The Unwanted Conundrum
In an education session panel preview, a host of experts takes on
the problem of unwanted alarms.
INTRODUCTION BY MARTY AHRENS
THE EARLY WARNING PROVIDED BY fire detection has provided countless numbers of people with alerts to pre-fire conditions, and with a few extra moments of critical escape time in the event of fires. But unwanted alarms are another matter.
In 2014, U.S. fire departments responded to almost 2.5 million false alarms, almost twice the total number of reported fires and five times the number of structure fires. Many of these were triggered by commercial monitored connections, including residential buildings such as apartment buildings, hotels, and dormitories. The share of alarms coming from occupancies with automatic fire alarms has increased over time as these systems have become more common.
Unwanted alarms have also taken an increasing toll on the nation’s fire service in the form of fuel costs, apparatus wear and tear, risk of collision and injury during response, and a growing complacency when responding to automatic alarms. By 2009, the fire service had had enough. The International Association of Fire Chiefs (IAFC) submitted 43 proposals for changes to commercial automatic fire alarms for changes to the 2013 edition of NFPA 72®, National Fire Alarm and Signaling Code. In final action on the proposals in 2011, the committee accepted nine of the 43 in principle, either entirely or in part.
In frustration, Jack Parow, chief of the Chelmsford, Massachusetts, Fire Department and president of the IAFC, contacted Ken Willette, who headed NFPA’s Public Fire Protection Division. Ken and other NFPA staff met with the chiefs to clarify NFPA’s process and to begin building better communication between the fire service and the technical committee. These discussions led to the Fire Alarm Response and Management Summit a few months later, which was probably the first time that many members of the alarm industry had been in the same room with firefighters who actually respond when alarm systems are activated. It was clear that the chiefs and the detection industry had different definitions of false or unwanted alarms. A detection system that activates in response to burned food, after all, is doing what it is designed to do. But fire chiefs don’t want to send a full structure-fire response to an incident that the occupant was able to handle before the fire department even received the alarm. Everyone gained a better understanding of the issues confronting the varied stakeholders.
As a result of the summit, NFPA created the Fire Service Guide to Reducing Unwanted Fire Alarms, a free tool available from NFPA. The Fire Protection Research Foundation (FPRF), through NFPA’s Fire Analysis and Research Division, developed a risk-based decision support tool to assist fire departments in managing unwanted alarms. Both efforts included input from the industry and the fire service. When the Fire Analysis and Research Division was developing the tool in 2012, we could not find any national data on the percent of commercial fire alarm activations that were real emergencies. We were dismayed to learn that, according to three different local sources, only 0.4 to 1.1 percent of commercial automatic alarms received were for actual fires, and many of those were already out by the time the fire department arrived.
Five years after the summit, unwanted alarms are still a problem. At the same time, we are learning more about how much faster fires can grow in modern homes and how much less time people have to escape. To address these parallel problems, a panel discussion has been organized for the Conference & Expo in Las Vegas that will offer a variety of viewpoints on the issue; the following perspectives are provided by most of the panel’s participants, as well as by NFPA’s staff liaison for NFPA 72. Instead of a conventional question and answer session, our panel will end with three questions for the larger group: What other work has been done on this topic? What are the most important research needs on this topic? What should/could NFPA do to help?
This is an ongoing conversation, and it’s important that you make your voice heard. The public input period for the 2019 edition of NFPA 72 is open until June 29. To add comments, visit the NFPA 72 document information webpage.
The industry perspective
Thomas P. Hammerberg, SET, CFPS
Technical director, Automatic Fire Alarm Association, Inc.
Unwanted alarms directly impact fire and life safety in many ways, and create a negative view of fire alarm systems by the authorities having jurisdiction (AHJs) and end users.
In my experience, both with the Automatic Fire Alarm Association (AFAA) as past president and executive director as well as in previous experience installing and maintaining fire alarm systems in the Orlando, Florida, area, I feel that the problem is not the fault of the equipment. The equipment is reliable if used properly. In too many cases, the industry experiences problems with poor design by unqualified individuals, poor installation by contractors, and lack of urgency on the part of owners to maintain the systems.
Most of these problems boil down to money. Engineers say they don’t have the budget to properly address fire alarms, and therefore leave many design aspects to the contractors. The contractor selected is often the low bidder working on a very tight schedule. Neither lends itself to quality and a reliable system. In addition, many contractors do not properly train their employees because they don’t see the value in it.
Owners must be reminded that they are responsible for keeping their systems in working condition and tested per NFPA 72. They may not legally be able to perform the work, but they are still responsible to see that it gets done.
