Fire Protection In Mission Critical Facilities
Note that evacuating the protected space is a general recommendation for any potential fire event.
- By Lance D. Harry, P.E.
- Aug 25, 2008
The use of fire protection technology in “mission critical” facilities has long been a standard for most engineering, IT, and facilities professionals. Whether it be as simple as portable extinguishers or as complex as high sensitivity smoke detection coupled with clean agent suppression systems, some sort of fire protection is a must.
Historically speaking, mission critical facilities (MCFs) have garnered an elevated level of fire protection awareness because of two factors:
• MCFs involve a collection of highvalue assets, typically electronics, that have significant monetary value but are often even more valuable operationally. The cost of replacement and downtime associated with damage to these assets can be astronomical.
• MCFs inherently involve a greater level of risk than most commercial space because of the presence of both a constant ignition source (electricity) and a plentiful supply of fuel (generally, plastics, as in printed circuit boards (PCBs)).
The MCF itself can take on a number of different identities. The most common example in today’s business environment is the data center. Data centers are often the hub of a business’ operations, handling anything from internal communications (e-mail) to vendor/customer data, external order handling, and financial transaction processing. Data centers are found in all segments of the business landscape. Financial, telecommunications, and large manufacturing most commonly have numerous and extensive data center assets. Other examples of MCFs are network control centers, process control rooms, laboratory facilities, power generation facilities, testing environments, etc. The MCF is really defined by the company and owner. If the assets and operations are of particular value, it can certainly be deemed mission critical.
Fire protection in these facilities can be varied. Traditionally, systems have been sought that provide the greatest level of protection for the least cost. Level of protection can be loosely evaluated through two elements: extinguishing fire rapidly and effectively, and also minimizing associated damage to the protected assets. It is important to realize that, in most cases, MCF protection involves both structural protection (generally recommended throughout a given building) and asset protection (supplemental protection for the high-value asset).
In the ’60s, ’70s, and ’80s, halon compounds were prevalently used for the MCF application. Halon 1301 was the most common clean agent of the era. Some would argue halon 1301 was used more liberally than it should have been.However, its effectiveness in extinguishing fire and minimizing damage to the protected space was exemplary. Unfortunately, halon compounds contain either bromine or chlorine as one of their primary elements. Both elements are known ozone depletion contributors and were thus targeted by the Montreal Protocol, originally signed in 1987.The Montreal Protocol banned the production of ozone-depleting compounds, including halons, in most developed countries. Halon is still available today in recycled form, however. For example, in the United States, it is legal to recharge existing halon 1301 systems with recycled halon purchased from third-party suppliers and recycling groups. In the countries of the European Union, however, halon systems are no longer permitted and are required to be removed from service.
Fire Protection Evaluation
Fire protection system selection is a complex process that can involve a number of different entities, including internal company resources (financial, engineering, IT), outside consultants, insurance representatives, architects/ engineers, and the local authority having jurisdiction (AHJ). It is first important to understand the difference between code-required protection systems (most often, sprinkler systems) and supplemental asset protection.
Evaluating the need for supplemental protection should start with an analysis of the facility in question. Several different factors should be taken into consideration:
Analysis of the facility. The physical characteristics of the facility should be noted. In new construction, particular requirements often can be included, such as tightly sealed windows/doors and venting if necessary (see below for a discussion of various clean agents). In older facilities being outfitted with a suppression system, room integrity is of critical importance. If a room is overly “leaky” and will likely not achieve a level of integrity that is desired (or is too costly to do so), other strategies may need to be sought. Most local fire protection contractors can assist in a determination of room integrity for a given facility.
Hazard analysis.Understand the fire hazards within the space. It is most important to differentiate between class A (common combustibles such as plastics and fabrics) and class B (flammable liquids). Most MCFs contain only class A hazards, but a thorough review should be performed. This is also a good opportunity to investigate and evaluate the desire for “Emergency Power Off,” or EPO.Most fire protection professionals and codes recommend power be shut off to the electronic assets in the protected space prior to discharge of the system. In a facility where power is not shut off prior to discharge,some additional design considerations may need to be implemented. Consult your AHJ and system manufacturers for additional guidance in this area.
