PERFORMANCE TEST REPORT Report No.: F6398.01-121-24 Test Date: March 9 th & 10th, 2016 Rendered to: Metal Building Manufacturers Association Cleveland, Ohio ASSEMBLIES: Metal Roof Panels and Clay Roof Tiles with Type 1 Photovoltaic Module TEST METHOD: Fire Test Program Developed by MBMA Task Group Using UL 1703 Section 31 & ASTM E108 Procedures with Alternate Roof Covering Materials This report contains in its entirety: Cover Page: 1 page Report Body: 11 pages Photographs: 8 pages
Page 1 of 11 1.0 Report Issued To: Metal Building Manufacturers Association 1300 Sumner Avenue Cleveland, Ohio 44115-2851 2.0 Test Laboratory: Architectural Testing, Inc., an Intertek company ( Intertek-ATI ) York, Pennsylvania 17406 717-764-7700 3.0 Project Summary: 3.1 Introduction: These fire testing standards aim to measure relative fire characteristics of roof coverings with photovoltaic modules and panels mounted above or on the roof covering under simulated fire scenarios which originate outside the building. Under controlled laboratory conditions, the behavioral response of materials, products or assemblies as affected by heat and flame are described. The performances of the roof covering systems are described only under specific conditions. Information is not provided by these tests that are applicable to any scenarios other than the specific conditions experienced during testing. Information is not provided by these tests that are applicable to actual fire situations because of the inherent differences between the classes as it pertains to fire source and fire application; no comparison between the classes exist. Results from tests are applicable to the specifics of the test and the aspect in which the tests were conducted, and are not applicable to similar materials or the results of those materials when used in concert with other materials. 3.2 Roof Assemblies: Metal Roof Panels and Clay Roof Tiles with Type 1 Photovoltaic Module 3.3 Compliance Statement: Results obtained are tested values and were secured by using the designated test method(s). This report does not constitute certification of this product nor an opinion or endorsement by this laboratory. It is the exclusive property of the client so named herein and relates only to the specimens tested. This report may not be reproduced, except in full, without the written approval of Intertek-ATI. 3.4 Test Dates: March 9 th & 10 th, 2016 3.5 Test Sample Source: The roof assembly components used in testing were supplied by various members of the UL Non-Combustible Roof Group of the Metal Building Manufacturers Association. Members included S-5!, SolarCity, SunEdison, and Intertek-ATI.
Page 2 of 11 3.0 Project Summary: (continued) 3.6 Test Method: 4.0 Test Details: UL 1703-15, Standard for Safety Flat-Plate Photovoltaic Modules and Panels Section 31 Fire Tests with alternate roof coverings ASTM E108-07a / UL 790-04, Standard Methods for Fire Tests of Roof Class A and Class B procedures for Calibration, Spread of Flame, and Burning Brand 4.1 Specimen Description: Class A test decks for the Spread of Flame and Burning Brand tests were constructed of 2x4 lumber and 15/32 inch Type AC grade plywood with face and back veneers of Douglas Fir and finished with the components outlined below: 4.2 Spread of Flame: 4.2.1 Baseline Tests Test Deck 40 inches wide by 8 feet long. Underlayment 15 lb. asphalt felt paper attached directly to the test deck with 5/16 in. staples on nominal 12 inch centers. Each layer overlapped the next layer by 2 inches. Roof Covering Class A three tab shingles were attached with 1-1/4 inch roofing nails. 4.2.2 Metal Panel Tests Test Deck 40 inches wide by 8 feet long. Underlayment One layer of GAF VersaShield was attached directly to the test deck with 5/16 inch staples on nominal 12 inch center spacing. Each layer overlapped the next layer by two inches. 30 lb. asphalt felt paper was installed directly above the VersaShield. Each layer of felt paper used a two inch overlap. Roof Covering Back section of deck comprised of steel 29 gauge galvalume metal roofing panels attached with #12 by 1 inch long metal roofing screws spaced 8 inches applied to the deck. Panels were installed directly under a Type 1 PV module. PV module measured 65 inches high by 39 inches wide by 2 inches thick. Module was mounted to the deck using S-5! hardware creating a 5 inch space between roof covering surface and bottom of the module. Front section of deck exposed to the flame plume was comprised of asphalt shingles and 15 lb. felt underlayment roof system (described in Section 4.2.1 and specified in UL 1703) to create a fire scenario at the metal roof panel and PV module interface.
