How to apply NFPA 110-2016
Consulting engineers who specify emergency power equipment understand that installations for mission critical facilities, such as hospitals and data centers, are required to comply with NFPA 110: Standard for Emergency and Standby Power Systems in conjunction with NFPA 70: National Electrical Code.
- Explore changes to NFPA 110: Standard for Emergency and Standby Power Systems.
- Define the important aspects of changes to this electrical code.
- Understand the importance of emergency and standby power systems.
NFPA 110: Standard for Emergency and Standby Power Systems covers requirements for emergency and standby power systems whose reliable operation is necessary, in varying degrees, to ensure the safety of human beings.
Updates to NFPA 110 occur on a 3-year cycle, like many of the NFPA codes. A number of changes were made in the 2016 edition. In general, these changes were incremental: new requirements were added, old requirements were modified, and some were clarified.
Of particular interest are changes to Chapter 8: Routine Maintenance and Operational Testing, as those changes affect all installations covered by NFPA 110. A section on record keeping was added in 2016.
A few definitions are important in understanding the changes to NFPA 110. These two definitions are fundamental to NFPA 110, as they describe the components of the system to which the code applies:
- Emergency power supply (EPS): The source of electric power. In practice, the EPS is typically the generation system. In paralleled systems, the EPS includes the generators and the paralleling gear (see Section 3.3.3).
- Emergency power supply system (EPSS): The EPS, together with its distribution, conductors, controls and monitoring devices, and transfer switches. The EPSS extends from the generators to the load contacts of the transfer switches (Section 3.3.4).
Two other definitions necessary to address the 2016 changes are:
- Level 1 systems are those whose failure could result in death or serious injury (Section 4.4.1). Examples of loads served by Level 1 systems are egress lighting, fire alarm systems, and fire pumps.
- Level 2 systems are those whose failure is less critical to human life or safety (Section 4.4.2). Annex A: Explanatory Material adds that Level 2 systems typically serve loads whose loss could create hazards or hamper rescue efforts. Examples of loads served by Level 2 systems are heating and cooling systems, sewage-disposal equipment, and certain industrial processes.
Section 1.1.3(6) states that NFPA 110 covers only Level 1 and Level 2 systems, whose operation is necessary to protect human beings. Requirements sometimes vary depending on the hazard should the equipment fail to provide power when necessary. No further levels are defined, as NFPA 110 does not cover systems that are less critical than Level 2.
The changes between the 2016 edition and the previous editions of NFPA 110 can be described as incremental. Changes affecting the design of standby power systems, which will affect only new systems installed under the 2016 edition, include new requirements for fuel systems. The most extensive changes to the code are additions to Chapter 8, describing routine testing and maintenance requirements, and will affect both new and existing systems. These additions include a new section on generation and retention of testing and maintenance records. A new requirement describes installation acceptance testing requirements for paralleling gear. Some of these changes are implemented in the code itself and are directly enforceable. Other changes are clarifications included in the explanatory material, which provide information regarding the intent of the code.
Automatically actuated valves in fuel systems
New to the 2016 edition is Section 7.9.13, which prohibits automatically actuated valves in the fuel supply or return lines. The intent is to avoid inadvertent interruption of the fuel supply by unintended or improper operation of a valve or its controls. Automatic control of fuel valves adds a layer of complexity to the fuel system and creates a single point of failure for the EPS.
There is no corresponding prohibition of manually operated valves. These valves are typically used to isolate portions of the fuel, allowing for continued operation of the system when one of its components requires maintenance. Section 7.9.11, unchanged from the previous edition, requires that manual valves provide a visual indication of their status. There is no direct requirement that manual valves be supervised to ensure that they are left in the open position after maintenance activities.
Clarification of requirements for multiuse fuel tanks
Section 5.5.1 prohibits using the fuel supply of a Level 1 system for any other purpose. In both editions, this prohibition is amended in paragraph 184.108.40.206, which permits the fuel supply to serve other systems, provided that other systems can't draw down the fuel level to a point that would compromise the continued operation of the EPSS.
In the 2016 edition, additional language permits passive features designed to ensure that an appropriate level of fuel is available for the EPSS. The 2013 requirement can be read to require that the fuel-use rate of other systems must be so low that it can't materially affect the amount of fuel available to the EPSS. The additional language in the 2016 edition permits other measures to limit the fuel drawdown. An example of such a passive measure might be the installation of the second fuel intake at a high level in the tank, so that other systems would lose their fuel supply before the continued operation of the EPSS was compromised.
