Fast and reliable index to protect the synchronous generators against loss of field incidence
2020; Institution of Engineering and Technology; Volume: 14; Issue: 24 Linguagem: Inglês
10.1049/iet-gtd.2020.0742
ISSN1751-8695
Autores Tópico(s)Computational Physics and Python Applications
ResumoIET Generation, Transmission & DistributionVolume 14, Issue 24 p. 6019-6026 Research ArticleFree Access Fast and reliable index to protect the synchronous generators against loss of field incidence Ali Rostami, Department of Electrical Engineering, University of Kurdistan, Sanandaj, IranSearch for more papers by this authorNavid Rezaei, Corresponding Author n.rezaei@uok.ac.ir orcid.org/0000-0002-8570-8301 Department of Electrical Engineering, University of Kurdistan, Sanandaj, IranSearch for more papers by this author Ali Rostami, Department of Electrical Engineering, University of Kurdistan, Sanandaj, IranSearch for more papers by this authorNavid Rezaei, Corresponding Author n.rezaei@uok.ac.ir orcid.org/0000-0002-8570-8301 Department of Electrical Engineering, University of Kurdistan, Sanandaj, IranSearch for more papers by this author First published: 02 November 2020 https://doi.org/10.1049/iet-gtd.2020.0742AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onEmailFacebookTwitterLinked InRedditWechat Abstract This study presents a new index for fast and reliable protection of the synchronous generator (SG) against loss of field (LOF) incidence, focusing on the operation time, reliability in clustering LOF, power system disturbance (PSD), and practical application. To this end, all allowable electrical parameters at SG output are acquired and thoroughly analysed. Analytical analysis of the parameters indicates that reactive power (QSG) and load angle (δSG) parameters have better conditions to improve the mentioned factors. Hence, based on the δSG increment and QSG decrement during LOF events, a new index based on the ratio of deviations of the δSG over QSG (ΔδSG/ΔQSG) is proposed as a suitable criterion owing to remarkable fast sensitivity to LOF and PSD events. In the proposed algorithm, analysis of the events by the proposed ΔδSG/ΔQSG index is dependent on ΔVSG < 0 condition. Extensive simulation studies demonstrated that the proposed LOFPS is a straightforward method to immediately and securely distinguish between LOF and PSD events, and it also has minimum operation time than other methods. Moreover, numerical data affirmed the greater changes of the ΔδSG/ΔQSG index during LOF events than those in PSD events, giving rise to the easy clustering of both LOF/PSD events. 1 Introduction The synchronous generator (SG) has a major role in the proper operation and stability of the power systems. It must be equipped with various protection relays to identify all faulty circumstances precisely and quickly without any mal-operations. Loss of field (LOF) protection relay is one of the key protections to protect the SG unit against LOF events and differentiates from the other power system disturbance (PSD) events. The LOF event has occurred under some conditions such as accidental tripping of the excitation breaker and open- or short-circuit of the excitation system [1, 2]. The LOF failure occurrence and extension of this situation can result in some serious negative effects on the appropriate operation of the both power system and SG unit [3, 4]. These negative effects for the SG units include over-loading of the field winding, over-heating of the stator end-winding, and torque oscillations, leading to some damages on this high-cost equipment, especially when it operates under high-loading conditions. The LOF incidence in the large-scale SGs can lead to voltage collapse at a wide area of the power system because it will absorb a large amount of reactive power from the remaining SGs in the power system [5]. Hence, due to the adverse effects of the LOF occurrence on both SG and power systems, it is crucial to develop a LOF protection scheme (LOFPS) that can quickly and reliably protect the SG against LOF incidence. The most commonly used method for LOF protection is an impedance-based relay. The initial version of the impedance-based relay suggested a single zone with a negative-offset mho characteristic [6]. To reduce the relay operation time under high-loading conditions, it is improved by using two negative offset mho characteristic zones [7]. However, these LOFPSs may not operate properly during partial LOF incidence. To overcome this problem, two negative offset mho elements along with a positive offset mho element are implemented in the second zone [8]. Other types of impedance-based LOFPSs, which are not being widely considered, are R–X with the directional element, admittance-based LOFPS including conductance and susceptance components [9, 10]. The offered adaptive mho-based LOFPS in [11] presents a