How to select a clean agent fire suppression system
Fire protection engineers should understand which clean agent fire suppression system is most appropriate for a mission critical facility.
- Review the codes and standards that dictate the specification of clean agent fire suppression systems.
- Obtain an overview of fire suppression options available to engineers.
- Gain insights into specifying a clean agent system in a mission critical facility.
How do you protect mission critical assets that can be damaged more by the protection methods themselves than the fire event? Fire protection measures for mission critical facilities have changed over the years due to a variety of reasons, such as product development, environmental concerns, and occupant safety.
When protecting high-value mission critical assets, one should consider the code path as well as the protection options available. There are several codes and standards that dictate when total flooding gaseous clean agent systems can be installed and how these systems are installed.
Before going into detail, it is important to note the difference between a code and a standard. Codes identify when a system is required or can be used, and standards indicate how to install the systems. Typically, codes are adopted into law by jurisdictions, and standards are referenced by those codes for design and installation purposes. As an example, the International Building Code (IBC) may require the installation of a fire sprinkler system. The IBC then references a specific edition of NFPA 13: Standard for the Installation of Sprinkler Systems, which prescribes the requirements on how to design and install the sprinkler system.
Much like the sprinkler system example, codes are used to identify when a gaseous clean agent system is required or when it can be used for protection. This article will reference the most up-to-date published codes and standards for simplicity. Note, however, that many jurisdictions still reference earlier editions of codes and also adopt amendments to the model building codes, which must be taken into consideration before the start of any project. Although some codes do allow the substitution of gaseous clean agent systems for sprinkler systems with authority having jurisdiction approval, it is not recommended practice for most applications. The substitution is not recommended because sprinkler systems are primarily designed for life safety purposes whereas gaseous clean agent systems offer protection of assets as their main objective.
The most common model codes for fire protection provisions in the United States are the IBC and the International Fire Code (IFC) published by the International Code Council (ICC). In addition to ICC-based requirements, some jurisdictions may adopt NFPA 5000: Building Construction and Safety Code and/or NFPA 1: Fire Code.
Generally speaking, the IBC mirrors the requirements of the IFC, so the discussion will be based on code sections from the IBC. Section 202 of the IBC defines a clean agent as an “electrically nonconducting, volatile, or gaseous fire extinguishant that does not leave a residue upon evaporation.”
Section 904 of the IBC, entitled Alternate Automatic Fire-Extinguishing Systems, states that “Automatic fire-extinguishing systems installed as an alternative to the required automatic sprinkler systems of Section 903 shall be approved by the fire code official.” Gaseous clean agent systems are specifically noted in IBC Section 904.10. This section states: “Clean agent fire-extinguishing systems shall be installed, maintained, periodically inspected, and tested in accordance with NFPA 2001 and their listing.”
As you can see, the IBC allows for the use of gaseous clean agent systems as a substitute for sprinkler systems when approved by the fire code official. Again, this substitution should be carefully considered, as a sprinkler system is typically designed for life safety and building protection while a gaseous clean agent system is designed for asset or contents protection. Sprinkler systems are designed for extended suppression, with water-supply-duration requirements based on the hazard of contents plus the ability for the fire department to supplement supply to extend the duration. This extended duration gives the fire department ample time to respond to an event while the sprinkler system is continually operational. Gaseous clean agent systems, however, expel all of the gas at one time and are required to maintain a minimum concentration for only 10 minutes.
Much like the IBC, NFPA 5000 lists clean agent extinguishing systems as alternate systems. According to NFPA 5000 Section 55.5.1, “In any occupancy where the character of the fuel for fire is such that extinguishment or control of the fire is accomplished by a type of automatic extinguishing system in lieu of an automatic sprinkler system, such extinguishing systems shall be installed in accordance with the standard referenced in Table 55.5.1.” Table 55.5.1 references gaseous clean agent systems as a fire suppression system and NFPA 2001 as the installation standard.
While this article concentrates on common building code requirements within the United States, it is always important to evaluate local and or international adoptions to determine specific requirements. It also may be important to recognize the insurer of the equipment or property, as some insurance companies, such as FM Global, may have additional requirements.
