Stray Current Control of Direct Current-Powered Rail Transit Systems: A Guidebook (2020)

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.

8 Literature Review A literature review was conducted to identify the body of knowledge, both domestic and international, which pertains to principles, procedures, methods, and criteria for achieving, documenting, and maintaining acceptable levels of stray current. This review will help the reader understand stray current, the process of its evolution, the associated corrosion, and the mitigation methods from the early days of the introduction of electrified rail systems to present-day design. The literature and articles reviewed date back to 1916 (5) and cover a range of national and international findings on this topic. The literature research includes the study of tech- nical journals, conference papers, books, design criteria documents from transit agencies, and the review of significant related articles and reports by research institutes and agencies or organizations. The list of various public and private agencies and sources includes, but is not limited to: the Transpor tation Research Board; the Transportation Technology Center; the Transit Cooperative Research Program; the American Public Transportation Association (APTA); the American Railway Engineering and Maintenance-of-Way Association (AREMA); Unified Facilities Criteria; and Australian, European, and South Asian Standards. The corrosion control criteria documents from various national transit agencies were used to develop a plat- form for comparative analysis and standard comparison. Some of the reports, in particular TCRP reports, gave an insight to a wealth of unparalleled history and background informa- tion on track-related research (6), rail base corrosion detection and prevention (7), and light rail transit design (8). In the United States, the Corrosion Society published recommendations on mitigation methods in the form of a report in 1921 (1). This report included some detailed mitigation techniques and construction methods. These techniques and methods were based on the study of transit systems at that time in Europe and America and included the following countries: • Germany—Earth Current Commission’s Recommendations (recommendations as adopted by the gas, water, and railway interests of Germany in 1910–1912). • France—Regulations by the Minister of Public Works (1911). • England—British Board of Trade Regulations (1894–1912). After reviewing recommendations from the previously mentioned countries and the research conducted by the local transit agencies in the United States, the Corrosion Society suggested that further testing and guidelines were warranted. Nothing substantial was done, however, until the 1950s and then the 1960s when the sensitivity of the stray current topic increased again (3). The 1960s was followed by another period in the 1980s and 1990s, when the National Aeronautics and Space Administration’s Technology Utilization and Industry Affairs Division conducted a research project for the U.S. Department of Transportation and C H A P T E R 2

Literature Review 9 produced a manual on corrosion control (9) and the first reference book (10). This reference book presents a compilation of more than 30 technical papers on stray current corrosion until 1994. The literature review conducted for TCRP Project D-16 provides a synopsis of the technical methods used to control stray current over the years and the recent advancements nationally and internationally. The literature review has shown that stray current is not as significant an issue in AC-traction power as it is in the DC-powered traction system (11), which is usually supplied by the overhead catenary system or the third rail. The following text summarizes the various concepts and topics covered under the literature review. 2.1 Corrosion and Corrosion Rate Corrosion could be defined as the deterioration of a material, commonly referred to as rust- ing (primarily when the metal is steel and iron), due to its interaction with the environment, that is, air, water, or soil. The practical definition is the tendency of the metal to revert to its natural oxide state (12). In simple terms, the corrosion process is a natural chemical or electrochemical reaction between a metal and its surroundings, in which the metal is oxidized (loses electrons), resulting in its progressive degradation. The corrosion tendency varies for different metals due to the varying energy content of the elements in their metallic state and is highly dependent on the surrounding environment. The process of corrosion requires four elements: electrolyte, anode, cathode, and a metallic path. Oxidization (loss of electrons) takes place at the anode-forming ions while reduction (gain of electrons or decrease in oxidation state) takes place at the cathode, which causes the anode to dissolve while the cathode remains intact. Electrolyte is defined as a solution of acids, bases, or salts containing free ions through which the electric current flows. The process of corrosion involves more than one oxidation and one reduction reaction (also referred to as anodic half-reaction and cathodic half-reaction). At least one such reaction must take place at the anodic surface for corrosion to occur so that the ions are formed and electrons are released. In case of electrolysis of underground structures, the moisture in the soil along with its dissolved acids, salts, and alkalis acts as the electrolyte, whereas electrodes are the metal utility pipes (13). Equations 1 and 2 represent a typical anode and cathode reaction, in which oxidation occurs when current leaves the rail to earth (anode reaction), whereas reduction occurs when current returns to the rail (cathode reaction): ( )→ +++ −2 oxidation at anode (1)M M e ( )+ + → −O 2H O 4 4OH reduction at cathode (2)2 2 e where M is the element involved (steel) and e is electrons. For corrosion to take place, both reactions need to occur at the same time. Checking if the damage is caused by a uniform attack or a localized attack measures the level of corrosion. In a uniform attack, the mass of metal corroded per unit of the surface area will define the damage, whereas in a localized attack the depth of penetration on a metal will define the cor- rosion rate (12). Corrosion rate is a function of many variables and thus in most cases it cannot be calculated without making some assumptions. The corrosion rates can be calculated using Faraday’s law by measuring the corrosion current flowing between the anode and cathode (the two ends) of the metal. Though the laws of physics for stray current corrosion are the same

10 Stray Current Control of Direct Current-Powered Rail Transit Systems: A Guidebook as those for galvanic corrosion, the metal loss is much faster due to the large amount of stray current leakage (14). See Equation 3. =Corrosion_rate (3)I nF corr where Icorr = corrosion current density in A/m2, F = Faraday’s constant 96,490 C/mole (Coulombs per mole of electrons), and n = number of electrons transferred per molecule of a metal corroded. To put things into perspective, one ampere of DC current that is constantly flowing from a metallic structure for 1 year will result in the dissolution of 20 pounds of iron, 75 pounds of lead, 22 pounds of copper, or 6.5 pounds of aluminum (15). To illustrate, the electrical decomposition of iron was calculated in Equation 4 by using Faraday’s Law (11) as ( )( )=Faraday’s Law: (4)m Q F M n where m = mass of substance liberated in grams (g), Q = total charge transferred in the reaction in Coulombs (C), M = molar mass of the substance in g/mole (M for iron = 55.845 g/mole), and n = number of electrons transferred per molecule of a metal corroded (n for iron = 2). One ampere will be equivalent to a flow of 1 Coulomb of charge per second (1A = C/s). Assuming mass of iron altered due to the flow of one ampere of stray current for a year and using the number of seconds in a year (Q = It, where I is the current in amperes and t is time in seconds), the resultant loss in mass of iron comes out to be 20.1 pounds (9,126 grams). From a practical viewpoint, it is not so much the total DC current that is important for the severity of corrosion but the local current density, because it is the local attack causing reduction in strength of structural members more than the uniform corrosion. Experimental studies have shown the effect of current density and duration of current discharge on corrosion of reinforcing steel and prestressing steel. According to a 2001 study supported by the Federal Highway Administration (FHWA) and the National Association of Corrosion Engineers (NACE), the annual cost of metallic corrosion in the United States is $276 billion. Only part of this cost is directly attributable to stray current corrosion. Significant savings can be achieved if proper inspection and corrosion management practices are employed (16). For the reliability and safety of public infrastructure, productivity of systems, minimal impacts to the environment, and economic competitiveness, it is essential to understand the fundamental cause of corrosion. Once understood it is possible to find the most effective options for cor- rosion mitigation and to generate best guidelines and principles along with regular inspection techniques. There are many forms of corrosion depending on the type of metal, the surrounding envi- ronment, and the length of exposure to the environment; however, this guidebook will focus only on the types of corrosion caused by stray current from the DC-powered transit system. Though DC-electrified transit systems are the main cause of stray current, there is another form of stray current called telluric currents. Telluric currents are caused by transient

