Electric Arc and associated Hazards in the Rail Transit Industry Are we up to date with current developments?

Transcription

1 Electric Arc and associated Hazards in the Rail Transit Industry Are we up to date with current developments? Dev Paul, P.E. AECOM Oakland, CA Abstract: Electrical arcing faults are inherent characteristics of a power system configuration; they can be minimized by proper design but can t be eliminated. The industry is developing innovations in electrical system designs, equipment, and protection to limit the arc flash hazard, especially in the case of ac power system design [1]-[6]. To increase personnel safety and to minimize equipment damage due to arcing faults, recently many codes and standards including NEC and OSHA require that an arc flash hazard analysis be performed on a power distribution system. For ac power system arc flash hazard analysis, industry is currently using IEEE Standard 1584 [7] and NFPA 70E [8]; however, there are no industry standards or guidelines for performing arc analysis for dc power systems. This paper lists the current research and development work undertaken by an IEEE/NFPA Committee on arc hazard analysis [15] [29] [30] with very limited research on dc system. In addition, this paper provides a simplified approach to performing dc arc hazard analysis in rail transit systems for dc traction power substations. Key words: Arc hazard, arc energy, energy label grounding system, electric shock, personnel protective clothing I. INTRODUCTION All rail transit projects utilize either ac or both ac and dc power system equipment, and therefore arc hazard analysis mandated by codes must be performed to minimize equipment damage and hazards to personnel. This paper provides a comprehensive list of technical references to understand ac power system equipment design and protection improvements from arc hazard analysis and associated incident energy. Although industry realizes that a dc arc is more dangerous than an ac arc, very little research has been done so far for dc arc hazard analysis. The dc arc in a traction power system has been the subject of research by AIEE Substations Committee in 1957 [9], which recommended use of a high-resistance grounding (HRG) protection scheme for the dc enclosure frame fault. This committee cited many other industrial dc projects and concluded that dc arc is very dangerous. This scheme required keeping dc enclosures completely isolated/insulated from ground with the exception of single-point grounding to earth by HRG protection schemes. Later, transit industry professionals had divided opinions on the use of HRG schemes, considering that the higher touch potential with respect to ground of the dc enclosure may endanger personnel due to electric shock. This led to the use of a low-resistance grounding (LRG) protection schemes at some transit systems. The subject of choosing either a HRG or a LRG protection scheme for dc equipment has been debated by transit professionals in many APTA technical conferences [10] [11] [20]. The definition of HRG and LRG for ac power system grounding [26] should not be confused with the dc equipment grounding protection described here using the same terminology. The technical requirements for dc equipment grounding (HRG or LRG) are entirely different from the requirements for ac power system grounding. All modern dc transit systems are normally operated ungrounded, intentionally kept isolated (floating) from ground to minimize dc stray current [23]. For safety reasons, in dc transit systems, rail-to-ground monitoring devices are used to ground the rail for short duration when rail voltage with respect to ground becomes excessive during fault conditions or train bunching [25]. It is known that a LRG protection scheme for dc equipment enclosure grounding will cause higher dc short-circuit current and incident energy during arcing fault, and thus the engineer should review resulting equipment damage and safety of personnel. Currently both grounding schemes are being used by the rail transit industry for grounding traction power substation equipment enclosures. The choice depends upon the 1

