FRAME Calculation examples book. - FRAME Fire Risk Assessment ...

FRAME Calculation examples book.

F.R.A.M.E. 

Content. 

Introduction............................................................................................................... 4 

FRAME calculations for real fires. .................................................................................. 5 

Case study 1. Bekina Indurub Kluisbergen (BE) , 1986. ............................................... 6 

Case study 2. Dupont Plaza Hotel Fire Puerto Rico, December 31, 1986.......................... 7 

Case study 3. Hotel International at Zurich (CH) , February 14, 1988 ............................. 8 

Case study 4. Hotel fire in Kristiansand (NO) , September 5 th , 1986 ............................... 9 

Case study 5. Honda Spare parts centre, Ghent (BE), October 29 th 1987. ..................... 10 

Case study 6. Fire at “Bourse de Bruxelles” (BE) on November 30, 1990....................... 11 

Case study 7. Furniture factory fire Kortessem (BE) , October 1991 ............................. 12 

Case study 8. Imperial Foods Plant Fire, Hamlet (NC), September 1991........................ 12 

Case study 9. Department store fire Luton Bedfordshire (UK) January 1996. ................. 14 

Case study 10. General warehouse fire New Orleans, Louisiana March 21, 1996. ........... 15 

Case study 11. Dusseldorf Airport (DE) fire April 1996................................................ 18 

Case study 12. The Sight and Sound Theater (US) fire on February 28, 1997. .............. 21 

Case study 13. Unifast Braine-le-Comte (BE) Fire January 11 th , 1999. ......................... 23 

Case study 14. Hartford (CT) nursing home fire February 26 , 2003. ............................ 25 

Case study 15. A clue for arson : Office in Bristol (UK), June 1 st , 1989........................ 26 

Case study 16. A clue for arson : Europneu Nivelles (BE) December 20, 2001. ............. 27 

Other fire cases evaluated with FRAME v.2 ............................................................... 28 

FRAME version 2. Renovation projects........................................................................ 30 

Case study 17. A tower building, 5 th floor with HVAC installation .................................. 30 

FRAME version 2. Fire protection design. .................................................................... 31 

Case study 18. Bulk storage of plastic pellets in PVC big bags...................................... 31 

Case study 19. Fire safety of a historical building. ...................................................... 32 

Case study 20. Fire safety for an existing school building............................................ 34 

Case study 21. Sweets manufacturer safety improvement program.............................. 36 

Case study 22. Specialty chemicals unit safety improvement program.......................... 37 

Case study 23. University Library fire safety concept.................................................. 38 

FRAME version 2. Inspection and verifications. ............................................................ 40 

Case study 24. Temporary theatre installation........................................................... 40 

Case study 25. High rack storage ............................................................................ 41 

Case study 26. Fire safety for a drink manufacturing plant. ......................................... 42 

Other risk verifications by FRAME v.2. ...................................................................... 44 

FRAME version 2. Equivalency of concepts. ................................................................. 46 

Case study 27. General Chemicals warehouse. .......................................................... 46 

Case study 28. Medical Education Centre. ................................................................. 47 

Other equivalent concept calculations....................................................................... 48 

Non-industrial risks with FRAME v1. ........................................................................... 50 

Case study 29. Small clinic, 2 nd floor. ...................................................................... 50 

Case study 30. Company canteen and restaurant, 6 th floor.......................................... 51 

Case study 31. Hotel inspection report..................................................................... 52 

Other non-industrial risks........................................................................................ 53 

Plastics industries with FRAME v.1 .............................................................................. 54 

Case study 32. Medical supplies storage ................................................................... 54 

Case study 33. Medical gloves manufacturing, storage ............................................... 55 

Case study 34. Medical supply sterilization plant. ....................................................... 56 

Case study 35. Medical gloves manufacturing, production unit..................................... 57 

Case study 36. Bubble foam manufacturer - storage. ................................................. 58 

Case study 37. Bubble foam manufacturer – production unit. ...................................... 59 

More plastics industry case studies........................................................................... 60 

Textile industry with FRAME v.1.................................................................................. 62 

Case study 38. Stockings factory - production ........................................................... 62 

Case study 39. Stockings factory - storage................................................................ 63 

Case study 40. Ready made clothing factory ............................................................. 64 

Case study 41. Jute spinning plant. .......................................................................... 65 

Case study 42. Textile wholesale trader .................................................................... 66 

2

F.R.A.M.E. 

Case study 43. Synthetic fibre spinning mill survey. ................................................... 67 

More textile industry case studies. ........................................................................... 68 

Floor coverings industry with FRAME v.1...................................................................... 69 

Case study 44. Floor covering factory ....................................................................... 69 

Case study 45. Floor covering tufting........................................................................ 70 

Case study 46. Floor covering, latex coating.............................................................. 71 

Case study 47. Floor covering tufting........................................................................ 72 

Case study 48. Carpet manufacturing, storage .......................................................... 73 

Case study 49. Wool carpet weaving ........................................................................ 74 

Woodworking industry with FRAME v.1. ....................................................................... 75 

Case study 50. Woodworking shop........................................................................... 75 

Case study 51. Furniture manufacturer..................................................................... 76 

More woodworking industry case studies................................................................... 77 

Food industry with FRAME v.1. ................................................................................... 79 

Case study 52. Biscuit bakery.................................................................................. 79 

Case study 53. Chocolate bar production ................................................................. 80 

Case study 54. Chocolate production ........................................................................ 81 

Case study 55. Food additive (seasonings) manufacturer ............................................ 82 

Case study 56. Frozen vegetables plant .................................................................... 83 

Case study 57. Frozen vegetables storage................................................................. 84 

Case study 58. Large slaughterhouse inspection . ...................................................... 84 

Case study 59. Animal feed production, mixing unit, basement.................................... 86 

Case study 60. Animal feed production (bagging) ...................................................... 87 

Case study 61. Animal feed production (top of mixing tower) ...................................... 88 

Case study 62. Storage silos ................................................................................... 89 

More food industry case studies............................................................................... 90 

Electric and electronics industry with FRAME v.1........................................................... 91 

Case study 63. Electronic print manufacturing, building with several levels ................... 91 

Case study 64. Radio assembly hall.......................................................................... 92 

Case study 65. Electronic parts sub assembly hall...................................................... 93 

Case study 66. Special lamps manufacturing hall. ...................................................... 94 

More electric and electronic industry case studies ...................................................... 95 

Other industries with FRAME v.1................................................................................. 96 

Case studies 67. Paint manufacturer, finished products warehouse. ............................. 96 

Case study 68. Paint manufacturer, lacquer-manufacturing unit................................... 97 

Case study 69. Candle manufacturer. ....................................................................... 98 

Case study 70. Honeycomb material manufacturer..................................................... 99 

Case study 71. Book printing .................................................................................100 

Case study 72. Car seats assembly plant. ................................................................101 

Case study 73. Oil regeneration plant......................................................................102 

Case studies 74. Metalwork industry........................................................................103 

Other industry case studies. ...................................................................................104 

Real fire cases with FRAME V.1..................................................................................105 

Case study 75. Fire at Menuiserie Paul Ottignies (BE) , December 1988.......................105 

Other fire cases verified with FRAME v.1. .................................................................106 

FREME....................................................................................................................107 

Case study 76. Summerland fire, Isle of Man, August 2 nd , 1973.................................107 

3

F.R.A.M.E. 

Introduction. 

This examples book gives an overview of FRAME calculations made during a 20 years period. 

The purpose is to illustrate the various applications that can be made of the FRAME method. 

Look at the various cases and see how the influence of a single aspect of the fire risk can be 

overlooked or otherwise compensated by other favourable elements. A building can be safe for 

the occupants and yet be poorly protected for property. 

When the FRAME risk assessment was made for risk survey purposes, the identification data 

have been rendered anonymous to protect the privacy of the building owners, and no one shall 

conclude from these data that the FRAME calculation can be linked to a particular building. It is 

quite possible that most of the buildings have been modified since the time they have been 

visited, the FRAME calculation has only relevance in relationship with the picture at the 

moment of the survey. 

A number of FRAME risk assessments were made for a real fire cases, based on public reports 

that are mentioned in the table. These calculations were to check if that kind of fire could 

have been expected by use of FRAME. It is possible that the public reports do not contain all 

the information needed for a calculation. In those cases, “neutral” assumptions have been 

made and the input data used have no aggravating impact on the FRAME calculation. 

FRAME version 2 calculations. 

The tables give the summaries of the input data and calculation results as found in the report 

generated by the software. 

FRAME version 1. 

FRAME version 1 was primarily used on industrial risks. The formula for the evacuation time 

factor considered only the escape route inside the compartment, which resulted in most cases 

in rather short evacuation times. The number of exit paths was considered in the escape factor 

calculation. 

The analysis of a number of real fires in non-industrial occupancies indicated that the factors 

for occupants’ risk should be modified for buildings with a moderate to high occupancy load 

(more than 1 person per 5 m²). This resulted in modifications of the calculation of the 

evacuation time factor t and the escape factor U in FRAME version 2 with a consequent 

increase of the risk value for such cases. For industrial buildings, the difference with version 1 

remained minimal. 

The property value was compared with 1985 insured values, which results in a higher value for 

factor c, and the salvage factor considered the “business share” as part of the protection. 

These elements were also modified in version 2 and the application of that version to the same 

cases would result in slightly higher values for the activities fire risk. 

The FRAME version 1 was supported by a BASIC program, the tables were made with the data 

on the printouts of the reports. 

FREME. 

The predecessor of FRAME was “FREME – fire risk evaluation method” . This was a 

development of the Swiss Gretener method combined with a risk calculation for the occupants. 

A number of calculations according to this early version are also included, in this example 

book. 

4

FRAME calculations for real fires. 

F.R.A.M.E. 

Professional magazines like the “NFPA Journal”, Fire prevention and the ANPI magazine or 

“revue du feu belge” publish regularly fire report with enough details to make a FRAME 

calculation. These fire reports were used during the development of the FRAME method as test 

cases and afterwards as verification. 

The calculations indicated a need to improve the method for the occupants’ risk assessment in 

non-industrial buildings and resulted in the FRAME version 2. None of the calculations made 

with FRAME version 2 have shown a discordance between the real fires and the risk 

assessment. 

The following cases studies are a selection of real fire reports that have been used to check the 

validity of FRAME. The main reason to make these checks was to verify if the FRAME 

calculations are confirmed by reality : can the method give and adequate warning for high 

risks before a fire occurs, and is a good protection as good as assumed 

Most of the fire reports relate to catastrophic fires, and the unfortunate consequences were 

“foreseeable” by the FRAME risk assessment. 

In many cases, a catastrophic fire like the one that occurred on 11 April 1996 in the Düsseldorf 

Airport, provokes the S.A.B.E.N.A. Syndrome: "Such A Bloody Experience Never Again". It 

touches building owners and authorities alike, and the result is a building concept where any 

imaginable and available protection features is included without much consideration for the 

necessity and for the economics of such a design. This apparently happened also in Düsseldorf, 

as can be seen in that case study. 

A few reports are found that describe that the available fire protection was adequate, and the 

FRAME calculation complies with the result. The Luton fire report is such a case. 

Some fires resulted in less damage than could be expected from the risk assessment, which 

means that the risk assessment may be on the conservative side. 

ALL fires where the damage was more important than indicated by the risk assessment were 

caused by arson. This means that for “suspicious” fires the FRAME risk assessment can give a 

clue for arson. 

5

F.R.A.M.E. 

Case study 1. Bekina Indurub Kluisbergen (BE) , 1986. 

This first calculation was made for an insurance inspection on April 2 nd 1986, and nearly two 

years LATER, the factory was destroyed by fire… 

Item explanation (sub) factor Value Result 

Type of occupancy rubber boots manufacturing 

Fire load immobile combustible construction Qi 1000 

Fire load mobile Qm 2000 q=1.77 

Temperature rise T 200 

Average dimension m 0.5 

Reaction to fire M 3 i=1.13 

Length l 110 

Width b 45.7 g=1.76 

Level E 0 e=1.00 

Height of room h 6 

Ventilation k 0.01 v=1.01 

Access direction Z 2 

Height difference H 0 z=1.05 

Potential Risks P= 3.74 P1=2.13 P2=2.12 

Activation factor 

Main: rubber industry a 0.4 a=0.5 

Heating in room 

Electrical: ok 

Explosion: no 

Secondary: none 

0.1 

0 

0 

0 

Occupants Number: 50 

Mobility factor: 1 

Exits X 8 

Exit directions k 4 

Evacuation time factor 

t=0.03 

Content factor: 6.25 M. euro c=0.0 

environment factor r r= 0.60 

dependency factor d d= 0.30 

Acceptable Risks A= 1.07 A1= 0.47 A2 =0.80 

Water supplies limited W=0.57 

Normal protection limited training – no hose 

N = 0.60 

reels –no alarm 

transmission – 

Special Protection volunteer station S= 1.16 

Fire resistance Structure: 60 

Walls: 60 

Ceiling: 0 

Partitions:30 

f= 48.8 F=1.47 

Escape protection sub compartments U=1.89 

Salvage sub compartments y=4 Y=1.22 

Calculated Risks: R= 3.63 R1= 2.41 R2=3.00 

Conclusion: 

inadequately protected risk. 

With this risk assessment, the insurance company refused to take the contract for a better 

price than the existing policy. The fire was reported in the ANPI magazine n° 89 of February 

1988, as an “instructive fire”. (See below). Damage was then about 250 million BEF ( = 6.25 

Million Euro). As the fire started at 1h 38 in the night, few people were present and luckily 

everyone escaped. 

The factory was well insured both for property and business interruption and was completely 

rebuilt. 

6

F.R.A.M.E. 

Case study 2. Dupont Plaza Hotel Fire Puerto Rico, December 31, 1986 

This very large hotel (ground floor 23.000 m²) had a fire resistive construction, but no fire 

detection. A fire of criminal origin caused several deaths and large property loss due to the 

lack of adequate fire safety provisions. This FRAME calculation was made with the data found 

in the NFPA fire report. 


Type of occupancy large hotel 

Fire load immobile Qi 100 





Length l 163 

Width b 143 g=4.45 








Main: residential 

a 0 

a=0.25 

Heating : air conditioning 

Electrical: no control 

Explosion: no 

Secondary: yes 

0.1 

0.1 

0 

0.05 



+risk of 

panic 

Exits X 10 



t=0.61 

Content factor: 40 M. dollar c=0.19 


dependency factor average d d= 0.3 


Water supplies 

W=1.00 

Normal protection no proper alarming, 

N = 0.49 

training, hose reels 

Special Protection full time station S= 1.48 


F=1.80 

Walls: 60 

Ceiling:120 

Partitions:60 

Escape protection horizontal evacuation 

U=1.62 

50% 

Salvage y=0 Y=1.00 


Conclusion: 

poor protection 

These results indicate a design prone to catastrophe and it happened. On the afternoon of 

December 31, 1986 a fire occurred and resulted in 97 fatalities and between 6 and 8 million 

dollars of damage. A detailed report is available through www.nfpa.org 

7

F.R.A.M.E. 

Case study 3. Hotel International at Zurich (CH) , February 14, 1988 

This high-rise hotel (28 floors) had a fire resistive construction and automatic fire detection. A 

fire in the restaurant at the 24 th floor resulted in 6 deaths and total destruction of that 

compartment. The FRAME calculation was made with the data found in the June 1988 ANPI 

magazine n° 91, “instructive fire” report nr 137. 

The fire case was used in the development of the FRAME version 2, as the calculation by 

FRAME version 1 did not indicate such a high risk for the occupants. 


Type of occupancy restaurant in high rise hotel 






Length l 26 

Width b 19 g=0.62 

Level E 24 e=1.89 


Ventilation k 2 % v=0.88 






Heating : ok 


Explosion: no 

Secondary: yes, cooking 

a 0 

0 

0 

0 

0.1 

a=0.10 



+risk of 

panic 

Exits X 10 



+ no clear 

exits 

t=0.63 

Content factor: 5 M. euro c=0.00 





Normal protection 

Special Protection 

no proper alarming, 

training, hose reels 

fire detection, 

professional fire brigade 


Walls: 60 

Ceiling:60 

Partitions:0 

Escape protection 

fire detection, horizontal 

evacuation 50% 

W=1.00 

N = 0.66 

S= 2.65 

F=1.69 

U=4.32 

Salvage 

Y=1.47 


Conclusion: 

poor protection 

These results indicate a situation with a low risk for property but a high risk for the occupants. 

The 3 main factors that caused the catastrophe were the combustible decoration, the lack of 

organisation and training of the personnel, and the location at the 24 th floor. The 

8

F.R.A.M.E. 

corresponding factors in the FRAME calculation are : the fire spread factor i and the 

environment factor r, the level factor e and also low value of the normal protection N. 

Case study 4. Hotel fire in Kristiansand (NO) , September 5 th , 1986 

Reference : Fire Prevention n° 198, April 1987 

In the early morning of 5 September 1986 a severe fire swept through the Hotel Caledonian in 

Kristiansand in the south of Norway. The fire resulted in the death of fourteen guests and 

injury to fifty others. In the event, the fire protection features of the hotel proved inadequate 

and staff training not very effective. 

The hotel, constructed in 1968, complied with the building regulations in force at the time. It 

also complied with the Norwegian Hotel Fire Precautions Act 1963 that, among other 

requirements, ordered that the entire premises be protected by an automatic fire alarm 

system. Unfortunately, the system installed comprised only heat-activated detectors but it did 

have a direct link to the fire brigade. It was also linked to a tape system that relayed 

information to guests in case of fire. 

The lower three storeys of the building contained the reception area, restaurants, discotheques 

and conference facilities. The third storey in the tower housed the administration offices and 

the fourth to twelfth storeys contained two hundred guest rooms. 

On the night of the incident, there were 113 guests in the hotel. Two night porters were on 

duty. At around 0440 hours the porter checking security was alerted by the fire alarm. He 

immediately took the lift to the first storey reception area where he was greeted by flames and 

smoke near the stairwell. Together with the other porter, he ran through the restaurant into 

the street where both men found themselves locked out. 

The fire brigade in Kristiansand did receive a telephone call at 0440 hours with the message 

'Fire in the reception of Hotel Caledonian' and seconds later the automatic fire message came 

through. The brigade responded by immediately. On arriving at the hotel three minutes later, 

the firemen were confronted by a very serious fire. 

The emergency staircase led down to the reception area that was aflame. The first firemen to 

arrive concentrated their initial effort in this area as they believed that the stairway was filled 

with trapped guests who could not get through the burning reception area. However, the two 

jets initially available proved no match for the intense heat. It was also established that none 

of the guests had reached the emergency stairs due to smoke logging of corridors outside the 

bedrooms. 

