Fact File 57 Fire Detection & Alarms

Fact File 57 Fire Detection & Alarms Report on tests conducted to demonstrate the effectiveness of visual alarm devices (VAD) installed in different conditions

Report on tests conducted to demonstrate the effectiveness of visual alarm devices (VAD) installed in different conditions INTRODUCTION... 3 PROCEDURE... 4 RECORDED RESULTS... 5 DEMONSTRATION 1... 5 DEMONSTRATION 2... 5 DEMONSTRATION 3... 6 DEMONSTRATION 4... 6 DEMONSTRATION 5... 7 AVERAGE OF RESULTS... 8 AVERAGE OF 5 DEMONSTRATIONS... 8 SUBJECTS IN DIRECT VIEW AVERAGE OF 5 DEMONSTRATIONS... 8 SUBJECTS IN INDIRECT VIEW GRAPHICAL REPRESENTATIONS... 9 FIGURE 1...10 FIGURE 2...10 FIGURE 3...11 FIGURE 4...11 FIGURE 5...12 FIGURE 6...12 FIGURE 7...13 FIGURE 8...13 CONCLUSIONS...14 TABLE 1...15 TABLE 2...16 ANNEX A - SUPPLEMENTARY FIELD DEMONSTRATION...18 2 of 25

INTRODUCTION The report contained in this Fact File has been prepared jointly by the Building Research Establishment (BRE) and the Fire Industry Association as part of drafting a code of practice for visual alarm devices used for fire warning. This information should be used to supplement the Code of Practice for visual alarm devices used for fire warning; CoP 0001. This report describes the tests that were conducted to assess the effectiveness of VADs in warning persons of the presence of a danger and tabulates the results obtained. To represent a broad range of application situations, VADs were used of varying output, giving white or red and using either a xenon flashing source or LED pulsing source (both 1Hz). Although the tests were conducted in order to assist in the preparation of guidelines for the application of VADs approved to the new European standard, EN 54-23, it was only possible to use currently available non-approved VADs. This does not imply, however, that the devices used could not be approved in due course. Different room sizes with natural and/or artificial background ing levels were also used. In order to facilitate analysis of the results and reach conclusions as to the effectiveness of different illuminance, the report also gives graphical representations of the scores obtained. The process also involved a case study to investigate the effectiveness of VADs in an actual field application. For this additional demonstrations of factors affecting the effectiveness of VAD were conducted in a working installation, at a large hotel in the City of London. The results of this investigation are covered in Annex A to this report. This report was compiled in 2009 by the BRE/FIA Joint Task Group tasked to draft LPS1652 Code of Practice for Visual Alarm Devices used for fire warning following tests carried out earlier that year. 3 of 25

PROCEDURE A 'score card' method was used. Eight to 10 participants, aged between 30 and 70 were spread approximately equally around the perimeter of the room. For each test, each participant scored the device for effectiveness as follows: 9 highly effective, 7 moderately effective, 5 acceptably effective, 3 not really effective or 1 for absolutely ineffective. The effectiveness of each VAD, in each situation, was scored in both direct view and in indirect view. Direct view was taken as the VAD being in the field of view of the participant (which was not necessarily in direct line depending upon his position in the room). In the indirect view, the participants faced the wall of the room. The tables below show the raw scores and the score for each test. A summary table is also given. Two series of tests were conducted, the first at BRE on 9 February 2009, the second at Gent in Leicester on 2 March. During the first series of tests, two rooms were used to give various conditions of room size and background ing levels. In each test, ceiling mounted or wall mounted VADs were turned on individually. During the second series of tests, the size of the room dictated that two VADs be turned simultaneously at opposite ends of the room (see Figure 1 for room arrangement). The ambient level measurements were taken in the middle of the room and the of two lux meters were used for each reading. The level of the room was calculated by taking the of six readings - one from each of the four walls, the ceiling and the floor. 4 of 25