Section 220.127.116.11 specifically permits passive features to prevent other systems from endangering the fuel supply to the standby power system. It doesn't mention active features, such as automatically operated valves, for that purpose. Presumably, an automatically operated valve serving a second system wouldn't meet the intent of the code because such a valve may fail in the open position as easily as it might fail closed, jeopardizing the fuel supply to the standby power system.
Interlocks for transfer switches
Section 6.2.3 requires a method of preventing inadvertent interconnection between any two separate power sources, either by mechanical interlocks or by some other means approved by the authority having jurisdiction (AHJ). The intent of this requirement is to protect the EPSS from damage due to connection to an unsynchronized power system. Historically, some AHJs have interpreted this requirement as a general prohibition against any interconnection between sources, whether inadvertent or intentional. While Section 6.2.3 remains unchanged in the 2016 edition, a new entry in Annex A: Explanatory Material clarifies that intentional interconnections are not strictly prohibited by the code and may be applied with the approval of the AHJ.
The most common type of transfer switch, an "open-transition" switch, briefly disconnects the load from both sources when it operates, even when switching between two live sources. A transfer from the alternate source to the normal source, as typically happens after the normal source is restored to service, results in a very short outage on the order of roughly 1/10 second.
This disturbance on the electrical system can be ameliorated with a "closed-transition" switch, which briefly connects the two sources before releasing its contact with the alternate source, typically for less than 1/10 second. Switch controls monitor the phase angle of each system and initiate the transition when the two sources are quite close to synchronization-when the phase angle between them is nearly zero and changing slowly.
A "soft-loading" system connects the two sources in parallel and controls the power output of the local source to smoothly transition the load from one source to another. When the load has been fully transferred to the destination source, the other source is disconnected. A soft-loading system can provide a very smooth transition between live sources, at the cost of increased complexity. A soft-loading system is a bona fide paralleling system, with all its attendant operational requirements. Such a system is normally applied where parallel operation with the utility is required for some other purpose, such as peak shaving, demand response, or to facilitate load testing of large systems.
Generally, closed-transition switches and soft-loading systems must be approved by the electric utility as well as by the AHJ. Some utilities will require complex protection systems before for even the briefest interconnection, while others will be more permissive. Designers should work with both AHJs and utilities to determine requirements for closed-transition switches or soft-loading systems before applying them.
Verification of paralleled systems
For systems consisting of paralleled EPSs, Section 18.104.22.168.3 requires that the control functions for paralleling generators and for load shedding must be verified against the system design documentation. This requirement is new to the 2016 edition. The previous addition did not address load shedding, and for paralleled systems, required only that load testing be performed with the quantity of EPSs intended to operate simultaneously.
At first look, this addition brings a layer of complexity to initial acceptance testing. Paralleling and load-shedding strategies can be complicated, with multiple contingencies. A thorough test of those functions will involve selectively disabling individual EPSs to simulate unit failure, and varying system loads to trigger load-shedding functions. Thorough testing of all functions, under all possible failure conditions and a reasonable number of load conditions, could become onerous for large systems.
In practice, though, this new requirement will have limited effect. Typical industry practice calls for extensive start-up testing of parallel systems, and most systems are commissioned by an independent authority. To a degree, this requirement codifies existing industry practice.
This requirement applies only to the initial acceptance testing of the EPS. Chapter 8: Routine Maintenance and Operational Testing contains a new provision, described below, calling for verification that the paralleling gear controls "will operate as intended."
Changes to Chapter 8
A number of changes were made to Chapter 8: Routine Maintenance and Operational Testing. These changes will impact all systems covered by NFPA 110. Section 1.3: Application states that the testing and maintenance requirements of Chapter 8 apply to existing systems as well as new systems, while the remainder of the code applies only to new systems.
Owners of systems that require regular certifications, such as hospital systems, can expect to see their system testing and maintenance records inspected by their accrediting agencies, once those agencies adopt the 2016 edition.
Paralleling gear testing and maintenance
The 2016 edition adds a requirement for paralleling gear in Section 8.3.5(5), which requires verification that "system controls will operate as intended." No such requirement was included in previous editions. This requirement is the companion to Section 22.214.171.124.3, described above, which requires verification of system controls, including paralleling and load shedding, during initial acceptance testing.
An exhaustive test of system control functions would require a considerable effort and would be disruptive to the operation of an existing facility. The operational testing requirements of Chapter 8 generally are designed to minimize the operational impact on an existing facility. This new requirement offers no specifics. In practice, it may be as simple as observing that the system is capable of bringing all units online in parallel, with selective load shedding demonstrated by taking a unit out of service while serving a simulated load. It remains to be seen how AHJs will interpret this requirement, and how it will be enforced.