Tests results analysis and summary for Heathrow Airport

Baggage Tunnel

1.     Introduction

1.1      General Situation

The baggage transfer tunnel at Heathrow is primarily used to convey traveller’s baggage between Terminal 4 and the other main terminal buildings (Terminals 1 to 3). The main tunnel basically consists of 2 sections:

1) Main luggage rail on which the baggage trolleys travel at speed.

2) Personnel walkway - Used for maintenance/service purposes.

At each end of the main tunnel, there are also 4 ‘Spiral sections’ that transport the baggage to and from ground level, these are also to be considered for fire protection.

The peculiarities of the baggage tunnel are: considerable amount of power, lighting, control and signal cables and optical fiber cables existed in the tunnel, and electrical facilities such as power distribution and control cabinets are installed at a defined interval in the tunnel; the baggage rail and walkway is partitioned with a safety nets, and the net wobbles under the effect of the tunnel airflows; the baggage trolleys travel at speed on the rail carrying baggage of all shapes and sizes; the baggage trolleys acceleration controllers possess overheating hazards; the tunnel is operating continuously with controlled airflow rate, and up to a maximum of 7.5m/s; and a very dusty environment when supplying air into and extracting air from the tunnel.

1.2      Analysis of fire hazards

1.2.1      A variety of power supply, electrical, and control cabinets are installed in the tunnel which are the probable sources of fire.  Of those electric fires under 10KV, more then 70% are caused by overloading and deteriorated electrical components in the power distribution and control cabinets.  A bigger problem is all the cables and optical fibers connecting the power distribution and control cabinets are probably the main route for the spreading fires.  If this kind of situation is not controlled in time, the devastation could rapidly spread to various terminals where the tunnel is connected to.  On top of that, the connection cables may have overheating or short-circuitry problems which cause and spread fires.

1.2.2      In the baggage tunnel, the baggage itself is one of the most probable sources for fires.  There are few potential hazards involve: Firstly, the electrical fire may halt the baggage trolleys and spread the fire using the baggage as a burning agent; Secondly, the baggage may contain flammable substances which self-ignite under certain specific conditions and spread in the tunnel; Lastly, the baggage trolley may be jammed at the acceleration controller which causes overheating and fire breakout.

1.2.3      The partition safety net in between the baggage rail and walkway which stretches throughout the tunnel is also a probable fire source and a likely spread path for fire.

1.3      Basic demand of fire detection in baggage tunnel

1.3.1      Detect fire in a shortest possible time and provide instant information on the zone and location of fire to control personnel, allowing them with ample time for firm decision making and to adopt proper fire prevention strategy.

1.3.2      When fire breakout, there must be provision of real-time information on fire spreading pattern, process, zonal position, flame conditions and so on, to offer an excellent reference to aid subsequent rescue and evacuation actions.  And to achieve that, images of acceptable clarity are required.

1.3.3      This new fire detection system should operate effectively under all kinds of airflow rate and direction, and it should not be affected by the dust particles in the tunnel which could lead to false alarm.

1.3.4      The detection system should have excellent sensitivity. Especially when there are high risks electrical facilities in the key protection areas, high sensitivity and rapid detection become a crucial requirement.

1.3.5      Other than detecting for smoke, ability to detect flame is also very important.  This flame detection becomes very critical to let rescue team gain control under extreme situations.

1.3.6      In addition to fire detection, the system should also provide necessary security surveillance features, such as detection of trolley stoppage, because unusual system malfunction is a good indication of other major problems which immediate attentions and rectifications are required.

3.     Summary and Suggestion

3.1      Test results synthesis

3.1.1   Test 1--- Smoke test carried out with Fans supplying air to the TBS tunnel and no baggage trolleys running

3.1.5  Test 5 --- Smoke test carried out with Fans extracting air from the TBS tunnel, baggage trolleys running and beacons operating
                                                                                                

3.2.1      Base on the test result, adopting the installation (as shown in Figure 1 and 2) with two opposite facing detectors seems able to achieve better detection performance despite the effects from all existing conditions in the baggage tunnel, e.g. lit or unlit condition, baggage trolleys running or not, and various airflow rate and directions.  However, the system proposed by InnoSys is to have only one detector for every 50m, the analysis for the two opposite facing detectors system (which requires two detectors for every 50m) was not included in this report.  Besides, the total system cost for the latter configuration is relatively much higher.