RECORDED RESULTS Demonstration 1 Location: BRE, Building 24, Room 108: 2.4 m (H) x 3.5 m x 4 m, (14.0 m 2, 33.6 m 3 ) ambient level: 400 Lux Room details: 4 matt creamy white walls, white ceiling tiles, brown carpeted floor Conditions: 4 Florescent tubes on (each 58W white ) Device/s & output level Colour Location C = ceiling W= wall Effectiveness of warning in direct view of VAD Score for direct view Effectiveness of warning in indirect view of VAD Score for indirect view score LED 1cd Red C 7x3, 2x5 3.44 5x1, 4x3 1.89 2.67 Xn 1cd Red C 1x3, 8x1 1.55 9x1 1.00 1.11 Xn 3cd White C 2x3, 7x5 4.55 4x3, 1x1, 4x5 3.67 4.11 LED 4.5cd Red C 9x5 5.00 2x1, 6x3, 1x5 2.78 3.89 Xn 15cd White C 4x7, 5x9 8.11 2x5, 4x7, 3x9 7.22 7.67 Xn 15cd White W 1x7, 9x9 8.80 3x5, 5x7, 2x9 6.80 7.80 LED 3cd Red W 2x5, 8x7 5.20 3x1, 3x3, 4x5 3.20 4.90 LED 2cd Red W 8x5, 1x7 5.22 5x1, 4x3, 1x5 2.20 3.71 Demonstration 2 Location: BRE, Building 24, Room 108: 2.4 m (H) x 3.5 m x 4 m, (14.0 m 2, 33.6 m 3 ) ambient level: 7.5 Lux Room details: 4 matt creamy white walls, white ceiling tiles, brown carpeted floor Conditions: Natural (1 window) Device/s & output level Colour Location C = ceiling W= wall Effectiveness of warning in direct view of VAD Score for direct view Effectiveness of warning in indirect view of VAD Score for indirect view score LED 1cd Red C 3x5, 7x7 6.40 6x3, 3x5, 1x7 4.00 5.20 Xn 1cd Red C 8x5, 2x3 4.60 8x3, 1x5, 2x1 3.10 3.71 Xn 3cd White C 8x7, 2x9 7.40 5x5, 4x7, 1x9 6.20 6.80 LED 4.5cd Red C 5x7, 5x9 8.00 4x5, 5x7, 1x9 6.40 7.20 Xn 15cd White C 10x9 9.00 10x9 9.00 9.00 Xn 15cd White W 10x9 9.00 10x9 9.00 9.00 LED 3cd Red W 6x7, 4x9 7.80 5x5, 5x7 6.00 6.90 LED 2cd Red W 5x5, 5x7 6.00 6x3, 4x5 3.80 4.90 5 of 25

Demonstration 3 Location: BRE, Building 1, Room 6: 3.75 m (H) x 5 m x 8 m, (40 m 2, 150 m 3 ) ambient level: 150 Lux Room details: 4 walls (2/3 grey, 1/3 white), white painted ceiling, blue carpeted floor Conditions: Lights on 8 Circular florescent tubes Device/s & output level Colour Location C=ceiling W= wall Effectiveness of warning in direct view of VAD Score for direct view Effectiveness of warning in indirect view of VAD Score for indirect view score LED 1cd Red C 9x3, 1x5 3.2 10x1 1 2.10 Xn 1cd Red C 2x1, 8x3 2.6 10x1 1 1.80 Xn 3cd White C 5x3, 5x5 4.0 10x1 1 2.50 LED 4.5cd Red C 8x5, 2x7 5.4 10x1 1 3.20 Xn 15cd White C 4x5, 6x7 6.2 7x3, 3x5 3.6 4.90 Xn 15cd White W 5x7, 5x9 8.0 5x5, 5x7 6.0 7.00 LED 3cd Red W 6x5, 4x7 5.8 5x1, 5x3 2.0 3.90 LED 2cd Red W 5x5, 5x7 6.0 10x1 1 3.50 Demonstration 4 Location: BRE, Building 1, Room 6: 3.75 m (H) x 5 m x 8 m, (40 m 2, 150 m 3 ) Room details: 4 walls (2/3 grey, 1/3 white), white painted ceiling and blue carpeted floor Conditions: Natural from 2 large south facing windows ambient level: 92 Lux Device/s & output level Colour Location C = ceiling W= wall Effectiveness of warning in direct view of VAD Score for direct view Effectiveness of warning in indirect view of VAD Score for indirect view score LED 1cd Red C 9x5, 1x7 5.2 3x3, 7x1 1.6 3.40 Xn 1cd Red C 9x3, 1x5 3.2 10x1 1.0 2.10 Xn 3cd White C 9x5, 1x7 5.2 3x1, 7x3 2.4 3.80 LED 4.5cd Red C 4x5, 6x7 6.2 4x1, 4x3, 2x5 2.6 4.40 Xn 15cd White C 5x7, 5x9 8.0 2x3, 2x5, 4x7, 1x9 5.3 6.94 Xn 15cd White W 1x7, 9x9 8.8 1x3, 4x5, 4x7, 1x9 6.0 7.40 LED 3cd Red W 3x5, 6x7, 1x9 5.1 4x1, 5x3, 1x5 2.4 3.75 LED 2cd Red W 4x5, 6x7 4.2 8x1, 2x3 1.4 280 6 of 25

