of the line resistance (R) is often significant in low voltage (400 V) distribution networks. However, on the high voltage side of the network (11 kV and higher), the X/R ratio is normally high and the effect of the feeder resistance may be ignored.
Therefore, normal variations in the power generation of RES may cause voltage variations, which will be experienced by other consumers connected to the PCC (a substation that supplies several points is referred to as the, PCC).
20.6.2. Voltage control using tap changers on distribution transformers
Control of transformer’s tap changer is the most popular form of voltage control at all voltage levels in a power system. This is based on changing the turns ratio of a transformer, hence the voltage in the secondary circuit is varied and voltage control is obtained. The tap changer is usually placed on the high voltage winding as the current is lower and this minimizes the current handling requirements and stress during operation of the tap changer.
Figure 20.14a shows a schematic diagram of an off-load tap changer that requires disconnection of the transformer when the tap setting is to be changed. Most transformers have an OLTC, which is shown in basic form in Figure 20.14b. In the position shown, the voltage at the LV side is at its maximum and the current divides equally in to the two halves of the coil L, resulting in zero resultant flux and minimum impedance. To reduce the voltage at the LV side, S1 opens and the total current passes through the other half of the reactor L. Selector switch A then moves to the next contact and S1 closes. A circulating current now flows in L superimposed on the load current. Then, S2 opens and B moves to the next tapping; S2 then closes and the operation is complete.
Figure 20.14 Distribution transformers.
To avoid large voltage disturbances, the voltage change between taps is normally small, about 1.25% of the nominal voltage. The total range of tapping varies with transformer usage; a typical figure for generator transformers is +2% to −16% in 18 steps.
Voltage control using OLTCs is usually based on voltage measurement at the transformer location. Some OLTCs also measure the current through the transformer to adjust for the variation in load currents (line compensation). Existing controllers for line compensation are designed to measure the current in one direction only, as power flow in existing power networks is in one direction.
As explained in Section 20.5.2, with significant penetration of distributed RES, power flow in the network may be altered, with potential reverse power flow and voltage rise, which falls beyond the control limits of the OLTC, as shown in Figure 20.15. As can be seen, the voltage rise is most significant at the remote end of the feeder where the generator is connected. The amount of voltage rise depends on the level of RES penetration level, operating p–f, loading conditions (peak or off peak) and distance from the main transformer. Violation of statutory upper voltage limit is likely to occur at minimum loading condition and this worsens when the generator is operating at lagging p–f (supplying reactive power).
Figure 20.15 Voltage rise due to distributed generation.
This rise may be felt at the point of the RES connection and possibly at the PCC with other sensitive loads. Too many distributed RES connected to the network increase the risk of voltage rises. Therefore, the philosophy of voltage control in future distribution networks will need to change in order to allow for large penetration of distributed RES.
20.6.3. Voltage fluctuation (flicker)
When connecting distributed RES to the grid, one needs to ensure that the system does not interfere with other users of the power network (flicker, harmonics, transients, fault in-feed, etc.). Voltage fluctuations or “flicker” are rapid changes in voltage magnitude within the statutory limits of the usual slow variations of voltage (±5% of nominal value). These fluctuations can cause noticeable variations in lighting (flicker) and interrupt the operations of some electronic controllers. Research has shown that some loads are very sensitive to certain fluctuations of voltage magnitude such that a fluctuation of 0.5% at a frequency of 5–6 Hz will cause annoying flickers, as shown in Figure 20.16 [23].
Voltage fluctuations can occur due to rapid change in current flow in the grid. Power generated from distributed RES (e.g., wind and wave) can be variable and intermittent with frequent and repetitive changes. Also, clouds passing by a PV system may produce a rapid change in its power output. Such generation results in a large change in active and reactive power exchanged with the grid and may give rise to voltage fluctuations at the PCC.
Figure 20.17 shows two alternative ways of supplying a 1 pu purely inductive load XL. This load may be connected to busbar A or B, both of which are connected to supply point C via transformers T2 and T3. The effect of switching on this load on the voltage level at busbar A where another (sensitive) load (load 1) is already connected is analyzed. To simplify the analysis, load 1 is assumed to be very small relative to the inductive load under consideration.
Figure 20.17 Two alternative ways of supplying a 1 pu purely inductive load.
