Residential sprinkler and high-pressure water mist systems. Tests in a living room scenario

Full text

(1)Magnus Arvidson Ida Larsson. Residential Sprinkler and High-Pressure Water Mist Systems Tests in a Living Room Scenario. SP Swedish National Testing and Research Institute SP Fire Technology SP REPORT 2001:16.

(2) Magnus Arvidson Ida Larsson. Residential Sprinkler and High-Pressure Water Mist Systems Tests in a Living Room Scenario. SP Swedish National Testing and Research Institute SP Fire Technology SP REPORT 2001:16.

(3) 2. Abstract The performance of three different types, commercially available, residential sprinklers and a high-pressure water mist system nozzle was evaluated in a living room scenario. The sprinklers were of the recessed pendent, concealed and horizontal sidewall type. A living room was constructed using wood studs and ceiling joists. The ceiling height of the living room was either 2,5 m or 5,0 m. The living room was attached to a smaller bedroom. The ceiling surface and ceiling gas temperatures were measured. The tests with the sprinklers were conducted at a nominal discharge density of either 2,05 mm/min (0,05 gpm/ft2), 4,1 mm/min (0,1 gpm/ft2) or 6,1 mm/min (0,15 gpm/ft2), respectively. The results indicate that residential sprinklers generally will reduce the severity of the fire compared to the non-sprinklered case. However, the performance of the tested residential sprinklers was very dependent on the type of sprinkler and the way the fire was initiated. Key words: Residential sprinklers, water mist system, large-scale fire tests. Sökord: Boendesprinkler, sprinkler, vattendimma, fullskaleförsök.. SP Sveriges Provnings- och Forskningsinstitut SP Rapport 2001:16 ISBN 91-7848-860-5 ISSN 0284-5172 Borås 2001. SP Swedish National Testing and Research Institute SP Report 2001:16. Postal address: Box 857, SE-501 15 BORÅS, Sweden Telephone: +46 33 16 50 00 Telex: 36252 Testing S Telefax: +46 33 13 55 02 E-mail: info@sp.se.

(4) 3. Contents Abstract. 2. Contents. 3. Preface. 4. Sammanfattning. 5. 1 1.1 1.2. Introduction Background Objective of the test series. 7 7 7. 2 2.1 2.2 2.3 2.4 2.4.1 2.4.2 2.5. Test set-up Test facility Instrumentation and documentation The sprinklers and the water mist nozzle used for the tests The fire scenarios The upholstered chair scenario The simulated furniture scenario Test programme. 8 8 8 9 11 11 11 12. 3 3.1 3.2 3.3 3.4. Test results Test results for the recessed pendent sprinkler Test results for the concealed pendent sprinkler Test results for the horizontal sidewall sprinkler Test results for the Ultra Fog water mist nozzle. 14 14 17 19 24. 4 4.1 4.2. Discussion and conclusions The performance of the different sprinklers Other observations and conclusions. 27 27 28. References. 29. Enclosure 1 – The test room (2,5 m ceiling height) Enclosure 2 – The test room (5,0 m ceiling height) Enclosure 3 – The simulated furniture Appendix A – Measurement graphs.

(5) 4. Preface The test programme described within this report was financed by VINNOVA (previously named NUTEK), the Swedish National Board for Industrial and Technical Development, under their research programme for residential sprinklers, project AIS 7. The AIS 7 programme on residential sprinklers was started early 2000 and has got a twoyear frame. There are three main objectives of the programme: • To develop a reliable and economically feasible sprinkler protection concept for residential occupancies. • To explain and promote the benefits of residential sprinklers for professionals as well as for the public. • To prepare a handbook with installation requirements and guidance regarding specific building code trade-offs. The programme is carried out by the Swedish Institute for Wood Technology Research (Trätek), the University of Lund and the Swedish National Testing and Research Institute (SP) in close corporation with the building industry, building products manufacturers, sprinkler contractors and fire protection consultants..

(6) 5. Sammanfattning Sprinkler i bostäder har aktualiserats i samband med att flervåningshus med bärverk av trä blev tillåtna enligt en ny svensk byggnorm från år 1994. Målsättningen med boendesprinkler är att de i första hand skall rädda liv genom att skydda mot övertändning av lägenhetens inredning. Med sprinkler kan man acceptera vissa byggnadstekniska avsteg från byggnormens krav. Eventuellt kan lägre krav ställas på valet av inredningsmaterial vilket medger ett flexiblare materialval med mer synligt trä invändigt och på fasad. De försök som redovisas i denna rapport finansierades av VINNOVA (f d NUTEK) och genomfördes inom ramen för projekt AIS 7, ”Boendesprinkler” samt projektet ”Boendesprinkler för etagevåningar” (projektnr P14168-1). Den primära målsättningen med försöken var att jämföra olika typer av sprinkler och prova olika vattentätheter i ett rum möblerat som ett vardagsrum. Försöken genomfördes i en uppställning som speglade brand i ett 20 m2 ”vardagsrum” förbundet med ett mindre ”sovrum”. Två försöksserier genomfördes, i den första var vardagsrummets takhöjd 2,5 m, i den andra var den 5,0 m. Avsikten med den andra försöksserien var att simulera en lägenhet med dubbel rumshöjd förbundet med ett ”sovloft”. Väggar och tak där branden anlades bekläddes antingen med spånskiva (huvuddelen av försöken) eller med gipsskiva. Två olika primära brandkällor användes, den första var en kommersiell fåtölj, som antändes antingen vid golvnivå, invid fåtöljens sida så att väggpanelen säkert skulle komma att involveras i branden eller på fåtöljens sittdyna. Den andra typen av brandkälla var en ”simulerad” möbel som gav en högre strålningsintensitet mot väggen än fåtöljen, vilket satte sprinklernas förmåga att väta väggytorna i rummet på prov. Försöken med sprinkler kompletterades med försök med ett system med ”vattendimma”. Dessa system dimensionerades med ett nominellt vattentryck om 75 bar, alltså betydligt högre än för sprinkler, och finfördelar därför vattnet till betydligt mindre vattendroppar. Totalt genomfördes 36 försök (av vilka några dock inte rapporterades). Försöken visar att det är stora skillnader mellan de olika sprinklerna, vilket i första hand beror på skillnader i aktiveringstid på grund av skillnader i termisk tröghet mellan de tre olika sprinklerna. Försöken visar också att väggarnas ytskikt har stor betydelse för sprinklernas förmåga att dämpa eller kontrollera branden. Däremot spelar takets ytskikt inte särkilt stor roll, risken för att takets ytskikt skall antända är liten om sprinklersystemet är dimensionerat för att dämpa branden. Rummets takhöjd har relativt liten inverkan på dimensioneringen av sprinklersystemet, faktum är att bränderna i flera fall var lättare att dämpa eller att kontrollera i rummet med 5,0 m takhöjd. Försöken visade att det var den ”simulerade” möbeln var det svåraste brandsceneriet för sprinklerna, eftersom denna brandkälla gav en betydligt högre strålningsintensitet mot väggen än fåtöljen och dessutom tillväxte betydligt snabbare. Man kan notera att det är en stor risk för att sprinkler i ett angränsande rum aktiverar. Denna risk är inte alls beroende av vilken vattentäthet systemet är dimensionerat med, eftersom sprinkler i det angränsande rummet i flera fall aktiverade innan sprinklern i vardagsrummet..

(7) 6. Siktbarheten i rummet blev ganska snabbt ett problem, vilket försvårar utrymning. När fåtöljen, som ofta hade ett mycket långsamt initialt brandförlopp, användes som brandkälla fylldes rummet med en tunn, vit rök. I ett senare skede bildades tjockare, svart rök, som trycktes ned mot golvnivå när sprinklern aktiverade. Däremot överskreds inte den kolmonoxidkoncentration, 3000 ppm (0,3 vol%) som normalt har använts som gränsvärde för boendesprinkler, i något av försöken. Försöken med systemet med ”vattendimma” visar att dessa system behöver ha samma egenskaper som en sprinkler, det vill säga att vattensprayen når högt upp på rummets väggar. Försöken visar även att kylförmågan hos vattendropparna är bättre jämfört med sprinkler, vilket var förväntat. En annan observation som gjordes var att systemet drog ned en större andel frisk luft till branden, vilket fick branden att brinna med en högre intensitet..

(8) 7. 1. Introduction. 1.1. Background. Residential sprinklers have been used since the beginning of the 1980’s in the USA, but have not become very common in Sweden. There are several reasons for this, however, efforts are now undertaken to adopt and implement the residential sprinkler technology to Swedish conditions. One of the primary driving forces for using residential sprinklers is the new focus on multi-storey buildings constructed mainly from wood. Historically, such buildings have been prohibited by the Swedish building code, unless the maximum number of floors is restricted to two. An important reason for this is several major city fires during the previous centuries. However, with the introduction of a performance based building code approach a few years ago, higher buildings are allowed if it can be shown that the level of safety for the occupants is not reduced. Obviously, one way to improve the safety is to use sprinklers. Eventually, this would permit the use of exposed wood both for the interior as well as the exterior (facades).. 1.2. Objective of the test series. Previously, an initial evaluation of different residential sprinkler has been conducted [1] together with an extensive compilation of well-documented residential sprinkler fire tests [2]. The primary objective of these tests was to provide guidance regarding the appropriate water discharge densities as a function of the type of residential sprinklers used and its coverage area. The ultimate goal of the tests will serve as a baseline for the development of a recommendation regarding the installation and design of residential sprinklers for Swedish conditions. In addition to the sprinkler tests, tests with a high-pressure water mist nozzle were conducted. During the first series of test, tests were conducted inside a living room having a ceiling height of 2,5 m. During the second test series, the ceiling height of the living room was 5,0 m high. The sprinklers were installed in accordance with the design and installation manual of the manufacturer and were of the recessed pendent, concealed and horizontal sidewall type. Underwriters Laboratories, Inc. listed all three sprinklers for use in residential occupancies per NFPA 13R..

