Train the Trainer 101: Practical Underground Safety: Handling Neutrals and Rescue
Over the years I spent as a lineman, I did my share of underground installation and maintenance work. During my years in safety, I have seen the expansion of safety processes associated with underground, especially in response to the most recent OSHA changes. Not all of the changes have been effective, and that’s why we’re now going to spend some time addressing several underground safety questions Incident Prevention frequently receives. We’ll look at the rules and practices and what works from a practical perspective.
Handling URD Neutrals
This will not come as news to most of you, but for more than 60 years we have been splicing URD concentric neutrals during underground repairs without isolating the neutral or bonding across the open neutral in the ditch. That is something no lineworker would do on an overhead neutral, yet hardly any readers will be able to recall a time when someone was injured making neutrals in URD. Now, as OSHA’s language and expectations are more defined regarding grounding for personal protection, industry better recognizes current flowing in grounded systems, and employers are looking for ways to create equipotential and grounding during underground maintenance. For the most part, it’s not going well. The two questions I hear most are, why should we ground and how do we do it?
Let’s first address why we should ground, even though that is not a simple question to answer. Sometimes, as you will see later in this article, isolation may be preferable to grounding. But for starters, if you’re not convinced there is an issue that needs a protection plan, you won’t work very hard or smart when it comes to how you should do it.
There are several reasons why we should be concerned with risks associated with neutrals in underground. One is that, with the exception of delta systems, just like overhead neutrals, whether the cable is in a loop or a radial feed, neutrals are connected in a system. By design, interconnection of every neutral increases reliability for each individual neutral, ensuring a neutral path will be available for voltage quality, stability and fault clearing. But the very nature of interconnection means that the neutral current in your neutral may not be electrically associated with the phase current in the cable you are dealing with. Neutral current, like any current, will flow in every available path, inversely proportional to the resistance of the path. In other words, most current will stay on the main feeders since they usually have the larger neutral conductors, pole bonds and therefore lowest resistance. However, all you need is one bad double-dead-end jumper connection and that current will divide among all the remaining higher-resistance paths. In the most extreme case, it could mean that the ground rod in the 50-kVA padmount transformer at the end of a short single-phase radial is the main path to ground for a feeder circuit that has an open neutral. And no matter how many times you open the dropout supplying that radial, the ground rod at the end is still going to have feeder neutral current on it. This extreme example is to show that all neutrals, regardless of the status of the phases associated with a particular neutral, are subject to neutral current. Delta systems are the exception but have the same risks. In delta systems, transformers are connected in such a way that the neutral for the services is established locally at the transformer by grounding. In the delta systems I am familiar with, there are two modes of concentric application in underground. One is that the concentrics are grounded and serve as a shielding for cable insulation integrity. In the other, though the feed from overhead is delta, the underground systems are connected in a typical wye configuration and the concentrics function as a system neutral in the local underground loop. In each system, concentrics are subject to current flow by virtue of the bond between the neutral bushing and ground. Unless the concentric is isolated by disconnection from the system, it will have current flowing on it. The same fundamentals apply for fault current and induction current. So, there is risk. The issue with this particular risk is that you cannot quantify it as negligible or minor, even if you measure it with an amp meter during your job hazard analysis, simply because the system itself is dynamic. Conditions change. What might seem minor now could be fault current a few minutes later.
Bare Concentric vs. Jacketed Cables
The visual difference between bare and jacketed concentrics is obvious. In my discussions with crews and companies, I nearly always find that conditions and relative risks were not considered and may, in some cases, be a greater risk to employees depending on the scenario. There is one great benefit to bare concentric cable. When direct buried, bare concentric is in intimate contact with earth. This contact helps to discharge fault current under some conditions. In normal operation, relative to conductivity of the earth surrounding the cable, most current remains with the lower-resistance electrical connections. But as we have learned over the years, bare concentric, relative to conductivity and mineral composition of the earth, frequently becomes greenish-blue stripes on the cable. The cable may continue to operate with the earth, providing the symmetrical shielding necessary to control insulation leakage, but the ground rods in padmount transformers are the only source of return path for the respective loads. Jacketed cable will protect the concentric conductors from corroding, but it also isolates the concentric from the earth, so you lose any fault dissipation that may have occurred through direct contact with earth.
