Compatibility of HVDC Transmission Lines with Signal Systems. Rod Perala, PhD Jennifer Kitaygorsky, PhD

Compatibility of HVDC Transmission Lines with Signal Systems Rod Perala, PhD Jennifer Kitaygorsky, PhD Electro Magnetic Applications Inc. Lakewood CO 303-980-0070; rod@ema3d.com AREMA Railway Exchange Orlando, FL August, 2016 ABSTRACT 4017 words plus 13 figures @ 250 words (3250)=7267 words Innovations in HVDC transmission technology have made it possible to consider the compatibility of sharing railroad corridors with buried HVDC transmission lines. The objective of this paper is to present a summary of the compatibility issues and their solutions. We know of no such existing shared corridors, but in recent years we have performed corridor compatibility studies for two different Class I railroads involving proposed specific designs and installations. This paper is based on these studies. The results support the favorability of such installations. Topics we have considered include the following: HVDC Transmission Line Design considerations Interference from AC currents caused by Converter Stations Corona Noise Interference Corrosion Effects from Stray DC Currents Lightning Effects Fault Current Effects Transient Effects from Energizing/De-energizing the Line Personnel Hazards We show that the important aspects of HVDC transmission line design relevant to corridor compatibility include: HVDC Bipole design Pole conductor shielding Filtering at converter station interfaces Robust cable design 1 HVDC Transmission Line Description We provide here the relevant design features that are most important for compatibility evaluation. First, a typical burial configuration is given in Figure 1.1, which shows two pairs of cables, for a total of four cables in all. The baseline distance from the rail is 15 m (49.2 ft.) and the minimum distance between the pairs is 4 m (13.1 ft.). AREMA 2016 1199

Figure 1.1 Typical cable burial configuration Typical burial depth is ~1.5m A typical possible cable is shown in Figures 1.2. The main features include: A high density polyethylene dielectric jacket A metallic shielding layer ~700 mil cross-linked polymer insulator Longitudinal water barrier A radial water barrier Figure 1.2 A typical HVDC buried cable (http://www.prysmiangroup.com/en/business_markets/markets/hv-andsubmarine/downloads/datasheets/hvdc_a4_low.pdf) 1200 AREMA 2016

The general architecture of an HVDC system is a bi-pole with metallic return, similar to that shown in Figure 1.3. The significant features of this design along with the cable design include: The pole conductors are very nearly balanced in terms of voltages and currents. There are very nearly zero DC leakage currents in the cable shields and in the earth during normal operation. The cable shields completely shield the pole conductor electric fields from the outside world. Figure 1.3 Typical bi-pole design 2 Interference from AC currents (converter noise) on the Pole Conductors Voltage Source Converters (VSC) can produce converter noise at the drive points of the pole conductors. It is possible to completely mitigate this by application of large capacitors to ground at the converter outputs. For example, a 400 mf capacitor filter on the converter output will have an impedance, Z c, defined as: Z c =-1/(j2πfC), where f is the frequency of the converter noise and C is the value of the capacitor, in this case 400 mf. This value of capacitance effectively shorts out any converter noise on the conductors to the shield ground. For example, the capacitor impedance at 100 Hz is about 4 mω, which provides effective filtering. Higher frequency noise is attenuated even more. Other factors also mitigate any noise that might be on the pole conductors: The noise would be shielded by the pole conductor coaxial shields. Any coupling from the pole conductors to a track would be mostly in the common mode, and the grade crossing warning systems (and other signal systems) are designed to work with differential modes. This provides mitigation by common mode rejection (CMR). 3 Corona Noise and Interference We expect no corona noise from the HVDC line design. The basic cable construction creates large electric fields only within the cable shield, eliminating any possibility of corona (which in any event would be contained within the cable shield). AREMA 2016 1201

