IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 26, NO. 2, APRIL

Transcription

1 IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 26, NO. 2, APRIL Practical Power Quality Charts for Motor Starting Assessment Xiaoyu Wang, Member, IEEE, Jing Yong, Member, IEEE, Wilsun Xu, Fellow, IEEE, and Walmir Freitas, Member, IEEE Abstract The impact of motor starting on power quality can be assessed using detailed computer simulation studies. However, not every motor installation case needs such an extensive assessment. Utility planners are interested in quick evaluation of the potential impact of a motor installation proposal. Based on the findings, they can then determine if detailed case studies and what types of case studies are necessary. This paper presents three charts for motor starting planning according to three power quality concerns. These concerns are the amount of voltage drop caused by motor starting, the compliance to the ITIC curve, and the compliance to the IEC flicker meter limits. These charts can help utility planers to conduct quick and first-cut assessment of a motor starting situation. They also reveal the key factors affecting the motor starting related power quality concerns. The principles behind these charts are explained. Examples are given to show how to use them for quick assessment of motor starting impact. Index Terms Flicker, motor starting, power quality (PQ), voltage sag. I. INTRODUCTION P OWER-QUALITY issues caused by induction motor starting have been recognized and investigated for a long time. It is known that direct motor starting typically produces voltage sags with a duration longer than 30 cycles, especially when a large size motor with high inertia load is connected to a weak power system [1] [6]. Such long duration sags can lead to a wide range of sensitive equipment to drop out [7]. In addition, voltage flicker may arise because of frequent or sporadic motor starting. Whenever an induction motor is connected or started, system planning engineers are interested to know if the motor starting would result in the unacceptable voltage sags and flicker according to the restriction of the established power acceptability curves or standards, such as the Computer & Business Equipment Manufacturers Association (CBEMA) Manuscript received October 14, 2009; revised July 12, 2010; accepted November 21, Date of current version March 25, This work was supported by Alberta power industry. Paper no. TPWRD X. Wang is with the Department of Electrical Engineering, State Key Lab of Power Systems, Tsinghua University, Beijing , China ( xiaoyuw@tsinghua.edu.cn). J. Yong is with the State Key Laboratory of Power Transmission Equipment and System Security and New Technology, Chongqing University, Chongqing , China ( yongjingcq@yahoo.com.cn). W. Xu is with the Department of Electrical and Computer Engineering, University of Alberta, Edmonton, AB T6G 2V4, Canada ( wxu@ualberta. ca). W. Freitas is with the Department of Electrical Energy Systems, University of Campinas, Campinas , Brazil ( walmir@ieee.org). Color versions of one or more of the figures in this paper are available online at Digital Object Identifier /TPWRD curve [8], the Information Technology Industry Council (ITIC) curve [9], and the IEC/IEEE flicker standards [10]. The voltage sag and flicker effects of motor starting can be assessed by detailed system simulations [11], [12]. However, this process is complicated and time-consuming in practice. In many cases, utilities and motor users are more interested in a simpler and faster method for a first-cut understanding of the potential impact of a motor starting situation [13], [14]. With limited information available, the method shall reveal whether an induction motor application may cause any power quality problem. If the results indicate violations of voltage sag or flicker limits, a detailed investigation can then be undertaken. A simple method utilizing the constant energy criterion to estimate the voltage sag impact of induction motor starting was described in [15]. Although the method simplifies the motor starting analysis, it is not intuitive enough for quick motor starting assessment. This is partially due to the fact that two key parameters, motor size and system short-circuit level, are not explicitly used. In addition, the applied specific energy curve [16] has relative large error compared with the ITIC curve, which was recently designated to replace the CBEMA curve. In North America, the historic flicker curves [17], [18] have been gradually replaced by the IEC flicker assessment method [19]. The IEC method requires the implementation of a sophisticated flicker meter scheme which is designed mainly for measurement purposes. Little work has been done on how to apply the IEC flicker meter to the planning process. In this paper, a systematic and practical method to evaluate the effect of voltage sag and flicker generated by induction motor starting is presented. Analytical approaches are used to derive charts that can provide a quick and first-cut understanding of the potential