new concept to improve the impedance-based relay using a capability curve limit. All these impedance-based LOFPSs may mal-operate under some PSD events. The proposed LOFPS in [12] utilised reactive power parameter for the detection of both LOF/PSD events. The main defect of this parameter is the wrong detection during the operation of the SG at the unit or near unity power factor. The proposed method in [13] utilises the resistance seen from the viewpoint of the SG unit and asserts that its changes will be negative following the LOF events. On the other hand, the proposed second derivation method of currents in [14] has proved that it remains positive after LOF incidence. However, the defined features of the proposed parameters in [13, 14] are not acceptable under all operating conditions of the SG unit. Although the proposed setting-free LOFPS in [15] does not require any threshold value to distinguish between LOF/PSD events, it may not operate properly under some PSD events and cannot properly detect all LOF events. The proposed internal-voltage-based LOFPS in [16] that can be implemented using the SG common measurable parameters is not able to properly cluster all LOF/PSD events. The proposed methods in [17, 18] utilise the flux linkage and rotor signal estimation, respectively. These LOFPSs need more sensors and additional devices and also the proper operation of the proposed parameters is extremely dependent on the accuracy of the measured parameters. The proposed method in [19] uses the voltage, active, and reactive powers to protect the SG against LOF events. This method needs more time to operate and identify the happened events properly. In [20], an Artificial Neural Network (ANN)-based LOFPS is suggested for proper identification of LOF/PSD events. This method is costly as it uses several parameters and needs more measuring devices. The outlined method in [21] exploits multiplying voltage, reactive power, and load angle parameters as an effective LOFPS. The main drawbacks of this method include: (i) multiplying three variables leads to more complexity and a little time delay for proper operation under various operating conditions of the power system, (ii) because the variations of the selected parameters are not in a similar direction (e.g. voltage decrement and load angle increment), multiplying these variables with different changing directions cannot properly address all events, leading to sending wrong tripping command. Meanwhile, setting the threshold value for this scheme is quite difficult. The recently published paper [22] presents a worthwhile current-based differential technique, and also in [23] the performance of some industrial LOFPSs using a realistic model in the Real Time Digital Simulator (RTDS) has been evaluated. This paper presents a new index for fast and secure protection of the SG against LOF incidence considering operation time, reliability in clustering LOF and PSD events, and practical application. A comprehensive analytical analysis of all allowable parameters shows that two parameters including QSG and δSG have better performance than other parameters and can be more effective in designing an exact LOFPS. Hence, a new index such as ΔδSG/ΔQSG is suggested as the best index owing to noticeable sensitivity to LOF and PSD events. Meanwhile, using the ΔδSG/ΔQSG index is designed to be dependent on ΔVSG < 0 condition. Simulation studies show the superiority of the proposed LOFPS over other methods. To indicate the comparison between various research works in the literature and the proposed index in this paper, the taxonomy of the literature is presented in Table 1. According to Table 1, the novel contribution of the proposed work with respect to previously published papers can be demonstrated. It should be noted that for each factor that has been met by the LOFPS, a double-star sign is inserted, and if it cannot completely be met by the LOFPS; a single-star sign is intercalated. Table 1. Taxonomy table of the LOF protection schemes References Operation time Reliable clustering Practical application this paper ** ** * [11] * * ** [13] ** * * [14] ** * * [15] * * ** [17] * ** * [19] * * ** [21] ** * * The main contributions of this paper are as follows: Proposing a novel LOFPS to identify both LOF and PSD events faster than other schemes. Providing a LOFPS to cluster the LOF and PSD events precisely without any mal-operations. Suggesting an economical and practical viable LOFPS without needing any additional devices. The remaining part of this paper consists of four sections. The proposed LOFPS is introduced in Section 2 and is thoroughly analysed in detail. Performed simulation studies are presented in Section 3 to illustrate the efficiency and applicability of the introduced LOFPS. Section 4 describes the comparison to other available LOFPS. Eventually, the conclusion is outlined in Section 5. 