Previewing NEC 2017 changes
The current version of NFPA 70: National Electrical Code (NEC) is the 2014 edition. Though this version has not been adopted in all jurisdictions, this article reviews a project that is being designed to meet the 2014 code. A few key updates to the NEC in 2017 and their potential impacts to future designs are also highlighted.
- Understand NFPA 70: National Electrical Code (NEC) and when to apply it.
- Discover how to work with the authority having jurisdiction to apply the NEC in specific cases.
- Learn how to size a service to an industrial facility using alternate methods allowed by the NEC.
- Recognize upcoming changes to the 2017 NEC.
Designing industrial facilities requires diligence, long hours, close adherence to the National Electrical Code (NEC), and often, coordination with the authority having jurisdiction (AHJ). For industrial facility electrical designs to be safe and economical, the engineer plays a critical role in designing a properly sized system that avoids waste and oversizing.
An approach CH2M is taking on a recent industrial project includes steps to gather necessary client data, size the service in coordination with the AHJ, and ensure that the 2014 NEC is properly applied. A few key updates to the NEC in 2017 and their potential impacts to future designs are also highlighted.
Gather facility design basis info
The first step in the process is to gather the fundamental design information from the client so that a functional and code-compliant design could be achieved. After several client meetings and extensive data gathering, the client provided the following major design criteria:
- The design is for a greenfield facility that closely resembles an existing plant.
- Outages can be tolerated once a year, with the objective of limiting complete outages to one event in 5 years. This must be counterbalanced with the expense of construction.
- Energized work is to be avoided; where possible, provide the ability to reduce the incident energy to less than 1.2 cal/cm2.
- Liquid-filled outdoor transformers are preferred over larger (1-MVA and larger) transformers.
- Provide 2 MVA of service capacity for future expansion
- The location of the site has not been confirmed, but based on the probable location, the design must follow the 2014 NEC.
When the design boundary conditions have been determined, the next critical step is to determine the service size. The NEC has prescriptive rules for calculating the size of new services. In practice, it is rare to calculate the service size for an industrial facility using the standard methods shown in the NEC. These would more typically be used for commercial applications. For most industrial processes, sizing the overall service for anything close to the connected load would result in a system that is much larger than required, costly, and wasteful. Chapter 2 of IEEE 141: IEEE Recommended Practice for Electric Power Distribution for Industrial Plants discusses this topic in more detail and is an excellent reference.
There are several sections of the NEC that can be used to justify a smaller service size. The most important section is Article 90.4, which states, "By special permission, the AHJ may waive specific requirements in this code or permit alternative methods where it is assured that the equivalent objectives can be achieved by establishing and maintaining effective safety." If this section is used, it is incumbent upon the engineer to provide concrete evidence that the system he or she has designed is safe and the size can be justified. In the case of our plant, we are fortunate that there is an existing plant that runs a nearly identical process. Using 12 months of utility data, existing client metering, and collected field data, we've determined the loading on the existing facility to be:
- Connected load: 5.0 MVA
- Maximum demand: 3.0 MVA
- Average demand: 2.5 MVA.
However, there are several key differences existing between the two sites that prevent us from using the existing facility information without modification:
- The new site has different climactic conditions that increase the cooling load due to increased use of chillers. Our helpful mechanical engineer was able to estimate the increase in mechanical loading as a 20% increase in peak mechanical loading over the design basis and an approximately 10% increase over the average.
- Due to process-control changes, there will be a higher use of process equipment. As a result, the ratio of the anticipated average demand over the maximum demand is greater for the new facility. In practice, this makes little difference to the design internal to the plant, but it is important information to provide to the utility.
- The new plant will be larger, and additional lines of similar equipment will be installed.
The projected load data for the new facility is expected to be:
- Connected load: 14 MVA
- Maximum demand: 10 MVA
- Average demand: 9 MVA.
At this point, it should be obvious that the connected load has nearly tripled, yet all of our demand numbers have increased; i.e., we do not appear to be taking credit for additional load diversity. After closely examining the loads in the new facility, we determined that the majority of the equipment would operate coincidently and would not result in an increase in diversity. Based on the collected data and allowing for the additional 2 MVA for future load growth, our service is sized for 12 MVA.
Most AHJs will allow industrial services to be sized under engineering judgment using similar methods, as described above. In the event that you do not have an AHJ that is willing to approve an approach similar to this, there are several code sections that may allow you to reduce the size of both services and feeders.
NEC Article 220.60 allows you to use the largest load of two or more noncoincident loads for both feeder and service calculations. In addition, NEC Article 430.26 allows " ... [for] motors operating on duty cycle, intermittently, or from all motors not operating at one time, the authority having jurisdiction may grant permission for feeder conductors to have an ampacity less than specified in 430.24 ... " Lastly, for existing services, the engineer can use the provisions of Article 220.87 and calculate the service load by multiplying the maximum measured demand over the last year by 125% and then adding the new load.