impact of a motor starting situation. The paper is organized as follows. Three motor starting guideline charts are proposed in Section II. The use of the charts is also explained in this section. Sections III V present the development process of the guideline charts for three types of power quality concerns: voltage drop, voltage sag and voltage flicker, respectively. Sensitivity studies results are also presented in these sections. Since the proposed methods are based on simplifications, electromagnetic transient (EMT) simulations are used in Section VI to validate the proposed charts. The conclusions are summarized in Section VII. II. PRACTICAL CHARTS FOR MOTOR STARTING ASSESSMENT The direct motor starting applies the network voltage to the motor terminal directly. Fig. 1 shows the schematic diagram of a real-life direct motor starting case. This case involves a 25 kv /$ IEEE

2 800 IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 26, NO. 2, APRIL 2011 Fig. 1. Diagram of direct motor starting in power distribution systems. Fig. 2. Voltage drop chart for motor starting assessment. Fig. 3. Voltage sag chart for motor starting assessment. distribution feeder typically seen in rural North America. The parameters of this system are listed in Appendix A. A 1.1 MW (1500 hp) induction motor is planned to be connected to the feeder. The voltage quality at the point of common coupling (PCC) is of concern during the starting process of the motor in Fig. 1. Three guideline charts proposed in this paper can be used to quickly assess the motor starting from the perspective of voltage drop, voltage sag and flicker limitations. The voltage drop chart is designed to address the concern on the amount of voltage drop caused by motor starting. The voltage sag chart considers both the voltage drop amount and the sag duration according to the ITIC curve. The flicker chart is developed based on the IEC flicker limit. A. Use of Voltage Drop Chart According to utility practices [20], a motor starting procedure that causes a voltage drop of 5% or more at the PCC during the starting process is not acceptable. The voltage drop chart is developed for utility planners to quickly determine if a motor installation scenario will result in unacceptable voltage drop (i.e., greater than 5%). The proposed chart is a set of curves whose x-axis is the motor capacity and y-axis is the system short-circuit level at the PCC. Fig. 2 displays the simplest and a conservatively constructed chart, which has only one curve. This curve indicates a boundary above which the impact of direct motor starting can be considered as insignificant in terms of voltage drop amount, while below which direct motor starting could cause a voltage drop exceeding 5%. Detailed motor starting analysis is therefore recommended for the latter case. As an example, we consider the case shown in Fig. 1 where the system short-circuit level is 28 MVA at the PCC and a 1.1 MW motor is to be installed at the PCC through a step down transformer. A point (1.1 MW, 28 MVA) can be drawn in the chart, which is shown as a red circle in Fig. 2. This point resides in the region with power quality (PQ) concern. Thus it can be concluded that the 1.1 MW motor is likely to cause excessive voltage drop at the PCC when the motor is directly started. Consequently, detailed motor starting analysis is recommended to further investigate the voltage drop problem. B. Use of Voltage Sag Chart Considering the voltage drop amount alone is not sufficient to understand the impact of motor starting, since the severity of a voltage sag event is a function of both the voltage drop amount and its duration. If a motor has a large inertia constant, the motor starting process can be long enough to cause a voltage sag problem. The voltage sag chart that can take into account both factors is then proposed. The basic idea of the chart is that a motor starting event should not cause violation of the ITIC curve. Fig. 3 shows the voltage sag chart developed for direct motor starting. In this chart, the x-axis is the ratio of the motor size to the short-circuit level at the PCC where the voltage sag is assessed for motor starting. The y-axis is the motor inertia constant. Like the curve in the voltage drop chart, the curve in the voltage sag chart also indicates a boundary below which the impact of direct motor starting can be considered as insignificant in terms of its voltage sag impact, while above which direct motor starting could cause a problem. Detailed motor starting analysis is therefore recommended for the latter case. Some examples are given as follows to illustrate the use of the voltage sag chart. 1) Case 1: Motor Starting Will Not Cause Power Quality Concern: In the first example, assume that the ratio of the motor