2 opinion and judgment of the transit authority or the project s design consultant. Currently, IEEE has a task of developing a new standard on DC Traction Power System Grounding, and therefore it is important for the committee to resolve this issue. It is the author s opinion and recommendation to use HRG from an arc hazard and safety point of view. The concern of higher touch potential and associated electric shock concern may be compromised by proper design of the substation dc equipment enclosure, insulating barriers between the enclosure and other adjacent metallic cubicles, insulation under and around the enclosure, and mandated use of personnel protective equipment (PPE) and arc flash training. Enforcement of the arc flash hazard training and PPE use has been mandated by OSHA and other regulatory agencies. Additional research and development work is needed by the suppliers of dc equipment to establish data needed to perform dc arc hazard analysis for rail transit systems. This paper includes a simplified approach to perform dc arc hazard incident energy analysis inside a dc switchgear enclosure configured inside a prefabricated metallic traction power substation housing. II. RAIL TRANSIT POWER SYSTEMS Rail transit systems around the world fall into four major categories as follows: 1. High Speed Rail 2. Heavy Rail System 3. Light Rail System 4. People Movers High speed rails, in general, use a single phase ac electrification system. Utility company power at 115 kv or 230 kv is used to power ac traction power substations located along the tracks that provide 25 kv, single-phase, 60 Hz for the overhead contact wire. Other single phase ac traction power systems using different voltages with different frequencies (such as 12 kv at 25 Hz, 12.5 kv at 60 Hz, and 15 kv at 16-2/3 Hz) are also in use at some high speed rail operating systems. Heavy rail systems may employ ac bulk power substations to interface with the utility at high voltage (HV) to step-down to medium voltage (MV). The MV power distribution system is installed along the tracks to supply power to the dc traction power substations (TPSS). Instead of bulk power substations and using MV distribution along the tracks, some systems use direct MV supply from a utility at individual traction power substation sites. Generally, a dc positive, third rail system then provides dc power to the vehicle propulsion system. All traction power substations house both the MV and the DC switchgears, sometimes in separate houses and sometime in the same house. All these metallic houses are built to meet minimum working clearances. Analysis of arc hazard should be performed for an arcing fault within these switchgear cubicles. Light rail systems use utility ac power supply and dc power and an overhead contact system (OCS) to deliver dc power to the vehicles by pantographs mounted on the vehicles. Likewise, people movers are similar to light rail systems but may use smaller substations and guided rails for the positive and negative dc or ac power distribution system. From the above cited rail transit systems, it is clear that power distribution systems in rail transit systems use both ac HV systems, ac MV systems, ac LV systems, and dc power distribution systems in both the outdoor and indoor configurations. Equipment used for such electrification systems are built to current ANSI, NEMA, IEEE, UL, CSA and IEC standards. Recently, arc hazard and safety concerns have opened doors for innovative protection and equipment design features. Industry has performed much research in ac power systems to make a safer design from the unpredictable threat of arcing faults and associated equipment damage and personnel hazards. This paper provides reference to the technical papers on ac arc hazard analysis that promote innovations in design features that can be applied to transit system ac power systems. In addition, this paper provides the current status of IEEE research on dc arc hazard analysis [15]. However, research, new equipment development and testing are needed for a new IEEE/APTA standard on dc arc hazard analysis. III. AC ARC HAZARD ANALYSIS Technical papers [1] [2] [12] [15] [16] [24] provide algorithms for arc flash calculations, sometimes referring to equations in IEEE 1584 [5] and NFPA 70E [7]. The physics of arcing faults in electrical equipment is included in [16]. A two-part paper covers system design to reduce arc flash incident energy in a multi-voltage-level ac distribution system [1] [2]. Based upon the current standards listed below and with the help of other technical papers, a design engineer 2