Overall risk assessment. A thorough risk assessment comes down to an understanding of potential harm versus likelihood of an incident occurring. It is standard to use a risk matrix to assist in the determination of how a given facility may be categorized from a risk perspective (see Table 1).
Fire in an MCF is generally viewed as a critical or serious impact, while the probability of occurrence is either low or medium. Even with this, the risk category would either be moderate or high. Further, risk professionals agree that moderate/high-risk categories are areas that should be addressed as a priority.
Once a determination has been made that supplemental fire protection is appropriate, it is natural to evaluate the two most common strategies with respect to cost versus levels of protection:
Pre-action sprinkler systems. A pre-action sprinkler system is water based and incorporates several operations in order to minimize the risk associated with accidental discharge and water damage. Three steps typically occur during a pre-action sprinkler event:
1. Spot type smoke detectors sense a combustion event.An alarm is triggered and sent to the system control panel, which interprets the alarm and then sends a signal to a solenoid-controlled valve, opening the valve. The open valve allows water to fill the pipe to a second valve.
2.With a combustion event progressing, heat builds at the ceiling of the protected space.While most sprinkler heads are “rated” for a temperature of between 135 degrees F and 165 degrees F (57 degrees C to 74 degrees C), that does not mean when the temperature around the head reaches that number, the head will discharge water. Sprinkler heads utilize an element of some sort (typically, a glass bulb or fusible metal link) to sense temperature. These devices are very reliable but take time to heat to the appropriate temperature. It is not uncommon for the surrounding space to be as much as 500 degrees F (260 degrees C) before the element breaks. This effect is commonly referred to as thermal lag.
3. With at least one sprinkler head “open,” the piping first discharges compressed air or nitrogen present in the pipe. The loss of pressure in the piping causes a pressure switch to be triggered and opens the second in-line valve.Water is then allowed to flow freely into the piping network and discharged from any sprinkler head whose element has been broken.
It is worth noting that even in a pre-action system, a significant level of heat is required to activate the system.While the level of risk associated with false discharges or damaged heads is significantly minimized in this strategy, the level of protection to the assets is still based on heat detection and water discharge. Use of water-based systems should be considered as part of the overall risk profile within the MCF.
Waterless/clean agent systems. Clean agent systems typically require two actions to occur prior to an actual discharge:
1. Spot smoke detectors or other smoke detection devices provide a signal to the suppression system control panel. Two separate alarms (often, one photoelectric type and one ionization type) are required prior to a countdown to discharge of the agent. This is commonly referred to as a cross-zoned detection strategy.
2. The control panel initiates a timed countdown, typically 30 seconds, before signaling the agent storage container to discharge the agent into the protected space. A delay is provided while occupants can vacate the facility and other preparations (mechanical door closures, automatic dampers, etc.) can be made prior to discharge. Note that evacuating the protected space is a general recommendation for any potential fire event and is not due to the discharge of the agent. Agents discussed below are all safe to potential occupants when designed in accordance with U.S. and international standards.
Many strategies involve the use of both a water-based system (for structural protection) and a waterless system (for asset protection) in the same protected space. This strategy provides the most complete fire protection scenario in any given facility. It follows that this also requires the greatest financial investment by the end user.
For more information on the differences in protection levels provided by clean agents versus water-based systems, visit www.e1 .greatlakes.com/wfp/common/jsp/index.jsp. Click on the Technical Information link and then on the Comparison Testing link.
Waterless Fire Protection Systems
After the evaluation process above yields the need for waterless fire protection, additional choices and evaluation will be necessary. Myriad options are available in the three major system categories:
1. System controls. The control system is the “brain” of the fire protection system. Control panels are available in several varieties, including intelligent versus convenwww.tional and suppression-specific panels versus general fire alarm panels.