3 WANG et al.: PRACTICAL PQ CHARTS FOR MOTOR STARTING ASSESSMENT 801 size to the system short-circuit level is at the PCC. For the illustrated system shown in Fig. 1 the motor will be 0.9 MW (1250 hp) as the system short-circuit level is 28 MVA at the PCC. The motor has an inertia constant of 1 s. Then a point C (0.032, 1) can be found in the region without PQ concern in the chart shown in Fig. 3. As a result, this point (point C) will not cause a voltage sag problem at the PCC when the ITIC curve is applied for direct motor starting. 2) Case 2: Motor Starting Will Cause Power Quality Concern: In Fig. 3, point D is related to the point with and. This point is above the boundary curve and a detailed analysis needs to be conducted for this point. 3) Case 3: Terminal Points of the Boundary Curve: The points at the terminals of the boundary curve (point A and point B) represent two extreme conditions. For point A, the value of is If is less than 0.022, the ITIC curve will not be violated no matter what the inertia constant is. On the other hand, if is greater than 0.052, which is related to point B, the ITIC curve will always be violated no matter how small the inertia constant is. Note that motor starting usually lasts more than 0.5 second for medium and large motors and for small motors usually one does not need to do a detailed analysis. The classification of a motor as small, medium and large can be done based on [21]. Accordingly, an induction motor with 2, 4 or 6 poles is classified as small, medium or large when the nominal power is smaller than 0.75 kw (1 hp); higher than 0.75 kw and smaller than 370 kw (500 hp); or larger than 370 kw (500 hp), respectively. C. Use of Voltage Flicker Chart Another motor starting concern is the voltage flicker. The flicker level caused by motor starting is determined not only by the voltage drop amount but also by the frequency of occurrence of starting. Another complication is that the human element is involved when determining the permissible amount of flicker voltage. In recent years, the IEC flicker meter has gained wide acceptance, which should be applied to motor starting assessment [10]. In this paper, an approximate formula that connects the IEC flicker meter results (i.e., the short-term flicker severity index and the long-term flicker severity index levels) with the frequency of motor starting and the corresponding voltage drop amount at the PCC is used to derive a motor starting flicker chart. The flicker chart consists of two curves: the curve and the curve. The chart shows, for a given voltage drop amount, what is the highest frequency of repetitive motor starting that will lead to the violation of the IEC flicker meter limits. Fig. 4(a) shows the flicker curve for motor starting. In this figure, the x-axis is the number of motor starting per 10 minutes and the y-axis is the voltage drop amount at the PCC in per unit. The time interval 10 minutes is exploited when expressing the motor starting frequency. This is because the value of is calculated every 10 minutes by the IEC flicker meter. From Fig. 4(a) one can see that when the repetitive motor starting event is located above the boundary flicker curve, a flicker problem occurs for this motor starting. The limit of for the boundary flicker curve is recommended by Fig. 4. Voltage flicker chart for motor starting assessment. (a) P (b) P = 0:7 curve. = 0:9 curve. the IEC standard [23], where is required to be less than 0.9 for utility medium voltage (MV) system planning purposes. Fig. 4(b) shows the (MV planning level [23]) curve for the motor starting cases within the starting frequency less than once per 10 minutes and more than once per two hours. The time interval of 2 hours is defined by IEC for calculations. Similar to the curve, the boundary flicker curve also indicates whether a detailed flicker analysis should be performed or not. Examples of how to use the flicker chart are given as follows. 1) Case 1: Example For the Curve: It is known that the motor connected to the PCC in Fig. 1 is started 3 times per 10 minutes and that the voltage drop amount at the PCC for each start is 4.8%. The question is if the motor will cause the flicker problem according to the IEC flicker standard. Fig. 4(a) displays that the above motor starting event is related to point A which is located inside the region with flicker concern. Thus, the flicker problem could be incurred, and a detailed flicker analysis is needed for this case. However, if the voltage drop amount is decreased to 3.5% for each time (point A in Fig. 4(a)), the flicker problem caused by motor starting will not occur. 2) Case 2: Example For the Curve: The objective is to assess the motor starting flicker at the PCC with the following conditions: and motor starting frequency is