There is also a lack of enforcement of inspection, testing, and maintenance (ITM) procedures of fire alarm systems for a growing number of reasons. Budget cuts and cutbacks on training are two big problems. The Automatic Fire Alarm Association offers free memberships to AHJs as well as free webinars and discounts on seminars to try to address the education problem.
I have found from my personal experience that proper ITM, along with a good working relationship with fire prevention personnel, can help solve problematic systems. Each group going in their own direction does not solve the problem. When trying to persuade end users to allow repair of faulty systems, I have had them tell me that when the fire department makes them fix it, they will deal with it then.
We need better information on problematic systems to determine the cause of unwanted alarms and a quick response to get the problems corrected promptly. Unfortunately, there is no simple solution. I believe that to find solutions, end users must be engaged in the conversation with AHJs, contractors, and designers. We need more research, but I do not believe we need more requirements in the codes. They are already there. We just need all parties to apply them properly.
The AHJ’s perspective
Anthony Apfelbeck, CFPS, CBO,
FM, CFO, FIFireE
Fire marshal and building official,
City of Altamonte Springs, Florida
As the fire marshal and building official of a medium-sized jurisdiction, unwanted alarms impact us in three major ways. First, there is the added risk exposure to our citizens that is created by unnecessary responses to these incidents and the loss of fire department unit availability while committed to the unwanted alarm call. Second, there is a direct and indirect cost for each of these unit responses to the jurisdiction and, ultimately, the taxpayer. Finally, unwanted alarms create a loss of fire alarm credibility with the building occupants, in turn encouraging them to ignore the fire alarm as being unreliable.
To reduce these impacts, we need to overcome a significant challenge, which is the lack of available evidence-based studies on effective prevention methods to reduce the number of unwanted alarms. That data shortfall was clearly shown in the unwanted alarm studies recently conducted by the Fire Protection Research Foundation.
Our local experience in attempting to develop such successful evidence-based studies has not proven to be very fruitful. As an example of our experience, the jurisdictions in Seminole County decided to test two hypotheses that could potentially reduce the number of alarm responses in our county through encouraging a greater use of inspection, testing, and maintenance (ITM).
The first tested approach was to institute earlier intervention by fire prevention bureaus and earlier notification of business owners in the hope that more timely ITM follow-up would reduce the number of responses. The second tested approach was the institution of a fee schedule for repeat unwanted alarm responses by the fire department as a method to encourage a greater ITM focus on unwanted-alarm-prone systems. Unfortunately, after completion of both tests, the data indicated that neither of these programs proved successful in reducing the number of unwanted alarms.
For the future, I believe we need to take a deeper look at the unwanted alarm issue with a much more granular study than we have been able to do with studies based on National Fire Incident Reporting System data and cause analysis conducted by fire departments. For example, let’s put two fire alarm experts in a battalion chief’s vehicle for a week in a number of large jurisdictions responding to these unwanted alarms. Let them study, in detail, why these systems are creating unwanted alarms on a system-by-system basis. Once we understand the root causes, we can use that data to determine the appropriate code, standard, and operational responses to combat the problem.
The industry perspective, part II
Executive director, Central Station Alarm Association
There is no easy answer to reduce unwanted alarms. Implementing solutions to the core causes impacts a wide range of stakeholders. A “whack-a-mole” singular approach will not lead to meaningful results. Firefighting resources are being unnecessarily dispatched to systems installed last month, as well as systems installed in the last century. For many of these dispatches, the system is functioning properly. A successful program needs to look holistically at the issue with a very wide-ranging view.
In 2009, Jeff Johnson, fire chief in Tualatin Valley, Oregon, and then-president of the International Association of Fire Chiefs (IAFC), requested that the Central Station Alarm Association work with the IAFC to tackle the unwanted alarm issue. I chaired the subsequent committee, which was composed of four fire chiefs and four tenured automatic fire alarm professionals. The committee’s efforts led to the fire chiefs making dozens of individual proposals for changes to the 2013 edition of NFPA 72.
Unfortunately, relatively few of the chiefs’ proposals were accommodated in the 2013 version of NFPA 72. I have had many code experts tell me it can take two to three code cycles to implement changes. If you do the math, that’s a six- to nine-year endeavor, not to mention the time it takes for municipalities to put a new version of the code in force.
While the eventual end result of the committee’s work was disappointing, some good came from the effort. The challenges to overcome became very clear. A need to enhance goal congruency between firefighting services and the code writers was clearly evident.