(9) 8. 2. Test set-up. 2.1. Test facility. All the fire tests were conducted in a 4 m wide by 5 m long “living room” which was attached to a small “bedroom” sized 3 m by 3 m. The rooms were constructed using wood studs and ceiling joists. The living room had a doorway opening, 0,9 m wide and 2,3 m high, positioned centric at on of the short sidewalls. The short sidewall also had a front glazing for observation purposes. The bedroom was connected to the living room with a doorway of the same size. The depth of the lintel of the two doorway openings was approximately 0,22 m. The ceiling height of the two rooms was 2,5 m. During the second test series, the living room had a ceiling height of 5,0 m. In this case, the bedroom was connected to the living room with a doorway that had its low edge 2,5 m above floor level. Obviously, a staircase would be needed in an actual apartment to go between the two rooms, but such staircase was not used in the tests. The depth of the lintel of the doorway opening between the two rooms was also in this case approximately 0,22 m. To withstand a series of consecutive fire tests with limited need for repair in between the tests the walls and the ceiling consisted of one layer of 12 mm ”Promatect-H” noncombustible board. Either nominally 10 mm thick particle board or nominally 13 mm thick gypsum board (three tests only) was used as the outer lining material on the corner walls and ceiling where the fire was arranged. A 1200 mm by 1200 mm ”Promatect-H” non-combustible board was placed under the fire source in order to protect the concrete floor. Schematic drawings of the two test set-ups are provided in enclosures 1 and 2, respectively.. 2.2. Instrumentation and documentation. The test facility was instrumented with thermocouples to measure ceiling gas temperatures above the fire as well as close to the installed sprinkler or nozzle. In addition, temperatures were measured inside the bedroom at the two positions of the gas sampling probes, see below. All the thermocouples were of type K (chromel-alumel) made from 0,5 mm wire welded together. Ceiling surface temperatures were measured with a type K thermocouple that was welded to a small (diameter = 12 mm) and 0,3 mm thick copper plate. The copper plate was glued to the ceiling surface. In addition to the temperature measurements discussed above, the possibility for having an activation of an additional sprinkler was determined. This was made possible by using a special thermocouple with thermal characteristics similar to a ”fast response” sprinkler. The thermocouple was positioned centric inside the bedroom, with its bottom 50 mm below the ceiling. The temperatures at this position were also measured with a 0,5 mm wire thermocouple..

(10) 9. Table 1 Channel Ch 31 Ch 32 Ch 33 Ch 34 Ch 35 Ch 36 Ch 30 Ch 39 Ch 40 Ch 38 Ch 7 Ch 6 Ch 5 Ch 37 Ch 71 Ch 72 Ch 73 Ch 61 Ch 74. Measurement points and associated channels. Location 500 mm from corner walls Thermocouple flush with ceiling surface (copper plate) Thermocouple 10 mm below ceiling Thermocouple 50 mm below ceiling 1000 mm from corner walls Thermocouple flush with ceiling surface (copper plate) Thermocouple 10 mm below ceiling Thermocouple 50 mm below ceiling Other Thermocouple close to the glass bulb of the sprinkler/nozzle Special thermocouple with characteristics of a ”fast response” sprinkler in bedroom Thermocouple adjacent to Ch 39 Inside bedroom, 0,6 m above floor level Thermocouple at sampling point inside the bedroom Oxygen (O2) concentration Carbon Monoxide (CO) concentration Carbon Dioxide (CO2) concentration Inside bedroom, 1,6 m above floor level Thermocouple at lower gas sampling point inside the bedroom Oxygen (O2) concentration Carbon Monoxide (CO) concentration Carbon Dioxide (CO2) concentration Water pressure and water flow rate Water pressure Water flow rate. The Oxygen (O2), Carbon Monoxide (CO) and Carbon Dioxide (CO2) concentrations were measured at a fix position, 1,0 m inside the bedroom. Gas was sampled at 0,6 m and 1,6 m, respectively, above the floor. The sampled gas was dried before it was transmitted to the measurement equipment, which means that the possible reduction of the O2-concentration from water vapour was not determined. The time delay for the response of the gas analysis equipment was approximately 15 seconds. The water pressure was measured with a pressure gauge close to the installed sprinkler/nozzle and the water flow rate was measured after the pump.. 2.3. The sprinklers and the water mist nozzle used for the tests. Three different types of sprinklers were used for the test programme, a recessed pendent sprinkler, a concealed pendent sprinkler and a horizontal sidewall sprinkler. The sprinklers were installed in accordance with the design and installation manual from the manufacturer. Underwriters Laboratories, Inc. listed all three sprinklers for use in residential occupancies per NFPA 13R..

(11) 10. The recessed and concealed pendent sprinklers were installed at a distance from the closest walls of 2,44 m. This equals a coverage of 4,88 m (16 ft) by 4,88 m (16 ft) or 23,8 m2 (256 ft2). The sidewall sprinkler was positioned centric on the short sidewall, above the doorway opening i.e. opposite the fire source, or, alternatively, on the short sidewall opposite the doorway opening. In both cases the coverage area equalled the area of the living room, i.e. 20 m2 (215 ft2). All sprinklers were of the frangible bulb type with 3 mm glass bulbs and a nominal temperature rating of 68°C (155°F). The recessed pendent and the horizontal sidewall sprinkler had a nominal K-factor of 63,4 L/min/bar1/2 (K=4,4), respectively. The concealed pendent sprinkler had a nominal K-factor of 60,5 L/min/bar1/2 (K=4,2). The recessed pendent sprinkler was height adjusted to its nominal mid-height, a nominal distance of 29 mm from the escutcheon plate. Given the thickness of the plate, the distance measured from the ceiling to the bottom of the deflector was nominally 33 mm. The sprinkler was orientated such that the plane of the frame arms was parallel to the right hand sidewall. The separable two-piece design of the cover plate and mounting cup of the concealed sprinkler provides for 12,7 mm of vertical adjustment, to reduce the accuracy to which the length of the fixed pipe drops of an actual system must be cut. For these tests, the pipe drop was prepared such that the sprinkler could be adjusted to its lowest position. From the standpoint of activation, this was probably the most advantageous adjustment. The sprinkler was orientated such that the plane of the frame arms was parallel to the right hand sidewall. The horizontal sidewall sprinkler was installed with the top of its deflector 125 mm below the ceiling, which is mid-height of the recommended distance of 100 to 150 mm. A recessed escutcheon plate was used with the sprinkler and the sprinkler was adjusted to the nominal mid-level distance from the plate, 32 mm. The pipe-work was made from DN 25 mm steel pipe, connected via a DN 34 mm hose to the pump. The desired water flow rate was set discharging the actual sprinkler type (with the deflector removed) before each test. As the far end of the pipe-work was fitted with a pressure gauge and the water flow rate was constantly measured and monitored, manual adjustment of the pump during the tests was possible. The Ultra Fog high-pressure water mist nozzle was installed at two positions, either 2,0 m from the walls (16 m2 coverage area) or 1,5 m from the walls (9 m2 coverage area). The nozzle was fitted with a frangible bulb with a 2 mm glass bulb and a nominal temperature rating of 57°C (135°F). The nozzle had a nominal K-factor of 1,76 L/min/bar1/2 (K=0,13)..

(12) 11. 2.4. The fire scenarios. 2.4.1. The upholstered chair scenario. In order to simulate a representative living room fire scenario a commercial, upholstered chair was used as the primary fire source. The chair had a timber frame with arms, seat and back going down to the ground. The chair had polyether foam filling with a density measured to be 29 kg/m3 and a fabric made from 63% cotton and 37% linen. The fabric had a measured area weight of 380 g/m2. The average overall weight of one chair was approximately 21,5 kg. The ignition source consisted of a cube, 60 mm by 60 mm by 75 mm, made of pieces of insulating fibre board, soaked with 120 mL of Heptane, and wrapped in a polyethylene bag. Ignition took place at the left hand side of the chair, with intention to allow the fire to involve the wall panelling prior to the activation of the sprinkler. In addition, the chair itself prevented the water spray from directly hitting the base of the fire. During these tests the chair was positioned with its side 100 mm from the left hand sidewall of the living room and with its back 50 mm from the back wall. Some tests were also conducted where ignition took place centric at the seat, close to the bottom of the backrest. This allowed a direct hit of the water spray from the sprinkler and the combustible wall panelling never became involved in the fire. During these tests the chair was angled 45°, such that the corners of the back of the chair touched the two walls.. 2.4.2. The simulated furniture scenario. The simulated furniture fire source consisted of an 800 mm by 800 mm plywood panel, nailed to a wood frame constructed from 45 mm by 45 mm wood sticks. A 100 mm thick polyether foam cushion was glued to the plywood panel and the cushion was covered with a cotton fabric that was stapled along the edges to the wood frame that held the plywood panel. The polyether foam had a nominal density of 33 kg/m3 and the fabric had a nominal area weight of 390 g/m2. The total average weight of the plywood backing, its wood frame, the foam cushion and the fabric was approximately 10 kg. The foam and the plywood backing were held in vertical position by wood supports that provided for a 45 mm distance measured from the floor to the underneath of the foam. A drawing of the construction of the simulated furniture is given in enclosure 3. The front face of the foam cushion was positioned facing the left hand sidewall, at a 100 mm distance. The distance measured from the edge of the foam to the backwall was 50 mm. The ignition source was similar to the one described above and was positioned at the floor, at the centreline of the foam. This simulated furniture fire source exposed the adjacent wall (100 mm distance) with a considerably higher heat flux rate as compared to the upholstered chair. The fact that the cotton fabric not was as heavy as the fabric used on the chair made the fire penetrate to the polyether foam faster and the fire develop more rapidly. The heat release rate of the simulated furniture (only) has previously been.

