Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
18 Developing the Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems A. Research Plan This section of the Final Report consists of a summary of the problem being researched and the project objectives. A.1 Research Problem Statement Improving the reliability of railway electric systems and preventing electrical outages due to insulator flashover and insulation failure are difficult because of aging infrastructure and the lack of continuous monitoring of the state of the systems. Leakage currents resulting from insulator failure or contact of nonconductive surfaces, such as animals or trees, with the conductor fall below the detectable level of conventional relays. Because these faults are often caused by energized conductors within public reach, they pose a great threat to personal safety and property security. More often than is desirable, the service that electric railway systems provide to commuters is interrupted due to electric power disruptions. To minimize these service dis- ruptions there is a need to design and develop a smart sensor network system to do predictive electric system monitoring and to keep faults from occurring. A.2 Research Objective and Scope The University of Akron (UA), in collaboration with Exacter, Inc., and M2, Inc., developed a research program that aims to develop a health monitoring system for DC railways. The proposed solution is based on determining the high frequency behavior of DC railways and characterizing the RF emissions from the network. A.3 Research Approach The goal of the project is to develop low-level leakage current detection methods using two approaches. The first approach involves injecting a high frequency signal into a segment of the power line nonintrusively, and determining the high frequency impedance of the selected line segment. Segmentation of the line is achieved by applying a virtual blocker at each end of the chosen segment. The high frequency impedance of that segment is thus linked to the condition of the cable and the level of the leakage current in the segment. The second approach measures the RF emissions from the power network, and relates the condition of the network to the frequency content of those emissions. Other than these two proposed approaches, to the best of our knowledge, there has been no other method developed to determine the low-level leakage currents in DC electric railway systems. A P P E N D I X
Developing the Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems 19 In Phase I, the research team conducted a literature survey of the existing technologies and standards for leakage current modeling and prediction in transit railway systems. The team developed a model and a simulation of the electric transit system to test the feasibility of the proposed injection method in determining the level of the leakage current. Based on the developed model, the team tested the feasibility of the injection method, with positive results. The research team had several meetings with transit agencies, in Cleveland and Philadelphia, and have done field testing for further verification and validation of the proposed method. For the second approach, the Exacterâs AC emission measurement system was modified to DC systems and tested at the UA Labs. The team installed the measurement units on Greater Cleveland Regional Transit Authority (GCRTA) transit cars for data collection. In Phase II, the team finalized the development of the active clamp injector units and deployed them on a GCRTA transit railway segment for testing. Similarly, Exacter RF emission units were modified based on the tests and deployed as well on a GCRTA transit car. For Phase II, the research team prepared TCRP Research Report 211: Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems. B. Accomplishments and Significant Findings This section of the Final Report addresses the work accomplished for Phase I and Phase II of the project. B.1 Task Status Task 1: Survey of DC-Powered Rail Transit Systems Review of Standards. The research team acquired the American Railway Engineering and Maintenance-of-Way Association (AREMA) manual for railway engineering, and National Fire Protection Association (NFPA) 130: Standard for Fixed Guideway Transit and Passenger Railway Systems. A summary follows. i) IEEE Standard 80 The Institute of Electrical and Electronics Engineers (IEEE) Standard 80 is concerned with outdoor AC substations, either conventional or gas-insulated. Distribution, transmission, and generating plant substations are included. With proper caution, the methods described in this standard also apply to indoor portions of such substations, or to substations wholly indoors. The IEEE Standard 80 does not cover the grounding problems particular to DC substations (which are used in DC railway systems) (1). ii) National Electrical Code The National Electrical Code (NEC), or NFPA 70, is a regionally adoptable standard for the safe installation of electrical wiring and equipment in the United States. It is part of the National Fire Codes series published by the NFPA, a private trade association. Despite the term ânational,â it is not a federal law. It is typically adopted by states and municipalities in an effort to standardize their enforcement of safe electrical practices (2). Railway applications (railway power, signaling, and communications wiring) are not covered by this standard (2). iii) National Electrical Safety Code The National Electrical Safety Code (NESC), or ANSI Standard C2, is a United States standard of the safe installation, operation, and maintenance of electric power and communication utility systems including power substations, power and communication overhead lines, and power and communication underground lines. It is published by IEEE. âNational Electrical Safety Codeâ and âNESCâ are registered trademarks of the IEEE. According to NESC, the rated dry flashover voltage should not exceed the values in Table 1 (3).
20 Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems iv) NFPA 130, Standard for Fixed Guideway Transit and Passenger Rail Systems The NFPA 130 covers life safety from fire and fire protection requirements for fixed guideway transit and passenger rail systems, including, but not limited to, stations, train ways, emergency ventilation systems, vehicles, emergency procedures, communications, and control systems. This standard has information about the safety inside the vehicles or the train like motor pro- tection, insulation between the wires, and requirements for firefighting, but it does not cover the insulation of the overhead lines or fault detection in the railway system (4). v) AREMA AREMA is a North American railway industry group. It publishes recommended practices for the design, construction, and maintenance of railway infrastructure, which are requirements in the United States and Canada. AREMA Chapter 12, Rail Transit Chapter 12, Rail Transit, is one of three chapters constituting the Passenger and Transit sectors of the rail industry. The other two are Chapter 11, Commuter and Intercity Rail Systems, and Chapter 17, High Speed Rail Systems. These three chapters cover civil and structural engineering, mechanical and electrical engineering, train control and communication systems, track and related infrastructure, passenger stations, maintenance shop and yard facilities, and rail vehicles. Chapter 12 is generally used for rail systems that carry only passengers on dedicated or closed rights of way and that fall under the purview, in the United States, of the Federal Transit Administration (FTA) of the U.S. Department of Transportation (U.S. DOT). In some systems, the right of way is shared by freight railroads and passenger rail transit vehicles. In these circumstances, both the Federal Railroad Administration (FRA) and the FTA are involved. The chapter has information about construction and maintenance of the railway system as it pertains to the safety of the passengers in trains and stations and the safety of the people passing by the transit system, but it provides no information about the electrical system leakage current or fault detection (5). AREMA Chapter 33, Electrical Energy Utilization Chapter 33, Electrical Energy Utilization, provides major guidelines in design, operation, protection, and safety of the electrical system of transit systems. The guidelines provide important information about the interfaces of the control and sensing units. Section 3 of Chapter 33 provides standards about the recommended voltages to be used at different points of the electrification systems. The guidelines are very important for selecting and designing any sensors for interfacing into the system. Section 5 provides detailed information about the railway electrification compatibility with the signal systems, especially the sensors requiring injection into the line requirements to comply with these guidelines. Clearances for third rail configurations are provided in Section 2. Catenary systems design, interconnection, ampacity ratings, and sizing for various envi- ronmental and operating conditions are covered in Section 4. Section 7 provides important guidelines in grounding and bounding to protect signal and power system integrity. This section is important in modeling the leakage current on the negative side of electrification systems. Section 12 covers the power supply interface, basic system configuration including substation locations, sectionalizing, overall power system control, and protection and contingency operations. This information is useful in integrating the sensor into the control and operation of the electrification system. 6.9 13.2 13.2 55 23 75 34.5 100 46 125 69 175 Nominal voltage (kV) Rated dry flashover voltage (kV) 0.75 5 2.4 20 Table 1. Dry flashover voltage for insulators.