Page 3 of 11 4.0 Test Details: 4.1.1 Clay Tile Tests Test Deck 40 inches wide by 8 feet long. Underlayment 30 lb. asphalt felt paper was attached directly to the test deck with 5/16 in. staples on nominal 12 inch centers. Each layer overlapped the next layer by 2 inches. Clay Roof Tiles Two types of tiles were utilized in the test program: flat and barrel shaped. Back section of deck comprised of the clay roof tiles with a 3 in. head lap between courses and were installed with 1-1/4 in. roof nails. Tiles were installed directly under a Type 1 PV module. PV module measured 65 inches high by 39 inches wide by 2 inches thick. Module was mounted to the deck using SunEdison hardware creating a 5 inch space between roof covering surface and bottom of the module. Front section of deck exposed to the flame plume was comprised of asphalt shingles and 15 lb. felt underlayment roof system (described in Section 4.2.1 and specified in UL 1703) to create a fire scenario at the metal roof panel and PV module interface. 4.3 Burning Brand: Test Deck 40 inches wide by 52 inches long with a 1/8 inch full length 40 inch horizontal joint at 22-1/2 inches and a 1/8 inch partial 22-1/2 inch vertical joint centered at 20 inches and located on the far side of the deck during exposure. Underlayment One layer of GAF VersaShield was attached directly to the test deck with 5/16 inch staples on nominal 12 inch center spacing. Each layer overlapped the next layer by two inches. 30 lb. asphalt felt paper was installed directly above the VersaShield. Each layer of felt paper used a two inch overlap. Metal Roof Panels Steel 29 gauge galvalume metal roofing panels attached with #12 by 1 inch long metal roofing screws spaced 8 inches applied to the deck. Panels were installed directly under a Type 1 PV module. PV module measured 65 inches high by 39 inches wide by 2 inches thick. Module was mounted to the deck using S-5! hardware creating a 5 inch space between roof covering surface and bottom of the module. Clay Roof Tiles Two types of tiles were utilized in the test program: flat and barrel shaped. Clay roof tiles with a 3 in. head lap between courses and were installed with 1-1/4 in. roof nails. Tiles were installed directly under a Type 1 PV module. PV module measured 65 inches high by 39 inches wide by 2 inches thick. Module was mounted to the deck using SunEdison hardware creating a 5 inch space between roof covering surface and bottom of the module. 4.4 Storage Information: Decks were stored in the fire laboratory prior to testing. Typical laboratory conditions are 60-80 F and 40-65% relative humidity.
Page 4 of 11 4.0 Test Details: (continued) 4.5 Moisture Content: Moisture content of the lumber and plywood were verified prior to testing. Plywood was less than 8% and the lumber was within 8-10%. 4.6 Equipment Calibration: A Fire Test Apparatus as described in ASTM E108 was used to generate 12 ± 0.5 mph air current and flame temperature of 1400 ±50 F. Air speed and flame temperature of the Fire Test Apparatus were calibrated prior to testing. A gas burner with flame temperature of 1630 ±50 F was used to ignite the brands for testing. See Section 5.1 for more information on equipment calibration. 4.7 Spread of Flame Procedure: After calibration of equipment, a test specimen described in Section 4.1.1 was placed into the steel framed holder for testing at a 5:12 slope and refractory mortar applied at the leading edge of the deck. The fire test apparatus was turned on and the specimen was subjected to continuous flame and air current for duration of 10 minutes. This procedure was repeated for both test decks and observations were documented. 4.8 Burning Brand Procedure: Brands were oven conditioned at 105 to 120 F for a minimum of 24 hours prior to testing. Class A brands consisted of three layers of twelve 1 inch by 1 inch by 12 inch strips of Douglas Fir forming a grid of 12 inch square by 2-1/4 inch thick and weight verified to be within 2000 ±150 grams. The brand was ignited by exposing each 12 inch by 12 inch face for 30 seconds, each 2-1/4 inch by 12 inch face for 45 seconds, and the 12 inch by 12 inch faces again for 30 seconds into the gas burner. Total duration of ignition exposure was 5 minutes. Class B test brands consisted of a grid 6 inches square and approximately 2-1 4 inches thick made of dry Douglas Fir lumber free of knots and pitch pockets. Use 18 nominal 1 by 1 by 6 inch strips, dressed on all four faces to 3 4 by 3 4 inches, and placed in three layers of six strips each with strips spaced 1 4 inch. The brands weight was verified to be within 500-650 grams at the time of test. The brand was ignited by exposing each 6 by 6-inch face for 30 seconds, each 2 1 4 by 6-inch face for 30 seconds, and each 6 by 6-inch face again for 30 seconds. After the brands were ignited, they were placed onto the test specimen described in Section 4.1.3 of the E108 and Section 31.2.1.2 of the UL 1703 at a 5:12 slope and secured in place with No. 18 soft iron wire. Class A brand were positioned with its upper edge 25.5 inches from the leading edge of the test deck and centered laterally with respect to the vertical joint in the test deck. The brand is to be placed so that the strips in both the upper and lower layers are parallel to the direction of air flow. Class B brand were positioned with its upper edge 24 inches from the leading edge of the test deck and centered laterally with respect to the vertical joint in the test deck. Brands were placed so that the strips in both the upper and lower layers are parallel to the direction of air flow. The testing apparatus without flame but only air current was turned on and the specimen was tested for 90 minutes or until the brand was completely consumed and the exposed and underside surfaces of the test deck did not exhibited signs of combustion. This procedure was repeated for all test decks and observations were documented.