(13) 12. determined to approximately 200 kW. When the combustible walls are involved in the fire, the peak heat release rate is approximately 800 kW, given that the ceiling height is 2,5 m. These heat release rates are comparable to a smaller type upholstered chair. Compared to the upholstered chair, the advantage of the simulated furniture fire is that it probably has more repeatable fire characteristics and the fact that the cost is far less.. 2.5. Test programme. The test programme for the first series of tests (2,5 m ceiling height) consisted of a total of 18 tests. The tests with the sprinklers were conducted at a nominal discharge density of either 2,05 mm/min (0,05 gpm/ft2), 4,1 mm/min (0,1 gpm/ft2) or 6,1 mm/min (0,15 gpm/ft2), respectively. In addition to the tests with the sprinklers, three tests were conducted using the Ultra Fog high-pressure water mist nozzle. The nozzle was positioned such that its coverage area was less compared to the sprinklers, on the other hand the nominal discharge density was only in the order of 1,1 mm/min (0,03 gpm/ft2). The second series of tests (5,0 m high living room) consisted of a total of 14 tests. These tests were conducted in a similar way as the tests inside the living room with 2,5 m ceiling height. The test programmes are summarised in the tables below. Table 2 Test 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22. Test programme for the tests conducted in the 2,5 m high living room.. Fire source Upholstered chair (at seat) Upholstered chair (at seat) Upholstered chair (left hand side) Upholstered chair (left hand side) Upholstered chair (left hand side) Upholstered chair (left hand side)* Simulated furniture Simulated furniture Not reported Simulated furniture Simulated furniture Simulated furniture Simulated furniture Simulated furniture Simulated furniture Not reported Not reported Not reported Upholstered chair (at seat) Upholstered chair (left hand side) Upholstered chair (left hand side) Upholstered chair (left hand side)**. Type of sprinkler Recessed pendent sprinkler Recessed pendent sprinkler Recessed pendent sprinkler Recessed pendent sprinkler Concealed pendent sprinkler Concealed pendent sprinkler Recessed pendent sprinkler Recessed pendent sprinkler. Discharge density (mm/min) 4,1 mm/min 2,05 mm/min 4,1 mm/min 2,05 mm/min 4,1 mm/min 4,1 mm/min 2,05 mm/min 4,1 mm/min. Ultra Fog water mist nozzle Ultra Fog water mist nozzle Horiz. sidewall sprinkler Horiz. sidewall sprinkler Horiz. sidewall sprinkler Horiz. sidewall sprinkler. 1,1 mm/min 2,2 mm/min 4,1 mm/min 6,1 mm/min 6,1 mm/min 4,1 mm/min. Ultra Fog water mist nozzle Horiz. sidewall sprinkler Horiz. sidewall sprinkler Horiz. sidewall sprinkler. 1,1 mm/min 4,1 mm/min 4,1 mm/min 4,1 mm/min. *) Conducted with gypsum board on the corner walls. **) Conducted with another type of upholstered chair..

(14) 13. Table 3 Test 23 24 25 26 27 28 29 30 31 32 33 34 35 36. *). Test programme for the tests conducted in the 5,0 m high living room.. Fire source Upholstered chair (left hand side) Upholstered chair (left hand side) Upholstered chair (left hand side) Simulated furniture Simulated furniture* Simulated furniture Simulated furniture Simulated furniture Upholstered chair (left hand side) Upholstered chair (left hand side) Upholstered chair (left hand side)* Upholstered chair (left hand side) Upholstered chair (left hand side) Simulated furniture. Type of sprinkler Recessed pendent sprinkler Recessed pendent sprinkler Recessed pendent sprinkler Horiz. sidewall sprinkler Horiz. sidewall sprinkler Horiz. sidewall sprinkler Horiz. sidewall sprinkler Horiz. sidewall sprinkler Horiz. sidewall sprinkler Concealed pendent sprinkler Concealed pendent sprinkler Concealed pendent sprinkler Ultra Fog water mist nozzle Ultra Fog water mist nozzle. Conducted with gypsum board on the corner walls.. Discharge density (mm/min) 4,1 mm/min 2,05 mm/min 4,1 mm/min 4,1 mm/min 4,1 mm/min 6,1 mm/min 4,1 mm/min 6,1 mm/min 4,1 mm/min 4,1 mm/min 4,1 mm/min 4,1 mm/min 1,1 mm/min 1,1 mm/min.

(15) 14. 3. Test results. The test results and observations are summarised in the following chapters. Temperature and gas measurement graphs for all tests are given in Appendix A.. 3.1. Test results for the recessed pendent sprinkler. The two initial tests were conducted at a nominal discharge density of 4,1 mm/min and 2,05 mm/min, respectively. Ignition took place at the seat of the chair, i.e. the fire was fully exposed to the water spray. The times to activation were, with respect to the random nature of a fire, comparable. The tests results were also very similar; the fire was quickly suppressed in both tests. As expected, the weight loss of the chair was less in test 1, when a higher discharge density was used. The wall lining never became involved in the fire in any of the two tests. Tests 3 and 4 were conducted using the same discharge densities as in the first two tests. However, the fire was initiated at the side of the chair. It was expected that this position would be more challenging because of the shielded nature of the fire and the fact that the walls would become involved in the fire prior to the activation. The tests showed that the initial fire development was rather slow. It took a while for the fire to penetrate through the fabric and reach the polyether foam filling on the side and the armrest of the chair. It was not until the fire reached the side of the backrest that the fire accelerated and the heat flux became intense enough to ignite the adjacent wall. Flames had reached to approximately half the height of the wall when the sprinkler activated. Fire damage to the walls as well as the weight loss of the chair was slightly more severe at the lower discharge density. The “sprinkler” in the bedroom “activated” during one of the tests, test 2 as its temperature reached 70ºC. However, the measured difference in temperature was not significant if the results for the four tests are compared. All the test results with the recessed pendent sprinkler type, when the upholstered chair was used as the fire source, can be considered successful..

(16) 15. Table 4. Test results for the tests with the recessed pendent sprinkler. Ignition took place either at the seat or at the bottom, left hand side of the upholstered chair.. Test no. Date of test Wall and ceiling lining Ignition position of chair Type of sprinkler Desired flow rate [L/min] Coverage area [m2] Nominal density [mm/min] Activation time [min:s] Weight of chair before test [kg] Weight of chair after test [kg] Weight loss of chair [kg] Damage to walls Selected measurement data Ch 31, flush w/ ceiling surface Ch 34, flush w/ ceiling surface Ch 30, close to sprinkler Ch 39, “sprinkler” in bedroom Minimum O2 conc. [vol%] Maximum CO conc. [vol%] Note:. 1 2 3 4 2001-03-07 2001-03-07 2001-03-08 2001-03-08 Particle board Particle board Particle board Particle board At seat At seat Left hand side Left hand side Recessed pendent Recessed pendent Recessed pendent Recessed pendent 98 L/min (26 gpm) 49 L/min (13 gpm) 98 L/min (26 gpm) 49 L/min (13 gpm) 23,8 m2 (16 ft x 16 ft) 23,8 m2 (16 ft x 16 ft) 23,8 m2 (16 ft x 16 ft) 23,8 m2 (16 ft x 16 ft) 4,1 mm/min 2,05 mm/min 4,1 mm/min 2,05 mm/min (0,1 gpm/ft2). (0,05 gpm/ft2). (0,1 gpm/ft2). (0,05 gpm/ft2). 04:58 21,98 kg 20,84 kg 1,14 kg None. 05:23 21,33 kg 19,59 kg 1,76 kg None. 11:43 22,11 kg 20,48 kg 1,63 kg Slight. 11:12 21,07 kg 17,54 kg 3,53 kg Partly. 99°C 80°C 126°C 66°C 20,1 vol% 0,018 vol% -. 109°C 86°C 123°C 70°C 20,0 vol% 0,036 vol% -. 105°C 81°C 120°C 62°C 20,3 vol% 0,091 vol% -. 124°C 88°C 130°C 66°C 20,1 vol% 0,153 vol% -. Tests 7 and 8 were conducted with the simulated furniture as the fire source. This fire source exposed the adjacent wall (100 mm distance) with a considerably higher heat flux rate as compared to the upholstered chair. The fact that the cotton fabric not was as heavy as the fabric on the chair made the fire penetrate to the polyether foam faster and the fire develop more rapidly and repeatable. Test 7 was conducted using a discharge density of 2,05 mm/min. The sprinkler activated moments after the flame reached to the ceiling and had an initial effect on the fire, however, the water was not able to suppress the fire on the walls. Approximately four minutes after the activation of the sprinkler the fire reached its peak and flames protruded across the ceiling. About a minute later the material of the fire source was consumed and the fire intensity decreased. Almost all of the polyether foam was consumed and the walls were burnt and blackened all the way to the ceiling. The “sprinkler” in the bedroom “activated” during the test as its temperature reached 134ºC. The second test, test 8, was conducted at a discharge density of 4,1 mm/min. The fire spread a little faster to the adjacent wall in this test, which made the sprinkler activate at an earlier stage. However, visually the fire size was comparable to test 7. The course of events after the activation was not very dramatic, the fire in the wall panelling was suppressed rather quickly and the fire was confined to the source ignition after about a minute. The fire damage to the adjacent wall was minor, the area of the corner was not damaged at all, and more than half of the polyether foam was left after the test. The temperature measured at the “sprinkler” in the bedroom reached 56ºC, i.e. the “sprinkler” never activated. In summary, the results of test 8 can be considered very successful..