NFPA 2001: Standard on Clean Agent Fire Extinguishing Systems regulates the requirements for components, system design, and application of systems (local or total flooding) and outlines inspection, testing, maintenance, and training requirements. This standard is used by industry professionals to ensure gaseous clean agent systems are being designed, installed, tested, and maintained.
There are other standards that may be referenced when doing work in jurisdictions or internationally. One such standard is ISO 14520-1:2015, published by the International Organization for Standardization. This standard covers the design, installation, inspection, testing, and maintenance of gaseous clean agent systems.
Finally, it is important to note that the standards require systems to be listed and/or approved by a third-party agency. To ensure that gaseous suppression systems comply with listings or approvals, it is important to follow the applicable design and calculation methods for the agent being specified. Unlike water-based systems, gaseous suppression requires manufacturer-specific criteria and software, which typically cannot be fully designed by most consulting or specifying engineers.
Gaseous clean agent systems, typically used to protect electronics, can be used to protect a wide variety of hazards. Some of these mission critical hazards include:
- Military and government facilities where loss of equipment and a long downtime can lead to lapses in defense systems
- Control centers for national air space
- Missile-guidance centers
- Airline flight-control servers
- Consumer-product credit card transaction computers
- Intelligence servers
- Satellite communication systems
- Research and development computing data hubs
- Engine test cells with accompanying data.
Types of clean agents
NFPA 2001 provides two main categories of gaseous clean agent systems. Inert gases, such as IG-541, extinguish fires by depleting oxygen, and halocarbon-based agents, such as HFC 227ea, extinguish fires on a molecular level in the presence of heat, fuel, and oxygen. Note that halocarbons disrupt the chemical reaction in the fire tetrahedron. Both types of systems have their pros and cons when protecting mission critical applications. When comparing systems, it is important to consider the cost to refill, shelf life, and environmental concerns.
Inert gas clean agent systems are readily suited for multiple hazard-enclosure applications, such as engine test cells or more than one server room in a building. One bank of agent cylinders can be used to protect multiple hazards in adjacent areas. This can result in less total agent, thereby minimizing overall cost. This is accomplished by using selector valves, separate control heads, and individual piping networks. When protecting multiple hazards with inert gas, maximum and minimum agent concentrations should be evaluated for each protected enclosure under normal system operation. Halocarbon agents typically are only capable of protecting single hazards with each cylinder or bank of cylinders.
The weight and required storage area of the suppression agent cylinders are typically smaller with halocarbon-based systems as compared with inert gas systems. Increased weight can have structural implications that should be accounted for during design. More inert gas cylinders will typically be required than when using halocarbon cylinders for protection of the same volume. This can be a driving factor on, for example, lease space and building-use planning. It is also an important consideration in retrofit applications in existing buildings, where large areas for cylinder storage are often difficult to find.
Cylinder pressures for halocarbon-based clean agents are typically lower than for inert gas agents. While the agent pressure drops drastically during release, there are still piping implications to be considered. Some inert gas systems require higher pressure-rated piping and fittings. The increase of room pressure is also greater with inert gases. Pressure relief venting of the protected enclosure is more common with inert gas systems than with halocarbon systems.
Most fire protection engineers fail to assign adequate importance to agent density. Agent density directly relates the ability for an agent to remain in an enclosure after discharge. The denser an agent is, the more it will migrate from the protected enclosure and result in lower retention times. Table 1 compares some of the more common agent’s densities to air.
While inert gases reduce oxygen to a level that humans can survive and fire cannot, the halocarbon-based agents extinguish fires while heat, fuel, and oxygen are present. NFPA 2001 and other standards require that the gaseous clean agent is maintained at a minimum concentration threshold within a room for the duration of protection; this is usually referred to as the “hold time.” The latest requirement is 10 minutes.
The hold time is intended to allow the gaseous clean agent to extinguish the fire and reduce the potential for reignition due to the presence of heat, fuel, and oxygen. To ensure that the required hold time is achieved, the room must be sealed as much as possible to reduce areas from which gaseous clean agents can leak. Sealing of the room is referred to as room integrity. Typically, walls in the protected area are continuous to the roof or floor/ceiling assembly above, or to a “hard” gypsum ceiling. Dampers or HVAC equipment controlled by the fire alarm system may be necessary to maintain the room integrity. Caulking to address holes and other leakage paths in walls is usually necessary. Finally, seals around doors equipped with automatic closers may be needed to minimize leakage.