2 Proposed methodology To provide an exact analysis of the operation procedure of the proposed LOF protection index, the sample system is shown in Fig. 1. Fig. 1Open in figure viewerPowerPoint Single-line diagram of the sample power system and SG with the voltage control system 2.1 Operational principle of the proposed scheme There are various parameters in the SG output that can be utilised for identifying LOF and PSD events. The utilised parameters must reliably and immediately detect all LOF and PSD events under various conditions, though only some of them can be suitable for such circumstances. For example, the parameters, which are dependent on inertia of the SG, have deviated slowly and cannot protect the SG against various LOF/PSD events quickly. On the other hand, some of the electrical parameters that have similar changing patterns during both LOF/PSD events cannot properly differentiate between them. Hence, to select the most suitable parameters for LOF/PSD detection, a set of allowable parameters from SG is provided in Table 2. In the following, all parameters presented in Table 2 are thoroughly analysed considering the most important design LOFPS items including operation time, reliability in classifying both LOF/PSD events as well as practical applications. Operation time: To design a fast LOFPS, it is necessary to select the parameters that changes faster than other parameters. The electrical parameters which are independent of inertia of SG have quickly deviated from other inertia dependent parameters. Among the presented parameters in Table 2, PSG, fSG, ω, dω, θ, Id, Iq, and ISG are directly and indirectly affected by the inertia of the SG unit. For example, PSG, fSG, ω, and dω are directly imposed by inertia constant, while θ, Id, Iq, and ISG are indirectly affected by it. It should be noted that the mentioned currents are affected by both the voltage control system (inertia independent) and frequency control system (inertia dependent). Hence, these parameters can be removed from the candidate parameters to design a proper LOFPS with minimum operation Practical applications: Economical issues and simple execution are the other factors that should be considered to design a simple and cost-effective LOFPS. Hence, the measurement devices and other additional sensors for the selected parameters are the most important issues in this regard. For example, voltage and current transformers are commonly installed at the SG output and as a result, VSG, ISG, dq-components of these parameters as well as PSG and QSG parameters are easily allowable without the need for any additional devices. Nevertheless, some other parameters such as Idmd, Idmq, FFd, FFq, and Te parameters need additional devices and sensors for measuring and implementation. Meanwhile, using several parameters for LOF/PSD detection requires installing various measurement devices, making the LOFPS highly-costly. On the other hand, the accuracy of the measurement devices may be negatively affected in the presence of noise, which can lead to mal-operation. Hence, the simple and cost-effective LOFPS is the method that uses a few parameters, making it more desirable than complex ones. Therefore, these parameters are also removed from the candidate parameters to design an efficient LOFPS. Reliable detection of the LOF/PSD events: Using a single parameter for detecting the LOF/PSD events is not reliable and may mal-operate under some conditions. Hence, to cover all operating conditions of the SG and power systems and to prevent any wrong operations, a combination of some parameters is implemented. Although utilising several parameters for LOFPS leads to the reliable classification of both LOF/PSD events, it also leads to some complexity. Hence, a few most effective parameters should be adopted in combination with each other. Table 2. All allowable signals from the SG unit to design the LOFPS Signal group Signals symbols and definition output electrical parameters VSG voltage at SG output ISG current at SG output PSG active power at SG output QSG reactive power at SG output δSG load angle between ESG and VSG fSG frequency at SG output dq0 components Id d-component of stator current Iq q-component of stator current Idmd d-component of damper winding current Idmq q-component of damper winding current FFd d-component of field flux FFq q-component of field flux Vd d-component of stator voltage Vq q-component of stator voltage mechanical parameters ω rotor speed d ω rotor speed deviation θ rotor mechanical angle Te electromagnetic torque Based on the above-mentioned discussions, the remaining parameters for designing a fast and reliable LOFPS consist of VSG, QSG, and δSG. However, using each of these parameters alone or an inappropriate combination of them, even with a suitable threshold value