1、AN-04-1 1-4 (RP-1108) Full-Scale Fire Tests for Cables in Plenums G.D. Lougheed, Ph.D. Member ASHRAE C. McCartney M. Kanabus-Kaminska, Ph.D. ABSTRACT In a joint research project involving ASHRAE and the National Research Council Canada (RP-I I08), the efect of $res involving communication cable inst
2、alled in air-handling plenums was investigated. The project included suweys in North American ojce buildings to determine the type and quantities of cable in return airplenums andjre scenarios that could potentially ignite the cables. It also includedfire tests performed at three scales: small, medi
3、um, and full. The bench- scale tests were conducted using a cone calorimeter: In addi- tion, medium-scale tests were conducted using a modijedstan- dard roomjre test facility. Tests conducted in this facility were used to determine the behavior of communications cable with exposure to air heated to
4、200“C, 325“C, and 450“C, as well as direct flame impingement. Final data from the tests are used to investigate the potential effect ofjres involving cables in plenum spaces on the hazard to building occupants in compartments contaminated by smoke distributed through a building HVAC system. INTRODUC
5、TION The use of ceiling voids for unducted return ventilation air is an increasingly common practice in modem commercial buildings (Clarke et al. 1993). It is also common practice to route communication cables through hidden voids in build- ings. In those cases in which the void space is also used a
6、s part of the normal HVAC system, there is the potential, in the case of a cable fire, to spread heat and smoke to inhabited parts of the building. With the rapid increase in computer-based information technology, there is a corresponding rise in the demand for cabling to support it. It is estimated
7、 that computer usage is increasing at a rate of 20% per year, and local area networks (LANs) are recabled approximately every three years (Fardel1 1998). This new cabling may be installed over multiple layers of older cables, potentially resulting in high fuel load in concealed spaces. The potential
8、 increase in cable loads in plenums resulting from the increased use of computers and re-cabling of LAN networks has raised concerns in the regulatory community (Clarke and Gewain 2000). Specific concerns regarding the potential impact on life safety of exposed LAN cables installed in above-ceiling
9、return air plenums resulted in ASHRAE initiating a research project with the National Research Council Canada. The objective of this project was to evaluate the hazard to human life of computer and communi- cation cable fires in return air plenums above ceilings and to develop information that can b
10、e used as input to performance test standards and codes. In the initial phase of the project, surveys were conducted in North American office buildings to determine the type and quantities ofcable in return air plenums, and fire scenarios that could potentially ignite the cables. In addition, bench-
11、scale and medium-scale tests were conducted with cables used in plenum spaces in North America. The results of the prelimi- nary investigations were discussed in a previous paper (Lougheed et al. 2002). These preliminary studies were used to select cable types, cable loads, and fire scenario for a s
12、eries of full-scale tests using a facility set up specifically for this Gary D. Lougheed is a senior research officer and Cam McCartney and Malgosia Kanabus-Kaminska are technical officers in the Fire Risk Management Program, Institute for Research in Construction, National Research Council Canada,
13、Ottawa, Ontario. 652 02004 ASHRAE. Table 1. Characteristics of New Communication Cables Cable Rating Category Sheath Material Insulation Material A B C D E CMP FT4 CMP CMP CMP CMP CMP CMP CMP FT4 5 5 5 5 5 5 3 3 5 5 project. The fire scenario selected for the full-scale tests was a fully developed f
14、ire in the compartment below the plenum. A primary objective of the plenum cable fire project was to develop data on the amount of smoke and other combustion products produced by the cables and the potential effect of this smoke on tenability conditions in a target room. In this paper, the data prod
15、uced by bench-, medium-, and full-scale fire tests are used to determine the effects of the smoke in the target room on the ability of occupants to evacuate. CONE CALORIMETER TESTS There is a broad range of data communication cables available in the North American market. Bench-scale tests were cond
16、ucted using a cone calorimeter to provide an initial evaluation of the fire performance of representative cables. The bench-scale tests were conducted using the ASTM E 1354 cone calorimeter (ASTM 1997), with a heat flux of 50 kW/m2. Specimen mass loss and smoke production were recorded. Any gases pr
