Designing for selective coordination

By following basic guidelines, both fusible and circuit breaker systems may be selectively coordinated without the engineer ever having to perform a detailed selective coordination study.

By Kenneth L. Lovorn, PE, Lovorn Engineering Assocs., Pittsburgh January 24, 2012

In the competitive environment of today’s engineering world, saving every possible design hour is the goal of nearly every engineering firm. One of those ways has been to require the contractor to provide the selective coordination study (SCS) as a part of its construction bid instead of including it in the engineering design contract. While the actual cost of this study does not change no matter who provides it, it is a small portion of the electrical construction budget.

While selective coordination is very important for maintaining service continuity in facilities, the National Electrical Code (NEC) mandates selective coordination for only a few vital systems such as emergency and legally required standby systems. Specifically, Article 700.27 and 701.27 state, “overcurrent devices shall be selectively coordinated with all supply side overcurrent protective devices.” This does not say anything about coordination down to 0.1 sec, 0.5 sec, 1 sec, 0.01 sec, or any other arbitrary point; it says, “shall be selectively coordinated.”

The issue of coordination down to only 0.1 sec creates a potentially life-threatening situation in which all power to a hospital may be lost due to a short circuit on a 20-amp branch circuit. With failure of power to infusion pumps, ventilators, and other life-sustaining equipment, patients could die just because an electrical distribution system was not fully, selectively coordinated. (Note: The author has seen two instances in hospital situations in which the lack of a selectively coordinated system resulted in total power loss to large sections of the hospital, jeopardizing patients. The description of one of these instances was published in Consulting-Specifying Engineer, January 2010, page 24.)

At the 2011 NFPA annual meeting, there was a floor debate concerning adoption of the selective coordination requirements in the 2012 NFPA 99 Healthcare Facilities Code. The debate concerned the issues surrounding full coordination (for the full range of overcurrents) or partial coordination (down to only 0.1 sec). While the NEC Technical Committees responsible for selective coordination requirements in the NEC support full coordination and strongly opposed partial coordination for several NEC cycles, a Technical Committee of NFPA 99 passed a requirement for only partial coordination. During the annual meeting floor debate, some of the proponents of partial coordination contended that selective coordination is too restrictive and there are no instances of a lack of coordination ever having been a problem (jeopardized life safety). The floor vote was to retain the NFPA 99 partial coordination requirement. (Note: This author is opposed to partial coordination for systems powering life safety related loads. The engineer’s paramount focus should be the safety, health, and welfare of the public, which, for life safety loads, means selective coordination for the full range of overcurrents.)

Because selective coordination is required, what does the engineer do to assure that the electrical distribution system that he designs can actually be selectively coordinated (full coordination)? Just because the engineer specified that the contractor is responsible for the SCS does not mean that the contractor or his coordination study subcontractor can actually selectively coordinate the system. The following is a recipe for designing an electrical distribution system that can be selectively coordinated without having to perform a SCS.


There are three cases for designing a system that can be selectively coordinated:

  • All fused devices in the distribution system  
  • All circuit breakers in the distribution system
  • A mixture of fused devices and circuit breakers.

Fuses—For the case in which all devices are fused, the easiest method is to use the fuse manufacturer’s selectivity ratio guide (see Figure 1). Using this guide, in most cases the engineer needs to select the type of fuse and ampere rating that is required for the specific load or circuit. This includes considering conditions such as motor starting and degree of current-limitation desired or needed. It really depends on the fuse types selected, but in most cases typical system design layouts are easy to do with fuses and achieve full selective coordination. The engineer just needs to ensure that the ampere rating ratios of the designed/installed fuses are equal to or greater than the published ratios.

For example, if the upstream device is a 1200-amp, KRP-C Class L fuse and the downstream device is a FRS-R Class RK5, dual-element fuse, the ratio from the guide is 4:1. That means that the maximum RK5 fuse downstream from the 1200-amp fuse would be 300 amps (1200/4 = 300). However, a better alternative for the downstream fuse would be a LPJ Class J dual-element fuse with a ratio of 2:1 with the KRP-C Class L fuse. This means then the 1200-amp Class L fuse will selectively coordinate with a 600-amp (or less) LPJ Class J fuse. For each different condition, the engineer can use the ratio from the guide and determine the size of the fused switch and the type of fuse. When the contractor runs the coordination study, he can select fuses that will coordinate, without concern that the system could not be selectively coordinated.