Conventional control systems involve the collection of multiple circuits or “zones” of protection.Various panels can accommodate different numbers of zones. Each zone is a circuit, with detection devices located on each circuit.When a detection device goes into alarm, the circuit is closed. The panel recognizes the signal and proceeds to the appropriate action (initiation of a suppression system, initiation of audible or visible alarms, etc.). Conventional systems are inherently labor intensive, requiring individual maintenance and testing for each device.Additionally, when in alarm or trouble, there is no way of determining which device has been triggered without investigating each individual device in the zone. Conventional control panels provide a base level of functionality and are the most cost effective options for a control system.
Intelligent systems provide state-of-theart controls for a given fire protection system. Intelligent systems allow the owner and servicing personnel to identify each circuit device in trouble or alarm, at the control panel,without surveying each individual device. Each device has a unique address on a given circuit, enabling this identification ability. Intelligent systems offer an enhanced level of functionality and will typically cost more than a conventional panel system.
Fire suppression panels are generally designed and installed specifically for the release of a fire suppression system (such as a sprinkler valve or clean agent container). Suppression panels are fully functional with various methods of detection and devices, such as manual pull stations, abort switches, monitoring switches, etc., and they come in both conventional and intelligent varieties, as mentioned above.
Fire alarm panels generally include a broader level of functionality than a suppression- releasing panel and include several key features, such as voice evacuation systems and ability to communicate with thirdparty central monitoring stations (typically referred to as “central station”). In some cases, panels may act as both a fire alarm panel and a fire suppression-releasing panel, but it is generally not the norm. Manufacturers of panels often make it difficult to interact with manufacturers of suppression systems for proprietary and commercial reasons. Consequently, it is most common to see both suppression-releasing panels and fire alarm panels within a given facility.
2. Hazard detection. Detection of a hazard within the protected space can come in a variety of forms, including water, heat, and smoke detection. For the purposes of this discussion, we will focus on the most common method of hazard detection when dealing with clean agent fire protection systems— smoke detection. There are several technologies available for the detection of smoke within an MCF.
Spot smoke detection is the most common variety of smoke detection available. Technology in this area has been greatly improved in just the past five to 10 years, making detectors more reliable and more sensitive when necessary and less apt to false alarms.While completely eliminating false alarms is virtually impossible, the latest technology significantly reduces this risk.
The two most common varieties of spot smoke detection are ionization and photoelectric:
• Ionization detection is based on the detection of smoke through a reduction in current across a charged surface. Smoke particles passing through an ionization chamber will attach themselves to ions in the chamber created by small amounts of alpha radiation and disrupt the current being generated. The electronics of the detector sense this drop in current and, when sufficient, will initiate an alarm condition. Ionization detectors will respond most quickly to large flaming fires, which produce smaller particles of combustion.
• The most common type of photoelectric detector uses a light beam source to detect how much light is being scattered, through the size and frequency of smoke particulates present.When the light is scattered by the particles sufficiently onto a photocell, a current is generated and an alarm condition results. Photoelectric detectors will respond most quickly to smaller, smoldering fires, which produce relatively larger particles of combustion.
Both technologies are commonly used around the world. Both are reliable and proven methods for spot smoke detection. In a cross-zoned strategy of detection, as mentioned above, a designer will typically use one of each type to create a rapid detection scenario for any potential fire hazard.
The second method of smoke detection used in MCFs is air sampling smoke detection or HSSD (high sensitivity smoke detection). HSSD systems are typically magnitudes more sensitive than conventional spot detectors for several reasons. First, HSSD uses small but powerful fans to draw air from the protected space through small “sampling points” and back to a central detection unit. This enables the detector to “see” the potential products of combustion substantially faster than with spot detectors, which rely on interior air currents to bring the smoke to the detector. Second, HSSD systems generally rely on technology that can discriminate between particle types, thus determining both particle size and number prior to an alarm condition. Finally, HSSD systems can be programmed to accommodate a variety of sensitivity requirements.
HSSD systems are applicable as a replacement for spot type detectors but are more commonly used as a supplemental level of early warning in a highly sensitive environment. Both spot type and HSSD type detection will interface with any type of panel discussed above. Spot detectors are recommended for initiation of a system discharge, while HSSD systems are recommended for early warning detection only but can be used to initiate a system discharge if desired.
