Development of a common European system for fire testing of pipe insulation based on EN 13823 (SBI) and ISO 9705 (Room/corner test)

Full text

(1)Björn Sundström Jesper Axelsson. Development of a common European system for fire testing of pipe insulation based on EN 13823 (SBI) and ISO 9705 (Room/Corner Test). SP Swedish National Testing and Research Institute Box 857, SE-501 15 BORÅS, Sweden Telephone: + 46 33 16 50 00, Telefax: + 46 33 13 55 02 E-mail: info@sp.se, Internet: www.sp.se. SP REPORT 2002:21 ISBN 91-7848-871-0 ISSN 0284-5172. SP Swedish National Testing and Research Institute SP Fire Technology SP REPORT 2002:21.

(2) 2. Abstract The European union is harmonising testing and classification of building products. For reasons of free trade building products will be classified in the same way in all European countries. Euroclasses are introduced for reaction to fire. The classes are based on test data from test methods; the most important test is the Single Burning Item, the SBI Test. A large-scale reference scenario, the Room/Corner Test, ISO 9705, was used to identify suitable limit values for linings when tested in the SBI Test. However, so-called linear products, for example pipe insulation, were not easily covered by this system due to technical problems in testing and classification. Therefore the European Commission opened a possibility for taking a special decision on these products. CEN/TC 88 therefore started a project with the aim of finding a technical solution how to test and classify pipe insulation in a way useful for European harmonisation. Important findings of this work were that the Room/Corner Test can be used as the reference scenario for testing of pipe insulation, that the SBI Test correlates very well with the Room/Corner Test, that test results are stable for various small changes in procedures, and that a classification system for pipe insulation on a European level can be created.. Key words: pipe insulation, fire engineering, fire technology. Sveriges Provnings- och Forskningsinstitut SP Rapport 2002:21 ISBN 91-7848-871-0 ISSN 0284-5172 Borås 2002. Swedish National Testing and Research Institute SP Report 2002:21. Postal address: P O Box 857, SE-501 15 BORÅS Sweden Telephone: +46 33 16 50 00 Telefax: +46 33 13 55 02 E-mail: info@sp.se Internet: www.sp.se.

(3) 3. Table of contents Abstract. 2. Table of contents. 3. Acknowledgements. 4. Sammanfattning. 5. 1 1.1 1.2. The rationale of the study Classification of linings Considerations for pipe insulation. 7 7 8. 2 2.1 2.2. Selection of ISO 9705 as reference scenario ISO 9705 - the test Mounting configuration. 9 9 10. 3 3.1 3.1.1. Using the SBI Test procedure Mounting Frame for mounting the pipe insulation. 11 11 12. 4. Basic principles for the analysis. 13. 5 5.1 5.2 5.3 5.3.1 5.3.2. Test results and data analysis Products tested Test data from ISO 9705 and the SBI (EN 13823) Analysis of correlation between ISO 9705 and the SBI FIGRA SMOGRA. 15 15 16 18 18 19. 6 6.1 6.1.1 6.1.2 6.2 6.3 6.4 6.5. Sensitivity analysis Fire load Variation of fire load by changing the number of pipes Variation of the fire load by changing insulation thickness Influence of threshold values for FIGRA in the SBI Vertical and horizontal mounting in the SBI Attachment of the pipe insulation in the SBI Comparison of test data of material in tube form with flat form in the SBI. 21 21 21 25 28 29 32 32. 7. Conclusions. 36. 8. References. 37. A1 A1.2. Detailed test data and product parameters EPS test. 39 44. A2. Room/Corner Test mounting details. 46. A3 A3.1 A3.2. Calculation of test parameters Room/Corner Test parameters SBI Test parameters. 53 53 55. A4. ISO 9705 HRR and SPR plots. 56. A5. SBI (Single Burning Item) HRR and SPR plots. 80.

(4) 4. Acknowledgements The work was initiated and co-ordinated through CEN/TC 88/WG 10/Ad hoc group fire-tests. The Ad hoc group leader Ulrich Rohr, Armacell, and the members are highly acknowledged for their support and active work in the project. The authors would like to thank their colleagues from SP Fire Technology, especially Patrik Johansson, Per Thureson, Lars Pettersson and Patrick Van Hees for their participation in the project..

(5) 5. Sammanfattning Inom EU pågår en harmonisering av provning och klassifikation av byggnadsprodukter. För att främja fri handel skall byggnadsprodukter klassificeras på samma sätt i alla europeiska länder. Inom brandområdet introduceras ett system av brandklasser som kallas Euroclass. Klasserna är baserade på provningsdata från testmetoder och den viktigaste metoden är SBI, Single Burning Item. För att identifiera lämpliga gränsvärden för material som testas i SBI användes ett fullskaligt referensscenario, Room/Corner Test. När det gäller ytskikt finns klassningssystemet färdigt men s.k. linjära produkter, t.ex. rörisolering, täcks inte på ett bra sätt av systemet för ytskikt på grund av tekniska och praktiska problem med provning och klassificering. Av denna anledning har europakommissionen öppnat en möjlighet till att ta fram speciella beslut angående dessa produkter. CEN/TC 88 har därför startat ett projekt med målet att finna en teknisk lösning som möjliggör harmoniserad brandprovning och klassificering av rörisolering i Europa. Viktiga resultat från arbetet är att SBI kan användas som huvudtestmetod för rörisolering med Room/Corner Test som referensscenario, att resultaten från SBI korrelerar väl med dem från Room/Corner Test, att resultaten är stabila för små ändringar i provningsprocedurer samt att ett klassifikationssystem för rörisolering kan skapas på europeisk nivå..

(6) 6.

(7) 7. 1. The rationale of the study. Classification of reaction to fire performance of construction products is laid down in the European Commission Decision 2000/147/EC. The decision covers a wide range of building products, mostly linings with a relatively small number of test procedures. It was also recognised that not all building products could be tested and classified in their end-use conditions in a fully satisfactory way considering their fire performance. Therefore a footnote in 2000/147/EC was left for so called linear products. The footnote says: “The treatment of some families of products, e.g. linear products (pipes, ducts, cables etc.), is still under review and may necessitate an amendment to this decision”. Guidance paper G, issued by the Commission 14 Dec 1999, covers the case when a product family cannot be correctly assessed by using the ordinary tests and classification. An appeal procedure that can be used by producers against a given classification is described. Linear products are mentioned in guidance paper G. Pipe insulation and jacketing are considered to be linear products and this report presents work done on a technical solution for testing and classifying pipe insulation and jacketing. NOTE: A jacketing is applied on site. A facing is factory applied.. 1.1. Classification of linings. The major test procedure and definition of classification criteria according to 2000/147/EC relies on the SBI1 test. The most important parameter used for the main classification is FIGRA describing the fire growth rate. The actual limit values used for the Euroclasses are based on work done in a so-called reference scenario. The reference scenario is taken to reflect the real fire behaviour in such a way that it can be used for classification purposes. The reference scenario for linings for the SBI Test is the Room/Corner Test, ISO 97052. The Room/Corner Test is a room fire test that allows products to be tested in conditions that are close to end-use. In defining the Euroclasses the EU group of regulators, RG, proposed performance requirements as well as the corresponding values for FIGRA also for the Room/Corner Test. The requirements are explained in the Standing Committee document, CONSTRUCT 98/296, see Table 1-1.. Table 1-1. Proposed limit values for the Euroclasses of lining products tested in the reference scenario, ISO 9705 the Room/Corner Test. FIGRA (RC) criterion (kW/s) ≤ 0,16. > 7,5. Burning Behaviour in Reference Scenario No flashover plasterboard or better No flashover No flashover at 100 kW No flashover before 2 min at 100 kW Flashover before 2 min. -. -. ≤ 0,5 ≤ 1,5 ≤ 7,5. EUROCLASS A1 A2 B C D E F. Data in large scale according to the Room/Corner Test for more than 60 building products was analysed to support the limit values given above. For the reference scenario used for linings FIGRA (RC) is calculated during the period from ignition to flashover as the peak HRR of the fire.

(8) 8 excluding the contribution from the ignition source divided by the time at which this occurs. For the SBI Test FIGRA (SBI) is calculated in principle as the maximum of the function HRR/t. Correlating the FIGRA (RC) with the FIGRA (SBI) results in a value of about 95% for linings.. 1.2. Considerations for pipe insulation. We note the significant correlation between flashover in the Room/Corner Test and the different Euroclasses for linings. However, the mounting of pipe insulation in real-life conditions is not comparable to the mounting of linings. The surface area covered by the pipe insulation is far less. The resulting fire growth due to the involvement of pipe insulation is largely governed by flame spread along individual or several lines. In contrast to linings the expected fire in pipe insulation will therefore not typically lead to flashover in a room due to the insulation alone. Therefore, testing of representative installations of pipe insulation cannot use the flashover phenomenon as a performance criterion. In fact a continuous measure of the fire growth rate is needed as a basis for classification. This will allow criteria for classification to be selected in a way to reflect hazard for these types of products. Thus the Room/Corner Test can be seen as a large-scale reference test, in this report called the reference scenario, when used with the modifications that are needed for testing pipe insulation. The basic principles for the analysis in this report are further discussed in Chapter 4..