All of the above contributes to our relatively safe history of handling neutral connections during cable preparation. When we park an elbow or ground a live front, we have effectively grounded the phase conductor, but we have also put the phase conductor in parallel with the concentric neutral. If we ground the cable on both ends, we have helped to reduce the problem of capacitance by putting the phase conductor and neutral in series through the ground connection. So, there are any number of issues that can create current flow and present risks to the worker. Then why haven’t there been more reports of electrical injuries? It appears much of the protection we receive is coincidental to the typical connections we use in operating the system.
Just like every neutral in a wye system is connected, in underground loop systems almost every open neutral is also bypassed by the other parts of the loop. If you have an open neutral at the middle transformer on B-phase loop, that open can be described as “jumpered” by the continuous connection of the neutral associated with A-phase and C-phase loops. The same condition exists in a three-phase loop in a ditch. B phase may be burned in two, but A and C – if still together – are acting like jumpers.
A warning: Do not use this discussion as means to disregard the risks of open neutrals. Here we are simply observing the peculiarities of the system that may be part of the reason we have had some protection from neutral current flow risks in the past. But here is another part of the equation: open circuit voltage. When we open B-phase neutral, we would expect to see an open circuit voltage, and we will. However, that open circuit voltage is a result of resistances and current flow in the system. The neutrals on the associated loops have resistances associated with connections and length of cable. Yet in a grounded system, those resistances have to be pretty high in order to develop a voltage drop across the open neutral high enough to penetrate the electrical resistance of your skin. So, the current available may be high enough to present a risk, but the voltage present at the open is not high enough to penetrate the natural resistance of your skin. The result may be that you think there is not risk, and that would be a mistake. Everyone knows concentrics corrode and connections fail. The same coincidence of connections that keeps voltage low on one open neutral can fail and make the next open neutral deadly.
A similar bypass jumper coincidence at a cable-splicing operation is created by the phase conductor. When we repair a cable, we usually ground it at both ends. That puts both the phase and neutral in parallel. For the most part, we prep the phase-conductor splice while holding the cable, making few contacts directly with the phase conductor. Once we join the phase conductor, we effectively jumper the open concentric, again creating a jumper. The resistance necessary to develop a voltage across the open concentric is very low; in fact, it is the difference between conductor resistance and length. Obviously, then, if the open is at the center of the length of cable that is grounded at each end, the resistance would be about equal end to end. But when both ends of the cable are grounded, the phase conductor and neutral are bonded together, creating a continuous conductor so that both the open concentric and the bypassing phase are equal in conductor length resistance. In this case, the phase conductor acts as a bonding jumper across the open concentric. That condition has not necessarily solved all the risks of current flow, but it does explain why there is very low open-circuit voltage reducing the risk of electrocution. All of these issues become more critical during current rise associated with fault conditions.
Total Isolation
IEEE has a couple papers I am familiar with that will help employers develop safety procedures for working in underground. The first is “Worker Safety During Various Maintenance Procedures Common to Underground Residential Distribution,” written by C.C. King and published in 2003. The other is “Worker Protection While Working De-Energized Underground Distribution Systems,” written by IEEE ESMOL Subcommittee 15.07 and published in 2004. Both papers are available through IEEE. The ESMOL paper describes cable isolation, a method of protection I was criticized for recommending years ago. Many readers wrote to iP, rejecting the method because they thought isolation was a violation of OSHA’s requirement to ground a cable to work it. However, in 1910.269(n)(8), OSHA does permit cable isolation for protection of employees. Isolation, of course, requires disconnecting the neutral, and that in itself can create issues for system-voltage stability under some conditions. If used, isolation has to be done selectively by professionals who know how to identify the risks.