If the inductive load XL is connected to busbar A and is switched on, the voltage at busbar A will be,
(20.10)That is, when the inductive load XL is switched on, a voltage drop of approximately 9% is produced at busbar A. Note that as load 1 is assumed to be very small, the voltage at busbar A before connecting the inductive load is equal to 1.0 pu (grid voltage). Clearly, this voltage drop will also be experienced by other consumers on this busbar (load 1). However, if load XL is connected to busbar B, the same voltage drop will be produced at busbar B, but only a fraction of this voltage drop will be seen at busbar A, through the voltage changes produced at busbar C (the PCC). The voltage at busbar C is:
(20.11)Therefore, when load XL is connected at busbar B and switched on, it will produce less than 4% voltage drop at busbar A.
These two alternatives represent a change in the PCC for the loads (load 1 and load XL) from busbar A to busbar C. That is, moving the PCC to a higher voltage level will significantly improve the quality of supply in the presence of disturbing loads.
Variations in voltage level at the PCC can be minimized by providing a “strong” system, that is, one with lower source impedance 
, 
, and 
, or by using voltage control devices. Slow acting devices such as automatic voltage regulators for generators, transformers, OLTCs, and conventional reactive power compensators are relatively slow and may not provide smooth control. Hence, they are inadequate for compensation of rapid fluctuations in voltage level (flicker). To deal with such fluctuations fast-acting reactive power compensators are necessary, such as those that employ power electronics (FACTS) technology [24].
, , and , or by using voltage control devices. Slow acting devices such as automatic voltage regulators for generators, transformers, OLTCs, and conventional reactive power compensators are relatively slow and may not provide smooth control. Hence, they are inadequate for compensation of rapid fluctuations in voltage level (flicker). To deal with such fluctuations fast-acting reactive power compensators are necessary, such as those that employ power electronics (FACTS) technology [24].20.7. Power quality and harmonics
20.7.1. Definitions and impact of RES
Power quality (PQ) is a generic term often used in relation to unwanted disturbances of the electricity supply. A PQ event (disturbance) is defined as any deviation (steady state or transient) of voltage or current waveforms from a pure sinusoidal form of a specified magnitude [25]. PQ events, as defined by IEEE Standards 1159–1995, include steady-state events (long-term) abnormalities in the voltage/current waveform as well as transient, short and long duration events that are sudden abnormalities with time scales ranging from a few nanoseconds to several minutes [26,27]. These standards define harmonics as “sinusoidal voltages or currents having frequencies that are integer multiples of the frequency at which the supply system is designed to operate (termed the fundamental frequency; usually 50 or 60 Hz)” [27]. Harmonics are superimposed on the fundamental component and cause waveform distortion.
In the recent years, there has been an increased number of PQ related problems. This is mainly due to the rapid growth in the use of nonlinear loads, mainly power electronic equipment that employs switching devices such as switched mode power supplies, variable speed drives, converters for connection of distributed RES, etc. This equipment usually includes sophisticated control, which is also sensitive to PQ disturbances. The most common PQ problem associated with the use of power electronics converters used for interfacing distributed RES to the grid is to do with harmonics, which is described in this section.
Power electronics converters used for grid connection and control of distributed RES are usually designed and tested to comply with existing standards and therefore should not produce significant PQ events. However, the interaction of different converters during normal operation may result in amplification of certain harmonics or events, which will be experienced by other consumers connected to the PCC. The consumer has a right to receive a supply free from significant distortion. Equally, the user must not cause distortion, which will reflect upon other consumers.
20.7.2. Harmonics effects and solutions
The effects of harmonics and the degree of susceptibility to harmonics depend upon the characteristics of the equipment (those that generate harmonics and those that are affected by them) and also on the specifications of the supply network. In general, possible effects of harmonics can be summarized as follows [28]:
1. Extra losses (I2R), heating and overloading of power system equipment, for example, cables and transformers. In three-phase four conductor systems, triplen harmonic components add in the neutral conductor, causing overloading of the neutral conductor unless it is oversized.
2. Harmonics giving rise to resonance between the inductive and capacitive impedances in the system at harmonic frequencies. Consequently, high current or voltage may appear at points in the system causing damage to equipment.
3. Maloperation of control and protection equipment normally designed to function with a sinusoidal wave, for example, protective relays and controllers of adjustable speed drives or any electronics equipment that use the mains zero crossing point as a timing signal.
4. Interference with communication systems caused by induced noise from harmonic currents and voltages. Telephone systems may be subject to interference where the lines run adjacent to power lines carrying harmonic currents.
5. Metering and instrumentation affected by the presence of harmonics, particularly if resonant conditions occur.
6. Rotating machines experiencing overheating, noise, and torque pulsation in the presence of excessive harmonic distortion. The possible shaft torsional-vibration problems can be harmful to the motor loads, particularly for critical processes. Harmonics can also set up resonant conditions if the natural frequency of the mechanical system is excited by them.
