Vapour & Fire Control Testing of FOAMGLAS PFS System (Gen 1) on LNG and LPG. Final Report

Transcription

1 Walker House, George Street, Aylesbury, Bucks, HP20 2HU, United Kingdom. Tel: (01296) Fax (01296) Web: Vapour & Fire Control Testing of FOAMGLAS PFS System (Gen 1) on LNG and LPG 2014

2 REVISION RECORD ISSUE DATE ORIGINATOR REVIEWED APPROVED DESCRIPTION DRAFT KM PW Issued for Review REV 1 - KM PW Revised incorporating client comments FINAL KM PW RPI Ref: P February 2014

3 EXECUTIVE SUMMARY Falck Emergency Services UK, also trading as Resource Protection International (RPI), an Independent Specialist Fire Consultants, was appointed by Total Oil Company to undertake a series of tests. The tests evaluated the effectiveness of passive firefighting systems, in suppressing LNG and LPG fires and vapours. Resource Protection International (RPI) was commissioned by Total to undertake this series of tests in October 2013 and the tests were undertaken at the Centro Jovellanos, Asturias, Spain, during the week commencing Monday 14 October The testing team was led by Paul Watkins of RPI. Four tests were conducted and the results of the tests featuring the use of the passive material (FOAMGLAS PFS System (Gen 1)) were compared to the baseline data obtained from the reference tests. FOAMGLAS PFS System (Gen 1) is effective in concrete LNG containment pits, as demonstrated by the test results: under the ideal experimental conditions, it was apparent that there was a significant reduction in radiant heat and a relatively short fire control time. The purpose of this report is to provide the results of the series of tests that were conducted by RPI, outlining all observations and subsequent conclusions. It should be noted that the conclusions may be updated and/or revised further, following a more detailed analysis of the results. Based on what has been observed so far, with the test work having been undertaken, it is possible to make some general statements about the potential applicability of the FOAMGLAS PFS System (Gen 1) under testing. It should be noted that any findings of the experiments were based on experimental conditions; thus, any practice should be adjusted to the varying conditions. 1. LNG and LPG reference tests were performed, in order to compare the test results with regards to the FOAMGLAS PFS System (Gen 1) with baseline results. 2. During the continuous discharge of the LNG, the vapour cloud appeared larger than when the LNG flow was stopped. This occurred when LNG was released and a fraction of the discharged LNG flashed directly, without creating the pool. Thus, the LNG vapour largely derived from the flashed LNG. This did not occur when the LNG flow was stopped and the only source of the LNG vapour was the evaporating pool. This was confirmed by observing the size of the white vapour cloud during both phases. 3. FOAMGLAS PFS System (Gen 1) was effective in reducing radiant heat flux. The observation and experiment results demonstrate that the FOAMGLAS PFS System (Gen 1) controls fire gradually after ignition. 4. With regards to FOAMGLAS PFS System (Gen 1), only the top of the cubes were damaged. It was also observed that the bottom layer of the polyethylene green bags was protected. 5. FOAMGLAS PFS System (Gen 1) provided a consistent coverage, which helped to reduce the flame size of the LNG and LPG and ensured that the fire was in a steady state within a short space of time. This result can be explained by the fact that the burning rate was significantly lessened by reducing the exposed area of the LNG or LPG pool. RPI Ref: P February 2014

4 6. FOAMGLAS PFS System (Gen 1) does not depend on any additional procedure or supplemental application during the fire. RPI Ref: P February 2014

5 Table of Contents EXECUTIVE SUMMARY INTRODUCTION THE TEST SITE AND INSTRUMENTATIONS LNG test pit LPG test pan and delivery system Heat Flux Gauges Cooling system Gas Detection Meteorological station Data acquisition Visual Records Other Logistics TEST PROGRAMME Reference tests FOAMGLAS PFS System (Gen 1) tests TEST PROCEDURE Procedure of reference tests Procedure of FOAMGLAS PFS System (Gen 1) tests THE EXPERIMENTAL WORK AND RESULTS Test 1: LNG reference test Test 2: LNG test using FOAMGLAS PFS System (Gen 1) Test 3: LPG reference test Test 4: LPG Test using FOAMGLAS PFS System (Gen 1) COMPARISON OF RESULTS WITH THE REFERENCE TESTS Test 2:LNG test incorporating the FOAMGLAS PFS System (Gen 1) Test 4: LPG test incorporating the FOAMGLAS PFS System (Gen 1) 37 7 SUMMARY AND RECOMMENDATIONS RPI Ref: P February 2014