(9) 9. 2. Selection of ISO 9705 as reference scenario. 2.1. ISO 9705 - the test. The reference scenario chosen is the international standard Room/Corner Test, ISO 9705, see Figure 2-1. ISO 9705 is also in the process of becoming a CEN standard. Smoke measurement. Gas analysis (O2, CO, CO2) Flow measurement. Exhaust gases. Exhaust hood 3mx3mx1m. 2,4 m. Gas burner. Doorway 0,8 m x 2,0 m. 3,6 m. m 2,4. Figure 2-1. The Room/Corner Test, ISO 9705. The Room/Corner Test was first published by ASTM in 19823 and then by NORDTEST in 19864. The international standard, ISO 9705, was published in 1993. The Room/Corner Test is a large-scale test method for measurement of the burning behaviour of building products (linings). The principle output is the occurrence and time to flashover. As for the SBI a direct measure of fire growth (Heat Release Rate, HRR) and light obscuring smoke (Smoke Production Rate, SPR) are also results of a test. The building product is subjected to a fire attack that is close to a real fire under end-use conditions. A door opening ventilates the room. The ignition source is a gas burner that is placed in one of the room corners. The burner heat output is 100 kW for the first ten minutes and then 300 kW for another ten minutes. The peak HRR of the ignition source may be seen as representing a large waste paper basket (100 kW) or a small, upholstered chair (300 kW). In comparison, an upholstered sofa may give a maximum HRR in excess of 2000 kW. HRR and SPR are measured continuously. The test is stopped after 20 minutes or when flashover is occurring..

(10) 10 Experience of testing products has been gained during at least 10 years of work with the Room/Corner Test. A considerable amount of information on product burning behaviour in this method is available and the thermal conditions during a fire test have been carefully mapped5,6,7.. 2.2. Mounting configuration. A lining is normally mounted on three walls and on the ceiling of the compartment. A pipe insulation however is not a lining and must be mounted in a different way. In this study several different ways of mounting were tried in the Room/Corner Test. In total four different mountings were tried, called version 1-4. In all versions the pipe insulation was mounted on the ceiling and along the walls of the test room but in different amounts and configuration. Comparative tests in all versions justified the choice of Version 1, shown in Figure 2-2 below, for the main testing. Version 2 was discarded early in the study being too similar to Version 1. The comparison is further discussed in Section 6.1 and detailed drawings of the mounting in all versions can be found in Appendix A2.. Figure 2-2. Schematic drawing of mounting Version 1. The roof and left wall of the ISO 9705 room have been removed for better view..

(11) 11. 3. Using the SBI Test procedure. The Single Burning Item1, SBI, is the major test procedure for classification of linings, see Figure 3-1.. Figure 3-1. SBI, the Single Burning Item, a new test procedure used for classifying building products in a harmonised European system. SBI is of intermediate scale size. Two test samples, 0,5 m x 1,5 m and 1,0 m x 1,5 m are mounted in a corner configuration where they are subjected to a gas flame ignition source. The heat output from the burner is 30 kW and the burning behaviour of the samples is evaluated during 20 minutes. A direct measure of fire growth (Heat Release Rate, HRR) and light obscuring smoke (Smoke Production Rate, SPR) are principal results from a test. Other properties such as the occurrence of burning droplets/particles and flame spread are also observed. The tests in this project were mostly performed according to prEN 13823 and some according to EN 13823. However, there is no difference to the prEN procedure and the EN procedure for the tests in this project. Therefore the term EN 13823 is used throughout this report.. 3.1. Mounting. In the SBI a simple approach for mounting was selected because it is not realistic to test all the variety of combinations. The pipe insulation was mounted side-by-side using spacing, which enhance upward flame spread. A few experiments were run with alternative mounting including horizontal sections (Lshaped), this is described in Section 6.3. Pipe insulation was generally tested at a standard dimension of 22 mm bore with 25-30 and 50 mm insulation thickness mounted on steel pipes. The 22 mm bore was chosen as it is the common pipe diameter produced and marketed by all the product groups for the building equipment application. The major part of the tests were run at an insulation thickness of 2530 mm because products in this range are produced and marketed by all product groups. The maximum insulation thickness that can be practically tested in the SBI is 75 mm (with the 22 mm bore). Jacketing products were tested on mineral wool substrates..

(12) 12. 3.1.1. Frame for mounting the pipe insulation. A steel frame with vertically mounted steel pipes having an outer diameter of 21.3 mm (ten on the large wing and five on the short wing) is used. Pipe insulation was fixed on the steel pipe according to the product task group specifications. The pipe insulation covered the entire steel pipe. Against the frame the 10 mm thick calcium silicate backing board was placed, according to prEN 13823.. 25. The distance between the backing board and the pipe insulation and the spacing between each pipe insulation was 25 mm, see Figure 3-2.. 25. backing board. 25. Figure 3-2. Schematic drawings of the mounting of the test specimen in the SBI in the case of 25 mm insulation thickness. The same principle can be applied to larger thicknesses up to 75mm. Front view to the left and top view to the right..

(13) 13. 4. Basic principles for the analysis. The reference scenario for linings is the Room/Corner Test, ISO 9705, (see Section 1.1). The work on linings showed that FIGRA (Room/Corner Test) and FIGRA (SBI) correlates with about 95 %. Further there is a broad agreement between the occurrence of flashover in the Room/Corner Test and the limit values of FIGRA (SBI) used in the Euroclasses. In contrast to linings the phenomenon of flashover due to the fire growth of the product in the Room/Corner Test cannot be the basic criterion for correlation or the basis for a future classification system for pipe insulation. Pipe insulation will in reality appear together with other construction products and add to the fire load already present. Only the fire load caused by pipe insulation is considered in this work. Thus from a fire safety point of view a flashover caused by pipe insulation only when tested in the Room/Corner Test would present a considerable hazard. Consequently, for pipe insulation we must use test data, which are more distinguishing than the flashover criterion when finding possible Euroclasses. Therefore, the correlation study between the Room/Corner Test and the SBI is based on FIGRA defined as in the SBI Test, i.e. as the maximum fire growth rate whenever it occurs during a test. The parameter is called FIGRARC PIPE. (for detailed definition see Appendix A3). In the Room/Corner Test, the test procedure was designed to fulfil the following aims: - The mounting must be representative of end-use. - The test procedure must give results that are insensitive to small variations in mounting, etc. - The test and mounting should be easy to perform. - The results should be reproducible and repeatable. - It must be possible to define classification criteria allowing for a large-scale test to be done in cases where the SBI Test will not give representative data. In the SBI Test, pipe insulation could theoretically be tested either as tubes or in flat shape as sheets. The tube shape was considered as the most appropriate form for a number of reasons: - The SBI Test needs to model the local ignition and fire growth for pipe insulation mounted in its end-use conditions. - Exposed surface area dynamics are different for tubular shapes than for flat surfaces (in reality, flames typically encompass the full circumference of an insulated pipe). - There is a greater available volume of material for a given thickness of insulation in tubular structures than in flat structures. Hence tubular shapes provide higher potential loadings compared to linings. - For some pipe insulation products it is impossible to manufacture flat specimens. In addition to these general justifications, the rationale for the specific selection of pipe diameter and thickness is set out in Section 3.1. Apart from the obvious practical reasons set out there, the fire performance parameters of the insulation thickness was covered. It is interesting to note that in the SBI Test, the surface area exposed to the burner increases with insulation thickness when maintaining a constant distance between the outer circumferences of adjacent insulated pipes. The relationship is shown in Figure 4-1. However, above 30 mm thickness, the graph levels out. The graph illustrates the potential problem in permitting the testing of lower insulation thickness, since considerably lower exposed surface areas would be involved, and the worst case in this respect is no longer maintained. For this reason the optimum would be to test products in the range of 25 mm thickness up to the maximum allowable in the SBI, 75 mm. This will produce data sets of maximum versatility..

(14) 14. 2.5. Area exposed to burner (m2). 2.0. 1.5. 1.0. 0.5. 0. 10. 20. 30. 40. 50. 60. 70. 80. 90. Thickness (mm). Figure 4-1. Relationship between area of pipe insulation sections exposed to burner and thickness of pipe insulation, for a given bore of 22 mm. Area exposed to burner means the area of the pipes that are within the extent of the burner dimension..

