Testing Today's Telecom Equipment with Yesterday's Fire-Resistance Methodology

New fire-resistance test methods and specifications are needed to properly test current telecom designs. Several cases are examined that reveal the need for revising current fire-resistance test standards.

Telecommunications equipment designed for use in network facilities, such as the Regional Bell Operating Companies (RBOCs) central offices, has changed significantly in the last five years since GR-63-CORE was last revised. New technologies and an increasing demand for speed and reduced chassis size have changed the nature and design constraints of telecommunications equipment. For example, many of these new systems now incorporate design elements such as midplanes, smart fans, compact internal power supplies, thermal cutoff switches, heat shielding, and firewalls.

In the typical setup for fire-resistance testing, the heat flux transducers are placed to the left and right of the unit to be tested.

These new elements have created new challenges for designers who must ensure that these telecommunications systems meet fire-resistance requirements. In many cases, tests must be modified or compromised to conform to the new designs.

Much of the equipment deployed within RBOC facilities must meet many requirements to ensure a reliable and safe network. Among these requirements are those that telecommunications equipment manufacturers must comply with, including the Network Equipment Building System (NEBS) requirements detailed in GR-1089-CORE and GR-63-CORE. The fire-resistance requirement is defined in GR-63-CORE, section 4.2.¹

The evolution of the technology driving the designs of network systems has forced the need to review the fire-resistance test specified in GR-63-CORE, section 4.2 in light of designs incorporating advanced technologies. In fact, ANSI Committee T1E1.8 has begun addressing this issue in order to update and revise ANSI T1.319, the specification from which the GR-63-CORE is derived.² Test organizations and industry representatives are contributing their experience and research to this committee at ANSI. This article presents some of the key issues driving the need to create a fire-resistance test that can be easily adapted to today's equipment and beyond. This article also examines the challenges and issues both equipment manufacturers and testing organizations face as they interpret the requirements and apply them to current system designs.

Updating the Methodology

A primary element of the fire-resistance test is the use of a line burner to simulate a flame within a system. The line burner used for this test—and the methane flow rate used to control the flame profile—is no longer universal to all of the equipment that must conform to the fire-resistance criteria. The need to develop an updated test methodology is critical to address the variety of equipment types now available. Existing test methods need to reflect the variety of equipment types to which they are being applied. Work toward revising the methodology has begun through contributions from RBOCs, test labs, and manufacturers. Research into areas such as the heat-release characteristics of today's circuit cards, line burner applications, and alternative test methods are being examined. The end result of this research will be fire-resistance test methods that simulate more accurately the fire hazard model posed by current telecommunications equipment. Updated methods will help ensure safe and reliable operation of telecom equipment in central office environments. However, until such methods are developed, approved, and accepted by ANSI and the telecommunications industry, manufacturers and testing organizations must adapt the existing GR-63-CORE test methods to current equipment types.

New System Designs, Old Methods

The following scenarios describe design issues that are not currently addressed by GR-63-CORE.

Smart Fans. In the past, test units have used hot-wired fans to avoid powering up the chassis, which typically would contain dead circuit cards. Some in industry argue that this method is flawed because it fails to investigate the potential of a fire shutting down the fans and the consequences of such an event. Therefore, many test labs, with verbal input from RBOCs and industry, sometimes require chassis to operate the fans from internal power, which normally means powering the fans off the mid- or backplanes. Some designs, however, incorporate smart fans to suppress a fire by manipulating the airflow through a system. These smart fans typically use a thermal cutoff switch to detect a significant rise in temperature that may indicate a fire. The theory is that the fans turn off, thereby suffocating the fire within the chassis. Although this method is not always an ideal solution, if used, it must still be tested to ensure that the smart fan systems work as designed. The chassis containing this technology must be able to operate these smart fans during the test.

Many chassis also utilize a midplane design to minimize vertical height. This shortened design typically means that circuit cards located in the back of the unit are 3–4 in. deep. The line burner does not fit into these circuit card slots because the burner itself is approximately 6 in. long. Omitting tests on these circuit card slots would not sufficiently investigate the fire propagation hazard posed by the equipment under test. To accommodate the length, the preferred method is to use aluminum tape to cover any line burner holes that do not fit into the circuit card slot. Some believe this method is flawed because the flame height of the methane flow profile is much larger than the heat release and flame height of the actual circuit cards, and thus this method causes a severe overtest of the shortened card slots. A scaled line burner profile that accurately reflects the heat release and flame height of these circuit cards is being investigated.

Shallow circuit cards located in the back of a midplane chassis design cannot accommodate a 6-in.-long line burner.