6 1 INTRODUCTION The release of flammable gases, such as Liquefied Natural Gas (LNG) or Liquefied Petroleum Gas (LPG), may result in the formation of a flammable vapour cloud that is often dense, depending on the ambient conditions (particularly wind speed and direction). If this cloud encounters a source of ignition within the surrounding environment, a fire may occur and this might have catastrophic consequences for nearby facilities. The characteristics of the resulting fire are dependent upon the release conditions and the environment into which the vapours are released. Various scenarios may occur, including flash fires, fireballs, pool or jet fires or even vapour cloud explosions, if the cloud is released into a congested area. LNG and LPG pool fires emit high thermal radiation, which may prevent firefighters approaching and extinguishing the fire. Expansion foams and dry chemicals are the conventional methods used to control and suppress vapour clouds and fire resultant from LNG and LPG. Expansion foams are used to reduce radiant heat to a level where the LNG/LPG pool fires are controlled, thus allowing firefighters to approach and extinguish the fire, using dry chemicals. Applying expansion foam to the surface of LNG can reduce radiant heat from a fire: this is because the water drained from the foam forms a frozen layer, which acts as an insulation blanket, reduces the evaporation rate and subsequently decreases radiant heat. The application of these methods depends on a number of factors, such as foam application rate, the location of the generator and the LNG/LPG containment design. Recently, there have been a number of studies conducted, as attempts to fill the gaps increased the need for LNG/LPG field experiments involving expansion foams. Various materials, such as the FOAMGLAS PFS System (FOAMGLAS is a registered trademark of Pittsburgh Corning Corporation in the United States and various other countries), have been introduced in the last decade, in order to suppress fires resulting from flammable gases and reduce the effect of radiant heat on nearby objects by reducing the size of the fire. In addition, the FOAMGLAS PFS System is free of the disadvantages associated with traditional fire-fighting foams and thus can help overcome the various difficulties that fire fighters face in using high expansion foam, such as: 1. The requirement of large quantities of water, foam and dry chemicals to ensure complete control and extinguishment of large LNG/LPG fires. 2. The necessity for the continuous application of high expansion foam, as the foam blanket can be weakened by fire and it may be necessary to replace it with a new one. FOAMGLAS PFS System is a non-flammable dry foam material that acts as a floating barrier in insulating the surface of a burning liquid. The density of FOAMGLAS PFS System is less than one third of the density of LNG (115 kg/m3 (+/-15%)) and it has a completely closed cell structure, in order to prevent the absorption of LNG/LPG during contact. FOAMGLAS PFS System can be easily arranged in the shape of the containment of the spill. The work undertaken in this report was performed by Resource Protection International (RPI). Data was recorded, with regards to investigating the performance of FOAMGLAS PFS System (Gen 1) in suppressing and controlling the fires of LNG/LPG. To date, relatively few experimental investigations have been undertaken, particularly in terms of LPG fire control. During the week commencing Monday 14 October 2013, RPI carried out a programme of tests, which were sponsored by Total Oil Company. The tests were performed at Centro Jovellanos, Asturias, Spain. RPI Ref: P February 2014

7 The principal objectives of the experimental programme, as outlined in this report, were to investigate the effectiveness of the FOAMGLAS PFS System (Gen 1) on the reduction of radiant heat flux pertaining to LNG/LPG pool fires. There was also a further aim to investigate the effectiveness of the FOAMGLAS PFS System (Gen 1) in suppressing the evolution of vapour from pools of the aforementioned hydrocarbons. Specifically, the purpose of the experiments was to obtain data on LNG/LPG fires and vapours, including vapour thermal radiative output and fire size, in terms of the use of the FOAMGLAS PFS System (Gen 1). In order to achieve the above objectives, four tests were conducted; in which radiant was measured using four TG F radiometers, manufactured by Vatell. Vapour concentration was measured using Crowcon Gasman gas detectors calibrated for methane (for the LNG tests) and propane (for the LPG tests). In addition, weather conditions, including wind speed, wind direction, temperature and humidity, were continuously measured throughout the duration of the tests. The Centro Jovellanos (CJ) industrial/marine fire training facility was used as a base for the tests. The CJ facility LNG pit (with dimensions of 2m x 2m x 1.2m depth) was used for the LNG tests, while a specially fabricated stainless steel pan, with delivery system, was used for the LPG tests. The dimensions of the pan were 1m x 1m x 0.6m depth. Measurements of radiant heat without use of the FOAMGLAS PFS System (Gen 1) were recorded for both LNG and LPG, in order to compare the results with the measurements gained when the FOAMGLAS PFS System (Gen 1) was used. General observations and the test results clearly indicated a reduction in radiant heat when the FOAMGLAS PFS System (Gen 1) was used, which was evident by a decrease in flame size. The purpose of this report is to deliver the results of the experiment and any observations. RPI reserve the right to modify the contents of the report, based on the on-going analysis of the test results. RPI Ref: P February 2014

8 2 THE TEST SITE AND INSTRUMENTATIONS The tests were carried out at Centro Jovellanos, Asturias, Spain. The test site comprised of an LNG storage facility, an LNG release system and an LNG test pit, into which the LNG was discharged. The experiments were monitored from a control room located to the south of the LNG test pit, approximately 100m from the LNG release point. All sited instrumentation was wired back to the control room, using appropriate cables. The data generated during the tests were recorded on a computer, which was also located in the control room. During the tests, all personnel, with the exception of look-outs, were positioned near the control room. LNG Test Pit Figure 1: The Test Site at Centro Jovellanos, Asturias, Spain LNG storage Supply line 2.1 LNG test pit Figure 2: LNG Storage and Delivery System A diagram of the LNG pool fire test pit is shown in Figure 3: the dimensions of the pit were 2m x 2m x 1.2m depth. The pit was constructed from concrete and was reinforced with weld mesh and rebar. RPI Ref: P February 2014

9 1m high (removable) safety fence around the pit Grade 150mm LNG Pit 2m x 2m x 1.2m deep Walls 150mm thick, reinforced with weld mesh and RE-BAR Figure 3: The LNG Test Pit 2.2 LPG test pan and delivery system A square stainless steel pan was used for the LPG tests, as shown in Figures 4 and 5 below. The dimensions of the pan were 1m x 1m x 0.6m depth. The LPG was delivered to the test pan using a special delivery system that was fabricated particularly for this purpose, as shown in Figure 6. Two LPG cylinders (12.5kg) were emptied into the test pan at the same time and were easily replaced, in order to achieve the desired level of LPG liquid. Each cylinder outlet pipe was fitted with a manually operated valve and the outlets from the two cylinders were manifolded together, into a single long discharge pipe. The discharge pipe was mounted horizontally, at a height of about 0.5 m above ground, and was orientated towards the LPG test pan. Figure 4: LPG Test Pan RPI Ref: P February 2014