3. Suppression systems. If the control system is the “brain”of a fire protection system, the suppression system is the muscle. The suppression system is deployed for the early extinguishment of a potentially catastrophic fire event in an operationally or monetarily critical space. The suppression system is made up of a multitude of components: valves, piping, nozzles, cylinders, and the suppression agent itself. It is important to recognize the importance of the agent, as well as the system as a whole, and how the agent is delivered to the protected space.We will address each of these topics independently:
As mentioned above, halon 1301 was the standard by which all others are compared. The extinguishment characteristics of halon are outstanding, and its safety for both people and the protected space is exceptional. However, its production ban has led the industry to a number of halon alternative agents that are available today.
(Note: Brand names are used below due to their universal recognition. Many different chemical formulas, names, and other designations can be used for each of these compounds. See the chart below for additional designations.)
FM-200® is the brand name given by a specific manufacturer for the compound HFC-227ea, or heptafluoropropane. FM- 200 is a hydro-fluorocarbon (HFC) compound and is the most widely used and recognized halon alternative on the market today. It has been accepted/approved, tested, installed, and extinguished fires around the world. FM-200 extinguishes fire through the absorption of heat and does not significantly reduce the oxygen concentration within the space. An FM-200 discharge does not leave a residue or harm people or equipment in the protected space. FM-200 is the most common choice for applications such as data centers, telecommunications facilities, record storage, cleanrooms, etc. The extinguishment times and storage space required are very similar to halon.Use concentrations for FM-200 are typically between 6.25 percent and 8 percent by volume. FE-13™ is also an HFC compound and is a lesser used, niche agent in the market today. FE-13 has several unique characteristics, including a very low boiling point (allowing it to vaporize when discharged at very low temperatures) and a high vapor pressure (allowing the discharge to be quite energetic). These two features make FE-13 a good choice for low-temperature applications (both storage and protected space), as well as particularly high ceiling applications. Use concentrations for FE-13 are typically around 20 percent by volume.
FE-25™ is a newly available halon alternative. FE-25 had long been used for unoccupied spaces but is now applicable to all traditional occupied space applications through the use of Physiologically Based Pharmacokinetic (PBPK) modeling. PBPK is a simulation methodology used to predict human response and bloodstream absorption in the exposure to these types of compounds. The two major standards organizations that establish guidance for use of these systems (ISO and NFPA) have both adopted the PBPK method for the determination of toxicology thresholds at which these agents may be used safely in an occupied space. Prior to the adoption of the PBPK model, FE-25 had a use concentration slightly above the first threshold of toxicology, called the No Adverse Effect Level (NOAEL). Presently, use concentrations are slightly below the thresholds determined using the PBPK methodology. The compound is also an HFC and extinguishes fire in much the same mechanism as both FM-200 and FE-13. Use concentration for FE-25 is typically between 8.0 percent and 9.0 percent by volume.
3M™ Novec™ 1230 Fire Protection Fluid is the most recently developed agent on the clean agent market.Novec 1230 fluid is a fluorine- based compound but not an HFC like FM-200,FE-13, and FE-25.Novec 1230 fluid extinguishes fire by the absorption of heat and thus does not reduce oxygen concentrations within the protected space, similar to all three HFC compounds mentioned above. The unique characteristic of Novec 1230 fluid is that it has both zero ozone depletion potential (again, similar to FM-200, FE-13, and FE-25) and an extremely short atmospheric lifetime, which contributes to its low global warming potential (GWP). HFC compounds typically have a moderate atmospheric lifetime measured in tens of years, while Novec 1230 has an atmospheric lifetime of approximately five days. In applications with end users who are particularly sensitive to environmental considerations, Novec 1230 is a good selection.Use concentration for Novec 1230 fluid is typically between 4.0 percent and 6.0 percent by volume.
Note: All of the agents listed above are considerably more environmentally friendly than halon 1301. Also, in most jurisdictions, clean agent fire suppression is considered a “non-emissive”use of these compounds and, as such, the end user evaluating the various options typically determines environmental considerations.