The fire spread rapidly and soon involved the first four storeys of the hotel. Firemen began 

rescuing guests with ladders, breathing apparatus teams entered the hotel through broken 

windows to assist guests onto the ladders. Conditions were extremely difficult with heavy 

smoke pouring from the building. 

The fire was thought to have been caused by an electrical fault in the wiring to a ceiling lamp 

located in the stairway between the first storey reception and the ground storey restaurant. 

The lower storeys were found to have been inadequately compartmented, with gaps between 

fire resisting walls and floors. The walls and ceilings had combustible linings that assisted fire 

spread. 

FRAME calculation 

The report of Fire Prevention n°198 , April 1987, did not give al the features of the building, so 

a FRAME calculation can only be made with some guesswork for the missing parameters. 

However, this first rough approach gave as a result R=1.00, R1=2.54, R2=1.65. These values 

9

F.R.A.M.E. 

are an indication that property damage will be limited, but that it is likely that a serious fire 

would make victims. 

Case study 5. Honda Spare parts centre, Ghent (BE), October 29 th 1987. 

This modern spare parts warehouse of the renowned car manufacturer was located at 10 min. 

from my home. On that morning, it burned completely down in about 4 hours. The offices, the 

boiler house and the electrical substation, which were separated from the main building by 

firewalls, were saved. The people in the building were informed by the fire detection and alarm 

system and could all leave the building in due time. There were no victims. It took the 

company about two years to rebuild a new (overprotected) warehouse. During that time the 

company had to distribute its operations over several rented warehouses in the area. 

The origin of the fire remains unknown. A defective lighting fixture in the rack storage has 

been indicated as the most probable cause. This 31.500 m² large warehouse was used as 

European spare parts distribution centre of a car manufacturer. Spare parts for all current 

types were stored here, and orders from garages and stockists were handled here and 

dispatched within 24 hours. In the building there was 6.7 m high rack storage and order 

picking areas. Some 160 persons worked in the building. 

The building had a steel structure with light concrete external walls. Roof height was 7.5 m. 

Fire protection consisted of an automatic fire detection system, smoke vents at a ratio of 2 % 

of the floor area, extinguishers and hand hoses. The plant had a first intervention team, the 

local fire brigade is a full professional fire brigade at 10 min distance from the plant. The plant 

had no independent water supply, but is located at +/- 300 m from the sea canal Ghent- 

Terneuzen. 

This case study was first documented in the technical file DT73 évaluation des risques, (1988) 

of ANPI Belgium that explained the FRAME version 1. The Risk calculation gave an initial risk 

Ro =10.24, which means that compartimentation will be needed to reduce the property risk to 

an acceptable level. The calculated values for this situation were : 

R = 5.95 , R1 = 0.47, R2 = 1.79 

These results mean that a total loss could be expected, without victims (R1 is OK) but also 

with serious business interruption consequences. AND THAT IS WHAT HAPPENED ! 

Major factors that contributed to this situation: 

-very large compartment g =5.13 

- high fire load q = 2.03 

- water supply not immediately available W = 0.66 

The case study was recalculated with FRAME version 2 , and because of the changes in the 

factor t formula, slightly different results were obtained: 

R = 6.63, R1 = 0.45 and R2=1.98. The conclusions are the same. 

10

F.R.A.M.E. 

Case study 6. Fire at “Bourse de Bruxelles” (BE) on November 30, 1990. 

Reference : ANPI magazine “revue belge du feu” n°104, February 1991, étude n° 150. 

The Bourse de Bruxelles building is located in the centre of Brussels and was built in 1873- 

1876. It is classified as a historical building. The building is 80 m long and 44 m wide, and 

has a basement, a ground level and 3 upper levels. It was the seat of the Brussels stock 

exchange market. 

In the morning of November 30, 1990 a by-passer saw smoke coming out of a window and 

called the fire brigade, which is located nearby at 6h21. The fire was extinguished at 7h30. It 

was located at the first floor, which was extensively damaged. The Brussels stock exchange 

was temporary transferred to an other location for some days. The origin of the fire was 

probably due to an electrical fault ( included in FRAME in factor a). 


Type of occupancy Bourse de Bruxelles, Brussels Belgium, 1990 

Fire load immobile wood partitions Qi 300 





Length l 80 

Width b 44 g=1.83 

Level 

ground floor + open E 1.8 e=1.34 

galleries 







Main: trade 

Heating : ok 


Explosion: no 




a 0 

0 

0.1 

0 

0 

during 

working 

hours 

Exits X 10 



a=0.10 

t=0.07 

Content factor: 2 M. Euro c=0.00 




Water supplies adequate W=1.00 

Normal protection limited training, N = 0.90 


permanently staffed 

station 


Walls: 60 

Ceiling:60 

Partitions:0 

S= 1.48 

F=1.49 

Escape protection horizontal evacuation U=1.80 

11

F.R.A.M.E. 

50% 

Salvage 

financial data protected, 

Y=1.41 

repairs possible, 

relocation possible 


Conclusion: 

Inadequate level of protection for property and people, luckily 

the fire occurred outside working hours. 

The ANPI report indicated as major cause for the extent of the damage, the lack of automatic 

fire detection in the building. 

Adding automatic fire detection to the calculation changes the values of the protection factors : 

S becomes 2.65, F becomes 1.44, U becomes 3.23. The risk values are then R= 0.77 R1= 

0.89 and R2=0.55. The FRAME calculation confirms the ANPI expert advice. 

Case study 7. Furniture factory fire Kortessem (BE) , October 1991 

Reference: ANPI Magazine n°109, February 1992 

On October 1991, a fire destroys half of “SCHRIJNWERKERIJ MEUKO" in Kortessem Belgium. 

The fire started at 9h.05 in the morning during repair work in an old paint unit (area C) and 

caused 40 million BEF ( approx. 1 M. Euro) damage. 

The furniture factory had a floor area of +/- 3.000 m², steel construction with concrete and 

stonewalls. Building A was used as storage for finished goods, woodworking machines were 

located in F, G and H. Building E contained furniture in production, and the offices were in D. 

The building was protected with fire extinguishers only and the staff could not cope the fire 

with them. The Hasselt fire brigade was called immediately and arrived at the plant in less 

than 15 minutes. The local water supply consisted of a pond at 300 m distance and was 

insufficient to cope with the fire. Reinforcement came with tank trucks from Hasselt and Genk, 

and the fire was under control at 19.00 hrs and extinguished at 23.45 hrs. 

The report in ANPI magazine indicated that only sprinkler protection would have been 

adequate for this type of building. The report permitted a scanty FRAME calculation with the 

following results : Property risk R = 3.38 , occupants’ risk : 2.67 , business risk 3.12 and 

Initial risk Ro = 1.76: This means that the FRAME calculation also indicates that sprinklers are 

the recommended protection in accordance with the ANPI expert advice. 

The FRAME risk assessment “ as is” shows a higher damage potential than what happened. In 

fact, the rapid call for the fire brigade, who came from the beginning with heavy equipment 

resulted in saving half of the building. 

Case study 8. Imperial Foods Plant Fire, Hamlet (NC), September 1991 

References : NFPA Alert Bulletin 91-4 and ANPI magazine 111. 

On Tuesday, September 3, 1991, at approximately 8:15 am, a fire occurred at the Imperial 

Foods Processing Plant in Hamlet, North Carolina resulting in 25 fatalities and 54 injuries. The 

intense fire quickly spread products of combustion throughout the plant causing employees to 

search for available exits. Although many of the estimated 90 occupants escaped without 

incident, others found exterior doors unavailable and sought alternative means of escape. Not 

all of those who remained were able to be rescued and many perished. 

12

F.R.A.M.E. 

Background 

The fire occurred in a one-story, unprotected non combustible building of approximately 

33,000 sq. ft. The non sprinklered, windowless building had undergone several changes in 

ownership and numerous alterations over the years including several additions to the building 

and changes to its interior partitions. 

The plant operation included preparing and cooking chicken pieces for distribution to 

restaurants. During a normal shift, approximately 90 employees would be involved in the 

various phases of production. Portions of the plant included a large preparation area, packing 

and storage of preparation materials and finished product. Several of these storage areas were 

cooler or freezer units. 

The facility contained no effective barriers to prevent the spread of products of combustion. 

The building had numerous exterior openings for shipping and receiving of goods and for 

personnel doors, but the personnel doors lacked various components, sizing, or arrangements 

that prevented their utilization as a component of a properly designed exiting system. 

The Fire Incident 

The fire originated in the processing area of the plant at approximately 8:15 am. The cause of 

the fire was determined to be the ignition of hydraulic oil from a ruptured line only a few feet 

from a natural gas-fueled cooker used in preparation of the chicken. The ignition resulted in a 

rapidly developing and spreading fire that quickly distributed thick, black smoke throughout 

the building. The intense fire also impinged upon a natural gas regulator, which soon failed and 

added to the fuels being consumed. Occupants within the affected areas had little or no time to 

evacuate before their way was blocked or their visibility obscured by the thick smoke. 

Employee Actions at the Time of the Fire 

The employees in the processing room attempted to reach safety by moving through structural 

openings. Employees outside of the processing room were alerted to the fire by other 

employees shouting and by smoke. A cluster of occupants found an inoperable personnel entry 

door located next to the loading dock. Within one (1) minute of discovering the fire, these 

employees found that the smoke had obscured their visibility... 

Others took refuge in the cooler when they found that they could not move past the fire area. 

Employees in other portions of the building were not as severely exposed to the fire. 

Fire, Rescue, and Emergency Medical Services Operations 

The Hamlet Fire Department was notified of the fire when an employee drove several blocks to 

the fire station and advised fire personnel on duty that there was a serious fire at the plant. 

First responding units were immediately confronted with 14 casualties outside the south end of 

the plant. The first actions of arriving fire fighters was to administer first aid to the injured. 

First arriving fire fighters observed evidence of a severe fire burning within the building, 

prompting requests for mutual aid assistance from neighbouring fire and rescue departments. 

Analysis of this incident illustrates the importance of adhering to basic fire safety principles in 

order to ensure life safety. Among these principles are the need for automatic fire detection 

and fire suppression systems, a fire alarm system for alerting occupants, proper exit design 

(including an adequate number, properly sized and arranged exits), and the need for occupant 

fire safety training. 

FRAME Risk assessment 

A FRAME calculation was made with the elements available from the NFPA Alert Bulletin nr 91- 

4. The results were R=2.51 R1= 3.46 and R2=3.10. The catastrophic fire was foreseeable ! 

Major elements that contributed to this drama were the lack of compartimentation between 

hazardous and non hazardous areas, the lack of openings for smoke evacuation, the lack of 

13

F.R.A.M.E. 

adequate warning and the lack of training of the people. In the FRAME calculation, these 

considerations are reflected by the value of the following factors: 

the area factor g, the ventilation factor v, the activation factor a, the normal protection N, the 

special protection S and the escape protection U. 

Case study 9. Department store fire Luton Bedfordshire (UK) January 1996. 

A fire that was controlled by sprinklers caused only £315,000 damage to this £5m 

department store. 

This report was published in "Fire Prevention" n° 287 , March 1996. Based on this report , a 

FRAME calculation was made to check if the conclusions of the report are reflected in the risk 

assessment . It is clear from the calculation that the use of sprinklers for this kind of 

occupancy is a proper engineering choice; in addition, it proved very valuable, as the origin of 

the fire was probably arson. FIVE SPRINKLER heads activated and Bedfordshire Fire and 

Rescue Service were automatically alerted when a fire broke out in this department store. 

The FIRE REPORT 

Fire fighters arrived to see smoke coming from the first floor of the building and made a forced 

entry. A breathing apparatus team entered the smoke-logged first-floor storage area, a 

2000m2 undivided room. Using a thermal imaging camera, the team found the fire in a stack 

of display material against an external wall. The sprinklers had contained the fire and the fire 

fighters were able to extinguish it using a hose reel. 

They then cleared the area of smoke using a fan carried by the crew to prevent any further 

damage and began salvage operations. The seat of the fire was below a window, which had 

been broken. It appears that intruders climbed a fixed-ladder fire escape close to the window 

in an attempt to enter the building. Investigators believe that they broke the window and then 

lit the fire because bars inside the window prevented them entering. 

The storage area was full of racking which almost reached the low ceiling. It was protected by 

sprinkler heads at 2m intervals along the aisles between the racks. The brigade believes that 

the fire might well have spread rapidly throughout the first floor if the sprinkler system had not 

operated. The combined value of the building and stock, both of which could have been lost, 

was about £5m. The store was closed for only one day following the fire. 


Type of occupancy Department store Luton UK, 1996 






Length l 45 

Width b 45 g=1.83 

Level level 1 E 1 e=1.22 







Main: storage 

Heating : ok 


Explosion: no 


a 0 

0 

0 

0 

0 

a=0.0 

14

F.R.A.M.E. 



Exits X 2 



t=0.05 






Normal protection fire brigade at 15 min, 

N = 0.81 

limited training 

Special Protection permanently staffed 

S= 3.07 

station, sprinklers 


F=1.19 

Walls: 30 

Ceiling:30 

Partitions:0 

Escape protection Sprinklers, max. 300, 

U=4.53 

smoke vents 

Salvage 

repairs possible, 

Y=1.28 

relocation possible 


Conclusion: 

adequate protection 

The damage of this (incendiary) fire was 6 % of the total value. This fits well with the Probable 

maximum loss (with the protection working) according to the FRAME formula PML = 10 R % or 

8 %. 

Case study 10. General warehouse fire New Orleans, Louisiana March 21, 

1996. 

THIS MASSIVE warehouse fire burnt for six days before being brought under control by the 

New Orleans Fire Department. The damage, which was caused by an arson fire and a 

subsequent electrical fire, could possibly have been avoided. Complications involving the 

sprinklers were to blame, according to the NFPA. 

The 380 m by 278 m large warehouse and distribution facility stored closeouts and 

discontinued items in two rack areas. The building structure consisted of unprotected steel I- 

beam columns and girders, and roof support was provided by unprotected metal bar joists. The 

roof was constructed of a single-ply membrane with rigid insulation over a corrugated metal 

deck. Part of the building was an office area and segregated storage area that occupied 19,021 

m². The roof over the main warehouse was on two levels. The lower level ranged from 11m to 

12m. The upper level, which covered the higher of the two rack areas, had a maximum height 

of 22m. The main warehouse floor, which measured 86,400 m², did not have any fire 

separation barriers. 

On the south side commodities were delivered by truck, on the north side, they shipped out. In 

the east sections of the building, there was a low portable rack storage area. These racks were 

not equipped with any in-rack sprinkler systems. The ceiling was 22m high in this area. In the 

middle of the building were high-bay rack storage areas. The racks in these areas were 19m 

high, 31 m long and 1.2m deep. There was a mixture of commodities stored on these racks, 

including wicker baskets and furniture; rugs; polyfil pillows; cardboard boxes; duvets; towels; 

stacked plastic chairs and plastic bags. In addition to the racks, the main warehouse area 

housed a variety of conveyor systems and a mechanized retrieval system. 

15

F.R.A.M.E. 

The building was equipped with 30 overhead sprinkler systems and 17 in-rack sprinkler 

systems. The facility was supplied by a 10 " looped water main connected to a 20" municipal 

water main. A 1,135,600 litre water tank provided an additional water supply. Two 9,462 

litre/min fire pumps, one electric and one diesel, were connected to the fire protection system. 

The electric pump was connected to the water mains and the diesel pump was connected to 

the water tank. 

The sprinkler system over the area of the first fire was designed to supply a density of 18 litre/ 

min/ ml. However, there were no in-rack sprinklers in this area. The ceiling was located about 

22m above the floor and 16m above the highest level of racks. 

A fire alarm system monitored the sprinkler systems, valves and fire pumps. Alarm and trouble 

signals were transmitted to an on-site security office that was continually monitored. Alarm 

signals were transmitted to an off-premises supervising station that was responsible for 

notifying the fire department. At the time of the first fire, there were 15 employees in the 

building. 

The first fire, reported at 5.32am, quickly involved the combustible materials being stored in 

the 6.4m high portable racks. When the fire was discovered, it was reported to have fully 

involved the three-tier rack, and was extending to the ceiling. The rapid fire spread can be 

attributed to the combustible nature of the products being stored and their storage 

configuration. Employees tried unsuccessfully to fight it using portable fire extinguishers and 

occupant-use fire hoses. The overhead sprinkler system, that was located approximately 15m 

above the top level of the racks, activated at 5.32am, but did not extinguish or control the fire. 

The fire was extinguished by the fire department using an interior attack. 

The New Orleans Fire Department responded at 5.37am with an initial attendance of six 

appliances. The first appliance arrived at 5.40am and reported heavy smoke showing. Eight 

additional units were dispatched at 5.42am. The fire was reported to be coming through the 

roof at 5.43am. Fire fighters began an interior operation to contain the fire. It was declared 

under control at 8.44am, and extinguished at 11.54 am. 

Once fire control had been achieved, all the sprinkler systems in the facility were manually 

shut down by closing the individual valves on the risers. According to the New Orleans Fire 

Department, this was because numerous sprinkler heads had opened throughout the 

warehouse, causing extensive water damage. At 2.20pm, electrical power was restored to the 

conveying system within the racks in the south-central portion of the building. Damaged wiring 

in this area arced and ignited combustible materials. As soon as the second fire was 

discovered, facility personnel began opening the recently shut valves on the ceiling and in-rack 

sprinkler systems. The catastrophic damage that followed is attributed to the fact that the 

sprinkler systems were shut down throughout the entire facility. Consequently, the sprinkler 

system was overwhelmed, and the fire department had to mount a defensive operation. 

This second fire was declared under control on Wednesday, 27 March 1996 at 9.08pm, six 

days after the first fire. 

The warehouse and distribution portion of the building, and its contents, were destroyed by the 

second fire. There was minimal fire and water damage to the area located on the other side of 

the fire separation wall. There was some structural damage to the walls due to the stresses 

placed on this portion of the building by the collapsing roof on the east side of the fire 

separation wall. On this side, structural collapse was widespread. 

The devastation caused by the fires was the result of a number of factors, including: 

Excessive clearance between the ceiling sprinklers and the top of the storage racks in the area 

of fire origin for the first fire, lack of in-rack sprinkler systems ,shutting off all the sprinkler 

systems in the building following the first fire restoration of electrical service without 

16

F.R.A.M.E. 

evaluating the damage to the electrical system. This action was determined to have caused the 

second fire. 

