Clean agent fire suppression for mission critical facilities

Clean agent systems have changed over the years due to a variety of reasons, including safety and product updates. Know the design parameters and codes and standards that dictate the specification of clean agent systems.
By M. Lee Draper III, PE, Koffel Associates, Columbia, Md. May 22, 2017

This article is peer-reviewed.Learning objectives:

  • Know the codes and standards that guide fire protection engineers when selecting clean agent fire suppression systems.
  • Understand the reasons clean agent systems might be specified in a building.
  • Learn which mission critical facilities require clean agent fire suppression systems.

In the 30 years since the signing of the Montreal Protocol of 1987 ended the viability of halon 1301 as a fire suppression agent, a concerted effort to develop a replacement clean suppression agent has been underway. The struggle to develop a halon alternative led multiple chemical corporations, such as 3M and DuPont, in many different directions, leaving the codes and standards field to develop guidelines for the use of these newly developed agents.

NFPA 2001: Standard on Clean Agent Fire Extinguishing Systems was first released in 1994 as a response to industry and governmental calls for clean agent selection, usage, and installation guidelines. Rather than develop a standard for each separate type of agent in development, NFPA authored the standard in such a way that it would apply to all clean agents. NFPA 2001 identifies clean agents as “electrically nonconducting, volatile, or gaseous fire extinguishants that do not leave a residue upon evaporation.”

These properties make clean agents especially useful for fire protection of data centers and other mission critical facilities, along with archive-storage and other special-purpose facilities in which water-based fire protection systems could damage room contents or cause major business disruption.

Table 1: Some common clean agents and their chemical makeups. All graphics courtesy: Koffel AssociatesWhile clean agents can be used in local-application scenarios, the vast majority of clean agent systems are total-flooding systems. In a total-flooding system, the agent is dispersed throughout the entire volume of the space being protected to ensure that a minimum agent concentration is achieved throughout that volume, thus extinguishing the flame. Popular from the 1960s until the halt of their production in 1994, halons are extremely efficient at extinguishing fire and relatively cost-effective to produce and employ. However, they have been shown to negatively impact the environment. Thus, the environmental impacts of replacement agents are a primary concern and are measured using three objective criteria:

  • Global warming potential
  • Ozone-depletion potential
  • Atmospheric lifetime.

Due to legislative measures driven largely by these environmental concerns, perfluorocarbons and hydrochlorofluorocarbons are being rapidly phased out of use, leaving halocarbon compounds and inert gases as the current clean agent types of choice. Looking beyond environmental impacts, new clean agents also are judged based on toxicity, their decomposition when released in a fire environment, and extinguishment effectiveness per mass, as well as cost-effectiveness and marketing capabilities. From a safety perspective, any agent included in NFPA 2001 has been evaluated in a manner equivalent to the process used by the U.S. Environmental Protection Agency’s Significant New Alternatives Policy (SNAP) program for total-flooding agents.

As previously noted, halocarbon compounds and inert gases make up the majority of the current total-flooding clean agent market. While they both adhere to the general definition of a clean agent, they achieve flame extinguishment in different ways and have unique benefits and disadvantages. Halocarbons extinguish flames by reducing flame temperature to the point that a flame can no longer be sustained. Common halocarbons include hydrofluorocarbons, such as HFC-125 and HFC-227ea, and perfluoroketones, such as FK-5-1-12. Inert gases and combinations thereof extinguish flames by reducing the ambient air temperature and/or decreasing the oxygen level in the hazard volume below minimum levels required to sustain a flame. Some popular inert gas clean agents include IG-01, IG-55, and IG-541 (see Table 1). Though outside the scope of NFPA 2001, carbon dioxide systems can sometimes be an option in mission critical facilities.

Figure 1: This close-up is of a two-tank FK-5-1-12 manifold system installed in a small data center room. Only one cylinder is active at a time. Note the keyed switch to manually change the active cylinder (in green in the photograph). How to select a clean agent

With the numerous clean agent options available to the consulting engineer, it is prudent to consider a number of different factors when specifying a clean agent for a mission critical facility. One of the most important factors is cost. Most available clean agents and their required nozzles are highly proprietary; therefore, finding reliable pricing information is almost always a matter of contacting a sales representative for the agent in question.

In addition, the quantity of agent required to achieve the desired concentration will differ between agents, such that a price-per-pound analysis is not always a useful comparison. Again, contacting a sales representative is typically the best way to get an accurate estimated cost.

Additional considerations include:

  • The type of hazard being protected
  • Local availability of the agent
  • Available space for agent cylinders
  • Presence of an existing halon system.

Some clean agents are designed with the intent to extinguish deep-seated Class C fires (i.e., electrical), such as those in a server rack, while other agents have been developed to protect Class A fires consuming solid fuels (i.e., paper and plastic). Testing has shown that the discharge of some clean agent systems can have a negative effect on the performance of hard disk drives in data center applications. The probability of hard drive problems is higher in some situations than others, but as a general rule, halocarbon systems tend to cause less interference with disk drives than inert gas systems. If hard drive performance is a primary concern in the facility being protected, further analysis to determine the risks involved may be required.