Page 5 of 11 4.0 Test Details: (continued) 4.9 List of Official Observers: Name Samuel Truthseeker Vincent Sagan Ryan Devine Dustin Haddock Scott Gingrich Ben Green Robert George Company SunEdison Metal Building Manufacturers Association SolarCity S-5! Attachment Solutions Intertek-ATI Intertek-ATI Intertek-ATI 5.0 Calibration: 5.1 Calibration Information: Date: 03-08-2016 Average Wind Speed: 12.5 mph Ambient Temperature: 64 F Average Flame temperature: 1392 F Date: 03-09-2016 Average Wind Speed: 11.8 mph Ambient Temperature: 61 F Average Flame temperature: 1395 F Date: 03-10-2016 Average Wind Speed: 11.9 mph Ambient Temperature: 67 F Average Flame temperature: 1395 F
Page 6 of 11 5.0 Calibration: (continued) Base line Spread of Flame tests Test Flame Spread Deck #1 Deck #2 Deck #3 Average PV Offset Distance 46 inches 47 inches 48 inches 47 inches 35 inches 6.0 : Spread of Flame with Metal Roof Covering Specimen #1 00:01 Flame On Test Starts. 2:24 Ignition Ignition of the asphalt shingles. 3:57 1 foot Flame spread reaches 1 foot. 4:44 2 foot Flame spread reaches 2 feet. 6:07 3 foot Flame spread reaches 3 feet. 10:00 End End of test After Test No sustained surface flaming was observed passes 3 feet; PASS Spread of Flame with Metal Roof Covering Specimen #2 00:01 Flame On Test Starts. 1:37 Ignition Ignition of the asphalt shingles. 1:53 1 foot Flame spread reaches 1 foot. 3:04 2 foot Flame spread reaches 2 feet. 6:07 3 foot Flame spread reaches 3 feet. 10:00 End End of test After Test No sustained surface flaming was observed passes 3 feet; PASS
Page 7 of 11 6.0 : (continued) Class A Burning Brand with Metal Roof Covering Specimen #3 00:01 Brand Placed Start of test. 2:04 Burn through Ignition on the underside of the PV module. 3:31 Smoke Smoke emits from the underside of deck at horizontal saw kerf. 6:15 Glass Breaks Brand burns thru PV module breaks glass and falls on roof deck. Brand 20:57 Brand is fully consumed. Consumed 29:00 End of Test End of test no signs of combustion. After Test No ignition of the product or underlayment was observed; PASS Class A Burning Brand with Metal Roof Covering Specimen #4 00:01 Brand Placed Start of test. 2:15 Burn through Ignition on the underside of the PV module. 4:12 Smoke Smoke emits from the underside of deck at horizontal saw kerf. 5:50 Glass Breaks Brand burns thru PV module breaks glass and falls on roof deck. 20:30 End of Test End of test no signs of combustion. After Test No ignition of the product or underlayment was observed; PASS. Class B Burning Brand with Metal Roof Covering Specimen #5 00:01 Brand Placed Start of test. 00:51 Smoke Smoke emits from the saw kerfs joint on underside of the deck. 17:04 Brand Brand consumed. 45:00 End Test No smoke or evidence of combustion - end test. After Test No sustained flaming on the underside of the deck was observed; PASS.
