N. K. Roy and H. R. Pota
Abstract Distributed generation (DG) is gaining popularity as it has a positive environmental impact and the capability to reduce high transmission costs and power losses. Although the integration of renewable energy-based DG will help reduce greenhouse gas emissions, it will rely heavily on new ways of managing system complexity. As traditional distribution networks were not designed to accommodate power generation facilities, various technical issues arise in the integration of distributed energy resources (DERs) into grids. This chapter presents an analysis of the major obstacles to the integration of green energy into power distribution systems (PDSs). Static and dynamic analyses are carried out with solar photovoltaic (PV) generators connected to different test systems to gain a clear understanding of the effect of PVs in PDSs. The results are compared with the existing utility standards to determine the critical issues in the integration of PVs into PDSs. A novel H? based control methodology is proposed to ensure grid code-compatible performances of PV generators. During the controller design, special attention is given to the dynamics of the load compositions of distribution systems. It is found that the proposed controller enhances the voltage stability of distribution systems under varying operating conditions.
Keywords Composite loads Distributed generation D-STATCOM Voltage stability Photovoltaic generator Robust control and uncertainty
N. K. Roy (H) • H. R. Pota
School of Engineering and Information Technology, The University of New South Wales, PO Box 7916, Canberra, ACT 2610, Australia e-mail: n. roy@student. unsw. edu. au
H. R. Pota
e-mail: h. pota@adfa. edu. au
J. Hossain and A. Mahmud (eds.), Renewable Energy Integration, Green Energy and Technology, DOI: 10.1007/978-981-4585-27-9_10, © Springer Science+Business Media Singapore 2014
10.1 Introduction
Power supply reliability, efficiency and sustainability are the major concerns in the development of future energy systems. Many countries of the world have set their targets for power generation from green energy to reduce greenhouse gas emissions, as shown in Table 10.1 [1]. To reach the goal of green energy, distributed generation (DG) is a promising option and, therefore, is receiving a great deal of interest.
Photovoltaic (PV) and wind energy-based systems are the most important green energy resources. Research on the integration of solar PV in distribution systems is still in its infancy. At present, the main concerns about high penetration levels of solar PV generation in a distribution system are the effects of intermittency on the system protection when multiple sources are connected to a radial feeder or network [2, 3]. Close attention has been paid to modeling generators and their associated controls, and distribution system equipment. A model of a PV array is proposed in
[4] which uses theoretical and empirical equations, together with data provided by the manufacturer, solar radiation, cell temperature and other variables, to predict the current-voltage curve. To study interactions of PV generators within the power system, a model of PV generator developed based on experimental results [5] suggests that the maximum power point tracking (MPPT) part of the control system of a PV generator dominates the dynamic behavior of the system. A mathematical model suitable for stability analysis that includes the nonlinear behavior of grid – connected PV modules is presented in [6]. Simulations which include the entire power converter are performed in [6] to support the mathematical analysis and it is concluded that the system is more susceptible to instability under high loading levels, i. e., when operating close to its maximum power point.
Grid integration of solar PV systems is gaining more interest than traditional stand-alone systems because of the following benefits:
• under favorable conditions, a grid-connected PV system supplies the excess power, beyond the consumption required by the connected load, to the utility grid;
• it is comparatively easy to install as it does not require a battery system because the grid is used as a backup;
• no storage losses are incurred; and
• it has potential cost advantages.
It is expected that grid-connected PV systems in medium-voltage networks will be commercially accepted in the near future [7]. Therefore, it is necessary to accurately predict the dynamic performance of three-phase grid-connected PV systems under different operating conditions in order to make a sound decision on the ancillary services that need to be provided to utilize their maximum benefits without violating grid constraints. Thus, the spread and growth of solar and other distributed renewable energy has led to significant modeling and engineering analyses of distribution systems. Although DG has several potential benefits, the
|
Country |
Target (%) |
Year |
|
Australia |
20 |
2020 |
|
Austria |
34 |
2020 |
|
Belgium |
13 |
2020 |
|
China |
15 |
2020 |
|
Denmark |
30 |
2025 |
|
Finland |
38 |
2020 |
|
France |
23 |
2020 |
|
Germany |
18 |
2020 |
|
Netherlands |
14 |
2020 |
|
New Zealand |
90 |
2025 |
|
Spain |
20 |
2020 |
|
Sweden |
49 |
2020 |
|
UK |
15 |
2020 |
|
US |
25 |
2025 |
|
Table 10.1 Renewable energy targets in different countries |
connection of it in the existing distribution network will increase the fault level of the system. The impact of DG on the local voltage level can be significant during a contingency. Typical contingencies on a distribution network can occur in the form of single or multiple outages, such as unplanned losses of generators or distribution feeders. Several internal and external causes are responsible for equipment outages [8]. The internal causes arise from phenomena, such as insulation breakdown, over-temperature relay action or simply incorrect operation of relays. The external causes result from some environmental effects, such as lightning, high winds and icy conditions or non-weather related events, such as a vehicle or aircraft coming into contact with equipment or even human or animal direct contact. These contingencies can result in partial or full power outage in a distribution network unless an appropriate control action is taken.
