by Ed McConnell and Tom Nestor, SSOE Group
Identifying the risk of arc flash is not easy, but it is essential so utility workers can protect themselves. More than 2,000 workers were injured or killed by electrical shock and arc flash in 2010, according to the U.S. Bureau of Labor Statistics. Since then, no significant improvements have been shown in electrical-related fatalities.
Arc flashes typically are caused by a short circuit in an electrical system that results in an uncontrolled release of energy.
Third-party electrical systems analysis programs, however, help engineers identify which areas of the electrical system pose the greatest risk of injury. When the final report on incident energy values is complete and delivered, utilities must be assured that all possible scenarios have been considered so employees can be protected. The following key approaches provide solutions to two challenges of conducting arc flash studies.
|This excerpt from a data collection form shows how order is given to data collected. © SSOE|
Data collection typically accounts for 30 to 50 percent of the total cost of the overall study and is essential in developing an accurate arc flash study.
Assemble team. The team must employ and use people who:
- Are qualified according to OSHA definitions of a “qualified individual.” These definitions include being familiar with the equipment, knowing what voltage level with which they will be working, and knowing how to use the proper personal protective equipment (PPE).
- Are knowledgeable of the electrical system and qualified to operate it.
- Follow the standards that are established by the engineer, can interpret the data and log it correctly on data collection forms.
- Can open an electrical box, determine what size wire is connected, identify the specific fuse information and determine settings for circuit breaker systems, including adjustable trip settings.
- Bring a basic knowledge of arc flash evaluation and are trainable based on a solid knowledge of electrical systems.
Collection tools. Consistency in collection improves the flow of the overall process and often relies on using a team standardized data collection form that matches well with the data entry to be done later. This consistency is critical when an electrical contracting firm separate from the engineer is used for data collection. Interpreting the physical equipment data and converting it to an accurate one-line diagram is critical for entering the data correctly into the study software. The software builds an electrical model of the system. Accurate connections and interfaces are critical to produce accurate study results. Photographic documentation is also important because the ability to visualize the facility ensures data collection is more precise and speeds the compiling of information for the model. It also allows the engineer to visualize or remember the system in the office. In a study at a chemical company, it was unclear without photos whether circuit breakers were individual or in a panel board. A photo would have shown that what was designated as a panel was actually individual, molded case circuit breakers. This affected the modeling phase and required different input and more labels. Photographs should show how conduits are routed, whether into the ground or ceiling, and the accessibility of components. Wire sizes and lengths should be identified. Do their paths follow a direct route or do they run underground or up a wall and over a roof? Photos show how equipment fits into a space. A wide-angle shot that shows an electrical panel mounted on the wall with a conduit up to the ceiling clarifies whether the distance is 15 or 30 feet above the panel. A close-up of the equipment also can help identify the nameplate and model number. It should be taken with no flash so a bright spot does not cover information printed on reflective surfaces.
A good set of plan drawings also identify feeder and cable lengths. Drawings show where cables are horizontal and vertical and distances between feeders. Short feeder lengths are easy to estimate and do not affect study values significantly; however, longer feeders can be harder to estimate and have a greater significance in the study because of the higher impedance in the cable. The number of feeders in the system determines the cost of data collection. In a hospital, there might be 1,000 devices that need to be labeled. If it takes six minutes each to calculate feeder lengths, this part of the data collection can take up to 100 hours.
Preplanning. Data collection can become complicated when the subject is a large, complex electrical distribution system. The study can consist of several buildings, or there might be a number of system ties. These complex studies require careful preplanning and identifying a person familiar with the site to work with the data collection team. Some owners recommend a contractor who worked regularly on the site to assist in data collection. In complex facilities, electrical rooms sometimes are in obscure, difficult to access locations. Instead of using a door off the main corridor, the electrical entry might be inside a closet and up a ladder. Alternatively, the electrical equipment might be in the mechanical room but separated from the mechanical equipment by a wall or partition. It saves time to have someone on the team who is familiar with the electrical locations. The same person can help fill in important preplanning information that is missing from existing plan drawings and one lines, such as recent system changes.
Preplanning includes knowing the specific requirements of certain areas of the plant or building and orienting those who are going into the location. Some areas that require access for data collection have special PPE requirements. For example, when evaluating electrical equipment in the batch house of a glass plant, each collector must wear a respirator or a mask. This rules out people with beards because they are unable to wear respirators. Another example is some electrical equipment in an area designated as a confined space. In this case, the person must be qualified to go into such a space and be familiar with all the rules this entry entails.
Even with preplanning, in hazardous locations it might not be apparent that equipment is inaccessible while the process is operating. With a system shutdown, access still might be possible but could be encumbered by a switch cover secured with 14 bolts that normally could be opened with a latch if it were in a nonhazardous location. More time is required to get into a breaker in a hazardous location even if it is shutdown. To do an assessment under these conditions, hot work and additional permits might be required.
Ensure the right, knowledgeable employees on the site are available to provide access. This is a significant issue during spring break on a college campus, for example, when maintenance staff is a skeleton crew. Someone on that crew with keys must be available during shutdown.