Code writing is by necessity a facts-based process, and there can be no argument that there is a lack of empirical data on the subject. Nevertheless, automatic fire alarm experts with decades of experience in the industry have a good understanding of many of the root causes for false dispatches. While not academic-level research, their learned knowledge is not anecdotal opinion.
Another primary challenge is dealing with the inaccurate perception that many of the false dispatch solutions are counter to the “time is everything” philosophy and therefore threaten life safety. Nowhere is that more apparent than the discussions about alarm verification. It’s a complex issue when timelines associated with detection technology, sprinkler activation, notification, and firefighter turnout and response codes are all considered.
And then there are the financial realities involved. Fire systems, like every complex device known to man, require maintenance to remain fully functional. Ordinances that require inspection, testing, and maintenance will keep systems performing properly. Ordinances that financially force end user behavior usually have a positive impact. A requirement for point identification for new installations will generate valuable data, aiding in short-term corrections as well as establishing the path to real empirical data.
Nothing worth doing is easy. Solving the unwanted alarm issue will enhance life safety as well as improve efficiencies for every stakeholder involved in fire protection. It takes a team.
The public education perspective
Fire and life safety specialist
Rapid City, South Dakota, Fire Department
As a fire and life safety educator, I am interested in what people do and don’t do when a fire alarm sounds. I’m hearing an increasing number of examples of people not evacuating when an alarm sounds. In most situations, there is not a direct threat to their safety and their decision not to evacuate can be justified. Sometimes I’m the one teaching them to not evacuate unless they have more information.
We teach people to put a lid on their cooking fire and use a fire extinguisher. If you are at greater risk of injury while evacuating, especially when an alarm is warning you of a neighbor’s burning food, you are taught that you may be better off staying put. This is especially true for older adults and people with disabilities living in high-rise apartments protected by fire sprinklers. It is only hindsight that tells us if we should praise the person who contained their cooking fire or blame the person who died attempting to fight a fire.
Even more daunting, as a public safety educator I’m up against human nature. Research has indicated that the more often someone experiences a nonemergency alarm activation, the less likely they are to evacuate during the next alarm. I cannot convince people to take every alarm seriously when it probably isn’t an emergency.
With all of this in mind, the Rapid City Fire Department set out to increase the credibility of our alarms. Our first step has been identifying why alarms are sounding. We are two years into our special study, reviewing over 2,500 incidents identified as either a National Fire Incident Reporting System False Alarm & False Calls (all incident types in the 700 series) or a Cooking fire, confined to container (incident type 113). So far, incident breakdown has been revealing: in 18 percent of incidents, the detector misinterpreted a substance as smoke, such as construction dust or an aerosol sprayed by the occupant; 18 percent were cooking smoke with no fire, damage, or need for rescue; 15 percent were supervisory or trouble signals (and we respond to many supervisory signals in an attempt to provide good customer service); 10 percent were unreported fire drills, unreported maintenance of the system, or a misinterpretation of the alarm code such as calling 911 for a trouble signal; 10 percent were caused by direct damage to the device or sprinkler line; and 5 percent were a malfunction of the system. In addition, 8 percent were emergency exits and pull station activations; 3 percent were carbon monoxide and other gas that was clear upon arrival, and other unwanted alarms; and in 13 percent of the incidents we could not determine the cause.
Those numbers illustrate the challenges we all face, and they especially highlight the problems that unwanted alarms can pose for public educators.
The fire service perspective
Deputy chief and fire marshal
Plano, Texas, Fire Rescue
Plano Fire Rescue is a lot like other fire departments across the county, and we struggle with the growing number of unwanted alarms each year. It is to that point that the International Association of Fire Chiefs (IAFC) began to take interest, and proposed numerous changes to the 2013 edition of NFPA 72. Most of the proposals were turned down. There was some frustration on the part of the chiefs, and NFPA hosted a summit to discuss the problem.
From that summit, two very important things emerged. The first offer by NFPA was to produce a guide for the fire service, and in 2012 NFPA produced the very beneficial Fire Service Guide to Reducing Unwanted Fire Alarms. This is a free download and it is very user friendly. The second project, taken on by the Fire Protection Research Foundation (FPRF), was to develop a risk-based decision support tool to assist fire departments in managing unwanted alarms. The tool was not designed to make any suggestions regarding response models or the number of personnel and apparatus to send, but it could assist with beginning the conversation regarding response options. The tool is based on concerns expressed, both positive and negative, regarding a strategy of alarm verification with a delay in response by the fire department. The proposal from the IAFC to allow a 90-second delay for verification prior to dispatch created debate among the fire service. What would the delay look like? What would the outcome be if we did delay in terms of life and property? Could we justify such an action to the public?
