Printer Friendly

Suppression progression: an IT room is now safe not only from fire but also from the damage an outdated suppression system might have done.

Anyone who has ever spilled coffee on his or her computer keyboard knows the damage that liquid can do to electronics. Multiplying that damage by thousands--or millions--of dollars provides an estimate of the harm that can be done if fire sprinklers discharge in a room filled with expensive IT equipment. Companies avoid that exposure by using so-called "clean agents," fire suppression agents that are safe for electronic equipment. But Halon, perhaps the most famous of these, creates well-known environmental hazards, and Environmental Protection Agency rules forbid new Halon-based systems from being installed. Consequently, companies in need of new or replacement systems must look to one of the new generation of agents that are more environmentally friendly and safer to use. These include Ansul's Inergen (a mixture of inert gases) and chemical agents like Great Lakes Chemical Corporation's FM-200.

LAST YEAR, JASON WHITT, SAFETY and environmental coordinator at Leco, a St. Joseph, Michigan-based manufacturer of analytical lab instruments, struggled with the question of which agent to use. Whitt says that when the company acquired a great deal of new IT equipment to beef up its computer room, his team decided to take a close look at changing its fire suppression system. That system consisted of a Halon-protected subfloor (where computer cables and electrical wiring are run, and where a fire is most likely to break out) and a sprinkler system. Besides the fact that the Halon system was now obsolete, there were other considerations as well. "The Halon protected the subfloor but not the room itself," says Whitt. "We also realized that with the sprinkler protection we had, probably one of the worst things that could happen would be for those sprinklers to go off."

Whitt also recognized that IT staff members were nervous about health considerations in the event of a fire and a discharge by the Halon system. He wanted a way to protect the company's IT investments while keeping staff safe.

Whitt contacted a local dealer, Vanguard Fire & Security Systems, and asked their advice. Special hazard engineer Micah Dickinson met with Whitt and discussed several clean-agent options. The last option Dickinson described was a clean agent called Sapphire, invented about five years ago. Although it has been used for some time in Europe, it was only approved for use in the United States in January 2004. Dickinson told Whitt that he had witnessed a demonstration in which a salesperson immersed his cell phone in the Sapphire liquid, then pulled it out and made a call with no damage done to the phone.

Sapphire is stored as a liquid in tanks, and in its liquid state it looks and feels like water. Other clean agents are also stored in liquid form, but Sapphire has a higher boiling point (49[degrees] C, compared, for example to FM-200, which boils at -16.4[degrees] C). Consequently, it is easier to store and handle, as when tanks need to be refilled, Dickinson says.

Sapphire evaporates soon after it is discharged, a fact that was proven to Dickinson when he saw the sales rep's cell phone dry almost immediately after being removed from the agent. Dickinson explained that the agent is nontoxic under normal conditions, although a Material Safety Data Sheet on the product warns that "prolonged or repeated exposure" may cause liver problems; it also advises that persons who get the substance on their skin or in their eyes must use water to clean the agent off, though it is not expected to cause any significant irritation. The agent's ozone depletion level is negligible.

These factors had Whitt almost sold on the Sapphire system. Dickinson and some Vanguard technicians came to the Leco facility to take room dimensions to calculate how much suppression agent would be needed, as space was a final area of concern for Whitt. The area to be protected was estimated at some 5,000 cubic feet, and the space available for tanks of suppression agent was limited. The results of Vanguard's calculations tipped the scales for Whitt in favor of Sapphire. The footprint needed for the Sapphire system would be much smaller than for other agents.

"Whereas Inergen might have taken eight tanks to protect the room, Sapphire took only two," Whitt says. Even better was that the price was reasonable. "We were expecting it to be significantly more expensive than Inergen or FM-200," Whitt says. "What we found is that it was actually cheaper than one of the quotes we got for FM-200 and almost identical to Inergen."

That assuaged any remaining concerns. Whitt decided to go with Sapphire, making him one of the first companies in the United States to use it.

Fan test. The next step was for Vanguard technicians to come in and conduct a fan pressurization test to find any air holes and monitor the pressure that the room could maintain. This step was needed because clean agents must be able to reach certain concentrations to extinguish a fire and prevent a reflash; these concentrations can only be reached if a certain level of pressure is maintained in the room, which would not be possible if air was rushing in or out around a door or ceiling tile.

While the significant escape routes for the agent must be sealed, the room should not be made completely airtight, explains Dave Pelton, director of global marketing for Tyco Fire and Security. The room "should also allow venting somewhere to accommodate any excessive pressure built up upon discharge of the agent," he says.

During a fan test, a specialized computer program is used to gauge how long the agent would need to be held for maximum effectiveness. Technicians put into the program information about the quantity of the agent, the volume of the room, and other specifics including whether or not air circulation systems will be shut off.

Before the first fan test was done, Whitt took some preliminary steps to make the room more leakproof, such as putting in door gaskets and caulking holes in certain areas such as where cables ran through walls or the floor. Then, Dickinson and his team came in for the test. They set up a large device that included a fan that would be placed in the doorway and a thick cover that would go over the rest of the doorway. The fan was turned on and air was blown into the room to pressurize it. Dickinson kept his eye on the gauge used to indicate the amount of pressure achieved. For a test to be successful, the room needed to maintain a certain pressure for about 10 minutes, Dickinson says.

At the same time, a technician moved around the room with a smoke pencil, a hand-held device that releases a smoke-like substance when squeezed. He used the pencil to put puffs of smoke in different areas to look for any leaks. These would be indicated by smoke being sucked up or blown down through a hole in a ceiling tile, for example.