Developing the Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems 21 DC Systems and Insulation Failure Mechanisms. Due to the lack of previous work about leakage currents in the positive side and failure in railway systems, the research team reviewed the literature on DC power line transmission systems and their failure mechanisms. Insulators are vital components of overhead transmission lines and substations. Insulator and transmission line faults caused by aging of the line, contamination, and lightning largely determine the performance of the system. Insulator failures cause flashovers, but before flash- overs occur, elevated leakage currents would pass through the insulators. Surface contamination of insulators is caused by the ambient salt content in coastal areas, pollution in industrial areas, and dust in agricultural and rural areas. Leakage currents usually occur when the insulator string is under high humidity conditions and bird excreta. The insulator will heat up at the activation of the line, and so as the insulator goes through wet/dry cycles, a dry film forms and grows on the insulator surface (6, 7). Although contamination affects both HVAC and HVDC, the response to the contamination differs between AC and DC insulators. Cheng et al. (8) mention three reasons for the different response: 1. The accumulation process of contaminants is different. DC insulators attract more pollution than their AC counterparts under similar weather conditions. 2. DC outputs a constant voltage or current, which further increases the flashover problem on DC insulators. Scintillations are therefore harder to extinguish on DC insulators. 3. DC flashovers also occur at a lower voltage level than AC flashovers; hence, DC flashovers occur more regularly. A study has been done to relate the leakage current levels to insulator condition (9). Laboratory experiments were conducted on a 69 kV post insulator with a unified specific creep- age distance (USCD) of 46 mm/kV. Figure 1 relates the leakage current to the severity of the pollution. Beyond a certain leakage current, the probability of flashover in the insulator increases significantly. Experiments showed that the maximum leakage current across the insulator can be expressed in terms of pollution severity by the formula in Equation 1: 1.45 (1)0.77Ihighest = Î³ Figure 1. Leakage current as a function of pollution severity.
22 Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems where Ihighest and Î³ are expressed in amp, and by the equivalent salt deposit density (ESDD), in mg/cm2, respectively. Using this formula, a relationship between leakage current and pollution severity can be plotted for a specific insulator. This information could relate the condition of the insulator to the leakage current, provided there is an easy way to measure the leakage current in real time. Task 2: Performance Specifications and Compliance with Current Practices The research team found no practices for leakage current detection in electric transit systems. As part of the survey, the team summarized some of the current practices for protection schemes, such as overcurrent protection, distance protection, and differential protection. i) Overcurrent Protection Overcurrent protection relies on the measurement of the main current passing through the electric transit network. Overcurrent exists when the current exceeds the equipment rating or the capacity of a conductor. This may be due to an overload, short circuit, di/dt (current fluctuations), or ground fault. Overcurrent devices protect conductors and equipment from overcurrent. If the measured current through the branch exceeds the maximum allowed current, then the circuit trips and opens the circuit breaker. This method is commonly used in railway systems. The main disadvantage of this method is that the setting for the tripping current must always be above the maximum operation current in the system, and thus it can detect no high impedance faults or any leakage currents. ii) Distance Protection The distance protection relay detects both voltage and current to determine the line impedance. It is typically used in AC grids to protect a specified zone in the transmission, as shown in Figure 2. A fault on a circuit will generally create a sag in the voltage level. If the ratio of voltage to current (impedance) measured at the relay terminals falls within a predetermined level, the circuit breaker will operate. This is useful for reasonably long lines because their oper- ating characteristics are based on the line characteristics. This means that when a fault appears on the line, the impedance setting in the relay is compared with the apparent impedance of the line from the relay terminals to the fault. If the relay setting is below the apparent impedance, it is determined that the fault is within the protection zone. Although the project targets DC systems, there is an AC component to unfiltered 12-pulse DC rectification systems. While this method works well for stationary systems, it cannot be applied to the railway system because the main loads in this system (the trains) are moving, and thus the fault impedance cannot be determined accurately (10). Figure 2. Distance protection with three protection zones.
Developing the Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems 23 iii) Differential Protection The main principle in the differential protection approach is that the input and output currents of the protected zone are equal, except in the case of internal faults (the algebraic sum of all currents in the protection zone is ideally zero in all cases except for internal faults). Figure 3 shows the general differential protection scheme. The current sensors on either side of the protected element are connected by pilot wires, while the operating coil of the relay is placed between the pilot wires. The circuit will open when the difference between the measured currents is higher than the set point. Applying this method to the railway system is both difficult and complicated because of the moving load in the system. A high level of communication must be used between all the trains and the main stations (11). Problems with the conventional protection methods: â¢ All these methods cannot detect leakage currents and high impedance faults. â¢ All these protection methods use the current and the voltage of the system itself; hence, the sensors need to be rated to sense the large current and voltages with acceptable resolution. â¢ These protection methods mentioned cannot be used when the system is deenergized. These techniques rely on measuring the operating voltages and currents, which are not functional when the system is off. Task 3: Classification of Tools and Practices The proposed active clamp sensor will be measured by its i. System level performance, ii. Operations and maintenance levels by transit system personnel, and iii. Point level near the public. i) System Level Performance The proposed monitoring solution can be narrowed to a desired segment of the transit system with no changes in the arrangement of the sensors. The heart of the monitoring method is based on real-time impedance measurement and analyses of the power line at the high frequency range. The proposed solution is cost-effective, nonintrusive, reliable, and capable of tracking impedance variations to detect faults in the transit system. Unlike other protection methods, loading will not affect the measurements. The method is based on nonintrusive high frequency signal injection into the middle section of the desired power line segment, via a sensor in the middle of the segment, and blocking the injected signal with two other sensors at opposing ends of the segment. The blocking technique is based on current injection through magnetic coupling MZE Figure 3. Differential protection scheme.