4 Long Term Corrosion Effects from Stray DC Currents Galvanic corrosion to pipelines and other buried infrastructure can be caused by stray DC earth currents. The source of these earth currents, unbalanced DC currents from the HVDC pole conductors, would flow into the earth via the earth grounding system. There are two possible sources of DC current injection into the earth: Leakage currents through the cable dielectric Leakage currents through the transient arresters located at the converters Leakage currents through the cable polymer insulator can be estimated by using Ohm s law to compute the current between the cable core conductor and the surrounding metallic screen with an impressed voltage of 300 kv. The typical cable of Figure 1.2 will be used for these calculations. The conductivity of the insulation is approximately 10-15 Siemens/meter. We use the cable geometry from the cable to compute the resistance of 100 km of cable to be 7.3x10 9 Ohms. For 300 kv, the total leakage current in a cable 100 km long is about 100 µa. The transient arresters located at the converters are MOV (metal oxide varistor) stacks, and the leakage current is likely less than 1 ma. We note that for each of the leakage currents estimated above, there is also a very nearly equal negative leakage current from the other pole conductor. Because the bi-pole design is very nearly balanced, the net leakage current is much less than the numbers given above. The above estimates by themselves are of no significance, and when the line balance is also included, the net leakage currents are of no consequence for DC corrosion. 5 Lightning Effects Our concerns for lightning damage to the buried cables include: Puncture of the cable outer layers to the central pole conductor might cause an immediate power fault. Puncture of the outer layers might not create an immediate power fault, but moisture ingress into the cable might eventually cause a fault. We first wish to estimate the likelihood of strike attachment to the cables of Figure 2.1, in units of strikes per year per 100 km of length. This is equal to the product of the number of flashes to ground per square km (km 2 ) and the cable exposure area in km 2. The exposure area is the product of the width of the exposure (estimated to be ~18m plus 4m separation between the two trenches = 22m) and the total length of the buried cable (100 km) and is approximately 22m x 100 km ~ 2.2 km 2. Data from the National Lightning Detection Network can be used to estimate the flash density in the corridor. Example density maps are shown in Figure 5.1. We assume the cables are located in the green area, with 3 to 5 flashes to ground/ km 2 per year. Therefore, with these assumptions, the cables are estimated to be exposed to lightning at a rate of about 6 to 10 flashes/year. For a lifetime of 30 years, we expect 180 to 300 strikes close to the cables. Not all of these flashes will attach to the cable. Lightning has a 2% peak current level of about 200 ka, with an average current of about 30 ka. If we assume that only flashes above average will attach to the cable we now have ~90 to 150 flashes that will attach to the cable over a 30 year lifetime. 1202 AREMA 2016

Figure 5.1 Lightning ground flash density for the years 1992 95. (Richard E. Orville and Alan C. Silver, Lightning Ground Flash Density in the Contiguous United States: 1992 95, Monthly Weather Review, Volume 125) Not all of these attaching flashes will cause significant damage to the cable. In order to have an effect, these flashes must do the following: Puncture the outer 4.5 mm jacket. The jacket likely has a DCWV of about 100 kv, and an impulse voltage rating of about 200 kv. The 0.2 mm aluminum moisture barrier must be punctured so that water can ingress though it. Measurements of aluminum puncture by lightning are reported in Figure 5.2 below, but for 6 times the thickness (curve 5). This chart shows that much less than 50 C of charge can create a 30 mm 2 hole. The 2% level for charge is about 200 C, so we see that a strike having only a small fraction of the worst case charge could create a hole in the Al layer. The longitudinal swell tape moisture barrier will significantly inhibit moisture longitudinal movement. The copper screening layer would have to be degraded to damage the underlying XLPE insulation. The XLPE insulation can withstand 300 kv DC over its lifetime, and probably has an impulse standoff voltage of more than 700 kv. AREMA 2016 1203

Figure 5.2 Metal puncture dependency upon charge transfer (F.A. Fisher, J.A. Plumer, and R.A. Perala, Lightning Protection of Aircraft, Lightning Technologies, Inc., 1990) Our evaluation of the lightning risk is summarized as follows: We can expect that 100 km of cables will be directly struck by lightning approximately 90 to 150 times over a 30 year lifetime. The buried cables robust features include o The 4.5 mm jacket o The radial aluminum moisture barrier o The longitudinal swell tape moisture barrier o The copper screening layer o The 18 mm XLPE insulation around the core conductor If the cable outer materials are punctured by lightning, it would take a long but presently unknown time for water ingress to have an effect, if any. We expect that the cables could be physically damaged by lightning over its lifetime. We also expect that the chances of this damage causing a fault current or another adverse effect are small. We recommend that the cable supplier perform some lightning testing of the damage to their cables. 6 Fault Current Effects The transmission line design is robust and fault current events will likely be rare. Nevertheless, such events are possible from the following causes: Accidental damage to cables by heavy construction and maintenance equipment Train derailment or other disaster Vandalism or terrorism Water or other liquid ingress into the cable Dielectric breakdown caused by manufacturing deficiencies (voids, material defects) Damage from animals 1204 AREMA 2016