Demonstration 5 Location: Gent, Leicester, Conference 1+2: 15.4 m x 7.1 m x 3 m (H), (109.3 m 2, 328 m 3 ) Room details: 4 walls (2/3 cream), white ceiling tiles and grey carpeted floor Conditions: Artificial ing from 12 x 4 (36 W) fluorescent bars ambient level: 222 Lux Device/s & output level Colour Location C = ceiling W= wall Effectiveness of warning in direct view of VAD Score for direct view Effectiveness of warning in indirect view of VAD Score for indirect view score LED 1cd Red C 9x5, 1x7 5.2 10x1 1.0 3.10 Xenon 1cd Red C 8x3, 2x5 3.4 6x1, 4x3 1.8 2.60 Xenon 3cd White C 2x3, 6x5, 2x7 5.0 3x1, 7x3 2.4 3.70 LED 4.5cd Red C 2x3, 2x5, 6x7 5.8 1x1, 6x3, 3x5 3.4 4.60 Xenon 15cd White C 5x5, 5x7 6.0 3x3, 6x5, 1x7 4.6 5.30 Xenon 15cd White W 2x7, 8x9 8.6 1x3, 6x5, 3x7 5.4 7.00 LED 3cd Red W 1x3, 7x5, 2x7 5.2 4x1, 6x3 2.2 3.70 LED 2cd Red W 3x3, 7x5 4.4 10x1 1.0 2.70 Additional tests with higher output Xenon flashing VADs Xenon 75cd White C 5x7, 3x9 7.75 5x5, 3x7 5.75 6.75 Xenon 30cd White W 3x7, 7x9 8.4 4x5, 6x7 6.2 7.30 Xenon 75cd White W 10x9 9.0 5x7, 5x9 8.0 8.50 Xenon 110cd White W 10x9 9.0 10x9 9.0 9.00 7 of 25

AVERAGE OF RESULTS The s for all device types over all demonstrations are shown in the two tables below, one for VADs in the field of vision of the subjects and the other for indirect illumination where subjects were facing walls or other illuminated surfaces: of 5 demonstrations - Subjects in direct view (VAD illuminating the field of vision) Device/s & output level Colour Location C = ceiling W= wall Demo 1 Demo 2 Demo 3 Demo 4 Demo 5 score LED 1cd Red C 3.44 6.40 3.2 5.2 5.2 4.7 Xenon 1cd Red C 1.55 4.60 2.6 3.2 3.4 3.1 Xenon 3cd White C 4.55 7.40 4.0 5.2 5.0 5.2 LED 4.5cd Red C 5.00 8.00 5.4 6.2 5.8 6.1 Xenon 15cd White C 8.11 9.00 6.2 8.0 6.0 7.5 Xenon 15cd White W 8.80 9.00 8.0 8.8 8.6 8.6 LED 3cd Red W 5.20 7.80 5.8 5.1 5.2 5.8 LED 2cd Red W 5.22 6.00 6.0 4.2 4.4 5.2 Additional tests with higher output Xenon flashing VADs Xenon 75cd White C 6.75 6.75 Xenon 30cd White W 7.30 7.30 Xenon 75cd White W 8.50 8.50 Xenon 110cd White W 9.00 9.00 of 5 demonstrations - Subjects in indirect view (VAD illuminating facing walls, floor) Device/s & output level Colour Location C = ceiling W= wall Demo 1 Demo 2 Demo 3 Demo 4 Demo 5 score LED 1cd Red C 1.89 4.00 1 3.40 1.0 1.9 Xenon 1cd Red C 1.00 3.10 1 2.10 1.8 1.8 Xenon 3cd White C 3.67 6.20 1 3.80 2.4 3.1 LED 4.5cd Red C 2.78 6.40 1 4.40 3.4 3.2 Xenon 15cd White C 7.22 9.00 3.6 6.94 4.6 5.9 Xenon 15cd White W 6.80 9.00 6.0 7.40 5.4 6.6 LED 3cd Red W 3.20 6.00 2.0 4.50 2.2 3.2 LED 2cd Red W 2.20 3.80 1 3.80 1.0 1.9 Additional tests with higher output Xenon flashing VADs Xenon 75cd White C 6.75 6.75 Xenon 30cd White W 7.30 7.30 Xenon 75cd White W 8.50 8.50 Xenon 110cd White W 9.00 9.00 8 of 25