The results of the FRAME calculation for this case are: 

The area factor g = 8.61 

The initial risk Ro = 18.87 

R = 3.17 , R = 0.94 , R = 1.50 : the property is NOT adequately protected, but the risk for the 

occupants is acceptable 

With this case study three features of the FRAME method and program can be illustrated: 

First, the results of the calculation (and the fire itself) show that the building is not adequately 

protected, although it is equipped with sprinklers. FRAME has a built-in balance between risk 

factors and protection, which makes that very large compartments with high fire loads are not 

considered as acceptable, even with sprinkler protection. 

Second, the user of the program will not be allowed to introduce the 22 m value for the ceiling 

height, as the program accepts only a 15 m maximum. There are two technical reasons for this 

limitation : one is that sprinkler activation will be delayed and sprinkler protection can become 

ineffective when there is too much clearance between the ceiling and the fire seat, as this case 

illustrates dramatically. The second reason is that in such high spaces like in atria, 

stratification of the smoke layer can occur, making smoke detection, smoke venting and 

sprinklers ineffective. By refusing values above 15 m for the ceiling height, the program makes 

the user aware of the fact that he is confronted with a special situation where particular care is 

required in the design of the fire protection. 

The third interesting feature is that the value of Ro gives the designer an early warning that 

the risk is too large to be properly protected . Very large risks shall be provided with 

redundant protection systems, which means in practice that the building has to be divided in 

compartments and sprinklered as well. 

For this particular case, the report as publicised in Fire Prevention (UK) issue 300 of June 

1997, expressed concerns about the Sprinkler system performance, as it was not designed 

according the requirements of the NFPA codes. 

The sprinkler systems never had a chance to control the second fire because the main control 

valves were shut off after the first fire was extinguished. A sprinkler system's ability to control 

a fire is greatly reduced if it cannot operate during the initial stages of fire development. By 

the time the second fire was discovered and efforts were made to open the valves, the fire had 

grown to a size where the sprinkler discharge would not have been effective. 

While properly designed, installed and maintained sprinkler systems have an excellent record 

of controlling fires, it can be necessary to employ a more comprehensive fire protection plan, 

which accounts for uncertainties, when extremely large values are at stake. This is exactly the 

intention of the built-in balance of risk factors in the FRAME method. 

System impairment 

During overhaul operations, all of the ceiling and in-rack sprinkler systems throughout the 

building were shut down, despite the fact that fire damage was limited to a relatively small 

area in the north-eastern portion of the building. According to the fire department, it was 

necessary to shut off each riser because a number of sprinkler heads had opened throughout 

the facility following the first fire. 

As the electricity was restored, the second fire broke out in an area remote from the first and 

there was no water in the sprinkler systems to control it. The fact that each sprinkler system 

was shut off individually contributed significantly to the resulting destruction of the building. 

17

F.R.A.M.E. 

Utility Restoration 

The restoration of electricity was done before the electrical system could be fully evaluated for 

damage, and before the sprinkler system was made operable. Failure to isolate or repair the 

damaged section prior to restoring power was another key factor in contributing to the ignition 

of the second fire. It is imperative that services are restored in a planned and coordinated 

manner, and only after a complete survey of the damage has been taken. There are several 

hazards associated with an uncoordinated restoration of utilities. The potential risk to rescue 

personnel and other workers operating in the area during overhaul and restoration is 

significant. Utility systems that have been damaged during a fire should be inspected and 

repaired before being restored. It is also imperative that everyone involved with the overhaul 

and restoration operations be aware of what actions are being taken by the different agencies 

and services involved. 

Case study 11. Dusseldorf Airport (DE) fire April 1996. 

During the worst fire in the history of Düsseldorf and the incident with the highest loss of life in 

the Federal Republic of Germany since the end of World War 2,17 civilians died, 72 people 

suffered serious injuries, and several hundred people suffered minor injuries. The cost of the 

damage is still not completely known, as some areas of the airport have not been restored to 

full operation. Estimates vary between 300 million and 1 billion German Marks (approximately 

U.S. $200 million to $600 million). 

The fire began at approximately 3:31 p.m., about the time someone reported seeing sparks 

fal1ing from the ceiling in the vicinity of a flower shop at the east end of the arrivals hal1 on 

the first floor. When two fire fighters from the Airport Fire Brigade responded at 3:33 p.m., 

they detected an odour inside the building and asked that an electrician respond, as problems 

with the motors on the automatic doors in the area had been reported in the past. 

At 3:38 p.m., however, smoke was seen coming from the vents in the flower shop, and the 

ceiling began to glow and drop burning embers. All airport apparatus and personnel were 

requested and were on the scene by 3:40 p.m. 

At 3:58 p.m., a very rapid fire build-up occurred throughout a large area of the first level of 

the terminal, and the Düsseldorf Fire Brigade was called. 

By the time the fire was extinguished, 700 personnel from 12 different rescue services or 

municipalities had responded to the incident on 215 pieces of apparatus. The fire was finally 

declared under control at 7:20 p.m., 3 hours and 49 minutes after the first report of sparks 

was called in. 

German authorities determined that the fire began when a welder working on expansion plates 

in a roadway above the lower level of the terminal building ignited the polystyrene insulation 

used in the void above the ceiling on the first level. The smoke and flames spread throughout 

the first level, then extended to the second level through unprotected open stairwells and 

escalator openings. The fire did significant damage in the vicinity of the stairwells, and heavy 

smoke damage throughout approximately two-thirds of the second and third levels. Smoke 

also spread to the fourth level through unprotected escalator openings. 

The airport was completely shut down for 3 1/2 days following the fire. Limited operations 

were restored the Monday after the blaze, and the airport was back to 90 percent operations 

as of July 1, 1996. Tents and hangars were used as temporary terminals. 

With the data found in the German report of the "112"-magazine and its NFPA translation, the 

following FRAME calculation has been made. 

18

F.R.A.M.E. 


Type of occupancy Dusseldorf Airport Ankunfsthalle, 1996 





Reaction to fire M 1.50 i=0.95 

Length l 560 

Width b 70 g=3.85 

Level level 1 E 1.5 e=1.30 







Main: non industrial 

Heating : air conditioning 


Explosion: no 


a 0 

0.1 

0 

0 

0 



Exits X 55 



a=0.1 

t=0.61 






Normal protection limited training N = 0.90 



station, private airport 

brigade 


Walls: 60 

Ceiling:60 

Partitions:0 

S= 4.12 

F=1.51 

Escape protection signalisation, fire brigade U= 2.29 

Salvage safeguarded data Y=1.34 


Conclusion: 

inadequate protection 

It is a pity to find out from the FRAME calculated vales that such fire catastrophe could be 

expected and that the fire protection for this building was clearly unsatisfactory. The R- Value 

= 2.55 for property means that a more than 100 % loss can be expected, which means that 

the fire can develop beyond the compartment of origin (which it did) , the R1-Value = 2.01 

means a unacceptable high risk level for the people inside ( 17 killed, many injured!). And the 

R2-value gives a rather moderate but still unacceptable level for business interruption . 

The new Airport 2000+ concept. 

Reference : BRANDSCHUTZ -Deutsche Feuerwehr-Zeitung nr. 7/2001. 

19

F.R.A.M.E. 

The central airport building was completely renewed and provided for a larger flow of 

passengers with more comfort and more services. The new fire safety concept addresses all 

weaknesses found during the analysis of the 1996 event, and uses the following main goals : 

minimising fire loads, 

compartimentation, 

early detection of fire, 

smoke control, 

safeguarding the evacuation 

The data available in the reference article were enough to recalculate the fire risk for the new 

concept. The results of this calculation show a very low residual risk, which goes beyond the 

standard "good" level proposed by FRAME. In the next table , the improved values are in bold. 


Type of occupancy Dusseldorf Airport 2000+ concept 






Length l 250 

Width b 90 g=3.61 

Level level access improved E 0 e=1.00 


Ventilation smoke venting system k 1.16 % v=0.94 





Main: non industrial a 0 

a=0.0 

Heating : ok 


Explosion: no 


0 

0 

0 

0 



Exits X 72 



t=0.37 







Special Protection automatic detection, 

S= 14.64 

sprinklers, permanently 

staffed station, private 

airport brigade 


F=1.08 

Walls: 60 

Ceiling:60 

Partitions:0 

Escape protection see above U= 12.04 

Salvage 

safeguarded data and 

Y=1.63 

repair 


Conclusion: 

inadequate protection 

20

F.R.A.M.E. 

The pendulum swung to the other side as could be expected: Based on the FRAME calculations, 

it is possible to say that in the Dusseldorf Airport, the fire protection design was not properly 

balanced: before the catastrophe, it was too poor, afterwards a costly overdone design was 

chosen.... 

Case study 12. The Sight and Sound Theater (US) fire on February 28, 1997. 

This fire has been extensively documented in report 097 of the major fires investigation 

project, published by the United States Fire Administration. The following information comes 

from the report overview: 

“On the morning of January 28, 1997, in the Lancaster County , Pennsy1vania township of 

Strasburg, a fire caused the collapse of the state-of-the-art, seven year old Sight and Sound 

Theater and resulted in structural damage to most of the connecting buildings. The theater 

was a total loss, valued at over $15 million. 

The stage area was undergoing renovation and the theater was closed to the public, however, 

approximately 200 people, construction staff and employees were in the building at the time 

the fire started. Although the theater was built to conform to a two hour, fire rated assembly 

code requirement, many other fire protection features that could have assisted in saving the 

structure and reducing the damage were not present. 

Further contributing to the resultant loss, was the failure of the alarm system to notify the fire 

dispatch communications center and the lack of an adequate, readily available water supply. 

The volunteer fire departments that responded were faced with difficult fire conditions and 

tactical challenges for which they had not been adequately trained, and were without the 

benefit of adequate pre-planning. Local fire service suggestions for built-in fire suppression and 

smoke ventilation systems during the pre-construction plan review phase were ignored. 

While the incident was influenced by many conditions and situations that contributed to the 

large fire loss, fortunately there was no loss of life and only six minor injuries to the 

construction company staff. If the 1,400 seat capacity auditorium had been full, the situation 

could have been catastrophic.” 

The following issues have been identified in the report 

summary of key issues 

issues comments FRAME 

Fire origin Welding operations caused the fire. added as risk factor in the 

activation factor a 

Sprinkler System 

Waived 

Alarm System Failure 

Employees Delayed 

Report of Fire 

Structural Failure 

The requirement for sprinklers in the high 

hazard storage area under the stage was 

offset for a central alarmed smoke 

detector system and 2-hour fire walls. 

The alarm system failed to notify the 

county fire communications center and 

contributed to a delayed response. 

The fire was discovered by theater 

employees who used fire extinguishers for 

several minutes calling 911. 

Construction on the stage floor damaged 

the sprayed-on fire-resistant coating of 

see special protection 

factor S =1.80 , calculation 

of initial risk R0 also 

indicates the need for 

sprinkler protection 

Thermal detection included 

in special protection S and 

escape factor U 

intervention time 

estimated at 15 minutes, 

not included in the 

calculation, fire resistance 

21

F.R.A.M.E. 

steel structural members. The rapid fire 

spread caused early structural failure of 

the stage floor and contributed to fire 

extension 

assumed adequate 

Water Supply 

Eliminated 

Lack of 

Compartmentalization 

Lack for Exterior Fire 

Stream Access 

Inadequate Staff 

Training 

Pre-Fire Planning 

Fire Department 

Tactical operations 

Lack of Local Fire 

Code 

The water supply pond originally intended 

to provide water for fire suppression was 

eliminated to accommodate the addition 

of a prop manufacturing building. 

Several additions to the original theater 

connected the main structure to the 

maintenance buildings. The additional 

construction increased the overall size of 

the complex and compromised the 

compartmentalization. The high hazard 

storage area was not discretely separated 

from the rest of the structure. 

Lack of windows prevented exterior fire 

streams from being effective until the roof 

collapsed. 

The theater staff and the fire department 

had not trained together on managing a 

fire emergency in this technically and 

tactically challenging complex facility 

The fire department was not familiar with 

the new additions and the increased 

potential risk to fire-fighter safety. 

The initial fire attack was conducted with 

under-sized hand lines, inadequate for the 

heavy fuel load in the building 

There was no mandate for local 

government, fire officials, and local 

building owners to coordinate to ensure 

fire safe occupancies. 

This situation causes a 

very low value for the 

water supply factor W 

Size of building gives area 

factor g 

this aspect results in a 

severity increase by the 

value of the ventilation 

factor v 

lack of training in normal 

protection N 

not included in FRAME 



The FRAME calculation is based on the information in that report , and confirms the potential 

for a catastrophic fire. According to FRAME, the following improvements would be required: 

sprinkler protection with adequate water supply and outside fire loop with hydrants 

smoke venting system for the room 

eliminating the high combustible material from the main compartment 


Type of occupancy sight and sound theatre fire , Pennsylvania, 1997 






Length l 61 

Width b 61 g=1.83 


Height of room h 12.2 

Ventilation k 0 v=1.16 

22

F.R.A.M.E. 





Main: public 

Heating : ok 


Explosion: no 

Secondary: welding 

a 0 

0 

0 

0 



+risk of 

panic 

Exits X 29 



0.1 

a=0.10 

t=0.36 

Content factor: 15 M. dollar c=0.04 




Water supplies no distribution network W=0.54 



limited training, fire 

brigade at 10 min 

fire detection, small fire 

brigade 


Walls: 60 

Ceiling:60 

Partitions:0 

N = 0.81 

S= 1.80 

F=1.73 

Escape protection fire detection U=2.08 

Salvage not verified Y=1.00 


Conclusion: 

poor protection with potential for catastrophe 

Case study 13. Unifast Braine-le-Comte (BE) Fire January 11 th , 1999. 

This fire was described in detail in the ANPI magazine n°144 of February 1999. 

On the morning of January 11 th , a fire broke out in this plant that produces cloth fastening 

systems . It is located in an old metal and brickwork building , the main building is approx. 

1500 m² large, with several smaller buildings attached. Total site area is 5600 m². The 

production process requires some use of flammable liquids , which are stored in a separate 

room with CO2 protection. The plant has a partial fire detection system in critical areas, linked 

to a central station, and has a first intervention team . Local fire brigade is part time staffed 

with a complement of volunteers. 

The fire probably started by a fault of an electric heating system in a galvanizing unit with 

polypropylene tanks. This area did not have fire detection, but the fire brigade was called by 

the central station after an automatic fire alarm from an other area. Although there was much 

smoke, the fire brigade could locate the seat of the fire from the information given by the staff 

The fire was extinguished with foam. One production line was destroyed , but production could 

be resumed on the other line after one day. 

A FRAME calculation with these data was made and gives the following result : R= 1.70, R1= 

0.77, R2=1.08. This means a 50 % loss estimate, which is more than was the case, but the 

ANPI report indicated that the fire brigade action was very effective due to the precise 

information given by the management on the seat of the fire. ANPI recommended to extent 

the fire detection system to the whole plant, which would result in an improvement of the risk 

23

F.R.A.M.E. 

factors to R=0.97 R1=0.43 and R2=0.69. The FRAME calculation is in line with this expert 

advice. 


Type of occupancy unifast, braine le comte fire on 11.01.1999 






Length l 100 

Width b 56 g=1.98 

Level platforms E 0.20 e=1.05 



Access direction limited access Z 1 




Main: metal work 

Heating : ok 

Electrical: no regular 

checks 

Explosion: no 

Secondary: flammables 

a 0 

0 

0 



Exits X 5 



0 

0.1 

a=0.10 

t=0.04 







limited training, nearest 

station at less than 10 

min 

N = 0.90 

Special Protection part time fire brigade S= 1.34 


Walls: 60 

Ceiling:60 

Partitions:60 

F=1.56 

Escape protection max. 300 persons, sub 

U=3.56 

compartments 

Salvage 

partial detection, data 

Y=2.28 

protected 


Conclusion: 

adequate protection for staff and business, 50 % damage to be 

expected on property. 

24

F.R.A.M.E. 

Case study 14. Hartford (CT) nursing home fire February 26 , 2003. 

NFPA fire investigations made a report on a nursing home fire on February 26, 2003 in 

Hartford (CT) with 16 fatalities. A summary of the report can be downloaded at 

http://www.nfpa.org/assets/files//PDF/FIHartfordSummary.pdf and the full report is available 

to NFPA members. 

A fire that broke out in the early morning hours of February 26, 2003 in a patient room at a 

nursing home in Hartford, CT resulted in sixteen fatalities and dozens of injuries. At the time of 

the fire there were 148 patients being cared for at the facility. The fire damaged several 

patients’ rooms and a wing of the facility. 

A master box alarm notified the Hartford Fire Department of the fire at the complex at 2:40 

a.m. A first assignment arrived within four minutes and found a fire in a patient room in one 

wing of the building with staff removing patients from that wing and other affected areas of 

the facility. 

Fire fighters and the facility staff were faced with not only a growing fire but with heavy 

volumes of smoke and numerous non-ambulatory patients exposed to the smoke and heat. 

Dozens of ambulances responded to the scene to transport patients to hospitals and other 

facilities. The fire was declared under control at 3:30 a.m. 

When the fire was extinguished the initial death toll was ten patients. 6 other victims died 

from injuries and smoke inhalation in the following days. Investigators determined that the fire 

was caused when a patient ignited bedding on her bed in the room of fire origin, with a lighter. 

The data found in this report were used to make a FRAME calculation. 


Type of occupancy Hartford nursing home on February 26, 2003 



Temperature rise immobilised residents T 100 


Reaction to fire limited M 1 i=1.05 

Length l 180 

Width b 27 g=1.41 

Level E 0.0 e=1.00 







Occupants 

Main: nursing home 

Heating : ok 


checks 

Explosion: no 

Secondary: 

uncontrollable smokers 

Number: 486 (code 

density) 

Mobility factor: 

a 0 

0 

0.1 

not 

mobile, 

limited 

perception 

of risk 

Exits X 10 



0 

0.1 

p =10 

a=0.20 

t=0.82 

25

F.R.A.M.E. 






Normal protection limited training, nearest 

N = 0.90 

station at less than 10 

min 

Special Protection part time fire brigade S= 1.34 


F=1.28 

Walls: 30 

Ceiling:30 

Partitions:30 

Escape protection sub compartments, 

U=1.80 

adequate signalisation 

Salvage sub compartments Y=1.10 


Conclusion: 

A catastrophe could be expected. 

Notes. 

1. The report indicates a capacity of 148 residents. With staff and visitors an occupancy load of 

350 persons could be estimated, below the 486 based on code occupant load numbers. A lower 

number of occupants would still give an unacceptable high risk level. 

2. The smoke detectors in corridors were not accounted for as those are not the critical areas 

(the rooms should be considered as critical) . 