10 0.6m LPG Test Pan 1 m 1 m LPG Test Pan 1 m Figure 5: LPG Test Pan Figure 6: LPG Delivery System RPI Ref: P February 2014

11 2.3 Heat Flux Gauges During the experiments, four heat-flux gauges were used to measure the rate of heat flux from the fire. The heat flux gauge used was type TG9000-9F, manufactured by the Vatell Corporation (as seen in Figure 7). The gauges are water cooled, when constructed as a radiometer: convective heat flux is blocked from reaching the sensor by a window: this allows radiation only to be measured. This type of gauge has a maximum operating temperature of 200 C, regardless of the heat flux applied. The heat flux gauges had a 120 field of view and were tilted downwards, at an angle of 12 o below the horizontal line, so that they pointed towards the base of the flame: this allowed the approximate maximum radiant heat flux at each location to be recorded. In the experiments, the gauges were calibrated in accordance with the manufacturer s recommendations and instructions and the output of the heatflux gauge was measured in mvs. The gauge is designed to provide an approximate output signal of 10 mv, when exposed to the specified heat flux range. A conversion factor was obtained, so that the measured voltages were converted to true heat flux, in W/cm². Each heat flux gauge was mounted on a tripod stand located at height of 1.2m above the base of the flame. The gauges were placed at the corners of the LNG pit, at a distance of 5m from the pit. With regards to the LPG tests, two gauges were used and were located at a distance of 2m away from two sides of the test pan. The gauges were water-cooled, in order to avoid damaging them. Before each of the tests, all cables from the heat flux gauges to the data loggers and from the data loggers to the computers were checked, as was the water cooling system. The heat flux gauges were cleaned, as they were exposed to vapour clouds, smoke and dry chemicals on a number of occasions. The results of the tests were analysed, in order to obtain measurements of radiant heat flux, using the calibration data supplied with each of the heat flux gauges. The results obtained confirmed that there was no deterioration of the heat flux gauges, in terms of them being exposed to dry chemicals or vapour clouds. Figure 7: Heat Flux Gauge Type TG9000-9F RPI Ref: P February 2014

12 2.4 Cooling system Water cooling was recommended to prevent damage to the heat-flux gauges, in the event of them being exposed to fire for long periods of time. A water cooling system was provided, consisting of two cooling water pumps, rubber tubes and two water tanks. The water tanks and pumps were located close to the control room, approximately 100m from the LNG pit/lpg pan, and the rubber tubes close to the pit were protected by thermal protective material. 2.5 Gas Detection There are two types of gas detectors that may have been used to measure LNG and LPG gas concentration in the air during the experiment: point gas detectors and portable gas detectors. Point gas detectors, such as the Searchpoint Optima Plus model, are designed for industrial application, in terms of the detection of any methane leaks and the concentration of the gas. The use of this type of detector involves suction of the gas by applying a vacuum into the cell of the gas detector. While it was possible to install the gas detector directly within the field for the experiment, this configuration is not flexible, due to the size and the weight of this type of gas detector. The movement of LNG/LPG vapour dispersion is very dynamic, depending on wind speed and direction; thus, it is important to set up the gas detector in order to ensure easy relocation. As a result, it is recommended that all gas detectors are placed together, with the measurement cells connected with tubes of at least 30m in length. The tubes should also be of the same type and length, in order to increase accuracy. This type of gas detector is relatively expensive and the setup is complicated. However, an advantage is that the concentration profile can be easily obtained in the form of (time/concentration). The second type of gas detector is the portable gas detector, and it was this type that was used in the experiments. The application of portable gas detectors includes measuring gas concentration at certain points where the point gas detectors are not available, in addition to ensuring the safety of the person conducting the experiment. Portable gas detectors have the capability to store data, using their built-in memory. The disadvantages of this type of detector are that the power is disconnected when the reading reaches overrange, in order to avoid damage to the sensor, and the reading is produced in the form of peaks. Fourteen gas detectors were used to investigate the effectiveness of FOAMGLAS PFS System (Gen 1) in suppressing the evolution of vapour from the LNG and LPG pools. The detectors used were Crowcon Gasman personal gas detectors, as shown in Figure 8, and were mounted on pole stands. They were located in the downwind direction of the LNG pit/lpg pan, at various distances. The vapours were measured with and without the use of FOAMGLAS PFS System (Gen 1). RPI Ref: P February 2014

13 Figure 8: Crowcon Gasman Gas Detector 2.6 Meteorological station Ambient conditions during the tests, particularly wind-speed, may have had an effect on the measurement of heat flux and the evaporation rate of LNG/LPG; thus, a meteorological station was used to monitor such conditions. Wind-speed and direction, humidity, barometric pressure and temperature were measured. The temperature and humidity sensors and the 3-cup anemometer, featuring a wind vane, were positioned 1.5m above ground. Figure 9: The Meteorological Station 2.7 Data acquisition Radiant heat flux was one of the most important measurements obtained from the pool-fire tests. For all experiments, data were recorded using the data logger MS6D, as shown in Figure 10. Two data loggers were used and each was connected to two heat-flux gauges. The voltage output from each test was measured in millivolts (mvs) and the conversion factor for the sensor was applied, in order to yield heat flux in W/cm 2. The data loggers for each of the RPI Ref: P February 2014