GRAPHICAL REPRESENTATION The graphs in Figures 1 to Figure 6 show the effectiveness obtained for each type of VAD (ceiling or wall mounted), in both direct and indirect view of the subjects. These have been derived from the demonstration results. Another set of graphs summarising the results against background ambient are also shown. In these graphs, the X axis represent the scores given and the Y axis represents the illuminance given by the VAD at the furthest point from it in the test room. This illuminance is calculated using the claimed effective output in candela provided by the manufacturer. No VADs tested were to EN 54-23 (the standard has not yet been released), which makes all the readings very much worst case. In each case the effective candela is divided by the square of the maximum distance from the VAD in each room, to produce the illuminance, assuming an even distribution of output. The following illuminance distances have been calculated: For ceiling mounted VADs Room 1 (BRE - Building 24, Room 108) 3.58m Room 2 (BRE - Building 1, Room 6) 6.03m Room 3 (Gent - Conference rooms 1+2) 6.04m For wall mounted VADs Room 1 (BRE - Building 24, Room 108) 4.55m Room 2 (BRE - Building 1, Room 6) 8.65m Room 3 (Gent - Conference rooms 1+2) 8.73m Note: Room 3 (Gent) was split into two equal-size rooms. The graphs in Figure 7 to Figure 14 show what illuminance is required with different levels of ambient to achieve the acceptable effective level, ie an score of 5. Figures 9 and 10, for the ceiling mounted VAD, and Figures 13 and 14, for the wall mounted VAD, give an indication of the effect of ambient on the effectiveness of the visual warning. For these graphs, the trend curve equation is given in order to facilitate extrapolation to other X values. 9 of 25

Figure 1 - Illuminance required with different background (ceiling mounted VAD - direct) Figure 2 - Illuminance required with different background (ceiling mounted VAD - indirect) 10 of 25

Figure 3 - Illuminance required for acceptable effectiveness with ambient (ceiling mounted VAD - direct) Figure 4 - Illuminance required for acceptable effectiveness with ambient (ceiling mounted VAD - indirect) 11 of 25

Figure 5 - Illuminance required with background (wall mounted VAD - direct) Figure 6 - Illuminance required with background (wall mounted VAD - indirect) 12 of 25

Figure 7 - Illuminance required for acceptable effectiveness with ambient (wall mounted VAD- direct) Figure 8 - Illuminance required for acceptable effectiveness with ambient (wall mounted VAD- indirect) 13 of 25

CONCLUSIONS, the test conducted gave sufficient data for use as a basis of establishing general rules in the application of VADs for warning of a fire situation. The data gathered on effectiveness in different situations can be used in the determination of factors that might be applied to modify the coverage distance of specific EN 54-23 approved VADs. For example: where reliance on direct line of sight is possible; where the ambient levels can be determined. The extent and coverage of the test programme was however, by its nature, limited. One of the conclusions must be, therefore, that further tests would be beneficial. For example, the following experimental parameters could be further investigated: a. behaviour of VADs in very high ambient levels b. using more samples with a greater range of output to increase statistical accuracy c. using a different cross-section of people, eg including persons with hearing difficulties and a d. wider age range e. using fully characterised VADS, preferably when EN 54-23 approved devices become available. According to the results of the tests carried out, the minimum required output from a VAD for effective warning varies significantly with the background ambient level. The degree of variation is not uniform throughout the range of ambient considered (0-1,000 lux). It also depends whether the VAD is wall or ceiling-mounted and whether indirect or direct viewing of the device is considered. These results can be used to establish the minimum required output from a VAD for effective warning in different ambient levels (see Table 1). An analysis of Table 1 indicates the following: a. In all situations, wall mounted VAD were found to be more effective across the range of ambient level considered and should be preferred for general applications. b. As ambient increases, the relative effectiveness of wall mounted VAD when viewed directly seemed to falls sharply in relation to the decrease in effectiveness when viewed indirectly. For example, above 800 lux ambient levels, VAD viewed indirectly seems to be as effective as or, even more effective than those viewed directly. This may be due to a number of factors such as experimental uncertainties (ambient was taken as an of measured values in six directions), subjective scoring and room characteristics. Although this test indicator can safely be used for ambient levels up to 600 lux, a more rigorous assessment of effectiveness will be required for the higher levels. 14 of 25