(15) 15. 5. Test results and data analysis. More than 25 large-scale tests and a large number of intermediate-scale SBI tests have been performed. In large-scale three different mounting techniques were studied. The products tested are representative of the market as well as showing a wide range of burning behaviour. Details of test results and mounting procedures are given in Appendix A1.. 5.1. Products tested. Products made of the materials listed in Table 5-1 were tested according to the SBI and ISO 9705 in this programme. For most of the products groups several variants were tested. Further product details are given in Appendix A1. Table 5-1. Product identification. Product Aluminium foil faced Cellular Glass (CG)*, pipe sections PVC film jacketing on mineral wool pipe sections as substrate Unfaced Polyethylene Foam (PEF) extruded tubes tubes made from cross-linked sheets Unfaced Flexible Elastomeric Foam (FEF) extruded tubes Aluminium foil faced Mineral Wool (MW), pipe sections Semi-rigid Polyurethane extruded pipe sections + PVC foil facing Unfaced Extruded Polystyrene foam (XPS), pipe sections Rigid Polyurethane / Polyisocyanurate Foam (PUR; PIR), pipe sections, unfaced or Aluminium foil faced Unfaced Melamine Foam (MF), pipe sections Unfaced Flexible Elastomeric Foam (FEF), sheets. M-no.. M1 M2 M3 M4. M5 M6 M7 M8 M9 M10 M11. (tested only for comparison with tubes). Unfaced Flax Fibre, pipe sections. M12. Expanded Polystyrene (EPS), unfaced or aluminium foil faced. M13. NOTE: EPS pipe insulation was tested at a thickness of 21 mm and for information the results can be found in Appendix A1.2. * Cellular glass is A1 according to deemed to satisfy list but was included to have a wider range of fire behaviour..

(16) 16. 5.2. Test data from ISO 9705 and the SBI (EN 13823). Table 5-3 and Table 5-4 give the main test results. Results from Room/Corner Tests in mounting version 3 and 4 can be found in Section 6.1. Individual SBI tests are presented in Appendix A1. The numbering of the products is on the form Mx:0y, where “x” represents one of the product groups in Table 5-1 and “y” represents a specific product from that group. The national classifications given for several European countries in Table 5-2 were provided by the supplier of the product. The individual product sample was not tested for the national class prior to the tests in this program. For definition of the parameters in the tables see Section 5.3 and A3. Table 5-2. National classification of the products tested. Product no. M1:01. National classification D. GB. F. NL. I. S. A. CH. DIN 4102 part 1. BS as indicated below. Essai par rayonnement. NEN as indicated below. UNI 9174+8457. BBR 5:511. Ö-Norm B 3800 part 1. VKF. A1. -. M0. Pass. -. -. -. 6.3. -. P III. B1, Q2, Tr1. 5.2. -. P III. B1, Q2, Tr1. 5.2. -. P III. B1, Q2, Tr1. 5.1. -. -. -. 5.3. Class 1. P II. B1, Q2, Tr1. 5(200°C).2. Class 1. P II. B1, Q2, Tr1. 5(200°C).2. (6064). M2:01 M3:01. B1 B1. Passes sources A+B. -. Class 2 (6065). (BS 476, part 12). M3:02. B1. Passes sources A+B. -. Class 2 (6065). (BS 476, part 12). M3:03. B1. Passes source A. -. Class 2 (6065). (BS 476, part 12). M3:05 * M3:06 * M3:07 * M4:01. B1. -. -. Class 2. M5:01. B1. -. -. Class 2. (6065) (6065). M5:02. B1. -. -. Class 2 (6065). M5:05 M5:06. B2 B2. M5:07. B2. Class 1. M2. -. Class 1. P II. B1, Q2, Tr1. 5(200°C).3. -. -. Class 1. P II. B1, Q2, Tr1. 5.2. M1 M1 M1 M1. Class 2. -. -. B1, Q2, Tr1 -. 5.3. M1. Class 1. -. -. -. -. (BS 476, part 7). Class 1 (BS 476, part 7). M6:01 *** M7:01 M8:01 M8:02 M8:03 M9:01. B2 B1 B1 B1 B1. Class 1 (BS 476, part 7). M9:02. -. Class 1. (6065). (BS 476, part 7). M9:03 *** M9:04 *** M10:01. B1. Class 0. (6065). M1. -. -. -. -. -. M1. Class 1. -. -. -. -. M1. Class 2. -. -. -. -. (BS 476, part 6+7). M11 ** M12:01 M13:01. B2 B1. M13:02. B2. Class A (BSI 3837-part 1). -. (6065) (6065). * Products are non-commercial prototypes added to get a broader basis. Thus no national classification is known. ** Not tested in Room/Corner Test. *** No national classification exists..

(17) 17. Table 5-3. ISO 9705 Room Corner Test results for mounting Version 1. TSP Peak HRR THR from prod not sm. excl. burner (kW/s) (m /s ) (MJ) (m2) (kW) M1:01 0.063 0.58 10 254 55 M2:01 1.4 135 67 3500 184 M3:01 0.83 14.4 174 1360 518 M3:02* 1.13 14.3 148 1310 700 M3:03* 4.9 21.8 182 1630 700 M3:05 0.59 8.3 248 2160 499 M3:06 0.47 5.7 131 959 235 M3:07 0.67 4.7 222 1200 698 M4:01 2.7 112 90 3080 310 M5:01 0.16 87.2 24 6630 107 M5:02 0.21 47.9 16 4980 86 M5:05 4.7 58.6 123 2877 594 M5:06 1.38 64 140 4630 361 M5:07 1.6 75 115 5640 339 M6:01 0.0 3.6 4.7 810 45 M7:01 6.7 132 39 5900 491 M8:01 0.10 0.91 36 441 68 M8:02 0.79 13 86 1760 475 M8:03* + 0.72 33 76 2730 700 M9:01 0.37 49 32 3330 89 M9:02 0.44 29 34 3150 83 M9:03 1.25 98 60 3540 131 M9:04 + 2.05 112 213 12400 295 M10:01 0.44 13.2 39 1580 83 M12:01* 7.8 68 52 471 900 * Flashover. The fire parameters are given only for the time period up to flashover. Therefore the values of THR and TSP are not directly comparable to the other results. + Insulation thickness = 50 mm NOTE: Explanation of parameters in Appendix A3. Product no. FIGRA. SMOGRA. RC PIPE. RC PIPE 2 2.

(18) 18. Table 5-4. SBI Test results on the products tested with vertical mounting, average values. Product no. FIGRA 0.2 (W/s) 0.0 421 459. FIGRA 0.4 SMOGRA THR600 (MJ) (W/s) (cm2/s2) 0.0 0.0 0.0 417 791 4.2 459 30 28. M1:01 M2:01 M3:01 M3:02 * M3:03 1184 1184 M3:04 M3:05 365 342 M3:06 190 158 M3:07 349 317 M4:01 676 666 M5:01 184 162 M5:02 185 157 M5:03 730 693 M5:05 1970 1970 M5:06 674 647 M5:07 612 600 M6:01 207 156 M7:01 2290 2290 M8:01 20 20 M8:02 66 66 M8:03 + 142 142 M9:01 393 292 M9:02 82 73 M9:03 281 249 M9:04 + 466 349 M10:01 160 0.0 M12:01 2400 2400 * Deleted due to uncertain data ** Corrected for smoke clogging + Insulation thickness = 50 mm. TSP600s (cm2) 21 286 228. LFS (edge). No. of tests 1 1 2 3 2 2 2 2 2 3 2 3 2 2 3 4 2. 59. 102. 822. 59 27 7.0 602 902 389 340** 342 376 689 0.0 2610 10. 14 10.6 15 21 5.5 5.1 16 56 9.0 13 1.5 13 1.4. 227 142 89 615 647 477 390** 731 294 1273 30 996** 66. no no no no no no no no no no no no no no no no no no no. 47 127. 8.5 18.3. 643 1572. no no. 3 3. 525 80. 4.0 2.7. 203 168. no no. 2 2. 578 618. 4.3 9.1. 216 510. no no. 3 3. 50 100. 1.1 140. 57 178. no no. 1 2. 5.3. Analysis of correlation between ISO 9705 and the SBI. 5.3.1. FIGRA. As earlier discussed in Chapter 4, the principle of using only flashover as point of hazard is not sufficient for pipe insulation. Instead we need to interpret the large-scale test data on a continuous basis. Using the same principle for calculating FIGRA in the ISO 9705 test as in the SBI we find the best correlation between the two test methods. For the Room/Corner Test the definition of FIGRARC PIPE is the following: Calculate FIGRA following the same principle as for the SBI, i.e. maximum of HRR/time. There is no averaging of the HRR data. The HRR threshold value for calculating FIGRA is 50 kW, no THR threshold value is used. See Appendix A3 for a more detailed explanation. For SBI the definition of FIGRA is according to the standard1. The standard defines two different THR threshold values, 0.2 MJ and 0.4 MJ. In the correlation the 0.2 MJ threshold is used for FIGRA < 500 and the 0.4 MJ threshold for FIGRA ≥ 500..

(19) 19. 3000 2. R = 0.93. FIGRA SBI (W/s). 2500. 2000. 1500. 1000. 500. 0 0. 2. 4. 6. 8. 10. FIGRARC PIPE (kW/s). Figure 5-1. FIGRA correlation between Room Corner Test and the SBI. In Figure 5-1 we see that the correlation is good and of the same order as for linings. We note that a large portion of the products on the market has values of FIGRA less than 1000 in the SBI Test.. 5.3.2. SMOGRA. We now consider the smoke production. One way to do that is to use the same definition for SMOGRA as for FIGRA for the Room/Corner Test. That is to use the maximum of the function SPR/time as the SMOGRA value. In analogy with FIGRARC PIPE we define SMOGRARC PIPE . This also means that the definitions are the same as for the SBI but with threshold SPR=0.3 m2/s, see Appendix A3. The correlation between ISO 9705 and the SBI is shown in Figure 5-2. There is a clear dependence but also one outlier (material M7:01) can be identified. This outlier should however not present any problems since the result is in the highest range in both tests..