Literature Review 11 geomagnetic activity. The influence of telluric currents on structures is for a limited duration due to nonlocalized discharge areas and thus is rare to find. Likewise, currents caused by other systems and operations are not discussed further in this guidebook, and the focus is on stray currents caused by the DC transit system. 2.2 Stray Current and Stray Current Corrosion by Transit Systems The operating current for the electric traction power supply flows through the overhead catenary system or the third rail to the vehicle and returns to the substation through the return circuit. The return circuit includes numerous conductors that help complete the path of the return current to the substation. Running rails are the most widely used conductors for the return of electric current. Since perfect insulation does not exist and rails have finite resistance, the return current leaks into the earth and finds its way to the substation via the path of least resistance. A handful of transit systems use a fourth rail system for the return of current, which is typically an insulated conductor fourth rail, electrically isolated from the running rails and the surrounding soil. This fourth rail collects the current and returns it to the substation. The alternative paths of least resistance that the return current may take include metallic utility lines, other metallic structures, reinforcement in the slab structure, and the soil itself. This current that takes the path of the least resistance (other than the rail) is called stray current and can be defined as the current that flows in the unintended path. Stray current corrosion is the corrosion that this stray current causes along its path (Figure 2). In a DC-powered rail transit system, stray current will follow any path of least resistance on its way to the traction power substation (TPS). This causes extreme corrosion to the metallic structures where it leaves the conductor. Hence, when there is a continuous flow of electric current, measures need to be taken to contain it at the source by providing suitable insulation or by using other means of rail isolation. This will prevent the current from flowing into the conductor earth. Stray current affects all the metallic components that are under the track, including the reinforcement steel supporting neighboring structures and rail track metallic components. Risk of stray current flow from the rail to other metal structures is greater when the potential difference between the rail and other metals is higher, which occurs in low resistivity soils. Stray Current Leaves Pipe and Enters Substation Corrosion Pipe/Utility Line Soil Current Return Path Substation Train Soil Cathode Region Overhead Catenary (or third rail) Anode Region Stray Current Enters Pipeline (Rail Used for Return Path) Rail Current Return Path (+) (•) Figure 2. Stray current corrosion path.

12 Stray Current Control of Direct Current-Powered Rail Transit Systems: A Guidebook To reduce this stray current, the rail-to-earth potential should be as uniform as possible over the entire length of the utility pipeline and the utility line needs to be electrically continuous (8). Stray currents are hard to detect since they are irregular because of varying dynamic rail traffic. The conventional method is to record the pipeline potential in the suspect areas for at least 24 hours. Corrosion rate depends on the current level (intensity), duration of current, and the properties of the metal. Figure 3 is a simple circuit model demonstrating the basic com- ponents affecting the levels of stray currents generated by a DC traction power system (17). The components within the simple circuit model include the following: RN is the resistance of negative return circuit, RP is the resistance of positive circuit, RL is track-to-earth resistance at the load end, RS is track-to-earth resistance at the source end, IT is train operating current, IN is current return through the rails, IL is leakage current to earth at the load end, IS is current returning to substation through earth, VS is substation voltage, VGL is track-to-earth voltage at the load end, VGS is track-to-earth voltage at the substation location, and VN is voltage developed across Rn by In. Equations 5, 6, and 7 present the relationship between the voltages VGL, VGS, and VN. ≈ + × (5)V R R R VGL L L S N ≈ + × (6)V R R R VGS S L S N = × (7)V I RN N N Corrosion rate is directly proportional to stray current and is more severe when focused on a small area. Unlike natural corrosion, however, stray current corrosion is independent of oxygen concentration and pH and is mainly related to the DC currents from rail transit (18). Although generally referred to as electrolysis, it is the process whereby chemical changes take place in the electrolyte when DC flows through a metal. The entire process of corrosion of underground metals is accelerated because of stray current. Rail-based corrosion gets worse (expedites) due to electrolysis caused by DC at the contact point with wet debris (mud or slime) that builds up under the rail base and also due to deicing salts (7). Stray current not only corrodes neighboring utilities but also affects the metallic structure of the transit system itself. Based on a 1990 report prepared and provided by NACE, Battelle Memorial Institute, and the U.S. Department of Commerce, the cost of corrosion caused by stray current was estimated to be $500 million annually (19). This number does not take into account the costs associated with signal problems and primarily accounts for the losses TRAIN SUBSTATION RP IT IT IN IL IS VGSVGL VS + - RSRL RN VN Figure 3. Simple circuit model illustrating stray current components (17).

Literature Review 13 to DC-powered transit agencies and detrimental effects to surrounding infrastructure and utilities. A recent TCRP study in 2007 on “rail base corrosion detection and prevention,” suggests that the steel used in the fabrication of rails can hold up to the effects of the environ- ment (galvanic corrosion); however, DC significantly affects the corrosion rate and makes the rails less corrosion resistant (7). Table 1 shows an estimated average stray current leakage by a transit system. 2.3 Traction Power Transmission of electric power has always been along the track by means of an overhead wire (see Figure 2) or at ground level by means of a third rail mounted on the insulators (extra rail) close to the running rails. (Safety concerns about a ground-level third rail are not discussed.) AC systems use overhead wires whereas DC systems can use either an overhead wire or a third rail; both are common. Current supplied to the train from the substation depends on the size and the number of train cars. Both AC and DC overhead systems require at least one collector attached to the train so it can always be in contact with the power supply. With economics and cost being the deciding factors in the selection of the train’s circuit return path, running rails have been used in most rail transit systems as the return conductor for the return of traction power to the substation. The running rails are at earth potential and connect to the substation. There have been long, ongoing debates since the inception of electric traction on which supply system is better, AC versus DC. The scope of this report is not to determine whether a DC or AC transit system is better; however, it is worthwhile to understand the basic variation between the two systems. The general rule has been that AC is for longer distance commuter and high-speed rails, whereas DC traction is used for shorter distances like metropolitan and suburban lines. In the early days, the AC-powered vehicle had to carry a transformer onboard to convert the high voltage to a lower system voltage. This transformer was quite heavy and therefore for smaller trains carrying smaller passenger capacities, it was inefficient in terms of weight per vehicle. However, the introduction of AC motors around 1965 eliminated the issue of converting the current into DC (21). For DC trains, the transformer is located at the substations, along with the rectifier, to supply DC power, thus increasing the efficiency of the vehicle, which represents a balanced weight/cost versus passenger capacity for short distances and makes it reliable as there is less equipment to fail. Throughout the globe, over half of all electric traction systems still use DC (21). How- ever, based on the literature reviewed so far and on the advancements made in traction power, AC traction power is the preferred current in countries building new rail systems, including Track-to-Earth Potential (volt) Track-to-Earth Resistance [ohms – (Ω)] per 1,000 ft. (305 m) of Track (2 Rails) 10 25 100 250 500 1,000 12,500 10 2.000 0.800 0.200 0.080 0.040 0.020 0.002 20 4.000 1.600 0.400 0.160 0.080 0.040 0.003 30 6.000 2.400 0.600 0.240 0.120 0.060 0.005 40 8.000 3.200 0.800 0.320 0.160 0.080 0.006 50 10.000 4.000 1.000 0.400 0.200 0.100 0.008 60 12.000 4.800 1.200 0.480 0.240 0.120 0.010 70 14.000 5.600 1.400 0.560 0.280 0.140 0.011 80 16.000 6.400 1.600 0.640 0.320 0.160 0.013 90 18.000 7.200 1.800 0.720 0.360 0.180 0.014 100 20.000 8.000 2.000 0.800 0.400 0.200 0.016 Table 1. Estimated stray current leakage (in amps) by a transit system (20).