A higher PV penetration level could possibly cause instability problems when a large percentage of the system load is supplied by PVs. Therefore, it is becoming more important to understand the behavior of a DG-integrated system under disturbances with practical distribution network loads since variations in loads physically close to generators are a large fraction of the generation. A practical system load is a combination of various types of loads and it is referred to as a composite load. The accurate modeling of loads is a difficult task due to several factors, such as [9]:
• large number of diverse load components;
• ownership and location of load devices in customer facilities not directly accessible to the electric utility;
• changing load composition with time of day and week, seasons, weather and through time;
• lack of precise information on the composition of the load; and
• uncertainties regarding the characteristics of many load components, particularly for large voltage or frequency variations.
|
Table 10.2 Interconnection system response to abnormal |
Voltage range (pu) |
Clearing time (s) |
|
V < 0.5 |
0.16 |
|
|
voltages |
||
|
0.5 < V < 0.88 |
2.00 |
|
|
1.1 < V < 1.2 |
1.00 |
|
|
V > 1.2 |
0.16 |
Load modeling is qualitatively different from generator modeling in many aspects. Generally, only the aggregate behavior of load is required for power system stability studies rather than a whole collection of individual component behaviors [10, 11]. Thus, one cannot escape the necessity to analyze the impact of bus load compositions in the distribution system behavior.
It is known that the majority (more than 60 %) of power system loads are induction motors, the impact of which must be taken into account during network analysis [12]. A higher proportion of induction motors in the composite loads could cause voltage stability problem which could disconnect DG units from the network as per the current utility practice [13, 14] which demands that a system voltage should recover to an acceptable level after a disturbance within the time indicated in Table 10.2. The unnecessary disconnection of generators reduces the expected benefits of DG and should be avoided because of the increasing importance of DG. In this context, it is particularly important to supply reactive power to the dynamic load to maintain system stability, thereby keeping DG units connected.
Although inverter-connected PV systems have their own reactive power capability, they are not allowed to operate in voltage control mode to avoid controller interactions [13, 14]. Moreover, in order to contribute more real power into a system, small-scale PV units are operated at unity power factor (pf) [15]. It is not advised to use PV inverters with a variable pf because, at high penetration levels, this may increase the number of balanced conditions of load demands and generations and, subsequently, increase the probability of islanding which is a safety hazard [7]. If DG units are not allowed to regulate voltage, additional sources of reactive power need to be installed at critical locations to supply the reactive power of a system.
It is well-known that a static synchronous compensator (STATCOM) has excellent performance in terms of its response speed and capabilities to reduce system power loss and harmonics, improve voltage level and stability, and decrease occupation area [16, 17]. The internal controls of a Distribution STATCOM (D-STATCOM) play a very important role in maintaining the system voltage. The use of suitable control methods in a D-STATCOM may offer a better performance along with making possible tracking of the desirable references more efficiently. Conventional controllers for D-STATCOMs are mainly PI controllers [18, 19], the tuning of which is a complex task for a nonlinear system with switching devices. To avoid the limitations of PI controllers, a linear quadratic regulator (LQR) method is used in [20] to design a STATCOM controller which has a superior performance. Compared to the LQR method [20], which uses states as feedback, linear quadratic Gaussian (LQG) controller is more realistic as it can be designed using only measurable outputs and state variables estimated from them [21]. However, linear controllers designed based on the given operating point are not suitable in the event of large variations in the system model. To improve the performance of an LQG controller, model mismatches or uncertainties can be bounded by an Иж norm which provides robust closed-loop stability as well as optimal performance [22]. As distribution networks have different types of loads, a controller designed without regard to a tight bound on variations in load compositions will not lead to a satisfactory performance. However, this important issue has not been considered in the existing literature. The superior dynamic performance of the D-STATCOM compared to many other compensating devices encourages further refining its control scheme to achieve robust performance without enhancing its reactive power capacity.
This chapter makes three contributions: (a) an investigation of the static voltage stability of distribution networks through Q-V analysis for different contingencies;
(b) an examination of the impact of different load compositions on the dynamic behavior of distribution networks; and (c) a novel D-STATCOM controller design which is robust to variations in load compositions in a practical distribution system.
The organization of this chapter is as follows. Section 10.2 presents the static voltage stability analysis of the system. The mathematical model of the system for dynamic simulation is described in Sect. 10.3. Section 10.4 demonstrates the impact of various load compositions on the distribution system. The controller design is given in Sect. 10.5 and the performance of the designed controller is evaluated for different operating conditions in Sect. 10.6. Concluding remarks are given in Sect. 10.7.