14 heat flux gauges were located as close to the gauges as possible and were covered with a protective lid, in order to shield them from the flames. The signal from the data loggers was transmitted to the control room by Ethernet cables, which were protected by thermal resistant sleeves where they were likely to be exposed to fire or radiant heat. 2.8 Visual Records Figure 10: Data Logger MS6D Two digital video cameras were used to produce visual recordings of the tests. The cameras were positioned to the side and either upwind or downwind of the test pit/pan, in order to provide different views, in accordance with the requirements of each test. These cameras can be linked to a computer, so that individual frames may be downloaded for analysis. 2.9 Other Logistics Cables Cables were required for the heat flux gauges and data and power supply and it is estimated that 400m of cable was used. Tripods Tripods were utilised to position heat flux gauges, gas detectors and the weather station and it is estimated that twelve tripods were used. Fire Resistance Insulation This was required to protect the cables, wires and water cooling tubes close to the LNG pit and the LPG pan. Fire resistant insulation sleeves were used. RPI Ref: P February 2014

15 3 TEST PROGRAMME The total experimental programme consisted of four individual tests. Radiant heat flux and the vapour concentration of the LNG and LPG pools were measured before and after the FOAMGLAS PFS System (Gen 1) was applied. Vapour concentration (%LEL) was measured downwind, using Crowcon Gasman detectors placed at various distances from the LNG test pit and the LPG test pan. The distances were subject to change, depending on the direction of the wind. The gauges were placed at the corners of the LNG pit, at a distance of 5m from the pit. With regards to the LPG tests, two gauges were used and were located at a distance of 2m away from two sides of the test pan. 3.1 Reference tests As the main aim of the experimental programme was to evaluate the performance of the FOAMGLAS PFS System (Gen 1) in suppressing fire pertaining to LNG and LPG fuels, it was also necessary to conduct tests without the incorporation of this system (reference tests), in order to collect baseline data that could be compared against the measurements taken when the FOAMGLAS PFS System (Gen 1) was used. Thus, two reference tests were conducted: one for the LNG and one for the LPG. A further reference test for the LNG was conducted in the LPG pan, in order to compare the radiant heat associated with LNG with that of the LPG. In terms of the concentration of the vapour cloud, this was measured after the discharge of the LNG/LPG liquid. The fuel was then ignited and radiant heat flux was measured. 3.2 FOAMGLAS PFS System (Gen 1) tests Two FOAMGLAS PFS System (Gen 1) tests were performed: the LPG test was conducted in the LPG pan and the LNG test was conducted in the LNG pit. In both tests, the FOAMGLAS PFS System (Gen 1) was placed into the test pit/pan and vapour concentration was measured, upon the discharge of the LPG/LNG. The fuel was then ignited and radiant heat flux measured. The fire was extinguished using dry powder. Test No Type of Test LNG Tests Measured radiant heat flux and vapour concentration without product (LNG reference test-1) in the LNG test pit. Measured vapour concentration and heat flux with the use of the product. LPG Tests Measured heat flux and vapour concentration without the use of the product. Measured vapour concentration and heat flux with the use of the product. Product None FOAMGLAS PFS System (Gen 1) None FOAMGLAS PFS System (Gen 1) 4 TEST PROCEDURE Radiant heat flux from the LNG/LPG pool fires, with and without the use of the FOAMGLAS PFS System (Gen 1), was recorded. Vapour concentration measurements for the LNG/LPG fuels were also taken before and after application of the FOAMGLAS PFS System (Gen 1). With regards to the tests conducted in the LNG pit, four heat flux gauges were placed at the corners of the LNG pit and seven gas detectors were placed close RPI Ref: P February 2014

16 to the pit, in order to measure vapour concentration. The detectors were located at various distances, in a downwind direction, to cover the majority of the area where the vapour cloud was dispersed. For the LPG tests, two heat flux gauges and six gas detectors were used and were placed close to the LPG pan. Weather conditions were recorded for all tests. Two camcorders were used to record the events of the tests and the size of the flames from the fires. The test procedures are set out below. 4.1 Procedure of reference tests All equipment was checked, ensuring that the heat flux gauges were located at a distance of 5m from the edge of the LNG test pit and 2m from the LPG pan, in the case of LPG tests. Atmospheric conditions were recorded, the dataloggers were checked (for recording data) and the cooling system was ready. The cooling pumps were activated and it was ensured that the water flowed through the heat-flux gauges at the minimum required flow rate of 60mL/minute. The gas detectors were switched on. The data loggers began to record data and baseline data. For the LNG test, the fuel was released until there was a 750 kg 1.5 m 3 pool of fuel in the test pit. For the LPG test, the fuel was released until there was a 37.5kg (0.7 m 3 ) pool of fuel in the test pan. Measurements of vapour concentration (%LEL) were taken downwind for 15 to 20 minutes. The gas detectors were moved to a safe area. The LNG or LPG was ignited and the resulting fire was allowed to achieve steady burning conditions. Radiant heat flux data were collected for 5 to 15 minutes. The fire was extinguished. A further 3 minutes of data were collected after the final extinguishment of the fire. The data loggers were then switched off. 4.2 Procedure of FOAMGLAS PFS System (Gen 1) tests The test began with the placement of the FOAMGLAS PFS System (Gen 1) in the LNG test pit or the LPG test pan, when the test pit/pan was completely empty. All equipment was checked, ensuring that the atmospheric conditions were being recorded and that the data loggers were ready to record the data. The heat-flux gauges were located at the designated distances from the edge of the test pit or the test pan and the gas detectors were ready. The gas detectors were switched on. Data logging began and 3 minutes of baseline data were collected. In terms of the LNG, the fuel was released on top of the FOAMGLAS PFS System (Gen 1) layer, until there was a 750 kg 1.5 m 3 pool of fuel in the pit. For the LPG, the fuel was released on top of the FOAMGLAS PFS System (Gen 1) layer until there is a 37.5 kg (0.7 m 3 ) pool of fuel in the pan. Measurements of vapour concentration (%LEL) were taken downwind. The measurement continued for a time period of at least 15 minutes. The gas detectors were moved to a safe area. The LNG or LPG was ignited and time allowed for the resulting fire to achieve steady burning conditions. Radiant heat flux data was recorded for at least 30 minutes. The final step was the extinguishing of the pool fire after being controlled (or suppressed) by the FOAMGLAS PFS System (Gen 1). This step was completed RPI Ref: P February 2014