14 Stray Current Control of Direct Current-Powered Rail Transit Systems: A Guidebook high-speed lines, primarily due to much higher reliability and reduced maintenance require- ments of AC traction motors (21). A high voltage direct current transmission system has been used for economy and power flow control but has yet to be used for a rail transit system. Some of the commonly used and proven traction systems in use by transit systems are the following: • DC 600 volt, 750 volt, 1,500 volt, and 3,000 volt overhead catenary; • DC 600 volt and 750 volt third rail; and • AC 16.7 hertz 15 kilovolt and 50 hertz 25 kilovolt overhead catenary. Table 2 illustrates some of the common advantages and disadvantages of both power systems based on the literature search. It is important to mention here that although this study focused on the damage caused by DC, leakages of AC at industrial facilities have also been suspected to corrode buried metallic structures. The general perception among corrosion engineers is that corrosion caused by AC is less severe than corrosion caused by DC. Thus, the possibility of damage to any metallic utility line resulting from AC corrosion seems small. 2.4 Soil Resistance—Corrosion and Earth Conduction A mitigation method in minimizing stray current leakage is to keep the rail-to-earth resistance high by electrically insulating the rail from the surrounding pavement/earth especially through urban/suburban streets and pedestrian crossings. This insulation inhibits the stray currents from entering the soil and causing corrosion of metals in the surrounding structures. Soil resistivity is used to gauge the degree of corrosion in underground utility lines. The literature review demonstrates that areas with low earth resistivity values result in an increased corrosion risk affecting metal pipes and other infrastructure in the absence of any stray cur- rent mitigation and collection system (13). In comparison with other conductors like copper and steel, earth is a poor conductor of electricity. However, it will change to a good conductor provided the area of the path of the current is large, which in turn lowers earth’s resistance. The resistance of the surrounding earth will generally be larger than the pipe resistance and the pipe-to-earth resistance and is heavily dependent on the soil type, temperature, and moisture content. Though resistivity of soil changes with the type of soil, it is difficult to give an exact value of soil resistivity and thus soil resistivity is defined in a wide range of values (see Tables 3 and 4). The amount of moisture content, soil content, and chemical constitution drastically affects soil’s resistivity (22). This includes the effects of chlorides and sulfates on soil caused by the deicing Table 2. Comparison of traction power systems. Train Handling Purchase Cost Delivers traction power at higher voltage, which then delivers power over longer distances and allows for less frequent substations. Delivers traction power at low voltage, which then allows for tighter clearance and requires more frequent substations. Challenged by its need to receive power from high voltage. Challenged by stray current. Ease of maintenance AC Traction Key Factors DC Traction Key Factors Draws unbalanced power from two of the three phases. Draws balanced power from the utility supply.

Literature Review 15 salts found on the track. For most soils, the pH value falls within the range of 5 to 8 and is generally not considered to be the dominant variable affecting corrosion rates (though higher acidic soils present serious corrosion risks) (7). An increase in temperature will also decrease the resistivity of soil whereas the resistivity will increase as the temperatures fall below freezing. This seasonal variation makes it difficult to assume a fixed value for earth’s resistivity and the only safe way to establish the correct resistivity values is to measure them. Resistivity of the soil is a significant factor in the determination of the most effective and efficient stray current collection system. Since so many elements factor into the resistivity of the soil, including temperature, soil resistivity studies must be performed and a worst-case scenario should be used for the design of a collection system. The allowable earth potential gradient development over a given length from the rails is determined by using the soil resistivity levels. The resistance (R) of any system of electrodes to earth can be theoretically calculated by Equation 8 (22): = ρ (8)R L A where r = the resistivity of the earth in ohm-cm, L = the length of the conducting path, and A = the cross-sectional area of the path. Additionally, the four-point method, also known as Wenner’s method, is the most commonly utilized method, depending on the depth, to determine the soil resistivity (R = r/2pa) (23). A study by Pham et al. in 2001 presented an earth potential gradient model that measures the potential developed between two points in the earth. The magnitude of this Soil Resistivity, Ohm-cm (range) Corrosivity Rating a >20,000 Essentially noncorrosive 10,000–20,000 Mildly corrosive 5,000–10,000 Moderately corrosive 3,000–5,000 Corrosive 1,000–3,000 Highly corrosive <1,000 Extremely corrosive aParticularly to chloride and sulfates. Table 4. Corrosivity ratings based on soil resistivity (7). Soil Description Average Resistivity, ohm-cm (range) Well-graded gravel, gravel–sand mixtures, little or no fines 60,000–100,000 Poorly graded gravels, gravel–sand mixtures, little or no fines 100,000–250,000 Clayey gravel, poorly graded gravel, sand–clay mixtures 20,000–40,000 Silty sands, poorly graded sand–silts mixtures 10,000–50,000 Clayey sands, poorly graded sand–clay mixtures 5,000–20,000 Silty or clayey fine sands with slight plasticity 3,000–8,000 Fine sandy or silty soils, elastic silts 8,000–30,000 Gravelly clays, sandy clays, silty clays, lean clays 2,500–6,000a Inorganic clays of high plasticity 1,000–5,500a aThese results are highly influenced by the presence of moisture. Table 3. Soil resistivity range (22).

