The Skinny on Switchgear and the New NEC

By Keith Lane, P.E., RCDD/NTS Specialist, LC, LEED AP, Vice President, Engineering, SASCO, Seattle September 1, 2005

It must be understood that the short time delay from not having an instantaneous trip can allow fault currents to flow for several cycles (typically up to 30 cycles), which can subject the electrical system to high mechanical and thermal stresses.

The system equipment must also be coordinated to ensure it can withstand this tripping time delay. Switchboard construction will handle its rated fault current for only three cycles, while switchgear construction can handle its rated fault current for 30 cycles. There is, however, one particular manufacturer that I know of who makes a hybrid type of switchboard that will handle the rated fault current for 30 cycles. But in general, if the engineer chooses to go the route of providing breakers with no instantaneous function, the system must use more expensive switchgear construction. Furthermore, this type of construction will require rear access, which will typically increase the size of electrical rooms and decrease leasable space.

In addition, the instantaneous portion of the time-current curve can only be turned off on ANSI-rated power breakers. Typically, if the system utilizes molded case or insulated case breakers, the breakers utilize an instantaneous override function for high-fault current levels. In this situation, selective coordination can be lost if the available fault current is above the preset override current level.

For these reasons, as well as the fact that many small breakers do not have the option of a “no instantaneous trip” setting, breakers without instantaneous tripping are more suited for larger systems. Smaller emergency and standby systems are not well suited for this type of selective coordination between breakers. If you have a short circuit on a 20-amp thermal magnetic branch circuit that is protected on the supply side by a thermal magnetic 100-amp panel main breaker, there is typically no way to verify for sure which breaker will trip first. Either or both breakers may trip, which would indicate that the electrical distribution system does not utilize selective coordination.

With zone interlocking, selective coordination can be achieved without removing the instantaneous trip setting. Zone interlocking is a protective function that will minimize the extent of an outage. The interlock function can be utilized for a ground fault or a phase fault. During a fault condition with a pre-programmed and predetermined current level at the load side of a downstream breaker, the downstream breaker can be set to trip instantly. This breaker can be programmed to send a restraining signal to the upstream breaker not to trip. The restraining signal can be coordinated between multiple levels of overcurrent protection devices. This system would require additional electronic relays, control wiring between the relays and engineering and analysis to determine the correct settings to ensure full selective coordination and to avoid nuisance tripping.

This type of system would certainly add significant cost and complexity to the electrical distribution system and would not be suitable or cost-effective for a majority of medium and small emergency electrical distribution systems with multiple 20-amp single-pole breakers.

On the other hand, I can’t see engineers and designers across the country placing fused switches and fused panelboard in facilities with emergency or standby distribution systems. Fused systems will take up considerably more space than a typical panelboard. Additionally, most maintenance personnel currently enjoy the ease of resetting a breaker instead of ensuring that spare fuses are available and replacing spent fuses in the event of a short circuit or overload. A lack of good maintenance protocols can actually reduce site availability and uptime if spare fuses are not available after an overload trip or short circuit.

Another factor with utilizing fused distribution includes the possibility of single-phasing motor loads from the loss of one of the three-phase conductors.

Due to the critical nature of the emergency and legally required standby electrical distribution systems, selective coordination has been mandated as a revised code for these systems in the 2005 National Electrical Code.

I do agree that emergency distribution systems should require a high degree of site availability and uptime and should strive toward full overcurrent protection and selectability between protective devices. This selectability can be accomplished with both fuses and circuit breakers when they are selected appropriately and when overcurrent protective device settings have been coordinated.

Selective coordination of emergency and legally required standby systems with supply side overcurrent protective devices will provide for a more reliable electrical distribution system, but I think local code authorities should think long and hard about the potential far reaching ramifications of the full implementation of this code change. There may be some middle ground.

For example, the main breakers only, in the main emergency or standby distribution panels feeding the emergency and standby sub-distribution panels, could have the instantaneous portion of the breaker turned off.

The remainder of the downstream breakers could then be the normal type of thermal magnetic breaker. In this situation, under the worst-case scenario of a fault current at the downstream side of a 20-amp breaker exceeding the instantaneous setting of the upstream sub-distribution breaker, only the specific emergency or legally required standby branch distribution panel and branch breakers would be subject to a complete outage. The main emergency and standby electrical distribution system could remain functional to feed the other emergency and standby branches. In this situation, only the main gear would be subject to the expense of the ANSI power breaker, switchgear construction and the extra space required for the larger gear.

Figures 1 and 2 represent an 800-amp ANSI power breaker with the instantaneous setting turned off from either the generator or the normal power feeding into an 800-amp automatic transfer switch. The generator main breaker may not have to be an ANSI power breaker with the instantaneous portion turned off, if the generator cannot provide 8,000 amps of fault current. The amount of fault current that can be provided by the generator will be based on the reactance of the alternator in the generator. If the fault current at the generator main breaker is less than 8,000 amps, a standard breaker with the instantaneous setting of 10 times the continuous current can be utilized.

FIGURE 1

In Fig. 1, the 800-amp breaker feeds an 800-amp automatic transfer switch that then feeds an 800-amp main lug-only (MLO) panel filled with 225-amp breakers. These 225-amp breakers feed 225-amp MLO panels filled with 20-amp breakers that in turn feed 20-amp, 277-volt emergency lighting circuits. FIGURE 2

Fig. 2 also illustrates an 800-amp ANSI power breaker with no instantaneous setting and a 225-amp and a 20-amp breaker with instantaneous settings at 10 times the constant current. A fault on the downstream side of the 20-amp breaker of less than 2,250 amps will only take out this breaker. A fault of greater than 2,250 amps could take out both the 20- and 225-amp breakers. A fault on the downstream side of the 20-amp breaker should at no time take out the 800-amp breaker. A fault would limit the outage to one branch panel under the worst-case scenario. Depending on the available fault current in the system, the location of the fault on the 20-amp circuit and the nature of the fault (bolted or arcing), a fault downstream of the 20-amp breaker could very likely only take it out in this configuration.

Additionally, a coordination study could be mandated to tweak the system settings of the breakers to ensure that the maximum available fault current is coordinated with the instantaneous settings of the breakers. For example, if the calculated maximum available fault current at a 100-amp breaker was only 7,000 amps, and the overlap of the breakers occurred between 3,000 amps and 4,000 amps, then only the most severe fault currents would take out both breakers. A fault with any amount of impedance, other than a bolted short, in this case, would probably not rise to the level of the instantaneous portion of the 400-amp breaker. Fault current calculations are based on a worst-case scenario of a bolted short.

The reality is that most fault currents are based on an arcing short. In a bolted short, no additional impedance is introduced into the system, and the level of fault current will be based only on the impedance in the electrical distribution system to the point of the fault. These impedances include components like conductors and transformers. An arcing fault introduces additional impedance into the distribution system that will lower the actual fault current flowing through the short.

Based on a publication from one breaker manufacturer, 80% of faults are what would be considered low to very low arcing faults that would produce about 10% of the fault current seen from a bolted short. The publication also indicated that about another 15% of faults would be considered low to medium faults that would produce about 30% of the fault current seen from a bolted short. This indicates that a vast majority of faults are nowhere near the worst-case scenario from a fault current calculation simulating a bolted short circuit.