(20) 20. 3000. 2. 2. SMOGRA SBI (cm /s ). 2500. 2000. 1500. 1000. 500. 0 0. 50. 100 2. 150. 2. SMOGRARC Pipe (m /s ). Figure 5-2. Correlation between SMOGRA SBI and SMOGRARC PIPE. Alternatively we can use the SMOGRA as defined in the forthcoming CEN standard for ISO 9705, defined as maximum SPR divided by the time when it occurs. This approach together with a total smoke production approach was tested but no improved correlation was found. It appears that the correlation of SMOGRA between the Room/Corner Test and the SBI Test is not similar in precision as for the FIGRA correlation. However, the agreement is good enough for identifying groups of fire performance and class limits for smoke that will produce similar classification in both tests..

(21) 21. 6. Sensitivity analysis. 6.1. Fire load. To investigate the effect of fire load in the tests two studies were made. One analysed the effect of the number of pipes in the Room Corner Test and one studied the effect of higher insulation thickness both in the SBI and the Room Corner Test.. 6.1.1. Variation of fire load by changing the number of pipes. A number of large-scale tests were performed with different amounts of pipe insulation mounted (with the same thickness). Three different product families were tested. In Figure 6-2 - Figure 6-4, results from different versions of mounting are shown. The different mountings are summarised in Table 6-1 and schematic pictures with perspective views of the mounting versions can be seen in Figure 6-1. Complete results are given in Table 6-2.. Table 6-1. The different versions of mounting used for the test in the Room/Corner Test. Version of mounting *. Fire load approximately. Comments. Surface tested compared to linings (approx.). Version 1. 90 m of pipe insulation. Standard version, see Figure 6-1 and Section A2.. 65 %. Version 3. 180 m of pipe insulation. Double fire load, see Figure 6-1 and Section A2.. 130 %. Like version 3 but with more 140 % tubes mounted in the ceiling close to the burner. An air gap of 10 cm between the pipe insulation and the ceiling was introduced to enhance flame spread, see Figure 6-1 and Section A2. * Version 2 was discarded being too similar to Version 1. Version 4. 190 m of pipe insulation.

(22) 22. Version 1. Version 3. Version 4 Figure 6-1. Schematic drawings of the different mounting versions used. The roof and left wall of the ISO 9705 room have been removed. The burner is placed in the innermost right corner of the room. See Figure 2-1 for room dimensions..

(23) 23 Product M3:01 was tested when mounted in versions 1, 3 and 4, see Figure 6-2. 1000. 800. HRR (kW). Version 1 Version 3 Version 4. Product HRR at flashover. 600. 400. 200. 0 0. 5. 10. 15. 20. Time (min). Figure 6-2. HRR data for product no M3:01 when tested in the Room/Corner Test according to mounting version 1, 3 and 4. The FIGRARC PIPE values are 0.83, 0.97 and 1.25 kW/s respectively. All three tests on M3:01 behave similarly during the first 7 minutes, with the Versions 3 and 4 producing a higher HRR than Version 1. At the 300 kW level both tests with the higher fire load produce a flashover while the Version 1 test peaks just above 500 kW. The Version 3 test has a high peak at the end of the 100 kW level which is not seen in the other two versions. Product M5:01 was tested when mounted in versions 1 and 3, see Figure 6-3. 1000 Version 1 Product HRR at flashover. HRR (kW). 800. Version 3. 600. 400. 200. 0 0. 5. 10. 15. 20. Time (min). Figure 6-3. HRR data for product no M5:01 when tested in the Room/Corner Test according to mounting version 1 and 3. The FIGRARC PIPE values are 0.16 and 0.23 kW/s respectively..

(24) 24. The difference between results from the two fire loads is small. At the early stage of test there is no difference at all. When the heat output from the burner is increased after 10 min the HRR from the product remains relatively small. Product M7:01 was tested according to mounting versions 1 (standard), 3 and 4. The results are shown in Figure 6-4. 1000 Version 1 800. HRR (kW). Version 3. Product HRR at flashover. Version 4. 600. 400. 200. 0 0. 5. 10. 15. 20. Time (min). Figure 6-4. HRR data for product M7:01 when tested in the Room/Corner Test according to mounting version 1, 3, and 4. The FIGRARC PIPE values are 6.7, 10 and 6.9 kW/s respectively. The initial peak HRR becomes higher with the higher fire load but still not enough to cause flashover at the early stages of the test. In addition the large surface area covered with tubes close to the burner, version 4, did not give increased flame spread in the early stages of the test. Instead a flashover occurred after increasing the heat output from the ignition source after 10 min. We conclude that the selected scenario for mounting in the Room/Corner Test is not very sensitive to variations in fire load caused by the increased number of pipes for the three types of products tested, i.e the corresponding variation in FIGRA is not excessive, see Table 6-2. The higher load scenarios are naturally more likely to exhibit flashover but this difference will mainly show during the higher burner level, as we can see in Figure 6-2 and Figure 6-4. The flashover phenomenon is not very important for the classification of pipe insulation, in contrast to surface materials where the flashover is the essential parameter. The strategy of this work is to classify pipe insulation according to heat release parameters, e.g. the FIGRA index and THR. This approach is very sensitive leading to robust results in terms of classification. An increase of fire load would not give any further information other than changing the class limits as the ranking order according to fire performance in most cases would prevail. See further Section 1.2 and Section 6.1.2..

(25) 25. Table 6-2. Room/Corner Test results for tests with different mounting versions. FIGRA. SMOGRA. RC PIPE. RC PIPE 2 2. (m /s ). THR from prod (MJ). TSP not sm. (m2). Peak HRR excl. burner (kW). M3:01 – ver. 1 0.83 M3:01 – ver. 3 * 0.97 M3:01 – ver. 4 * 1.25. 14.4 8.5 1.9. 174 139 102. 1360 1530 297. 518 700 700. M5:01 – ver. 1 M5:01 – ver. 3. 87.2 64. 24 78. 6630 11430. 107 158. 132 650 471. 39 185 84. 5900 11660 7050. 491 883 700. Product. (kW/s). 0.16 0.23. M7:01 – ver. 1 6.7 M7:01 – ver. 3 10 M7:01 – ver. 4 * 6.9. * Flashover NOTE: Calculations are made only to the time of flashover. Therefore the integrated values (THR, TSP) are sometimes low despite the higher fire load.. 6.1.2. Variation of the fire load by changing insulation thickness. A separate test series was conducted to analyse the influence of insulation thickness. In this series two products were tested, one thermoplastic, XPS, and one thermosetting, PIR. The products were tested at two thicknesses, 25 and 50 mm, with a 22 mm bore. They fall into the product groups M8 and M9 and are coded in Table 6-3. Tests were made both in the SBI and in the Room/Corner Test and the main results are presented in Figure 6-5 - Figure 6-8 and in Table 6-4 - Table 6-5 below. The mounting in Room/Corner Test was made according to the standard procedure Version 1. In the SBI the mounting was made with 25 mm spacing between the pipe insulation and the backing board, see Section 3.1. The results show that the fire performance parameters vary as a function of thickness. Thicker insulation produce higher values for both the XPS and the PIR insulation and the trend is consistent in both the SBI and the Room/Corner Test (see Figure 5-1). The fire performance parameters are in most cases approximately doubled due to doubled thickness. This suggests that a higher fire load due to higher thickness is more serious than a higher fire load due to a larger number of pipes with 25 mm thickness, as discussed in the section above. However, the data indicate that the thickness should be a parameter in any classification system.. Table 6-3. Products tested for thickness dependence. Product XPS – 25 mm XPS – 50 mm PIR – 25 mm PIR – 50 mm. M-no M8:02 M8:03 M9:03 M9:04.

(26) 26. 1000 Product HRR at flashover. HRR (kW). 800. XPS 25 mm. 600. XPS 50 mm 400. 200. 0 0. 5. 10. 15. 20. Time (min). Figure 6-5. HRR data for M8:02, M8:03 when tested in the Room/Corner Test at different thickness.. 1000 Product HRR at flashover. HRR (kW). 800. PIR 25 mm. 600. PIR 50 mm. 400. 200. 0 0. 5. 10. 15. 20. Time (min). Figure 6-6. HRR data for M9:03, M9:04 when tested in the Room/Corner Test at different thickness..

(27) 27. 100 XPS 25 mm test 1 XPS 25 mm test 2 XPS 25 mm test 3 XPS 50 mm test 1 XPS 50 mm test 2 XPS 50 mm test 3. HRR from Product (kW). 80. 60. 40. 20. 0 0. 500. 1000. 1500. Time (s). Figure 6-7. HRR data for M8:02, M8:03 when tested in the SBI at different thickness.. 50 PIR 25 PIR 25 PIR 25 PIR 50 PIR 50 PIR 50. HRR from Product (kW). 40. mm test 1 mm test 2 mm test 3 mm test 1 mm test 2 mm test 3. 30. 20. 10. 0 0. 500. 1000. 1500. Time (s). Figure 6-8. HRR data for M9:03, M9:04 when tested in the SBI at different thickness..