4 802 IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 26, NO. 2, APRIL 2011 is (i.e., cosine of the angle of the motor equivalent impedance for slip equal to 1), then in the motor base, we have (4) Fig. 5. Equivalent circuit for voltage drop amount assessment. 2 times per hour. In this case, the motor starting frequency is less than once per 10 minutes so that the curve cannot be used. The curve will be applied for this case. Point B in Fig. 4(b) can be found to represent the motor starting case. Fig. 4(b) reveals that point B is above the curve and within the region with flicker concern. As a result, the flicker problem could happen. The boundary curves of the proposed guideline charts are established using different power quality limits at the PCC. All the charts are developed by considering the worst case voltage violation conditions and the motor starting characteristics. Details on the derivation of the curves are described in the following sections. III. DEVELOPMENT OF THE VOLTAGE DROP CHART A. Chart Derivation In order to construct the voltage drop chart, one has to execute two steps: (1) determine the system parameters based only on the motor size and on the system short-circuit level at the PCC and (2) create the chart based on a simple formula with the relationship between the system short-circuit level and the motor size. The system presented in Fig. 1 will be used as an example. Fig. 5 shows the equivalent circuit of the example system. Initially, the following parameters are usually known for the motor starting analysis: Motor size (MVA). System short-circuit level at the PCC (MVA). In order to evaluate the motor starting impact the impedance parameters shown in Fig. 5 must be determined based only on the available information. First, we represent the system equivalent impedance based on the ratio and the system ratio, which are known. Thus, and referred to the motor size can be calculated as follows: where is the system ratio. If we assume that the rated capacity of the transformer is times the motor size, and that the transformer percent impedance referred to the transformer base is, then the transformer impedance referred to the motor size is Assuming that the inrush current of the motor is times the motor rated current and that the initial motor power factor (1) (2) (3) Also we can assume and (where and are the stator resistance and reactance, and are the rotor resistance and reactance, respectively, and the shunt magnetization impedance is neglected). Based on the above circuit, the voltage level at the PCC,, can be estimated, which can be symbolically represented as Since we are interested in determining the system size for a given or threshold, the above equation can be rearranged to show the relationship of as a function of and and other parameters. Based on the procedure shown in Appendix B, such a relationship can be established as: where In these expressions, represents the open circuit voltage at the PCC and represents the PCC voltage level at the instant of motor energization. The short-circuit level for different voltage drop limits imposed by the distribution utility can be obtained using these expressions. For the chart developed in this paper and (i.e., ) are used. Other parameters used for generating the basic chart shown in Fig. 2 are:,,, 6,. B. Sensitivity Analysis The development process of the chart shows that the chart is sensitive to the system ratio, transformer reactance (determined by the parameters of and ), motor starting power factor, motor starting inrush current coefficient. Sensitivity studies on these factors were conducted, and the results are reported in the following figures. Fig. 6 presents the chart for different ratios of the short-circuit level seen from the PCC. This figure shows that the lower the ratio is, the higher the maximum allowable motor size is for the same. This phenomenon is expected as the lower the ratio, which is equivalent to a smaller reactance and a higher resistance, the lower the reactive power losses in the system impedance. Consequently, the reactive power delivered to the motor can be reduced, allowing the increase of the maximum motor capacity. Fig. 7 presents the chart considering different sizes of the step-down transformer (5) (6) (7)

5 WANG et al.: PRACTICAL PQ CHARTS FOR MOTOR STARTING ASSESSMENT 803 Fig. 6. X=R ratio sensitivity (k = 6; cos ' = 0:2; = 1:2; = 5%). Fig. 8. Inrush current factor sensitivity ( =10; cos ' =0:2; =1:2; = 5%). Fig. 9. Motor power factor sensitivity ( =10;k =6; =1:2; = 5%). Fig. 7. Transformer size sensitivity (k = 6; cos ' = 0:2; = 10; = 5%). connecting the motor, which is represented as a percentage of the motor size. A transformer reactance of 5% in the transformer base was used and the transformer size is times the motor size. Fig. 7 reveals that larger transformers will limit the connecting motor capacity for the same. This result is expected since the higher the factor is, corresponding to a lower value of transformer impedance (referred to the motor base), the higher the inrush current drawn by the motor is. Thus, the voltage drop through the system impedance is higher. Consequently, the maximum motor size is reduced. The inrush current factor was analyzed in Fig. 8, which indicates that the lower the inrush current factor is, the higher the maximum motor size is. This result is expected since the lower the inrush factor, the higher the motor impedance, in accordance to (4) and (5). As a result, the inrush current drawn by the motor is smaller. Therefore, the maximum capacity of the motor can be further increased. In Fig. 9 the initial motor power factor is analyzed. This factor reflects the relationship between and of the motor equivalent circuit in Fig. 5 for a slip equal to 1. The result displayed in Fig. 9 reveals that the curve will not be considerably changed when the initial motor power factor is small. Fig. 10. Motor starting voltage sag mapped to the ITIC curve. IV. DEVELOPMENT OF THE VOLTAGE SAG CHART Fig. 10 shows the voltage sag boundary of the bottom ITIC curve. On this curve, voltage levels above 80% of the nominal are acceptable if their duration is less than 10 seconds, and voltage levels between 70% and 80% of the nominal are acceptable if the duration is less than 0.5 second. The curve further reveals that voltage levels above 90% of the nominal