24 Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems into the railway power line to cancel the flow of the high frequency injected signal. The new method requires no direct connections to the overhead power line. ii) Operations and Maintenance Levels by Transit System Personnel One smart sensor (working as an injector) injects the high frequency signal into the line while the signal is blocked by the two other smart sensors (working as blockers). The injected signal propagates through the railway power line. The impedance of the line at high frequency is monitored by tracking the current resulting from the injected signal. The difference in the premeasured and actual impedance provides information about the health condition of the power line. Since the power consumption of the sensor is in the range of 10 W, even for continuous monitoring, the sensor can be easily powered by a battery. Energy harvesters can be included into the sensor to limit the maintenance associated with battery replacement. The sensor is effective even when the transit system is powered off, which makes it easier for the maintenance team to monitor the condition of the transit system before powering it up again. iii) Point Level Near the Public The voltage injected by the monitoring system will be in the range of 5 V to 10 V, therefore the system will be very safe and will not pose any danger to the public. At the suggestion of the TCRP Project D-17 panel, the research team focused its efforts in this task more toward developing the tools needed to develop new sensors for the application at hand. The railway transit system has been modeled and incorporated into the Matlab Simulink simulation environment to test the performance of the new sensor technology. Electric Railway Modeling. To model a railway electrical system, the resistance of the tracks and the resistances between the tracks and the ground are assumed to be uniformly distributed throughout the railway. The train is modeled as a current source whose current depends on the operating condition of the train, such as accelerating or braking. The main challenge in modeling the stray current is that it does not leak from specific points on the track, but rather throughout the track. Therefore, the stray current may be treated in a similar way as the current of the power system transmission lines, where the voltage and the current at a point x on the line are given by Equation 2 and Equation 3 (12), (13), (14): ( ) = + â (2)1 2i x C e C eyx yx ( )( ) = â + â (3)0 1 2u x R C e C eyx yx for 0 < x < 1 0= =y R R R RRG G where, u(x) represents the potential of the conductor at point x, i(x) is the current in the conductor at point x, R is the longitudinal resistance of the conductor (â¦/km), and RG is the leakage resistance between the conductor and ground (â¦ km). Equations 2 and 3 provide the exact amount of current and touch voltage at any point on the line. However, it is impractical to implement them using physical hardware because it requires an infinite number of paths to ground (15), (16), (17). With a railway system emulator, the length of the conductor is not long because the typical distance between any two stations in a railway system is within 3 km to 8 km. Therefore, with a limited number of paths to ground,
Developing the Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems 25 an approximate model can be obtained. As explained in (18), (19), (20), the expression for the highest value for the stray current for a grounded system is given by Equation 4: 2 (4) 2 =I IR l R stray G where, I is the train current and l is the distance between every two stations. To determine the number of ground paths that would be sufficient to obtain a model that is accurate enough and feasible for practical implementation, a comparison study was conducted. This study yielded the results in Table 2. In this table, the maximum stray current was obtained for different ground paths and different distances between two stations. The currents using an infinite number of paths are calculated using Equation 2. From Table 2, the maximum values of the stray current differ from the exact current at station lengths of 6 km and 10 km, with a percentage error in both cases of less than 0.001. Because in a typical railway system the distance between two stations is typically 1 km to 2 km, only two paths to ground turn out to be sufficient to effectively simulate grounded railway systems. This leads to the model structure in Figure 4. The performance of the proposed circuit was analyzed using MATLAB/SIMULINK. Table 3 includes the simulation parameters used for this analysis. In this table, the variable I represents the train current, RO is the overhead line resistance, R is the resistance of the rail tracks, RG is the resistance between the rail and ground, E is the voltage of the railway system, and D is the distance between the two stations. While the train is moving from Station 1 to Station 2, Figure 5 shows the total stray current leaking from the rail, while Figure 6 shows the touch voltage of the rail. # of ground paths Current at 1 km (A) Current at 3 km (A) Current at 6 km (A) Current at 10 km (A) 2 0.0265 0.5622 2.2449 6.2112 8 0.0265 0.5622 2.2457 6.2170 14 0.0265 0.5622 2.2458 6.2174 Current for infinite paths 0.0265 0.5622 2.2458 6.2176 Table 2. Maximum stray current for different modeling paths. Figure 4. Impedance model of the railway system.
26 Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems 1,000 A 0.08 0.04/km /km 80 km 1,100 V 4 km Table 3. Parameters values used in the simulation. Figure 5. Total stray current leaking from the rail as a function of the train distance from the substation. Figure 6. Rail touch voltage as a function of the train distance from the substation. Insulator Modeling. For DC electric railways, the impedance of the insulator is infinitely large. However, because the proposed technology requires the injection of the high frequency current into the line, the research team had to model the insulators and incorporate the model into the simulations. The research team expected that the nonintrusive high frequency active clamp sensor could determine the condition of the insulator effectively. To achieve this objective, the team simulated the performance of the sensor based on the high frequency insulator models determined experimentally.
Developing the Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems 27 Insulators have two metallic ends to suspend the conductor to the tower. There is a non- conductive material between the two ends to keep a high level of insulation between conductor and ground. The insulator structure and impedance model are illustrated in Figure 7. The impedance model of the insulator has a characteristic similar to that of a capacitor, and thus it is possible to express its impedance as in Equation 5: (5)= wÎµ Z d j A ins where d is the length of the insulator, A is the area of the two conductive plates, and Îµ is the dielectric constant, which depends on the materials used to build the insulator (21). Pollution effects can be added to the insulator impedance as in Equation 6: = + â (6)Z Z Z Z Z ins tot ins poll ins poll To see the effect of pollution on the impedance of the insulators, laboratory experiments have been conducted. A test setup was developed at the UA Alternative Energy Lab (AEL), as shown in Figure 8. The impedance between the positive line and ground is measured using a highly accurate LCR meter (Agilent E4980A) at different frequencies, with and without the effect of pollution. Pollution around the insulator is implemented by applying Kaolin white clay (recommended by IEEE testing procedures), as shown in Figure 9. The measurement frequency is swept from 20 Hz to 1 MHz. Figure 10 shows the impedance of the insulator, with and without contamination. The team has also tested the impedance of the insulator at higher humidity conditions. Figure 10 presents the impedance of the insulator under various conditions and frequencies. There is a clear difference in the impedance measurements between the uncontaminated and contaminated insulators. It was found that the impedance difference for the two test conditions peaks to a 50% difference at around 10 kHz. The difference was much more significant when the team compared the high frequency impedances of the wet and dry insulators. Integrated DC Railway Modeling. With the insulator modeled for high frequency signal injection, the overall electric railway system was modeled and simulated. Figure 11 presents the integrated electric railway model. The simulation is provided in Task 4. Figure 7. Insulator structure and the impedance representation with pollution.
Figure 8. Insulator test setup for high frequency impedance measurement. Figure 9. Testing the effect of pollution on the insulator using Kaolin white clay. Figure 10. The impedance of the insulator with and without contamination at various frequencies.