17 by applying dry chemical for a few seconds, until the pool fire was completely extinguished. A further 3 minutes of data were collected after the final extinguishment of the fire. The data loggers were then switched off. RPI Ref: P February 2014

18 5 THE EXPERIMENTAL WORK AND RESULTS A summary of each of the tests is provided below. There are notes on each individual test, along with a schematic diagram of the test set-up and details of any test in which the set-up or procedure differed. The radiant heat flux data were analysed using the calibrations supplied by the manufacturer and the measured data were converted into radiant heat flux measurements using the calibration parameters provided for each of the heat flux gauges. The concentration of the vapour taken into account was the peak point, the highest concentration measured, over measurement prior to ignition. The charts below show the heat flux measured by each of the gauges. Further data relating to flame size and behaviour were produced through an analysis of the video records. 5.1 Test 1: LNG reference test-1 The test began with the discharge of a certain quantity (750 kg 1.5 m 3 ) of LNG liquid into the pit, forming a depth of approximately 6 inches of LNG liquid (this assumption is based upon previous preparatory tests and takes into account the initial boiled-off quantity of LNG liquid). Gas detectors were place as shown in Figure 11. Gas concentration was measured for approximately 22 minutes. The gas detectors were then removed and the LNG was ignited, in order to measure radiant heat flux for a period of approximately 6 minutes. The fire was then extinguished, using dry chemical. The table below features the timeline of the test. Tim (min) Action 0:00 Data logging began 01:00 Cooling pumps started 03:00 Gas detectors switched on 05:00 LNG valve opened and 750kg of LNG liquid discharged into the pit 17:00 LNG valve closed 27:00 Gas detectors moved to a safe place 30:00 LNG ignited 36:00 The fire was extinguished, using a dry chemical Weather conditions were constantly measured for the duration of the test and the average values for these are shown in the table below. Wind Speed (m/s) Wind Direction Temperature ( o C) Humidity (%) 0 to 0.4 South The gas detectors were placed as shown in the diagram below. As soon as the LNG liquid touched the concrete base of the pit, it vaporised. A visible white condensate cloud was then produced, as the very cold vapour condensed the water content in the air. The white vapour cloud was observed to disperse in all directions, as a result of the very low wind speed. White cloud formation is RPI Ref: P February 2014

19 largely dependent upon the humidity of the ambient air, so it may or may not represent the actual size of an LNG cloud. After several minutes, the water spray curtains were turned on, in order to disperse the cloud, while the LNG continued to be released into the test pit. Figure 12 below features a photograph depicting the test. D G C A B 2m 4m 2m E F 6m 2m The Test Pit 2x2m 2m Figure 11: Diagram showing the layout of the gas detectors for test 1 (the reference test-1) Figure 12: LNG is released into the test pit and gas concentration is measured RPI Ref: P February 2014

20 Gas concentration (%LEL) 5m 5m Vapour and Fire Control Testing of FOAMGLAS PFS System (Gen 1) on LNG and LPG The figure below shows the LEL % peak points A B C D E F G Gas detectors Figure 13: Gas detector measurements for the LNG reference test-1 (test 1) The diagram below shows the layout of the radiometers. R1D1 R2D1 R1D2 R2D2 D1 D2 Radiometers Dataloggers N LNG Storage Tank LNG Discharge Line R1D1 R1D2 5m D1 The Test Pit 2x2m D2 5m 5m 5m R2D1 R2D2 Figure 14: Layout of Equipment for the measuring of radiant heat flux PC RPI Ref: P February 2014

21 Figure 16 below shows the radiant heat flux measurements and it is apparent that the highest measurement was recorded by radiometer R2D1: this was due to changing wind conditions (speed and direction), as the wind-speed in Test 1 (the reference test) was between 0 to 0.4 m/s towards the south. Although wind speed was relatively low, it was demonstrated that the wind does have a bearing on the radiant heat flux received by the radiometers: it causes the flame to change position and thus affects the distance between the flame and the radiometer. This subsequently resulted in a lower surface emissive power and view factor, which, in turn, had an effect on the radiant heat flux received by the radiometer. As can be seen in the photograph below, the flame was almost vertical, due to the low wind speed. The height of the flame was less than 10m. Figure 15: Radiant heat being measured during the LNG pool fire reference test (test 1) RPI Ref: P February 2014

22 Radiant Heat Flux (kw/m 2 ) Vapour and Fire Control Testing of FOAMGLAS PFS System (Gen 1) on LNG and LPG 9 R1 D1 R2 D1 R1 D2 R2 D Time (min) Figure 16: Radiant heat measurement for test 1 RPI Ref: P February 2014

23 5.2 Test 2: LNG test using FOAMGLAS PFS System (Gen 1) FOAMGLAS PFS System (Gen 1) was placed into the LNG pit, as demonstrated in Figure 17. A quantity (750 kg 1.5 m 3 ) of LNG liquid was discharged into the pit and gas concentration measured for 18 minutes, upon the release of the LNG. The LNG was then ignited and radiant heat flux was measured for 22 minutes, at a distance of 5m from the pit. The radiometers were then moved to a distance of 2m from the pit and radiant heat flux recorded for a further 16 minutes. The table below displays the timeline of the test: Tim (min) Action 0:00 Data logging began 02:00 Cooling pumps started 03:00 Gas detectors switched on 06:00 LNG valve opened, in order to discharge 1.5 m 3 of LNG liquid into the pit 14:00 LNG valve closed 24:00 Gas detectors moved to a safe place 26:00 LNG ignited 38:00 Radiometers moved 2m away from the test pit 54:00 The fire was extinguished using a dry chemical FOAMGLAS PFS System (Gen 1) was placed in the LNG pit, as seen in the photograph below: the first layer comprised of 9x8 blocks, while the second layer was about 10% of the bottom layer (5x2 blocks). Figure 17: FOAMGLAS PFS System (Gen 1) RPI Ref: P February 2014