6 804 IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 26, NO. 2, APRIL 2011 are considered as always acceptable. Fig. 10 also plots several motor starting events marked as. As the motor size increases, the voltage drop amount increases and the trace will cross the 80% threshold. The results also reveal that the motor starting cases often involve sag durations between 0.5 to 10 seconds or higher. So the region of concern is the 80% threshold line and the boundary between the 80% and the 90% threshold lines. Now let s consider the barely acceptable motor starting cases that result in 20% voltage drop. If the motor inertia constant increases without causing voltage drop change, the sag event points will move rightward as it takes more time to start the motor. Point corresponds to the largest inertia constant that can be accepted without violating the ITIC curve. The voltage sag impact chart shows this largest acceptable inertia constant as a function of the motor and system sizes while the 20% voltage drop amount is not violated. With the above background introduction, the development of the voltage sag chart can be described as follows. First, we know that for a given system short-circuit level at the PCC, we can use (8) to calculate the voltage drop amount of the PCC caused by motor starting Fig. 11. Motor starting H 0 S curve for S =10MVA. where is the function defined in (7). Equations (8) shows that each motor size is related to a value of if is known. Since, from the previous explanation, we have (8) (9) thus, we can get a motor size range from (8). is the motor size causing and is the motor size causing. Then, for each motor size within the range,a boundary motor inertia constant can be calculated from the following equations by setting as 10 seconds: (10) where is the motor starting time or the voltage sag duration, is the electromagnetic torque, is the mechanical torque of the load, is the electric rotor speed, and is its steady-state value. The torques are expressed in the per unit system with the motor size as the base system capacity. The mechanical torque usually has the following form: (11) where is the coefficient for the load torque which changes with the rotor speed during the motor starting process and reaches the rated motor torque at steady state [22]. is determined by. From (10), we know that the motor starting time is proportional to once the motor size is decided. Based on (11), a motor inertia constant range can be obtained for the motor size range. is the Fig. 12. Impact of inertia constant on motor starting time. boundary value for, and is the boundary value for. A boundary curve shown in Fig. 11 can then be plotted for with a given value. When the system short-circuit level is changed, different curves can be plotted. However, further study shows that these curves can be unified if the x-axis in Fig. 11 is expressed as the ratio of. Then, a generalized curve like the one shown in Fig. 3 can be drawn for motor starting voltage sag assessment. The largest motor starting time of 10 seconds is used to get the boundary value. The meaning of this boundary value is that if a motor has inertia constant larger than the boundary value when and are known, the motor starting time will be over 10 seconds. This duration will result in the motor starting event being located in the No Damage Region of the ITIC curve. The voltage sag problem will then occur. In Fig. 12, points C, D, E cause the same voltage sag magnitude 0.85 pu because they have the same values of and. However, they have different values. Point D is related to the boundary value. The inertia constant value at point C is smaller than the boundary value, thus, it is in the safe region. Point E has an inertia constant value larger than the boundary value, consequently, point E is not allowed for motor starting.