The best way to deal with harmonics is not to generate them. Power electronics converters may be designed to produce low or no harmonics by using appropriate control, for example, using high pulse number, pulse width modulation (PWM) control, etc. If this proves difficult or expensive then filters may be used. Filtering of harmonics involves the provision of a low impedance path to ground for the harmonic frequencies of interest, while providing high impedance for the basic mains frequency. Such a circuit can be made using an LCR “acceptor,” where the inductor is in series with a capacitor. At resonance the impedance of the circuit is basically that of the resistive component only, which can be made low. An example of this procedure, for rejecting the seventh harmonic of 50 Hz (350 Hz), carried out through OrCAD simulation, is shown in Figure 20.18.
Figure 20.18 Harmonic current flow.
(a) Power circuit and (b) harmonic spectrum.
(a) Power circuit and (b) harmonic spectrum.
20.7.3. Limits of harmonic current distortion
Under the provisions of IEEE 519, recommended limits for maximum harmonic current distortion (in percent of the maximum demand current IL) for individual order odd harmonics are given in Table 20.1 [29]. Here, TDD is the total demand distortion defined as “the ratio of the root mean square of the harmonic content, considering harmonic components up to the 50th order and specifically excluding interharmonics, expressed as a percent of the maximum demand current” [29].
Table 20.1
Current distortion limits (in percent of IL) for general distribution systems (120 V to 69 kV)
| Order | 3 ≤ h < 11 | 11 ≤ h < 17 | 17 ≤ h < 23 | 23 ≤ h < 35 | 35 ≤ h ≤ 50 | TDD |
| Limit | 4% | 2% | 1.5% | 0.6% | 0.3% | 5% |
Adapted from Ref. [29].
TDD, total demand distortion.
Even harmonics are limited to 25% of the odd harmonic limits.
Under the provisions of the UK G5/4, recommended limits for maximum voltage distortion in a 400 V system are given in Table 20.2 [30].
Table 20.2
Planning levels of harmonic voltages in 400 V systems
| Odd harmonics (nonmultiples of three) | Odd harmonics (multiples of three) | Even harmonics | |||
| Order “h” | Harmonic voltage (%) |
Order “h” | Harmonic voltage (%) |
Order “h” | Harmonic voltage (%) |
| 5 | 4 | 3 | 4 | 2 | 1.6 |
| 7 | 4 | 9 | 1.2 | 4 | 1.0 |
| 9 | 3 | 15 | 0.3 | 6 | 0.5 |
| 13 | 2.5 | 21 | 0.2 | 8 | 0.4 |
| 17 | 1.6 | >21 | 0.2 | 10 | 0.4 |
| 19 | 1.2 | 12 | 0.2 | ||
| 23 | 1.2 | >12 | 0.2 | ||
| 25 | 0.7 | ||||
| >25 | 0.2 + 0.5(25/h) | ||||
Adapted from Ref. [27].
Adapted from [27] Table 2, page 10.
IEC Recommendations for voltage harmonic levels at low and medium voltage levels are given in Table 20.3, with a maximum total harmonic distortion of 8% [31].
Table 20.3
Compatibility levels for individual harmonic voltages in low and medium voltage networks (percent of fundamental components) [3]
| Odd harmonics (nonmultiples of three) | Odd harmonics (multiples of three) | Even harmonics | |||
| Order “h” | Harmonic voltage (%) |
Order “h” | Harmonic voltage (%) |
Order “h” | Harmonic voltage (%) |
| 5 | 6 | 3 | 5 | 2 | 2 |
| 7 | 5 | 9 | 1.5 | 4 | 1.0 |
| 11 | 3.5 | 15 | 0.4 | 6 | 0.5 |
| 13 | 3 | 21 | 0.3 | 8 | 0.5 |
| 17≤ h ≤ 49 | 2.27 × 17/h – 0.27 | 21< h ≤ 45 | 0.2 | 10 ≤ h ≤ 50 | 0.25 × 10/h + 0.25 |
20.8. Regulations for connection of distributed RES to the grid
When connecting a distributed RES to the grid, one needs to ensure that the system does not interfere with other users of the power network (flicker, harmonics, transients, fault in-feed). Also, ensure that the system does not present a danger to network operation (loss of mains disconnection). The owner of the distributed RES needs to ensure that the system is disconnected from the grid, properly shuts down, and protects itself under grid fault conditions. This is done if any of the following exceeds a preset level, frequency, ROCOF, or voltage level. Also, the owner needs to ensure that the system is safe for installers, operators, and users (e.g., earthing, compliance with standards, labeling, isolation, etc.).