17、oduced were sampled and analyzed using standard gas analyzers for O, CO, and CO, to determine the heat release rate. In addition, the combustion gases were analyzed using an FTIR spectrometer to measure other combustion byproducts, including HF and HCl. For the cone tests, the cable was cut into 100
18、 mm lengths. These cable lengths were placed side by side on metal wire mesh in the cone calorimeter holder. The ends of the cables were unsealed. Ten cable types were purchased on the open market. The characteristics of the cables are summarized in Table 1. Each cable is referenced by a letter desi
19、gnation. Tests were also conducted with previously used cables removed from buildings during refurbishment. The results for these cables were comparable to those for the new cables (Lougheed et al. 2003a). For this paper, the results for only the new cables are used for the hazard analysis discussed
20、 in the next section of the paper. The cables were selected from the major North American manufacturers to represent a cross section of cable types and PVC PVC PVC PVC PVC PVC PVC PVC PVC PVC Perfluoropolymer Poly olefin Fluoropol ymer Perfiuoropolymer Perfluoropolymer Perfluoropolymer PVC PVC Perfl
21、uoropolymer Polyolefin ratings presently used in office buildings. The cables tested included eight that were labeled as CMP and, thus, meeting the requirements for use in air-handling plenums in the U.S. and some jurisdictions in Canada. In addition, two cables were labeled as FT4, meeting the requ
22、irements for the remaining jurisdictions in Canada. The information on the sheathing and insulation materials provided in Table 1 is based on the test results. Specifically, the FTIR measurements provided information on the production of various gases that would typically be produced by PVC-, fluoro
23、polymer-, and perfluoropolymer-based materials. A summary of the cone calorimeter results for each cable is provided in Table 2. The results provided in the table are the average of three tests. The test results (Table 2) indicate that there is a wide vari- ation in the performance of the ten commun
24、ication cables. However, based on total heat output, the communication cables can be grouped as follows: Group 1 (Cables A, D, E, and F) had total heat outputs of 5.8-8.2 MJ/m2. All the cables in this group used per- fluoropolymer materiais as the cable insulation. Group 2 (Cables C, G, H, and I) ha
25、d total heat outputs of 13.3-16.9 MJ/m2. The cables in this group used a variety of materials for the cable insulation, including PVC (Cables G and H), fluoropolymer (Cable C), and perflu- oropolymer (Cable I). Group 3 (Cables B and J) had total heat outputs of 38.7- 44.1 MJ/m2. These were the two F
26、T4 cables that used polyolefin materials for the insulation. The lowest total heat outputs were for the Group 1 cables, and the Group 3 cables had the highest heat outputs. The smoke production rates for the cables were measured using a He-Ne laser-based system. As with the heat release rates, the s
27、moke release rate results were similar for the cables within a group. There was, however, a difference in the amount of smoke produced by the cables in the different groups. The ASHRAE Transactions: Symposia 653 Table 2. Summary of Cone Calorimeter Results with 50 kW/m2 Exposure for New Cables Ignit
28、ion Initial Mass Time Mass Loss Peak HRR Total HR Peak SPR Total Smoke Cable Rating (g) (SI (g) (kW/m2) (MJ/m2) (m2/s) (m2) A CMP 63.9 111.3 27.5 46.7 7.2 0.030 3.1 B FT4 60.0 52.0 31.0 198.7 44.1 O. 125 19.6 C CMP 62.7 80.7 25.4 87.5 13.3 0.075 6.4 D CMP 68.6 227.7 27.3 43.9 5.8 0.025 3.0 E CMP 64.
29、7 48.3 14.7 62.4 8.2 0.015 1.8 F CMP 68.8 42.0 12.4 73.0 6.2 0.024 1.8 G CMP 60.5 66.3 28.3 111.7 16.9 0.056 7.8 H CMP 60.4 69.7 26.4 89.4 13.3 0.054 6.6 I CMP 64.7 33.8 26.9 88.6 14.6 0.041 5.9 J FT4 55.8 45.5 30.5 170.6 38.7 0.108 14.9 HRR - Heat release rate Total HR - Total heat release SPR - Sm
30、oke production rate total smoke produced by the cables in Group 1 was 1.8-3.1 m2, compared with 5.9-7.8 m2 for the cables in Group 2 and 14.9- 19.6 m2 for those in Group 3 (Table 2). Gas samples were extracted from the exhaust duct for analysis using an FTIR spectrometer. The FTIR was calibrated wit
31、h gas standards for CO, CO, HCl, HF, and COF,. Using these calibrations, the FTIR spectra were analyzed to provide time-dependent measurements for the gas concentrations produced in the tests. Figure 1 shows the heat release rate, smoke production, and gas species results for the tests with Cable D.