Table 1: Minimum required output from VAD for effective warning in different ambient levels Mounting location Wall Ceiling level (lux) Minimum required level (lux) (at application illuminance distance) Direct 0-100 0.13 0.015 100-200 0.18 0.03 200-400 0.31 0.10 400-600 0.55 0.37 600-800 0.98 1.35 800-1000 1.75 4.8 0-100 0.22 0.06 100-200 0.35 0.10 200-400 0.88 0.35 400-600 2.20 1.15 600-800 5.40 3.80 800-1000 13.6 12.6 NOTE Figures shown in italics have been obtained by extrapolating from the trend line on the appropriate graphs in section 4. 15 of 25

Using the trend line equation in Figures 3, 4, 7 and 8, acceptable VAD illumination for different background levels can be derived as shown in Table 2 below. Note: In figures where two curves are given, the equation giving the closest match (higher R2 value) has been used. Table 2: Acceptable VAD illumination for different background levels Wall mounted VAD Ceiling mounted VAD VAD o/p (lux) level (lux) VAD o/p (lux) level (lux) Direct Direct 0.05 2) 283 0.05 2) 90 0.1 43 392 0.1 2) 222 0.2 254 502 0.2 77 354 0.4 1) 467 611 0.4 1) 229 486 0.6 591 674 0.6 317 563 0.8 679 720 0.8 330 618 1) 0.4 lux (Lumens/m 2 ) corresponds to the output requirements of EN 54-23. 2) Insufficient VAD output to obtain effective warning at ambient levels other than darkness. In relation to the VAD illumination required by EN 54-23, ie 0.4 lux, Table 2 shows that, as an approximation, the following factors can be applied to the illumination required from VADs: a. When indirect illumination is required, the illumination from wall mounted VAD: 1) can be reduced by a factor of 4 (to 0.1 lux) for ambient levels below 40 lux 2) can be reduced by a factor of 2 (to 0.2 lux) for ambient levels between 40 lux and 3) 250 lux 4) should be unchanged (at 0.4 lux) for ambient levels between 250 lux and 470 lux 5) should be increased by a factor of 1.5 (to 0.6 lux) for ambient levels between 470 lux 6) and 590 lux 7) should be increased by a factor of 2 (to 0.8 lux) for ambient levels between 590 lux and 8) 680 lux. b. When direct illumination is possible, the illumination from wall mounted VAD: 1) can be reduced by a factor of 8 (to 0.05 lux) for ambient levels below 280 lux 2) can be reduced by a factor of 4 (to 0.1 lux) for ambient levels between 280 lux and 390 lux 3) can be reduced by a factor of 2 (to 0.2 lux) for ambient levels between 390 lux and 500 lux 4) should be unchanged (at 0.4 lux) for ambient levels between 500 lux and 610 lux 5) should be increased by a factor of 1.5 (to 0.6 lux) for ambient levels between 610 lux and 670 lux 6) should be increased by a factor of 2 (to 0.8 lux) for ambient levels between 670 lux and 720 lux. 16 of 25