is just useful for identifying some LOF/PSD events and may identify some other conditions wrongly. Hence, it is crucial to design a combination of these parameters in the best way to avoid any mal-operation during all LOF/PSD events. The sample dynamic variations of the selected parameters have been presented in Fig. 2, which are acquired from the SG in Fig. 4 during the LOF event. Fig. 2Open in figure viewerPowerPoint Typical dynamic deviations during the LOF event incidence Based on the typical simulation results and discussions in [14, 16, 21], the most important findings associated with the selected parameters are provided as follows: The changes of VSG and QSG after LOF occurrence are in a similar direction (i.e. decrement variation), while the variation of the δSG parameter is in the increment direction. The changes in the magnitude of QSG and δSG are remarkable than other parameters such as VSG and are not seriously affected by changing the loading levels of the SG unit. The changes in the magnitude of the parameters are affected by changing the loading level of the SG unit so that the changes in magnitude will be small under light loading conditions. As a result, the QSG and δSG parameters, owing to the fast response, are used as the main parameters to design an exact LOFPS index. Considering the δSG increment and QSG decrement after LOF incidence, the ratio of δSG to QSG is suggested. In fact, by an increase in δSG and a decrease in QSG parameter during LOF events, the variations of the suggested index ΔδSG/ΔQSG will be more visible. It should be noted that the VSG parameter due to decreasing changes during LOF events is utilised as the activator to evaluate the variation of the proposed index. 2.2 Flowchart of the proposed LOFPS The suggested algorithm to detect LOF/PSD events based on the analysis provided in the previous section is shown in Fig. 3. First, the voltage, current, and load angle from the SG target are measured, and reactive power is calculated. Whenever any decreasing variation is observed in VSG (ΔVSG < 0), this situation can be one of the LOF or PSD events. Hence, to address this issue properly, the ΔδSG/ΔQSG index is calculated and the magnitude of the ΔδSG/ΔQSG variations is measured for 200 ms. If the magnitude of the ΔδSG/ΔQSG is greater than β, the LOF event can be identified. Fig. 3Open in figure viewerPowerPoint Flowchart of the proposed LOFPS To determine the appropriate time interval for monitoring as well as the threshold values of the proposed LOFPS, exhaustive simulations including both LOF and PSD events have been carried out. The results show that, given a margin of confidence, all LOF and PSD events can be easily detected in 200 ms. Moreover, the best threshold value (β) that can exactly cluster all events is 1.2 × 105 deg/p.u. 2.3 Mathematical analysis of the proposed LOFPS In this section, to indicate the efficiency of the proposed LOFPS of this paper, an analytical evaluation based on the configuration is presented in Fig. 1. According to Fig. 1 and based on the rules of the power flow between two buses, the QSG parameter between the internal and terminal voltage of the SG can be written as ((1)) Simplifying (1) by (2) and (3) gives the final equation of δSG as presented in (4) ((2)) ((3)) ((4)) As is shown, the variation of δSG is dependent on ESG, VSG, and QSG parameters. It is clear that ESG is greater than VSG and its decrement is smaller than VSG, while QSG decrement is noticeable than VSG and ESG, leading to a remarkable decrement in the argument of cos−1. Hence, the δSG parameter will be increased after the LOF incidence. On the other hand, based on QSG presented in (1), by an increase in δSG and decrease in ESG and VSG, the QSG decrement due to more decrease in the first part of (1) and increase in the second part, will be remarkable than other parameters. Deviations of the selected parameters including QSG and δSG can be expressed as (5) and (6), respectively ((5)) ((6)) where is the base value of the QSG parameter to obtain per-unit value. Therefore, by using the ratio of the selected parameters for identifying both LOF/PSD events, we have ((7)) Based on the above analysis, the most important features of the proposed index in comparison with the introduced LOFPS in [21] can be summarised as follows: Since the variation of VSG and QSG is in the decrement direction and δSG is in the increment direction, the final direction (increment/decrement) and amount of changes of the offered (QSG × δSG × VSG) index in [21] are completely dependent on conditions of the LOF event (such as the loading level). Hence, setting a proper threshold value to identify both LOF/PSD events is quite complex. Nevertheless, the changes in the proposed index of this paper during all LOF events will be always increasing, and setting the threshold value to distinguish LOF/PSD events will be simple. The proposed LOFPS in [21] is due to the variations in the changing pattern of the selected parameters, which requires a coefficient to increase the clarity of variations of the proposed index. Meanwhile, the clarity coefficient should be adapted for various power systems and SG units, leading to some complexity in LOFPSs. However, the requirement of this clarity coefficient is thoroughly removed by the proposed index in this paper, due to the use of per-unit value of the QSG, which causes to be multiplied in variation of the selected index (as shown in (7)). As a result, the visibility of the offered index during all LOF events is quite guaranteed. 