(28) 28. Table 6-4. Room/Corner Test results for tests with different thickness. Product. FIGRA. SMOGRA. RC PIPE. RC PIPE 2 2. (m /s ). THR from prod (MJ). TSP not sm. (m2). Peak HRR excl. burner (kW). 13 33 98 112. 86 76 60 213. 1760 2730 3540 12400. 475 700 131 295. (kW/s). XPS – 25 mm XPS – 50 mm * PIR – 25 mm PIR – 50 mm. 0.79 0.72 1.25 2.05. * Flashover Table 6-5. SBI Test results for tests with different thickness. Averages values. Product. FIGRA FIGRA 0.4 SMOGRA 0.2 (W/s) (W/s) (cm2/s2). THR600 (MJ). TSP600s (cm2). LFS (edge) No. of tests. XPS – 25 mm XPS – 50 mm PIR – 25 mm PIR – 50 mm. 66 142 281 466. 8.5 18.3 4.3 9.1. 643 1572 216 510. No No No No. 6.2. 66 142 249 349. 47 127 578 618. 3 3 3 3. Influence of threshold values for FIGRA in the SBI. The threshold values used in the SBI Test for calculating FIGRA may influence the value considerably if a product has the tendency to ignite fast but produce a relatively small peak of HRR. The effect of different THR threshold values on test data is shown in Table 6-6. Table 6-6. The effect of selecting different threshold values for calculating FIGRA (SBI). For comparison the value of FIGRA 0,2 MJ value is set 100 %. Product no M1:01 M2:01 M3:01 M3:03 M3:05 M3:06 M3:07 M4:01 M5:02 M5:03 M5:04 M5:05 M5:06 M5:07 M6:01 M7:01 M8:01 M9:01 M9:02 M10:01 M12:01. *not computed. FIGRA 0.2 FIGRA 0.3 FIGRA 0.4 FIGRA 0.5 FIGRA 0.6 No of tests 100% 100% 100% 100% 100% 1 100% 100% 99% 98% 96% 1 100% 100% 100% 100% 100% 2 100% 100% 100% 100% 100% 3 100% * 94% * * 2 100% * 83% * * 2 100% * 91% * * 2 100% 100% 98% 97% 92% 2 100% 88% 85% 82% 82% 3 100% 99% 95% 90% 86% 2 100% 94% 88% 79% 71% 2 100% 100% 100% 100% 100% 3 100% * 96% * * 2 100% * 98% * * 2 100% 94% 75% 49% 25% 3 100% 100% 100% 100% 100% 1 100% * 100% * * 2 100% * 74% * * 2 100% * 89% * * 2 100% * 0% * * 1 100% * 100% * * 2.

(29) 29. It is clear that most data is not changed when the threshold value is varied. However, there are some three categories of products that are affected, series 5, series 6 and series 9. The products concerned all show very fast ignition and a subsequent rather small peak HRR. However, it is not critical for the correlation with the Room/Corner Test which threshold value is selected. For pipe insulation it is therefore recommended to use the work on linings where a combination of thresholds of 0,2 MJ and 0,4 MJ is used.. 6.3. Vertical and horizontal mounting in the SBI. A small number of experiments where made to investigate if horizontal flame spread will have a major influence of SBI Test results, see Figure 6-9. The special testing frame (Section 3.1.1) could not be used in this case and pipe hangers were therefore used. The results are presented in Figure 6-10 to Figure 6-12 and summarised in Table 6-7.. Figure 6-9. Example of horizontal and vertical mounting in the SBI. In this particular case pipe hangers were used..

(30) 30. 50 M8:01 Vert test1 M8:01 Vert test2 M8:01 vert+hor test1 M8:01 vert+hor test2. HRR from Product (kW). 40. 30. 20. 10. 0 0. 500. 1000. 1500. Time (s). Figure 6-10. HRR data for experiments where horizontal flame spread was compared to vertical flame spread, M8:01 series.. 25 M9:01 Vert test1 M9:01 Vert test2 M9:01 Vert+hor test1. HRR from Product (kW). 20. 15. 10. 5. 0 0. 500. 1000. 1500. Time (s). Figure 6-11. HRR data for experiments where horizontal flame spread was compared to vertical flame spread, M9:01 series..

(31) 31. 10 M9:02 Vert test1 M9:02 Vert test2 M9:02 vert+hor test1. HRR from Product (kW). 8. 6. 4. 2. 0 0. 500. 1000. 1500. Time (s). Figure 6-12. HRR data for experiments where horizontal flame spread was compared to vertical flame spread, M9:02 series. Table 6-7. Average FIGRA values for the two mounting procedures. Product. M8:01 M9:01 M9:02. Vertical mounting FIGRA 0.2 FIGRA 0.4 (W/s) (W/s) 20 20 393 292 82 73. Vert + hor mounting FIGRA 0.2 FIGRA 0.4 (W/s) (W/s) 84 84 275 157 69 52. Although these products show a spread in results we can identify two types of burning behaviour. For the non-thermoplastic material, series M9, there is a tendency to a smaller peak HRR for the vertical/horizontal configuration. This could be due to a change in airflow pattern caused by the horizontal sections. Series M8 represents a thermoplastic product. In this case the peak HRR is larger for the test that includes horizontal mounting. The reason for this was mainly that pipe hangers had to be used for mounting the pipes and insulation, see Figure 6-9. The pipe hangers and the bend itself kept melting material from sliding down to the floor. Thus a larger amount of material was burning during the horizontal test compared to the vertical orientation. We see from the horizontal tests that some difference in results can be obtained. Thermoplastic products can show different results but that was mostly due to the stopping of material from melting to the floor rather than horizontal flame spread. The test will also be more expensive and difficult to perform with horizontal mounting; the complex mounting is likely to decrease reproducibility. As the tests with vertical mounting have proven to correlate sufficiently with the full-scale tests there is no clear reason to complicate and raise the price of the test. Therefore the conclusion is that mounting should be vertical only..

(32) 32. 6.4. Attachment of the pipe insulation in the SBI. The experiments with the different mounting techniques revealed that results would be influenced for thermoplastic materials as a function of number and location of pipe hangers during a test. Data from non-thermoplastic products is not affected. The pipe hanger can keep material in position instead of allowing it to melt to the floor. The amount of material kept in position and thus burning will vary with the number and location of pipe hangers and is expected to be very sensitive. This effect seen in the SBI is not expected to be so pronounced in large scale. Therefore pipe hangers should not be introduced in the SBI as it will only decrease the repeatability without adding any useful information. Products that may slide down during test can be fixed to the pipe at the top of the specimen. For intermediate positions a wire or a tape can be used for cases where the pipe insulation sections are shorter than the specified specimen length.. 6.5. Comparison of test data of material in tube form with flat form in the SBI. Insulation of very large and rectangular ducts may include pipe insulation material in such quantities that it is comparable with an ordinary wall lining and therefore the product should be tested as such. It is therefore of interest to analyse test data from the same material tested both as a lining, flat material, and as a tube, circular material, see Figure 6-13 - Figure 6-17. The results are summarised in Table 6-8. Some of the products tested were not available in flat form and had to be artificially constructed from tube material. This can have a quite big influence on the behaviour of the product since some properties might change. 500 Sheet test1 Sheet test2 Tube test1 Tube test2 Tube test3. HRR from Product (kW). 400. 300. 200. 100. 0 0. 500. 1000. 1500. Time (s). Figure 6-13. HRR data for tube (25 mm) and sheet (19 mm) tests on the same material (M3:03)..

(33) 33. 50 Sheet test1 Sheet test2 Tube test1 Tube test2. HRR from Product (kW). 40. 30. 20. 10. 0 0. 500. 1000. 1500. Time (s). Figure 6-14. HRR data for tube (25 mm) and sheet (19 mm) tests on the same material (M5:06).. 70 Sheet test1 Tube test1 60. Tube test2. HRR from Product (kW). 50. 40. 30. 20. 10. 0 0. 500. 1000. 1500. Time (s). Figure 6-15. HRR data for tube (25 mm) and sheet(19 mm) tests on the same material (M5:07)..

(34) 34. 25 Sheet test1 Sheet test2 Tube test1 Tube test2 Tube test3. HRR from Product (kW). 20. 15. 10. 5. 0 0. 500. 1000. 1500. Time (s). Figure 6-16. HRR data for tube (25 mm) and sheet (19 mm) tests on the same material (M11:01).. 25 Sheet test1 Sheet test2 Tube test1 Tube test2. HRR from Product (kW). 20. 15. 10. 5. 0 0. 500. 1000. 1500. Time (s). Figure 6-17. HRR data for tube (25 mm) and sheet (19 mm) tests on the same material (M11:02)..