Historically, fault current damage to railroad wayside systems can extend for several miles. We wish to provide an estimated scope of damage for the HVDC buried cables.. We have identified the following types of possible faults: A short from a pole conductor to its metal return conductor A short from a pole conductor directly to the earth A short from a pole conductor to another pole conductor Only the first fault scenario above is relevant. Inspection of the cable design shows that it is impossible to have a pole to pole short, or a pole to earth short, without involving the cable sheath. We therefore only consider here a short from a core conductor to its concentric metallic screen. Our analysis is based on a cable similar to that shown in Figure 1.2 We assume that the cable is spliced every 900 ft (at each spool length) and the shield grounded at every splice with an earth ground of impedance 25 Ω. We have used our computational electromagnetic (CEM) codes EMA3D and MHARNESS (www.ema3d.com) to compute the fault currents on the sheath, fault currents injected into the earth grounds, and the currents induced in the signal system track wires. In the interest of clarity and brevity, the details of these simulations are not provided here. However, the 3D simulation model is shown in Figure 6.1. When a fault occurs, a current transient (Figure 6.2(a)) within the cable shield with amplitude ~14 ka (=300kV/21.3 Ω) will propagate away from the fault location in both directions. This current flows on the pole conductor and also in the opposite direction on the sheath conductor. The cable metal sheath, which is the parallel combination of the wire screen and aluminum moisture barrier, has a finite resistance per unit length (~2.0 10-4 Ω/m, also called the transfer resistance). We note here that for cable shields in general, there is often also a transfer inductance in series with the transfer resistance. However, the continuous aluminum moisture barrier reduces the transfer inductance to very nearly zero. Figure 6.1 EMA3D numerical model used for coupling of fault currents into a typical wayside signal installation AREMA 2016 1205

The transient sheath current creates a resistive voltage drop (electric field) of 2.6 V/m along the sheath outer surface. This electric field drives a current on the sheath outer surface (Figure 6.2(b)) and also creates a voltage difference between sheath grounding points along the line, and from the sheath to the surrounding earth. There are two possible ways in which the fault current couples to railroad signal systems: Currents are injected into the earth at the metallic screen grounding points which are located at splices. It is the current injected into the earth grounding point closest to the fault location that is of interest because this is the largest. Currents flow on the metallic sheath outer surface. These currents are caused by voltages which are the product of the fault currents (which also flow on the inside of the sheath) and the sheath resistance. This results in a line current source parallel to the nearby rails and induces current on them. 2500 Cable 1 Outer Sheath Current 2000 current (A) 1500 1000 500 0 0 0.01 0.02 0.03 0.04 time (sec) (a) (b) Figure 6.2 (a) Fault current transient flowing on the cable core conductor. Inset shows the leading edge of the current pulse; (b) outer sheath current 0 Current Injected to Ground at Fault Location 0 Voltage to Earth Ground at Fault Location -50-1000 current (A) -100 voltage (V) -2000-3000 -150-4000 -200 0 0.002 0.004 0.006 0.008 0.01 time (sec) -5000 0 0.002 0.004 0.006 0.008 0.01 time (sec) (a) (b) Figure 6.3 (a) Current injected into the earth ground; (b) Voltage on earth ground 1206 AREMA 2016

We note that the current shown in Figure 6.3(a) is the current from the fault propagating in one direction; there is also an identical current flowing in the other direction. Therefore, the total current injected into the earth is twice that of Figure 6.3(a), or ~400 A. The currents on the cable sheath together with the current injected into the ground represent the currents that can couple to nearby rails. EMA3D analysis of the configuration shown in Figure 6.1is performed to predict the amount of coupling to the rails and track wires. Typical track wire currents are shown in Figure 6.4, and are on the order of 6 amps. This is orders of magnitude less than the few thousand amps of fault current from overhead AC power lines. These are also orders of magnitude less than lightning currents, which are mitigated by arresters. The HVDC fault currents are therefore of no consequence. Figure 6.4 Typical Fault currents induced on track wires 7 Transient Effects from Energizing/De-energizing the Line When the HVDC line is energized/de-energized, the transient currents might couple to signal systems. The path for coupling a transient to railroad signal systems can be determined by using the same simulation models described above for fault currents. The coupling to signal systems is determined by the following: Current rise/decay times Transient current amplitude At least two events are of interest: Energizing the line De-energizing the line by a sudden collapse of power on the AC side of the converter The rise times of the energizing currents can be controlled by the operator and are measured in seconds or minutes, which is much slower than that of the fault current described in Section 3.6. Also, the amplitude of the energizing current is only about 10% of the fault current. Therefore the effects of energizing/de-energizing are bounded by the fault current effects. If there were a sudden loss of AC power, there is a very large capacitor (400 mf) and other design features which will keep the line energized with a very slow decay time. AREMA 2016 1207