16 Stray Current Control of Direct Current-Powered Rail Transit Systems: A Guidebook potential will have a direct link to the stray current effect on buried utilities. The earth potential gradient is calculated by using Equation 9 (17): = ρ π ln (9)• 1 2 E I l d d where I = current from source (amperes), d1 = distance from source of structure, d2 = distance from source of structure, and l = length of current source (parallel rail), cm. Equation 9 has the following limitations, as acknowledged by the authors: • Soil resistivity is assumed uniform within the length. • Earth potential gradients are assumed not to be distorted due to the presence of pipes and are unchanged between any two points. Occasionally chemical treatment of the soil is carried out before construction of the tracks to stabilize the soil resistivity and to provide a stray current collection system. Chemicals like sodium chloride, magnesium sulfate, copper sulfate, and calcium chloride have been known to have been used temporarily in the past to reduce the soil resistivity. A reduction of resistivity by 15% to 90% depending on the type of soil could be achieved by chemical treatment. This treatment is carried out mostly on high resistivity soils to ensure an effective low resistance grounding system and/or stray collection system (22). 2.5 DC Traction System Grounding (Earthing) DC traction power design includes three different earthing systems: the solidly bonded or grounded, the floating or ungrounded, and the diode-bonded systems. Most of the older transit systems used the solidly grounded system; however, the literature research that follows shows that it caused more problems than it solved. Consequently, floating, automatic grounding systems, and diode-bonded systems, emerged to satisfy the conflicting requirements of stray current and touch potentials. In a DC traction system, however, it is still a challenge to com- pletely stop stray current leakage and reduce the rail voltage at the same time. Thus, a suitable traction power design and selection of an appropriate grounding scheme are essential to reduce the rail voltage and the stray current leakage. To have a clear understanding of the subject it is important to realize the difference between system grounding and equipment grounding. System grounding refers to grounding of current conductors of the DC negative return system whereas equipment grounding refers to the grounding of enclosures of the rectifier unit and DC switchgear (24). Equipment grounding is not within the scope of this literature review. The main objective of system grounding for all the transit systems is to offer the continuity of a safe power supply. This includes not exposing any human being to electric shock in the vicinity of the earthen installation and minimizing DC stray current during normal and fault conditions. To achieve this objective, transit agencies use the following grounding methods along with their limitations. 2.5.1 Grounded or Solidly Bonded System A grounded or solidly bonded system is characterized by the direct metallic connection of the rectifier negative bus to the local ground grid at the substation. Absence of insulation on

Literature Review 17 the running rails is an optional characteristic. This system permits the unregulated flow of stray current where stray currents will leave the running rails along the entire length and will return at the substation ground grid using paths other than the running rails. This leakage of stray current increases the potential of corrosion and thus this system is not used in modern DC transit systems. See Figure 4. 2.5.2 Ungrounded or Floating System Unlike a grounded system, a floating system has no deliberate connection to earth and thus represents the other extreme of traction power design. High rail-to-earth resistance using rail boot, rail coating and rail fasteners restricts stray current. However, this could potentially result in increased running rail voltage, as compared with the grounded system, causing safety con- cerns for the public and transit agency staff. Moreover, during fault conditions, high electric potentials can develop between the platforms and the earth. Though these safety concerns present a downside to the system, this system is preferred over the other systems, and concerns have been addressed with the use of overvoltage protection equipment and platform insulation procedures. See Figure 5. 2.5.3 Diode-Grounded System In a diode-bonded system, the traction power substation is connected to the ground grid through a diode arrangement and stray currents can be collected (collection mat) and returned to the substations via the diode path. This system represents a compromise between grounded and ungrounded systems and is used to alleviate the problems in old grounded systems. In this system, the diode, in the negative return ground connection, will provide a low resis- tance path to permit the faster clearing of the fault currents. However, this also permits stray Figure 4. Grounded or solidly bonded system. Figure 5. Ungrounded or floating system.

18 Stray Current Control of Direct Current-Powered Rail Transit Systems: A Guidebook currents to return to the substation via the diode path, which can potentially increase stray current corrosion. Diode systems provide a unidirectional flow. This essentially means that they block the flow of current from the negative bus to the ground grid or collection mat. However, they allow the fault currents and the rail leakage currents back to the substation. Research shows that diode-earthing system may result in high touch potentials and stray currents at the same time (25). See Figure 6. 2.5.4 Additional Research on System Grounding A study presented in 2002 provided an account of DC traction power-system grounding practices in North America (24). The study first highlighted the difference between equip- ment and system grounding and then discussed the stray current leakage and personal safety affected by various system-grounding techniques. In addition to the three system-grounding schemes previously described, the author also presented the “automatic grounding switch” and “thyristor grounding schemes.” Figure 7 represents all the grounding schemes presented in the study. The thyristor grounding system will work as an ungrounded system under normal system operation and will ground the system only when an unsafe voltage occurs. Unsafe voltages may develop due to either train bunching load currents or due to positive (third rail)-to-earth faults. This gives the thyristor system an edge over a diode-grounded system, since diodes are always conductive (grounded system) under normal system operation and when there are small voltage differences between rail and earth. A study in 2005 (26) also presented the impacts of different grounding schemes. The study, with the help of a simulation model, demonstrated the advantages of a floating rail system. The study concluded that total stray current leaking from a floating system could be “four times” less than that in a grounded system. The authors highlighted the need to take steps to maintain safe levels of rail-to-earth voltages during fault conditions for the floating system. Two additional papers on the Taipei transit system published in 2006 (27) and in 2009 (28) carried out a detailed analysis of the grounding schemes and their effects on rail potential and stray currents. Simulation models were used to analyze one of the tracks (Blue line). The authors concluded that general ungrounded systems generate less stray current than diode-grounded systems whereas diode-grounded systems are used to reduce the stray current corrosion issues in old grounded systems. Based on the findings of the literature review, one can safely conclude that, to date, an optimal earthing setup that would decrease the stray current level and maintain touch potentials within Figure 6. Diode-bonded system.

Literature Review 19 safe limits is yet to be discovered. Numerous studies highlight the advantages and disadvantages of each scheme over the other; however, the effectiveness of grounding systems varies for different systems. Typical examples are two scientific papers presented by engineers of Railway Systems Consultants Ltd. (29), and Balfour Beatty Rail Projects Limited (30). Both studies concluded that the grounding schemes would have to be case specific and tailored to each application and to the dominant conditions. An optimal earthing setup is another area in which transit agencies need specific standards or guidelines to properly design the traction power system, but this is beyond the scope of this guidebook. 2.6 Effects of Stray Current on People and Animals The running rails of a DC transit system do not connect to structure earth or earth to avoid stray current leakage, which could potentially result in increased touch potentials. The local and time-dependent rail potential is the main reason for touch voltages under operating and fault conditions. Safety practices are followed to prevent electrical shock hazards to personnel who may be exposed to currents during system operation and maintenance and to the public Figure 7. Grounding system types (22).