Electrical bus ducts present another preplanning wild card. Usually mounted 15 to 20 feet above the floor, a system can have a concentration of plugs through a bus length of hundreds of feet. The team needs to know in advance if a man lift is required to access the plugs, especially if they cross production lines in a plant. As bus duct data must be collected during a shutdown, a process plant that runs 24/7 must alert the team when there is an opportunity for collection, which could be months away. When the study needs to be completed quickly, assumptions must be made based on the best information provided in preplanning.
Evaluating incident risk. Any unconfirmed assumptions that need to be made in data collection should have a conservative bias. When equipment areas are unavailable due to inaccessibility, the electrical engineer, maintenance manager or electricians can be relied on for temporary information. The assumptions can later be verified when the equipment is accessible, for example, if there is a shutdown six months later.
OSHA recognizes exceptions for equipment shutdowns. In a glass plant, for example, the electrical feed to a glass furnace might be shut off for two hours while glass sits. The collectors can wear safe PPE, such as 40-Cal suits, to look at high-energy equipment during this time. Often, though, it is wiser to make conservative assumptions rather than to put someone at risk for data collection. Conservative assumptions should lead to a higher level of risk. This leads to higher cost of PPE and more cumbersome access issues, but it ensures personnel will be protected against the incident energy that is possible.
|Time Current Curve (TCC) illustrates longer trip time for lower fault current and higher incident energy. © SSOE|
Sometimes during the data input phase, analysis software detects incomplete data. In a hospital project, for example, the model type of one circuit breaker in a substation was properly identified but another breaker type was not. As a result, all breakers were assumed to be of the type identified. When the original panel had been installed, extra spaces had been provided so additional breakers might not have been the same model. The difference in model type of a 30-amp breaker might not make much difference in the study results, yet the information must be verified as accurately as possible so the analysis will keep the electricians safe when working on the panels.
Accurate information helps electricians determine the proper PPE required for their locations. Conservative assumptions instead of actual data can add additional cost in providing a level of PPE that is not required. The results also can force the technicians working on a device to work in suboptimal conditions. Electricians do not want to have to put on 40-Cal (Level 4) suits because of the difficulty of working with tools and the uncomfortable conditions inside the suits. Many industrial plants require all maintenance people to wear 8-Cal clothing, which has a Level 2 arc flash rating. The risk category in which they work can vary from Level 0 to Level 2 throughout the day, but the convenience of always being ready for the higher level is important to them, and Level 2 clothing is comfortable. The Level 4 suits can be in more limited supply and are used rarely as necessary, which limits the extra cost that goes into expensive suits and their care.
Calculating incident energy hazard. Once the data is collected and entered in the electrical study software, the results of the arc flash study can be obtained. Usually the results are predictable and intuitive. That is, when there is a large transformer that provides a lot of short circuit current, the arc flash energy level is high. The study details how high that level will be.
There are some cases, however, where arc flash values are different than one would expect. It might seem obvious to assume that when short circuit current level goes down, the incident energy also goes down, this is a not true in all cases.
If a device is exposed to a lower short circuit current, it might take longer for that device to operate. Bus ducts can be a case in point. With a long bus duct, which has a high impedance, the short circuit current is higher at the source end than at the far end. A protective device at the far end experiences a lower short circuit current, thus it takes longer to open the circuit because of the time and current characteristics of the trip device. The arc lasts longer so energy is allowed to flow into the fault for a longer time.
Bus ducts should be considered a series of bus ducts instead of a single bus because of the impedance of the bus. A long horizontal bus duct may be considered in 50-foot increments or some distance that makes sense for the grouping of the connected equipment. In a hospital project where a vertical bus went from the ground to the 15th floor, each floor’s bus duct was considered a separate 15-foot section (the floor to ceiling length). The available incident energy in each section was different.
Many industrial plants, including most glass plants, have multiple power sources. Many have substations that are double-ended with a tiebreaker so the power can be fed all from source A, all from source B or half from each source. Other plants have a ring bus setup such that, depending on where in the ring the feed is opened, portions of the plant load can be fed from different sources. In any of these multiple-source systems, calculations must be made for all scenarios of the loads’ being fed. The short circuit currents may differ for various arrangements of the feeds, and thus the arc flash energy at any point in the system will vary. Often when such arrangements occur, the equipment is labeled with the worst-case scenario level of arc flash energy because the person going into the area does not always know which switches are opened and closed throughout the system.
Regardless of its type of electrical system, the owner must be assured that all possible scenarios have been considered, proper incident energy values have been determined and an accurate arc flash risk study has been conducted and compiled. Only then can an owner rest assured that employees who enter areas where electrical equipment is present are properly informed about the possibility of injury at work.
Ed McConnell, a registered professional engineer, is a master engineer at SSOE Group, a global engineering, procurement and construction management firm. With more than 35 years of experience, he specializes in facility electrical power distribution, large motor starting, lighting, electrical systems grounding, building grounding, building utilities electrical power and control, substations, security systems, communication systems, and fire detection and alarm systems. Reach him at 419.255.3830 or email@example.com.
Tom Nestor, a registered professional engineer, is an electrical engineer at SSOE Group. With more than 40 years of experience, he is responsible for the proper technical and administrative completion of projects, including facility electrical power distribution, large motor starting, lighting, electrical systems grounding, building grounding, building utilities electrical power and control, substations, security systems, communication systems, and fire detection and alarm systems. Reach him at 419.255.3830 or firstname.lastname@example.org.
PowerGrid International Articles Archives
View Power Generation Articles on PennEnergy.com