c. When indirect illumination is required, the illumination from ceiling mounted VAD: 1) can be reduced by a factor of 2 (to 0.2 lux) for ambient levels below 80 lux 2) should be unchanged (at 0.4 lux) for ambient levels between 80 lux and 230 lux 3) should be increased by a factor of 1.5 (to 0.6 lux) for ambient levels between 230 lux and 310 lux 4) should be increased by a factor of 2 (to 0.8 lux) for ambient levels above 310 lux and 330 lux. d) When direct illumination is possible, the illumination from ceiling mounted VAD: 1) can be reduced by a factor of 8 (to 0.05 lux) for ambient levels below 90 lux 2) can be reduced by a factor of 4 (to 0.1 lux) for ambient levels between 90 lux and 220 lux 3) can be reduced by a factor of 2 (to 0.2 lux) for ambient levels between 220 lux and 350 lux 4) should be unchanged (at 0.4 lux) for ambient levels between 350 lux and 490 lux 5) should be increased by a factor of 1.5 (to 0.6 lux) for ambient levels between 490 lux and 560 lux 6) should be increased by a factor of 2 (to 0.8 lux) for ambient levels between 560 lux and 620 lux. As effective illumination is the inverse square of the distance, reductions in the VAD illumination correspond to increases in its coverage distance as follows: a reduction in illumination by 8 corresponds to an increase in coverage distance by 2.8 a reduction in illumination by 4 corresponds to an increase in coverage distance by 2.0 a reduction in illumination by 2 corresponds to an increase in coverage distance by 1.4 Similarly, increases in the VAD illumination correspond to reductions in its coverage distance as follows: an increase in illumination by 1.5 corresponds to a reduction in coverage distance by 1.2 an increase in illumination by 2 corresponds to a reduction in coverage distance by 1.4 17 of 25

ANNEX A - SUPPLEMENTARY FIELD DEMONSTRATION Introduction Additional demonstrations of factors affecting the effectiveness of VAD were conducted in a working installation at a large hotel in the City of London. The following areas which had been fitted with VADs were assessed: bedrooms corridors meeting rooms plant rooms In each case, the assessment was conducted in the same way as described in the main report. For each demonstration, the layout of the room, the ambient parameters and the summary results of the assessment is given. Where applicable, additional comments are included to high specific conditions found during the demonstration or relevant to any lessons learned in the application of VADs. The installation used ceiling-mounted VADs with a nominal output of 4.5 cd device and wall-mounted VADs with a nominal output of 3.5 cd. Demonstrations Executive bedroom 701 Executive bedroom 701 is a large bedroom on the top floor of the hotel. It comprises living and bedroom areas, as well as a separate bathroom and a study alcove. Of the two smoke detectors installed, only the one above the bed had a VAD facility (see Fig A.2). Full size glass windows occupied the two external walls affording the room a potential high level of ambient. A very large entertainment system screen provides a significant reflective surface on one side of the living area (see Fig A.1). The height in the living areas was measured as 2.5m. Fig A.1 Picture of bedroom 701 showing the large entertainment screen 18 of 25

Fig A.2 Layout of Executive bedroom 701 RESULTS OBTAINED IN BEDROOM 701 Direct Direct 76.66 6.0 2.4 4.2 98.33 6.0 2.4 4.2 Comments 1. A single VAD above the bed was insufficient to ensure adequate visual warning in recessed areas of the room. Two VADs would have provided a more adequate coverage. VADs would have been more effective if centrally placed. 2. There was no effective visual warning in the bathroom. It would be necessary to fit an additional VAD in separated bathrooms where people with hearing difficulties are expected to be on their own. 19 of 25

FLOOR 7 CORRIDOR A demonstration was conducted in a 21.8 m section of corridor linking bedrooms on floor 7. The corridor section, which is separated by fire doors, uses one smoke sensor with VAD sited in proximity (20% of the length a corridor section) of one of the doors (see Fig A 3). The ambient level in the corridor was measured at below 5 lux. Fig A3 Layout of corridor in floor 7 RESULTS OBTAINED IN FLOOR 7 CORRIDOR Direct 4.6 lux 8.8 3.0 5.9 Direct Comments 1. In spite of the low level of ambient, the VAD arrangement in the corridor proved inadequate when warning depended on indirect illumination. 2. In corridor, specifically long ones and those separated in sections for smoke control purpose, it is recommended that VADs are installed such that they are in the direct field of view of people exiting bedrooms and walking towards one or other directions. This can be achieved by fitting a minimum of two VADs near each of the door leading to the next section of corridor. 3. In corridors with low ambient levels it is possible to significantly increase the nominal coverage distance of VADs while still achieving effective warning of occupants in direct view of the devices. 20 of 25