3 responsible for rail transit design can perform required engineering analysis for arc flash hazard for the ac portion of a transit system. The result of such analysis will provide a power system protection scheme that can limit the arc hazard to equipment, and improve safety to personnel by providing proper energy labels (hazard - warning signs) at equipment enclosures and use of PPE. Current Standards for ac Power System Analysis Currently there are only two standards, listed below, that provide guidance for ac arc flash hazard calculations. Other safety standards (third item below), provide guidance on specifying PPE tools and other equipment needed for safety of personnel exposed to live equipment. 1) NFPA 70E, revised in 2009 [8] 2) IEEE 1584, revised in 2002, IEEE 1584a, Amendment 1, 2004, IEEE P1584b/D2 Draft 2, unapproved, July 2009.[7] 3) Other safety standards include NESC C2, and OSHA; for additional list of safety standards see references in technical papers [1] [4] [17] [18] [19] [21] [22][28] Arc flash hazard analysis is based upon the following. Flash protection boundary distance Flash protection boundary: An approach limit at a distance from exposed live parts within which a person could receive a second-degree burn if an electrical arc flash were to occur. Flame-resistant PPE must be worn by personnel within the flash protection boundary. Incident energy Incident energy (IE) is the amount of energy impressed on a surface, a certain distance from the source, generated during an electrical arc event. This energy is generally expressed in calories/cm 2. The surface of concern impressed on is the worker s body, particularly the head and trunk. Incident energy is calculated using variables such as available fault current, system voltage, expected arcing fault duration, and the worker's distance from the arc. Data obtained from the calculations are used to select the appropriate flame resistant (FR) PPE, just as voltage level is used to select a class of rubber gloves. PPE risk category This is the minimum level of PPE required, based on calories per centimeter squared, as evaluated in IEEE Standard 1584 [7], to protect the worker from the thermal effects of the arc flash at 18 inches from the source of the arc. Table 1 below provides a relationship between incident energy, risk category and required PPE based upon the current industry practice of ac power system analysis. Min Incident Energy, cal/cm 2 Max Incident Energy, cal/cm 2 Risk Category Required Min Rating of PPE, cal/cm and above Not Available N/A Table 1. Ac power system, incident energy, risk category, and PPE This table may be used for dc arc and protection until new research on dc systems provides further guidance. IV. DC ARC HAZARD ANALYSIS Arcing fault (arc flash) in a dc power system is defined as the failure of the insulation (dielectric strength) at some specific location between the positive and negative polarities of the distribution system. Arcing fault is further characterized by ionization of the air involving conductive particles that generally establish a current path between positive and negative with unstable variable arc resistance. Such arcing fault can become bolted fault if the arc resistance becomes zero, or the arc may cease itself if the resistance becomes very high when dielectric strength becomes normal. Electrical faults are inherent characteristics of a power distribution system configuration; by proper design, they can be minimized but can t be eliminated. Research indicates that dc arc voltage is a function of the combination of contact materials and the gap between the contacts [15]. When the arc is short, the voltage is mainly determined by the anode and cathode drop (around 20V; see Fig. 1 a below). When the arcs are long, and the current not too low, the arc voltage tends to be in the order of 10V/cm (see Fig.1 b). In case of a dc traction power system, if the arc occurs inside the dc switchgear enclosure, it will be from the positive polarity (copper, 3