3 Simulation studies and discussions A schematic diagram of the system study is depicted in Fig. 3, whose simulation data are reported in Table 3. Table 3. Simulation data of the system under study Equipment Quantity Value generator nominal power 200 MVA line voltage 13.8 kV Xd, Xd′, Xd″ 1.305, 0.296, 0.525 p.u. Xq, Xq′, Xl 0.474, 0.243, 0.18 p.u. Td′, Td″, Tqo″ 1.01, 0.053, 0.1 s stator resistance 0.0028544 p.u. inertia coefficient 3.2 s transformer nominal power 210 MVA V1n 13.8 kV V2n 230 kV R1 = R2 0.0027 p.u. L1 = L2 0.08 p.u. Rm = Lm 500 p.u. transmission lines R0, R1 0.413, 0.1153 Ω/km L0, L1 0.00332,0.00105 H/km C0, C1 5.01×10−9, 11.33×10−9 F/km main power grid voltage 230 kV short-circuit level 1000 MVA As shown in Fig. 3, the system study consists of a 200 MVA SG unit, which is connected to the main power grid through a step-up transformer and two transmission lines. Meanwhile, it has two loads connected on both sides of the power transformer. To illustrate the efficiency of the proposed LOFPS, a comprehensive simulation study regarding the relevant standards is performed under various LOF and PSD events. It is worth mentioning that in all simulation studies, LOF and PSD events have occurred at t = 3 s. It is worth mentioning that the provided simulation results include the maximum recorded samples for the ΔδSG/ΔQSG index during the specified processing time in tabular form, and then, for instance, some of the provided data have been graphically presented. 3.1 LOF occurrence tests The efficiency of the LOFPSs can be seriously affected under light- or high-loading conditions of the SG as well as the power factor under loading conditions. Hence, to perform an in-depth evaluation of the suggested LOF protection index, two various categories of LOF tests are carried out on the system studied and shown in Fig. 4. The first category consists of light-loading conditions of SG with various power factors. The second category includes high-loading conditions of SG with various amounts of power factors. The corresponding simulation results are provided in Table 4. According to the presented data, in either category of light- or high-loading levels of the SG, the value of the proposed index has exceeded the threshold value. As a result, the proposed LOF protection index reveals quickly and securely all the situations as a LOF event occurs. To indicate the dynamic deviation of the proposed index during LOF incidence, some of the results presented in Table 4 are graphically depicted in Figs. 5 and 6. Table 4. Obtained results for LOF incidence considering various loading levels of SG with different power factors Test type SG loading, MW, MVAr ΔδSG/ΔQSG, deg/p.u. Detection time, s light loading 1 100 + j20 3.53 × 105 3.092 2 80 + j15 3.47 × 105 3.090 3 60 + j10 3.14 × 105 3.088 4 40 + j8 2.81 × 105 3.083 5 30 + j4 1.68 × 105 3.078 6 30 − j4 1.57 × 105 3.076 7 40 − j8 2.76 × 105 3.081 8 60 − j10 3.42 × 105 3.086 9 80 − j15 3.57 × 105 3.088 10 100 − j20 3.84 × 105 3.091 high loading 11 180 + j35 12.8 × 105 3.110 12 160 + j30 7.98 × 105 3.102 13 150 + j25 4.15 × 105 3.091 14 130 + j20 3.91 × 105 3.087 15 120 + j15 3.67 × 105 3.079 16 120 − j15 3.26 × 105 3.076 17 130 − j20 3.84 × 105 3.086 18 150 − j25 4.12 × 105 3.091 19 160 − j30 9.27 × 105 3.098 20 180 − j35 12.1 × 105 3.100 Fig. 4Open in figure viewerPowerPoint Single-line diagram of the test system Fig. 5Open in figure viewerPowerPoint Simulation results for the LOF incidence during light loading situations of the SG unit Fig. 6Open in figure viewerPowerPoint Simulation results for LOF incidence during high-loading situations of the SG unit The following analysis can be performed on the obtained results in Table 4 and Figs. 5 and 6. As can be seen in Figs. 5 and 6, although the proposed ΔδSG/ΔQSG index is changed after a small time delay (around 100 ms) after LOF occurrence, all LOF incidences can be properly identified just within 200 ms owing to the remarkable spike in this index. According to the data reported in Table 4, the loading conditions of the SG unit do not have any effects on the proper operation of the suggested index. It means that during all LOF events under various loading conditions a significant spike has been observed in the proposed index, which is greater than the threshold value. Dynamic variations of the proposed index in Figs. 5 and 6 indicate that increasing the loading levels of the SG unit leads to the time delay of the ΔδSG/ΔQSG index to be greater than the light-loading ones. However, the time delay will be around 100 ms for the worst circumstances. Based on Table 4, the detection time of the LOF events under high-loading conditions is longer than light-loading ones, though all LOF events can be detected within the specified period of study. 