3. New nursing homes of this type would require sprinkler protection according to NFPA 101. 

Adding sprinkler protection would reduce the risk levels to : R=1.10 , R1=1.40 and R2=0.66 . 

This means that FRAME does not consider adding sprinklers as adequate enough. Additionally 

automatic smoke detection would be required to get an acceptable level of safety for the 

occupants. This is because of the low mobility of the residents which requires a fast detection 

and reaction. 

On the other hand, if the 4 wings of the building were properly separated by compartment 

walls and fire doors, whereby the number of people at risk would be below 100, a smoke 

detection system would provide adequate protection according to FRAME. 

Case study 15. A clue for arson : Office in Bristol (UK), June 1 st , 1989. 

Reference : Fire Prevention 232, September 1990 

FRAME calculation : 


Type of occupancy Bristol computer suite 1989 



Temperature rise computers T 100 



Length l 100 

Width b 30 g=0.74 



Ventilation k 0.5 % v=1.02 


26

F.R.A.M.E. 




Main: computer room 

Heating : air conditioned 


Explosion: no 

Secondary: no 

a 0 

0.1 

0 

0 

0 

a=0.10 

Occupants 

Number: 300 (daytime) 


Exits X 6 



t=0.10 








limited training brigade at 

10 min 

automatic detection, 

daytime staffed station 


Walls: 60 

Ceiling:60 

Partitions:30 

Escape protection detection, max. 300, 

protected stairs, 

signalisation 

N = 0.90 

S= 1.98 

F=1.50 

U=3.23 

Salvage 

halon system for 

Y=1.34 

computers 


Conclusion: 

adequately protected for people and property 

The value of R =0.97 would indicate a probable loss of 10 % . The difference with the real loss 

gives a clue that arson could be the cause of the fire. 

Case study 16. A clue for arson : Europneu Nivelles (BE) December 20, 2001. 

Reference : ANPI Magazine n°157. 

The building of the Europneu car tyre supplier in Nivelles was completely destroyed by a fierce 

fire on December 20, 2001. The inquiry discovered several origins of fire, which clearly 

indicated that this fire was caused by arson. Unfortunately, the company went bankrupt two 

years later. 


Type of occupancy Europneu Nivelles, Belgium : tyre storage 






Length l 42 

Width b 20 g=0.74 




Access direction one side access Z 1 

27

F.R.A.M.E. 




Main: tyre trade 

Heating : gas in room 


Explosion: no 

Secondary: no 

a 0 

0.2 

0 

0 

0 

a=0.20 



outside 

working 

hours 

Exits X 2 



t=0.03 






Normal protection no alarm, limited training 

N = 0.66 


Special Protection daytime staffed station S= 1.28 


Walls: 60 

Ceiling:30 

Partitions:0 

F=1.46 

Escape protection none U=1.12 

Salvage not verified Y=1.00 


Conclusion: 

the lack of an alarm system increases the risk for the occupants. 

The value of R =1.30 would indicate a probable loss of 20 % . The difference with the real loss 

(100%) gives a clue that arson could be the cause of the fire. 

Other fire cases evaluated with FRAME v.2 . 

Description R R1 R2 

On November 28, 1998 a fire occurred in a dancing in 

Gothenburg , Sweden, killing 63 and wounding 213 

teenagers. It was caused by arson committed by a youth 

from the same community. The fire was started in the 

basement of an overcrowded discotheque and made, due to 

unfortunate circumstances, devastating progress. (see 

http://www.nfpa.org/assets/files/PDF/gothenburg.pdf ). 

1998 0.55 2.46 0.45 

The FRAME calculation was made with the data found in 

various reports on this fire. A first calculation with 400 

occupants and a single exit resulted in negative values for A, 

indicating a totally unacceptable situation. A second 

calculation was made with an occupant load of 150 persons , 

which is the maximum capacity for the single available exit . 

The values obtained were : 

P= 0.67 P1= 1.66 and P2 = 0.61 

A= 1.25 A1=0.85 A2=1.40. 

W= 1.00 N= 0.49 S= 1.34 F=1.49 U= 1.63. 

The incendiary nature of the fire explains the discrepancy 

between the risk value for the property and the real damage 

caused by the fire. 

However, the total lack of normal protection, alarm devices 

28

F.R.A.M.E. 

and exit capacity still results in a unacceptable high level of 

risk for the occupants. 

29

F.R.A.M.E. 

FRAME version 2. Renovation projects. 

The following FRAME calculations were made to assess the fire risk in building renovation 

projects. The purpose was to verify if the buildings had an adequate level of safety and what 

can be recommended to improve it when necessary. 

Case study 17. A tower building, 5 th floor with HVAC installation 


Type of occupancy Tower building, technical floor 2003 






Length l 38 

Width b 14.4 g=0.57 



Ventilation no ventilation k 0.0 v=1.12 


Height difference H 17.8 z=1.05 



Main: HVAC room 

Heating in room, gas 


Explosion: no 


a 0 

0.2 

0 

0 

0 



(at that 

floor) 

Exits X 2 



a=0.2 

t=0.08 

c=0.2 

Content factor: 

1.5 M. euro 

long term equipment 


dependency factor high d d= 0.55 





fire brigade 15 min to get 

on that floor 


station, fire detection, 

highly reliable water 

supply 


Walls: 120 

Ceiling: 120 

Partitions:120 

N = 0.90 

S= 4.54 

f= 120 F=1.86 

Escape protection detection, max. 300 U=3.07 

Salvage none y=0 Y=1.00 


Conclusion: 

high level of protection, no improvements required 

30

F.R.A.M.E. 

FRAME version 2. Fire protection design. 

The following FRAME calculations were made to select a suitable fire protection concept for 

various types of buildings. 

Case study 18. Bulk storage of plastic pellets in PVC big bags 

This warehouse complex is composed of a number of adjoining warehouses, separated by fire 

walls, so that some of these compartments have only outside access from one side. The 

warehouse owner stores “big bags” with plastic pellets produced in a nearby plant, and ships 

them to the users. The content of the warehouse does not belong to the warehouse owner but 

to his clients. Because of the low value of building and content , investing in sprinkler 

protection does not give a fast return, so an alternative protection was sought , while 

accepting a slightly larger property risk and corresponding insurance premium0 

The proposal was to subdivide the complex in compartments of max. 13 000 m², to equip the 

warehouses with smoke detection and smoke vents and to install a dedicated water supply for 

fire fighting on the site. 

The first FRAME calculation was made for a 10 800 m² compartment, which was only 

accessible from one side. 


Type of occupancy Bulk storage of plastic pellets in PVC big bags 2002 




Average dimension m 1 


Length l 150 

Width b 72 g=2.66 



Ventilation 

210 m² smoke vents, k 0.0361 v=0.90 

540 m² windows 

Access direction to long façade only Z 1 

Height difference H z=1.15 



Main: storage 

No heating 


Explosion: no 


a 0 

0 

0 

0 

0 



Exits X 12 



t=0.05 



dependency factor low d d= 0.10 


Water supplies limited quantity W=0.81 

Normal protection fire brigade at 10 -15 min N = 0.90 


volunteer brigade, fire 

detection, water tank on 

a=0 

S= 3.23 

31

F.R.A.M.E. 

site 


Walls: 60 

Ceiling: 15 

Partitions:0 

Escape protection Detection, max. 300, 

smoke vents actuated 

mixed 

concrete 

steel 

F=1.24 

U=3.39 

Salvage relocation, data protected Y=2.08 


Conclusion: 

high level of protection for people and business, higher risk for 

property acceptable 

A second calculation was made for a 13 000 m² warehouse that was accessible on two sides . 

The calculated risks were R= 1.11 R1=0.52 and R2 =0.37 , which shows a comparable risk 

level. 

A third calculation was made for an undivided warehouse of 21 600 m² with sprinkler 

protection instead of smoke detection. The calculated risks were R= 0.98 , R1=0.89 and R2= 

0.29 . This gives a better risk, but the cost of the sprinklers was much higher than that of 

improving the quality of the existing separations to a firewall type construction. 

A fourth calculation was made for a compartment of 4000 m² size, were the plastic pellets 

were bagged in PVC bags. 

Potential risks were lower as the fire load and the area were smaller : 

P = 2.41 P1= 2.02 P2 = 1.46 

The values of the acceptable risks indicated a higher exposure , because of the activity : 

A = 1.25 A1 = 1.05 A2 = 1.20 

The protection degrees were basically the same as the concept was unchanged. 

The balance of all elements resulted in the following risk levels : 

R= 0.65 R1= 0.73 R2 = 0.36 

The concept was considered as adequate for the plant. 

Case study 19. Fire safety of a historical building. 

The manager of a 13-15 th century monastery, used as museum and cultural centre wanted to 

know what the level of fire safety was of this building. 

Construction : 2250 m² in a U-shape, 3 levels, very thick stone and masonry walls, wooden 

floors, some with tiles, slate roof on massive oak structure. No effective compartimentation 

because of two monumental stairs joining the three levels. 

The building is accessible on one side only, the other sides are adjacent buildings and a river; 

the city water supply is only a 3" pipe; The building is equipped with extinguishers, hose reels, 

a fire alarm system and partial fire detection. Notification to the fire brigade is by the staff. 

The fire load is low, except for a library with old books, located in the North wing; this library 

has a wooden floor and a decorated wooden ceiling, which is also the floor of meeting room at 

the upper level. This room can receive about 150 persons and has one adequate and one 

limited exit to stairs. 

In the central wing, there is a small restaurant and a kitchen. 

32

F.R.A.M.E. 

Maximum occupancy is 500 persons for meetings or conferences. The building has several 

exits, but some exit doors turn to the inside. 

A first FRAME calculation was made for the existing situation: 


Type of occupancy Museum and Cultural Centre 2001 






Length l 150 

Width b 15 g=0.88 





Height difference H -1 z=1.05 






Explosion: no 

Secondary: kitchen 

a 0 

0.1 

0 

0 

0.1 



no clear 

plan 

Exits X 8 



risk of 

panic 

a=0.2 

t=0.59 

c=0.22 


20 M. euro , unique 

building 




Water supplies limited network W=0.86 

Normal protection no training N = 0.81 



station, river nearby 


Walls: 60 

Ceiling: 30 

Partitions: 

S= 1.71 

F=1.70 

Escape protection partial detection, fire 

U=1.63 

brigade 

Salvage relocation possible Y=1.16 


Conclusion: 

Risk too high for persons, mainly due to the absence of a 

warning system and clear evacuation plan. 

As a temporary measure, it was proposed to limit the number of people that could be in the 

building at the same time, i.e. avoid that the full meeting capacity of the building was used. 

A FRAME calculation with 250 persons at the ground level alone, adequate signalisation and no 

catering resulted in acceptable R values of R= 0.53 R1=1.06 and R2 =0.60. 

33

F.R.A.M.E. 

On the long term, an improvement plan was proposed based on a subdivision of the building 

in 3 main compartments. The heavy walls have adequate structural fire resistance, the 

openings can be protected by fireproof glass partitions in the corridors with limited impact on 

the architectural look of the building. Automatic smoke detection should be installed in all 

areas. 

The improvement proposal was supported by more FRAME calculations. 

The central part of the building of 1350 m² ( factor g= 0.81) could still have + 500 persons 

present, but with automatic detection, better evacuation possibilities and good signalisation 

the escape factor u= 3.56 and the resulting risk values are : 

R = 0.46 R1 = 0.93 and R2 =0.43. 

For the northern wing, where the historical library is located, a calculation for a concept with 

fire detection alone indicated that the existing stairs were inadequate for the maximum 

capacity of the meeting rooms in that wing. As it is practically impossible to add an evacuation 

stair to that wing, the seating capacity should be limited to max. 100 persons to achieve the 

R1 < 1 goal. 

As an alternative, it was proposed to create a false floor above the wooden ceiling of the 

library install recessed sprinklers for library through the ceiling . This would provide a sound 

and fire barrier between the library and the meeting rooms above it, and reduce the overall 

fire risk, resulting in risk values of R=0.25 R1=0.90 R2= 0.17. 

It was necessary to convince the museum management that in case of a fire , the water 

damage to the library by sprinklers would be less than that caused by the fire brigade who 

would eventually have to use river water to fight a fire from the outside. 

Unfortunately, it was not clear whether sufficient funds would be available to realize such an 

improvement programme. 

Case study 20. Fire safety for an existing school building. 

The management of this school was worried about the fire safety in their school, after a fire 

service inspection that pointed out several weak points compared with modern fire safety 

requirements . The school occupies a 19 th century built front building, but later wings were 

added in 1970s. The financial means of the school for improvements were limited, any major 

building work takes at least six years before the plans are approved and funds obtained. 

A proposal was made to evaluate the fire safety with FRAME and to find out which 

improvements could be made with a limited budget on a short time. Safety of the users was 

the prime concern, any improvement to provide better protection for the building was of lesser 

concern. 

A survey of the building was made and showed that the fire load was low and no hazardous 

materials were found in the classrooms. The school was well also organised. Fire exit drills 

were regularly made and documented. 

The FRAME calculation for the existing situation gave the following result : 

34

F.R.A.M.E. 


Type of occupancy Existing school 2006 

Fire load immobile wood panels Qi 100 





Length l 130 

Width b 18.5 g=1.09 

Level ground floor +2 

E 1.60 e=1.32 

communicating levels 


Ventilation 

ample single glass k 0.0347 v=0.79 

windows 





Main: school 

central heating 


Explosion: no 

Secondary: smokers 

a 0 

0 

0 

0 

0.1 



Exits X 8 

Exit directions all exits needed k 1 


a=0.1 

t=0.47 





Water supplies to few hydrants W=0.95 


no manual alarm boxes, 

not enough 

extinguishers, no hose 

reels, fire brigade at 10- 

15 min. 

N = 0.54 



Walls: 60 

Ceiling: 30 

Partitions:0 

mixed 

construction 

F=1.46 

Escape protection fire brigade, drills U=1.63 

Salvage activity can be relocated Y=1.80 


Conclusion: 

Too high risk for people. It should be noted that the exit time for 

the drills is much less than estimated by factor t. 

The recommendation was to install fire extinguishers, an automatic addressable fire detection 

and alarm system and to provide smoke curtains between the corridors and the staircases. 

These improvements can be realised in a short time and do not need any major investment or 

disturbance of the school operations. 

35

F.R.A.M.E. 

These improvements result in higher values for the normal protection ( N=0.66) , for the 

special protection ( S=1.63 and F = 1.45) and for the escape protection (U = 2.65) . The 

calculated risk levels are R =0.62 , R1=0.81 and R2 =0.44 . This means that an adequate level 

of safety can be obtained for all 3 aspects of the fire risk assessment. 

Case study 21. Sweets manufacturer safety improvement program. 

A sweets manufacturer asked fire safety consultant in 2004 for a fire risk analysis and 

improvement program for his factory. The consultant used FRAME to support his findings and 

recommendations. A total of 7 FRAME calculations were made for this job. 

The plant has his own water storage tank for fire fighting with an electrical driven fire pump 

and a hydrant network, but no sprinklers. The nearest fire station ( day time manned) is at 10 

-15 min. from the plant. The plant has a guard service outside working hours and a small first 

intervention team during working hours. 

A first series of calculation was made for the production unit. 

The building is 125 m long and 67 m wide, mainly ground floor but with upper levels for about 

30 % of the ground floor area. Construction is mainly concrete and brickwork, and ceilings are 

at an average of 4 m. The fire load is estimated at 900 MJ/m² . For hygienic reasons, there are 

no openable windows or smoke vents. About 60 persons per shift work in the factory, and 

there are 8 exit doors. The value of the factory is estimated at 5 Million Euro. 

Some part of the production process is heated by steam or hot water and there is a zone with 

an occasional dust hazard without proper dust removal system. 

The 1 st calculation was made for the factory “as is” and gave the following results : R =2.17 , 

R1=1.58 and R2 = 1.91, which indicate an unacceptable high risk level. As Ro, the initial risk 

was = 2.60 , sprinkler protection was recommendable 

A 2 nd calculation was made with sprinklers, and resulted in the following risk levels: R=1.01, 

R1=0.80 and R=0.73 or a well protected risk. 

As the investment in sprinklers was beyond the financial possibilities of the plant, an 

alternative improvement program was proposed: Solve the dust problem by adding a dust 

extraction system and compartimentation of that part of the process and installation of an 

addressable automatic fire detection (rate of rise detectors mostly). This resulted in the 

following risk levels R=1.31 R1=0.82 and R2=0.96 

This would lead to an acceptable level of risk for the personnel and the business and also 

reduce the property loss expectancy to about 20 % for an affordable investment. 

The next calculations were made for the finished good warehouse. 

This was a 82.5 m x 32.5 m large steel building, 11 m high with a rack storage of palletised 

goods. The mobile fire load was estimated at 12000 MJ/m². It was accessible from 3 sides. 

No smoke venting was provided. Warehouse value was estimated at 1 million Euro. 

The building has no heating, the electrical installation was code compliant, handling operations 

were done with electrical lift trucks with a battery charger station outside the warehouse. 

The impact of the warehouse on the business is limited: warehouse capacity is available at 

other locations too. 

36

F.R.A.M.E. 

The FRAME calculation for the warehouse “as-is” gave the following results : R = 2.22 R1=1.00 

and R=0.66. The conclusion was that the protection level was adequate for the personnel and 

the activity, but that a total loss could be expected in the event of a larger fire. 

The next calculation was made for the addition of a manually operated smoke venting system, 

as often recommended by the local fire service. The effect was minimal : R= 2.06 R1= 0.93 

R2= 0.63. The main reason is that late detection and arrival of the fire brigade is still possible. 

The next step was to foresee automatic fire detection (beam detectors) linked to a monitoring 

station and smoke venting. This would reduce the risk level to R= 1.29 R1=0.57 and R2 = 

0.38. This would reduce the property loss expectancy to about 20 % of the warehouse for a 

moderate investment. 

Finally a risk calculation was made for a sprinkler protected warehouse: This would reduce the 

risk level to R= 0.95, R1=0.59 and R2 =0.26. This is a well protected risk but requires a 

large investment in sprinklers. 

As the plants’ financial capability was limited, the moderate investment in fire detection and 

the dust collection system was approved, even if the property loss expectancy remained on the 

high side. 

Case study 22. Specialty chemicals unit safety improvement program. 

A manufacturer of chemicals has on a large site a small unit where a specific type of chemicals 

is made with a booming success. As long term contracts were made with the customers, the 

safety officer wanted to use FRAME to check if this unit was properly protected. 

The risk analysis was made with 3 FRAME calculations. 