24 The average values of the weather conditions were as follows: Wind Speed (m/s) Wind Direction Temperature ( o C) Humidity (%) 0 to 1 WNW to N In this test, six gas detectors were used (A, B, C, D, E and G), with their layout as in Figure 18 below. The gas concentration recorded varied between the detectors, in accordance with changes in wind direction. G A 4m 1m C 1m The Test Pit 2x2m 1m B 2m 1m E D Figure 18: Diagram showing the layout of gas detectors for LNG FOAMGLAS PFS System (Gen 1) test (standard configuration) Figure 19: LNG is released into the test pit and gas concentration is measured RPI Ref: P February 2014

25 Gas concentration (% LEL) Vapour and Fire Control Testing of FOAMGLAS PFS System (Gen 1) on LNG and LPG A B C D E G Gas detectors Figure 20: Measurements of the gas detectors for the LNG FOAMGLAS PFS System (Gen 1) Figure 21 shows the size of the LNG flame, which was considerably decreased, while Figure 22 highlights the variation in radiant heat flux over time. Figure 21: The LNG flame 15 minutes after ignition RPI Ref: P February 2014

26 Radiant Heat Flux (kw/m 2 ) Vapour and Fire Control Testing of FOAMGLAS PFS System (Gen 1) on LNG and LPG R1 D1 R2 D1 R1 D2 R2 D Radiometers moved to 2m Time (min) Figure 22: Heat flux measurements pertaining to the use of FOAMGLAS PFS System (Gen 1) Figure 23 shows the FOAMGLAS PFS System (Gen 1), after the fire was extinguished by dry chemical. Figure 23: shows the FOAMGLAS PFS System (Gen 1) after the fire was extinguished, using a dry chemical. RPI Ref: P February 2014

27 5.3 Test 3: LPG reference test The LPG tests were performed in a specially fabricated pan, with delivery system. The pan measured 1m x 1m x 0.6m depth and the quantity of LPG used in each test was about (37.5 kg 0.07m 3 ). Two radiometers (R1 D1 and R2 D1) were used to measure radiant heat flux, and these were positioned as in the diagram below. LPG was measured using five Crowcon Gasman gas detectors calibrated for propane. The locations of the gas detectors and radiometers were as shown in the diagrams below. E B C 1 x 1 Steel Pan A The detectors are at 0.2m distance from the pan and 0.7m height D Figure 24: Layout of the gas detectors (LPG reference test). R1 D1 N 2m 2m 1 x 1 Steel Pan R2 D1 Figure 25: Layout of the radiometers (LPG reference test) RPI Ref: P February 2014

28 A quantity of LPG (37.5 kg 0.07m 3 ) was discharged into the test pan, and gas concentration was measured for about 19 minutes. The LPG was then ignited and radiant heat flux was measured for about 8 minutes, at which point the LPG was consumed and the fire burnt out. The table below shows the timeline of the test Tim (min) Action 0:00 Data logging began 05:00 Cooling pumps started 05:00 Gas detectors switched on 09:00 LPG valve opened 24:00 LPG valve closed 28:00 Gas detectors moved to a safe place 31:00 LPG ignited 38:00 LPG consumed and fire self-extinguished Three LPG cylinders were used to form the LPG liquid pool, as seen in the photograph below. Figure 26: LPG discharged into the test pan Average weather conditions during the test were as follows: Wind Speed (m/s) Wind Direction Temperature ( o C) Humidity (%) 1 S to SSW RPI Ref: P February 2014

29 Gas concentration (% LEL) Vapour and Fire Control Testing of FOAMGLAS PFS System (Gen 1) on LNG and LPG The figure below shows the gas concentration measurements for the LPG reference test: A B C D E Gas detectors Figure 27: Gas concentration measurements for the LPG reference test Figure 28 displays the LPG flame from the LPG reference test. Figure 28: LPG flame (reference test) Figure 29 below shows the variation in radiant heat flux. Looking at the figure, it is apparent that the measurements from radiometer R1 D1 were higher than RPI Ref: P February 2014

30 Radiant Heat Flux (kw/m 2 ) Vapour and Fire Control Testing of FOAMGLAS PFS System (Gen 1) on LNG and LPG those from radiometer R2 D1. This was due to the wind blowing in a southerly direction, tilting the flame towards radiometer R1 D1. R1 D1 R2 D Time (min) Figure 29: The radiant heat measurement for LPG reference test. RPI Ref: P February 2014

31 5.4 Test 4: LPG Test using FOAMGLAS PFS System (Gen 1) The FOAMGLAS PFS System (Gen 1) was placed in the test pan, as seen in Figure 30, and 37.5kg of LPG liquid was discharged into the pan. Gas concentration was measured for 14 minutes, and a spot measurement of 40% LEL was undertaken. The LPG was then ignited and radiant heat was measured for 64 minutes. The gas detectors location was as shown in the diagram below and the radiometers were placed as in the LPG reference test. The table below shows the timeline of the test Tim (min) Action 0:00 Data logging began 04:00 Cooling pumps started 06:00 Gas detectors switched on 09:00 LPG valve opened 29:00 LPG valve closed 30:00 Gas detectors moved to a safe place 31:00 LPG ignited 95:00 Data logging ceased 134:00 Fire was extinguished, using water Figure 30: FOAMGLAS PFS System (Gen 1) placed in the LPG test pan RPI Ref: P February 2014