(35) 35. Table 6-8. Average FIGRA values for sheet and tube tests on five materials. Product. M3:03 M5:06 M5:07 M11:01 M11:02. FIGRA 0.2 (W/s) 686 832 676 133 202. Sheet FIGRA 0.4 (W/s) 638 709 675 44 97. FIGRA 0.2 (W/s) 1184 674 612 106 128. Tube FIGRA 0.4 (W/s) 1184 647 600 81 107. Of the five materials in the table above, M3:03 is the only thermoplastic product, the other four are non-thermoplastic products. Some interesting differences are seen for these test series. For the non-thermoplastic materials the sheet form shows a faster growth rate of the HRR and then a faster decline to a lower level of HRR than the tube form. FIGRA for the sheet material is more sensitive to a change in the threshold value. The subsequent large HRR from the tubes can be explained by the fire load. The tubes will supply approximately 50 % more material as fuel compared to a sheet if the tubes are mounted as specified in Section 3.1. This will result in a long lasting fire on a higher level compared to the sheets. For the thermoplastic product the difference in HRR is very significant, see Figure 6-13. The fire growth rate, FIGRA, when testing the tubes is much larger and the tubes produce significantly more heat than the sheets. It appears that melting and burning combined produce very different results depending on whether the material is in tube or in flat form. In this case a serious underestimate of the fire hazard could have been the case if flat form data were used. The sheets were mounted with nails and washers, preventing it from sliding down onto the burner. Table 6-8 also shows how FIGRA varies with different threshold values. FIGRA for the sheet material is more sensitive to a change in the threshold value than FIGRA for the tube material. This limited investigation shows that there will be different results if a product is tested in flat form compared to the tube form. Test data will also not be so sensitive to small changes in calculation procedures when using the tube form. As shown earlier there is a good correlation between the Room/Corner Test and the SBI for the tube form. This is partly due to the fact that SBI test in tube form model the local ignition and flame spread in the reference scenario. Another important argument is that for some products it may be very difficult to prepare flat samples. For these reasons and the fact that tube form is closer to reality, pipe insulation should be tested in tube form in the SBI procedure..

(36) 36. 7. Conclusions. It may be concluded that 1. For the purpose of fire testing and classification of pipe insulation in a common European system both the Room/Corner Test, ISO 9705, and the SBI Test, EN 13823, can be used. 2. The Room/Corner Test can be used as reference scenario for the burning behaviour of pipe insulation requiring only that the products be mounted as in end-use conditions outlined in this report. In all other aspects the standard ISO 9705 apply. 3. In the Room/Corner Test, the pipe insulation should be mounted along the walls and ceiling, according to “Version 1” in this report. Test results with higher fire loading due to the number of pipes do not add to the information about a products fire performance, insulation thickness does. 4. In the SBI, the pipe insulation should be mounted in tube form on vertical steel pipes covering the same area as a lining. This will represent end-use conditions. Testing the products in sheet form gives a lower fire load, less stable results and does not allow the same level of fire performance differentiation as for tube tests. As for the Room/Corner test, insulation thickness is a factor. 5. Test data on variations in mounting and calculation procedure indicate that the results are not sensitive to small variations. This is important for a system with many users. However, if further work is required then the reproducibility of the SBI procedure should be studied. 6. The test results in the Room/Corner Test and the SBI Test for pipe insulation correlate equally well as those for lining products (R2 = 0.93) and the great number of tests provides a solid basis for the correlation. 7. Starting from the reference scenario performance requirements can be applied for pipe insulation. These performance requirements can then be identified as limit values for Euroclasses in the SBI. Thus a system for testing and classification of pipe insulation can be created, which is believed to fulfil industry and regulatory requirements..

(37) 37. 8. References. 1. European Standard – Reaction to fire tests for building products – Building products excluding floorings exposed to the thermal attack by a single burning item. EN 13823:2002 (E).CEN Central Secretariat, Brussels 2002. 2. International Standard – Fire tests -- Full-scale room test for surface products. ISO 9705:1993(E). International Organization for Standardization, Geneva, 1993. 3. 1982 Annual book of ASTM standards, part 18 - Proposed Method for Room Fire Test of Wall and Ceiling Materials and Assemblies. 4. Surface Products: Room Fire Test in Full Scale, NT FIRE 025, Helsinki 1986.. 5. Sundström, B., Full-Scale Fire Testing of Surface Materials. Measurements of Heat Release and Productions of Smoke and Gas Species, Technical Report SP-RAPP 1986:45, BORÅS 1986. 6. EUREFIC Seminar Proceedings, Interscience Communications Ltd, London, ISBN 0 9516320 19.. 7. Östman, B., Nussbaum, R., National Standard Fire Tests in Small Scale Compared with the Full-Scale ISO Room Test, Träteknikcentrum Rapport I 870217. 8. J. Axelsson, B. Sundström, U. Rohr, Development of a common European system for fire testing and classification of pipe insulation, Ninth International Interflam Conference Proceedings, Volume 1, p485-494, Interscience communications Ltd, ISBN 0 9532312 8 3..

(38) 38. Appendix.

(39) 39. A1 Detailed test data and product parameters Large-scale tests according to ISO 9705 were performed. The test procedure was identical to the procedure used when developing the Euroclasses for linings but the mounting was taking account of the end-use conditions for pipe insulation. Version 1 of the mounting in Table A1-1 below corresponds to the selected solution for mounting. Table A1-1. Room/Corner Test data on the products. All mounting versions. Product no. FIGRA. FIGRA. SMOGRA. SMOGRA. RC PIPE. RC. RC. (kW/s). (kW/s). RC PIPE 2 2. M1:01 M2:01 M3:01 M3:01* M3:01* M3:02* M3:03* M3:04 M3:05 M3:06 M3:07 M4:01 M5:01 M5:01 M5:02 M5:03 M5:05 M5:06 M5:07 M6:01 M7:01 M7:01 M7:01* M8:01 M8:02 M8:03*+ M9:01 M9:02 M9:03 M9:04+ M10:01 M12:01*. 0.063 1.4 0.83 0.97 1.25 1.13 4.9 0.28 0.59 0.47 0.67 2.7 0.16 0.23 0.21 1.7 4.7 1.38 1.6 0.0 6.7 10 6.9 0.10 0.79 0.72 0.37 0.44 1.25 2.05 0.44 7.8. 0.06 0.28 0.81 0.84 1.08 0.66 0.63. (m /s ) 0.58 135 14.4 8.5 1.9 14.3 21.8. 2. 2. (m /s ) 0.54 20 6.2 8.5 1.9 6.9 4.2. THR from prod (MJ) 10 67 174 139 102 148 182 50 248 131 222 90 24 78 16 273 123 140 115 4.7 39 185 84 36 86 76 32 34 60 213 39 52. TSP (m2) 254 3500 1360 1530 297 1310 1630 590 2160 959 1200 3080 6630 11430 4980 7960 2877 4630 5640 810 5900 11660 7050 441 1760 2730 3330 3150 3540 12400 1580 471. Peak HRR excl. burner (kW) 55 184 518 700 700 700 700 118 499 235 698 310 107 158 86 537 594 361 339 45 491 883 700 68 475 700 89 83 131 295 83 900. Version of mounting. 1 1 1 3 4 1 1 3 ** 0.54 8.3 4.3 1 0.21 5.7 2.1 1 0.61 4.7 4.7 1 0.33 112 14.2 1 0.16 87.2 25.9 1 3 0.11 47.9 18.1 1 3 ** 4.7 58.6 54 1 0.50 64 17 1 1.60 75 71.2 1 0.0 3.6 1.7 1 6.5 132 130 1 3 4 0.10 0.91 0.60 1 0.40 13 12.5 1 0.72 33 29.3 1 0.14 49 15 1 0.12 29 14 1 0.20 98 15.8 1 0.43 112 31.0 1 0.10 13.2 9.3 1 6.8 68 68 1 * Flashover. The fire parameters are given only for the time period up to flashover. Therefore the values of THR and TSP are not directly comparable to the other results. ** Not tested in Version 1 or 4. + Insulation thickness = 50 mm NOTE: Explanation of parameters in Appendix A3.. We can see from table 1 that a wide range of products and burning behaviour is represented. The data includes products causing flashover both on the 300 kW and 100 kW burner levels, corresponding to product HRR of 700 kW and 900 kW respectively, as well as very small fires of 45 kW maximum. SBI tests were performed on all products. In Table A1-2 product details and mounting details in the SBI are presented. Results from all vertical tests are presented in Table A1-3..