20 Stray Current Control of Direct Current-Powered Rail Transit Systems: A Guidebook and animals. To avoid impermissible effective touch voltage, in certain cases, voltage-limiting devices are installed (e.g., at passenger stations) to achieve equipotential bonding between return circuit and earth. Stray current effects are also considered for the safety hazards they impose on people and animals who come in direct contact or in contact through a conductive part with the affected structure or the connected equipment. The severity of the electrical shock that a person receives depends on several factors such as the potential level and duration of the exposure, the human body and skin conditions, the path and magnitude of the current conducted by the human body, and the general health of the person prior to the shock. The permissible touch voltage therefore is subdivided into long-term values for operation and short-term values for fault conditions, including specific short-term operational situations. In depots and workshops, protection against impermissible touch voltage can only be achieved by direct earthing of the running rails, which potentially causes stray currents to leak. 2.7 History of Stray Current Corrosion and Methods of Mitigation Stray current corrosion has been a source of concern for the transit authorities and utility companies since the inception of electrified rail transit systems in the United States. The cor- rosion problem was originally believed to be caused by a chemical mix of the soil; however, after some research it was evident that the soil alone cannot cause the severe corrosion that was noticed in the rail base and nearby utilities. A conclusion was reached that the leakage of the traction current is the cause of this corrosion. The problem of corrosion caused by stray current in the United States was noticed within 10 years of the first DC-powered rail line in 1888 (31) in Richmond, Virginia. Since then, the control of stray current has been critical in the United States. By that time, Germany, France, and England had also already observed the effects of rail cor- rosion caused by stray current. Some of the worst stray current problems are found in the older DC transit systems and are fundamentally due to the following factors: • Poor insulation of running rails from the earth, • Improper and wide-spaced substations causing voltage drops in the rail, • Small rail cross sections of running rails resulting in high electrical resistance, and • Lack of maintaining a good return system. The stray current problem is tied to the fundamental design of DC-electrified rail transit systems, in which the running rails and return cables carry the return current to the substations. Considering the magnitude of the stray current corrosion impact, it becomes imperative to provide proper mitigation measures to control the leakage and the upkeep of the tracks by following a maintenance plan. 2.7.1 Historical Development 1890s to 1950s To address stray current problems and to provide the best mitigation options possible, during this era corrosion committees and the engineering community conducted numerous studies. Many of those recommendations were implemented on the newer designs at that time, with varying results, including some adverse effects on nearby utility lines, thus making it necessary to conduct further studies. It was in 1921 when some corrosion and engineering solutions

Literature Review 21 were recommended by the corrosion committee to reduce the leakage and the severity of stray current corrosion. Some of the measures that were successfully developed to control stray current leakage and corrosion follow: • Use of properly bonded joints (welded joints), cross bonding, and heavy rails for good track conductivity; • Use of high electrical roadbed resistance to earth and insulated negative return feeders; • Use of maximum number of traction power substations to reduce the return current distance, consistent with system economy; and • Use of three-wire traction power system. These four mitigation and control techniques are described in further detail. Bonded joints, cross bonding, and heavy rails. The use of heavy rail sections and suitably bonded rail joints was one of the earliest implemented mitigation methods for the control of stray current. The evolution of rail sections and steel along with other metal composition has continued throughout the years and across the globe. With time, rail sections have been improved in cross section, length, and the method of joining two sections of rail. With joints being the weak link in the track system, various methods of connecting rail lengths were experimented with before concluding that welded joints provide conductivity equal to or greater than continuous rail and are less subject to failure compared to other forms of rail joints. Thermite welds were the most common kind of welds used by the transit agencies during those times (1). Welding of rail lengths was thus acquired as the standard form of construction, especially in embedded rails, within the light rail transit systems. This has not only proved instrumental in the reduction of stray current but also improved the performance of rail. Cross bonding between single track and parallel track rails was installed to ensure rail connectivity and to equalize the current flow between the rails, thus reducing voltage drop (rail potential). In the United States, cross bonding was placed at a distance of 500 feet (152 meters) on urban railways and 1,000 to 2,000 feet on suburban railways. In Germany, cross bonds were provided every 328 feet (100 meters). In France, they were placed every 160 to 328 feet (50 to 100 meters) and, in England, they were placed every 120 feet (36.6 meters) (1). Resistance to earth and insulation of negative return feeders. Resistance of the ground immediately in contact with the rail depends primarily on the type of ground material that is in contact with the rail. Measures were taken to insulate the track from the earth to reduce the stray current process, thus reducing the corrosion of the base of the rails and other grounded steel structures for the sections of the rail that were embedded in the ground in urban areas. The corrosion committees suggested maintenance of the tracks to keep vegetation out of the tracks. The committee also suggested maintenance to keep the tracks clean, dry, and dirt and salt free to help keep the resistivity of the rail-to-earth high by keeping them insulated from earth (1). Well-drained broken stone ballast or gravel ballast was recommended for use in the non- embedded sections for its much higher resistance to stray current as compared with concrete. However, authorities in Germany and England were of the view that leakage of current cannot be reduced by roadbed construction. In the United States, it was recognized that well-drained crushed stone ballast had a resistance from 2 Ω to 5 Ω per 1,000 feet of single track. In comparison, the resistance of solid concrete ballast in contact with the rails and also earth roadbeds in which the ties are embedded was only from 0.5 Ω to 1.5 Ω per 1,000 feet of single track and 0.4 Ω for 1,000 feet of double track. Moreover, it was also established that resistance in dry weather may be three or more times higher than in wet weather per 1,000 feet of single track (1).

22 Stray Current Control of Direct Current-Powered Rail Transit Systems: A Guidebook Non-insulated negative return feeders were widely used in early constructions, especially where track bonds could not be well maintained. Supplementary conductors were installed in parallel with the track and connected to the track at frequent intervals to carry the current to the negative feeders and to ensure continuity of the return circuit. However, it was soon detected that these buried bare conductors increase the contact area between the return circuit and the earth, therefore counteracting the significance of their need (5). Later it was also deduced that the use of frequent substations along the route provided more economical increase in the track current drainage points compared with the use of insulated negative feeders. Maximum number of traction power substations. Another stray current mitigation technique that saw more advancement earlier on was increasing the number of substations consistent with economy. This technique reduces the feeding distances and the amount of current to be returned to any one point, resulting in the reduction of track voltage drop, thereby reducing the amount of current that will stray away from the rails. Technology papers by the Bureau of Standards on leakage of currents from electric railways, issued in 1916 (5) explain the importance of reducing the feeding distance to minimize stray current leakage for both grounded and ungrounded systems by using Equations 10 and 11. = +( ) ( ) δ − δ (10)i Ae Ber x r x = + − (11)i Ae Beax ax where = δa r and A and B are the integration constants. Formula for using boundary conditions for ungrounded system: at the beginning of the line x = 0 and the current I = 0, whereas at x = L the current in the tracks must be ioL. Thus, the total leakage current up to any point x can be calculated by using Equations 12 and 13. = − (12)1i i x io ( ) ( )= − sinh sinh (13)1i i x i L aL axo o Formula for using boundary conditions for grounded system: at the beginning of the line x = 0 and the current I = 0, whereas at x = L since the track is grounded, the leakage resistance between the track and earth is zero and the current in the tracks will be io = di /dx. Thus, the leakage current up to any point x can be calculated by using Equations 14 and 15. = − (14)1i i x io ( ) ( ) = − sinh cosh (15)1i i x i ax a aL o o where, in Equations 10 through 15, io = originating current per unit length of line assumed uniformly distributed, i = total current in rails at any point distant x from the outer end of the line,