7 WANG et al.: PRACTICAL PQ CHARTS FOR MOTOR STARTING ASSESSMENT 805 V. DEVELOPMENT OF THE VOLTAGE FLICKER CHART Motor starting can occur once per hour as well as several times per week. is calculated every 10 minutes. As a result, IEC has suggested an analytical method in which a shape factor is defined for motor starting voltage characteristics [23]. With the shape factor definition, the motor starting flicker can be calculated by using the following procedure. The analytical calculation equation recommended by IEC is shown in (12) (12) where is the sum of the flicker impression times. If we assume that the motor starting voltage drop amount is the same for all the motor starts, the total flicker impression time can be expressed as follows: (13) where is the time of motor starting within the assessment time period. is calculated as follows [23]: (14) where is the maximum relative voltage change (the voltage drop amount) which is expressed as a percentage of the nominal voltage. Equation (14) is defined by IEC. For motor starting without special mitigation methods for inrush current reduction, the shape factor is normally about 1, as recommended in [23]. From (12) (14) the relationship between the motor starting times and the voltage drop amount can be obtained by Fig. 13. Simulation verification of the voltage drop chart. (a) Voltage drop chart ( = 3; = 6:8; = 6%; k = 5:5; cos ' = 0:3). (b) Motor starting simulation waveforms. (15) In the IEC standards, should be smaller than the planning level 0.9 in MV system and the evaluation time for is 10 min. With these two conditions, the boundary flicker curve with, shown in Fig. 4(a), can be plotted based on (15). Similarly, the value of can also be calculated by using (12) by replacing the evaluation time segment with 2 hours Then, (15) will be changed as follows for the (16) flicker curve: (17) The planning level limit for is suggested as 0.7 by IEC [23]. The flicker curve developed from (17) is shown in Fig. 4(b) where. When the maximum voltage drop amount is determined by the flicker curves, the maximum motor size related to this voltage drop amount can be checked from the flicker chart. VI. VALIDATION RESULTS The section presents the EMT simulation results executed in PSCAD/EMTDC to verify the accuracy and limitations of the proposed power quality charts for motor starting. The real case shown in Fig. 1 is employed to run the simulation. The field data of the distribution system are listed in Table I of Appendix A and the selected motor data [24] are shown in Table II of Appendix C. The comparisons are summarized as follows. A. Simulation Results for the Voltage Drop Chart In Fig. 13(a), the following four cases are analyzed with the voltage drop chart: Case 1 (1500 hp); ; Case 2 (500 hp); ; Case 3 (100 hp); ; Case 4 (500 hp);. The parameters used to plot the chart are 3, 6.8,, 5.5, and 0.3. From Fig. 13(a) one can see that Case 1 and Case 2 are in the region with PQ concern whereas Case 3 and Case 4 are in the region without PQ concern. The detailed motor starting situation for the four

8 806 IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 26, NO. 2, APRIL 2011 Fig. 14. Voltage sag chart determined by repetitive EMT simulations and by the proposed formulas ( = 10; cos ' = 0:2; = 1:2; = 5%). cases are simulated respectively and the simulation results are displayed in Fig. 13(b). In the simulations, the motor starting instant is 1 s and the 25 kv/4.16 kv transformer capacity is changed to maintain 6.8 for the different cases. The distribution system line impedances are also changed when is increased from 28 MVA to 56 MVA; however, is always kept as 3 in order to assure the same conditions in Fig. 13(a) and (b). Fig. 13(b) reveals that the PCC voltage for Case 1 and Case 2 is below 0.95 pu during the motor starting, whereas it is above 0.95 pu for Case 3 and Case 4. Thus, the voltage drop chart has successfully predicted the motor starting impact of the illustrated cases. In order to further validate the voltage drop chart, Fig. 14 shows the curve behavior for a variation of. In this figure, the curves determined by the proposed formulas are compared to the curves obtained by repetitive EMT simulations done with PSCAD/EMTDC. As one can see the results obtained with the proposed method are very accurate. Similar results are also obtained for the other sensitivity analysis presented in Part B of Section III. In the EMT simulations, the induction motor is represented by the traditional squirrel cage model, the transformer by the T-model and the equivalent system by a voltage source behind an impedance. The algorithm used to trace the curve by repetitive EMT simulations is as follows: Step 1) Set,, and ; Step 2) Set (step of motor capability) and (step of system short-circuit level); Step 3) Do short-circuit level under analysis; Step 4) If short-circuit level under analysis plot the points end; Step 5) Do motor capability under analysis; Step 6) Set: and using (1) and (2) (system parameters) using (3) (transformer parameter) and using (4) and (5), and (motor parameters); Fig. 15. Simulation verification of the voltage sag chart. (a) Voltage sag chart ( = 3; = 6:8; = 6%; k = 5:5; cos ' = 0:3). (b) Motor starting simulation waveforms. Step 7) Run a direct motor starting simulation by using PSCAD/EMTDC If Do Go to Step 6 If Store the points Do Go to Step 4. B. Simulation Results for the Voltage Sag Chart Fig. 15(a) shows the voltage sag chart based on the parameters set in Fig. 13(a). Point C and point D in Fig. 15(a) are all related to Case 1 in Part A of Section VI. The only difference between the two points is that for point C and for point D. From Fig. 15(a) one can see that point C is in the region without PQ concern and point D is in the region with PQ concern. This observation can be validated by the EMT simulation results shown in Fig. 15(b) where the motor starting time is 3.2 s for and is 14 s for. The voltage drop amount is the same (around 17%) for the two points. According to the ITIC curve displayed in Fig. 10, the case