32 LPG vapour concentration (% LEL) Vapour and Fire Control Testing of FOAMGLAS PFS System (Gen 1) on LNG and LPG D N E C A 1 x 1 Steel Pan 1m B Detectors A, B and D were at 0.2m from the pan and at 0.7m height Detector E was on the ground Figure 31: Layout of gas detectors for FOAMGLAS PFS System (Gen 1) test Average weather conditions during the test were as follows: Wind Speed (m/s) Wind Direction Temperature ( o C) Humidity (%) 1 S to SSE A B C D E Gas detectors Figure 32: Gas concentration measurements for the LPG test (FOAMGLAS PFS System (Gen 1) RPI Ref: P February 2014

33 Radiant Heat Flux (kw/m 2 ) Vapour and Fire Control Testing of FOAMGLAS PFS System (Gen 1) on LNG and LPG Figure 33: LPG flame 6 minutes after ignition R1 D1 R2 D Time (min) Figure 34: Radiant heat flux measurements for the LPG test incorporating FOAMGLAS PFS System (Gen 1) RPI Ref: P February 2014

34 6 COMPARISON OF RESULTS WITH THE REFERENCE TESTS A total of four experiments were conducted, in order to determine the effectiveness of the passive firefighting material (FOAMGLAS PFS System (Gen 1)) in suppressing LNG and LPG fires and vapours. All the tests were conducted in the LNG firefighting training facility at Centro Jovellanos, Asturias, Spain. For each LNG test, the LNG was discharged into a concrete pit; thus, LNG vapour was produced, as a result of the phenomenon of vapourisation. With regards to the LPG test, the LPG was discharged into a steel pan specially manufactured for this purpose. The experiments also focused on the effects of the FOAMGLAS PFS System (Gen 1) on LNG/LPG flame length; thus, all experiments were recorded, using digital cameras. The changes in LNG/LPG radiant heat were the main focus of these experiments. Therefore, radiant heat flux was measured at various distances from the test pit/pan. In order to demonstrate the effectiveness of the FOAMGLAS PFS System (Gen 1), the data collected from the tests were compared to the baseline data from the reference tests. This section outlines the comparison of the tests conducted with and without the use of FOAMGLAS PFS System (Gen 1). 6.1 Test 2:LNG test incorporating the FOAMGLAS PFS System (Gen 1) In tests 1 and 2, the LNG was continuously released from the LNG storage tank into the test pit, through an insulated discharge line, until there was an approximate 750 kg 1.5 m 3 pool of fuel in the test pit. Initially, the LNG was in the form of a vapour, until the discharge line was completely cooled down. The gas detectors in test 2 were closer to the test pit than in test 1, as shown in the previous section. Vapour produced from the evaporation of the fuel pool dispersed downwind, and vapour cloud and cloud concentrations were measured at various points, using gas detectors. The unit of measurement was % LEL (note that 100% LEL is equal to a 5% v/v methane concentration in the air). An analysis was conducted, in order to study the effect of the FOAMGLAS PFS System (Gen 1) on the reduction of radiant heat pertaining to LNG fires. The data collected from the four radiometers in tests 1 and 2 were averaged and compared, as shown in Figure 35. This figure shows how, during the first few seconds of the ignition of the LNG in test 2, radiant heat flux was at almost the same level as in the reference test: this was due to the fact that the LNG vapours that were already in the pit burned during this period. However, after 5 minutes, the flame visible above the pit became stable, remaining not much higher than 4-5 feet above the pit. Thus, heat radiation came only from this part of the fire. After 5 minutes, radiant heat flux for test 2 became stable, at around 1 kw/m 2 ; this infers an approximate radiant heat reduction of 80%, when compared to average radiant heat flux in the reference test. RPI Ref: P February 2014

35 Radiant heat flux (kw/m 2 ) Vapour and Fire Control Testing of FOAMGLAS PFS System (Gen 1) on LNG and LPG LNG reference test FOAMGLAS PFS System (Gen 1) Time (min) Figure 35: Comparison of radiant heat flux measurements in test 2 and test 1 The figure below shows a comparison of the height of the flame in tests 1 and 2. Figure 36: Comparison of LNG flame size in test 2 and test 1 During the reference test experiment, average wind speed was considerably low (almost 0 m/s), while, in test 2, average wind speed was between 0 and 1 m/s, in a northerly direction. Figure 37 shows the comparison of the results of test 1 (the reference test) and test 2 (using the FOAMGLAS PFS System (Gen 1)). The high concentration reading in test 2 was largely due to low wind speed and the short distance between the test pit and the gas detectors. In test 1, the LNG steadily spread out from the test pit in all directions, while, in test 2, the LNG largely travelled towards detector B. This explains the high measurements recorded by detector B in test 2. RPI Ref: P February 2014

36 Gas concentration (% LEL) Vapour and Fire Control Testing of FOAMGLAS PFS System (Gen 1) on LNG and LPG Despite the high LNG measurements recorded by the gas detectors in test 2, it was observed that the size of the vapour cloud, when using the FOAMGLAS PFS System (Gen 1), was considerably less than the size of the vapour cloud in the reference test. This may have been due to the fact that the spaces between the FOAMGLAS PFS System (Gen 1) units were the only avenue by which the LNG vapour could disperse into the atmosphere. In the reference test, vapours were generated and then moved upwards, over the entire surface area of the LNG pool. The FOAMGLAS PFS System (Gen 1) on the surface of the LNG pool in test 2 blocked the flow of the vapour, with such vapours flowing through the spaces between units of the FOAMGLAS PFS System (Gen 1). Without Foamglas With Foamglass A B C D E G Gas detectors Figure 37: Comparison of gas concentration measurements in test 2 and test 1 RPI Ref: P February 2014