Literature Review 23 e = potential difference between tracks and ground at any point distance x from the end of line, i1 = total leakage current up to any point, r = leakage resistance between tacks and remote earth per unit length of line, d = resistance of track per unit length of line, x = distance from outer end of line of any point under consideration, and L = total length of line. Making use of the above equations, Figure 8 defines the stray current curves. These curves show the effect of feeding distance on stray current for a defined load of 40 amperes per 1,000 feet, length of line 20,000 feet, leakage resistance of 0.4 Ω for 1,000 feet of double track, and track resistance of 0.004 Ω per 1,000 feet. Figure 8 allows a 10% increase in the resistivity of the track to account for cross bonding. The figure depicts the total current at any point on the line, stray current for a grounded and ungrounded bus with station at the end of the line, and then for an ungrounded bus with a station in the middle of the line reducing the feeding distance by half. It is observed that by providing the supply station at the middle of the line instead of at the end, the maximum value of the stray current can be reduced from 147 to 24 amperes (1). Equation 16 shows the maximum stray current for an ungrounded system. ( ) = − −    −max cosh 1 1 (16)1 1 2i i L aL u u u o where ( ) = sinhu aL aL Figure 8. Effect of substation spacing on stray current (1).

24 Stray Current Control of Direct Current-Powered Rail Transit Systems: A Guidebook Equation 17 shows the maximum stray current for a grounded system. ( ) ( )= −    max 1 tanh (17)1i i L v v o where v = aL Potential gradient on the tracks and potential difference between earth and rails can be calculated by using the preceding equations. The potential gradient and potential drop (respectively) for an ungrounded system are pre- sented in Equations 18 and 19. ( ) ( ) = δ sinh sinh (18)1E i L ax aL o [ ] ( ) ( )= δ − sinh cosh 1 (19)1E i L a aL aL o The potential gradient and potential drop (respectively) for a grounded system are presented in Equations 20 and 21. ( ) ( ) = δ sinh cosh (20)2E i ax a aL o [ ] ( ) ( )= δ − cosh cosh 1 (21)2E i aL aL o Figure 9 shows the overall voltage curves for the same line when the station is at the end, in the middle of the line (two stations), and at one third and three fourths of the total distance (three stations). The curves shown are based on a theoretical condition with no stray current Figure 9. Effect of substation spacing on voltage (1).

Literature Review 25 whereas the actual curves will be less since a portion of the current will leak to the earth (1). The electrolysis committees observed that the overall voltage reduces by the square of the feeding distance when the feeding distance is shortened (see Figure 9). Considering this marked effect on the reduction of the stray currents and overall potentials due to the reduction in the feed- ing distance, further detailed studies were conducted in the United States. The initiation of automatic and semiautomatic controls for substations made it economically feasible to increase the number of feeding points. Three-wire traction power system and other methods. This method was similar to the city power system, in which one trolley is negative and the other trolley is positive, and the tracks act as a neutral conductor. With proper application, this method not only reduced the stray current to one-half the value on some existing transit systems but also gave a better operating voltage for the cars (1). This method required the third and fourth rail to be a positive feed and negative return, respectively. However, because of the cost implications of adding a fourth rail or running two trolley poles in parallel on a single car, most transit agencies did not adopt this method. Besides the adoption of the aforementioned mitigation methods to control the leakage of stray current, further mitigation methods were warranted and thus procedures were adopted to protect the utility structures near the transit system that could be damaged by stray currents. These measures included surface coating of pipes, use of conduits in cable construction, use of insulating joints, pipe drainage, and interconnection of affected structures and the rail return circuit. Other measures included keeping new utility construction at a greater distance from rail lines, avoiding the crossing of rail lines, and placing utilities as deep as possible where the utilities must cross the tracks. Some of the methods proved effective, and some of the methods required more testing and development and were thus studied and investigated further. As with any other mitigation method, however, these procedures had limitations. While some methods that were originally recommended by the corrosion committee in 1921 are in use, the drainage bond mitigation technique was widely criticized later by the engineering community due to the variation of the conductivity of different types of pipes under different conditions, their material properties, and variety of joint types. Drainage bonds allow stray current to drain from underground structures through the switch back to the negative bus but prevent current flow in the opposite direction. A key fact to remember is that this use of diodes is different from diode grounding. In drainage, bond diodes and reverse switches are used to mitigate stray current corrosion on an affected structure. Insulated wires or cables are run from underground pipes or metallic structures to transfer and conduct the current from such structures to the substation (32), thus, potentially reducing the flow of current from such structures to earth and other conductors. Three types of drainage bonds in use are the following. Direct drainage bond: as the name suggests, it is a direct bond between the affected structure and the substation (or return circuit) and may include resistors. In a direct drainage bond, the current may flow in both directions. Therefore, direct drainage bonds may be used only when the potential at the connecting point of the bond to the DC current source is always more negative than the potential of the interfered structure, that is, the direction of current flowing in the bond will never reverse. Forced drainage bond: forced drainage bond includes a separate source of DC power to enhance the transfer of the stray current. Generally, forced drainage bond (also known as forced electric drainage) is used when a direct or unidirectional drainage insufficiently drains

26 Stray Current Control of Direct Current-Powered Rail Transit Systems: A Guidebook all stray currents from the affected structure, because the interfering structure does not have a sufficiently negative potential. The technique is used where the stray current originates from a DC traction system. For large and frequent voltage variations between the running rails and the interfered struc- ture, the drainage current and the potential of the structure will vary considerably. The potential of the affected structure can be maintained more negative than a preset value by the use of an automatically controlled forced drainage bond. When using this technique, the selection of a suitable site for the permanent sensing electrode must be done carefully. Unidirectional drainage bond: unidirectional drainage bond will include a diode to ensure that the current flows in one direction only. Therefore, the unidirectional drainage bonds may be used where the potential of the interfered structure is not always more positive than the potential of the DC source, for example, DC traction systems. Accurate design of drainage bonds is essential where excessive drainage might compound the problem and inadequate drainage might permit corrosion to continue. Even though drainage bonds are not popular, research shows they are still used by some transit agencies to avoid unsafe levels of track-to-earth voltages caused by stray current. 1960s to 1990s Taking advantage of the studies and detailed investigations conducted in the earlier era, most transit agencies adopted recommendations back then and augmented some of the miti- gation methods with the latest technological advances to further reduce the stray current to tolerable levels (once detected). More advancements were made in the areas of track-to-earth resistance, rail return circuit resistance, TPS distance, conductance of negative conductors, modification of surrounding underground utilities, location of track cross bonds, and magnitude of propulsion current. Design solutions, including the use of nonmetallic pipes for new utility lines, making metallic pipelines electrically continuous, installation of testing locations along new track construction, and maintenance solutions, were jointly recommended by rail transit agencies and utility companies. Significant adjustments were made to some rail transit systems, new and old alike, to keep the stray current leakage in check by decreasing the rail return circuit resistance and increasing the resistance of the rail-to-earth leakage path (3). Decreasing the resistance of the rail return path was achieved by • Increasing the cross-sectional area or size of the rail, which is achieved by using standard size rails ranging from 90 to 120 pounds (115 RE tee is the commonly used rail with a longitudinal resistance of around 40–80 mΩ/km). • Maintaining a continuous electrical path for the negative current by using continuously welded rails and welded cable bonds on special track work and frequent cross bonding (every 500 to 1,000 feet). • Decreasing the TPS spacing to 1 to 2 miles to reduce the voltage drop between the two substations. Increasing the resistance of the rail-to-earth leakage path, which is considered the most useful approach to mitigate the stray current leakage, was accomplished by undertaking the following measures (3): • Increasing the rail-to-earth distance by using well-graded, well-drained, and clean ballast, insulated track fasteners, and sealing compound or rail boot.