BEDROOM 707 Bedroom 707 is a medium size square bedroom on the top floor of the hotel. It comprises two areas - a bedroom with separate bathroom and the other a lounge. Each area was equipped with one smoke detector with a VAD facility (see Fig A4 and Fig A5). The living area had a 24 cm deep beam with the VAD mounting inside the recessed thus formed. In each area the VAD was off-centre relative to one of the room s axes. Tests were conducted in maximum ambient level with room s turned on and a lower level obtained by switching off the room s. The height in the living areas was measured as 2.5m. Fig A.4 Layout of lounge area Fig A.5 Layout of sleeping area RESULTS OBTAINED IN BEDROOM 707 - SLEEPING AREA Direct Direct 17.3 lux 8.09 4.72 6.41 30.66 lux 6.64 2.82 4.73 RESULTS OBTAINED IN BEDROOM 707 - LOUNGE AREA Direct Direct Comments 1. Having two VADs greatly improved the effectiveness of the warning in this dual usage room. As in room 701, additional warning in the bathroom needs to be considered. 2. In the lounge area, the position of a VAD behind a beam greatly reduced its effectiveness. 7.33 lux 7.73 3.91 5.82 14.17 lux 5.73 2.09 3.91 21 of 25

MEETING ROOM 6 A demonstration was conducted in an inner meeting room with no natural. The room was equipped with a single smoke detector with built-in VAD facility. The VAD was approximately central to the geometry of the room. Two of the walls exhibited large reflective surfaces - one being a large audio-visual monitor screen, the other a 'smoky' glass panel, also incorporating the entrance door to the room (see Fig A6 and Fig A7). Two different levels of ambient were obtained by using dimmers provided. The height in the living areas was measured as 2.5m. Fig A.6 Picture of meeting room 6 showing the VAD and reflective surfaces Fig A7 Layout of corridor in floor 7 RESULTS OBTAINED IN MEETING ROOM 6 Direct Direct 55 lux 6.45 2.27 4.36 1.2 9.0 4.45 6.72 Comments 1. As a meeting room where it is expected that occupants will be seating mostly around the table, the use of a single VAD in the middle of the room seemed to provide a sufficiently effective warning. 2. Either an increase in VAD output, or the use of two VADs with the same output, should be considered if effective warning by indirect illumination is deemed necessary. 22 of 25

PLANT ROOM - LEVEL 1 A demonstration was conducted in a plant room representing a narrow space with a bright artificial (neon) (see Fig A8 and Fig A9). The VAD was of the wall-mounted type but, in this case, it was mounted to the ceiling. Numerous pipes were hanging off the ceiling, affecting direct vision to the VAD from a number of observer positions. Smaller areas lead off the main space where the VAD was fitted. Fig A.8 Picture of plant room showing VAD and surrounding pipe work Fig A.9 Layout of plant room on Level 1 RESULTS OBTAINED IN PLANT ROOM - LEVEL 1 Direct Direct 280 lux 3.22 1.89 2.55 Comments 1. Although this was a small space, loss of effectiveness occurred where the observer's view was blocked by obstacles such as the pipework. 2. Partly due to high ambient level, the warning was ineffective in most observers' positions when having to rely on indirect illumination. 3. It could be expected there was no direct or indirect illumination from the VAD areas off the main protected space. 23 of 25

PLANT ROOM - GROUND FLOOR A demonstration was conducted in a plant room on the ground floor representing an (approximately) square space. The VAD was a wall mounted type sited just above electrical cabinets approximately in the centre of the wall which also had the entrance door to the room. The space was used as a pump room and characterised by heavy machinery and large vertical and horizontal pipes (see Figs A10, A11 and A12). There was a high level of background noise justifying the use of VAD as the main warning of a fire alarm. Fig A.10 Picture of ground floor plant room showing VAD and surrounding cabinets Fig A.11 Picture of ground floor plant room showing VAD and surrounding pipe work Fig A.12 Layout of plant room on ground floor RESULTS OBTAINED IN PLANT ROOM - GROUND FLOOR Direct 87.5 lux 5.66 1.89 3.77 Direct Comments 1. Generally, in this small space, the VAD was effective when viewed directly by observers. However, significant loss of effectiveness occurred where the observer's view was blocked by obstacles such as the pipework. 2. The warning was largely ineffective in most observer's positions when having to rely on indirect illumination. 3. Due to the particular use of the space, the installation would have benefited from additional VADs sited to avoid blind spots. 24 of 25

DISCLAIMER The information set out in this document is believed to be correct in the of information currently available but it is not guaranteed and neither the Fire Industry Association nor its officers can accept any responsibility in respect of the contents or any events arising from use of the information contained within this document. Tudor House, Kingsway Business Park, Oldfield Road, Hampton, Middlesex TW12 2HD Tel: +44 (0)20 3166 5002 www.fia.uk.com 25 of 25