Circuit breakers—Selective coordination of a system depends on selecting circuit breakers that may be coordinated, that is, they have the correct trip functions and are sized so that it is possible to nest the tripping curves, not stack them on top of one another. In the sample single-line (Figure 1), we have selected six tiers of breakers for coordination:

  • T1 – The main service breaker or the main substation breaker in main distribution panel (MDP)
  • T2 – The feeder breakers located within the main service equipment, MDP
  • T3 – The main breaker within a distribution panel (DP)
  • T4 – The feeder breakers within a distribution panel, DP
  • T5 –The main breaker within a branch circuit panel
  • T6 – Branch breakers within a branch circuit panel.

Because the main problem for coordination of circuit breakers is in the instantaneous region, something must be done to address this problem. The simple way to permit a number of breakers in series to be selectively coordinated is to eliminate the instantaneous element in the main breaker. The main circuit breaker in MDP (T1) will have long-time, short-time, and ground fault trip elements. This main breaker will normally be a power air circuit breaker. Molded case breakers and insulated case circuit breakers may have a short-time delay but normally also include an instantaneous trip or override, so they would not qualify for this application. (See sidebar, “Why use air circuit breakers?” below.)

The feeder breakers in MDP (T2) will also have miscoordination issues with downstream equipment in the instantaneous region. The elimination of the instantaneous elements for the feeder breakers can be handled in the same manner as the main breaker, which turns the feeders into power air circuit breakers. Therefore, these breakers will have long-time, short-time, and ground fault trip elements, just like the main breaker. This will also provide two-level ground fault for hospital or other critical applications.  

A main disconnect (T3) is commonly located in distribution panels to permit local disconnecting means. In the past, these breakers could be non-automatic breakers (molded case switches) and would disappear from the selective coordination study. Due to UL changes and the way that these molded case switches are listed, they now have instantaneous elements as a part of the unit, which will interfere with coordination. Therefore, these main disconnect devices should be eliminated and a lockout/tagout procedure should be used for maintenance of this panel.

The feeder breakers within panel DP (T4) may be designed as insulated case breakers. Most manufacturers can provide a high-range instantaneous (override) on these breakers (approximately 15x, although one manufacturer has insulated case breakers with a high-range trip of 20x) so that they will fully coordinate with the downstream breakers. These breakers will still have the long-time and short-time tripping elements along with the high-range instantaneous element. Ground fault typically is not required at this distribution level.

A main breaker (T5) also is located in branch circuit panels for the same reason as that in the distribution panel and should be eliminated for the same reasons.

The branch circuit breakers (T6) would have their thermal magnetic trip units, which is no change from the typical electrical distribution system.

Sizing circuit breakers

Circuit  breakers will still be sized, just as they always have been, taking in to account the anticipated load, the starting inrush current, the various demand factors, and the desired spare capacity. The difference that must be considered during design is the ratio between the upstream and downstream breakers. In every case, the upstream breaker (with short-time delay) should be at least twice the ampacity of the downstream one. In our example, if the main breaker in MDP is 2,500 amps, then the largest feeder breaker in MDP should be no larger than 1,200 amps. With this 2:1 ratio, selective coordination is possible, but it will be challenging to fit these breakers between the upstream transformer primary fuse and the downstream distribution panel breakers. Optimally, ratios of 3:1 or greater will make coordination easier.

That raises the question: what do you do about having downstream breakers that are larger than the permitted 2:1 ratio? For instance, the facility has a large chiller with a maximum overcurrent protection (MOCP) of 1,500 amps and the main breaker is 2,500 amps. The actual current draw will likely be 600 to 700 amps, so there is plenty of capacity on the 2,500-amp main breaker, but the 1,500-amp feeder breaker will not selectively coordinate with the main. In this example, there is a 1,200-amp breaker feeding a distribution bus duct. The remaining breakers in the MDP are smaller, with trip ratios greater than 3:1.