8 Personnel Hazards from a Normally Operating HVDC Cable System Here we demonstrate that there are no hazards to personnel from a normally operating cable system. The cable system design features provide personnel protection as follows: The cable, shown in Figure 1.2, indicates a coaxial design. The cable shields enclose the center conductor so that the electric fields from the energized conductor are 100% contained between the core conductors and the cable shields. Therefore, there is no human exposure to high electric fields or voltages. The bipole design uses two of these cables, as shown in Figure 1.3. One cable has the center conductor energized to +300 kvdc, and the other cable is energized to -300 kvdc. This design ensures that the currents in each cable are nearly perfectly balanced. The shields are periodically bonded together and to earth ground (~900 ft or so), so that the shields conduct nearly zero current. As a result, the shields have nearly zero voltage with respect to the earth. Therefore, there is no human exposure to hazardous voltages. The HVDC cable system does produce above ground static magnetic fields from the energized conductors. The units used in the literature for the strength of the magnetic field or magnetic flux density can be confusing. The strength of the magnetic field and magnetic flux density can be stated with different symbols and units. Therefore, it is useful to define their symbols and relationships as follows: A commonly used unit for the magnetic flux density B is the Tesla (T, also equal to 1 Weber/m 2 ). Another unit of B is the Gauss (G). (The conversion factor is 1 T = 10,000 G). The magnetic field H in non-magnetic materials (B = µ 0 H, where µ 0 ~ 12.56 x 10-7 Henrys per meter) is measured in Amperes per meter (A/m). As a point of reference, the earth s magnetic field varies between 25-65 µt (micro-tesla; 1 µt = 10-6 T), or equivalently 0.25-0.65 G, or 20-52 A/m, from equator to poles and is a static field to which everyone is exposed. We have computed the strength of the magnetic fields caused by the HVDC pole conductor currents using our electromagnetic solver EMA3D (www.ema3d.com) and the results are shown in Figure 8.1, The figure indicates the fields just above the cables where a person would walk and also just above the nearest rail; the results are specified in the various units discussed above. Figure 8.1 Magnetic fields from the HVDC pole conductors computed by EMA3D 1208 AREMA 2016

The results show that the magnetic field exposure of a human walking above the cables is slightly less than twice the earth s background magnetic field. The medical literature provides the following information about human being susceptibility to static magnetic fields: MRI (Magnetic Resonance Imaging) technology uses strong DC magnetic fields in the range of 0.2 3 T, and in medical research applications magnetic fields up to 10 T are used for whole patient body scanning (World Health Organization, Electromagnetic fields and public health: static electric and magnetic fields, March 2006). We note that MRI exposure is at least 2000 times larger than the exposure to the HVDC line. Persons moving in a field above 2T can experience symptoms such as vertigo, nausea, metallic taste in their mouth and perceptions of light flashes. These symptoms are temporary. Other more serious symptoms, such as abnormal heart rhythms, only occur in DC fields in excess of 8T (World Health Organization, Electromagnetic fields and public health: static electric and magnetic fields, March 2006). We note that these field levels are more than 20,000 times larger than the exposure to the HVDC line. The International Commission on Non-Ionizing Radiation Protection (www.icnirp.org) has the following limits on occupational exposure to DC magnetic fields: A time weighted average of 200 mt per day with a ceiling value of 2 T. A continuous exposure limit of 40 mt is given for the general public. We note that these field levels are more than 400 times larger than the exposure to the HVDC line. Long term health effects if any to DC fields are not known. (World Health Organization, Electromagnetic fields and public health: static electric and magnetic fields, March 2006). Static fields, such as the one from the HVDC line were considered not classifiable either due to insufficient or inconsistent scientific information (World Health Organization International Agency for Research on Cancer, Volume 80 Non-Ionizing Radiation, Part 1: Static and Extremely Low- Frequency (ELF) Electric and Magnetic Fields and Electromagnetic Fields and Public Health: Extremely low frequency and cancer, International EMF Project, Fact Sheet No. 263, October 2001). We therefore conclude that no adverse effects on humans should occur from continued exposure to the HVDC magnetic fields. We have also investigated the susceptibility of pacemakers to static magnetic fields: Medtronic pacemakers/defibrillators are designed to operate normally in static magnetic fields measuring 5 gauss. (Medtronic Patient Letter Version 1.0 19 Jun 09). Boston Scientific states, As described in the pacemaker and defibrillator instructions for use, exposure to strong magnetic fields > 10 gauss (=1 mt) may alter implanted device function. (Portable Multimedia Players and Implantable Pacemakers and Defibrillators, Boston Scientific, March 6, 2009). Research presented at the America Heart Association s Scientific Sessions 2008 indicated that a Field strength of 10 gauss at the site of the pacemaker or defibrillator has the potential to interact with the implantable device. (American Heart Association Scientific Sessions 2008, B. Prescott, MP3 Headphones Interfere with Implantable Defibrillators, Pacemakers, November 9, 2008). The Institute for Biomedical Engineering, University and ETH Zurich, Switzerland also cites a DC magnetic field strength susceptibility of 10 gauss or 1 mt. (S. Ryf, T. Wolber, F. Duru and R. Luechinger, Interference of neodymium magnets with cardiac pacemakers and implantable cardioverter-defibrillators: an in vitro study. Institute for Biomedical Engineering, University and ETH Zurich, Switzerland). Therefore, most literature sources cite DC magnetic field exposure less than 10 gauss or 1 mt as safe for pacemakers. The most conservative estimate is 5 gauss. We conclude that effects on pacemakers occur at fields more than 4 times greater than the exposure from the HVDC line. AREMA 2016 1209