Vault and Manhole Rescue
We’ve addressed rescue many times in iP, but it’s been a while and there are always readers who are new to the publication. Thus, we still get questions regarding rescue requirements and practices. Since the “Train the Trainer 101” articles are about practical safety applications, let’s first clear up the question of outside third-party rescue. “Third party” usually means technical rescue by professional firefighters. The issue is time. With few exceptions, when rescue is performed by a fire rescue group, it will take a minimum of half an hour before the first firefighter even begins to enter the space. That is far beyond the four minutes we try to stick to when performing an overhead rescue. If your co-worker is severely injured, firefighters will be performing a body recovery, not a rescue. If you have people working in underground vaults or manholes, you owe it to them to have an effective means of extraction from the space in order to perform lifesaving first-responder treatment.
Confined vs. Enclosed Spaces
Firefighters don’t know the difference between confined and enclosed spaces, but you should. There are three conditions that differentiate an enclosed space from a confined space or an unclassified space. They are defined in OSHA standard 1910.269(x). According to OSHA, an enclosed space is a “working space, such as a manhole, vault, tunnel, or shaft, that has a limited means of egress or entry, that is designed for periodic employee entry under normal operating conditions, and that, under normal conditions, does not contain a hazardous atmosphere, but may contain a hazardous atmosphere under abnormal conditions.” The three criteria are limited access, that the space is designed for periodic employee entry and that it may contain a hazardous atmosphere under abnormal conditions. Notice there is no reference to energized conductors. These spaces include manholes, buried vaults as well as aboveground vaults, and substation transformers or breakers designed to be entered for maintenance.
Hazardous atmosphere is the larger issue and a very real threat. As the standard states, the hazardous atmosphere is one that may occur from an abnormal condition. A hazardous atmosphere can be created in a safe manhole simply by blowing rope through a conduit. If there are hazardous gases in the conduit, blowing rope through it will introduce hazardous gas into the work space. Another source may be natural accumulation of hydrogen sulfide (H2S) or methane created by conditions outside the walls of the space. Still another may be failure of a cable introducing toxic gases into the atmosphere in the vault or manhole. In fact, the smoke from burning insulation is toxic, containing hydrogen chloride, phosgene that forms hydrochloric acid in the lungs and dioxins that form chlorine gas. This consideration should be a prime concern in rescue planning, especially in the case of aging equipment. A person exposed to the smoke from a post-equipment or cable failure will be instantly overcome and unable to perform self-rescue. A rescuer will not be able to look into, much less enter, a manhole belching toxic smoke. Similarly, a person exposed to H2S is in immediate danger of poisoning, and a person exposed to methane is in danger of oxygen deprivation, if not an ignition and explosion. Methane and a variety of other gases that can enter an enclosed space may also come from industrial sources. An individual exposed to any of these conditions needs to be effectively removed from the space to fresh air, and the exposed worker will not be able to do it on his or her own.
The Attendant
As OSHA paragraph 1910.269(t)(3)(ii) allows, an attendant may periodically enter a manhole or vault to assist workers in non-emergency situations. This is where employers make the first mistake. Many think that because it is enclosed, the attendant can enter and consider roped-off workers as a requirement of the permit-required confined space standard. That is not the case. In an emergency, the attendant at an enclosed space is prohibited from entering the space. If extraction from the space is required, the attendant must perform a non-entry rescue. In order for that to happen, the entrants must be attached to a rescue line. It’s that simple. The attendant who enters when the emergency is atmospheric may well become – as is usually the case – another victim.
In the case of transformers and breakers, entrants are not likely to be overcome by atmosphere. And in a case in which the attendant can assure himself that the space is safe to enter by testing, or if he is replaced by a second qualified attendant, he can enter to assist with rescue, but only under those assured conditions. Obviously, preplanned rescue procedures should include notification and assistance protocols before an attendant enters that space.
For more information, you can read “Train the Trainer 101: Enclosed Space Rescue” (incident-prevention.com/blog/enclosed-space-rescue/), originally published in the October 2012 issue of iP.
About the Author: After 25 years as a transmission-distribution lineman and foreman, Jim Vaughn has devoted the last 17 years to safety and training. A noted author, trainer and lecturer, he is director of safety for Atkinson Power. He can be reached at jim.vaughn@atkn.com.
Editor’s Note: “Train the Trainer 101” is a regular feature designed to assist trainers by making complex technical issues deliverable in a nontechnical format. If you have comments about this article or a topic idea for a future issue, please contact Kate Wade at kate@incident-prevention.com.