37 Radiant heat flux (kw/m 2 ) Vapour and Fire Control Testing of FOAMGLAS PFS System (Gen 1) on LNG and LPG 6.2 Test 4: LPG test incorporating the FOAMGLAS PFS System (Gen 1) During the LPG experiments, it was observed that the fire control time was significantly short; within a few seconds of ignition, the height of the flame was reduced by approximately 90%. The flame above the test pan became stable within a few seconds and remained not much higher than two feet. In addition, the LPG flame did not cover the whole surface area of the test pan. The fire lasted for about two hours, and was then extinguished using water. In terms of the unmitigated LPG, the fire lasted for only seven minutes for the same quantity of LPG and under similar weather conditions. It should be noted that the lower portion of the polyethylene bags were not completely burned: only the exposed outer portion of the units of the FOAMGLAS PFS System (Gen 1) were discoloured and damaged. Figure 38 below shows the significant reduction in radiant heat flux: there was an approximate 90% reduction in this, when compared to the LPG reference test Reference Test FOAMGLAS PFS System (Gen 1) Time (min) Figure 38: Comparison of radiant heat flux measurements in test 3 and test 4 Figure 39: Comparison of LPG flame size in test 3 and test 4 RPI Ref: P February 2014

38 Gas concentration (% LEL) Vapour and Fire Control Testing of FOAMGLAS PFS System (Gen 1) on LNG and LPG Gas detectors were placed so that the LPG vapours could be measured. LPG vapour, at atmospheric pressure and temperature, is 1.5 to 2 times heavier than air and would normally settle at ground level. The LPG vapours were not visible, however, and thus vapour cloud size could not be evaluated. In test 6, the gas detectors were placed relatively close to the LPG test pan and wind speed was slightly higher than in the reference test. LPG vapour concentration data were utilised, in order to assess the effectiveness of FOAMGLAS PFS System (Gen 1) in suppressing the dispersion of LPG vapour. The comparison was undertaken by comparing the maximum value of each gas detector, and Figure 40 below outlines the comparison of the vapour concentration measurements of test 6 and the LPG reference test. Without Foamglas With Foamglas A B C D E Gas detectors Figure 40: Comparison of gas concentration measurements in test 3 and test 4 RPI Ref: P February 2014

39 7 SUMMARY AND RECOMMENDATIONS Four tests were conducted, in order to investigate the effectiveness of LNG and LPG dispersion and pool fire suppression methodology incorporating the use of FOAMGLAS PFS System (Gen 1). The FOAMGLAS PFS System (Gen 1) was placed in the LNG test pit or the LPG test pan, into which a certain quantity of LNG or LPG was discharged. The vapour concentration of LNG/LPG was measured for a period of time and then the fuel was ignited, in order to allow for the measurement of radiant heat flux. The principal instruments used in the series of tests outlined in this report were designed to measure radiant heat flux from the LNG and LPG pool fires and the gas concentration of LNG/LPG. In addition, photographic and meteorological data were obtained. Radiant heat data collected from the tests pertaining to the FOAMGLAS PFS System (Gen 1) were compared to the baseline data collected from the reference tests, for the purpose of evaluating the performance of the FOAMGLAS PFS System (Gen 1) in suppressing pool fires (taking into account the various weather conditions for each individual test). The FOAMGLAS PFS System (Gen 1) significantly reduced radiant heat pertaining to both the LNG and LPG pool fires. This assertion is based on the experimental results, including the pool fire measurements and videos and photographs. In addition, it was observed that the FOAMGLAS PFS System (Gen 1) provided a coverage layer that helped to stabilise the fire, with no fluctuation; it also significantly reduced flame size. Furthermore, the FOAMGLAS PFS System (Gen 1) brought the LPG/LNG fires under control within a few seconds of ignition. The table below features a summary of the tests and the comments noted during the experiments. With a view to providing a better understanding of future LNG and LPG fire and vapour concentration suppression research and delineating the important information to be gathered, the following recommendations are suggested: 1. Hydrocarbon (H/C) IR cameras may be used to observe the actual dispersion of invisible vapours. LPG vapour is not visible and the hydrocarbon camera is able to capture the propane or butane cloud, enabling the observation of the movement of the cloud and the measuring of cloud size. 2. The use of point gas detectors to obtain a continuous gas concentration profile, rather than peak measurements. 3. Radiometers can be moved as close as possible to the fire while cooling is provided. RPI Ref: P February 2014

40 Report Ref. Test Date Test Test Test Test Material Used Reference Test FOAMGLAS PFS System (Gen 1) Reference Test FOAMGLAS PFS System (Gen 1) Fuel Vapour Measurement (min) Radiant Heat Measurement (min) Radiant Heat Reduction (%) Compared to Ref Test LNG LNG % LPG LPG % Comments The average radiant heat flux recorded by all radiometers was approximately 5 kw/m 2 The LNG flame gradually decreased, and only became steady after a period of more than 4 minutes. Consequently, average radiant heat flux decreased from 5 kw/m 2 to approximately 1 kw/m 2. The average radiant heat flux recorded by all radiometers was approximately 14 kw/m 2 The LPG flame above the test pan became stable within a few seconds upon ignition. The flame was considerably reduced to about 2 feet. The flame did not cover the whole surface area of the test pan. The fire lasted for about two hours, and was then extinguished using water The average radiant heat was 1.2 kw/m 2. RPI Ref: P February 2014