DGs must comply with relevant legislations in the country where it is installed, in addition to any national grid and distribution codes.
Current legislations for connection of DG in the United Kingdom are Engineering Recommendations G75, G59, and G83, which may be summarized as given in Table 20.4 [32].
Table 20.4
UK engineering recommendations that cover the connection of distributed generation to the electrical distribution network
| Engineering recommendations | Generator rated power (Prated) | Voltage at PCC |
| G75 | Prated ≥ 5 MW | Vrated ≥ 20 kV |
| G59 | 11.1 kW < Prated < 5 MW | 400 V ≤ Vrated ≤ 20 kV |
| G83 | Three-phase: Prated ≤ 11 kW | Vrated = 400 V |
| Single-phase: Prated ≤ 3.7 kW | Vrated = 230 V |
Adapted from Ref. [32].
In the United States, DG connection requirements may be summarized as given in Table 20.5, pursuant to FERC Order 2003a (Large Generator Interconnection) [33] and FERC Order 2006 (Small Generator Interconnection) [34]. The power limits are based upon the aggregate power output of a generating installation.
Table 20.5
US FERC distributed generation connection requirements
| Engineering recommendations | Generator rated power Prated | Installation requirements |
| FERC order 2003a | Prated > 20 MW | Full engineering evaluation of impact on system |
| FERC order 2006 | 2 MW < Prated < 20 MW | Full engineering evaluation of impact on system |
| Prated < 2 MW | Fast track restricted evaluation of impact on system | |
| Prated ≤ 10 kW | Application form-based permission for installation of certified inverter-based generators |
The cost of connection is dependent on several factors, including proximity to existing network connection point, level of reinforcement required to manage additional power flow, planning issues (e.g., overhead versus underground cables), capacity, and voltage level of connection in addition to any ancillary infrastructure requirements (protection, reactive power compensation, etc.).
20.9. Smart grid solutions
As described in Section 20.5, distributed RES can bring some environmental and commercial benefits as well as challenges to existing distribution networks. The impact of distributed RES is determined by the following factors, which need to be considered before connecting a system to the grid.
1. Type and capacity of the generator and the voltage level for connection to the grid (PCC).
2. Availability of power generated.
3. Variation of power generated with time and predictability of variation.
4. Reliability of the plant.
5. Technology and regulations for connection.
After consideration of these factors, if it is established that the distributed RES will negatively impact the grid, two options are possible to mitigate the impact. One is to restructure and reinforce the network but this could be prohibitively expensive. The other “smart” option is to employ new technologies such as active voltage control, demand-side management, and energy storage [35]. These are key technologies for the new concept of electricity networks of the future “smart grids” [35–37]. They will play an important role in controlling the power balance in future power networks and maximizing the energy output from distributed renewable energy generation. The smart grid requires advanced communication protocols in order to implement dynamic control and automation to allow the power network to operate closer to its capacity while maintaining system security and integrity.
As explained in Section 20.6.2, current OLTCs usually use local voltage measurements in order to bring the voltage at the transformer location towards a specified value and therefore provide an acceptable voltage profile across the feeder length. Although line compensation (measurement of current through the transformer) may help in adjusting voltage control based on load current variations, this “passive” control becomes less effective in the presence of distributed RES and may even malfunction if reverse power flow occurs since line compensation is designed to measure current in one direction only.
Figure 20.19 shows an example of how the voltage rise caused by the connection of distributed generation (described in Section 20.6.2, Figure 20.15) can be mitigated by using active voltage control. This is achieved if the transformer tap changer control is based on voltage signals other than the local substation (e.g., from the remote end of a feeder). Dynamic rating may also be implemented to avoid thermal overloading of transformers and feeders.
Figure 20.19 Active voltage control.
In active network control, the voltage at a number of locations is monitored (through remote monitoring units and smart meters), and the OLTC is adjusted to maintain all of the measured locations within the desired limits. This allows the vulnerable points in the network, such as the far end of a feeder, to be monitored and kept within statutory limits. Number of remote monitoring units can be minimized by using state estimation [38].
Problems
1. Comment on the potential benefits of using distributed renewable energy generation.
2. Comment on the potential technical problems of an existing distribution power network from large-scale connection of distributed renewable energy generation. Suggest and briefly explain possible solutions using new technologies that will help to mitigate these problems.
3. According to existing legislations, how should the protection of a DG respond to a short circuit fault on the main network?