4 anode) to cubicle enclosure (steel, cathode), to match the picture depicted in Fig. 1 b below. This dc arc will have the highest incident energy as compared to an arc in the open air on trackside of the dc traction power system. It is known that for the same magnitude of voltage, ac or dc, a dc arc will have higher incident energy. and the arc will dissipate a relatively constant amount of power. The load line may intercept the characteristic curve in two locations (a and b) as shown in Figure 3, but only point-b is stable [15]. The stable operating location is the point with the lowest arc voltage. Figure 3 10 mm gap voltage - current relationship and sample dc load line [15] Figure 1 Model of a dc arc. (a) Pictorial. (b) Voltage distribution [15] Figure 2 represents volt-ampere characteristics for arcs in air with copper electrodes having 1 to 200 mm gap [15]. The arc voltage versus arc current characteristic curve will shift upward as the gap increases in length. Basic equations for DC arc hazard analysis. The most dangerous area of a dc arc having maximum energy is when the dc arc occurs inside the dc switchgear or near the output terminals of the rectifier unit; see Fig. 5 for fault location D. It needs to be stated that as the arcing fault is controlled by the arc resistance, it is not necessary that the maximum arc hazard locations match the expected maximum dc short circuit current locations depicted in Fig. 5. The arc fault posing maximum hazard is certainly within the dc enclosure rather than outside the enclosure. Section VI of this paper provides imperial formulas for determining incident energy for arcing faults inside the enclosure and outside the enclosure. Figure 2 Current-voltage characteristics for DC arcs in air, with copper electrodes. [15] The arc operates at the intersection of this curve and the resistance load line of the circuit. Therefore, the arcing current will stabilize itself at a fixed point on the curve, Fig. 4 Typical trackside substation for heavy rail traction power substation [27] 4

5 I MAXE = (I SS /2) (1-T S /T C ) (1) Where T S = L S / R S and T C = L C / R C I SS L S R S L C R C prospective short-circuit current source (rectifier) inductance source resistance circuit inductance circuit resistance Fig.5 Bolted fault locations [27] Figure 5 provides possible locations of dc faults, bolted or arcing faults. Figure 6 shows bolted fault current characteristics for possible faults at different locations. Arcing fault current will be different than the bolted fault current and may not have high peak value, and fault current will be low due to arc resistance. To minimize tripping time and fault current hazard by bolted faults or arc faults, generally high-speed dc circuit breakers with current limiting effects are employed. The dc traction power substation arcing fault at location D can be modeled as a dc source in series with its internal system impedance (Z SYS ) and then simply shorted by arc impedance; see Fig. 7. Based upon the ratings of a traction power substation, system impedance can be calculated by equation (2), assuming voltage regulation is linear for an arc fault condition. Z SYS = V REG kv 2 /MW Ω (2) Where: V REG is the per unit voltage regulation of a traction power substation between the no-load to full-load conditions. This value is typically between 0.06 and per unit. kv is the traction power substation rated dc nominal voltage in kilovolts, generally between 0.6 kv and 1.5 kv, with 0.75 kv commonly used in the industry. MW is the rating of T-R units in parallel inside the substation; typical ratings range from 1.0MW to 3MW. Fig.6 Bolted fault currents, fault A or D with maximum fault current and energy [27] Bolted fault currents on various locations and their respective rate-of-rise times to reach maximum fault currents depicted in Fig. 6 depend upon the L/R time constant of the equivalent electrical circuit under fault condition. Maximum fault energy current (I MAXE ) for a bolted fault condition then can be calculated by using equation (1) below [27]. Figure 7 DC electrical equipment circuit for arc hazard analysis 5