3.2.2      Summarizing all the above analysis, installation as shown in Figure 3 or 4 is the proposed configuration for the tunnel.  For this configuration, only one detector is required for every 50m, and without changing the default configuration setting, the achieved result combining the probability for two detectors, among all the tests done in test 2~test 6, only one of the 3gm tests in test 4 did not trigger the alarm.  And among all the respond time, 80% less than 40s, 90% less than 60s, only 10% longer than 60s.  As for Test 7 where the smoke was initiated underneath the baggage rails, only one test was performed, but the probability from VISFD 01 rose to 36% indicating the system had captured the smoke information for analysis and only had the smoke lasted longer before the alarm would be triggered.  As for Test 9, where 3 tests were conducted under unlit condition, VISFD 01 had given pre-alarm twice and one of the tests with probability rose to 37%, implying that smoke information was capture for analysis but the smoke had not lasted long enough.

3.2.3      Base on the recorded video of all the tests, through adjustment to some self-adaptive parameters, 100% alarm requirement to all the conducted tests under all variable conditions can be achieved.  And the average respond time can be improved to 10~20s, and reference can be made to the VCD video footages.  This adaptive feature has made the system very versatile and cost effective.

3.2.4      Under the effect of various airflow rates and directions, the smoke drifts rapidly against the circular sidewall, which would only start to disperse after quite some distance.  Therefore, even if the upwind VISFD did not capture and alarm, the downwind VISFD is able to capture the sufficient smoke images for analysis to ensure effective detection.  As for the very difficult test 7, where the 9gm pellet was placed at the underside of the baggage rail, the downwind VISFD 01 was still able to capture the smoke information, and in reality, the cascaded downwind detectors would be able to capture even more noticeable smoke to ensure no missed alarm.

3.2.5      The running of baggage trolleys causes turbulence and disturbance to the movement and appearance of the smoke, but no significant impact to the functional features of the detection.

3.2.6      Results from all the test seems to indicate that the upwind facing detector has a comparatively better detection performance than the downwind facing detector.

3.2.7      When the baggage tunnel was unlit, AlarmEye^® AE series VISFD would switch on the attached IR source.  From then on, the CCD would capture mainly the scattering and absorption of the IR radiations, thus the amount of smoke information is significantly reduced.  However, the cascaded downwind detectors could still capture sufficient smoke information for detection and alarm.  The tests done on 23^rd October where the detectors are installed in a opposite facing manner, has shown that it is relatively easier for smoke detection because additional smoke information such as obstruction, refraction and scattering images are also available.

3.2.8      In general, from all the tests conducted using 3gm and 9gm smoke pellet, their difference in term of the amount of smoke had little impact on the detection performance.  However, when the smoke was initiated on the baggage rail side of the safety nets and under unlit condition, a prolong smoke initiation (e.g. a 9gm pellet) would be required for effective detection.

3.2.9      AlarmEye^® AE series VISFD is able to perform dependable and highly sensitive detection on fire flame, which is a crucial complement and assurance to fire detection in tunnel.

3.2.10   Throughout the tests, AlarmEye^® AE series VISFD was immune from all the probable false alarm sources such as the flashing emergency lights, swinging signs and wobbling of the safety nets, and not a single false alarm was reported.

3.2   Suggestion of system design and installation

3.3.1      It is recommended to install one detector for every 50m distance as shown in Figure 3 or 4, for better cost effectiveness.  And the detection reliability can be achieved through adjustment on the system adaptive parameters.

3.3.2      It is also recommended that an additional 5W, 30ºIR source to be installed 50m away on the opposite end of the detector, to increase IR image information obtained by the CCD so that the detection sensitivity can be improved.

3.3.3      Recommend to include flame detection as one of the basic features for the tunnel video image fire detection system, so as to upgrade system’s overall safety performance and dependability.

3.3.4      From the aspect of smoke dispersion, the current large grid safety nets is favorable for dispersion.  And the movement and wobbling of these nets does not cause any false alarm to AlarmEye® AE series VISFD.  Therefore, unless it was the original plan to replace the safety nets and these nets is made of fire-proof material, it is not necessary to replace all the nets to cater for the installation of video image fire detection system. 

3.3.5      It can be considered to have the detector installed in the upwind orientation (against the airflow direction), to improve the detection sensitivity.  One downside to this is the detector window would be exposed to more dusts contamination, as a result more frequent window cleaning is to be arranged.