(17) 16. Table 5. Test results for the tests with the recessed pendent sprinkler. The simulated furniture was used as the fire source.. Test no. Date of test Wall and ceiling lining Fire source Type of sprinkler Desired flow rate [L/min] Coverage area [m2] Nominal density [mm/min] Measured density [mm/min] Activation time [min:s] Weight of SF before test [kg] Weight of SF after test [kg] Weight loss of SF [kg] Damage to walls Selected measurement data Ch 31, flush w/ ceiling surface Ch 34, flush w/ ceiling surface Ch 30, close to sprinkler Ch 39, “sprinkler” in bedroom Minimum O2 conc. [vol%] Maximum CO conc. [vol%] Note:. 7 8 2001-03-08 2001-03-12 Particle board Particle board Simulated furniture Simulated furniture Recessed pendent Recessed pendent 49 L/min (13 gpm) 98 L/min (26 gpm) 23,8 m2 (16 ft x 16 ft) 23,8 m2 (16 ft x 16 ft) 4,1 mm/min 2,05 mm/min (0,05 gpm/ft2). (0,1 gpm/ft2). 02:56 9,12 kg 6,50 kg 2,62 kg Severe. 02:26 10,20 kg 9,09 kg 1,11 kg Minor. 450°C 244°C 132°C 135°C 18,8 vol% 0,134 vol% -. 138°C 72°C 128°C 56°C 20,4 vol% 0,058 vol% -. Test 23 and 24 were conducted inside the living room with 5,0 m ceiling height. The upholstered chair was used as the fire source and ignition took place at the bottom, left hand side. The first test, test 23 was conducted at a nominal discharge density of 4,1 mm/min. As experienced in the previous tests, the fire developed slow at the beginning, and did not accelerate until the wall panelling ignited. At this time, the room was almost completely filled with thin, white smoke. When the sprinkler activated, flames had reached to about half the height of the wall. The activation of the sprinkler suppressed the fire and damages to the walls and the weight loss of the chair was limited. The room was completely filled with smoke throughout the duration time of the test, which would have made an evacuation from the bedroom at the second floor difficult. The maximum temperature measured at the “sprinkler” inside the “bedroom” reached to 60°C. Test 24 was conducted at a nominal discharge density of 2,05 mm/min. In this test the fire spread to the adjacent wall was faster, with faster associated activation time of the sprinkler. Two primary differences were noted as compared to test 23, firstly, the water did not suppress the fire and, secondly, the water spray did not push the smoke layer down to the floor as fast. However, the room was completely filled with smoke after approximately 12 minutes from ignition. The damage to the walls were more severe and the weight loss of the chair higher for this test. Although the fire was only controlled, the temperatures measured at the ceiling never reached any high levels. The maximum temperature measured at the “sprinkler” inside the “bedroom” reached to 75°C and exceeded 68°C after 7 minutes from ignition..

(18) 17. Table 6. Test results for the tests with the recessed pendent sprinkler in the living room with 5,0 m ceiling height. Ignition took place the bottom, left hand side of the upholstered chair.. Test no. Date of test Wall and ceiling lining Ignition position of chair Type of sprinkler Desired flow rate [L/min] Coverage area [m2] Nominal density [mm/min] Activation time [min:s] Weight of chair before test [kg] Weight of chair after test [kg] Weight loss of chair [kg] Damage to walls Selected measurement data Ch 31, flush w/ ceiling surface Ch 34, flush w/ ceiling surface Ch 30, close to sprinkler Ch 39, “sprinkler” in bedroom Minimum O2 conc. [vol%] Maximum CO conc. [vol%] Note:. 3.2. 23 24 2001-05-21 2001-05-21 Particle board Particle board Left hand side Left hand side Recessed pendent Recessed pendent 98 L/min (26 gpm) 49 L/min (13 gpm) 23,8 m2 (16 ft x 16 ft) 23,8 m2 (16 ft x 16 ft) 2,05 mm/min 4,1 mm/min (0,1 gpm/ft2). (0,05 gpm/ft2). 11:26 21,34 kg 19,16 kg 2,18 kg Partly. 06:28 21,09 kg 17,20 kg 3,89 kg Severe. 88°C 71°C 109°C 60°C 19,9 vol% 0,102 vol% -. 103°C 90°C 110°C 75°C 19,0 vol% 0,126 vol% -. Test results for the concealed pendent sprinkler. At the first test with the concealed pendent sprinkler, test 5, the fire was ignited at the left hand side of the upholstered chair. Compared to the previous tests with this fire scenario the initial fire spread faster to the backrest of the chair. This in turn initiated the wall panelling at an earlier stage. Therefore, the time to activation was shorter, which obviously would not be expected with similar fire development. However, at the time of activation the fire had spread up the corner walls and flames reached to the ceiling, which indicated that the fire size was larger compared to the tests with the recessed pendent sprinkler. Initially, the water spray knocked down the flames at the walls, but as the fire continued at the side and on the backside of the chair, the fire redeveloped. Temperatures at the ceiling reached high and an actual sprinkler inside the bedroom would have activated. The damages to the walls were severe and the weight loss of the upholstered chair more than 4 kg. Overall, the performance of the sprinkler under these conditions and at the chosen flow rate can be considered inappropriate. The second test with the concealed pendent sprinkler was identical, but because of the bad performance in test 5, gypsum boards was used on the corner walls instead of particle board. Therefore, the fire developed slower and the sprinkler activated later. Because the walls contributed less to the fire spread, the chair was more involved in the fire prior to the activation. Moments before the activation, flames had spread to about half the height of the corner walls. The fire was initially suppressed by the water spray, but continued to burn inside the seat cushion, at the side, and at the back of the chair. The fire redeveloped slightly during the.

(19) 18. test, but the fire can be considered to be controlled, as the ceiling temperatures were reasonably low. However, also for this test the “sprinkler” in the bedroom activated. Table 7. Test results for the tests with the concealed pendent sprinkler. Ignition took place at the left hand side of the upholstered chair.. Test no. Date of test Wall and ceiling lining Ignition position of chair Type of sprinkler Desired flow rate [L/min] Coverage area [m2] Nominal density [mm/min]. 5 6 2001-03-08 2001-03-09 Particle board Gypsum board Left hand side Left hand side Concealed pendent Concealed pendent 98 L/min (26 gpm) 98 L/min (26 gpm) 23,8 m2 (16 ft x 16 ft) 23,8 m2 (16 ft x 16 ft) 4,1 mm/min 4,1 mm/min (0,1 gpm/ft2). Measured density [mm/min] Activation time [min:s] 07:27 Weight of chair before test [kg] 21,97 kg Weight of chair after test [kg] 17,88 kg Weight loss of chair [kg] 4,09 kg Damage to walls Severe Selected measurement data Ch 31, flush w/ ceiling surface 352°C Ch 34, flush w/ ceiling surface 244°C Ch 30, close to sprinkler 167°C Ch 39, “sprinkler” in bedroom 87°C Minimum O2 conc. [vol%] 17,0 vol% Maximum CO conc. [vol%] 0,153 vol% Note: 1) 1) The cover plate of the sprinkler fell off at 06:53. 2) The cover plate of the sprinkler fell off at 12:42.. (0,1 gpm/ft2). 13:07 21,46 kg 16,80 kg 4,66 kg Minor 139°C 99°C 151°C 85°C 19,2 vol% 0,176 vol% 2). Tests 32, 33 and 34 were all conducted inside the living room with 5,0 m ceiling height, with the upholstered chair as the fire source and at a nominal discharge density of 4,1 mm/min. Test 32 was conducted with particle board on the corner walls. As seen in the previous tests, the fire developed very slowly at the beginning and it was not until the walls became involved in the fire that it accelerated. The sprinkler activated late, flames reached almost to the ceiling, however, the fire at the walls was immediately suppressed. The temperatures measured at the ceiling and inside the bedroom were very high and even melted the sample (plastic) line for the gas sampling point 1,6 m above the bedroom floor. It was suspected that the activation of the sprinkler was delayed, probably because of a water leakage from the sprinkler fitting that had gathered on the inside cover plate of the sprinkler. The test was therefore repeated in test 34. Test 33 was a repeat of test 32, but with gypsum board instead of particle boards on the corner walls. This made the fire burn less intense. The “sprinkler” inside the bedroom reached to 68°C, i.e. it barely activated. A larger part of the upholstered chair was consumed in this test, which is expected due to the less flammable wall lining. The temperatures measured at the ceiling were generally low. Test 34 was a strict repeat of test 32, due to the suspected problems with the activation of the sprinkler, as mentioned above. It can be concluded that there were problems in test 32, the sprinkler activated at an earlier stage in test 34, the flames of the fire reached to about 3 m above floor when the sprinkler activated. The fire was immediately knocked down and temperatures at the ceiling were generally low. However, the sprinkler inside.