6 To understand the problem of a dc arc hazard, let us consider steady state current in the arc, so that we can use a fixed value of the resistance of the system and one particular value of arc resistance (R arc ) in the calculations. Any inductance in the system would tend to reduce the available power in the arc, so using resistance only instead of impedances would be a conservative assumption. From this steady state assumption, and assuming arc resistance is such that load current becomes practically zero, then arc current can be determined as follows: I arc = V sys / (R sys + R arc ) amperes (3) Before we go into the arc energy and associated hazard, we need to understand how we can minimize the magnitude of the arc current which will make arc energy minimum, along with minimizing damaging effects to equipment and hazard to personnel. A high-resistance grounding protection scheme of the equipment enclosure uses ,000 Ω resistance in series with the protection relay, whereas a lowresistance grounding protection scheme may use 1 Ω or less resistance. These indicated values of resistance may be different for each supplier as there is no fixed value of such resistance in the industry. Thus equation (3) will be modified using the grounding resistance of the protection scheme (R ps ) as follows: I arc = V sys / (R sys +R arc + R ps ) amperes (4) From equation (4) it is clear that arcing current will be low in case of a HRG as compared to a LRG scheme. It should be stated that all touch potential of the enclosure (E TOUCH ), voltage rise of the enclosure to remote earth will be high in case of a HRG scheme compared to a LRG scheme. This voltage may be indicated in equation (5) below; assuming substation ground grid resistance to remote earth is small compared to R ps. V TOUCH = I arc x R ps volts (5) From equation (5) it can be concluded that the frame fault arc current can be minimized by the enclosure frame fault protection scheme resistance. This will result in less arc-flash energy and hazard, especially if the arc current is detected by a frame fault protection relay to initiate tripping of the associated breakers and thus isolate arcing frame fault or bolted fault. High enclosure potential with respect to ground can be made tolerable by use of insulating floor and insulating gloves, insulating boots and PPE to provide safety to persons in case fault interruption is delayed for various reasons. As we have no control over the arc resistance and thus from arc hazard point of view, a HRG scheme may be better than a LRG scheme, whereas the persons may be safe due to reduced arc-flash hazard energy. V. DC ARC ENERGY CALCULATION Heat developed (damage) at the arc fault location can be described as follows: Damage = I arc 2 R arc dt watt-seconds or Joules (6) Where the arc fault current (I arc ), arc fault resistance (R arc ), and fault clearing time, are in amps, ohms, and seconds respectively. To convert to the calories used in some arc flash energy documents requires multiplying by the unit conversion of cal/j. To avoid performing integration which has been used to represent higher current at the instant of fault (t+) and then decreasing magnitude over time t, we can rewrite equation (6) as equation (7) as follows: Damage = I arc 2 R arc t watt-seconds or Joules (7) Where I arc is now steady current in amperes instead of a time varying current. Arc voltage drop (V arc ) will be I arc R arc. This is one concept of understanding arc voltage, that the arc voltage has a linear relationship between the arc current and arc resistance. Much research is underway to understand the behavior of the arc, arc voltage drop, and arc resistance. The purpose of this paper is to understand the heat generated by the arcing fault at the fault location and not to establish that this is the only way to represent the arcing fault and arc voltage drop. Current research provides the relationship between dc arc current and arc voltage as shown in Fig. 2 [15]. The heat generated at the arc may be represented by the modification of equation (7) to (8) using arc voltage drop as follows: Damage = V arc I arc t watt-seconds or Joules (8) Therefore, for a theoretical bolted fault condition with a zero fault resistance, equations (5), (6) and (7) represent practically zero damage at fault location (although there may be a trace of discoloring on the enclosure at the fault 6