Developing the Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems 29 Task 4: Comparison Metrics Development Through Simulations As in Task 3, the research team focused on developing the new sensors referred to previously. In this task, the team developed a condition assessment to distinguish between good and bad components that could be manifested in their high frequency behaviors. The railway power line was modeled according to DC power line system data, the insulators were modeled according to the experimental results obtained from both healthy and polluted insulators, and the train was modeled as a current source that can adapt to different types of operating patterns. Based on experimental measurements, the insulation impedance is modeled as a resistor and capacitor in parallel. The line impedance is modeled in a distributed fashion so the movement of the transit vehicle may be easily accounted for in the model. The parameters provided in Table 2 are used for the simulations, line capacitance per kilometer is taken as 25.48 nF. Figure 12 shows the leakage current through the healthy insulator as the transit vehicle moves through the railway. Figure 13 shows the leakage current through the contaminated insulator R7R5 RG1 Train Current V dc-1 V dc-2+ - + - R8R6 RG2 R1 R2 R3 R4 RR CC Figure 11. Integrated electric railway model. Figure 12. Leakage current through the healthy insulator as the transit vehicle moves through the railway.
30 Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems as the transit vehicle moves through the railway. Notice that the leakage current through the contaminated insulator is almost twice that of the healthy insulator. Although the leakage currents through the healthy and contaminated insulators are different, the magnitudes are small. With these techniques, it is difficult to determine these currents. An active clamping technique is used for leakage current detection. The impedance of the DC electric railway segment at high frequency contains information about the characteristics and operating conditions of the system. Real-time tracking of impedance changes can be used for leakage current detection and health monitoring of the electric railway segment. Figure 14 shows a smart sensor network to monitor the health condition of the railway network and to detect any fault in the bus, in real time, due to the poor health condition of insulators, conductors, and so forth. The smart sensor injects a high frequency signal into the bus while the signals are blocked by other smart sensors at the two ends of the desired segment. The injected signal propagates Figure 13. Leakage current through the contaminated insulator as the transit vehicle moves through the railway. High voltage DC bus Blocking signalBlocking signal Injected signal L i-1 L i Rail InsulatorInsulator Si-1 Si+1Si Figure 14. High frequency impedance detection for DC railway system.
Developing the Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems 31 through the bus. The impedance of the feeder, at high frequency, is monitored by tracking the injected signal. The difference between the premeasured and actual impedance provides infor- mation about the health of the DC bus feeder. Three smart sensors, one at each end of the desired segment and one at the transit car, block the propagation of the signal into undesired paths. Then the impedance of the segment at high frequency can be measured effectively. The research team simulated the movement of the transit car between two substations with healthy and contaminated insulators. The research team obtained the following simulation results, for different conditions, using the proposed active clamp technology: â¢ Both insulators are healthy, the measured impedance is 35.4 kâ¦ â¢ One insulator is polluted, and one is healthy, the measured impedance is 27.5 kâ¦ â¢ Both insulators are polluted, the measured impedance is 18.87 kâ¦ The simulation results proved the feasibility of the proposed technology to determine the condition of the network or the leakage current levels, because the leakage current is directly related to the health condition of the network. In this simulation, the high frequency injection method was also used to detect the leakage current through high impedance faults. The fault was simulated through a 50 â¦ impedance. The DC current drawn from the station was 22 A. Note this current is very large compared with that of high impedance faults, which can be between 1 A and 5 A. It will be very difficult to detect this leakage current with conventional protection methods. The DC impedance and the high frequency impedance were measured in both healthy and fault conditions to examine over current protection, distance protection, and the proposed high frequency impedance method. Figure 15 shows the DC current drawn from the station under both healthy and fault cases. Clearly, a 22 A current is hard to detect compared to that of the trainsâ rated current, which can reach over 1,500 A. In this case, 22 A will not be detectable. Figure 16 shows the ratio between the DC current in a healthy condition to the current in a fault condition. Figure 17 shows the DC impedance near the station in healthy and fault cases. Also, because of the train current, there is a small difference between the DC impedance in the two cases. Figure 15. DC current in both healthy and fault conditions.
32 Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems Figure 16. Ratio between the DC current in healthy and fault conditions. Figure 17. DC impedance in both healthy and fault conditions. Again, the 22 A leakage current is undetectable with the conventional over current protection schemes. Figure 18 shows the ratio between the DC impedance in a healthy condition to the impedance in a fault condition. Figure 19 shows the high frequency impedance near the station in healthy and fault condi- tions. Clearly, the impedance difference from 417 kâ¦ in the healthy condition to 405 â¦ in the fault condition is large. Basically, blocking the high frequency signal from the train creates a separate system, almost like an open circuit, which conducts only through the capacitance of the system. Thus, any fault or leakage current in the system, even in the order of 0.1 A, will create a short and change the high frequency impedance dramatically as illustrated in Figure 19. Since the system is operating in high frequency and isolated from the rest of the system, the large DC current drawn by the train does not affect the method.
Developing the Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems 33 Task 5: RF Emission-Based Technology The second approach is based on evaluating RF emissions from the deteriorated components of the electric railway system. The technology developed by Exacter, Inc., provides mobile surveys of the electric grid evaluating EMI (electro-magnetic interference) and partial discharge (PD) emissions (22). The project creates a system and method to continuously survey the electric railway, and identify the location of degraded insulators, connectors, and equipment for scheduled maintenance. During Phase I, Exacter RF sensors were placed in UAâs AEL, as shown in Figure 20. The algorithms were updated for DC emissions. A DC arc was generated and the performance of Figure 18. Ratio between the DC impedance in healthy and fault conditions. Figure 19. High frequency impedance in both healthy and fault conditions.
34 Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems the sensor in capturing the arc was evaluated. The DC algorithms were installed in the Exacter Technology Sensor and confirmed to function properly. A mobile sensor and sensing element were developed and installed on a transit car operating in the GCRTA network, as shown in Figure 21. Communication hardware and the software interface were established for continuous data capture for further analysis. The research team developed an auto-start/shutdown controller so the Exacter Technology sensors would require no interaction from railway personnel. The Cloud data collection system that transports data from the train sensors to Exacter for analytical review was established. The objective was to evaluate the possibility of measuring arc emissions from the traction motors to see if it is feasible to assist in a predictive insulation maintenance program to avoid motor flashover. Task 6: Data Collection and Analysis of RF Emission-Based Technology PD and EMI data were continuously collected as the GCRTA train was operating. The sensor system, emissions mapping software, and Cloud data concentration and storage network were proven very reliable. Data is collected by the sensors and sensing elements. The information is transferred to the Exacter analysis servers continuously on the Exacter Cellular Data Network. The information is evaluated upon receipt, and models for criticality were completed. An Analytical Dashboard provided visualization of the failure emissions. Final filtering and analysis algorithms were created to evaluate the collected emissions and to create criticality measures to prioritize maintenance. Task 7: Evaluation of Data Findings and the Location of Degrade Equipment RF Emission-Based Technology Data reduction and analysis was performed. The research team evaluated the best way to reduce background emission, which occurs as the train moves along. The trainâs motion causes Figure 20. RF sensor configured for DC emissions at UAâs AEL.