The 1 st calculation was made with the production unit “as is” . The building was 67.5 m x 31 m 

large and had a mixed concrete and laminated wood structure, with some 20 % platforms 

above floor level. Average height was 4.2 m and the plant had a large ventilation system 

(175.000 m³/h) . 

Structural fire resistance is 60 min for all main elements and for internal partitions. Internal 

communications lack fire doors. Part of the process uses flammable liquids in a closed circuit 

and therefore some areas were zone 2 classified . Process heating was indirect, a boiler was 

located in a separate compartment. The unit contained valuable special equipment, and the 

value was 28 million Euro. Added value ratio for the unit is estimated at 75 % of the turnover. 

Any fire would result in a long standstill for repairs, clean-up and re-qualification for 

high purity standards. A delay of more than 1.5 month in production would result in an 

interruption of supplies to the clients. 

The plant had a large and dedicated fire water supply system, all staff is trained in first 

intervention on fires, and there is also a small plant fire brigade during working hours. The 

nearest station is full staffed and at 10-15 min distance. The unit itself is equipped with 

automatic fire detection, extinguishers and hose reels. 

The first risk calculation resulted in the following : R= 0.49 , R1=0.55 and R2=1.41, indicating 

that the property and people protection was all right, but that the business interruption 

concern was justified. 

As the result of this analysis a risk improvement program was established: it included the 

addition of fire doors in the communications between the production and the purification area, 

creating two separate compartments, and the elaboration of a contingency plan to be able to 

obtain critical equipment and repairs on a very short notice. The main result of these 

improvements was that a fire could no longer endanger the whole production process and that 

37

F.R.A.M.E. 

the production interruption time could be brought back to less than the 1.5 month critical 

period. 

As compartimentation was part of the plan, two new FRAME calculations were necessary to 

check the impact of the risk improvement program. 

For the purification compartment (42.5 m x 30.5 m ) the results were : R=0.32, R1=0.43 R2= 

0.66 and for the synthesis unit ( 35 m x 23 m) the results were : R= 0.31 , R1=0.67 and 

R2=0.82 . 

These calculation results illustrate the effect of compartimentation on the risk level. 

Case study 23. University Library fire safety concept. 

Students of a fire engineering course checked the FRAME method with an analysis of a modern 

library building on their university campus. The buildings’ fire safety concept was developed 

with the input of the authorities and experienced engineers, so a “good” level of risk can be 

expected. 


Type of occupancy University Library and coffee shop 2006 

Fire load immobile wood panel decoration Qi 300 

Fire load mobile library Qm 2300 q=1.69 


Average dimension book shelves m 1 


Length l 60 

Width b 50 g=1.61 

Level main floor E 0 e=1.00 


Ventilation smoke vents 10 m² k 0.31 % v=1.09 


Height difference below surrounding land H -1.5 z=1.05 




Gas heating; 


Explosion: no 


a 0 

0.1 

0 

0 

0 



Exits X 20 



a=0.1 

t=0.08 

c=0.14 


10 M. euro , difficult to 

replace 







limited training, brigade 

at

F.R.A.M.E. 


F=1.49 

Walls: 60 

Ceiling: 60 

Partitions:60 

Escape protection automatic fire detection, 

U=3.56 

protected staircase, full 

time staffed brigade 

Salvage protected financial data, Y=1.10 


Conclusion: 

Acceptable risk level for all three aspects. 

39

F.R.A.M.E. 

FRAME version 2. Inspection and verifications. 

The following FRAME calculations were made for risk assessments of existing situations. 

Case study 24. Temporary theatre installation. 

A podium was built in an old shed for a summer theatre festival. The question was if the 

arrangement was safe enough. 


Type of occupancy Temporary theatre installation 2002 






Length l 50 

Width b 32 g=1.09 



Ventilation 60 m² windows k 0.0125 v=0.91 





Main: theatre 

No heating 


Explosion: no 

Secondary: smokers 

a 0 

0 

0.1 

0 

0.1 


Panic 


Exits X 20 



a=0.2 

t=0.09 







fire brigade at site, no 

hose reels , limited 

training 

N = 0.74 



Walls: 15 

Ceiling: 30 

Partitions:0 

mixed 

concrete 

steel 

F=1.35 

Escape protection max. 300, U=1.80 

Salvage none Y=1.00 


Conclusion: 

peoples’ safety only concern: acceptable risk 

40

F.R.A.M.E. 

Case study 25. High rack storage . 

Sprinklered high rack storage. 


Type of occupancy Sprinklered high rack storage 2002 





Reaction to fire M 2.00 i= 0.98 

Length l 148 

Width b 66 g=2.49 



Ventilation k 2% v=0.98 





Main: storage 

No heating 


Explosion: no 


a 0 

0 

0 

0 

0 



Exits X 2 



t=0.05 

Content factor: 10 M. euro , large stock c=0.1 




Water supplies limited quantity W=0.81 




brigade at +10 min. 

sprinklers with dedicated 

water supply , daytime 

staffed station 


Walls: 15 

Ceiling: 15 

Partitions:0 

mixed 

concrete 

steel 

a=0 

N = 0.81 

S= 4.54 

F=1.23 

Escape protection max. 300, sprinklers U=2.18 

Salvage 

data protected, repairs 

Y=1.48 

possible 


Conclusion: 

adequate protection for all 3 aspects. 

41

F.R.A.M.E. 

Case study 26. Fire safety for a drink manufacturing plant. 

The factory of this alcoholic drink manufacturer developed from a small distillery into a large 

mixing and bottling unit. During the development process modifications and extensions were 

numerous, so that the plant needed a thorough fire safety improvement program. The insurers 

practically imposed sprinklers for the whole plant, and the question was raised if this was 

enough to guaranteed also a level of fire safety that was acceptable for the authorities and 

that guaranteed also business continuity. 

A number of FRAME calculations were made to check the new situation. 


Type of occupancy Alcohol storage in (metal) tanks 2003 



Temperature rise alcohol T 50 

Average dimension tanks m 1 

Reaction to fire most metallic M 1 i=1.05 

Length l 85 

Width b 80 g=2.44 



Ventilation k 1 % v=1.02 





Main: storage 



control 

Explosion: zone 2 

Secondary: battery 

chargers 

a 0 

0.1 

0.1 



Exits X 6 



0.1 

0.1 

a=0.4 

t=0.04 

Content factor: 6 M. euro , c=0.0 








at +10 min 

sprinkler protection, 

dedicated water supply, 



Walls: 30 

Ceiling: 0 

Partitions:0 

steel 

structure, 

brick 

walls 

N = 0.81 

S= 5.00 

F=1.10 

Escape protection sprinklers, fire brigade, 

U=3.56 

max. 300 persons 

Salvage protected data, more Y=1.34 

42

F.R.A.M.E. 

than one unit 


Conclusion: 


A second compartment in the plant was the distillery and bottling unit. 


Type of occupancy Distillery and bottling unit 2003 



Temperature rise alcohol T 50 

Average dimension tanks m 1 

Reaction to fire most metallic M 1 i=1.05 

Length l 60 

Width b 20 g=2.44 

Level gallery level E 0.5 e=1.12 







Main: industry of 

flammables 

Heating; steam, 

separated boiler room 




a 0.4 



Exits X 2 



0 

0 

0.1 

a=0.5 

t=0.08 

Content factor: 6 M. euro , c=0.0 


dependency factor medium d d= 0.45 






at +10 min 





Walls: 60 

Ceiling: 60 

Partitions:60 


sprinklers, fire brigade, 

max. 300 persons, sub 

compartments 

N = 0.81 

S= 5.00 

F=1.41 

U=3.92 

Salvage 

protected data, sub 

Y=1.22 

compartments 


Conclusion: 


43

F.R.A.M.E. 

The second calculation shows that this older part of the plant, with a fire resistive construction, 

but less accessible, has quite different characteristics. 

A third and newer unit, on an other location, is used to produce cider from apple juice. This a 

basically a wet process in a 2000 m² large building with smoke vents and automatic fire 

detection, but no sprinklers. Based on the initial risk value of Ro= 0.48 , manual intervention 

means would have provided an adequate level of fire protection, but the detection and smoke 

vents were requested by the fire brigade. This is the standard requirement of the local fire 

brigade for industrial premises. FRAME was not used at the design stage and the values of the 

risk calculation (afterwards) are R=0.26 R1=0.25 and R=0.26, indicating a difficult to justify 

overdesign. 

Other risk verifications by FRAME v.2. 


A FRAME user submitted the following calculation for the 

Canteen and kitchen of his company. Company policy is to 

apply the most strict regulations as well as overall sprinkler 

protection. The plant has a dedicated fire water supply 

system, and there is plant fire brigade on the site. 

2003 0.09 0.24 0.04 

The compartment was 1100 m² large, 3.5 m high, concrete 

construction. Fire load was estimated at 600 MJ/m². The 

kitchen has an extraction system of 15000 m³/h and works 

with gas heated equipment and an extinguishing system in 

the kitchen. 550 persons can be present in the 

compartment, which has 3 exits with 6 exit units. 

Comment : The very low risk levels indicate an “oversized” 

fire protection concept. 

Airport buildings. 

The safety department of a large airport asked a specialised 

fire safety consultant to verify the level of fire safety in all its 

buildings. FRAME was used to detect rapidly the areas of 

concern. When the FRAME calculation showed a risk level 

above 1, some improvements were proposed. Any 

combination which would bring the risk levels below 1 was 

acceptable. 

2001 0.40 1.14 0.13 

A first calculation was made for a departure hall of 200 m x 

70 m and 8 m high, with a maximum occupant load of 3800 

persons, based on the actual use of the airport. 40 exit units 

were available, structural fire resistance was 120 min. and 30 

min for roof and facades. The hall was equipped with an 

addressable fire detection system and sprinklers. The airport 

has his own full staffed fire brigade. 

The first calculation gave R=0.40 R1=1.14 and R2=0.13 

A second calculation was made with the addition of a smoke 

venting system linked to the fire detection system., which 

reduced the risk levels to R= 0.34 R1=0.96 and R2=0.11. 

As expected for such a high hall, the impact of smoke venting 

is limited, but it brings the risk level down to below 1 with 

limited modifications to the building. 

+ 

smoke 

vents 

0.34 0.96 0.11 

44

F.R.A.M.E. 

A third calculation was made with the addition of 10 extra 

exit units .This brought the risk levels to R = 0.38, R1 = 0.98 

and R2 = 0.13 . This would give also give an acceptable 

level. 

A simulation of the evacuation indicated that the FRAME 

formula for t made a very conservative evaluation of the 

Required Safe Exit Time. With this in mind, the R1

F.R.A.M.E. 

FRAME version 2. Equivalency of concepts. 

The following FRAME calculations were made to verify the equivalency of alternate fire 

protection concepts. In most of these cases, the prescriptive requirements of an applicable 

code were not (entirely) followed. An alternative that suits more the fire safety goals of the 

owner / user and that considers specific building characteristics was proposed. Two 

calculations were made, one for the prescribed concept, the other for the alternative. The 

alternative was considered to be equivalent when the calculated risk levels of the alternative 

are equal or lower than the other. 

Case study 27. General Chemicals warehouse. 

For this warehouse, the code requirements foresee a fire resistive construction, where the 

owner and the insurer preferred a sprinkler protection. As the Ro value was =2.19 , the 

recommended protection per FRAME was sprinklers. 


Type of occupancy general chemicals warehouse 2000 






Length l 110 

Width b 46 g=1.77 



Ventilation 

windows + 1 % smoke k 2.7 % v=0.95 

vents 





Main: storage 

No heating 


Explosion: if leaks occur 


a 0 

0 

0 

0.1 

0 



Exits X 10 



a=0.1 

t=0.05 






Normal protection fire brigade at +10 min. N = 0.90 


sprinklers with dedicated 

water supply , volunteer 

brigade 


Walls: 15 

Ceiling: 15 

Partitions:0 

steel 

S= 5.25 

F=1.01 

Escape protection max. 300, sprinklers U=2.41 

46

F.R.A.M.E. 

Salvage 

data protected, repairs 

Y=1.28 

possible 


Conclusion: 

adequate protection for all 3 aspects. 

The code requirements foresee no sprinklers, but a general requirement of R30 construction 

and R60 separation of the rooms with flammable products. This option would result in S = 

1.55, F =1.31 and U = 1.80. The risk values are then R=1.97 , R1 =1.06 and R2 = 1.04. 

These values show clearly that code requirements are not aimed at property protection. The 

alternative gives much better safety levels and was accepted as equivalent. 

Case study 28. Medical Education Centre. 

This medical education centre was to be installed in an existing building. The local authorities 

recommended to modify the building in such a way that it would comply with the building 

requirements for new buildings, which implied fire proofing of the exiting steel structure. 

As an alternative it was proposed to install automatic fire detection, which would cost less and 

respect the art nouveau style interior. 

The risk level for the code complying situation was calculated and gave the following results: 

R= 0.54 R1= 0.77 and R2 =0.77. 


Type of occupancy medical education centre alternative 






Length l 36 

Width b 23 g=0.78 

Level E 1.60 e=1.32 


Ventilation windows k 1.45 % v=0.91 





Main: office 

a 0 

a=0 

heating: ok 


Explosion: none 


0 

0 

0 

0 



Exits X 3 



t=0.07 





Water supplies OK W=1.00 

Normal protection fire brigade at +10 min., 

no training 

N = 0.74 

47

F.R.A.M.E. 

Special Protection addressable automatic 

S= 2.41 

fire detection 


steel 

F=1.06 

Walls: 0 

Ceiling: 15 

Partitions:0 

Escape protection max. 300, detection, 

U=3.46 

protected staircase 

Salvage none Y=1.00 


Conclusion: 

The risk levels were lower than the code complying concept. 

There is equivalency. 

Other equivalent concept calculations. 

The next table gives an overview of other similar calculations that have been made to prove 

the equivalency of concepts with FRAME. 


4500 m² warehouse for combustible liquids, 7.5 m high , 

concrete construction . Fire load was estimated at 10950 

MJ/m². The building as +/- 1 % smoke vents, part of it is 

“zone 2 classified”, and public water supply is limited. There 

is a pre-action sprinkler system actuated by heat detectors. 

No internal separations exist. 

The code requirements (at the time the building was built) 

2005 

per 

1.02 

4.28 

0.87 

1.42 

0.45 

1.76 

are to have R30 fire resistive construction and sub 

compartments for the flammable liquids. 

Note that these code requirements are far below the FRAME 

recommended levels. 

code 

A 19 th century post office, which was externally protected as 2002 0.21 0.39 0.22 

historical building found a new destination as a civil court. 

The building was only 500 m² large, but had 20 % 

mezzanines, as the architect opened the upper floors of the 

building to bring more light in it. This did not fit with the 

prescriptive codes that the local authorities applied, unless 

some of the partition walls would be made of fire resistive 

glass. This would be very expensive and spoil the architects’ 

concept. 

As an alternative, sprinkler protection was proposed for the 

building, as well as automatic fire detection. 

The FRAME calculation indicated that installing both 

sprinklers and fire detection was overdone, but the project 

manager maintained both to avoid further discussions with 

the local fire authorities. 

A new housing project combined two 4 level apartment 

buildings of each 400 m² floor area with a covered internal 

court yard. The prescriptive code defines this arrangement 

as an atrium, that has to be fire separated from the adjacent 

buildings and the exits are not allowed to go through this 

atrium. Each level of each apartment block would then be a 

separate compartment. The code complying risk calculation 

considered only 36 persons to be evacuated per level. 

2004 0.54 0.89 0.55 

48

F.R.A.M.E. 

The proposal was to equip all the apartments with an 

automatic fire detection and alarm system, to install 1% 

smoke vents linked to the fire detection in the court yard, to 

link both buildings by a passage at each level so that every 

apartment would have two paths downwards. The FRAME 

calculation gave: 

A radiation calculation was also made to check that the 

opposite passage could be used safely. 

The alternative was accepted by the authorities. 

variant 0.36 0.87 0..36 

49

F.R.A.M.E. 

Non-industrial risks with FRAME v1. 

These calculations were made according to the first version of FRAME. 

Case study 29. Small clinic, 2 nd floor. 


Type of occupancy small clinic Belgium, 1988 






Length l 31 

Width b 27.5 g=0.85 









Heating : ok 


Explosion: no 


a 0 

0 

0.1 

0 

0.1 


Mobility factor: 8, people 

requiring help 

Exits X 2 


a=0.2 

t=0.18 







no alarm transmission, 

no hose reels, limited 

training 

N = 0.60 



Walls: 60 

Ceiling:60 

Partitions:60 

f= 60 F=1.56 

Escape protection Subcompartments, U=1.80 

Salvage business share 100 % y=5 Y=1.28 


Conclusion: 

just acceptable 

Recommendation: improve alarm system with automatic transmission to fire brigade 

50

F.R.A.M.E. 

Case study 30. Company canteen and restaurant, 6 th floor. 


Type of occupancy company canteen and restaurant, 1993. 






Length l 50 

Width b 12 g=0.54 









Heating : gas 


Explosion: no 

Secondary: kitchen 

a 0 

0.1 

0 

0 

0.1 


Mobility factor:1 

Exits X 2 


a=0.2 

t=0.03 





Water supplies limited pressure W=0.90 


Special Protection fulltime staffed station, S= 1.89 


Walls: 0 (glass) 

Ceiling:60 

Partitions:30 

f= 41 F=1.37 

Escape protection automatic detection in 

U=3.39 

kitchen 



Conclusion: 


51

F.R.A.M.E. 

Case study 31. Hotel inspection report. 

New hotel, fire resistive construction, automatic fire detection. 


Type of occupancy modern hotel, ground floor 1992 Belgium 






Length l 45 

Width b 20 g=0.76 




Access direction single access direction Z 1 





a 0 

a=0.15 

Heating : gas 


Explosion: no 


0.1 

0 

0 

0.05 



Exits X 3 


t=0.04 



dependency factor high added value d d= 0.50 



limited hydrant 

W=0.95 

connections 

Normal protection limited training – no 

N = 0.81 

extinguishers 

Special Protection addressable fire 

S= 2.65 

detection, professional 

fire brigade 


f= 60 F=1.51 

Walls: 60 

Ceiling:60 

Partitions:60 

Escape protection Subcompartments, 

U=6.08 

detection 



Conclusion: 

highly protected risk 

52

F.R.A.M.E. 

Other non-industrial risks. 

The next table gives an overview of a number of other non-industrial buildings that have been 

evaluated with FRAME. 