(20) 19. the bedroom “activated” prior to the actual sprinkler, 68°C was reached after 07:38, the maximum temperature was 75°C. The temperatures measured at the ceiling were generally low also in this test. Table 8. Test results for the tests with the concealed pendent sprinkler. Ignition took place at the left hand side of the upholstered chair. The ceiling height was 5,0 m.. Test no. Date of test Wall and ceiling lining Ignition position of chair Type of sprinkler Desired flow rate [L/min] Coverage area [m2] Nominal density [mm/min]. 32 33 34 2001-03-28 2001-05-28 2001-05-29 Particle board Gypsum board Particle board Left hand side Left hand side Left hand side Concealed pendent Concealed pendent Concealed pendent 98 L/min (26 gpm) 98 L/min (26 gpm) 98 L/min (26 gpm) 23,8 m2 (16 ft x 16 ft) 23,8 m2 (16 ft x 16 ft) 23,8 m2 (16 ft x 16 ft) 4,1 mm/min 4,1 mm/min 4,1 mm/min (0,1 gpm/ft2). Activation time [min:s] 13:57 Weight of chair before test [kg] 22, 97 kg Weight of chair after test [kg] 20,44 kg Weight loss of chair [kg] 2,53 kg Damage to walls Severe Selected measurement data Ch 31, flush w/ ceiling surface 620°C Ch 34, flush w/ ceiling surface 523°C Ch 30, close to sprinkler 670°C Ch 39, “sprinkler” in bedroom 255°C Minimum O2 conc. [vol%] 14,7 vol% Maximum CO conc. [vol%] 0,340 vol% Note: 3) 1) The cover plate of the sprinkler fell off at 13:56. 2) The cover plate of the sprinkler fell off at 13:05. 3) The cover plate of the sprinkler fell of at 07:30.. 3.3. (0,1 gpm/ft2). (0,1 gpm/ft2). 13:21 22,31 kg 17,25 kg 5,06 kg Minor. 07:45 21,11 kg 19,12 kg 1,99 kg Minor. 76°C 151°C 131°C 68°C 19,4 vol% 0,137 vol% 4). 109°C 77°C 137°C 75°C 19,6 vol% 0,089 vol% 5). Test results for the horizontal sidewall sprinkler. The tests with the horizontal sidewall sprinkler were conducted with the sprinkler positioned either above the doorway opening of the room or on opposite wall. The nominal discharge density was either 4,1 mm/min or 6,1 mm/min. The first four tests were conducted with the simulated furniture to get a first indication of the performance of the sprinkler, the remaining three tests with the upholstered chair. The two tests where the sprinkler was positioned above the doorway, tests 12 and 13, were similar with respect to the development of the fire and the activation time of the sprinkler. The sprinkler activated approximately 40 seconds after the flames had reached the ceiling. The fire on the wall was suppressed moments after the activation of the sprinkler, however, the fire inside the polyether foam continued to burn. Due to the fact that the smoke filled the room, any observations were difficult to achieve. All the foam of the simulated furniture was consumed in test 12 and about 80% in test 13. The “sprinkler” inside the adjacent room “activated” in both tests, however, the activation temperature of 68°C was barely exceeded. Higher temperatures were recorded in test 13, which probably was due to the fact that the sprinkler activated at a later stage in this test..

(21) 20. Tests 14 and 15 were conducted with the sprinkler positioned opposite the doorway of the room. As expected, the sprinkler activated at an earlier stage of the fire in this position. The fire on the wall was suppressed at the higher (6,1 mm/min) discharge density, but continued to burn at the lower density. In both tests, the majority of the polyether foam was consumed; the weight loss at the lower density was higher though. The “sprinkler” inside the adjacent room never reached any high temperatures, probably because the hot gases had to pass the water spray with the sprinkler in this position. A general observation is that the weight loss of the simulated furniture was higher for tests 14 and 15, i.e. when the sprinkler was positioned opposite the doorway opening. The ceiling surface temperatures measured at the position 1000 mm from the corner walls were, however, lower. Table 9. Test results for the tests with the horizontal sidewall sprinkler. The simulated furniture was used as the fire source.. Test no. Date of test Wall and ceiling lining Fire source Type of sprinkler Pos. of sprinkler Desired flow rate [L/min] Coverage area [m2] Nominal density [mm/min] Activation time [min:s] Weight of SF before test [kg] Weight of SF after test [kg] Weight loss of SF [kg] Damage to walls Selected measurement data Ch 31, flush w/ ceiling surface Ch 34, flush w/ ceiling surface Ch 30, close to sprinkler Ch 39, “sprinkler” in bedroom Minimum O2 conc. [vol%] Maximum CO conc. [vol%] Note:. 12 13 2001-03-20 2001-03-20 Particle board Particle board Simulated furniture Simulated furniture Horiz. sidewall Horiz. sidewall Above doorway Above doorway 82 L/min (22 gpm) 122 L/min (32 gpm) 20 m2 (16,4 ft x 13 ft) 20 m2 (16,4 ft x 13 ft) 4,1 mm/min 6,1 mm/min. 14 15 2001-03-21 2001-03-21 Particle board Particle board Simulated furniture Simulated furniture Horiz. sidewall Horiz. sidewall Opposite doorway Opposite doorway 122 L/min (32 gpm) 82 L/min (22 gpm) 20 m2 (16,4 ft x 13 ft) 20 m2 (16,4 ft x 13 ft) 6,1 mm/min 4,1 mm/min. (0,1 gpm/ft2). (0,15 gpm/ft2). (0,15 gpm/ft2). (0,1 gpm/ft2). 02:41 10,25 kg 7,89 kg 2,36 kg Severe. 02:30 9,85 kg 7,87 kg 1,98 kg Partly. 02:22 10,17 kg 7,54 kg 2,63 kg Partly. 01:49 10,12 kg 6,70 kg 3,42 kg Severe. 143°C 97°C 140°C 69°C 20,0 vol% 0,124 vol% -. 166°C 105°C 154°C 73°C 20,3 vol% 0,060 vol% -. 86°C 43°C 110°C 36°C 19,8 vol% 0,165 vol% -. 201°C 57°C 107°C 37°C 19,8 vol% 0,101 vol% -. To verify the results with the simulated furniture, tests 20 and 21 were conducted using the upholstered chair, ignited on the left hand side, as the fire source. The sprinkler was either positioned opposite or above the doorway opening. As observed in the previous tests, the initial fire development was very dependent on how fast the adjacent wall was ignited. Therefore, the time to activation is not any valid measure regarding how fast the sprinklers responded to the fire. However, from a visual judgement, the sprinkler activated at an earlier stage of the fire when the sprinkler was positioned opposite the doorway, as the sprinkler activated before the flames had reached to the ceiling. This observation is also verified by the temperature measurements close to the sprinkler. Test 22 was also conducted with an upholstered chair, but the chair was of another type than the chair used in all the other fire tests in the test series. The sprinkler was positioned above the doorway opening for this particular test. It turned out that the chair used for test 22 burned more rapidly. The reason was very likely because of the construction of the chair and the thickness of the fabric, which was less for this chair. It.

(22) 21. might also have been the result of the fire characteristics of the polyether foam, but this was not investigated in detail. Table 10. Test results for the tests with the horizontal sidewall sprinkler. Ignition took place at the left hand side of the upholstered chair.. Test no. Date of test Wall and ceiling lining Ignition position of chair Type of sprinkler Pos. of sprinkler Desired flow rate [L/min] Coverage area [m2] Nominal density [mm/min]. 20 2001-03-22 Particle board Left hand side Horiz. sidewall Opposite doorway 82 L/min (22 gpm) 20 m2 (16,4 ft x 13 ft) 4,1 mm/min. 21 2001-03-22 Particle board Left hand side Horiz. sidewall Above doorway 82 L/min (22 gpm) 20 m2 (16,4 ft x 13 ft) 4,1 mm/min. 22 2001-03-22 Particle board Left hand side Horiz. sidewall Above doorway 82 L/min (22 gpm) 20 m2 (16,4 ft x 13 ft) 4,1 mm/min. (0,1 gpm/ft2). (0,1 gpm/ft2). (0,1 gpm/ft2). 09:21 21,17 kg 19,28 kg 1,89 kg Partly. 02:30 22,10 kg 18,14 kg 3,96 kg Severe. 204°C 82°C 134°C 86°C 19,7 vol% 0,075 vol% -. 281°C 131°C 157°C 114°C 17,3 vol% 0,191 vol% 1). Activation time [min:s] 03:52 Weight of chair before test [kg] 22,12 kg Weight of chair after test [kg] 20,71 kg Weight loss of chair [kg] 1,41 kg Damage to walls Minor Selected measurement data Ch 31, flush w/ ceiling surface 79°C Ch 34, flush w/ ceiling surface 55°C Ch 30, close to sprinkler 115°C Ch 39, “sprinkler” in bedroom 42°C Minimum O2 conc. [vol%] 20,7 vol% Maximum CO conc. [vol%] 0,007 vol% Note: 1) The test was conducted with another type chair.. Test 25 was conducted inside the living room with 5,0 m ceiling height. The horizontal sidewall sprinkler was positioned above the doorway, 125 mm below the ceiling of the room and the water flow rate equalled 4,1 mm/min. The sprinkler activated when flames reached to about 3 m above floor level. The activation of the sprinkler almost immediately suppressed the fire at the wall, however, the fire in the chair continued to burn at a reduced rate. The temperature at the “sprinkler” inside the bedroom reached to 94°C, and exceeded 68°C after 09:04. Test 31 was similar to test 25, however, the sprinkler was positioned opposite the doorway opening. The fire developed a little slower in this test, however, the sprinkler activated when the fire size was less, flames reached to about half the height of the wall. Also in this case the flames at the wall were almost immediately suppressed, however, the fire in the chair continued to burn at a reduced rate. The temperature at the “sprinkler” inside the bedroom only reached to 55°C. The temperature levels at the ceiling were reasonable low in both of the tests and were immediately reduced as the sprinkler activated. It can be concluded that the 4,1 mm/min discharge density was sufficient also for the 5,0 m high living room set-up, when the upholstered chair was used as the fire source..