Availability of the desired clean agent can be a major selection factor in remote locations, such as mission critical facilities on military bases or in climates with extreme weather or security-related access concerns. The ability to replenish agent supplies can also become a major concern if the space is expected to be subject to regular discharges or if regulations prevent a facility from continuing operations until the clean agent suppression system is recharged and active.

Figure 2: Emergency halon-release switches are shown. Note that the switches are labeled to identify the room served.

A variety of agent cylinder pressures are used depending on the manufacturer; however, most agents require substantial floor space for storing agent-containing cylinders. If minimizing protection downtime following a discharge is desired, most clean agent systems can be designed with redundant cylinder banks connected by a manifold. The manifold is equipped with an integral bank-selection switch, either manual or automatic in actuation, which allows the redundant cylinder bank to be used as a supply for future discharges following discharge of the main bank. Essentially, such an arrangement allows the system to remain operational while waiting for the discharged cylinder bank to be replenished.

Although sometimes overlooked, the presence of an existing halon system can be a significant factor in the clean agent selection process. An analysis of the existing system is always recommended, as the size, type, and location of piping and halon cylinders can play an integral role in determining the level of renovation required to replace a halon system. In some increasingly rare cases, upgrading a halon system to a halocarbon system can be as simple as swapping agent cylinders and discharge nozzles and adding new control and detection equipment. It is always pertinent to discuss the goals and limitations of the project with the client, as many benefits of some clean agents, such as environmental friendliness or space savings, may add undesirable costs or complications.

Construction considerations

Regardless of the agent selected, the integrity of the room needs to be verified to ensure that the minimum-required concentration of agent, as dictated by NFPA 2001, is maintained for an adequate amount of time. In extreme cases, such as in construction types with very little anticipated leakage, it may be necessary to provide overpressurization vents such that the increased pressure caused by release and expansion of the agent does not damage the structure of the space being protected.

While not always a simple step-by-step process, the following outlines typical steps taken when designing a clean agent suppression system. Once a clean agent is selected, the minimum design concentration is determined using Chapter 5 of NFPA 2001. The minimum design concentration is based on a number of empirically determined extinguishing concentrations that are multiplied by various safety factors depending on the class of fuel in the space being protected. Also addressed are inerting concentrations, which must be employed when explosion or reflash concerns are present. The minimum design concentration is then used along with other variables, including the volume of the space to be protected and the specific volume of the clean agent being used, to determine the total-flooding quantity of clean agent.

Chapter 5 also presents the concept of design factors, which are used to increase the quantity of agent specified so as to compensate for any special conditions that would affect the extinguishing efficiency of the agent. Notable conditions requiring the use of design factors include altitude, the presence of tees, effects of unclosable openings, control of acid gases, and the potential of a hazard for reflash or explosion. The annex of the standard is an invaluable resource that provides a large amount of information regarding conditions that could result in the use of design factors.

Figure 3: A close-up of the existing halon cylinder bank is shown in a data center. No reserve tanks containing halon agent have been provided. The maximum amount of time that a system can take to achieve 95% of the minimum design concentration is dependent upon the type of agent used and the classification of the hazard. It ranges from 10 to 120 seconds. In any total-flooding clean agent system, 85% of the previously calculated adjusted minimum design concentration is required to be held at the highest height of protected content within the hazard for a period of 10 minutes—or for a time period sufficient to allow for response by trained personnel.

Finally, the routing of piping and nozzle choice all play integral parts in the clean agent system design process. Most clean agent distributors are more than willing to assist engineers in the design of a clean agent system. This support is invaluable to the designer, as distributors are likely to have specialized computer modeling programs that can help ensure a successful design.

Although less common in mission critical applications, NFPA 2001 does provide guidance for local-application systems in Chapter 6. Most of the same safety precautions that apply to total-flooding systems also apply to local-application systems, although the extent and location of the hazard arguably play a more central role in the design of local-application systems. Generally speaking, NFPA 2001 provides direction for design quantity, nozzle selection, and discharge time, although the standard relies heavily on listings and approvals for much of the design and installation process.

The search for a halon replacement has been underway for the past 30 years, though an ideal direct-replacement compound has yet to be found. Concurrently, the alternative fire suppression industry has transformed itself to become one of the most dynamic within the field of fire protection engineering. Fortunately, invaluable resources abound, and the consulting-specifying engineer can rely on excellent documents for guidance, such as NFPA 2001, the NFPA Fire Protection Handbook, and the Society of Fire Protection Engineers Handbook of Fire Protection Engineering.


M. Lee Draper III is a registered fire protection engineer with Koffel Associates