9 Personnel Hazards from an Internal Fault within the HVDC Cable System We define an internal fault as one which occurs entirely within the cable, and does not involve any external penetration of the cable shield from external forces or objects, such as a digging operation or derailment. This fault current event involves a short from a core conductor to its concentric metallic screen. Such an event could be caused by: Electrical breakdown in a void or imperfection within the XLPE dielectric An imperfection at a splice Section 6 provides the numerical simulation of such an event. The result is that a series of current pulses will be injected into the earth at the cable shield ground points, which exist about every 900 ft. The largest injected current and voltage to ground are located nearest the fault location and are shown in Figure 6.4. The voltage of more than 4.5 kv is unsafe. Smaller voltages will occur at other locations where the cable shield is earth grounded. If humans are in contact with the cable earth grounds, they would be exposed to these types of voltages. The fix is therefore to ensure that the earth ground cables and hardware are not available for human contact. 10 Miscellaneous Considerations There are other considerations that have to do with abnormal conditions, such as: Derailment Flooding Digging These can be mitigated by developing operating and emergency procedures and signage. 11 Conclusions The evaluation of buried HVDC power line compatibility with railroad corridors suggests that compatibility does exist, and is worth considering for implementation. List of Illustrations Figure 1.1 Typical cable burial configuration Figure 1.2 A typical HVDC buried cable Figure 1.3 Typical bi-pole design Figure 5.1 Lightning ground flash density for the years 1992 95 Figure 5.2 Metal puncture dependency upon charge transfer Figure 6.1 EMA3D numerical model used for coupling of fault currents into a typical wayside signal installation Figure 6.2 (a) Fault current transient flowing on the cable core conductor. Inset shows the leading edge of the current pulse; (b) outer sheath current Figure 6.3 (a) Current injected into the earth ground; (b) Voltage on earth ground Figure 6.4 Typical Fault currents induced on track wires Figure 8.1 Magnetic fields from the HVDC pole conductors computed by EMA3D 1210 AREMA 2016

Compatibility of HVDC Transmission Lines with Signal Systems AREMA 2016 1211 Rod Perala, PhD Jennifer Kitaygorsky, PhD Electro Magnetic Applications Inc. Lakewood CO 303-980-0070; rod@ema3d.com

1212 AREMA 2016 Background Innovations in HVDC transmission technology have made it possible to consider the compatibility of sharing railroad corridors with buried HVDC transmission lines. Objective: present a summary of compatibility issues and solutions. We know of no such existing shared corridors. In recent years we have performed corridor compatibility studies for two different Class I railroads. Results of these studies support the favorability of such installations. Topics HVDC Transmission Line Design Considerations Interference from Converter Noise Corona Noise Interference Corrosion Effects from Stray DC Currents Lightning Effects Fault Current Effects Transient Effects from Energizing/Deenergizing the Line Personnel Hazards Important Aspects of HVDC Transmission Line Design The General Architecture of an HVDC System HVDC Bipole design Pole conductor shielding Filtering at converter station interfaces Robust cable design The General Architecture of an HVDC System (continued) Typical Burial Configuration burial depth is ~1.5m The pole conductors are very nearly balanced in terms of voltages and currents. There are very nearly zero DC leakage currents in the cable shields and in the earth during normal operation. The cable shields completely shield the pole conductor electric fields from the outside world.