Literature Review 27 • Maintaining an ungrounded or diode-grounded negative circuit, though some researchers have observed that the rail life on diode-grounded transit system reduces to 20% of the normal life (10). • Isolating the track in yards and storage areas from the main track (33). For the safety of the agency staff from electrical shock in yards and shops, running rails were directly grounded and earthed back to the nearest substation via insulated cables. This resulted in excess leakage of stray current, since the only way back for the current to the substation was through the ground that contributed to the current leakage. Suggestions were made to provide a dedicated substation for the yard tracks and isolate them from the main- line tracks. Though some of the mitigation methods established in this era could be applied only to new transit systems (like grounding system, TPS spacing, or rail cross section), other mitigation methods could be applied to both old and new systems. Suggestions were made for a regular inspection, testing, and maintenance program following severe weather changes, street and pavement repairs, and after track and substation repair work to minimize the slippage of stray current leaks and avoid hefty cost repercussions in the form of utility line corrosions. Figures 10, 11, and 12 represent an example of a poorly insulated fastener connection, bro- ken rubber boot, and a missing insulated fastener clip and a rubber boot sleeve, respectively. Such careless mistakes can happen during construction, installation, and operation and were noticed during the stray corrosion testing conducted on tracks under construction. This is an example of a correct approach but of failed installation and maintenance, leading to stray cur- rent leakage. 2000s to the Present Design of earthing installations, TPS spacing, track-to-earth resistance, and return circuits for DC transit system are the most important defensive methods for controlling stray current and touch potentials. Most newer rail transit agencies have learned from past experiences and have started designing their rail lines with provisions for the control of stray current within the limits of their transit system by increasing the track-to-earth resistance. In the process, a few transit agencies have also incorporated testing and maintenance plans in their design criteria document. Tests like pipe-to-soil potential, track-to-earth resistance, track slab current measurements, and Figure 10. Cracked insulation cap on the fastener clip.

28 Stray Current Control of Direct Current-Powered Rail Transit Systems: A Guidebook cell-to-cell potential measurements are recommended by some transit agencies in their design criteria manuals. Various isolation techniques have been implemented by DC-powered rail transit agencies for the control of track-to-earth resistance in embedded tracks. This includes the use of rail boot. In the last two decades, the practice of rail boot usage has seen a significant increase by transit agencies in the United States for controlling the leakage of current in embedded track sections. However, experience has shown that the rail boot alone cannot always control the stray current leakage and that it is important to supplement the rail boot with additional stray current collec- tion and mitigation techniques. These methods reduce stray current corrosion by using various combinations of mitigation and collection techniques including, but not limited to, using elas- tomeric grout, insulating rail fasteners, embedding rail in troughs, providing current collection mats, and using collector cables. Figure 12. Missing insulation at clips and missing rail boot overlap. Figure 11. Broken rubber boot on the rail requiring excavation.

Literature Review 29 It is also evident that most of the mitigation methods and principles suggested by the cor- rosion committee that originated from the 1920s are still in use. Technical advancements have been made in mitigation methods, and new methods have been embraced by newer rail transit systems. However, the decision of when to use the relevant applicable method or a combination of methods and what level of stray current corrosion protection is needed still remains ambigu- ous for some rail transit providers. Despite recent technology advancements, the dynamic nature of the stray current problem renders it challenging to control it to a manageable level. Research has been conducted with innovative and experimental approaches like forcing the return current to return through a return conductor wire instead of rail or earth (34). However, the proposed return circuit decreases system efficiency. Moreover, transit agencies have started adding their own test facilities for the collection of stray current data in addition to utility company test facilities. Table 5 lists some critical stray control measures and principles identified in 1921 by the corrosion committee and still being used in present design supplemented by some advance- ment and recent developments. Following are various other methods or techniques that are used stand-alone or in combination with each other to achieve SCC (17): • Floating, diode earthed, and solidly earthed schemes; • Grounded systems and substations; • Floating returned rails; • Insulating pads and clips; • Insulating direct fixation fasteners; • Minimizing the stray current leakage path through rail or ballast contact by maintaining the ballast at a minimum of 1 in. below the bottom of the rails; • Cross bonding between rails and between tracks to maintain equal potentials of all rails; • Bonding rail jumpers at mechanical rail connections for special track work; • Insulating switch machines at the switch rods; • Utilizing separate traction power substations for the main line, yard, and shop; • Insulating the impedance bond tap connections from the housing case; • Maintaining as close substation spacing as practicable and as cost effective; • Placing substations near points of maximum train acceleration; • Increasing system nominal voltages; • Maintaining electrical continuity in tunnel liners and reinforcing steel; • Cathodic protection (CP); • Use of rail boots or insulating membrane for embedded rails; Description Corrosion Committee 1921 Currently Used Decreasing the Resistance of the Rail Return Path Rail size (cross-section area) X X Rail bonds X X Cross bonding X X Parallel conductors X X (rarely) Traction power substation X X Drainage bonds (case-by-case basis) X X (rarely) Increasing the Resistance of the Earth-to-Rail Leakage Path Track-to-earth resistance X X Ungrounded traction power substation X X Storage yard/main-line isolation X X Table 5. Stray current mitigation methods used for corrosion control.

