According to an article in the NY Times, the security employee monitoring the smoke alarm panel had only been on the job for three days. When the alarm came in the employee contacted a guard in the main church area and sent them to check the alarm. The guard responded back to the security employee that there was no indication of a fire. Unfortunately, it took approximately 30 minutes before they realized that the security employee had sent the guard to the wrong building – the fire alarm sensor was located in the attic area, above the main church. To get there the guard climbed 300 narrow steps but by then the fire was beyond controlling with a fire extinguisher. At that point the guard radioed the security employee to call the fire department.
Some interesting facts contained in the article: The “ponderous response plan” underestimated the speed with which the fire would spread in the attic area; To preserve the architecture, no sprinklers or fire walls had been installed; The security employee had not been replaced at the end of his eight-hour shift so was required to work a second shift; The control panel displayed a complicated string of letters and numbers – ZDA-110-3-15-1, that was code for a specific smoke detector among more than 160 individual detectors and manual alarms in the church.
Given the type and quantity of combustible materials in the upper portions of the church, the lack of sprinklers and fire walls, and the need for a person to physically respond in order to validate the alarm, any proper risk assessment should have concluded that the threat from fire was a high probability, high consequence event. The article did not indicate if or when the response plan was ever tested. Based on the events described in the article, my assumption is that it was not.