9 WANG et al.: PRACTICAL PQ CHARTS FOR MOTOR STARTING ASSESSMENT 807 TABLE I DISTRIBUTION SYSTEM PARAMETERS Fig. 16. Simulation verification of the voltage flicker chart. (a) Motor starting waveform. (b) Comparison of P (T =100ms;T = 100 ms; F =0:8). is in the No Interruption in Function Region (without PQ concern) and the case is in the No Damage Region (with PQ concern). As a result, the conclusions from Fig. 15(a) are confirmed in Fig. 15(b). It is worth noting that the range and the boundary values (point A and point B) in Fig. 15(a) are different from those in Fig. 3 due to the change of the parameters,,,, and. C. Simulation Results for the Voltage Flicker Chart From Section V, we know that the proposed voltage flicker chart was developed based on (12) (17); consequently, the accuracy of (12) (17) determines the validation of the proposed voltage flicker chart. In this section, the analytical motor starting flicker assessment method is verified by using a digital IEC flicker meter set up in Matlab/Simulink. Fig. 16(a) shows the motor starting waveform adopted by the analytical assessment method. Three main parameters, which are, (front time), and (tail time), are used to determine the shape factor of the waveform [25]. Fig. 16(b) shows the comparison result of the values obtained from the analytical method and the simulation method when is varied. In the analytical method, is determined by and according to Fig. 7 of [25]. A digital IEC flicker meter [19] is designed in Matlab/Simulink to run measure the of the motor starting waveform shown in Fig. 16(a). The motor starting frequency is once per minute. From Fig. 16(b) one can see the analytical result and the simulation result match well. VII. CONCLUSIONS Induction motor starting draws high inrush current from the system, which may result in an unacceptable voltage drop at the PCC. The analysis of this problem usually is done by using electromagnetic transient simulations, which is a very time-consuming process and requires elaborated data. Thus, three power quality charts were proposed in this paper to provide practical guidelines for motor starting assessment. The developed charts can be used to fast determine if a motor is permitted to be connected to the PCC based on the voltage drop amount criterion defined by the utilities as well as the ITIC curve and the IEC flicker standard. The proposed three charts, i.e., the voltage drop, the voltage sag and the flicker charts, were applied to several real cases and no case related to power quality issues was classified as power quality free. Due to the lack of space, these cases were not included in the paper; instead, we have decided to include only simple examples in order to facilitate the understanding of the proposed method. The results of these several cases showed that the proposed charts can be applied with confidence to separate the cases where the power quality is not a concern, not demanding further analysis, from the cases where the power quality may be a concern and more detailed investigation must be conducted. Therefore, the usage of these charts can significantly reduce the time spent by engineers when analyzing the installation of a new induction motor. The examples illustrated that the application of the guideline charts is straightforward and based on them one can know how the results may be interpreted. The sensitivity studies presented in Section III-B also showed which are the main factors affecting the impact of the motor starting process on the power quality issues. The analyzed factors were: feeder ratio, transformer size, inrush current factor and the initial motor power factor. The accuracy of the proposed charts was also verified by motor starting simulations in PSCAD/EMTDC and the IEC flicker meter in Matlab/Simulink. APPENDIX A Table I lists the parameters of the distribution system shown in Fig. 1. APPENDIX B This section presents the development procedure of the voltage drop amount equation shown in Section III. From Fig. 5 the voltage at the PCC can be derived as (18)