Developing the Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems 35 short bursts of arcing when the pantograph momentarily separates from the wires. This causes a strong broadband frequency burst, which lasts briefly in the time domain. The railway right of way runs parallel to 60 Hz electric distribution lines owned by First Energy. A 60 Hz sensitive Exacter sensor was included in the on-board Exacter Sensor Package. This sensor, which is readily available from Exactor, was used to sense and categorize 60 Hz emissions. The recorded DC emissions associated with the DC electric service network were identified as either transient conditions or steady state emissions. The steady state emissions, that is, those emissions that are repeatable and present at the same geospatial location, are the target leading indicators for DC system maintenance. Figures 22 through 25 present the train arcing data, DC fault and train arcing data, AC fault data, and AC fault and train arcing data, respectively, from the RF-based emission sensor. The team continued to receive data from the sensors installed on the GCRTA Red Line train. The Cloud-based network collected the uploads and processed the data automatically. Antenna installed on the transit car. RTA transit car on which the sensor is installed. Testing of the data collection for the RF emission-based sensor. Real-time data collection for data analytics. RF emission-based sensors for DC transit cars. Installation of the RF-based sensor by GCRTA personnel. Figure 21. RF emission-based sensor installation on a transit car at GCRTA.
36 Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems Figure 22. Example of train arcing data from the RF emission sensor. Figure 23. Example of DC fault and train arcing data from the RF emission sensor.
Developing the Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems 37 Figure 24. Example of AC fault data from the RF Emission Sensor. Figure 25. Example of AC fault and train arcing data from the RF emission sensor.
38 Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems The results show the locations of points of electric deterioration on the railways electric infrastructure. The team noticed consistent background RF noise mixed in with the electric infrastructure survey data. While we succeeded in filtering the data, we would like to collect electric infrastruc- ture data without the background noise to evaluate the filters. To do this, we must gather the field data from a platform that does not create the background noise. One alternative is to install the sensors on a diesel-powered maintenance vehicle to survey the electric infrastructure without the background noise created by the electric railway engine. Once this survey is completed, the team could evaluate the success of the data filters. The research team found that the ultrasonic acoustic emission locating equipment has not been as effective on the DC system as it is on the AC systems for which it was designed. The team could redesign the sensors so they will be more sensitive to DC source arcing, tracking, and leakage. Task 8: High Frequency Injection-Based TechnologyâPhase I As stated in the research approach section, the active clamp approach involves injecting a high frequency signal into the power line nonintrusively and determining the high frequency impedance of the selected line segment. Segmentation of the line is achieved by applying a virtual blocker at each end of the chosen segment. The high frequency impedance of that segment is thus linked to the condition of the cable and the level of the leakage current in the segment. In this task, a high frequency injector transformer was built around the nanocrystalline circular cut core for the transit system. The active clamp system measured high frequency impedances of up 20 kâ¦ at 100 kHz. Based on the tests the team conducted at GCRTA, the range for measurement was adequate in determining the segment impedance of the DC electric railway. The active clamp sensor has three main elements to measure the high frequency impedance of the selected segment in the transit system: â¢ A high frequency signal injector (to inject the high frequency voltage into the system); â¢ A current sensor that measures the current induced into the system to determine the high frequency impedance; and â¢ A blocking mechanism to block the high frequency current from entering the train or stations. Figure 26 presents the components of the active clamp sensor. The sensor can work as either the injector or the blocker. The critical part of the sensor is the high frequency transformer that works as a coupling element to the line for high frequency injection. The high frequency transformer was designed and built as part of the active clamp sensor. The first step of designing the nonintrusive voltage injector system was to determine the voltage level of the injected signal in the power line. The main factor in determining the required voltage level was the maximum impedance of the power line section the team wanted to measure (this includes the impedance of the line and the capacitance path to the ground) and the sensitivity of the current sensor being used. The current sensor should be able to measure the current in the line in the case of the maximum line impedance. From this measurement, the voltage level of the injected signal in the line was determined using Equation 7: = (7)2 min maxV I Z where Imin is the minimum current that can be sensed by the current sensor and Zmax is the maximum impedance of the power line section.
Developing the Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems 39 The core of the high frequency transformer must be designed so it does not get saturated by the DC current in the power line. This can be done by increasing the size of the core, choosing a material with high permeability (such as a nanocrystalline material), and adding an air gap if need be. Usually in the railway systems, with a maximum current higher than 1,000 A, using bigger cores and nanocrystalline materials are not enough, thus the air gap must be added. The air gap in the core can be determined by Equations 8 and 9: = â (8)maxI R Rc air = Î¼ â Î¼ (9) 2 g B A I lsat o max r where Imax is the maximum DC current in the power line, â is the flux in the core, Rc and Rair are the reluctances of the core and the air, respectively, g is the minimum air gap to prevent saturation, Bsat is the saturation flux density of the core, Î¼o and Î¼r are the permeability of the air and the core material, respectively, A is the cross sectional area of the core, and l is the circumference of the core. The number of turns in the primary (injecting side) is determined by Equation 10: = (10)2 1 1 V V N where V1 is the primary voltage, V2 is the secondary voltage, and N1 is the number of turns in the primary side. Figure 26. Components of the active clamp sensor.
40 Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems The next step is to determine the current needed by the system. This current is the sum of the current that reaches the secondary, the current dissipated in the winding, and the current needed to develop the flux in the core (especially in the air gap). Next, the capacitor between the driver and the core must be designed to achieve a resonance frequency between the capacitor and the primary inductor. This will boost the voltage in the primary without the need of a high voltage from the current driver. The value of the capacitance can be obtained by Equation 11: ( ) = p 1 2 (11)2C f L where C is the capacitance, f is the frequency, and L is the inductance of the primary. Where it is necessary to measure at different frequencies, a ladder of capacitors can be installed rather than one capacitor to achieve any needed value of capacitance. The maximum DC current is assumed to be 1,000 A. Given an air gap of 0.7 mm and using a material with Î¼r = 7,000, the team selected the core cross section area to operate in the linear region. Using this information, nanocrystal- line circular cut cores were selected and acquired. The injector was designed to avoid saturation up to 1,000 A in the railway line. The parameters used in the injector core are shown in Table 4. A 25 Î¼F capacitor is used to achieve resonance between the source and the core to increase the voltage injected into the line. Table 5 shows the injected voltage and current into the segments having different impedances. C. Nonintrusive Impedance Measurement Process C.1 Measurement Circuit The measurement process, pictured in Figure 27, starts with the current and voltage signals being passed through a signal processing chain. This hardware is used to condition the signal, so the microprocessor is relieved of most of the computational burden. The voltage and the current signals are first amplified using an instrumentation amplifier similar to that shown in Figure 28. Because it was unclear as to how good a signal would result, a potentiometer was used in both signals to adjust the gain to obtain nice clean waves. Experi- mentally, it was found there is no need to boost the voltage signal, so it was set to a unity gain by Turns Ratio Inner Diameter Outer Diameter Height g 3/1 44 mm 55 mm 10 mm 0.7 mm Table 4. Injector parameters. Frequency (kHz) (mV) (mV) Injected Current Primary (A) Segment Injected Current to Power Line (mA) 100 2,000 472 1.08 A 3 B 1.7 C 0.720 D 0.5 200 4,160 940 1.08 A 2.44 B 2.72 C 1.84 D 0.75 Table 5. Experimental test results of the voltage injector.