(23) 22. Table 11. Test results for the tests with the horizontal sidewall sprinkler in the living room with 5,0 m ceiling height. Ignition took place at the bottom, left hand side of the upholstered chair.. Test no. Date of test Wall and ceiling lining Ignition position of chair Type of sprinkler Pos. of sprinkler Desired flow rate [L/min] Coverage area [m2] Nominal density [mm/min] Activation time [min:s] Weight of chair before test [kg] Weight of chair after test [kg] Weight loss of chair [kg] Damage to walls Selected measurement data Ch 31, flush w/ ceiling surface Ch 34, flush w/ ceiling surface Ch 30, close to sprinkler Ch 39, “sprinkler” in bedroom Minimum O2 conc. [vol%] Maximum CO conc. [vol%] Note:. 25 2001-05-21 Particle board Left hand side Horiz. sidewall Above doorway 82 L/min (22 gpm) 20 m2 (16,4 ft x 13 ft) 4,1 mm/min. 31 2001-05-28 Particle board Left hand side Horiz. Sidewall Opposite doorway 82 L/min (22 gpm) 20 m2 (16,4 ft x 13 ft) 4,1 mm/min. (0,1 gpm/ft2). (0,1 gpm/ft2). 08:43 21,20 kg 18,59 kg 2,61 kg Partly. 13:08 21,23 kg 19,66 kg 1,57 kg Partly. 142°C 120°C 114°C 95°C 19,0 vol% 0,098 vol% -. 76°C 61°C 119°C 55°C 19,9 vol% 0,089 vol% -. Tests 26 – 30 were conducted with the simulated furniture as the fire source. Test 26 was conducted with a nominal discharge density of 4,1 mm/min. The sprinkler activated at the time the flames reached approximately 4 m up the corner walls. Visually it was observed that the activation of the sprinkler had no immediate effect on the size of the fire, although the temperatures at the ceiling were reduced. The temperature at the “sprinkler” inside the bedroom reached high, 179°C and exceeded 68°C after 02:31, i.e. before the activation of the sprinkler inside the living room. The fire size was gradually reduced as the combustibles of the fire source were consumed. The damages to the walls were quite severe. To determine the influence of the wall-lining, test 27 was conducted similar to test 26, but with gypsum board on the walls. This made the fire develop slower and the water spray reduced the fire at the wall soon after activation. Temperatures measured at the ceiling were significantly lower as compared to test 26. However, the “sprinkler” inside the bedroom “activated” previous to the sprinkler in the living room also in this test (at about 03:09). Test 28 was conducted with particle boards on the corner walls, but with a higher discharge rate, 6,1 mm/min. The sprinkler activated when the flames from the fire reached approximately 4 m above the floor, similar to test 26. However, in this test the fire was suppressed much faster, which limited the damage to the walls and reduced ceiling temperatures. As in the previous two tests the “sprinkler” inside the bedroom activated prior to the sprinkler in the living room, in this case at 02:54 from ignition..

(24) 23. Table 12. Test results for the tests with the horizontal sidewall sprinkler inside the 5,0 m high living room. The simulated furniture was used as the fire source and the sprinkler was positioned above the doorway opening.. Test no. Date of test Wall and ceiling lining Fire source Type of sprinkler Pos. of sprinkler Desired flow rate [L/min] Coverage area [m2] Nominal density [mm/min] Activation time [min:s] Weight of SF before test [kg] Weight of SF after test [kg] Weight loss of SF [kg] Damage to walls Selected measurement data Ch 31, flush w/ ceiling surface Ch 34, flush w/ ceiling surface Ch 30, close to sprinkler Ch 39, “sprinkler” in bedroom Minimum O2 conc. [vol%] Maximum CO conc. [vol%] Note:. 26 27 28 2001-05-22 2001-05-22 2001-05-23 Particle board Gypsum board Particle board Simulated furniture Simulated furniture Simulated furniture Horiz. sidewall Horiz. Sidewall Horiz. sidewall Above doorway Above doorway Above doorway 82 L/min (22 gpm) 82 L/min (22 gpm) 122 L/min (32 gpm) 20 m2 (16,4 ft x 13 ft) 20 m2 (16,4 ft x 13 ft) 20 m2 (16,4 ft x 13 ft) 4,1 mm/min 4,1 mm/min 6,1 mm/min (0,1 gpm/ft2). (0,1 gpm/ft2). (0,15 gpm/ft2). 02:49 10,16 kg 5,92 kg 4,24 kg Severe. 03:23 9,85 kg 6,76 kg 3,09 kg Partly. 03:05 9,94 kg 7,35 kg 2,59 kg Partly. 471°C 297°C 156°C 179°C 15,6 vol% 0,211 vol% -. 162°C 118°C 127°C 92°C 18,9 vol% 0,115 vol% -. 319°C 157°C 128°C 103°C 18,6 vol% 0,137 vol% -. The sprinkler was moved to the position opposite the doorway in test 29 and 30, but still 125 mm below the ceiling of the room. Test 29 was conducted at a nominal discharge density of 4,1 mm/min, test 30 at a 6,1 mm/min density. As expected, the sprinkler activated at an earlier stage of the fire at this position, flames reached to about 3 m above floor level. The activation of the sprinkler during test 29 had an initial effect on the fire at the corner walls, however, the fire redeveloped and was quite intense for a while. The temperatures measured at the ceiling were high, but the water spray cooled the gases enough to prevent the activation of the “sprinkler” inside the bedroom. The results were significantly improved when the discharge density was raised to 6,1 mm/min. The fire at the wall was almost immediately suppressed and the fire was confined to the simulated furniture. Damages to the walls were very limited. A general conclusion for the tests with the simulated furniture was that a nominal discharge density of 6,1 mm/min is necessary in order to effectively control or suppress the fire. The influence of the wall-lining material was also significant. It is clear that the discharge density can be reduced with a less flammable lining..

(25) 24. Table 13. Test results for the tests with the horizontal sidewall sprinkler inside the 5,0 m high living room. The simulated furniture was used as the fire source and the sprinkler was positioned opposite the doorway opening.. Test no. Date of test Wall and ceiling lining Fire source Type of sprinkler Pos. of sprinkler Desired flow rate [L/min] Coverage area [m2] Nominal density [mm/min] Activation time [min:s] Weight of SF before test [kg] Weight of SF after test [kg] Weight loss of SF [kg] Damage to walls Selected measurement data Ch 31, flush w/ ceiling surface Ch 34, flush w/ ceiling surface Ch 30, close to sprinkler Ch 39, “sprinkler” in bedroom Minimum O2 conc. [vol%] Maximum CO conc. [vol%] Note:. 3.4. 29 2001-05-23 Particle board Simulated furniture Horiz. sidewall Opposite doorway 82 L/min (22 gpm) 20 m2 (16,4 ft x 13 ft) 4,1 mm/min. 30 Particle board Particle board Simulated furniture Horiz. sidewall Opposite doorway 122 L/min (32 gpm) 20 m2 (16,4 ft x 13 ft) 6,1 mm/min. (0,1 gpm/ft2). (0,15 gpm/ft2). 9,75 kg 5,56 kg 4,19 kg Severe. 10,73 kg 8,26 kg 2,47 kg Minor. 277°C 329°C 111°C 58°C 15,2 vol% 0,281 vol% -. 83°C 101°C 129°C 50°C 19,9 vol% 0,137 vol% -. Test results for the Ultra Fog water mist nozzle. The first test, test 10, was conducted using the simulated furniture. The nozzle was positioned 2 m by 2 m from the walls, i.e. the coverage area was 16 m2. The nozzle activated when the flames reached to about half the height of the walls. Observations were difficult to make because of the obscuration by smoke. However, after the test it could be concluded that almost all of the polyether foam was consumed and that the wall that faced the fire source was burnt to a height approximately 1,5 m above floor. Test 11 was conducted with the coverage area of the nozzle decreased to 9 m2. The nozzle activated at an early stage of the fire, flames had reached about half the height of the walls. The fire at the walls was immediately suppressed, however, the fire in the polyether foam continued to burn until almost all of the foam was consumed. The damage to the walls was very minor and the temperatures measured at the ceiling were reduced as compared to test 10, and were kept very low. Test 19 was conducted using the upholstered chair as the fire source. For this test, the nozzle was once again positioned 2 m by 2 m from the walls, i.e. the coverage area was 16 m2. However, the system operating pressure was reduced to 60 bar. The chair was ignited at the seat, such that the fire was fully exposed to the water spray. The results can be considered successful, as the fire was limited to the chair, without any damages to the walls. Temperatures measured at the ceiling were very low..