10 808 IEEE TRANSACTIONS ON POWER DELIVERY, VOL. 26, NO. 2, APRIL 2011 TABLE II THREE-PHASE MOTOR PARAMETERS (4.16 kv) The equivalent system parameters can be calculated by (19) (20) Substituting and in (18) results in the following equation: (21) There are two solutions for the above equation. One is expressed by (7) and the other is (22). However, (22) will get a negative value because,, and. Thus, only (7) is considered in the analysis APPENDIX C Table II lists the motor parameters used in the simulations. (22) REFERENCES [1] M. H. J. Bollen, Understanding Power Quality Problems: Voltage Sags and Interruptions. New York: IEEE Press, [2] M. F. McGranaghan, D. R. Mueller, and M. J. Samotyj, Voltage sags in industrial systems, IEEE Trans. Ind. Appl., vol. 29, no. 2, pp , Mar./Apr [3] J. Lamoree, D. Mueller, P. Vinett, W. Jones, and M. Samotyj, Voltage sags analysis case studies, IEEE Trans. Ind. Appl., vol. 30, no. 4, pp , Jul [4] J. C. Gomez, C. Reineri, G. Campetelli, and M. M. Morcos, A study of voltage sags generated by induction motor starting, Elect. Power Compon. Syst., vol. 32, no. 6, pp , Jun [5] C. Hsu, H. Chuang, and C. Chen, Power quality assessment of large motor starting and loading for the integrated steel-making cogeneration facility, IEEE Trans. Ind. Appl., vol. 43, no. 2, pp , Mar [6] X. Liang and O. Ilochonwu, Induction motor starting in practical industrial applications, in Proc. IEEE Industrial and Commercial Power Systems Technical Conf., Tallahassee, FL, 2010, pp [7] R. F. McElveen and M. K. Tony, Starting high-inertia loads, IEEE Trans. Ind. Appl., vol. 37, no. 1, pp , Jan./Feb [8] G. Lee, M. M. Albu, and G. Heydt, A power quality index based on equipment sensitivity, cost, and network vulnerability, IEEE Trans. Power Del., vol. 19, no. 3, pp , Jul [9] ITI (CBEMA) Curve Application Notes. Washington, DC: Inf. Tech. Ind. Council, [10] S. M. Halpin, R. Bergeron, T. M. Blooming, R. F. Burch, L. E. Conrad, and T. S. Key, Voltage and lamp flicker issues: Should the IEEE adopt the IEC approach?, IEEE Trans. Power Del., vol. 18, no. 3, pp , Jul [11] S. M. Halpin and R. F. Burch, An improved simulation approach for the analysis of voltage flicker and the evaluation of mitigation strategies, IEEE Trans. Power Del., vol. 12, no. 3, pp , Jul Available: [12] X. Yang and M. Kratz, Power system flicker analysis by RMS voltage values and numeric flicker meter emulation, IEEE Trans. Power Del., vol. 24, no. 3, pp , Jul [13] X. Yang and J. Gauthier, How can flicker level be determined before a customer is connected to the electric grid, in Proc. IEEE Power Energy Soc. Gen. Meeting, Calgary, AB, Canada, 2009, pp [14] A. Polycarpou and H. Nouri, Validation of a proposed voltage sag prediction methodology for interconnected systems during motor starting, in Proc. 44th Int. Universities Power Engineering Conf., Glasgow, U.K., 2009, pp [15] J. C. Gomez and M. M. Morcos, Voltage sag effects on sensitive equipment due to starting cycles of induction motor, IEEE Power Eng. Rev., vol. 22, no. 8, pp , Aug [16] J. C. Gomez and M. M. Morcos, Voltage sag and recovery time in repetitive events, IEEE Trans. Power Del., vol. 17, no. 4, pp , Oct [17] IEEE Recommended Practice for Electric Power Distribution for Industrial Plants, IEEE Std (Revision of IEEE Std ), [18] 1C.5.1 Voltage Fluctuation and Light Flicker, PacifiCorp Engineering Handbook, [Online]. Available: File/File57576.pdf [19] Flickermeter Functional and Design Specifications, IEC Std , [20] Distribution Point-of-Delivery Interconnection Process Guideline- Standards of Service, Alberta Electric System Operator. [Online]. [21] Motors and Generators. VA: National Electrical Manufacturers Association MG1, [22] P. Kundur, Power System Stability and Control. New York: McGraw- Hill, [23] Assessment of Emission Limits for Fluctuating Loads in MV and HV Power Systems Basic EMC Publication, IEC Std , [24] G. S. Grewal, S. Pocsai, and M. M. Hakim, Transient motor reacceleration study in an integrated petrochemical facility, IEEE Trans. Ind. Appl., vol. 35, no. 4, pp , Jul./Aug [25] Limitation of Voltage Fluctuations and Flicker in Low-Voltage Supply Systems for Equipment With Rated Current 16 A Basic EMC Publication, IEC Std , Xiaoyu Wang (M 08) received the B.Sc. and M.Sc. degrees in electrical engineering from Tsinghua University, Beijing, China, in 2000 and 2003, respectively, and the Ph.D. degree in electrical and computer engineering from the University of Alberta, Edmonton, AB, Canada, in Currently, he is an Assistant Researcher at Tsinghua University. His research interests include distributed generation and power quality. Jing Yong (M 08) received the B.Sc., M.Sc., and Ph.D. degrees in electrical engineering from Chongqing University, Chongqing, China, in 1985, 1988, and 2007, respectively. She has been a Postdoctoral Fellow at the University of Alberta since Her current interest is power quality. Wilsun Xu (M 90 SM 95 F 05) received the Ph.D. degree from the University of British Columbia, Vancouver, BC, Canada, in He was an Engineer with BC Hydro from 1990 to Currently, he is a Professor and an NSERC/iCORE Industrial Research Chair at the University of Alberta, Edmonton, AB, Canada. His research interests are power quality and harmonics. Walmir Freitas (M 02) received the Ph.D. degree in electrical engineering from the University of Campinas, Campinas, Brazil, in From 2002 to 2003, he was a Postdoctoral Fellow with the University of Alberta, Edmonton, AB, Canada. Currently, he is an Associate Professor with the University of Campinas, Campinas, Brazil. His research interests are the analysis of distribution systems and distributed generation.