(40) 40. Table A1-2. Product and test details including mounting in the SBI. (continued on next page) Product no. Density Thickness (mm). Facing. M1:01. 120. 27. M2:01. 65-70. 0.35 (PVC) PVC 30 (glass wool). M3:01. 33. 26. No. M3:03. 32. 25. No. M3:05 M3:06 M3:07 M4:01. * * * 33. 25 25 25 26. No No No No. M5:01. 55. 25. No. M5:02. 55. 25. No. M5:03. 78. 25. No. M5:05. 65. 25. No. M5:06 M5:07 M6:01. 78 65 60. 25 25 25. No No Al foil. M7:01. 67. 20. PVC. M8:01. 33. 25. No. M8:02. 34. 25. No. M8:03. 34. 50. No. M9:01. 33. 25. No. M9:02. 33. 25. M9:03. 40. 25. Al, 40 µ No. M9:04. 40. 50. No. M10:01. 8-11. 25. No. M11:01. 60. 25. No. M11:02. 60. 25. No. M12:01. 225. 25. No. M13:01. 15. 21. No. M13:02. 15. 21. Al Foil. * Confidential (prototype). Al foil. Fixing. Steel straps Pipe hanger1 Pipe hanger2 Pipe hanger4 Frame3 Frame3 Frame3 Pipe hanger2 Pipe hanger1 Pipe hanger4 Pipe hanger1 Pipe hanger4 Frame3 Frame3 Metal wire Pipe hanger1 Metal wire Metal wire Metal wire Metal wire Metal wire Metal wire Metal wire Metal wire Pipe hanger1 Pipe hanger1 Metal wire Al tape at hor.joints Al tape at hor.joints. Vertical joints. Ends/horizontal joints. Lengths. 0.6 + 0.6 + 0.3 m PVC 1.5 m. Position. Type. Random. Not covered. Facing backing board n.a.. Butt, not adhered PVC overlapping n.a.. Not covered. n.a.. n.a.. Not covered. n.a. n.a. n.a. n.a.. n.a. n.a. n.a. n.a.. Not covered Not covered Not covered Not covered. n.a.. n.a.. Not covered. n.a.. n.a.. Not covered. n.a.. n.a.. Not covered. n.a.. n.a.. Not covered. 0.2 + 1.1 + 0.2 m 1.5 m. n.a. n.a. Random. n.a. n.a. Butt, not adhered PVC overlapping. Not covered Not covered Not covered. 1.5 m 1.5 m 1.5 m. Not covered. 0.2 + 1.1 + 0.2 m. Not covered. 1+0.5 m. Not covered. Not covered. 0.6 + 0.6 + 0.3 m 0.6 + 0.6 + 0.3 m 1+0.5 m. Not covered. 1+0.5 m. Not covered. 1+0.5 m. Not covered. 1+0.5 m. n.a.. Butt, not adhered Butt, not adhered Butt, not adhered Butt, not adhered Butt, not adhered Butt, not adhered Butt, not adhered Butt, not adhered n.a.. Not covered. n.a.. n.a.. Not covered. Random. Butt, not adhered Butt, not adhered. Not covered Al tape at horizontal joints. 0.2 + 1.1 + 0.2 m 0.2 + 1.1 + 0.2 m 0.6 + 0.6 + 0.3 m 1 + 0.5 m. n.a.. Al tape at horizontal joints. 1 + 0.5 m. Facing backing board Random Random Random Faced to burner Faced to burner Random Random Random. Facing backing board n.a.. Uncovered but PVC liner cover complete pipe. Not covered. 0.75 m + 0.75 m 1.5 m 1.5 m 1.5 m 1.5 m 0.75 m + 0.75 m 0.2 + 1.1 + 0.2 m 1.5 m. Not covered.

(41) 41 1. Two pipe hangers (22 mm) holding the steel pipe, at approximately 200 mm and 1300 mm height. No spacing between pipe insulation sections at pipe hangers. 2 Three pipe hangers (22 mm) holding the steel pipe, two at the ends of the pipe and one at approximately 750 mm height. Insulation discontinuous at the middle hanger with a gap of 25 mm, see Figure A1-1 below. 3 Seamless tubes mounted on pipe frame, see Figure 3-2. 4 Two pipe hangers (70 –75 mm) holding the steel pipe and the insulation at approximately 200 mm and 1300 mm height.. 25. 25. 25. Figure A1-1. Some products were mounted with pipe hangers and an insulation gap..

(42) 42 Table A1-3. SBI data for all tube tests with vertical mounting. (continued on next page) Product no. M1:01 M2:01 M3:01 - 1 M3:01 - 2 M3:02 * M3:03 - 1 M3:03 - 2 M3:03 - 3 M3:04 M3:05 - 1 M3:05 - 2 M3:06 - 1 M3:06 - 2 M3:07 - 1 M3:07 - 2 M4:01 - 1 M4:01 - 2 M5:01 - 1 M5:01 - 2 M5:02 - 1 M5:02 - 2 M5:02 - 3 M5:03 - 1 M5:03 - 2 M5:05 - 1 M5:05 - 2 M5:05 - 3 M5:06 - 1 M5:06 - 2 M5:07 - 1 M5:07 - 2 M6:01 - 1 M6:01 - 2 M6:01 - 3 M7:01 - 1 M7:01 - 2 M7:01 - 3 M7:01 - 4 M7:01 - 5 M7:01 - 6 M8:01 - 1 M8:01 - 2 M8:02 - 1 M8:02 - 2 M8:02 - 3 M8:03 - 1 + M8:03 - 2 + M8:03 - 3 + M9:01 - 1 M9:01 - 2 M9:02 - 1 M9:02 - 2 M9:03 - 1 M9:03 - 2 M9:03 - 3 M9:04 - 1 + M9:04 - 2 +. FIGRA 0.2 (W/s) 0 421 395 523. FIGRA 0.4 (W/s) 0 417 395 523. SMOGRA (cm2/s3) 0 791 28 32. THR600 (MJ) 0 4.2 28 28. TSP600 (cm2) 21 286 231 225. 1304 1402 846 355 375 190 190 349 349 688 664 164 203 186 185 183 753 707 2187 1928 1795 638 710 658 565 144 254 223 2401 2250 2325 2186 16 24 62 71 64 157 124 144 373 413 86 77 273 277 292 432 563. 1304 1402 846. 61 60 56 68 59 25 29 7.0 7.0 620 584 887 916 404 362 401 340*** 340*** 395 311 321 352 400 746 632 0.0 0.0 15 2805 2600 2369 2652 9.0 11 36 53 53 139 105 137 559 490 93 66 637 544 453 511 712. 104 112 90 13 15 12 9.2 15 15 23 19 5.4 5.6 5.2 5.1 4.9 15 16 59 57 52 6.9 11 12 13 1.5 1.5 1.4 13 13 12 14 1.0 1.7 6.7 10 8.8 20 17 18 3.0 5.0 2.7 2.6 4.7 4.1 4.1 7.9 11. 823 890 752 240 214 159 124 86 91 663 567 650 644 477 459 495 410*** 370*** 766 711 715 230 357 1255 1291 32 26 31 960 *** 1037 990 996 55 77 435 791 704 1770 1399 1547 195 210 206 130 244 193 212 407 707. 351 333 163 153 314 319 673 659 141 182 155 157 159 716 670 2187 1928 1795 594 700 658 541 118 193 156 2401 2250 2325 2186 16 24 62 71 64 157 124 144 251 333 72 73 240 216 292 290 424.

(43) 43. continued… Product no. M9:04 - 3 + M10:01 - 1 M11:01 - 1** M11:01 - 2** M11:01 - 3** M11:02 - 1** M11:02 - 2** M12:01 - 1 M12:01 - 2 M13:01 - 1 ++ M13:02 - 1 ++. FIGRA 0.2 (W/s) 403 160 121 103 94 125 130 2140 2660 61 92. FIGRA 0.4 (W/s) 334 0 94 73 75 100 114 2140 2600 61 92. * Uncertain data, deleted ** Not tested in the Room/Corner Test *** Corrected for smoke clogging + Insulation thickness = 50 mm ++ Insulation thickness = 21 mm. SMOGRA (cm2/s3) 631 50 930 860 780 730 800 125 74 11 5.7. THR600 (MJ) 8.9 1.1 5.0 4.2 5.0 4.5 4.2 146 134 9 7.6. TSP600 (cm2) 416 57 1090 900 780 760 990 199 156 35 43.

(44) 44. A1.2 EPS test Additional tests were made on EPS pipe insulation of 21 mm thickness. For information these test results are presented in Table A1-4 to Table A1-5 and Figure A1-2 to Figure A1-3 below. The product had a nominal thickness of 21 mm, a nominal inner diameter of 21 mm and a nominal density of 15 kg/m3. Table A1-4. ISO 9705 Room Corner Test results for M13:01 (EPS), mounting Version 1. Product. M13:01. FIGRA RC PIPE (kW/s) 0.07. SMOGRA RC PIPE (m2/s2) 0.5. TSP not sm. (m2) 310. THR from prod (MJ) 21. Peak HRR excl. burner (kW) 63. Table A1-5. SBI Test results for M13:01 (EPS). Product M13:01 M13:02. FIGRA 0.2 (W/s) 61 92. FIGRA 0.4 (W/s) 61 92. SMOGRA (cm2/s3) 11 5.7. THR600 (MJ) 8.9 7.6. TSP600 (cm2) 35 43. 1000 Product HRR at flashover. HRR (kW). 800. 600. 400. 200. 0 0. 5. 10 Time (min). 15. Figure A1-2. HRR data from full-scale test on EPS pipe insulation.. 20.

(45) 45. 5. 2. SPR (m /s) 60 s average. 4. 3. 2. 1. 0 0. 5. 10 Time (min). 15. 20. Figure A1-3. SPR data (60 s averaged) from full-scale test on EPS pipe insulation..