Developing the Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems 41 leaving the potentiometer circuit open. Then, the signal goes through a high pass filter to remove any low frequency signal and process only the high frequency signal. For filtering the SallenâKey topology was used, as shown in Figure 28. From the SallenâKey circuit, one can find the following by applying Equation 12: ( ) = + + (12) 3 4 1 2 3 1 2 3 4 V V Z Z Z Z Z Z Z Z Z out in For the high pass filter case, these values were selected to give a cut-off frequency of 20 kHz: = = = â¦ = â¦10 nF 10 nF 5.6 K 10 K1 2 3 4Z Z Z Z Then, the signal is amplified again through a programmable gain amplifier (PGA) (LTC6912CGN). This PGA, controlled by the microprocessor, is responsible for the process of auto-ranging the gain of the current signal or the voltage signal. After the PGA, the signal goes to an analog multiplier (MPY634KU) to be multiplied with a reference signal as explained in the measurement process. The output of the multiplication will have a high frequency component and a DC component that contain the required information. The signal will then go through a low pass filter using the SallenâKey topology of Figure 28. For the case of the low pass filter, these values were selected to obtain a cut-off frequency of 550 Hz: = â¦ = â¦ = =56 K 56 K 47 nF 22 nF1 2 3 4Z Z Z Z High Pass Filter PGA Low Pass Filter Sampled intoMicrocontroller 50 MHz Crystal Digitally Controlled Osc. Âµ Controller Analog Multiplier Âµ Controller VT or CT Pre- amp Figure 27. Signal processing chain. Figure 28. SallenâKey filter.
42 Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems The resulting output is a DC level signal that contains all the needed information. This signal then goes directly to the microprocessor. C.2 Program Structure For development purposes, the unit has been made to continuously update an LCD display with impedance magnitude and phase data for the sample it senses in its signal transformers. The main program process is shown in Figure 29. The microcontroller starts its infinite loop by initializing the hardware and taking an initial voltage and current sample. If the starting PGA gains and injection amplitude create saturation in either of the signal processing chains, the gains and amplitude are shifted in an auto-ranging procedure to bring both measurements out of saturation. C.3 Auto-Ranging The auto-ranging process is time consuming and severely lowers the deviceâs sampling rate. Thus, the sensor chooses to auto-range only when the measurement has changed sufficiently in magnitude from the original sample. If the impedance magnitude increases too much, the amplitude is too weak to get the measurement out of the noise floor, so either the injection amplitude or the current amplifier gain is increased, or both. Conversely, if the device senses saturation in the signal processing chain, the gains and/or injection amplitude are reduced until the signals are brought back within range. C.4 Signal Sampling and Curve Fitting The demodulation process eases the burden of an impractically high sampling rate from the ADC; however, demodulation has a disadvantage in that the true magnitude and phase informa- tion of the measured signal is obscured. Simple trigonometric principles may be used to extract this information out of the demodulated DC signal. Demodulation (and modulation) takes advantage of the trigonometric identity in Equation 13: [ ]( ) ( ) ( ) ( )= â â +sin sin 1 2 cos cos (13)A B A B A B The relevant signals in this application may be defined as m(t) = Am sin (wt + j), which gives the measured signal, and r(t) = Arsin (wt + q), which gives the reference oscillation. It is desired Start Device Initialize Hardware Auto-Range Signal Amplifiers Take New Measurement Relate Measurements to Calibration Tables Display Impedance Vector Still In Range? Yes No Figure 29. Main program loop.
Developing the Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems 43 to retrieve the amplitude and phase information from the measured signal, while it is assumed that the frequency of operation is already known. Multiplying these two signals results in the demodulated signal d(t), given by Equations 14 and 15: [ ]( )( ) ( )( )( ) ( ) ( ) ( ) ( )= = w + j â w + q â w + j + w + q1 2 cos cos (14)d t m t r t A A t t t tm r [ ]( ) ( )( ) = j â q â w + j + q1 2 cos cos 2 (15)d t A A tm r In this application the signal is then low pass filtered, so only the information at DC, dDC(t), is preserved. The DC signal is given by Equation 16: ( )= j â q +1 2 cos (16)d A A CDC m r There is a bias C added to the DC component such that the value does not become negative, because the hardware in this application cannot support negative signals at this stage. It can be assumed that Ar is known, because the reference signal is constant and generated on board. Three quantities must be determined in dDC â Am, j, and C. This can be accomplished by changing the phase of the reference signal q and then sampling dDC. Because there are three quantities to be determined, it takes three samples to characterize dDC. Specific phases q are chosen to ease the mathematics. Phases of q = 0, p 2 , and p are chosen to sample. For these phases, the corresponding dDC expressions can be seen as Equations 17 through 21: ( )= j +q= 1 2 cos (17)0d A A CDC m r = j + pï£« ï£ï£¬ ï£¶ ï£¸ï£· +q= p 1 2 cos 2 (18) 2 d A A CDC m r ( )= â j1 2 sin (19)A Am r ( )= j + p +q=p 1 2 cos (20)d A A CDC m r ( )= â j +1 2 cos (21)A A Cm r In Equations 22 through 24, these can be manipulated to show the following: = +q= q=p , (22)0C d dDC DC ( )( )= â + âq= q= p12 , (23)0 2 2 2 A A d C d Cm r DC DC
44 Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems and ( ) â â = j q= p q=p tan , (24)2 d C d C DC DC which can be solved for the quantities of interest. C.5 Circuit Calibration and Powering To calibrate the board, two relays (PCN-105D3MHZ) were used. The main function of these relays is to connect the signal coming from the pre amp directly to the measurement signal. This ensures there is a known signal amplitude and frequency signal connected directly to the measurement channel. Changing the magnitude using the DAC, we can obtain a different magnitude and calibrate the entire measurement range. The board is powered by a 12V DC main supply. A 3.3V DC-DC converter (JCD0512D03), a 5V DC-DC (JCD0512D05) converter, and 15V DC-DC (JCK4012D15) converters were used to power the different components. Upon selecting all the components and drawing the circuit diagrams, EAGLE software was used to design the PCB for the sensor, as shown in Figure 30. The PCB was made, and the components were populated on the board, as shown in Figure 31. To control the injector, a PIC24 was programmed using the Explorer 16 Development Board. To write the code and program the chip, MPLAB was used. The proposed method was designed to be nonintrusive and to work with live high voltage overhead lines. Also, the Figure 30. Injector PCB diagram.