Description Year R R1 R2 

Hotel technical floor: 1500 m² 4th floor (at 13 m height) of 1991 0.46 0.33 0.58 

a new hotel with the heating installation. The room is well 

ventilated , access is limited to 2 sides, protected by 

automatic fire detection, extinguishers and hose reels. Fire 

brigade at 5 min. distance : adequate protection 

Small hotel in older building. 360 m², 3 levels, fire resistive 1992 0.29 0.53 0.59 

structure, wooden floors, automatic fire detection, 

extinguishers and hose reels, fire brigade at -10 min. 

Adequate protection. Note: local code requires automatic fire 

detection if the building is not fire resistive 

Hotel entrance level, low fire load, 4700 m², concrete 

construction, 4 m high, air conditioned, kitchen as secondary 

activity, limited training, full time station at 5 min. , 

subcompartmentation , ample exits. 

automatic detection recommended in unoccupied areas to 

reduce risk for business interruption. 

1992 1.02 0.54 1.29 

53

F.R.A.M.E. 

Plastics industries with FRAME v.1 

Case study 32. Medical supplies storage 

This medical supplies company has various units, including high rack storage. They were 

looking for a competitive insurance price, based on the high level of fire protection they 

pretended to have. 


Type of occupancy medical supplies storage 1992 Belgium 


Fire load mobile Content Qm 9000 q=2.09 

Temperature rise mainly plastic parts T 100 



Length l 85 

Width b 40 g=1.49 








Main: storage 

a 0 

a=0.0 

Heating :none 


Explosion: not 

Secondary: no 

0 

0 

0 

0 



Exits X 4 


t=0.02 





Water supplies storage limited W=0.90 

Normal protection ok N = 1.00 

Special Protection owner controlled water 

S= 6.39 

supply for sprinkler 

system 


f= 50.6 F=1.29 

Walls: 15 

Ceiling:15 

Partitions:0 

Escape protection (not indicated) 

Salvage business share 10% y=5 Y=1.28 

Calculated Risks: R= 0.36 R2=0.14 

Conclusion: 

Well protected storage facility 

54

F.R.A.M.E. 

Case study 33. Medical gloves manufacturing, storage 


Type of occupancy Medical gloves manufacturing, storage 1992 






Length l 75 

Width b 65 g=2.03 








Main: storage 

Heating :none 



Secondary: no 

a 0 

0 

0 

0 

0 



Exits X 7 


a=0.0 

t=0.03 






W=1.00 


Special Protection sprinkler system S= 4.32 


Walls: 60 

Ceiling:60 

Partitions:30 

f= 56.25 F=1.41 

Escape protection ample exits, sprinklers U=7.39 


Calculated Risks: R= 0.48 R1=0.24 R2=0.14 

Conclusion: 

Well protected facility 

55

F.R.A.M.E. 

Case study 34. Medical supply sterilization plant. 

This plant combined a sterilization unit and two warehouses: one for incoming goods, one for 

outgoing goods. 


Type of occupancy Medical supply sterilization unit 1994 Belgium 






Length l 70 

Width b 70 g=2.10 








Main: industry 

a 0.2 a=0.4 

Heating : local 


Explosion: yes 


0.1 

0 

0.1 

0 



Exits X 4 


t=0.03 







Special Protection day time staffed station S= 1.34 


Walls: 60 

Ceiling:60 

f= 52.5 F=1.49 

Partitions:0 

Escape protection exits U=3.39 



Conclusion: 

inadequate protection level for property and business. 

The owner of this factory received the recommendation to separate the sterilization unit from 

the rest of the plant by a firewall. This would reduce the size of the sterilization compartment 

from 4900 m² to 1800 m² and lower the average fire load from 2000 MJ/m² to 1500 MJ/m² . 

This modification would reduce the risk levels to R=1.05 , R1=0.59 and R2 = 0.93 . 

56

F.R.A.M.E. 

Case study 35. Medical gloves manufacturing, production unit 


Type of occupancy Medical gloves manufacturing, production 1992 






Length l 105 

Width b 90 g=2.82 








Main: factory 

Heating :gas 




a 0.2 

0.2 

0 

0 

0.05 



Exits X 10 


a=0.45 

t=0.04 





Water supplies 10% lack W=0.99 


Special Protection sprinkler system S= 4.32 


Walls: 60 

Ceiling:60 

Partitions:60 

f= 60 F=1.44 

Escape protection ample exits, sprinklers U=6.70 


Calculated Risks: R= 0.91 R1=0.35 R2=0.75 

Conclusion: 

Adequately protected facility 

57

F.R.A.M.E. 

Case study 36. Bubble foam manufacturer - storage. 

The following case studies are made for several units of a bubble foam manufacturing unit. 


Type of occupancy bubble foam manufacturer - storage 1993 






Length l 48 

Width b 48 g=1.44 




Access direction limited Z 2 




Main: storage 

no heating 


Explosion: ok 


a 0 

0 

0 

0 

0 



Exits X 4 


a=0.0 

t=0.02 









brigade at + 10min 

small fire brigade, 

dedicated water supply 


Walls: 60 

Ceiling:60 

Partitions:0 

N = 0.70 

S= 1.55 

f= 52.5 F=1.48 




Conclusion: 

property risk to improve, consider automatic detection 

The adjacent converting hall has a slighter higher level of risk, and the same recommendation 

to install automatic fire detection was given. The production unit was a sprinkler-protected 

building, and the next FRAME calculation shows that it is a well protected building. 

58

F.R.A.M.E. 

Case study 37. Bubble foam manufacturer – production unit. 


Type of occupancy production unit 1993 




Average dimension film making m 0.01 


Length l 90 

Width b 40 g=1.51 




Access direction limited Z 3 




Main: plastics 

no heating 


Explosion: use of gas 


a 0.4 

0 

0 

0.1 

0 



Exits X 5 


a=0.5 

t=0.03 









brigade at + 10min 

small fire brigade, 

sprinklers, dedicated 

water supply 


Walls: 60 

Ceiling:60 

Partitions:60 

N = 0.70 

S= 4.54 

f= 60 F=1.43 

Escape protection exits , sprinklers U=9.43 



Conclusion: 


59

F.R.A.M.E. 

More plastics industry case studies. 

The next table gives an overview of a number of other plastic industry plants that have been 



PVC profiles production unit : 8800 m² single storey building, 1995 0.54 0.30 0.56 

4 m high, limited ventilation, fire resistive construction, large 

water supply, automatic fire detection, day time staffed 

station, several exits . 

Adequate protection level. 

same plant , other production unit of 12250 m² 1993 1.17 0.27 0.85 

PVC profiles storage hall: 9600 m² single storey building, 4 m 1995 1.12 0.23 0.41 

high, limited ventilation, fire resistive construction, large 

water supply, automatic fire detection , day time staffed 

station , several exits 

Larger hall and higher fire load give higher property risk 

Same plant , other storage of 5625 m² , 5.5 m high , better 1993 0.76 0.02 0.32 

ventilation 

Injection plastics production, 5000 m² steel and brickwork 1989 3.92 2.35 3.90 

construction, 6 m high, limited water supply, no detection, 


Unacceptable high risk level 

Plastic bag producer, storage building: 3700 m² 5 m high fire 1991 0.71 0.66 0.28 

resistive construction , limited water supply, small fire 

brigade , automatic detection. 

Adequate level of protection 

Plastic bag producer, printing unit, 2160 m², 4 m high , 1991 0.91 0.41 0.71 

adequate venting, fire resistive construction , limited water 

supply, small fire brigade , automatic detection. 

Adequate level of protection 

Plastic parts injection process. 3600 m², 8 m high single level 1995 1.26 0.97 1.35 

unit with platforms (E=0.2) ; steel and concrete construction, 

private and public fire brigade , no automatic fire detection. 

Automatic fire detection recommended to improve the 

property and business interruption risk. 

Injection plastic parts dispatch warehouse. 4000 m² , 15 m 1994 0.81 0.67 0.21 

high rack storage, concrete columns, steel roof construction , 

double water supply sprinkler protection 

PE- extruding unit. 7000 m² fire resistive construction, 5 m 1993 9.40 3.04 8.12 

high , no manual intervention means, no detection or 

sprinklers, daytime staffed station. 

totally unacceptable fire risk. 

Plastic coated textiles 9400 m², 5 m high 1.5 % venting, 1990 6.99 2.48 6.05 

limited water supply, lack of manual protection , volunteer 

brigade, fire resistive construction. 

unacceptable high fire risk. 

Latex sheet production, 1850 m², 8 m high, steel 

1990 1.46 0.17 0.85 

construction, no alarm transmission, no hose reels, small 

brigade, ample exit possibilities. 

acceptable risk for personnel safety and business interruption 

Car bumpers production, 7800 m² 10 m high concrete and 

steel construction, high fire load (5250 MJ/m²), 2 % venting, 

public water supply limited, sprinkler system with dedicated 

water supply, fulltime staffed station nearby, ample exits. 

Well protected. 

1995 0.80 0.05 0.47 

60

F.R.A.M.E. 

Pet bottle manufacturer, preform moulding unit. 3300 m², 

concrete construction , 7 m high, 2 % venting, 10 M Euro 

value and long term delivery machines. Limited public water 

supply, sprinkler protection with dedicated water supply, day 

time station at 5 min, all staff trained. 

FRAME used to design the fire protection : Ro = 1.84 

indicates need for sprinklers 

Without sprinklers : 

Same plant, bottle blowing hall , 3400 m², concrete 

construction , 7 m high, 2 % venting, 3 M Euro value, 

standard machinery. Automatic fire detection and linked 

smoke venting , dedicated water supply, , day time station at 

5 min, all staff trained. 

FRAME used to design the fire protection : Ro = 1.48 

indicates fire detection as economic solution. 

Same plant, warehouse for preforms and bottles. 7100 m², 

concrete construction, 7 m high 2 % venting, no heating , 5 

M euro value. Cost for sprinklers is prohibitive. Choice was: 

Automatic fire detection and linked smoke venting , dedicated 

water supply, , day time station at 5 min, all staff trained. 

Adequate level of protection. 

Without detection, unacceptable for property and BI : 

1992 0.67 

1.84 

0.09 

0.26 

0.50 

1.46 

1992 0.71 0.11 0.43 

1992 1.03 0.10 0.49 

2.80 0.13 1.68 

61

F.R.A.M.E. 

Textile industry with FRAME v.1. 

The fire risk in a number of textile factories has been assessed. The following examples show 

the wide variety of risk levels that exist in this industry. 

Case study 38. Stockings factory - production 


Type of occupancy stockings factory, production 1991 Belgium 




Average dimension small objects m 0.05 


Length l 33 

Width b 19 g=0.67 









a 0.2 a=0.2 

Heating : ok 


Explosion: no 


0 

0 

0 

0 



no exit 

signs 

Exits X 5 


t=0.03 





Water supplies low pressure W=0.90 


N = 0.74 

reels 

Special Protection fire detection, day time 

S= 1.89 



f= 60 F=1.54 

Walls: 60 

Ceiling:60 

Partitions:60 


U=8.98 

detection 



Conclusion: 


62

F.R.A.M.E. 

Case study 39. Stockings factory - storage 


Type of occupancy stockings factory, storage 1991 Belgium 




Average dimension carton boxes m 0.7 


Length l 27 

Width b 20 g=0.65 








Main: storage 

a 0 

a=0 

Heating : ok 


Explosion: no 


0 

0 

0 

0 



Exits X 3 


t=0.01 







N = 0.74 

reels 

Special Protection fire detection, day time 

S= 1.89 



f= 52.5 F=1.47 

Walls: 60 

Ceiling:60 

Partitions:0 


U=8.98 

detection 



Conclusion: 


63

F.R.A.M.E. 

Case study 40. Ready made clothing factory 


Type of occupancy ready made clothing factory 1990 Belgium 


Fire load mobile Qm 1500 q=162 




Length l 101 

Width b 49 g=1.80 

Level 10 % platforms E 0.1 e=1.03 







Main: factory 

a 0.2 a=0.2 

Heating : ok 


Explosion: no 


0 

0 

0 

0 



Exits X 7 


t=0.04 






Normal protection limited training – N = 0.81 

Special Protection small fire brigade, 

S= 4.12 

sprinklers 


Walls: 60 

Ceiling:15 

Partitions:0 

f= 31.8 F=1.21 

Escape protection sprinklers, exits U=3.92 



Conclusion: 

well protected risk 

64

F.R.A.M.E. 

Case study 41. Jute spinning plant. 


Type of occupancy jute spinning plant 1990 Belgium 






Length l 100 

Width b 40 g=1.56 

Level E 0.1 e=1.03 







Main: factory 

a 0.2 a=0.3 

Heating : ok 


Explosion: dust 


0 

0 

0.1 

0 



Exits X 5 


t=0.03 







N = 0.74 

reels 



f= 52.5 F=1.49 

Walls: 60 

Ceiling:30 

Partitions:30 

Escape protection Subcompartments, extra 

U=2.92 

exits 



Conclusion: 

inadequately protected. 

65

F.R.A.M.E. 

Case study 42. Textile wholesale trader 


Type of occupancy textile wholesale trader storage 1991 Belgium 






Length l 98 

Width b 64 g=2.16 

Level E 0.05 e=1.01 







Main: storage 

a 0 

a=0 

Heating : none 


Explosion: no 


0 

0 

0 

0 



Exits X 3 


t=0.10 






Normal protection no training – no hose 

N = 0.44 

reels –fire brigade at 15 

min 

Special Protection volunteer fire brigade S= 1.10 


Walls: 30 

Ceiling:15 

Partitions: 

f= 16.88 F=1.16 

Escape protection extra exits U=2.18 



Conclusion: 

poor risk with low probability of fire. 

The owner was recommended to install automatic fire detection and to add more exits to the 

building. This would result in a risk reduction to R =1.52 R1=0.15 and R2 =0.79. 

66

F.R.A.M.E. 

Case study 43. Synthetic fibre spinning mill survey. 


Type of occupancy synthetic fibre spinning mill , 1992 

no compartmentation 






Length l 190 

Width b 83 g=3.15 








Main: textile 

Heating : none 


Explosion: no 


a 0.2 

0 

0 

0 

0 



Exits X 10 


a=0.2 

t=0.05 




Acceptable Risks A= 1.16 A1= 090 A2 =076 




limited training –fire 

brigade at +10 min 

volunteer brigade, water 

storage on site 


Walls: 60 

Ceiling:0 

Partitions: 30 

N = 0.81 

S= 1.55 

f= 48.75 F=1.45 




Conclusion: 

protection adequate for staff, not for property and business. 

As this plant was already subdivided by internal walls, the first recommendation for 

improvement was to create compartments with the installation of fire doors. 

The largest compartment was then 85 m x 52 m , accessible from two sides. The new 

calculation gave for this compartment g=1.79, z= 1.00 , and initial risk Ro=1.18 . 

The resulting values for R = 1.54 , R1=0.53 and R=1.38 illustrate the risk reduction effect of 

the compartmentation. 

As further improvement, the installation of an automatic fire detection system was 

recommended. 

67

More textile industry case studies. 

F.R.A.M.E. 

The next table gives an overview of a number of other textile industries that have been 

evaluated with FRAME. The buildings or compartments have been selected to show the effect 

of unfavourable conditions and compensation by an increase in fire protection 


Weaving and dyeing unit, 7020 m², 5 m high steel and 1992 4.39 1.40 4.53 

concrete construction, limited access, limited water supply, 

limited training, small fire brigade , no hose reels, no 

automatic detection. 

Unacceptable high risk level: partial wet process perceived 

as no fire risk 

Ready made clothing 1500 m², fire resistive construction, 1990 0.79 0.78 0.74 

limited water supply, no alarm transmission to brigade , no 

detection. 

acceptable risk level 

Wool spinning unit, 11250 m², 5 m high fire resistive 1991 1.53 0.34 2.20 

construction , adequate ventilation, limited water supply, 

limited training, ample exits. 

Code compliant for personnel safety. 

Synthetic fibres spinning unit. 5400 m², 4 m high steel 1990 2.58 0.98 1.98 

and brickwork construction, adequate ventilation, lacks 

hose reels and training, no detection : code compliant for 

personnel safety, inadequate protection for property and 

business 

Jacquard weaving , 3900 m² , 6 m high, 1% venting, 

concrete construction, gas heated room, limited water 

supply, all persons trained in fire fighting, no alarm 

transmission, no hose reels, small fire brigade. Poorly 

protected risk. 

Improvement with manual fire fighting and alarm system 

1990 

improved 

4.06 

1.71 

1.78 

1.10 

4.38 

1.85 

gives acceptable level of protection for personnel safety 

and better protection for property. 

Clothes manufacturer, 8900 m², 8 m high concrete 

1992 2.07 0.50 1.49 

construction, no venting, limited public water supply, 

water tank for fire fighting, no hose reels, automatic fire 

detection, fire brigade at 5 min. ample exits. 

Adequate personnel protection . Building too large for 

protection by detection only. 

Synthetic fibre spinning mill, 12600 m² brickwork 

1992 3.98 0.46 4.03 

construction, 5 m high, subdivided by walls without fire 

doors, limited public water supply, day time staffed station 

at 5 min. 

Protection adequate for staff not for property and business 

For this plant, compartmentation by adding fire doors was improved 1.53 0.28 1.44 

the first improvement. The largest compartment would be 

92 x32 m and resulted in the corresponding values: 

Cotton fibre spinning mill, 12500 m², concrete building, 7 

m high , no venting, limited public water supply, all staff 

trained, full time station at 5 min, no sprinklers 

1989 

2.70 0.50 1.44 

with sprinkler protection 

0.90 

0.25 

0.42 

68

F.R.A.M.E. 

Floor coverings industry with FRAME v.1. 

A sub sector in the textile industry, which has a bad reputation among the insurers, because of 

some large fire losses, is floor-covering production. The process of this mass produced 

product requires large production halls and sprinkler protection is not the major concern of this 

industry. Here is a list of larger fires in this industry: 

1985 : Oostrozebeke , Belgium 

July 26, 1990: Arles, France 

May 7 th , 1997 : Assendelft Netherlands 

Nov 21 st 2000: Broadway, Haslingden, Lancashire , UK 

March 8 th , 2001: Hasselt –Overijssel, Netherlands 

September 2001: Waterford Ireland 

January 20, 2006: Yatala, Australia 

April 9, 2007 : Sydney Australia 

That such large fires can occur is no surprise when the following FRAME risk assessments are 

obtained for this industry. 