7 location). All of the energy of the fault will be dissipated as heat inside the equipment grounding resistor. The difficult part is to find the arcing time of the fault. It is logical to conclude that this arcing time is the arcing fault clearing time by the dc breakers, assuming arcing fault is efficiently monitored/ sensed and cleared in short duration. This paper will not go into the protection scheme of a dc power system, as it has been well understood by the dc equipment manufacturers; however, there is room for improvement by research from arcing fault and energy hazard points of view. Arcing power can be calculated by the following equation: P arc = V arc 2 /R arc watts (9) Arc energy will be arc power multiplied by arcing time. Using differential equations, we can find that the maximum power released in the arc is at the point where the resistance in the arc is equal to resistance in the system. This will be the case when the arc voltage is ½ the system voltage. Thus equation (9) for maximum arc power may be represented by equation (10). This also agrees with the conclusions of equation (1) for a bolted fault condition where all the fault energy will be across the small resistance of the circuit, which includes source and circuit resistance. P MAX = (V sys /2) 2 /R arc watts (10) To convert this to maximum energy (E MAX ), we need to add the time variable to maximum power (P MAX ). This variable time is the arc duration time (T ARC ) which is the arc fault clearing time in seconds by the dc switching device (main breaker or the feeder breaker depending upon the location of the fault, on the main dc bus or on load side of the feeder breaker). It should be noted that if this arcing fault is on the load side of the feeder breaker, then tripping of the feeder breaker only will not clear the fault unless a companion feeder breaker inside the adjacent substation is also opened simultaneously. Therefore, a transfer trip scheme appears to be necessary in the dc traction electrification system. E MAX = P MAX T ARC watt-seconds or joules (11) E MAX = 0.239[(V sys /2) 2 /R sys ] T ARC calories (12) Where: E MAX is total energy released at the maximum power point in calories, and R sys is the traction power system equivalent resistance in ohms. This will tend to become smaller due to adjacent substations in parallel, even if we neglect the effect of positive and negative dc power supply feeder system resistance from an adjacent substation to the local substation. To account for this, a constant (k 1 ) with value less than unity should be multiplied with R SYS equation (12). T ARC is the arcing time or the tripping time of both the local and the adjacent traction power substation feeder breaker. To account for perhaps delayed tripping of the feeder breaker inside the adjacent substation, either by the transfer tripping scheme or directly by (di/dt) rate-of-rise protection relay, another constant (k 2 ) with value greater than unity should be multiplied with T ARC in equation (12). Thus the modified equation (12) using two separate time constants described above will be equation (13). E MAX = 0.239[(V sys /2) 2 /k 1 R sys ] k 2 T ARC calories (13) VI. DC ARC FLASH INCIDENT ENERGY CALCULATION There are two cases of study: a) arc in the open area, and b) arc inside the enclosure. The published literature indicates that if the incident energy (IE) is needed at working distance d cm from the arc location, then the arc flash incident energy estimate in calories/cm 2 in open air can be estimated by dividing energy in calories, equation (13), by surface area of the sphere with diameter d cm as follows. Incident energy in open air Published research data [12] [15] indicate to use the following empirical expression: IE ARC = E ARC /4πd 2 calories/cm 2 (14) Where IE ARC is incident energy of the arc in calories/cm 2 at distance d in cm, and E ARC is energy of the arc in calories. Using equations (12) and (14), the incident energy for an arcing fault outside the substation may be calculated using equation (15) below: 7

8 IE MAX POWER = {0.239[(V sys /2) 2 /k1r sys ] k2t ARC }/ (4πd 2 ) calories/cm 2 (15) Incident Energy inside the DC Equipment Enclosure Published research data indicate [15] to use the following empirical expression (15): IE ARC = ke ARC / (a 2 +d 2 ) calories/cm 2 (16) Where k constant and d distances are indicated in table 2. It has been confirmed in previous research that the incident energy is increased toward personnel when arcing is inside the enclosure as compared to the incident energy when it is in the open air. Enclosure Width (cm) Height (cm) Depth (cm) a (cm) k (cm) Panelboard LV Switchgear MV Switchgear Table 2 Optimum values of a and k [15] It should be noted that all dc switchgear cubicles are generally 24 inch wide and thus