Developing the Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems 45 output from this sensor can serve as a direct tripping signal for the circuit breaker or simply as a warning signal for the maintenance team. The complete system is shown in Figure 32. To operate the system wirelessly, the system was powered by a 12V rechargeable battery (TURNIGY 2.2), while two wireless ZigBee communication modules were used to control and receive data from the board, as shown in Figure 33. One ZigBee module is connected to the Explorer 16 Development Board in the injector side and configured to be a coordinator node that can send and receive data from all the modules using the same address. The other module is connected to the laptop using the USB port and configured to be a router node to send data only to the coordinator. Both of these modules are configured and controlled using XCTU software. Figure 31. Completed injector PCB board. Figure 32. Complete system with the PCB connected to the micro controller and the clamp-on voltage and current transformer.
46 Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems C.6 Preliminary Field Tests at GCRTA The research team had several meetings with GCRTA. The team underwent GCRTAâs Rail Operations Rule Book training certification program. We configured the equipment to conduct field tests at GCRTAâs Brookpark facility. The facility has a feature that allows it to be connected to and disconnected from the rest of GCRTAâs DC railway system. We conducted tests to measure the high frequency impedance of the DC line segment at the facility. The facility has about a 200-ft DC line with two of the transit car segments connected to it. The impedance of the line segment, while two transit cars were connected, was measured at several frequencies. During the tests, the facilityâs overhead line was disconnected from the rest of the DC railway system. The measurements were done at the middle of the segment, which mimics the typical injector and blocker configuration for the active clamp sensor. Figure 34 shows the photos taken during field testing. The impedance of the DC railway segment taken at GCRTA Brookpark facility is shown in Figure 35. As seen from this figure, the impedance at a frequency of around 100 kHz is around 15 kâ¦, which is within the measurement range of the active clamp sensor that the research team developed. C.7 Final Field Tests at GCRTA The sensor was developed and tested in the lab before field testing it at the GCRTA facility. A second test was scheduled at GCRTA to test the active clamp sensor. The test was conducted at a GCRTA facility quite similar to their Brookpark facility. This facility was also almost 200 ft in length, so the results were expected to be within the same range as those of the first test. During the tests, the facilityâs overhead line was disconnected from the rest of the DC railway system, so GCRTA provided a diesel vehicle to be able to clamp the sensor to the overhead line. Figure 36 presents the active clamp sensor field testing at GCRTA. Because the team could not place the sensor exactly in the middle, there will be an increase in impedance; however, this increase is expected to be within the range of the device The test was done using the 100 kHz signal, and the sensor transmitted the data to the laptop using wireless technology. Figure 37 shows the test topology where the active clamp sensor is connected to the overhead line. The sensor injects the high frequency signal, measures the high frequency impedance, and sends the data to the laptop wirelessly. The test was done to represent five cases. Figure 33. The ZigBee wireless module and the battery.
Developing the Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems 47 Substation power supplies and the control panels at the GCRTA Brookpark facility. Impedance measurements of the line from the midsection with direct connection to the ground line. Impedance measurements of the line from the midsection with the power analyzer. Impedance measurements of the line from the starting end with the power analyzer. Initial setup of the equipment and instrumentation. Connection of the power analyzer leads to the power line. Figure 34. Field testing at GCRTA Brookpark facility. Figure 35. Impedance of DC railway segment taken at GCRTA Brookpark facility.
48 Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems Fault implementation in the mid-section of the line. Data logging of the sensor measurements through wireless communication networks. Initial measurements and fine-tuning of the active clamp sensor. Active clamp sensor core installations on the power line, via the diesel service car as a platform. Figure 36. Active clamp sensor field testing at GCRTA. sensor Diesel Vehicle HF injector Laptop Figure 37. Test topology at GCRTA.
Developing the Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems 49 Case A (healthy system) In Case A, the system is healthy, and there are no faults or leakage currents in the system, as shown in Figure 36. The impedance measured by the sensor was 20 kâ¦, which is consistent with the previous test using the impedance analyzer. Case B (fault far from the sensor) Figure 38 shows Case B. In Case B, the fault was generated at a point far from the sensor. The impedance measured with the active clamp sensor was 7.01 kâ¦. Case C (fault in the middle of the line) In Case C, a fault was generated in the middle of the section, as shown in Figure 39. The impedance measured with the active clamp sensor was 5.65 kâ¦. Case D (fault close to the sensor) In Case D, a fault was generated at a point close to the sensor, as shown in Figure 40. The impedance measured with the active clamp sensor was 4.3 kâ¦. sensor HF injector Laptop Diesel Vehicle Figure 38. Case B (fault on the far side). sensor HF injector Laptop Diesel Vehicle Figure 39. Case C (fault in the middle). sensor HF injector Laptop Diesel Vehicle Figure 40. Case D (fault at the close side).
50 Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems Case E (faults on both sides) In Case E, the overhead line was shorted on both sides of the sensor, as shown in Figure 41. The impedance measured with the active clamp sensor was 2.8 kâ¦. C.8 Summary of the GCRTA Tests The research team implemented the active clamp sensor on the DC power line. After tuning of the sensor, the research team obtained baseline impedance measurements. The research team then applied the faults at several places in the line. The faults were made using a direct connection through rusted ground connections. Then, the research team measured the overall impedance changes with the active clamp sensor. The impedance of the line varied from 2.8 kâ¦ to 20 kâ¦ depending on where the faults were generated. These results indicated that the active clamp sensor can detect faults, by measuring the impedance of the line that would relate the health conditions and the leakage current levels. D. Conclusions Throughout the project, the active clamp sensor was prototyped and configured for field testing. The prototype developed was tested at the GCRTA facilities. The active clamp sensor worked as expected. The sensor detected the various levels of fault conditions that relate to health conditions and leakage current levels. As a next step, it is important to develop impedance signatures of the lines that relate to various fault conditions and leakage current levels. There is also an opportunity to configure the sensor for the third rail system, which is more common and has known failures due to low-level leakage currents. The setup available for testing at GCRTA was for the pantograph system. The research teamâs sensor prototype was designed accordingly. The research team plans to modify the sensor for third rail systems. In the next phase of the project, the team will attempt to get testing conditions from GCRTA that will allow it to test the sensors for third rail systems. E. References and Other Resources 1. IEEE Standard 80, Guide for Safety in AC Substation Grounding, IEEE, Piscataway, NJ. (IEEE Standard 80 defines methods for calculating safe touch and step potential and is widely used in the transportation industry for traction power substation grounding.) 2. National Electrical Code, National Fire Protection Association, Quincy, MA. 3. National Electrical Safety Code, IEEE, Piscataway, NJ. 4. National Fire Protection Association (NFPA) 130, Standard for Fixed Guideway Transit and Passenger Rail Systems, Quincy, MA. 5. American Railway Engineering and Maintenance-of-Way Association (AREMA), Chapter 12, Rail Transit, Lanham, Maryland. sensor HF injector Laptop Diesel Vehicle Figure 41. Case E (Faults on both sides of the line).