Case study 44. Floor covering factory 


Type of occupancy floor covering latex coating unit , 1989 Belgium 






Length l 170 

Width b 90 g=3.23 








Main: manufacturing a 0.2 a=0.5 

Heating : gas heated unit 

Electrical: not ok 

Explosion: no 


0.2 

0.1 

0 

0 



Exits X 13 


t=0.05 






low pressure, limited 

W=0.77 

hydrant points 


N = 0.74 

reels 

Special Protection private owned water 

supply, volunteer fire 

brigade 

S= 3.73 

69

F.R.A.M.E. 


f= 52.5 F=1.40 

Walls: 60 

Ceiling:60 

Partitions:0 

Escape protection multiple exits, U= 1.89 

Salvage 

business share, CO2 100 % y=6 Y=1.24 

system on latex machine 


Conclusion: 

poor risk : large compartment, minimal protection 

Case study 45. Floor covering tufting 


Type of occupancy floor covering tufting 1992 France 






Length l 160 

Width b 120 g=3.90 

Level small platforms E 0.05 e=1.01 







Main: factory 

a 0.2 a=0.2 

Heating : ok 


Explosion: no 


0 

0 

0 

0 



Exits X 12 


t=0.06 







N = 0.63 

alarm transmission 



f= 26.25 F=1.24 

Walls: 30 

Ceiling:30 

Partitions: 

Escape protection ample exits U=2.93 



Conclusion: 

very poor risk : large compartment, no adequate structural fire 

resistance, only manual protection 

70

F.R.A.M.E. 

Case study 46. Floor covering, latex coating 


Type of occupancy floor covering latex coating unit , 1992 France 






Length l 145 

Width b 60 g=2.31 









Heating : gas heated unit 


Explosion: no 


0.25 

00 

0 



Exits X 7 


t=0.04 





Water supplies limited size network W=0.66 


N = 0.63 

alarm transmission 

Special Protection private owned water 

S= 1.48 

supply, volunteer fire 

brigade 


Walls: 60 

Ceiling:30 

f= 48.75 F=1.45 

Partitions:0 

Escape protection multiple exits, U= 3.56 



Conclusion: 

poor risk : large compartment, minimal protection 

71

F.R.A.M.E. 

Case study 47. Floor covering tufting 


Type of occupancy floor covering tufting 1992 Belgium 






Length l 200 

Width b 75 g=2.98 








Main: factory 

a 0.2 a=0.6 

gas heating in room 


Explosion: no 


0.3 

0.1 

0 

0 



Exits X 12 


t=0.06 







Special Protection small fire brigade S= 1.28 


f= 24.4 F=1.23 

Walls: 60 

Ceiling:15 

Partitions:0 




Conclusion: 

very poor risk : large compartment, no adequate structural fire 

resistance, only manual protection 

72

F.R.A.M.E. 

Case study 48. Carpet manufacturing, storage 


Type of occupancy carpet manufacturing, storage 1992 Belgium 






Length l 65 

Width b 55 g=1.73 








Main: storage 

a 0 

a=0.2 

Heating : ok 


Explosion: no 

Secondary: shrink 

wrapping 

0 

0.1 

0 

0.1 



Exits X 5 


t=0.02 






Normal protection no training – no 

N = 0.74 

extinguishers 

Special Protection small fire brigade S= 1.28 


f= 15 F=1.14 

Walls: 30 

Ceiling:0 

Partitions:0 




Conclusion: 

code complying building : occupants’ safety is OK, property 

protection inadequate for the high fire load. 

73

F.R.A.M.E. 

Case study 49. Wool carpet weaving 


Type of occupancy wool carpet weaving 1992 Belgium 






Length l 120 

Width b 45 g=1.79 








Main: factory 

a 0.2 a=0.2 

Heating : ok 


Explosion: no 


0 

0 

0 

0 



Exits X 6 


t=0.03 







N = 0.74 

reels 



f= 52.5 F=1.49 

Walls: 60 

Ceiling:0 

Partitions:60 

Escape protection Subcompartments, ample 

U=3.56 

exits 



Conclusion: 

code complying building : occupant’s safety is OK, property 

protection inadequate for the large building. 

74

F.R.A.M.E. 

Woodworking industry with FRAME v.1. 

This is also an industry where large fires have occurred. Using FRAME permits an insurer to 

select the better risks, or to show which premium benefit can be obtained by better fire 

protection. The next case studies illustrate the variety of risks in this industry. 

Case study 50. Woodworking shop 


Type of occupancy woodworking shop 1991 Belgium 






Length l 70 

Width b 55 g=1.77 

Level E 0.12 e=1.03 







Main: woodworking a 0.4 a=0.4 

Heating : ok 


Explosion: ok 


0 

0 

0 

0 



Exits X 5 


t=0.03 







N = 0.66 

reels – fire brigade at 15 

min 

Special Protection day time staffed station – 

S= 1.41 

water storage 

Fire resistance 

Structure:15 

Walls: 30 

Ceiling:15 

Partitions: 

steel f= 16.88 F=1.16 




Conclusion: 

inadequately protected. 

75

F.R.A.M.E. 

Case study 51. Furniture manufacturer 


Type of occupancy furniture manufacturer 1990 Belgium 






Length l 90 

Width b 47 g=1.69 

Level E 0.05 e=1.01 








Heating : ok 


Explosion: ok 


0 

0 

0 

0 



Exits X 6 


t=0.03 






Normal protection no alarm transmission N = 0.90 


S= 1.41 

near canal 

Fire resistance 

Structure:30 

f= 33.75 F=1.32 

Walls: 60 

Ceiling:30 

Partitions: 




Conclusion: 

code compliant for personnel safety, inadequately protected 

property. 

76

F.R.A.M.E. 

More woodworking industry case studies. 

The next table gives an overview of a number of other woodworking industries that have been 

evaluated with FRAME. The buildings or compartments have been sorted from high risk to low 

to show the effect of unfavourable conditions and compensation by an increase in fire 

protection 


Carpentry workshop of 2600 m², 6 m high. Steel 

1989 5.57 12.64 4.97 

structure, brick outside walls. Heating by waste furnace 

inside the shop. Limited water supply, no alarm 

transmission, no hose reels, daytime staffed station. 

Very low value for A1, indicating unacceptable risk 

MFD –board sawmill, 6400 m², 6 m high steel structure, 

moderate fire load density, secondary activity :packaging, 

access on 2 sides only, limited water supply, day time 

staffed station at 10 min. , limited training, ample exits. 

1993 

with 

4.79 

0.98 

0.61 

0.50 

3.10 

0.75 

Inadequate property protection. Recommendation for 

sprinklers to obtain adequate level of protection. (with 

additional sprinklers to protect steel columns). 

sprinklers 

Construction wood wholesaler –storage: 5200 m² building 1992 4.07 0.25 1.92 

steel structure, with high fire load (4000 MJ)m²) , 6 m 

high well ventilated, limited water supply, no detection, 

small fire brigade, small number of occupants, plenty of 

exits. 

Code compliant for personnel safety, very high risk for 

property. 

Furniture manufacturing, 5980 m², 4 m high, concrete 1995 2.87 1.13 3.05 

construction, 1% smoke vents, limited water network, 

limited training, no hose reels, no detection. 

The risk is considered to be too high. As improvement was 

recommended to subdivide in two compartments, to add 

smoke venting and to install hose reels and automatic fire 

detection. The result is shown in the following calculation. 

Woodwork shop , 2700 m² steel structure, brick walls, 1989 2.68 0.96 1.68 

limited water supply , small fire brigade, ample exits. 

Code compliant for personnel safety only. 

Furniture manufacturer 3290 m² single storey fire 

1990 2.11 1.59 2.12 

resistive building, limited water supply, daytime staffed 

station, no detection, limited exits. 

Inadequate protection level. 

Leather chair manufacturer, 5100 m², 5 m high, 5 % 1989 1.56 0.57 1.12 

venting, steel structure , gas heating inside unit, limited 

use of glues, very limited public water supply, sprinkler 

protection single water supply, station at 10 min., part 

time staffed. 

Sprinkler protection does not compensate enough for large 

compartment with high fire load and no fire resistance of 

structure. 

Paint shop for wooden windows ; 810 m² steel structure, 1989 1.53 0.69 1.04 

brick walls, limited water supply , small fire brigade, 

ample exits 

Sawmill of 1650 m², 5.5 m high single storey fire resistive 

construction, limited water supply, daytime staffed 

station, no detection, ample exits 

1990 1.39 0.52 0.97 

77

F.R.A.M.E. 

Code compliant for personnel safety only. 

Furniture dealer, 1525 m², 3.5 m high, concrete 

construction, 1% smoke vents, limited access, hose reels , 

no extinguishers, trained staff, automatic fire detection 

with transmission to fire brigade : well protected 

Furniture assembly hall, 3325 m² , 4 m high, concrete 

construction, 2% venting, limited water network, limited 

training, automatic detection, hose reels installed . 

Acceptable risk. 

1990 0.85 0.80 0.70 

1995 0.77 0.41 0.72 

78

F.R.A.M.E. 

Food industry with FRAME v.1. 

The next series of case studies give the risk assessments for a number of food factories. 

Case study 52. Biscuit bakery 


Type of occupancy biscuit bakery 1993 Belgium 






Length l 290 

Width b 40 g=2.14 

Level E 0.1 e=1.03 







Main: food industry a 0.2 a=0.4 

Heating : baking oven 


Explosion: ok 


0.2 

0 

0 

0 



Exits X 15 


t=0.07 






Normal protection limited training – fire 

N = 0.70 



S= 6.70 

water storage -sprinklers 


Walls: 60 

Ceiling:30 

Partitions:0 

f= 48.75 F=1.26 

Escape protection extra exits- sprinklers U=5.79 



Conclusion: 

adequately protected risk. 

79

F.R.A.M.E. 

Case study 53. Chocolate bar production 


Type of occupancy chocolate bar production 1992 






Length l 280 

Width b 90 g=3.73 

Level E 0.2 e=1.05 







Main: food industry 



Explosion: ok 


a 0.2 

0.1 

0.1 

0 

0 



Exits X 11 


a=0.4 

t=0.08 





Water supplies no water supply W=0.60 

Normal protection ok N = 1.00 


day time staffed station 

-automatic detection 


Walls: 60 

Ceiling:60 

Partitions:0 

S= 3.07 

f= 52.50 F=1.43 

Escape protection detection –exits U=2.93 



Conclusion: 

This factory has suffered a large fire in 2004. 

code compliant : ok for personnel safety, detection is inadequate 

for property and business protection of such a large building. 

80

F.R.A.M.E. 

Case study 54. Chocolate production 

This unit belongs to an other chocolate producer, but needs equally an improvement pf the 

fire protection. 


Type of occupancy chocolate (liquid) production 1992 






Length l 70 

Width b 65 g=1.99 








Main: food industry 

Heating : steam 


Explosion: ok 


a 0.2 

0 

0 

0 

0 



Exits X 8 


a=0.2 

t=0.03 






Normal protection fire brigade at 15 min N = 0.77 



Walls: 60 

Ceiling:30 

Partitions:0 

f= 48.75 F=1.46 




Conclusion: 

code compliant : ok for personnel safety, but not enough for 

property and business protection. 

81

F.R.A.M.E. 

Case study 55. Food additive (seasonings) manufacturer 


Type of occupancy seasonings manufacturer 1994 Belgium 






Length l 94 

Width b 43 g=3.73 

Level E 0.1 e=1.03 







Main: “wet” industry a 0 

a=0.1 



Explosion: ok 


0.1 

0 

0 

0 



Exits X 7 


t=0.03 









Walls: 60 

Ceiling:60 

f= 54.38 F=1.51 

Partitions:15 




Conclusion: 

adequate protection level 

82

F.R.A.M.E. 

Case study 56. Frozen vegetables plant 


Type of occupancy frozen vegetables plant 1989 Belgium 






Length l 44 

Width b 44 g=1.32 

Level E 0.2 e=1.05 







Main: “wet” industry a 0 

a=0.3 

Heating : cooking 


Explosion: ok 

Secondary: packing 

0.2 

0 

0 

0.1 



Exits X 5 


t=0.02 






limited network low 

W=0.60 

pressure 

Normal protection no training no hose reels N = 0.66 

Special Protection small fire brigade – large 

S= 1.63 

water pond 


Walls: 30 

Ceiling:15 

Partitions: 

steel f= 16.88 F=1.15 




Conclusion: 

code compliant: adequate for personnel, not enough for property 

83

F.R.A.M.E. 

Case study 57. Frozen vegetables storage 


Type of occupancy frozen vegetables storage 1989 Belgium 

Fire load immobile insulation! Qi 1500 




Reaction to fire metal sandwich panels M 1 i=1.10 

Length l 44 

Width b 31 g=1.03 








Main: storage 

a 0 

a=0.1 

Heating : no 


Explosion: ok 

Secondary: ammonia 

cooling 

0 

0 

0 

0.1 



no clear 

exit 

Exits X 3 


t=0.05 






limited network low 

W=0.60 

pressure 

Normal protection no training no hose reels N = 0.66 

Special Protection small fire brigade – large 

S= 1.63 

water pond 


steel f= 16.88 F=1.15 

Walls: 30 

Ceiling:15 

Partitions: 




Conclusion: 

code compliant: adequate for personnel, not enough for 

property, less impact on business 

Case study 58. Large slaughterhouse inspection . 

This case study was made for an insurance inspection as the broker looked for a low premium 

rate for this “wet process”. The answer was negative due to the size of the building with only 

elementary protection, which could result in a total loss. 


Type of occupancy Abattoir communal (slaughterhouse) 1990 Belgium 


84

F.R.A.M.E. 





Length l 150 

Width b 100 g=3.36 

Level E 0 e= 1.00 


Ventilation 0.5% k 0.005 v=0.99 





Main: low hazard 

Heating :gas 



Secondary: burning hides 

0 

0.1 

0 

0 

0.1 



Exits X 10 


a=0.2 

t=0.05 




Acceptable Risks A= 1.23 A1=1.05 A2 =1.03 

Water supplies not enough hydrants W=0.95 

Normal protection no education N = 0.81 



Walls: 60 

Ceiling:15 

Partitions:0 

f=46.8 

F=1.44 

Escape protection less than 300, extra exits U= 2.41 



Conclusion: 

very large building with only basic protection. Adequate for 

peoples’ safety, total loss possible 

85

F.R.A.M.E. 

Case study 59. Animal feed production, mixing unit, basement. 

These case studies are made for an animal feed production unit, where the fire risk is quite 

different in the various compartments. 


Type of occupancy animal feed mixing 1990 Belgium 






Length l 18 

Width b 15 g=0.47 

Level E -1 e=1.22 


Ventilation basement k 0 v=1.18 


Height difference H -4 z=1.05 



Main: high hazard 

a 0.4 a=0.6 

Heating :none 




0 

0 

0.2 

0 



Exits X 2 


t=0.01 



dependency factor very large turnover d d= 0.05 



network too small, low 

W=0.60 

pressure 

Normal protection lack of hose stations, fire 

N = 0.74 


Special Protection near canal, volunteer 

S= 1.89 

brigade 


Walls: 60 

Ceiling:60 

Partitions:0 

f= 52.5 F=1.47 

Escape protection U= 7.39 



Conclusion: 

small compartment , low level of protection for property. 

86

F.R.A.M.E. 

Case study 60. Animal feed production (bagging) 


Type of occupancy bagging unit 1990 Belgium 






Length l 40 

Width b 20 g=0.73 

Level intermediate platform E 0.2 e=1.05 


Ventilation 1 % k 0.01 v=1.02 






a 0.4 a=0.5 

Heating :none 




0 

0 

0.1 

0 



Exits X 4 


t=0.01 







W=0.60 

pressure 


N = 0.74 



S= 1.89 

brigade 


Walls: 60 

Ceiling:60 

Partitions:0 

f= 52.5 F=1.47 




Conclusion: 

small compartment , low level of protection for property. 

87

F.R.A.M.E. 

Case study 61. Animal feed production (top of mixing tower) 


Type of occupancy mixing tower 1990 Belgium 






Length l 18 

Width b 15 g=0.47 

Level intermediate platform E 13 e=1.80 


Ventilation 1 % k 0.01 v=1.02 






a 0.4 a=0.6 

Heating :none 




0 

0 

0.2 

0 



Exits X 1 


t=0.01 







W=0.60 

pressure 


N = 0.74 



S= 1.89 

brigade 


Walls: 60 

Ceiling:60 

Partitions:0 

f= 52.5 F=1.47 




Conclusion: 

Most sensitive part of the whole unit, cannot be reached easily. 

88

F.R.A.M.E. 

Case study 62. Storage silos 


Type of occupancy storage silos 1990 Belgium 






Length l 35 

Width b 13 g=0.52 








Main: storage 

a 0 

a=0.3 

Heating :none 




0 

0 

0.2 

0.1 



Exits X 1 


t=0.01 







W=0.60 

pressure 


N = 0.74 



S= 1.89 

brigade 


f= 52.5 F=1.47 

Walls: 60 

Ceiling:60 

Partitions:0 


U=7.39 



Conclusion: 

Most sensitive part of the whole unit, cannot be reached easily. 

89

F.R.A.M.E. 

More food industry case studies 

The next table gives an overview of a number of other food industries that have been 



capsule manufacturing unit. 6000 m² single storey building, 1994 2.52 1.08 2.90 

fire resistive construction with sub compartments, 6m high, 

no venting (for hygienic reasons) , limited water supply, 

small plant intervention team, no detection but 24hr/24hr 

working: code compliant protection, just OK for personnel , 

inadequate for property and business . 

Chewing gum production unit. 5200 m² single storey 

1994 0.87 0.48 0.79 

building, fire resistive construction with sub compartments, 

5m high, no venting (for hygienic reasons) , adequate water 

supply, small fire brigade, sprinkler protection. 

Adequate protection 

Bulk cocoa nut storage, high fire load, 10800 m², 8m high 1994 3.81 0.19 067 

single storey building, no venting, limited water supply, 

automatic fire detection, daytime staffed station, ample exits 

. 

Automatic fire detection is inadequate for such a large 

building and fire load. Most economical improvement would 

be compartmentalisation. 

Chocolates production unit, 6000 m² 3.5 m high first floor in 1994 2.01 0.97 3.02 

fire resistive building, access on 2 sides, adequate water 

supply, fulltime staffed station at 5 min. no detection, 

adequate exits. 

Combination of large compartment at 1 st floor , limited access 

and lack of automatic protection results in inadequate 

property protection. 

Meat preparation unit. 5700 m² 5 m high, protected steel 1995 1.63 0.62 1.38 

construction, no venting, adequate water supply, no hose 

reels, limited training, no detection, part time station at 5 

min. 

Adequate for personnel safety, 50 % damage possible ( e.g. 

from fire after working hours) 

Meat cutting plant, 10150 m² 6 m high, concrete 

construction, occasional explosion risk because of ammonia 

cooling system, no venting, limited water supply, part time 

station at 10 min. no alarm transmission, ample exits. 