Argonite® is different from the above four agents in that it is a blend of naturally occurring gases, not a fluorine-based compound. Argonite is referred to as an inert gas compound. Argonite is a 50/50 blend of argon and nitrogen. Argonite extinguishes a fire through the reduction of oxygen in the protected space such that combustion can no longer sustain itself. This oxygen reduction does not, however, preclude the use of Argonite in an occupied space because the oxygen concentration is reduced to a point at which combustion cannot be sustained, but human safety is maintained for a short exposure period. Argonite systems require the employment of dedicated venting systems because the quantity of agent is significant, in order to reduce oxygen levels sufficiently.
Inergen® is also a blend of naturally occurring gases, similar to Argonite. Inergen is a blend of approximately 40 percent argon, 52 percent nitrogen, and 8 percent carbon dioxide. The added CO2 in Inergen serves one primary purpose: CO2 increases the breathing rate in humans. In a reduced oxygen atmosphere, it is helpful to increase breathing rates to take in more oxygen. Inergen is also subject to venting requirements as mentioned above for Argonite.
Typical use concentrations for both Argonite and Inergen are 35 percent to 40 percent by volume, resulting in an oxygen concentration of between 12 percent and 14 percent.
For additional details and comparisons among the agents discussed in this article, refer to NFPA 2001.
Agent delivery systems
Discussion of and evaluation of the various agents available today can be time consuming and exhausting. Many pages have and can be written on a comparison of the various agents.What is presented above is simply a brief introduction into the most common halon alternatives.
What is typically not discussed at great length is a comparison of the equally important delivery of the agent to the protected space. Traditionally, the fluorine-based agents are stored in steel, welded containers and are pressurized with nitrogen. The nitrogen provides added pressure/energy to the container in order to propel the compressed liquid through the piping network when discharged.Nitrogen is super-pressurized into the containers at either 360 psi (25 bar) or 600 psi (42 bar), depending on local standards and specific application. MCF applications will almost always use 360 psi (25 bar) containers.
In the case of halon, the nitrogen is not overly soluble in halon 1301 and therefore provides a significant energetic “push” of agent through a piping network. Thus, halon was used in many piping networks that most standard delivery halon alternative systems cannot accommodate today. The elusive “drop in replacement” is something for which the special hazard fire protection industry continues to search. Most halon alternative systems today have rough guidelines upon which to base preliminary designs, including piping lengths and number of nozzles.
Inert gas systems are not as limited by discharge characteristics, but rather by storage space.While Argonite and Inergen systems can be several hundred feet from the protected space, storage area for the cylinders is 10 to 20 times that of halon and halon alternatives.
However, new technology in the delivery of fluorine-based agents allow very close to a true drop-in replacement for halon systems. One manufacturer offers a system in which the nitrogen used to “push” the agent is separated from the agent itself, in a completely separate container at the point of installation. When the system is activated; the nitrogen flows into the agent storage container and literately pushes the agent through the piping network, like a piston. The system is commonly referred to as “piston-flow.” The advantage of this type of system is a greater level of nitrogen pressurization, typically up to 400 psi (28 bar) but, more importantly, minimizing the level of mixing between the nitrogen and the agent itself. The result of this type of system is a considerable improvement in flow performance, which can rival the flow of halon and “fit” into existing halon systems.
If interested in this or other specific agent delivery system characteristics, you are encouraged to seek out manufacturers of these systems and discuss what is available.
Fire protection for mission critical facilities can be a complex and daunting topic. It is best to break up the task into several topics and begin to create manageable assignments out of each one.
Understand the value of the facility in question. Value can be defined in a number of ways, from asset value to operational value to historical or sentimental value.
Evaluate the level of risk in the facility in question. What has been done to mitigate these risks? What can be done? What should be done in order to adequately protect against a potential hazard, including fire, in the facility?
Investigate the options available, from water-based systems to waterless systems, and determine what is right for your facility and your business. Understand the unique and varying levels of protection your facility will get from each option. It is the business owner’s responsibility to fully understand what each system will provide for his or her business.
This article originally appeared in the August 2008 issue of Occupational Health & Safety.