4. In grid-connected PV systems, self-commutated voltage-source inverters can be either voltage or current controlled. Briefly describe the difference between the two in terms of operating principles and main features.
5. Briefly describe, with the aid of suitable diagrams, three arrangements for connecting WECSs to the electricity grid, giving the main components and features of each arrangement.
6. Define the concept of the smart grid and show how it can help in the integration of RESs into the grid.
7. A three-phase power electronic load rated at 150 kW at 0.8 p–f lagging is fed from a 400 V supply having a fault in-feed of 2 MVA. Assuming that the source impedance is purely inductive and ignoring the effects of other loads connected to the network, calculate the percentage voltage dip at the load supply point when the load is switched direct on line.
8. A 10 km feeder is used to supply a load connected at the far end of the feeder and has a maximum capacity of 4 MVA at 0.8 p–f lagging. The feeder has a series resistance r = 0.25 Ω/km per phase and an inductive reactance x = 0.1 Ω/km per phase. The sending-end voltage is maintained at 11.8 kV.
a. Estimate the magnitude of the load voltage at maximum loading conditions.
b. A 2.0 MW wind turbine employing an induction generator is connected at the far end of the feeder. Assuming minimum loading conditions of 1 MVA at 0.9 p–f lagging, estimate the load voltage when the generator is operating at 80% capacity and 0.9 p–f leading.
c. Comment on the results obtained and briefly explain how variations in wind speed can result in fluctuations in the system voltage. Suggest two possible methods to mitigate the effects of these variations.
9. A developer wishes to connect a 2 MW distributed RES to an existing distribution network. In order to do this, an underground cable is to be used to connect the generator to the nearest substation. A simplified equivalent circuit of the proposed system is shown in Figure 20.20. The existing network has a substation (PCC) with a breaker feeding other consumers, as shown in the figure. The prospective fault rating at the substation is 10 MVA at zero p–f lagging and the fault clearing capacity of the circuit breaker is 10.5 MVA. The impedance of the proposed DG (ZG) is j0.04 pu and the underground cable (ZL) is j0.06 pu.
The developer needs to verify that the circuit breaker will be capable of handling the prospective fault current after DG connection. Given a per unit base of 100 kVA;
a. calculate the per unit equivalent source impedance, ZN.
b. Calculate the fault power at the substation after generator connection.
c. Comment on the implications of the results obtained in (b) for the distribution network operator and the developer. What steps might the developer take to satisfy the network operator requirement with regard to the fault level?
d. If the impedance of the generator could be changed, what would be the minimum value required so that the fault power at the PCC does not exceed the circuit breaker capacity?
e. Apart from fault current, list four other issues that need to be considered before connecting additional embedded generation to an existing network.
10. For the distribution system shown in Figure 20.21, the feeder is 10 km long and has a series resistance r = 0.25 Ω/km per phase and an inductive reactance x = 0.1 Ω/km per phase. The load has after-diversity minimum and maximum demand of 0.5 and 4.0 MW, respectively both at 0.9 p–f lagging. The feeder voltage is controlled by the OLTC at the 66/11 kV transformer. Assume that the grid voltage is fixed at 33 kV and that the transformer has taps of ±12.5% in steps of 2.5% (Figure 20.21).
Usually, the target voltage (control signal) for the OLTC is the voltage at busbar S and assuming that the voltage at busbar R is to be maintained within the statutory limits of 11 kV ± 5% for both minimum and maximum loading conditions. Using a base power of 10 MVA and taking the voltage at busbar R as a reference;
a. determine busbar S voltage for both minimum and maximum loading conditions.
b. Determine, with justification, the initial tap position of the OLTC (i.e., the voltage that the OLTC should maintain at busbar S) in order to keep the line voltages at busbar R within the statutory limit.
c. A 2.0 MW wind turbine employing an induction generator is connected at busbar R. Assuming minimum loading conditions and tap setting as calculated in (b) and the generator is operating at 80% capacity and 0.9 p–f leading, determine whether the OLTC is capable of maintaining the load voltage at busbar R within the statutory limits (±5%). Briefly explain why connection of distributed generation may result in a voltage rise beyond the statutory limit.
d. Active voltage control may be used to provide better control of voltage levels in a network with distributed generation. Describe how the concept of active voltage control may be implemented for the system in Figure 20.20, explaining how this can help the integration of RESs into the smart grid.
Figure 20.20 A simple distribution network with a DG.
Figure 20.21 An 11 kV distribution feeder.
..................Content has been hidden....................
You can't read the all page of ebook, please click here login for view all page.