MS in Fire Protection Engineering
College of Engineering
Frederick Mowrer and Christopher Pascual
This culminating project has been submitted as part of the graduate program in Fire Protection Engineering at Cal Poly. It documents an Assisted Living and Memory Care Center’s compliance with applicable fire safety prescriptions contained in the 2019 California Building and Fire Codes (CBC and CFC). Performance-based methods incorporating deterministic design fires were then used to verify that the final building design and operating procedures met the life safety needs of its unique occupants.
The building under analysis was a 45,000 sq. ft, two-story, 58-bed residential care facility for the elderly. Occupants were all 60 years or older without acute medical conditions but with potential mild to severe mobility, sensory, and cognitive impairments. The fire- resistance-rated light-frame wood structure, its compartmentalized interior layout, and its active fire protection systems were found to satisfy the code provisions adopted by the local authority having jurisdiction. These included plentiful egress and exit capacity, localized fire and smoke containment, early smoke detection, audible and visual notification at levels appropriate to the occupants, and complete quick-response sprinkler coverage for life and property protection.
The priorities of the performance-based analysis were to check the adequacy of these code-compliant fire protection features, as well as to support housing accessibility and to inform staff training. These required realistic fire models to verify available safe egress times (ASETs), which were shorter for these residents than the general population due to their lower tolerances for heat and smoke exposure. Design fires took guidance from NFPA 101 Life Safety Code and the author’s research on the history of fatal care home fires. All fires were placed in residential wings using heat release data from calorimetry tests of residential furniture and mixed natural/ synthetic hydrocarbon contents in staff supply closets. Initial growth rates were between fast (0.0469 kW/s2) and ultrafast (0.1876 kW/s2), with peak heat release rates and embodied energies appropriate to the fuel packages but ultimately determined by ventilation conditions.
Model results supported the existing building design but showed that additional fuel control, compartmentation, detection/ notification, and automatic suppression would strengthen care staff’s response to and management of fires. Specifically, all rooms that communicate with residential corridors should have smoke detection and be fitted with door self-closers, following the findings of Performance Design Fires ‘B’ and ‘C.’ Where clients are housed also impacts their fire safety, so their facility intake forms/ health assessments should be used to guide placement— per Performance Design Fire ‘A,’ Assisted Living residents with the greatest cognitive, sensory, and locomotion disabilities should be housed closest to the lobby to receive prompt aid and minimize burns and smoke inhalation. These vulnerabilities also mean that sprinkler protection should be designed following the more rigorous commercial NFPA 13 standard as opposed to low- rise residential NFPA 13R, which was demonstrated in Performance Design Fire ‘D.’
Performance Design Fire ‘A’ was a nighttime living room furniture fire typical of all 40 Assisted Living dwellings. The occupant was assumed to be sleeping in the bedroom and not intimate with ignition; they were also capable of self-evacuation. Their required safe egress time (RSET) included a delay in waking to their low-frequency smoke alarm and traversing their unit to the corridor door, which totaled two minutes. At this time, the visibility through smoke was well below what would normally be accepted for design. The gasses at six feet above finished floor in the egress path were already too hot to move through (120°C), so the evacuee had to stoop, crouch, or even crawl, depending on the effectiveness of the sprinkler suppression. Since the sprinkler did temper heat, the asphyxiant fractional effective dose for incapacitation (FEDtot = 0.1) became the limiting tenability criteria; an especially respiratory-sensitive evacuee who took longer to find their door would have been incapacitated at two and a half minutes, but staff was expected to intervene by then. The slim margin for human error suggests that this scenario would benefit from a probabilistic assessment that includes ignition and suppression. A deterministic solution would be to regulate the flame spread and heat release of the furniture that residents bring in or are provided with.
In scenarios ‘B’ and ‘C,’ a mixed cellulose/ plastics design fire was placed in staff supply closets with doors open to the residential hallways in the Assisted Living and Memory Care wings. The door in Performance Design Fire ‘B’ was self-closing, so wedging it open represented an n = 1 managerial failure; the closet sprinkler was operational. The nighttime RSET of Assisted Living residents to reach an adjacent smoke compartment was three to four minutes, depending on their disability. The ASET was the time for the smoke layer to descend to six feet in the corridor, which was the only evacuation route. This occurred by a minute and a half for 44% of the dwelling units along the hallway, which was the earliest staff was expected to arrive and close the fire room door. Since visibility at the staff entrance to the corridor was below two meters, and required crouching or crawling to access the room, closing the fire room door was not a certainty. This scenario necessitated partial or full defend-in-place in the Assisted Living wing.
A similar result was found for the Memory Care wing in Performance Design Fire ‘C.’ A faulty sprinkler was an n = 1 device failure in this scenario because the closet door was not required to be self-closing. Occupants with dementia/ MNCD were assumed to be incapable of self-evacuation, and an RSET was not calculated for full staff evacuation of the wing, but it would have been much longer than the minute and a half ASET it took for smoke to descend to six feet in most of the corridor.
Performance Design Fire ‘D’ looked at ignition within a Memory Care dwelling and NFPA 13’s requirement for sprinklers in clothes closets, which goes beyond NFPA 13R. This model also assumed an n = 1 device failure of the sprinkler. In contrast with Design Fire ‘A,’ the RSET was the time it took for an attendant to rescue the fire room occupant. This was just over a minute; since the fire was shielded from the main room sprinkler by the closet door, the fire burned uncontrolled, and the heat became intolerable overhead (200°C) after a minute and a half. This slim margin for attendant error echoes the conclusions of Design Fire ‘A.’
A summary of ASETs versus RSETs and additional observations can be found in Chapter 11. Facility operator responsibilities, including fuel control, housekeeping, fire protection systems maintenance, and emergency preparedness plans, can be found in the fire safety plan in Chapter 12. These are primarily based on the requirements of the CFC and the findings of this report's prescriptive and performance chapters.