During the first test, a number of leaks were detected that prevented the room from reaching the right pressure for enough time. Whitt and his team spent a few days doing more leakproofing, such as clipping down ceiling tiles and changing light fixtures that were allowing air to escape. The room also had air-conditioning dampers, but they seemed to have no appreciable effect on room pressurization. If they had, Whitt says, he could have installed a mechanical damper switch that would have shut the damper when the fire-suppression system discharged.

Another fan test a week or so later showed that a few leaks still remained. "We were surprised at what we found," Whitt says. "We thought we had the place sealed up pretty well, but as we dug around a little further we found under the subfloor a few places where there was significant leakage." It took two more fan tests to achieve the right pressure and get the right results, Dickinson says.

Installation. Once the room was properly sealed, it was time to install the suppression system. There are two components to installation, according to Dickinson; putting in the piping and distribution network, and then installing the detection and control hardware. Dickinson estimates that the whole process of installation took about 80 hours over roughly two weeks.

Piping. The old subfloor piping from the Halon system was removed and new piping was cut and installed both above the ceiling and into the subfloor where nozzles would later need to go. At this point, the Sapphire cylinders were brought in and set securely in place.

Once the piping network was in, the Ansul computer program that modeled the setup of Leco's Sapphire system needed to be run again before the proper nozzles could be ordered. Dickinson says that this is because even minor changes made while installing the piping network can affect which nozzles are needed. "Each nozzle has a specific orifice that's drilled that [regulates] how long it will take the system to discharge and what kind of flow rates you're going to have," he says. These nozzles also range in size and position depending on the pipe size and the piping setup. The project was delayed for a few weeks while nozzles were ordered.

Detection and control. While the pipes were being put in place, other Vanguard technicians installed the detection and control equipment, such as the control panel, smoke detectors, manual pull stations, and abort stations. While some of the existing Halon fire suppression system components might have been compatible, Dickinson says that the equipment was already nearly a decade old, which he feared could lead to unreliability and unavailability of replacement parts. Whitt agreed that it would be best to change all the system components at the same time. In this phase, wire and conduit was run to connect and deliver power to the detection and control components.

Three automatic door closures were also installed at this time, Dickinson says, because two offices connected to the IT room didn't need protection but the staff members in those offices liked to keep their doors open. The project installed magnets that held the doors open, and the system was wired so that when the suppression agent was about to discharge, it would automatically shut the doors.

Another task was to connect the detection and control equipment to the security board that linked to Leco's security contractor, ADT. According to Whitt, there are separate relays for trouble alarms, high-heat alarms, and a release alarm which signals that the suppression system has released its agent and notifies the fire department to dispatch trucks. When work on the panel was completed, technicians worked with ADT to conduct dry tests of the relays to make sure that everything was functioning properly. Those tests went flawlessly, according to both Whitt and Dickinson.

Training. Vanguard is now helping to train Leco employees with regard to what different alarms mean and how employees should respond. Dickinson says that fires are often not handled properly because of human error, so during training he insists that anybody who might be in the room be familiar with what to expect when alarms go off. Anyone not tasked with responding to fire emergencies is told to evacuate.

Employees are also shown during training that brief exposure to the agent will not harm them. Nevertheless, they are urged to exit quickly, because burning wires in an IT room can cause harmful fumes. For those who have responsibility in emergency situations, training includes how abort stations work and how to use the panel functions. Once training is complete, Whitt says he expects that Leco will conduct fire drills (there have been no Sapphire releases yet at the facility).

Lessons learned. Clean agents aren't the answer to all fire-prevention requirements. Because they all require sufficient concentration of agent for a period of time, they work best in smaller rooms that can be sealed, rather than in large, open spaces. And, as Leco's experience shows, rooms typically need some construction or retrofitting before the agents can be effective. These agents are, therefore, only appropriate for sensitive assets within an enclosed space.

Whitt reflects that the most difficult and important part of changing to Sapphire was the fan pressurization test. "For anyone who might be considering this type of system, fan testing and sealing of your room is critical to the effectiveness of the system," he says. "I'd recommend they analyze and evaluate the room for its ability to hold pressure, and if they think it's a loose room, they might want to get an idea of how much they might end up having to spend in order to achieve an airtight room. It may be more than they're willing to spend on top of the cost of the system itself."

In addition to the roughly $23,000 it cost to install the clean-agent suppression system, Whitt estimates it cost another $2,000 to better seal the room, including purchasing doorframe gaskets and gasketed light fixtures. That amount could be dramatically different depending on the specifics of the room; for example, if an all-new drop ceiling with heavier tiles had been necessary, the cost would have been much higher, Whitt explains.

Dickinson adds that he advises clients to get a preliminary fan test to give them a better idea of what kinds of additions or alterations to a room might be necessary. "You can get into some rooms, and they don't pass for half a minute," he says. Changes needed could be so extensive that "you could end up building a new room," he adds.

As for Whitt, he's confident that extinguishing a fire in the IT room will no longer cause more problems than it's meant to solve. His next challenge is finding a way to keep coffee spills from damaging keyboards throughout the rest of the company.

Peter Piazza is assistant editor of Security Management.
COPYRIGHT 2004 American Society for Industrial Security
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2004 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Title Annotation:Information Technology
Author:Piazza, Peter
Publication:Security Management
Geographic Code:1USA
Date:Jul 1, 2004
Previous Article:A healthy approach to medical data: find out how one hospital streamlined secure data access to meet medical staff needs while complying with...
Next Article:Education, Exhibits, Networking.

Terms of use | Privacy policy | Copyright © 2021 Farlex, Inc. | Feedback | For webmasters