(26) 25. A general observation for the water mist nozzle tests were that a larger amount of fresh air was drawn to the fire, which made it burn more turbulent as compared to the sprinkler tests. It is also clear from the tests that high wall wetting is as desirable for a water mist nozzle as it is for a residential sprinkler in order to be successful in these kind of tests. Table 14. Test results for the tests with the Ultra Fog water mist nozzle. The simulated furniture as well as the upholstered chair (ignited at the seat) was used as the fire source.. Test no. Date of test Wall and ceiling lining Fire source Type of nozzle System pressure [bar] Desired flow rate [L/min] Coverage area [m2] Nominal density [mm/min] Activation time [min:s] Weight of SF before test [kg] Weight of SF after test [kg] Weight loss of SF [kg] Damage to walls Selected measurement data Ch 31, flush w/ ceiling surface Ch 34, flush w/ ceiling surface Ch 30, close to nozzle Ch 39, “sprinkler” in bedroom Minimum O2 conc. [vol%] Maximum CO conc. [vol%] Note:. 10 11 2001-03-19 2001-03-19 Particle board Particle board Simulated furniture Simulated furniture K=1,76 K=1,76 75 bar 75 bar 17,0 L/min (4,5 17,0 L/min (4,5. 19 2001-03-21 Particle board Upholstered chair K=1,76 60 bar 15,0 L/min (4 gpm). gpm). gpm). 16 m2 (13 ft x 13 ft) 1,1 mm/min. 9 m2 (10 ft x 10 ft) 1,9 mm/min. 16 m2 (13 ft x 13 ft) 0,9 mm/min. (0,03 gpm/ft2). (0,05 gpm/ft2). (0,02 gpm/ft2). 01:41 9,78 kg 7,20 kg 2,58 kg Partly. 01:54 10,74 kg 8,05 kg 2,69 kg Minor. 04:25 21,21 kg 18,89 kg 2,32 kg None. 200°C 116°C 130°C 57°C 19,2 vol% 0,132 vol% -. 115°C 90°C 114°C 39°C 19,1 vol% 0,123 vol% -. 79°C 58°C 102°C 44°C 19,9 vol% 0,128 vol% -. Tests 35 and 36 were conducted inside the 5,0 m high living room. Test 35 with the upholstered chair as the fire source and test 36 using the simulated furniture. The nozzle activated when flames reached to half the height of the ceiling in test 35. The fire was suppressed but not immediately, and the fire redeveloped slightly and burned at a reduced level (based on the temperature readings as the visibility was lost after the activation). Temperatures at the ceiling was, however, very low, and the “sprinkler” inside the bedroom peaked at only 47°C. Test 36 was conducted with the simulated furniture as the fire source; in all other respects the test was similar to test 35. The nozzle activated when flames had reached to half the height of the ceiling and the fire at the walls were initially suppressed. However, the fire redeveloped and burned very intense. Temperatures at the ceiling were high and the “sprinkler” inside the bedroom would have activated at 05:23, the maximum temperature was measured to be 73°C. The corner walls were severely burnt after the test..

(27) 26. Table 15. Test results for the tests with the Ultra Fog water mist nozzle. The simulated furniture as well as the upholstered chair (ignited at the left hand side) was used as the fire source. Ceiling height 5,0 m.. Test no. Date of test Wall and ceiling lining Fire source Type of nozzle System pressure [bar] Desired flow rate [L/min]. 35 2001-05-29 Particle board Upholstered chair K=1,76 75 bar 17,0 L/min (4,5 gpm). gpm). Coverage area [m2] Nominal density [mm/min]. 16 m2 (13 ft x 13 ft) 1,1 mm/min. 16 m2 (13 ft x 13 ft) 1,1 mm/min. (0,03 gpm/ft2). (0,03 gpm/ft2). 07:58 21,17 kg. 01:57 10,31 kg. 17,32 kg 3,85 kg Partly. 5,89 kg 4,42 kg Severe. 99°C 65°C 84°C 47°C 19,4 vol% 0,144 vol% -. 413°C 280°C 93°C 76°C 17, 3 vol% 0,211 vol% -. Activation time [min:s] Weight of chair/ SF before test [kg] Weight of chair/SF after test [kg] Weight loss of chair/SF [kg] Damage to walls Selected measurement data Ch 31, flush w/ ceiling surface Ch 34, flush w/ ceiling surface Ch 30, close to nozzle Ch 39, “sprinkler” in bedroom Minimum O2 conc. [vol%] Maximum CO conc. [vol%] Note:. 36 2001-05-29 Particle board Simulated furniture K=1,76 75 bar 17,0 L/min (4,5.

(28) 27. 4. Discussion and conclusions. The majority of the tests were conducted with particle boards on the corner walls, however, a few tests were conducted with gypsum boards as the wall lining. It can be concluded that the type of wall lining has a significant influence on the severity of the fire, given of course that the wall lining becomes involved. During the tests when the fire was ignited at the side of the upholstered chair and when the simulated furniture was used, the fire involved the wall panelling prior to the activation of the sprinkler. In addition, the fire source itself shielded the water spray from directly hitting the base of the fire, which makes the scenario challenging, although not unrealistic. The adjacent bedroom was instrumented with a thermocouple positioned centric at the ceiling. The thermocouple had thermal characteristics similar to a fast response sprinkler bulb and the intention was to determine if a sprinkler would have activated during the test. It was observed that the “sprinkler” inside the bedroom activated prior to the actual sprinkler in several of the tests with the concealed pendent and horizontal sidewall sprinklers.. 4.1. The performance of the different sprinklers. As expected, the recessed pendent sprinkler activated at an earlier stage of the fire compared to the other two types of sprinklers. The tests show that a nominal discharge density of 2,05 mm/min was sufficient to achieve a reasonably good protection when the upholstered chair was used as the fire source. The simulated furniture required a nominal discharge density of 4,1 mm/min to limit the damages to the corner walls, reduce the ceiling temperatures and to prevent the “sprinkler” inside the bedroom from activating. Two tests were conducted inside the 5,0 m high living room. The difference in performance was not significant, but the higher discharge density was required to prevent the activation of the “sprinkler” inside the bedroom. The concealed pendent sprinkler needs to be designed for a higher discharge density due to its low thermal sensitivity. All the tests were conducted with the upholstered chair as the fire source and at a nominal discharge density of 4,1 mm/min. This water density was not sufficient to suppress and control the fire when particle boards were used on the corner walls (2,5 m ceiling height). With gypsum boards on the walls, the fire was controlled, however, the “sprinkler” inside the bedroom activated also in this test. The difference in performance dependent on the wall lining material was not as significant for the tests inside the 5,0 m high living room. The fire was suppressed irrespective of the wall lining. Also in these tests the “sprinkler” inside the bedroom activated. The horizontal sidewall sprinkler was tested at two different positions, either above or opposite the doorway opening. A general observation is that the sprinkler activated relatively later at the position above the doorway opening, which is expected because of the longer distance from the fire source to the sprinkler. The first four tests were conducted with the simulated furniture as the fire source. This fire source required a nominal discharge density of 6,1 mm/min in order to limit the damages to the walls. Ceiling temperatures were never critical in any of these four tests..

(29) 28. The fire in the upholstered chair was easier to deal with, and a water density of 4,1 mm/min was appropriate to control the fire. One test was conducted with an upholstered chair of another type. This chair turned out to be more severe than the chair used throughout the rest of the tests series. The performance of the horizontal sprinkler inside the 5,0 m high living room was similar, 4,1 mm/min was sufficient to control the fire in the upholstered chair, but 6,1 mm/min was necessary to control the fire in the simulated furniture. Tests were also conducted using a high-pressure water mist nozzle. A general observation for the water mist nozzle tests were that a larger amount of fresh air was drawn to the fire, which made it burn more turbulent as compared to the sprinkler tests. It is also clear from the tests that high wall wetting is as desirable for a water mist nozzle as it is for a residential sprinkler in order to be successful.. 4.2. Other observations and conclusions. The design and installation guidelines of NFPA 13D and 13R require that the water supply is designed for two or four, respectively, sprinklers. However, the standards allow that the design can be reduced to the number of sprinklers installed inside a compartmented room. These tests show that there is a risk that additional sprinklers outside of a protected room activate in a residential fire scenario. The tests also show that a high discharge density not is a guarantee for the prevention of the activation of additional sprinklers. The “sprinkler” inside the adjacent bedroom “activated” prior to the sprinkler inside the living room in several of the tests. The activation of the sprinkler drove the smoke down to the floor, which reduced the visibility inside the room. The higher the water density, the faster this occurred. The influence of the wall-lining material was significant. It is clear from the tests that the discharge density can be reduced with a less flammable lining. During the ‘Los Angeles Sprinkler Test Program’ the maximum allowed eye-level concentration of CO was judged to be 3000 ppm (0,3 vol%) and the integrated value of CO concentration over the entire test period to be less than 43 000 ppm-min. It can be concluded that the maximum value of 3000 ppm (0,3 vol%) not was exceeded in any of the tests in this test programme..

(30) 29. References [1]. Arvidson, Magnus, “An Initial Evaluation of Different Residential Sprinklers using Heat Release Rate Measurements”, SP Report 2000:18, SP Swedish National Testing and Research Institute, Borås, 2000.. [2]. Arvidson, Magnus, “En sammanställning av väldokumenterade försök med bostadssprinkler (A compilation of experiences from well-documented residential sprinkler fire tests)”, SP Rapport 2001:03, SP Sveriges Provnings- och Forskningsinstitut, Borås, 2001 (in Swedish only)..

(31)

(32)

(33)

(34) A1. Appendix A. Test 1. Temperature (degrees C). 500 Ch 31, flush with ceiling Ch 32, 10 mm below ceiling Ch 33, 50 mm below ceiling. 400 300 200 100 0 0. 5. 10. Time (min). 15. 20. 25. Test 1. Temperature (degrees C). 500 Ch 34, flush with ceiling Ch 35, 10 mm below ceiling Ch 36, 50 mm below ceiling. 400 300 200 100 0 0. 5. 10. Time (min). 15. 20. 25. Test 1. Temperature (degrees C). 200 Ch 30, close to sprinkler Ch 39, "sprinkler" inside bedroom Ch 40, close to "sprinkler" inside bedroom. 150. 100. 50. 0 0. 5. 10. Time (min). 15. 20. 25.