fall in the category of MV Switchgear listed in Table 2. Therefore, incident energy for the arcing fault inside the traction power substation enclosure should be based upon using equation (17) as follows: IE ARC = 0.416E ARC / (95 2 +d 2 ) calories/cm 2 (17) Combining this with equation (15), the maximum incident energy at distance d cm toward personnel in front of the dc switchgear cubicle with internal arc flash hazard will be as shown in equation (18). IE ARC = 0.416{0.239[(V sys /2) 2 /k 1 R sys ] k 2 T ARC }/ (95 2 +d 2 ) IE ARC = { [(V sys /2) 2 /k 1 R sys ] k 2 T ARC }/ (95 2 +d 2 ) (18) In a dc power system, energy is stored in system inductance (L), and this stored energy is generally calculated by the following equation (19). Energy = 1/2L I 2 joules (19) Where: L is in henries and I is in amperes It is not established and clear to the author, how to factor such additional energy that seems to be added to the energy from the resistive arcing current [24]. For example, an arcing fault on the load side of the dc feeder breaker inside the dc switchgear at the traction power substation will not be cleared by tripping the local feeder breaker, unless the associated feeder breaker inside the adjacent traction power substation is tripped by either the transfer trip protection scheme or di/dt protection relay. It appears that the contribution of the arcing fault current from an adjacent traction power substation will be low but rich in system inductance stored energy due to long distance of the fault from its source. This needs further research to modify the incident energy equation (18) by the contribution of energy indicated in equation (19). For safety reasons, it becomes obvious that the transfer trip protection scheme may be needed in addition to (di/dt) as it will act as backup protection to a di/dt protection relay. VII. CONCLUSIONS 1. The use of a HRG protection scheme for the substation dc equipment enclosure appears to minimize the dc arc current, its incident energy and associated arc hazard and equipment damage. Dangerous touch potential of the enclosure exposed to a maintenance person may be acceptable, so long, as the current magnitude through a human body [25] can be made low by mandating use of insulating floors around the equipment and use of PPE which includes insulating gloves and insulating boots. 2. A transfer trip protection scheme may be desirable in all dc rail transit systems to clear the low level arcing fault current contribution from an adjacent substation, where the maximum incident energy calculation is presented in this paper. Industry has developed a di/dt relay for the dc feeder breakers which can reach its zone of protection for a fault inside the adjacent substation. However, if a transfer trip protection scheme is used, it may become either a primary or a backup protection to the di/dt protection depending upon the sensitivity of these two protection relays with respect to each other. 3. This paper provides empirical formulas to perform incident energy calculations in the rail transit dc power distribution system. These formulas are based upon the limited research data available. 4. Transit agencies, manufacturers and the IEEE technical committee responsible for dc equipment grounding standards for rail transit systems are requested to review the data and approach indicated 8

9 in this paper for analysis of dc arc in a dc traction power substation. 5. Further testing and research is recommended until we have sufficient data to create a new standard for dc power distribution system similar to IEEE Standard 1584 [7] and NFPA 70E [8] used for analysis of ac power system. ACKNOWLEDGEMENT The author would like to thank dc Arc Flash Working Group (EFCOG ESSG Brookhaven National Lab Oct. 4-8, 2010); Collaboration on Arc Flash Research and Testing Group [29]; and IEEE fellow Dr. P. K. Sen, for the research and valuable data [15] they have assembled for the industry on the dc arc hazard for industrial low voltage dc systems up to 250 V. Thanks go to Paul Forquer for bringing the issue of proper selection of a dc equipment frame fault protection scheme to the APTA committee responsible for developing a new IEEE Standard for the rail transit system on DC Equipment Grounding. Thanks to Andy Jones, Kinh D. Pham, Yu, Jian G, Steve Bezner, Benjamin W. Stell and Rey Belardo