The solution that will allow selective coordination between the main breaker and the downstream 1,200- and 1,500-amp breakers is to convert the MDP to a six-circuit rule switchboard. One main breaker would be rated at 1,200 amps to feed the bus duct, one breaker would be rated at 1,500 amps to feed the chiller, and the remaining main breaker would be rated at 800 amps and would serve the remaining MDP breakers in a distribution panel. For the sake of future loads, we would show at least two spaces suitable for 800-amp breakers.

Mixing fuses and circuit breakers

When the design begins to mix fuses and circuit breakers in the distribution system, designing for selective coordination is more difficult. The space requirements for this article do not allow for a detailed explanation of the analysis methods. IEEE papers provide analysis methods to determine coordination for mixing fuses and circuit breakers.

For a circuit breaker supplying fuses to be coordinated for a specific fault current, one IEEE paper method checks to determine if the peak current that the downstream fuse lets through for an available fault current is less than the peak let-through current that unlatches the upstream circuit breaker.

For fuses supplying circuit breakers, the curve plot generally is interpreted as noncoordinated when the fault current is greater than where the fuse curve intersects with the circuit breaker curve. Another IEEE paper discusses a selective coordination method where the let-through energy of a downstream current-limiting circuit breaker must be less than the upstream fuse’s melting energy.

If fuses are intended to be used in a series rated combination, due to low interrupting rating of circuit breakers, then selective coordination generally will not be achievable.

Interestingly enough, the use of fuses on the line side of a service transformer, which has a primary voltage of 4,160, 13,800, 23,000, or 35,000 V, are much easier to coordinate than circuit breakers in the same application. Primary fuses are designed to provide complete protection of the transformer damage curve while allowing the energizing current inrush without nuisance tripping. Even though it is much more adjustable, the circuit breaker interrupting curve does not lend itself completely to protecting the transformer, allowing for inrush current while coordinating with the downstream protective devices.

By following these simple guidelines, both fusible systems and circuit breaker systems may be selectively coordinated without the engineer ever having to perform a detailed selective coordination study. He can specify that the contractor is responsible for having the selective coordination study performed with the protective devices that are being provided for the project. And he can be confident that each of his designs may be selectively coordinated without a major redesign of the system.

Lovorn is president of Lovorn Engineering Assocs., Pittsburgh, and a member of the Consulting-Specifying Engineer Editorial Advisory Board.

Why selective coordination?

The NEC defines selective coordination as “Localization of an overcurrent condition to restrict outages to the circuit or equipment affected….” What this means to the user of a facility is that the failure of a small space heater or fractional horsepower motor will not open the feeder or main overcurrent protective device for the entire building. This may cause all work to stop and the building to be evacuated until the problem is located and removed and the main overcurrent protective device reset or replaced. In a building that has selective coordination, only the branch circuit overcurrent protective device opens for a fault on the branch circuit. Neither the feeder nor main overcurrent protective device will open.

Some engineers have taken the position that coordination down to 0.1 sec is all that is required for selective coordination. The result of partial coordination down to 0.1 sec is that the overcurrent devices will be coordinated only in the overload region and are permitted to be fully uncoordinated in the short circuit region. How can anyone believe that they have a selectively coordinated system if the overcurrent devices do not coordinate during a short circuit!?

Why use air circuit breakers?

When the facility is a hospital, data center, or other critical load that must have the highest power availability, the ability to substitute a replacement breaker for a failed unit is paramount. Power air circuit breakers are designed to be easily removed and reinstalled without interrupting power to the rest of the facility. Because of this, most mission critical facilities will use power air circuit breakers, which allow the engineer to design the distribution system for selective coordination, without an increase in cost or space requirements.

It is true that power air circuit breakers require the use of switchgear and cannot be installed within switchboard construction. However, the nominal increase in space requirements and cost is well justified when you are talking about the ability to design a safer, more reliable installation that can result in saving patients lives. Are we going to say that keeping the cost of the electrical distribution system down is worth the loss of two or three patients’ lives? Is any loss of life acceptable because the electrical distribution system is not as reliable as it can be?