The FPRF gathered interested parties to provide input for development of the tool. I served as a member of the committee representing the International Fire Marshals Association. I have also been involved with beta testing the tool since it was introduced. The tool only considered commercial property types as identified by the National Fire Incident Reporting System (NFIRS), including commercial and business, educational, institutional, residential (multifamily), industrial, and storage. A minimum of 5,000 alarm responses is needed, which may cover several years. It was determined that a national data set would be needed to assist smaller departments. The tool looked at various parameters to establish time credits and penalties relating to each response strategy. The results of the model are expressed in dollar loss amounts for injuries and deaths (civilian and firefighter), and direct and indirect loss. Three basic strategies are reviewed for each property type, and the model produces four outcomes, including a baseline where no verification is considered; alternative one, where a failed verification calls for a normal or full response; alternative two, where a failed verification calls for no response; and alternative three, where a failed verification calls for a reduced response.
I ran the report several times and the results were astonishing. In our community, data was analyzed from 6,186 automatic alarm calls over a five-year period. The calls ranged from a high of 3,291 in the commercial category to a low of 131 in the industrial category. Of the 6,186 calls, 20 were coded as structure fires in NFIRS. This resulted in a 99.7 percent nonfire situation. Of the 20 reported fires from automatic alarms in NFIRS, only two required any action by the fire department. As a result of this information, we changed the response model for Plano Fire Rescue and passed an alarm ordinance requiring alarm verification by the monitoring provider prior to notifying the fire department of an alarm condition.
The unwanted alarm problem is real and is not being resolved quickly enough. The fire service risks the lives of the pubic and firefighters every time they respond to a call; the U.S. Fire Administration reports 17 percent of firefighter deaths occur responding to, or returning from, a false or unwanted alarm. The tool is by no means perfect and needs more evaluation and work, but it is a great start toward addressing these kinds of problems. It is my hope that others will try the tool and provide their input as well.
The NFPA perspective
Richard Roux, staff liaison,
There are many types of unwanted alarms—that is, any alarm that occurs that is not the result of a potentially hazardous condition—occurring from different scenarios. NFPA 72 breaks these down into four primary types: malicious alarms, the unwanted activations of an alarm initiating device caused by a person acting with malice; nuisance alarms, the unwanted activation of a signaling system or an alarm initiating device in response to a stimulus or condition that is not the result of a potentially hazardous condition; unintentional alarms, the unwanted activation of an alarm initiating device caused by a person acting without malice; and unknown alarms, the unwanted activation of an alarm initiating device or system output function where the cause has not been identified.
Unwanted alarms are caused by many factors, and there is no single factor that can be eliminated that will yield zero unwanted alarms. The reduction of unwanted alarms is both an equipment issue and a people issue. Since the invention of the fire alarm system, designs and components that make up these systems have proliferated, improved, and continue to improve.
So do code provisions designed to minimize unwanted alarms. As we moved into preparation of the 2013 edition of NFPA 72, the International Association of Fire Chiefs proposed dozens of changes to the code, some significant and some minor. The technical committe reviewed and acted on each of these, some of which were accepted as proposed or somewhat modified. These changes yielded more than 60 comments or revisions. Again, these were reviewed by the technical committee, and some significant changes were made in the 2013 edition of the code.
Among these were changes in Chapter 10, “Fundamentals,” requiring supervising station operators and fire alarm system service providers to report to the authority having jurisdiction certain conditions of system impairment. Requirements for inspection, testing, and service personnel qualifications have been updated to better reflect the level of qualification needed for each type of activity. Changes have been made in Chapter 21, “Emergency Control Function Interfaces,” to address requirements for elevator recall when sprinklers are installed in elevator pits.
Additionally, changes have been made in Chapter 26, “Supervising Station Alarm Systems,” to address alarm signal verification, alarm signal content, and restoration of signals. These changes have been made in part to help emergency responders better manage issues related to unwanted alarms. Changes have been made in Chapter 29, “Single- and Multiple-Station Alarms and Household Fire Alarm Systems,” to address the connection of sprinkler waterflow switches to multiple station alarms and to add new requirements addressing the smoke alarm resistance to common nuisance sources.
As the development of the 2016 edition of NFPA 72 progressed, many changes were made pertaining to the reduction of unwanted alarms. Requirements were revised that addressed manual and automatic initiating devices and alarm signal verification.
Reducing unwanted alarms continues to be both an equipment issue and a people issue. Modern systems themselves all contribute to a reduction of unwanted alarms—the vast majority of today’s nuisance alarms can be attributed to improper application and lack of maintenance.