32、 For the communication cables, three acid gases (HCl, HF, and COF,) were found in addition to CO and CO, that are produced when carbon-based materials are burned. HCl was measured in the tests for ali ten cables. For most cables, the HCI was from the PVC used as the sheathing material. However, for
33、Cables G and H, PVC was used as the insulation material, and higher HCI concentrations were measured for these cables. HF was measured for those cables that used flouropoly- mer or perflouropolymer materials as the insulation material. This included Cables A, C, D, E, F, and I (Table 1). No HF was m
34、easured for the other four cables that used polyolefin or PVC as the insulation material. COF, was measured for those cables that used a perflouropolymer as the insulation material. These were Cables A, D, E, F, and I. The cone calorimeter tests were conducted in the open, with ambient levels of oxy
35、gen in the combustion air. For most materials, CO production is limited under such conditions. However, CO was produced in the tests with the communica- tion cables. The FTIR results were used to determine the total yields of CO, CO, HC1, HF, and COF, produced in the tests. A mass balance calculatio
36、n was carried out and indicated that 75% to 654 2M. . . . 1m 0.16 Time (s) Figure 1 Cone calorimeter results for Cable D. 90% of the mass loss was accounted for by these five gases (Lougheed et al. 2003a). The mass balance calculation does not include the mass loss as soot, H,O, volatile organic com
37、pounds (acrolein, benzene, toluene, etc.), and other minor components. HAZARD COMPARISON USING BENCH-SCALE DATA The potential impact of the various gases and smoke is dependent on the volume into which they are distributed (Babrauskas 2000). For the analysis in this section of the paper, a volume of
38、 i 5 m3 was selected. This is the approximate volume of air that flows through the cone calorimeter exhaust system in 600 s. This volume is also representative of a small room with dimensions of 2 m by 3 m by 2.5 m in height. One evaluation of the toxic fire hazard can be determined using the fracti
39、onal effective dose (FED) principle (Babraus- kas 2000; Purser 2002; Klote and Milke 2002). This approach ASHRAE Transactions: Symposia assumes that the toxic effects are linearly additive. The expres- sion used for the present calculations was (1) col I HCZ I HF I coF21 3000 3700 2000 750 FED = whe
40、re the terms in brackets denote the individual gas concen- trations (ppm) and the denominators represent the LC50 values for the same gases. The LC50 concentrations are assumed 30-minute concentrations for man (Babrauskas 1997; Mote and Milke 2002). An FED of 1 is the level that is lethal for the av
41、erage occupant. The FED values calculated for the new communication cables using Equation 1 are provided in Table 3. As shown in Figure 1, the gas species were produced at different times during the test. The hazard calculation in this section assumes that the total mass of all of the gas species an
42、d smoke is mixed together in the selected volume. The calcula- tion using Equation 1 provides an estimate of the relative hazard of the smoke produced from the cables based on the hypothetical volume and the selected criteria. The FED values for the new communication cables were similar for the thre
43、e groups of cables tested. Group 1 cables had FED values in the range of 0.098-0.229, compared with O. 1 16-0.200 and 0.072-0.12 1 for Group 2 and 3 cables, respec- tively. However, Cable E in Group 1 had better performance under the test conditions and had a relatively low FED compared to the other
44、 cables in this group. As a general trend, the lowest FED values were for the Group 3 cables, and the highest were for the Group 1 cables. The work by Jin (1 976) and others showed that the pres- ence of irritant gases could have a substantial effect on the abil- ity of people to safely evacuate. Th
45、e effects of irritant gases in terms of sensoryhpper respiratory and, to some extent, pulmonary irritation can be assessed using the fractional effec- tive concentration (FEC) concept (Purser 2002). The expres- sion used for the present calculations was where the terms in brackets denote the individ
46、ual gas concen- trations (ppm), and the denominators represent the irritant gas concentrations (ppm) that are expected to cause incapacitation in half the population (Purser 2002). The concentrations used in the denominator for HCl and HF are from Purser (2002). The analogous concentration for COF,
47、is a hypothetical value determined by reducing the LC, value by a factor of approx- imately 4 (Babrauskas 1997). An FEC of 1 is assumed to be the level at which the average occupant would be incapaci- tated by the irritant gases. The FEC values calculated for the new communication cables are provide
48、d in Table 3. The FEC values were similar for the three groups of cables (Group 1 with 0.234-0.556; Group 2 with 0.328-0.392; and Group 3 with O. 117-0.252). As with the FED values, Cable E had a relatively low FEC value compared with the other cables in Group 1 (0.234 versus 0.5). Otherwise, the ge
49、neral trend was for Group 1 to have the highest FEC value and Group 3 the lowest. An FED of 1 is the level lethal to the average occupant, and an FEC of 1 is the level at which the irritant gases would incapacitate the average occupant. In the context of fire scenarios, it is generally recommended that a threshold criteria of 0.3 be used (Purser 2002). At this level, it is expected that approximately 1 1.4% of the population will be susceptible to the exposure. The FED values for the smoke produced fiom the communication cables were 0.07-0.23. As such, the gas mixture in the hypo