Adequate for personnel safety, inadequate for property and 

business protection for such a large unit. Note: in this type of 

industry, a limited fire can result in a total loss of the 

content, for hygienic reasons. 

1990 2.29 0.84 2.34 

90

F.R.A.M.E. 

Electric and electronics industry with FRAME v.1 

Case study 63. Electronic print manufacturing, building with several levels 

A large number of production units were reviewed and FRAME was used to identify those 

where the level of fire protection needed improvement. 


Type of occupancy electronic print manufacturing 1994 Belgium 






Length l 39 

Width b 20 g=0.72 

Level 5 th floor E 5 e=1.60 


Ventilation exhaust system k 0.021 v=0.92 






Heating :none 


Explosion: no 


0 

0 

0 

0.1 



evacuation 

path not 

clear 

Exits X 2 


t=0.05 


5 M. euro , long term 

c=0.12 

equipment 






Special Protection automatic detection, 

S= 2.18 

professional fire brigade 


f= 37.5 F=1.33 

Walls: 0 

Ceiling:60 

Partitions:0 

Escape protection external stair U=3.92 



Conclusion: 

automatic fire detection considered to be adequate to obtain 

acceptable risk level. 

91

F.R.A.M.E. 

Case study 64. Radio assembly hall. 


Type of occupancy Radio assembly hall 1993 


Fire load mobile packaging material Qm 1500 q=1.57 




Length l 70 

Width b 60 g=0.72 








Main: manufacturing 

Heating :none 


Explosion: no 

Secondary: no 

a 0.2 

0 

0 

0 

0 



Exits X 7 


a=0.2 

t=0.03 








automatic detection, day 

time staffed station, plant 

brigade 


Walls: 30 

Ceiling:30 

Partitions:0 

S= 3.56 

f= 26.25 F=1.18 

Escape protection detection, exits, plant 

U=3.56 

and public brigade 



Conclusion: 

automatic fire detection and organisation considered to be 

adequate to obtain acceptable risk level. 

92

F.R.A.M.E. 

Case study 65. Electronic parts sub assembly hall. 


Type of occupancy Electronic metal parts sub assembly hall 1993 




Average dimension many small parts m 0.01 


Length l 120 

Width b 45 g=1.79 








Main: metal parts 

Heating :none 


Explosion: no 

Secondary: local welding 

a 0 

0 

0 

0 

0.05 



Exits X 12 


a=0.05 

t=0.03 






Normal protection brigade at +10 min N = 0.90 



day time staffed station, 

plant brigade 


Walls: 30 

Ceiling:30 

Partitions:0 

S= 2.53 

f= 26.25 F=1.21 

Escape protection exits, plant and public 

U=4.32 

brigade 



Conclusion: 

adequate protection level 

An other assembly hall of similar size, where plastic based parts were made and assembled, 

had higher values of calculated risks : R= 1.34 R1=0.83 R2=0.99. The rack storage of that 

plant obtained R=1.71 , R1=0.29 , R2 = 0.47. For these two buildings, an automatic fire 

detection system was recommended as improvement. 

93

F.R.A.M.E. 

Case study 66. Special lamps manufacturing hall. 


Type of occupancy Special lamps manufacturing hall 1994 






Length l 120 

Width b 50 g=1.92 









gas burners in the hall 


Explosion: no 


a 0.2 

0.25 

0 

0 

0 



Exits X 8 


a=0.45 

t=0.04 





Water supplies ok W=1.00 



heat detectors, dedicated 

water supply, day time 

staffed station, plant 

brigade 


Walls: 15 

Ceiling:60 

Partitions:0 

S= 4.32 

f= 41.25 F=1.29 

Escape protection exits, plant and public 

U=3.73 

brigade 



Conclusion: 

adequate protection level for property and personnel, too low for 

business. 

It was recommended to replace the older heat detector system by a modern rate-of-rise and 

addressable fire detection system to obtain a faster and more precise early warning. The 

improvement would result in R=0.64, R1=0.42 and R2=0.99. The benefit is mainly a 

reduction of the business interruption potential. 

94

F.R.A.M.E. 

More electric and electronic industry case studies 

The next table gives an overview of a number of other electric and electronic industries that 

have been evaluated with FRAME. 


Electric consumer products spare parts storage. 1500 m² 9m 1993 0.75 0.94 0.25 

high concrete and steel rack storage building and distribution 

centre. Fire load density 6000 MJ/m² Protected by single 

water supply sprinkler system, professional fire brigade. 

Adequate protection level 

Finished goods storage at radio plant. 3200 m² single story 1993 0.91 0.27 0.28 

high rack storage , high fire load 10500 MJ/m², fire resistive 

structure, detection, plant brigade, no sprinklers : protection 

considered adequate. 

Metal workshop for own machinery , 8500 m² single level 1994 1.06 0.59 0.81 

steel structure , plant fire brigade, dedicated fire water 

supply system, semi professional public brigade at 5 min. 

Level just acceptable, 

CD- player assembly, 3000 m² single level steel structure, 1994 0.40 0.20 0.42 

plant fire brigade, dedicated fire water supply system, semi 

professional public brigade at 5 min. 


Print assembly clean room, 1600 m², fire resistive 

construction, high added value, automatic fire detection, 

plant fire brigade, dedicated fire water supply system, semi 

professional public brigade at 5 min. 


1994 0.24 0.22 0.75 

95

F.R.A.M.E. 

Other industries with FRAME v.1. 

Case studies 67. Paint manufacturer, finished products warehouse. 

These case studies are examples of risk calculations made during an insurance inspection. 

They show how the risk in two different operations can be compared. 


Type of occupancy High bay warehouse for paint 1992 Netherlands 


Fire load mobile rack storage Qm 9000 q=2.09 




Length l 117 

Width b 30 g=1.33 

Level E 0 e= 1.00 


Ventilation 1% k 0.010 v=1.04 





Main: storage 

a 0.0 a=0 

Heating :none 




0 

0 

0 

0 



Exits X 3 


t=0.03 






quantity and network too 

W=0.74 

small 

Normal protection N = 1.00 

Special Protection sprinklers +fulltime 

S= 10.92 

staffed station + plant 

brigade 


f= 37.5 F=1.03 

Walls: 30 

Ceiling:5 

Partitions:0 




Conclusion: 

medium size rack storage , over protected (sprinkler protection 

for whole site) 

96

F.R.A.M.E. 

Case study 68. Paint manufacturer, lacquer-manufacturing unit. 


Type of occupancy lacquer manufacturing 1992 Netherlands 






Length l 63 

Width b 57 g=1.76 



Ventilation 2% k 0.02 v=0.98 






a 0.4 a=0.7 

Heating :none 


Explosion: yes 

Secondary: static 

electricity 

0 

0 

0.2 

0.1 



Exits X 8 


t=0.02 





Water supplies network too small W=0.81 

Normal protection N = 1.00 

Special Protection detection+ CO2 flooding 

system + fulltime staffed 

S= 9.91 

station + plant brigade 


Walls: 30 

Ceiling:30 

Partitions:30 

f= 30 F=1.01 




Conclusion: 

High hazard but highly protected risk 

97

F.R.A.M.E. 

Case study 69. Candle manufacturer. 

After a large fire at a colleagues’ plant, the candle manufacturer faced a large premium 

increase from his insurer and was looking for an other insurance company. 


Type of occupancy seasonings manufacturer 1990 Belgium 






Length l 67 

Width b 30 g=1.13 

Level E 0.06 e=1.02 







Main: candle industry a 0.4 a=0.1 

Heating : ok 


Explosion: ok 


0 

0 

0 

0 



no clear 

exits 

Exits X 6 


t=0.06 






Normal protection no training N = 0.74 

Special Protection professional fire brigade S= 1.34 


Walls: 60 

Ceiling:60 

f= 60 F=1.56 

Partitions:60 




Conclusion: 

not as bad as expected, provide training for personnel 

98

F.R.A.M.E. 

Case study 70. Honeycomb material manufacturer. 


Type of occupancy honeycomb material manufacturer 1994 






Length l 175 

Width b 85 g=3.13 

Level E 0.1 e=1.03 







Main: plastics industry 

local process heating 


Explosion: ok 


a 0.4 

0.25 

0 

0 

0 



Exits X 20 


a=0.65 

t=0.05 









sprinklers , dedicated 

water supply 


Walls: 60 

Ceiling:60 

Partitions:0 

S= 7.39 

f= 52.5 F=1.26 

Escape protection exits -sprinklers U=6.70 



Conclusion: 

adequate for property and personnel safety, may be improved 

for business risk by a contingency plan. 

99

F.R.A.M.E. 

Case study 71. Book printing . 


Type of occupancy book printing 1990 






Length l 88 

Width b 31 g=1.25 

Level E 0.1 e=1.03 







Main: printing 

heating : OK 

Electrical: not checked 

Explosion: ok 


a 0.2 

0 

0.1 

0 

0 



Exits X 6 


a=0.30 

t=0.03 









automatic fire detection 


Walls: 60 

Ceiling:0 

Partitions:30 

S= 2.41 

f= 48.75 F=1.42 

Escape protection exits detection U=3.23 



Conclusion: 

automatic fire detection was recommended to improve risk 

100

F.R.A.M.E. 

Case study 72. Car seats assembly plant. 


Type of occupancy car seats assembly 1994 






Length l 170 

Width b 119 g=3.95 

Level E 0.1 e=1.03 







Main: car seat assembly 

heating : gas 


Explosion: ok 


a 0.2 

0.1 

0 

0 

0 



unclear 

evacuation 

plan 

Exits X 12 

a=0.30 


t=0.19 





Water supplies ok W=1.00 




single water supply 


Walls: 30 

Ceiling:30 

Partitions:30 

S= 4.54 

f= 41.25 F=1.28 

Escape protection sprinklers U=3.23 



Conclusion: 

long exit paths, no clear evacuation plan are a threat for the 

people. The initial risk Ro= 6.11 indicating that the risk is too 

high. The first recommendation was to divide the unit into two 

compartments by adding fire doors to existing separation walls. 

Such a large plant should also have a sprinkler system with 

double water supply for reliability to obtain an adequate level of 

protection. 

101

F.R.A.M.E. 

Case study 73. Oil regeneration plant. 


Type of occupancy oil regeneration plant 1989 


Fire load mobile High Qm 2500 q=1.73 


Average dimension small tanks m 1 


Length l 20 

Width b 15 g=0.49 

Level E 0.3 e=1.08 







Main: oil treatment 

heating : in the room 




a 0.4 

0.2 

0 

0.1 

0 



Exits X 3 


a=0.70 

t=0.01 





Water supplies only one hydrant nearby W=0.56 

Normal protection no alarm transmission N = 0.66 

Special Protection fire brigade S= 1.28 


Walls: 30 

Ceiling:15 

Partitions:0 

f= 46.87 F=1.44 

Escape protection ample exit capacity U=5.00 



Conclusion: 

below standard normal protection and water supply need 

improvement. 

102

F.R.A.M.E. 

Case studies 74. Metalwork industry. 


Type of occupancy Welding shop 1989 


Fire load mobile rack storage Qm 200 q=0.98 



Reaction to fire M .5 i=0.68 

Length l 55 

Width b 17 g=0.71 

Level E 0.1 e= 1.03 


Ventilation 1% k 7 % v=062 





Main: Welding 

Heating :gas 




a 0.0 

0.1 

0 

0 

0 



Exits X 3 


a=0.1 

t=0.02 








limited training , station 

at 10 min 


station 


Walls: 60 

Ceiling:15 

Partitions:0 

N = 081 

S= 1.48 

f= 46.88 F=1.44 

Escape protection exits U= 3.23 



Conclusion: 

low risk, fire resistive construction requested by fire service was 

not justified. 

103

F.R.A.M.E. 

Other industry case studies. 


Car assembly plant. 100000 m², mainly concrete 

1994 1.11 0.25 0.43 

construction, 6 m high, 1.5 % venting capacity. Fire load 

estimated at 1200 MJ/m² , 400 persons present, adequate 

and ample water supplies, all staff trained, sprinkler 

protection with dedicated highly reliable water supply , full 

time professional fire station at 5 min., plant fire brigade 

Slightly inadequate property protection level can be explained 

by conservative fire load estimate, a more precise fire load 

calculation should be made. 

This is a very large plant, that needs (and has) all available 

protection, except automatic fire detection. 

Book storage, 5000 m² with 25% intermediate platforms, 1995 1.34 0.56 0.77 

steel construction 6.5 m high, 1 % smoke venting, access 

from 2 sides only, pallet wrapping as secondary activity, 

adequate water supply, limited training, full time station at 

5min. , automatic fire detection, 

Acceptable level of protection with increased property loss 

potential. 

Instruments manufacturer office and shop building, 5760 m², 1992 1.33 0.41 1.19 

mixed steel and brickwork construction, 4.5 m high, 1 % 

venting, gas heating, hose reels and extinguishers, no 

automatic detection, ample exits, full time station at 10 min. 

Improve property protection either by compartmentation 

between offices and shop or by automatic fire detection 

Instruments manufacturer (other site), 9350 m², mixed steel 

and brickwork construction, 5m high , 2 % venting, gas 

heating, secondary painting activity, hose reels, 

extinguishers, automatic detection, full time station at 5 min, 

ample exits. 

Adequate protection for all 3 aspects. 

1992 0.87 0.35 0.67 

104

F.R.A.M.E. 

Real fire cases with FRAME V.1 

Case study 75. Fire at Menuiserie Paul Ottignies (BE) , December 1988. 

Menuiserie Paul Ottignies. A fire destroyed this woodworking company on December 12 th 

1988. The fire was documented in the Revue belge du Feu n° 94 as “fire study n° 140” and the 

FRAME calculation was made with the data found in that report. 


Type of occupancy Woodworking 


Fire load mobile High Qm 2000 q=1.66 




Length l 32 

Width b 30 g=0.92 

Level E 0.1 e=1.03 





Potential Risks P= 1.86 P1=2.0.3 P2=1.12 



heating : none 




0 

0 

0 

0 



Exits X 3 


t=0.01 






Normal protection no alarm transmission, 

lack of hose reels, 

training, station at +10 

min 

N = 0.54 

Special Protection fire brigade S= 1.34 


Walls: 60 

Ceiling:15 

Partitions:0 

steel with 

brick walls 

f= 24.38 F=1.23 

Escape protection ample exit capacity 13.9 


Calculated Risks: R= 2.20 R1= 0.39 2.15 

Conclusion: 

total loss to be expected. 

105

F.R.A.M.E. 

Other fire cases verified with FRAME v.1. 

The following fire cases were also verified with FRAME v.1 

Ets DANIEL DOYEN Anderlecht (BE) on November 12 th 1981. 

Reference : Revue belge du Feu n° 61, fire case n° 97. 

This 2000 m² building with combustible construction was used a storage for car batteries, and 

was completely destroyed. The damage was about 200 million BEF of that time (approx. 5 M. 

Euro) 

The FRAME calculation, made with the data of the detailed report, gave the following results : 

P = q.i.g.e.v.z = 1.79 * 1.27 * 1.21 * 1.00 * 1.00 * 1.05 = 2.88 

A = 1.6 –a –t- c = 1.6 – 0.3 - 0.02 – 0.04 = 1.24 

D = W.N.S.F = 0.57 * 0.81 * 1.89 * 1.05 = 0.92 

R = P/ A*D = 2.53 

P1 = q.i.e.v.z = 2.38 

A1 = 1.6 –a – t – r = 1.6 -0.3 – 0.02 – 0.7 = 0.58 

D1 = N.U = 0.81 * 3.07 = 2.49 

R1 = P’/ A’ * D’ = 1.65 

P2= 1.62 A2 = 1.18 D2 = W.N.S.Y= 0.57* 0.81*1.89 *1.34 = 1.17 

R2 =1.60 

The loss is in line with the FRAME calculation 

106

F.R.A.M.E. 

FREME. 

The predecessor of FRAME was “FREME – fire risk evaluation method” . This was a 

development of the Swiss Gretener method combined with a risk calculation for the occupants. 

The formulas used were the same as in FRAME v.1 : 

For the property risk P = q.i.g.e.v.z A= 1.6 –a –t- c D= W.N.S.F 

R = P/ A*D 

For the occupant risk P’ = q.i.e.v.z A’ = 1.6 –a – t – r D’ = N.U R’ = P’/ A’ * D’ 

Case study 76. Summerland fire, Isle of Man, August 2 nd , 1973. 

Reference : This fire was discussed in detail in the professional press of that period. 

Information can be found on the Internet, see http://www.iomfire.com/main/Summerland.htm 

and http://en.wikipedia.org/wiki/Summerland_disaster . 

This fire incident was used as case study for FREME. Two calculations were made, one for the 

situation at the leisure centre as it was, “and as it should be” if regulations were followed and 

a Gretener/ FREME approach would be used during its design. 

For the “as it was” situation , the following results were obtained : 

P = q.i.g.e.v.z = 1.5 * 1.3 * 1.87 * 1.46 * 1.07 * 1.10 = 6.26 

A = 1.6 –a –t- c = 1.6 – 0.1 - 0.45 – 0 = 1.05 

D = W.N.S.F = 0.63 * 0.57 * 1.63 * 0.98 = 0.57 

R = P/ A*D = 10.39 

P’ = q.i.e.v.z = 1.5 * 1.3 * 1.46 *1.07 * 1.10 = 3.35 

A’ = 1.6 –a – t – r = 1.6 -0.1 – 0.45 – 0.7 = 0.35 

D’ = N.U = 0.57 * 1.48 = 0.84 

R’ = P’/ A’ * D’ = 6.52 

These results reflect the disaster that happened. A number of improvements were proposed : 

reducing the fire load, smoke vents, fire proofing of the structure, more and better exits, 

better water supplies, sprinkler protection . These improvements were included in the 

calculation “ as it should be” with the following results: 

P = q.i.g.e.v.z = 1.1 * 1.1 * 1.87 * 1.46 * 0.87 * 1.10 = 3.17 

A = 1.6 –a –t- c = 1.6 – 0.1 - 0.22 – 0 = 1.28 

D = W.N.S.F = 0.90 * 0.95 * 3.23 * 1.22 = 3.37 

R = P/ A*D = 0.73 

P’ = q.i.e.v.z = 1.1 * 1.1 * 1.46 *0.87 * 1.10 = 1.69 

A’ = 1.6 –a – t – r = 1.6 -0.1 – 0.22 – 0.4 = 0.88 

D’ = N.U = 0.95 * 4.54 = 4.31 

R’ = P’/ A’ * D’ = 0.42 

107

FRAME Calculation examples book. - FRAME Fire Risk Assessment ...

You also want an ePaper? Increase the reach of your titles

Delete template?

Save as template?