for timely review of the paper. Finally thanks to Sonia Paul for paper edits. REFERENCES [1] J.C Das, Protection Planning and System Design to Reduce Arc Flash Incident Energy in a Multi-Voltage Level Distribution System to 8 cal/cm 2 (HRC 2) or Less Part I: Methodology, IEEE Trans. Ind. Applications Vol. 47, No 1, pp , Jan./Feb [2] J.C Das, Protection Planning and System Design to Reduce Arc Flash Incident Energy in a Multi-Voltage Level Distribution System to 8 cal/cm 2 (HRC 2) or Less Part II: Analysis, IEEE Trans. Ind. Applications Vol. 47, No 1, pp , Jan./Feb [3] W.A. Brown and R. Shapiro, Incident Energy Reduction Techniques IEEE. Ind. Appl. Mag., vol. 15, no. 3, pp , May/June [4] Daleep C Mohala, Tim Driscoll, Paul S. Hamer, and Sergio A. R. Panetta, Mitigating Electric Shock and Arc Flash Energy A Total System Approach for Personnel and Equipment Protection IEEE/IAS PCIC Conference, pp , San Antonio, Texas, Sept , [5] George Roscoe, Marcelo E. Valdes, and Ray Luna, Methods for Arc-Flash Detection in Electrical Equipment IEEE/IAS PCIC Conference, pp , San Antonio, Texas, Sept , [6] James E. Mitchem, Mark Cross, and Daryld Ray Crow, Evolution of an Electrical Safety Culture IEEE/IAS PCIC Conference, pp , San Antonio, Texas, Sept , [7] IEEE Guide for Performing Arc-flash Hazard Calculations, IEEE Std [8] Standard for Electrical Safety in the Workplace, NFPA 70E, [9] IEEE Recommended Grounding Practices for Single- Polarity D-C Structures. AIEE committee Rep. P , Oct [10] P. Forquer, Equipment Grounding in Traction Power Substations, paper presented at the APTA Annual Conference, Miami, FL. 1993, Paper 67-R93. [11] D.C. Hoffman, Grounding of DC Structures and Enclosures, presented at the AIEE Winter General Meeting, New York, 1961 [12] Daniel R. Doan, Arc Flash Calculations for Exposures to DC Systems, IEEE Trans. Ind. Applications Vol. 46, No 6, pp , Nov./Dec [13] IEEE Guide for Testing Metal-Enclosed Switchgear Rated up to 38 kv for Internal Arcing Arcing Faults, IEEE C [14] R.H.Lee. The Other Electrical Hard: Electric Arc Blast Burns. IEEE Trans. Ind. Applications Vol. IA-18, pp , May/June [15] Ravel F. Ammerman, Tammy Gammaon, P.K. Sen, and John P. Nelson, DC-Arc Models and Incident- Energy Calculations, IEEE Trans. Ind. Applications Vol. 46, No 5, pp , Sept./Oct [16] David Sweeting, Arcing Faults in Electrical Equipment, IEEE Trans. Ind. Appl., vol. 47, pp , Jan./Feb [17] Charles F. Dalziel and W.R. Lee, " Lethal Electric Currents," IEEE Spectrum Vol. 6, No.2, Feb

10 [18] R. H. Lee, Electrical Safety in Industrial Plants, IEEE Spectrum., pp 51-55, June 1971 [19] Charles F. Dalziel, Electric Shock Hazard, IEEE Spectrum, pp , Feb [20] A. Kusko, J. J. LaMarca, and J.D. Glover, Grounding Practice for Electric People Mover Vehicles, IEEE/IAS Conference Record, pp , Sept [21] IEEE Std. 902, Guide for Maintenance, Operation, and Safety of Industrial and Commercial Power Systems [22] Dev Paul, "Operational Safety and Maintenance Considerations for People Movers, D-C Grounding Systems", APTA Transit Conference, New York, June 1998 [23] Dev Paul, "DC Traction Power System Grounding", IEEE Transactions on Industry Applications, Vol. 30, No. 5, May/June 2002, pp [24] Kinh D. Pham and Robert Jones, Arc Flash Hazard Analysis in Traction Power Substations, Proceedings of the 2009 ASME/IEEE Joint Rail Conference, March 3-5, 2009, Pueblo, Colorado. Powered Rail Transit System, APTA Conference 2010, Vancouver, British Columbia, Canada. [26] IEEE Std. 142, Recommended Practice for Grounding of Industrial and Commercial Power Systems (ANSI). [27] DC Switchgear by Whipp and Bourne, Google search on DC Arc Interruption and DC Switchgear. [28] H. Landis Floyd II, "Closing the Gaps in Arc Flash Hazard Mitigation: A Review of US, Canada, and EU Standards," IEEE copy right paper /09/2009. [29] John, B., Vogel, S., Sen., and Ammerman, R., Update on IEEE/NFPA Collaboration on Arc Flash Research and Testing, 2008 IEEE IAS Electrical Safety Workshop, March 18-21, 2008, Dallas Texas. [30] Jim White and Ron Widup, Update on the IEEE/NFPA Joint Collaboration Project on the Arc-Flash Hazard, Report on Comments (This year s NFPA 70E ROC) by Rd. Wei-Jen Lee of the University of Texas at Arlington, NETA World, pp [25] Dev Paul, How Effective Are Automatic Grounding Devices at the Floating Negative Rail System of a DC 10