30 Stray Current Control of Direct Current-Powered Rail Transit Systems: A Guidebook • Use of high resistivity concrete mix (chloride free) (18); • Epoxy coated reinforcement (not common) (35); • Use of current collection mat and collector cable (36); • Conducting regular testing of the transit system and nearby utilities; and • Maintaining an ongoing maintenance program that monitors rail-to-earth resistance values and keeps track-bed areas clean and well drained. While not all the methods can be implemented for every transit system, the research thus far has indicated that most of the design parameters like traction power, utility coatings, CP, substation spacing, and train headways for the transit system are standardized with the exception of track-to-earth resistance. Cathodic protection has been a popular and the most utilized mitigation technique to address both galvanic and stray current corrosion by the utility companies. Research has been carried out on the resistivity of concrete in the presence of stray current. The research covers the corrosion behavior of steel and steel fibers in concrete (18), corrosion damage of steel in concrete (37), and the corrosion rate of steel in concrete (38) in the presence of stray current. It is beyond the scope of this guidebook to go into detailed analysis of the resistivity of concrete and corrosion of steel in concrete. It is prudent, however, to share the findings of research that show that stray currents induce corrosion in steel in the presence of chlorides (39) whereas the risk of corrosion on steel fibers is low, primarily because the electrical connection between the steel fibers (manufactured fibers composed of stainless steel) is unlikely for the volume ratio of steel fibers (40). Research also shows that stray currents may aggravate fatigue damage of reinforced concrete (41). Moreover, it was observed during the construction of a local DC street car project that high resistivity concrete presents construction challenges. Challenges included special efforts for finishing, early-on shrinkage cracking of concrete, and extended curing times (42). A study by Yang et al. showed that stray current corrosion resistance of high resistive concrete with fly ash and powdered slag is more than five times that of regular concrete with 100-year-of- concrete design service life (43). 2.8 Rail Sections and Rail Boot There are two different applications for the tee rail (Figures 13 and 14). In one application the flangeway is constructed of grooved concrete and the other application includes the rubber boot to snap on to the flangeway section (creating a grooved section). The second application, which is the preferred one, addresses the problem of dirt and debris being collected in the con- crete flangeway because dirt and debris provide a conductive route for stray current. The rail boot snug fits both the girder rail and the tee rail and comes on a reel that is easy to deliver on the site along with other secondary joining materials. Depending on the boot manufacturer, the secondary materials may include epoxy grout, plastic ties, sealant to bond the rubber boot with the boot sleeve, duct tape, and the boot sleeve to connect and overlap two rubber boots at each end (44). Research has shown that the rail boot provides perfect insulation when first installed but with time, due to being in constant contact with the road traffic, different weather varia- tions, standing water, and rail traffic, it inevitably undergoes wear and tear and thus allows stray current to find a path to leak through it. Such current leakages are common in regions with moderate-to-heavy rainfall and in busy urban streets. Therefore, drainage design and track maintenance play a major role in achieving and maintaining SCC.

Literature Review 31 Special attention is also warranted for the proper placing of the rail boot and the subsequent concrete pour around it during construction. When the proper placing of the rail boot and the subsequent concrete pour around it during construction had not been done meticulously, there were instances where the track failed the safety test and the boot had to be reinstalled. Examples of such failure were observed during site visits to both local and international transit agencies, where multiple sections of track were dug up to reinstall the boot and clips. The damage to the rail boot results in degradation of the track-to-earth resistance and therefore is responsible for increased stray current leakage in a transit system. 2.9 Design Criteria and Standards A review of the design criteria or manuals of a sample of DC-powered transit agencies listed in Table 6 was completed to understand the source and origin of the limiting values listed in these documents and the isolation techniques adopted by agencies. However, none of the documents Figure 14. Tee rail with rail boot and concrete groove. Figure 13. Tee rail with grooved rail boot.

32 Stray Current Control of Direct Current-Powered Rail Transit Systems: A Guidebook mentioned the origin or the basis of the limiting values for the return voltage or the stray current. In some cases, it was not clear if the transit agencies conducted any initial baseline surveys to come up with these limiting values. These documents were either downloaded directly from the websites of the transit agencies or were provided by the traction or corrosion department at the agency. Some transit agencies did not have the design criteria document or elected not to provide any agency document. There is an American Society of Testing and Materials (ASTM) designation G165-99 that was issued in 1999 as a standard practice for determining rail-to-earth resistance. The authors found that some transit agencies or providers did not follow this practice. The authors reviewed the following English versions of European and Australian standards to understand what international transit agencies are using for SCC: • BS EN 50162:2004, British Standards Institution (BSI), Protection against corrosion by stray current from direct current systems, issued in 2004. • BS EN 50122-1:2011+A2-2016, BSI, Railway applications—Fixed installations—Electrical safety, earthing and the return circuit. Part 1: Protective provisions against electric shock. This standard issued in 2016 specifies the protective provisions in fixed installations related to electrical safety in AC and DC traction systems. • BS EN 50122-2:2010, BSI, Railway applications—Fixed installations—Electrical safety, earthing and the return circuit. Part 2: Provisions against the effects of stray currents caused by DC traction systems. This standard issued in 2010 explicitly deals with stray currents resulting from DC traction power and is the most applicable standard to this research. • BS 7430:2011, BSI, Code of Practice for Protective Earthing of Electrical Installations. This standard issued in 2011 gives guidance on the methods that may be adopted to earth an electrical system to limit the potential. • EP 12 30 00 01 SP, Electrolysis from Stray DC Current, RailCorp Engineering Standard— Electrical, Version 3.0, issued in 2012. • SPG 0709, Traction Return, Track Circuits and Bonding, RailCorp Engineering Standard— Electrical, Version 2.5, issued in 2011. • APTA RT-S-FS-005-03, Standard for Traction Electrification Stray Current Corrosion Control Equipment Inspection and Maintenance, 2004 draft. • NACE International, Task Group 297, Direct Current Operated Rail Transit Stray Current Mitigation, https://www.nace.org. The previous eight standards specify appropriate SCC measures that can be applied to DC systems along with some defense strategies against the effects of stray currents. However, these standards do not specify SCC testing or quality control methods. Reference Transit Agency Title Latest Revision 45 Houston METRO Design Criteria Manual April 2007 46 Phoenix METRO Design Criteria Manual January 2007 47 Denver RTD Design Guidelines & Criteria November 2005 48 New York City Transit Authority Corrosion Control Manual June 1984 49 Seattle Sound Transit Link Design Criteria Manual May 2011 50 Utah Transit Agency Design Criteria Manual July 2010 51 Portland, Oregon TriMet Design Criteria Manual January 2012 52 Washington, DC Streetcar Design Criteria Manual January 2012 Table 6. Design criteria manuals for transit agencies.

Literature Review 33 2.10 Chapter Summary With light rail transit (LRT) systems typically operating on embedded tracks in city streets, stray current corrosion is a major concern for track, utility, and other infrastructure owners near the DC-powered transit system. The literature review has shown that various mitigation measures and collection techniques are in use by transit providers. The literature review has also highlighted that based on site conditions, a floating system with ungrounded substations, shorter TPS spacing, and high track-to-earth resistance are the key design mitigation measures on newer systems. The stray current collection system is suitable for systems where the level of stray current produced by the transit system cannot be completely controlled by rail insulation. An excep- tion would be where stray current leakage is high. The goal of insulating the rail is to control the current at the source and minimize the stray current leakage. This “control at source” is achieved by reducing the distance between the TPSs, maintaining a continuous electrical path, using better coatings, cross bonding, and insulating track fasteners. The control at the source is also achievable via the use of rail boots, coating and insulating rail troughs, and isolating the tracks in yards and storages. However, research also has shown that there will be a certain amount of stray current leakage despite these control measures. This leakage usually happens after a few years of track life or even earlier if the tracks are not routinely maintained and tested. Stray current leakage cannot be completely eradicated, but it can be kept within acceptable levels and some of the newer transit agencies have been successful in achieving the desired levels of stray current. Many agencies, however, still struggle to control stray current corrosion. This situation may be improved through development and adoption of national standardized guide- lines, principles, and testing procedures and through developing and implementing regular testing and maintenance plans at the system level.