Statistical tests were also conducted to examine variations in the number of calls for service by day of the week, but no statistically significant difference was found. This is an unusual finding since such differences were found for data collected by NBS for other cities. It was not appropriate to tabulate the distribution of service calls by month of the year because of the limited sample. 3.2 Calls for Service by Type The calls for service were divided into three types: 1. 2. 3. Public Service and Special Services Public service calls involve a non-fire and non-hazard related service such as replacing a flagpole rope or rescuing a cat in a tree. If a hazard exists which the fire department is specifically equipped to handle, such as draining a water-filled basement or washing down a street after an accident, the call is designated as a special service call. False Alarms If an alarm is given with malicious intent when there is no actual fire, 3 the call is considered a false alarm. Suspected Fires, Real Fires, and Accidental Alarms A suspected fire is an alarm given with good intent because the caller mistakenly assumed a fire from such signs as smoke from a stove grease fire or burnt bearings in a furnace fan. A real fire consists of smoke and flames. An accidental alarm occurs due to such mechanical malfunctions as a water surge in an automatic sprinkler system. The average number of calls for service per day by type during the three time periods was tabulated and is shown in Table 2. Table 3 shows the distribution of the number of days having a given number of calls for service of the various types. For example, there were 24 days during the 123 day period when two Public Service calls occurred. There were five days when there were zero calls for service. This information is presented graphically in Figure 6. 3 An alarm is considered to have been given with malicious intent when no fire is found and the caller has left the scene. 3.3 Calls for Service by Service Time Service time is defined as the total time that a company devotes to an alarm. This includes travel time to the scene, time spent handling the alarm, and return travel time. The concept of service time is useful since knowledge of average service time by type of call provides a quantification of the different demands on company time. In addition, both the average number of calls per day and the service times are needed for the analysis in Section 4. Table 4 shows the average service time for the various types of calls; Table 5 shows the frequency distribution of service times for all calls. Engine and truck companies are defined here as the resources available to handle alarms. An important factor is the amount of time that these resources spend handling alarms. Table 6 shows the number of calls handled by each company, as well as the number of minutes each company spent handling calls during the 123 day period. The last column in the table was derived in the following manner. The number of minutes from the previous column was divided by the total number of minutes in the 123 day period (i.e., 123 days x 24 hours x 60 minutes=177,120 minutes). The result is the percentage of work time spent actually handling alarms. These percentages are averages for the total 123 days. 3.4 Calls for Service by Location It is Figure 7 shows the geographic distribution of calls for service in Alexandria, Virginia on a grid of 1500 foot square cells. evident that the calls over the 123 day period tend to cluster. This clustering has possible implications for fire station location, assignment, and unit relocation policies. The distribution by location of false alarms throughout the city is shown in Figure 8. The geographic distribution of calls for service could be further specified by time period and/or by type of call. Changes in patterns of alarms could then be correlated to demographic or seasonal changes to provide information useful for suggesting or evaluating such proposed new procedures as tactics to reduce false alarms. There are indications that the incidence of fire alarms is associated with the socio-economic characteristics of different areas of the city. For example, the New York City Rand Institute [2] has found a high correlation between fire department and police department calls for service, with respect both to time of occurrence and to location. Table 4 Average Service Time (in minutes) by Type of Alarm The long average service times for accidental alarms occur because the fire department must wait for the representative of the alarm company, determine that there is no fire, and find the cause of the automatic system malfunction. ** Nine cases were not included because of insufficient data. |