During construction, room integrity is measured by administering what is commonly referred to as a “door fan test.” This test uses specialized equipment to pressurize the enclosure and measure leakage rates, which then can be translated into hold times.
In today’s design and construction culture, there is no such thing as a cookie-cutter room. When designing or specifying clean agent systems, the engineer needs to be extremely detailed with enclosure geometry. Nozzle spacing, coverage heights, and enclosure volume can drastically impact the overall cost and feasibility of a clean agent system. Similar to sprinklers, clean agents have manufacturer-specified areas of coverage for each nozzle and maximum/minimum protected heights.
Sometimes, it may be more economical to use two back-to-back 180-deg nozzles with a larger nozzle throw than a single 360-deg nozzle with shorter throws. While typical server rooms may not have nozzle-placement concerns, industrial applications with two or more layers of nozzle coverage in a room with high ceilings is not uncommon.
Conversely, some subfloor applications in a typical server room may require additional nozzles based on a low enclosure height. Enclosure volume should be field-verified before design and after installation to evaluate whether adequate agent concentration will be achieved. If the volume increases from the volume used in the design, then the system may underperform. If the volume decreases with regard to the design volume, the agent may create a hazardous environment for the occupants.
Another issue rarely addressed is the exposure limits to the occupants of the space. While one would hope occupants would egress from a protected enclosure during a fire, we still need to consider the implications of the agents being discharged into an occupied space. Table 2 compares design concentrations for several popular agents to exposure limits. Note that some agents exceed recommended exposure limits at design concentrations. This should be a crucial factor in not only determining a proper agent, but also driving the importance of predischarge alarms for occupant evacuation.
Early detection is imperative for all types of clean agents. The systems are most effective when fires are in the incipient stage. Using heat-sensing detectors for system activation are not industry standard due to the delayed detection and the resulting larger fire size. This is especially important with halocarbon-based agents where agent thermal decomposition can occur, resulting in hazardous agent byproducts. As a result, smoke-sensing detectors that identify products of combustion in the incipient stage are typically used in clean agent systems.
Air-sampling detection and spot-type smoke detection are commonplace in the clean agent realm. When using ultra-sensitive air-sampling detection, multiple alarm thresholds should be considered to prevent false discharge. When spot-type detection is employed, it is good practice to design a cross-zoned system that requires two detectors to alarm in order to initiate agent release.
When evaluating control panels for the supervision and activation of gaseous clean agent systems, there are two types of control systems to consider:
- Conventional zoned panels/systems
- Addressable panels/systems.
Without diving deep into fire alarm system design concepts, the basic difference between the system types is the reporting of the initiating device location. A control panel for a zoned system will identify the general area (zone) in which an initiating device activation is received. A zone is defined by one circuit. So, if the desire is to have the system only activate when two detectors in the same area activate, you will need two or more circuits/zones in the same area.
An addressable system, however, will tell you exactly which device has activated. The panel can tell which device on a circuit is activated via addressable modules in the circuit. This allows cross-zoning detection with only one circuit. With the advancement of technology, it is likely that an addressable system will be installed for new designs, as the cost difference between an addressable and conventional system is negligible. Conventional panels are still widely used in single-enclosure applications where simplicity is preferred. Several manufacturers have more than one releasing circuit available on their conventional control panels. This allows for multi-enclosure protection or pre-action sprinkler solenoid actuation in conjunction with agent release.
Perhaps the most important consideration when choosing a control panel is the listing of the control panel. The listing must include fire suppression agent releasing and must be cross-listed with the solenoid used to release the gaseous clean agent into the piping network. This is an NFPA 2001 requirement to ensure the control panel and gaseous clean agent system communicate seamlessly.
Many systems in fire protection interact with one field or trade. For example, a sprinkler system is installed by a sprinkler contractor and a fire alarm system is installed by an electrician and fire alarm contractor. Clean agent systems require the interaction of multiple trades including detection/controls, piping, mechanical shutdowns, and room integrity, as outlined above. As a result, having a strong specification is imperative for coordination. In addition to the design, third-party testing is recommended.