(46) 46. A2 Room/Corner Test mounting details Four different mounting configurations were tried. The version chosen for the main testing is called standard Version 1 and is shown in Figure A2-1 to Figure A2-3. Version 2 was very similar to Version 1 and was discarded early in the project since it was considered to provide very little information other than Version 1. The Versions 3 and 4 represent higher fire loads than the selected standard version, Figure A2-4 to Figure A2-6. Versions 3 and 4 contains about 100 % more material and especially Version 4 includes large areas of tubes in the ceiling for flame spread. See Section 6.1 for further discussion on the choice of mounting and perspective pictures. The drawings in this chapter are for the case of 25 mm insulation thickness but the same principles can be applied to larger thickness, keeping the spacing between insulation surfaces to 25 mm.. Figure A2-1. Mounting Version 1, top view..

(47) 47. Figure A2-2. Mounting Version 1, detail of corner..

(48) 48. Figure A2-3. Mounting Version 1, front and side views..

(49) 49. Figure A2-4. Mounting Version 3, top view..

(50) 50. Figure A2-5. Mounting Version 3, front and side views..

(51) 51. Figure A2-6. Mounting version 4, top view. The pipe insulation was mounted with an air gap of 10 cm between the ceiling and the upper surface of the pipe insulation..

(52) 52. Figure A2-6. Mounting version 4, front and side views..

(53) 53. A3 Calculation of test parameters A3.1 Room/Corner Test parameters. Parameter. Explanation. Test start. 00:00 (min:s) ignition of burner.. End of test. 20:00 (min:s) after test start.. Peak HRR excl. burner, kW. Peak Heat Release Rate between test start and end of test, excluded contribution from ignition source.. Peak SPR, m2/s. Peak Smoke Production Rate between test start and end of test.. THR from prod, MJ. Total heat release from test start until end of test, excluded contribution from ignition source.. TSP, m2. Total smoke production (non-smoothed) from test start until end of test.. FIGRARC PIPE, kW/s. FIre Growth RAte Pipe index is defined as the highest value of the quotient between HRR excluding the contribution of ignition source and time. Threshold value 50 kW.. FIGRARC, kW/s. FIre Growth RAte index is defined as the maximum of HRR excluding burner divided by the time at which it occurs.. SMOGRARC, m2/s2 *. SMOke Growth RAte index is defined as maximum of SPRsmooth (see below) divided by the time at which it occurs and multiplied by 1000. Threshold value SPRsmooth 0.3 m2/s.. SMOGRARC PIPE, m2/s2 *. SMOke Growth RAte index is defined as highest value of the quotient between SPRsmooth and time, multiplied by 1000. Threshold value SPRsmooth 0.3 m2/s.. NOTE: All test parameters above are calculated only for the time period up to flashover. * Further information see below. Calculation of SPRsmooth SPRsmooth (t) is the average of SPR(t) over 60 seconds, calculated as in equation (1).. SPRsmooth (t ) =. SPR(t − 30s ) + SPR(t − 27 s ) + ... + SPR(t + 27 s) + SPR(t + 30s ) …(1) 21. During flashover, the first and the last minute of a test the calculation of SPRsmooth according to equation (1) does not apply, as the required 21 records are not available. For those cases the procedures given below apply. Beginning of test:.

(54) 54 For t = 0 s: SPRsmooth = 0 m2/s For t = 3 s: SPRsmooth = SPR average over the period ( 0s…6s) For t = 6 s: SPRsmooth = SPR average over the period (0s…12s) For t = 27s: SPRsmooth = SPR average over the period (0s…54s) For t = 30s: SPRsmooth is calculated according to equation (1) End of test: SPRsmooth is calculated according to equation (1) until the data point SPR(t+30s) is one record from the flashover point or, if there is no flashover, until SPR(t+30s) = 19min 57s. This means that there are no values of SPRsmooth given for the last 10 records, 30s, of a test. Flashover faster than 60s: In cases of flashover in shorter time than 60s, SPRsmooth is calculated as the time average over the entire actual time interval. The corresponding value of t is taken at the middle of the time interval. Comments on calculation of SMOGRA When maximum SPRsmooth occurs at the end of a test (at flashover or at 20 min) then the test time t is taken as the time when the last calculation of SPRsmooth was performed. For flashover time shorter than within 60s there is only one value of SPRsmooth and a corresponding t; these are use to calculate SMOGRA. Correction for smoke produced by the ignition burner is small and can be neglected. If the smoke production from the product is very small the index may become uncertain and be influenced by the smoke production from the burner. Therefore if the maximum SPRsmooth is less than 0,3 m2 s-1 SMOGRA is set to zero..

(55) 55. A3.2 SBI Test parameters Parameter. Explanation. Test start. Start of data collection.. End of test. 26:00 (min:s) after test start.. HRR30s, maximum, kW. Peak Heat Release Rate between ignition of the main burner and end of test (burner heat output excluded), as a 30 seconds running average value.. SPR60s, maximum, m2/s. Peak Smoke Production Rate between ignition of the main burner and end of test, as a 60 seconds running average value. Calculated similarly as in Room Corner.. FIGRA 0.2, Ws-1. FIre Growth RAte index is defined as the maximum of the quotient HRR30s(t)/(t-300s), multiplied by 1000. Threshold THR = 0.2 MJ. FIGRA 0.4, Ws-1. FIre Growth RAte index is defined as the maximum of the quotient HRR30s(t)/(t-300s), multiplied by 1000. Threshold THR = 0.4 MJ. SMOGRA, cm2s-2. SMOke Growth RAte index is defined as the maximum of the quotient SPR60s(t)/(t-300s), multiplied by 10 000.. THR600s, MJ. Total heat release of the sample during 300 s ≤ t ≤ 900 s. TSP600s, m2. Total smoke production of the sample during 300 s ≤ t ≤ 900 s. Additional information on the calculations in the SBI procedure can be found in1..

(56) 56. A4 ISO 9705 HRR and SPR plots The Heat Release Rate (HRR) plots are all excluding the burner contribution, which is 100 kW during the first 10 min and 300 kW during 10-20 minutes. The Smoke Production Rate (SPR) plots are 60-s averaged. The plots presented here are from tests with mounting version 1 except M3:04 and M5:03 which are only tested in version 3. The figures have no figure text but they can be identified by the axis and by the product code above each plot..

(57) 57. M 1.01 CT 1000 Product HRR at flashover. HRR (kW). 800. 600. 400. 200. 0 0. 5. 10. 15. 20. Time (min). M 1.01 CT 20. 2. SPR (m /s) 60 s average. 15. 10. 5. 0 0. 5. 10 Time (min). 15. 20.






(63) 63. M3.05 CT. 1000 Product HRR at flashover. HRR (kW). 800. 600. 400. 200. 0 0. 5. 10 Time (min). 15. 20. M3.05 CT. 20. 2. SPR (m /s) 60 s average. 15. 10. 5. 0 0. 5. 10 Time (min). 15. 20.











(74) 74. M7:01 CT. 1000 Product HRR at flashover. HRR (kW). 800. 600. 400. 200. 0 0. 5. 10 Time (min). 15. 20. M7.01 CT. 30 25. 2. SPR (m /s) 60 s average. 20 15 10 5 0 0. 5. 10 Time (min). 15. 20.




(78) 78. M10.01 CT. 1000 Product HRR at flashover. HRR (kW). 800. 600. 400. 200. 0 0. 5. 10 Time (min). 15. 20. M10.01 CT. 30. 2. SPR (m /s) 60 s average. 22.5. 15. 7.5. 0 0. 5. 10 Time (min). 15. 20.


(80) 80. A5 SBI (Single Burning Item) HRR and SPR plots The Heat Release Rate (HRR) plots are all excluding the burner contribution, which is 30 kW during the test The Smoke Production Rate (SPR) plots are 60-s averaged. The plots presented here are from tests with insulation in tube form and mounted vertically. The figures have no figure text but they can be identified by the axis and by the product code above each plot..

(81) 81. M 1.01 SB I 200. HRR (kW). 150. 100. 50. 0 0. 5. 10. 15. 20. 25. Time (min). M 1.01 SB I 5. 3. 2. SPR (m /s) 60 s average. 4. 2. 1. 0 0. 5. 10. 15. Time (min). 20. 25.



(84) 84. M 3.03 SB I 500. HRR (kW). 400. 300. 200. 100. 0 0. 5. 10. 15. 20. 25. Time (min). M 3.03 SB I 5. 3. 2. SPR (m /s) 60 s average. 4. 2. 1. 0 0. 5. 10. 15. Time (min). 20. 25.

(85) 85. M3.05 SBI. 200. HRR (kW). 150. 100. 50. 0 0. 5. 10 15 Time (min). 20. 25. M3.05 SBI. 5. 2. SPR (m /s) 60 s average. 4. 3. 2. 1. 0 0. 5. 10 15 Time (min). 20. 25.






(91) 91. M5.03 SBI. 200. HRR (kW). 150. 100. 50. 0 0. SPR smoke clogged. 5. 10 15 Time (min). 20. 25.

(92) 92. M5.05 SBI. 500. HRR (kW). 400. 300. 200. 100. 0 0. 5. 10 15 Time (min). 20. 25. M5.05 SBI. 5. 2. SPR (m /s) 60 s average. 4. 3. 2. 1. 0 0. 5. 10 15 Time (min). 20. 25.





No results found