3.2 Short circuit fault (SCF) tests The SCF incidence may mislead the efficiency of the LOFPSs to falsely identify the LOF event because the changing pattern of the selected parameters will be similar during both SCF and LOF events (e.g. VSG decrement, δSG increment). In this part, the efficiency of the proposed LOF protection index has been evaluated for various SCFs, i.e. single-, double-, and triple-phase faults with a duration of 100 ms under various loading conditions, though the triple-phase faults are commonly considered for evaluating the efficiency of the LOFPSs. The obtained results from the extensive simulation studies are reported in Table 5. As can be seen, during all types of SCF events, the changes in the magnitude of the proposed LOF protection index are smaller than the threshold value. As a result, the proposed LOF protection index is quite reliable and quick to differentiate the SCF events from the LOF events. The dynamic variation of the proposed index for some of the single-, double-, and triple-phase types of the SCF events has been depicted in Figs. 7–9. The following analysis can be performed for the detection of the SCF events by the proposed LOFPS in this paper. According to Table 5, although changes in the magnitude of the proposed ΔδSG/ΔQSG index during the three-phase fault are greater than single- and double-phase faults, it remains smaller than the specified threshold value for all tests. Hence, all SCF tests have been reliably identified as PSD events by the proposed index within 200 ms after the incidence. In Figs. 7–9, unlike LOF events, the offered ΔδSG/ΔQSG index does not have any time delay during the SCF events so that it has deviated immediately at the instant of SCF occurrence. As Table 5 shows, SCF incidence in the presence of various loading levels of the SG unit does not have any adverse effects on the proper operation of the proposed LOF protection index, and changes in the magnitude of the suggested index are smaller than the threshold value during all tests. Table 5. Obtained results for different short-circuit faults in the presence of various loading level of SG Test type SG loading, MW, MVAr ΔδSG/ΔQSG, deg/p.u. Detection time, s single-phase fault 1 180 + j35 1.34 × 104 3.200 2 160 + j30 4.29 × 104 3.200 3 130 + j24 4.12 × 104 3.200 4 100 + j18 3.97 × 104 3.200 5 80 + j12 3.73 × 104 3.200 6 80 − j12 3.85 × 104 3.200 7 100 − j18 3.45 × 104 3.200 8 130 − j24 2.67 × 104 3.200 9 160 − j30 2.15 × 104 3.200 10 180 − j35 1.83 × 104 3.200 double-phase fault 1 180 + j35 1.83 × 104 3.200 2 160 + j30 4.31 × 104 3.200 3 130 + j24 4.45 × 104 3.200 4 100 + j18 3.87 × 104 3.200 5 80 + j12 3.91 × 104 3.200 6 80 − j12 4.01 × 104 3.200 7 100 − j18 1.08 × 104 3.200 8 130 − j24 2.68 × 104 3.200 9 160 − j30 1.98 × 104 3.200 10 180 − j35 1.68 × 104 3.200 three-phase fault 1 180 + j35 8.23 × 104 3.200 2 160 + j30 8.74 × 104 3.200 3 130 + j24 8.96 × 104 3.200 4 100 + j18 9.16 × 104 3.200 5 80 + j12 9.83 × 104 3.200 6 80 − j12 8.21 × 104 3.200 7 100 − j18 7.67 × 104 3.200 8 130 − j24 7.12 × 104 3.200 9 160 − j30 6.29 × 104 3.200 10 180 − j35 5.18 × 104 3.200 Fig. 7Open in figure viewerPowerPoint Simulation results for the single-phase type of the SCF event Fig. 8Open in figure viewerPowerPoint Simulation results for the double-phase type of the SCF event Fig. 9Open in figure viewerPowerPoint Simulation results for the triple-phase type of the SCF event 3.3 Load injection tests The other PSD events that can create transient changes in the electrical parameters lead to the mal-operation of the LOFPS, which is the load injection event. In this part, the effects of load injection with various power rates on the efficiency of the offered LOF protection index are investigated. The obtained results from simulations are shown in Table 6. As can be seen in Table 6, in all tests of load injection, variation of the proposed ΔδSG/ΔQSG index is smaller than the threshold value. Consequently, the proposed index properly detects all tests as PSD events. For more clarity, some of the obtained results during both light and high load injections are plotted in Figs. 10 and 11. The following results explain the highlights of the proposed index performance under load
Referência(s)