Developing the Guidebook for Detecting and Mitigating Low-Level DC Leakage and Fault Currents in Transit Systems 51 6. Roman, M., R.R. Vanzyl, N. Parus, and N. Mahatho, Insulator leakage current monitoring: Challenges for high voltage direct current transmission lines, Industrial and Commercial Use of Energy (ICUE), Vol. 1, No. 7, pp. 19â20, Aug. 2014. 7. Ferreira, T.V., A.D. Germano, E.G. da Costa, J.M.G. Angelini, F.E. Nallim, and P. Mendonca, Naturally aged polymeric insulators: Washing and its consequences, Modern Electric Power Systems (MEPS), 2010 Proceedings of the International Symposium, Vol. 1, No. 5, pp. 20â22, Sept. 2010. 8. Cheng, T.C., F. Zedan, G.R. Elder, S.S. Low, J.N. Rippey, and G.D. Rodriguez, âEPRI - HVDC Insulator Studies: Part I Field Test at the Sylmar HVDC Converter Station,â IEEE Transactions. Power Apparatus and Systems, Vol. 100, No. 2, pp. 902â909, Feb. 1981. 9. Phillips, A.J., F.F. Bologna, J.M. Major, and C.S. Engelbrecht, Development and demonstration of low cost robust leakage current sensors for evaluating contaminated insulators. In International Symposium on High Voltage Engineering, Cape Town, South Africa, 2009. 10. Sinclair, A., D. Finney, D. Martin, and P. Sharma, Distance Protection in Distribution Systems: How It Assists with Integrating Distributed Resources, IEEE Trans. Ind. Appl., Vol. 50, No. 3, pp. 2186â2196, MayâJune 2014. 11. Sevov, L., S. Kennedy, R. Paes, and P. Ostojic, Differential protection for medium voltage pulse transformers, Petroleum and Chemical Industry Technical Conference (PCIC), pp.173â184, Sept. 8â10, 2014. 12. Yu, J.G., and C.J. Goodman, Modelling of rail potential rise and leakage current in DC rail transit systems, IEE Colloquium on Stray Current Effects of DC Railways and Tramways, pp. 221â226, Oct. 1990. 13. Sankey, J.E., and P.L. Anderson, Effects of Stray Current on Performance of Metallic Reinforcements in Reinforced Earth Structures, Transportation Research Record, No. 1675, 2007, pp. 61â66. 14. Yu, J.G., The Effects of Earthing Strategies on Rail Potential and Stray Currents in DC Transit Railways, In Proc. of the International Conference on Developments in Mass Transit Systems, London, UK, April 1998. 15. RodrÃguez, J.V., and J. Sanz Feito, Calculation of remote effects of stray currents on rail voltages in DC rail- ways systems, IET Electrical Systems in Transportation, Vol. 3, No. 2, pp. 31â40, June 2013. 16. Davis, A.J., Stray current control-an arbitrators view, IET Colloquium on Stray Current Effects of DC Railways and Tramways, 1990. 17. Natarajan, R., Analysis of grounding systems for electric traction, IEEE Transactions on Power Delivery, pp. 389â393, April 2001. 18. Charalambous, C.A., I. Cotton, P. Aylott, and N.D. Kokkinos, A Holistic Stray Current Assessment of Bored Tunnel Sections of DC Transit Systems, IEEE Transactions on Power Delivery, Vol. 28, No. 2, pp. 1048â1056, April 2013. 19. Kinh, D., S. Pham, R. Thomas, and W.E. Stinger, Analysis of stray current, track-to-earth potentials and substation negative grounding in DC traction electrification system, Railroad Conference, 2001. 20. Charalambous, C., I. Cotton, and P. Aylott, A simulation tool to predict the impact of soil topologies on coupling between a light rail system and buried third-party infrastructure, IEEE Transactions on Vehicular Technology, Vol. 57, pp. 1404â1416, May 2008. 21. Pasdar, A., Y. Sozer, and A. De-Abreu Garcia, Smart sensor network for power grid health monitoring, University of Akron, OH, United States Patent No. #10,261,119 B2, April 16, 2019. 22. Anderson, L., J.W. Beal, D.S. Ellis, and J. Lauletta, System and method for locating and analyzing arcing phenomena, Exacter, Inc., Columbus, OH, United States Patent No. #7912660, March 22, 2011.
Abbreviations and acronyms used without definitions in TRB publications: A4A Airlines for America AAAE American Association of Airport Executives AASHO American Association of State Highway Officials AASHTO American Association of State Highway and Transportation Officials ACIâNA Airports Council InternationalâNorth America ACRP Airport Cooperative Research Program ADA Americans with Disabilities Act APTA American Public Transportation Association ASCE American Society of Civil Engineers ASME American Society of Mechanical Engineers ASTM American Society for Testing and Materials ATA American Trucking Associations CTAA Community Transportation Association of America CTBSSP Commercial Truck and Bus Safety Synthesis Program DHS Department of Homeland Security DOE Department of Energy EPA Environmental Protection Agency FAA Federal Aviation Administration FAST Fixing Americaâs Surface Transportation Act (2015) FHWA Federal Highway Administration FMCSA Federal Motor Carrier Safety Administration FRA Federal Railroad Administration FTA Federal Transit Administration HMCRP Hazardous Materials Cooperative Research Program IEEE Institute of Electrical and Electronics Engineers ISTEA Intermodal Surface Transportation Efficiency Act of 1991 ITE Institute of Transportation Engineers MAP-21 Moving Ahead for Progress in the 21st Century Act (2012) NASA National Aeronautics and Space Administration NASAO National Association of State Aviation Officials NCFRP National Cooperative Freight Research Program NCHRP National Cooperative Highway Research Program NHTSA National Highway Traffic Safety Administration NTSB National Transportation Safety Board PHMSA Pipeline and Hazardous Materials Safety Administration RITA Research and Innovative Technology Administration SAE Society of Automotive Engineers SAFETEA-LU Safe, Accountable, Flexible, Efficient Transportation Equity Act: A Legacy for Users (2005) TCRP Transit Cooperative Research Program TDC Transit Development Corporation TEA-21 Transportation Equity Act for the 21st Century (1998) TRB Transportation Research Board TSA Transportation Security Administration U.S. DOT United States Department of Transportation
TRA N SPO RTATIO N RESEA RCH BO A RD 500 Fifth Street, N W W ashington, D C 20001 A D D RESS SERV ICE REQ U ESTED ISBN 978-0-309-48164-9 9 7 8 0 3 0 9 4 8 1 6 4 9 9 0 0 0 0