Monitoring Equipment. Another aspect of the current methodology that must be addressed is the specified placement of monitoring equipment. For example, when testing shelf-level products, the method requires that two thermocouples be attached to two circuit cards positioned 2 in. over the top of the shelf with the circuit card slot under investigation directly between the two raised circuit cards. The raised circuit cards are used to determine whether any fire escaping the top of the test sample could ignite circuit cards of another unit in a central office. The thermocouples measure the fire and heat temperature emanating from the top of the test unit. Two heat-flux transducers must be located on either side of the test unit at a height even with the top and 2 in. from the surface. The flux transducers help determine whether the heat flux from the test unit could ignite material next to the test unit in a central office. All of these instruments, along with those located in the exhaust duct, determine the flammability characteristics of the unit. The main debate surrounding the practice of using the two circuit cards, the thermocouple above the test unit, and the heat flux transducers is that this configuration may not properly measure the data. For example, this setup would fail to measure hot air, smoke, gasses, and fire exhausted from the sides, front, or bottom of a unit. Because the intent of monitoring equipment is to determine the risk of these elements, then the monitoring equipment should be located at all possible exhaust points. Placing them as the current test method calls for fails to properly assess the entire system for potential fire propagation.

Methane Ignition. Methane ignition for starting the test has been another subject of scrutiny. Typically a hole, approximately 0.75 in. in diameter, is drilled through the chassis, and the methane line burner is inserted through this hole. The methane can be lit either before sliding the burner through the hole or via propane torch after inserting the burner. A shearing effect that often extinguishes the methane source is common when lighting the methane before inserting the burner. This shearing effect complicates starting the test, especially when high velocity airflow is present inside the chassis. Although lighting the methane after the burner has been inserted may be more successful, this technique could inadvertently burn components with the ignition torch rather than the methane line burner. Both methods result in a nonessential hole in the chassis, which can allow flames to escape from the chassis or allow unwanted airflow into the chassis. It is debatable whether this setup increases the potential for failure or alters final test results.

To address this concern, NTS has investigated the use of an automatic sparking device to initiate the flame on the methane line burner. This device enables lighting the burner without introducing an external flame to the chassis in order to ignite the methane. Therefore, no unnecessary burning of the chassis occurs. Using this device also allows sealing the hole drilled in the chassis for the line burner, which preserves the firewall integrity of the face plates and prevents oxygen from entering the line burner. This solution has been proposed for the revised ANSI T1.319 specification under development.

Chassis Design. Most of the difficulties in performing the current GR-63-CORE fire-resistance test are a by-product of the drastic changes in the chassis design of today's telecommunications equipment. Many in the industry contend that the fire-resistance test is inappropriate to test current designs. The conservative point of view suggests that current test methods ensure a reliable and safe network; however, once equipment fails this test, manufacturers are forced to spend significant time and money on chassis redesign just to comply with the current standard—even though the unit may not be a fire hazard. One design that often fails this requirement is a small pizza box unit (no more than 2–3 in. tall) that includes fuse and patch panels. These designs are usually convection cooled, which provides top and bottom ventilation. This type of construction is susceptible to failure because fire can escape from the equipment and flaming drippings can fall through bottom ventilation openings. Such equipment should be evaluated for its ability to resist fire propagation. New test methods for evaluating these designs are being considered, and alternative ignition methods are under investigation.

Conclusion

These design issues present a sampling of the difficulties being addressed by the ANSI T1E1.8 committee in order to improve and update the ANSI T1.319 specification. In addition to the problems discussed in this article, many others are being addressed at the committee's quarterly meetings. Research into the heat-release characteristics of various sized circuit cards, alternatives to the methane line burner, and new pass/fail criteria and test instrumentation has been or is being considered. The committee is working with industry to identify technically based solutions to concerns. Those involved in telecom compliance have presented many valid concerns about the suitability of the current GR-63-CORE fire-resistance criteria for evaluating current telecom equipment.

Evaluating test results relies to a large degree on interpretation when applying this test methodology to newer types of equipment. Many test labs and individuals interpret the methodology in GR-63-CORE differently, which leads to significant discrepancies in how the tests are performed. Consistency of both the application and the test results between test labs should be expected. Achieving this requires that the telecom community agree on an accurate and repeatable specification. Some manufactures have submitted identical test specimens to various test labs and have received completely different results. In addition, guidance from RBOCs often influences a test lab's interpretation of the specification. These discrepancies present a case for specifications that ensure that consistent testing is performed, and relevant comparisons are made for future study. These specifications provide, in essence, a contractual procedure to fulfill the requests of customers.

When forced to use an outdated and ineffective test methodology, many problems arise. Test labs often must deviate from the specification, which leads to subjectivity and disagreement among manufacturers and RBOCs. Test labs try to enhance the telecom industry's ability to evaluate new product designs, provide consistent test results to manufacturers and RBOCs, and ensure a more reliable communication network. Doing so requires ongoing scrutiny of test methods and review specifications to determine whether they continue to apply to new product designs. The changes in current designs make it necessary to update the current fire-resistance test specification. Accomplishing this is the responsibility of all stakeholders in the telecommunications industry.

References

1. Bellcore GR-63-CORE, section 4.2, "Network Equipment Building Standard: Physical Protection," Telcordia Technologies, Morristown, NJ, October 1995.

2. ANSI T1.319-1995, "Fire Propagation Hazard Testing Procedure for Equipment," American National Standards Institute, Washington, DC.

Jason Crete is program manager/fire facility manager for National Technical Systems (Boxborough, MA). He is a member of the ANSI T1 standards committee and is involved with the revision and editing of the ANSI T1.319 fire-resistance test standard.

Back to November/December Table of Contents