AREMA 2016 1213 Typical Possible Cable A high density polyethylene dielectric jacket A metallic shielding layer ~700 mil cross-linked polymer insulator Longitudinal water barrier A radial water barrier Interference from Voltage Source Converters (VSC) Noise Mitigation with large capacitors to ground at the converter outputs For example, a 400 mf capacitor filter impedance Z c =- / j fc), f is the frequency of the converter noise C is the value of the capacitor, ~400 mf. For example, at 100 Hz Z c =-1/ j fc) = ~4 m, providing effective filtering. Other factors Noise shielded by conductor coaxial shields Coupling from the pole conductors to a track in the common mode; grade crossing warning systems (and other signal systems) designed for differential modes. This provides mitigation by common mode rejection (CMR). Corona Noise and Interference No corona noise expected Large electric fields exist only within the cable shield, eliminating external noise Long Term Corrosion Effects from Stray DC Currents Two possible sources of DC current injection into the earth: Leakage currents through the cable dielectric Leakage currents through the transient arresters located at the converters Leakage currents through the cable polymer insulator For 300 kv, and a cable 100 km long, current ~ is about 100 μa. Current through arresters: less than 1 ma. Bi-pole design very nearly balanced; net leakage current much less than the numbers given above. Lightning Effects Lightning Effects Puncture of the cable to the central pole conductor might cause an immediate power fault. Puncture of the outer layers might not create an immediate power fault, but moisture ingress might eventually cause a fault. Likelihood of strike attachment to the cables in 30 years per 100 km of length: 90-150 strikes (see paper)

1214 AREMA 2016 Lightning Effects Cable features that mitigate damage XLPE jacket Longitudinal water barrier Radial water barrier Copper screening layer XLPE inner insulation Damage difficult to quantify, but expected to be minimal Fault Current Effects Fault events possible from the following causes: Accidental damage to cables by heavy construction and maintenance equipment Train derailment or other disaster Vandalism or terrorism Water or other liquid ingress into the cable Dielectric breakdown caused by manufacturing deficiencies (voids, material defects) Damage from animals Most Common: A short from a pole conductor to its metal return conductor Fault Current Effects Fault current effects computed with our computational electromagnetic (CEM) codes EMA3D and MHARNESS (www.ema3d.com) Fault currents on the sheath Fault currents injected into the earth grounds, Currents induced in the signal system track wires See paper for details. Fault Current Numerical Simulation EMA3D numerical model used for coupling of fault currents into a typical wayside signal installation Fault Current into Earth Grounds Fault Currents on Track Wires 0 Current Injected to Ground at Fault Location 0 Voltage to Earth Ground at Fault Location -50-1000 current (A) -100 voltage (V) -2000-3000 -150-4000 -200 0 0.002 0.004 0.006 0.008 0.01 time (sec) -5000 0 0.002 0.004 0.006 0.008 0.01 time (sec) Note: Earth grounds should be insulated from human exposure. Typical Fault Currents Induced on Track Levels are small and of no consequence.

AREMA 2016 1215 Transient Effects from Energizing/Deenergizing the Line Two events of interest: Energizing the line De-energizing the line by a sudden loss of AC voltage Energizing rise times can be controlled by the operator, measured in seconds or minutes; much slower than that of fault currents Amplitude of the energizing current is only about 10% of the fault current. Effects of energizing/de-energizing are bounded by fault current effects For a sudden loss of AC power, a very large capacitor (400 mf) and other design features keep the line energized with a very slow decay time. Personnel Hazards from a Normally Operating HVDC Cable System: Electric Fields and Voltages Electric fields normal to the pole conductors are contained within cable shields: no human exposure to high electric fields. Shield currents create small inconsequential voltages on system periodic earth grounds. No human exposure concern Personnel Hazards from a Normally Operating HVDC Cable System: Magnetic Fields Magnetic fields from the HVDC pole conductors computed by EMA3D Magnetic field exposure of a human walking above the cables is slightly less than twice the earth s background magnetic field. No effects on humans or pacemakers/defibrillators Miscellaneous Considerations Other considerations Derailment Flooding Digging Mitigated by developing operating and emergency procedures and signage Conclusion The evaluation of buried HVDC power line compatibility with railroad corridors suggests that compatibility does exist, and is worth considering for implementation.