Initial design decisions
Based on the client's desire for reliability, reduced maintenance intervals, and arc flash reduction, several initial design decisions were made. These decisions (in bold), along with the key reasoning behind them, are briefly explained below:
- Medium-voltage switchgear configured in a main-tie-main configuration allows for loss of a utility service, utility maintenance, and maintenance of medium-voltage breakers and bus (see Figure 1).
- Outdoor liquid-filled 12.47-kV to 480 V ac transformers reduce building cooling loads and the building footprint. However, this created two issues to address:
- Long taps-the taps from the transformer to the switchgear were approximately 110 ft (see Figure 1). The provisions of NEC Article 240.21(C)(4), Outside Secondary Conductors, were used to allow for the long secondary taps, which included installing the conductors under at least 2 in. of concrete when under the building until the conductors were terminated on the main breaker.
- Incident energy levels-the arc flash incident energy levels were extremely high, so differential protection was added. The differential zone was designed to include the medium-voltage cable, the 12.47-kV to 480 V ac transformers, and the secondary of the transformers. The differential function is shown as ANSI device number 87 in Figure 1.
- Due to the 30-cycle rating, low-voltage switchgear allows the engineer to disable instantaneous trips on the main breaker. This results in greater selective coordination with downstream feeder breakers, but does increase the arc flash incident energy.
- Adding infrared (IR) windows to electrical equipment allows for IR scanning of breakers and terminations without opening equipment doors.
- Zone-selective interlocking (ZSI) at the switchgear level allows the trip devices to communicate so that the nearest upstream device to the fault operates without intentional delay. This can significantly reduce arc flash incident energies.
- High-resistance grounding (HRG) allows continuous plant operation during a single line-to-ground fault and reduces the probability of an arc flash event. The transformers shown in Figure 1 were specified with a 5-amp HRG system. For more discussion on this topic, refer to "Choosing between grounded and ungrounded electrical system designs" (Pure Power, Fall 2013, p.20) and "Well-grounded facilities." Refer to NEC Articles 250.36 and 408.3(F)(3) for HRG system code requirements (referred to as high-impedance ground in the NEC).
- NEC Article 240.87 requires arc energy reduction for breakers that can be rated or adjusted to 1,200 amps or higher via one of five methods:
- Differential relaying
- Energy-reducing maintenance switching with local status indicator
- Energy-reducing active arc flash mitigation system
- An approved equivalent means.
Differential relaying and ZSI have already been applied as previously discussed. In addition to those methods, the electronic trip units specified also include energy-reducing maintenance switching. The application of energy-reducing maintenance switching may impact selective coordination as required in Articles 700, 701, and 517. Use of other methods, such as ZSI, may be more appropriate on those systems. At the time of specification, breakers and trip unit manufacturers were not known, so the actual incident energy value that can be achieved is also unknown. The energy-reducing maintenance switching can be used to further reduce incident energy if the ZSI and differential protection do not achieve a sufficiently low value.
In addition to the typical process loads connected to the system in Figure 1, the new facility also contains emergency (NEC Article 700), legally required standby (NEC Article 701), and optional standby (NEC Article 702) loads. Article 700.10(B)(5)(a) requires that wiring from an emergency source to supply emergency and other loads is in, "Separate vertical switchgear sections or separate vertical switchboard sections, with or without a common bus, or individual disconnects mounted in separate enclosures shall be used to separate emergency loads from all other loads." This segregation is shown in Figure 2. The 2014 NEC introduced Article 700.8, which requires surge-protective devices on or in all emergency system switchgear, switchboards, and panelboards, so these were incorporated into the emergency system.
As what sometimes happens for these projects, the initial estimate came back with a higher cost than the client could support. To align the project budget with the client's business case, several critical modifications were made:
- Modify medium-voltage main-tie-main to a simple radial system, as shown in Figure 3. This reduced complexity, removed two 1,200-amp breakers, and reduced the building footprint. While a simpler system, this came at the expense of system redundancy and maintainability. Removal of the additional utility feed and the redundant main breaker will require that the plant take an outage for routine maintenance or in the event of loss of utility.
- Simplify the 480 V ac system to a radial system, which involved removing one 5,000-amp breaker per switchgear, transfer controllers in the switchgear, and the associated tie bus. Simplifying to a radial system allows for full use of the 12.47-kV to 480 V ac transformers because spare capacity does not need to be maintained to allow for supporting loads from both sets of switchgear on one transformer.
- Remove the emergency switchgear section. Because the NEC Article 700 switchgear section only supported emergency lighting loads, a listed lighting inverter was used instead to support the emergency lighting requirements. Placing the emergency lights on a lighting inverter allowed for removal of the emergency switchgear section, emergency automatic transfer switch, and associated cabling. Refer to Figure 2 to view the original configuration and Figure 4 for the revised standby power configuration.