(35) A2. Ch 7 - 0,6 m above floor level Ch 71 - 1,6 m above floor level. Ch 5 - 0,6 m above floor level Ch 73 - 1,6 m above floor level. Test 1. 1,2. 21. 1. 20. 0,8. 19,5. 0,6. 19. 0,4. 18,5. 0,2 0. 18 0. 5. 10. 15 Time (min). 20. 25. Test 1 0,02. Carbon Oxide %. 0,015 Ch 6 - 0,6 m above floor level Ch 72 - 1,6 m above floor level 0,01. 0,005. 0. 0. 5. 10. 15 Time (min). 20. 25. Carbon dioxide %. Oxygen %. 20,5.

(36) A3. Test 2. Temperature (degrees C). 500 Ch 31, flush with ceiling Ch 32, 10 mm below ceiling Ch 33, 50 mm below ceiling. 400 300 200 100 0 0. 5. 10. Time (min). 15. 20. 25. Test 2. Temperature (degrees C). 500 Ch 34, flush with ceiling Ch 35, 10 mm below ceiling Ch 36, 50 mm below ceiling. 400 300 200 100 0 0. 5. 10. Time (min). 15. 20. 25. Test 2. Temperature (degrees C). 200 Ch 30, close to sprinkler Ch 39, "sprinkler" inside bedroom Ch 40, close to "sprinkler" inside bedroom. 150. 100. 50. 0 0. 5. 10. Time (min). 15. 20. 25.

(37) A4. Ch 7 - 0,6 m above floor level Ch 71 - 1,6 m above floor level. Ch 5 - 0,6 m above floor level Ch 73 - 1,6 m above floor level. Test 2. 3. 21. 2,5 2. 20. 1,5. 19,5. 1. 19. Carbon dioxide %. Oxygen %. 20,5. 0,5. 18,5. 0. 18 0. 5. 10. 15 Time (min). 20. 25. Test 2 0,04 0,035 Carbon Oxide %. 0,03 0,025 0,02 0,015. Ch 6 - 0,6 m above floor level Ch 72 - 1,6 m above floor level. 0,01 0,005 0. 0. 5. 10. 15 Time (min). 20. 25.

(38) A5. Test 3. Temperature (degrees C). 500 Ch 31, flush with ceiling Ch 32, 10 mm below ceiling Ch 33, 50 mm below ceiling. 400 300 200 100 0 0. 5. 10. Time (min). 15. 20. 25. Test 3. Temperature (degrees C). 500 Ch 34, flush with ceiling Ch 35, 10 mm below ceiling Ch 36, 50 mm below ceiling. 400 300 200 100 0 0. 5. 10. Time (min). 15. 20. 25. Test 3. Temperature (degrees C). 200 Ch 30, close to sprinkler Ch 39, "sprinkler" inside bedroom Ch 40, clode to "sprinkler" inside bedroom. 150. 100. 50. 0 0. 5. 10. Time (min). 15. 20. 25.

(39) A6. Ch 7 - 0,6 m above floor level Ch 71 - 1,6 m above floor level. Ch 5 - 0,6 m above floor level Ch 73 - 1,6 m above floor level. Test 3. 3. 21. 2,5 2. 20. 1,5. 19,5. 1. 19. 0,5. 18,5. 0. 18 0. 5. 10. 15 Time (min). 20. Test 3. Ch 6 - 0,6 m above floor level Ch 72 - 1,6 m above floor level. 0,1. 25. Carbon Oxide %. 0,08 0,06 0,04 0,02 0. Carbon dioxide %. Oxygen %. 20,5. 0. 5. 10. 15 Time (min). 20. 25.

(40) A7. Test 4. Temperature (degrees C). 500 Ch 31, flush with ceiling Ch 32, 10 mm below ceiling Ch 33, 50 mm below ceiling. 400 300 200 100 0 0. 5. 10. Time (min). 15. 20. 25. Test 4. Temperature (degrees C). 500 Ch 34, flush with ceiling Ch 35, 10 mm below ceiling Ch 36, 50 mm below ceiling. 400 300 200 100 0 0. 5. 10. Time (min). 15. 20. 25. Test 4. Temperature (degrees C). 200 Ch 30, close to sprinkler Ch 39, "sprinkler" inside bedroom Ch 40, close to "sprinkler" inside bedroom. 150. 100. 50. 0 0. 5. 10. Time (min). 15. 20. 25.

(41) A8. Ch 7 - 0,6 m above floor level Ch 71 - 1,6 m above floor level. Ch 5 - 0,6 m above floor level Ch 73 - 1,6 m above floor level. Test 4. 3. 21. 2,5 2. 20. 1,5. 19,5. 1. 19. 0,5. 18,5. 0. 18 0. 5. 10. 15 Time (min). 20. 25. Test 4 0,16 0,14. Ch 6 - 0,6 m above floor level Ch 72 - 1,6 m above floor level. Carbon Oxide %. 0,12 0,1 0,08 0,06 0,04 0,02 0. 0. 5. 10. 15 Time (min). 20. 25. 30. Carbon dioxide %. Oxygen %. 20,5.

(42) A9. Test 5. Temperature (degrees C). 1000 Ch 31, flush with ceiling Ch 32, 10 mm below ceiling Ch 33, 50 mm below ceiling. 800 600 400 200 0 0. 5. 10. Time (min). 15. 20. 25. 15. 20. 25. Test 5 800 Temperature (degrees C). 700. Ch 34, flush with ceiling Ch 35, 10 mm below ceiling Ch 36, 50 mm below ceiling. 600 500 400 300 200 100 0 0. 5. 10. Time (min). Test 5. Temperature (degrees C). 200 Ch 30, close to sprinkler Ch 39, "sprinkler" inside bedroom Ch 40, close to "sprinkler" inside bedroom. 150. 100. 50. 0 0. 5. 10. Time (min). 15. 20. 25.

(43) A10. Ch 5 - 0,6 m above floor level Ch 73 - 1,6 m above floor level. Test 5. 21. 10. 20. 8. 19. 6. 18. 4. 17. 2. Carbon dioxide %. Oxygen %. Ch 7 - 0,6 m above floor level Ch 71 - 1,6 m above floor level. 0. 16 0. 5. 10. 15 Time (min). 20. 25. Test 5 0,16 Ch 6 - 0,6 m above floor level Ch 72 - 1,6 m above floor level. 0,14 Carbon Oxide %. 0,12 0,1 0,08 0,06 0,04 0,02 0. 0. 5. 10. 15 Time (min). 20. 25. 30.

(44) A11. Test 6. Temperature (degrees C). 500 Ch 31, flush with ceiling Ch 32, 10 mm below ceiling Ch 33, 50 mm below ceiling. 400 300 200 100 0 0. 5. 10. Time (min). 15. 20. 25. 15. 20. 25. 15. 20. 25. Test 6. Temperature (degrees C). 500 Ch 34, flush with ceiling Ch 35, 10 mm below ceiling Ch 36, 50 mm below ceiling. 400 300 200 100 0 0. 5. 10. Time (min). Temperature (degrees C). Ch 30, close to sprinkler Test 6 Ch 39, "sprinkler" inside bedroom 200 Ch 40, close to "sprinkler" inside bedroom 150. 100. 50. 0 0. 5. 10. Time (min).

(45) A12. Ch 7 - 0,6 m above floor level Ch 71 - 1,6 m above floor level. Ch 5 - 0,6 m above floor level Ch 73 - 1,6 m above floor level. Test 6. 3. 21. 2,5 2. 20. 1,5. 19,5. 1. 19. 0,5. 18,5. 0. 18 0. 5. 10. 15 Time (min). 20. 25. 30. 20. 25. 30. Test 6 0,2 Ch 6 - 0,6 m above floor level Ch 72 - 1,6 m above floor level. Carbon Oxide %. 0,15. 0,1. 0,05. 0. 0. 5. 10. 15 Time (min). Carbon dioxide %. Oxygen %. 20,5.

(46) A13. Test 7. Temperature (degrees C). 1000 Ch 31, flush with ceiling Ch 32, 10 mm below ceiling Ch 33, 50 mm below ceiling. 800 600 400 200 0 0. 5. 10. Time (min). 15. 20. 25. Test 7 800 Ch 34, flush with ceiling Ch 35, 10 mm below ceiling Ch 36, 50 mm below ceiling. Temperature (degrees C). 700 600 500 400 300 200 100 0 0. 5. 10. Time (min). 15. 20. 25. Test 7. Temperature (degrees C). 200 Ch 30, close to sprinkler Ch 39, "sprinkler" inside bedroom Ch 40, close to "sprinkler" inside bedroom. 150. 100. 50. 0 0. 5. 10. Time (min). 15. 20. 25.

(47) A14. Ch 7 - 0,6 m above floor level Ch 71 - 1,6 m above floor level. Ch 5 - 0,6 m above floor level Ch 73 - 1,6 m above floor level. Test 7. 3. 21. 2,5 2. 20. 1,5. 19,5. 1. 19. 0,5. 18,5. 0. 18 0. 5. 10. 15 Time (min). 20. 25. Test 7 0,14 Ch 6 - 0,6 m above floor level Ch 72 - 1,6 m above floor level. Carbon Oxide %. 0,12 0,1 0,08 0,06 0,04 0,02 0. 0. 5. 10. 15 Time (min). 20. 25. Carbon dioxide %. Oxygen %. 20,5.

No results found