Page 8 of 11 6.0 : (continued) Class B Burning Brand with Metal Roof Covering Specimen #6 00:01 Brand Placed Start of test. 00:48 Smoke Smoke emits from the saw kerfs joint on underside of the deck. 14:13 Brand Brand consumed. 21:36 End Test No smoke or evidence of combustion - end test. After Test No sustained flaming on the underside of the deck was observed; PASS. Spread of Flame with Clay Barrel Tile Specimen #7 00:01 Flame on Start of test. 01:25 Ignition Ignition of the asphalt shingles. 1:34 1 Foot Flame spread reaches 1 foot. 2:38 2 Feet Flame spread reaches 2 feet. 5:31 3 Feet Flame spread reaches 3 feet. 10:00 End Test End test. After Test No sustained surface flaming was observed passes 3 feet; PASS Spread of Flame with Clay Flat Tile Specimen #8 00:01 Flame on Start of test. 01:42 Ignition Ignition of the asphalt shingles. 1:45 1 Foot Flame spread reaches 1 foot. 2:53 2 Feet Flame spread reaches 2 feet. 3:55 3 Feet Flame spread reaches 3 feet. 10:00 End Test End test. After Test No sustained flaming on the underside of the deck was observed; PASS.
Page 9 of 11 6.0 : (continued) Spread of Flame with Clay Flat Tile Specimen #9 00:01 Flame on Start of test. 1:34 Ignition Ignition of the asphalt shingles. 1:36 1 Foot Flame spread reaches 1 foot. 4:36 2 Feet Flame spread reaches 2 feet. 5:09 3 Feet Flame spread reaches 3 feet. 10:00 End Test End test. After Test No sustained flaming on the underside of the deck was observed; PASS. Class B Burning Brand with Clay Barrel Tile Specimen #10 00:01 Brand Placed Start of test. 00:54 Cracked tile Tile below mounted burning brand cracks. 14:02 Brand Brand consumed. 26:30 End Test No smoke or evidence of combustion - end test. After Test No sustained flaming on the underside of the deck was observed; PASS. Class B Burning Brand with Clay Flat Tile Specimen #11 00:01 Brand Placed Start of test. 00:45 Cracked tile Tile below mounted burning brand cracks. 13:53 Brand Brand consumed. 20:00 End Test No smoke or evidence of combustion - end test. After Test No sustained flaming on the underside of the deck was observed; PASS.
Page 10 of 11 6.0 : (continued) Class A Burning Brand with Clay Flat Tile Specimen #12 00:01 Brand Placed Start of test. 2:24 Burn through Ignition on the underside of the PV module. 3:08 Dripping Dripping onto the roof deck from the underside of the PV module. 6:15 Glass Breaks Brand burns thru PV module breaks glass and falls on roof deck. 23:13 Brand Consumed Brand completely consumed. 27:00 End of Test End of test no signs of combustion. After Test No ignition of the product or underlayment was observed; PASS. Intertek-ATI will service this report for the entire test record retention period. Test records that are retained such as detailed drawings, datasheets, representative samples of test specimens, or other pertinent project documentation will be retained by Intertek-ATI for the entire test record retention period. This report does not constitute certification of this product nor an opinion or endorsement by this laboratory. It is the exclusive property of the client so named herein and relates only to the specimen(s) tested. This report may not be reproduced, except in full, without the written approval of Intertek-ATI. For INTERTEK-ATI: Scott Gingrich Senior Technician Fire Testing Matthew Freeborn Manager Fire Testing SDG:ddr Attachments (pages): This report is complete only when all attachments listed are included. Appendix-A: Photographs (8)
Page 11 of 11 Revision Log Rev. # Date Page(s) Revision(s) 0 03/23/2016 N/A Original report issue This report produced from controlled document template ATI 00538, revised 05/06/15.
Appendix A Photographs
Test Report No.: F6398.01-121-24 Photo No. 1 Spread of Flame Assembly with Metal Roof Covering Photo No. 2 Spread of Flame Assembly with Clay Flat Tile
Photo No. 3 Spread of Flame Assembly with Clay Barrel Tile Photo No. 4 Spread of Flame with Metal Roof Covering (During Test)
Photo No. 5 Spread of Flame with Clay Flat Tile (During Test) Photo No. 6 Spread of Flame with Clay Barrel Tile (During Test)
Photo No. 7 Class A Burning Brand Set-up on Metal Roof Covering (Pre-test) Photo No. 8 Class A Burning Brand on Metal Roof Covering (During Test)
Photo No. 9 Class B Burning Brand with Metal Roof Covering (During Test) Photo No. 10 Spread of Flame Burn Pattern with Metal Roof Covering (Post-test)
Photo No. 11 Class A Burning Brand with Metal Roof Covering (Post-test) Photo No. 12 Class A Burning Brand with Clay Flat Tile (Post-test)
Photo No. 13 Class B Burning Brand with Clay Flat Tile (Post-test) Photo No. 14 Class B Burning Brand with Metal Roof Covering (Post-test)
Photo No. 15 Class B Burning Brand with Metal Roof Covering Underdeck (Post-test)