Loading documentation...
Task 0 diagnostic and Task 1 terminology survey from project SI2.830108. Culled to DK/FI/NO/SE/UK/USA focus countries. Background to eu-firestat-final-report.md.
Closing data gaps and paving the way for pan-European fire safety efforts
| Field | Value |
|---|---|
| Project | SI2.830108 (European Parliament pilot project, DG GROW) |
| Published | 17 March 2021 |
| Contents | Task 0 diagnostic · Task 1 terminology · Challenges · Risk analysis |
Consortium (alphabetical): Efectis (lead), BAM, CFS-CTIF, DBI, Lund University, NFPA, University of Edinburgh, EuroFSA, VFDB.
Culled HeatWaves edition: country diagnostic sheets and appendices reduced to Denmark, Finland, Norway, Sweden, UK, and USA. Task 2 questionnaire and Comment Handling Document removed (see
eu-firestat-final-report.md).
HeatWaves edition notice: Markdown derived from EU FIRESTAT Progress Report 1 (March 2021, SI2.830108). Many country sections, annexes, and pages were removed for focus; figures and charts are not included (captions may remain; some chart axis/country label text from the PDF may appear as plain lines). For the authoritative full document and all graphics, use the published progress report and the July 2022 final report (DOI 10.2873/778991).
The aim of this pilot project is to map the terminology used and the data collected by the EU Member States regarding fire events, and to propose a common terminology and a method to collect the necessary data in each EU Member State with a view to obtain meaningful datasets (based on standardised terms and definitions).
This in turn would allow for knowledge-based decisions regarding fire safety at the MS and at the EU level regarding building fires (i.e. houses, apartment blocks, office buildings, commercial buildings, hospitals, schools and kindergartens, elderly homes, etc.).
This first progress report summarises the work carried out during the first twenty-two weeks (five months) of the project SI2.830108 financed by the European Parliament and commissioned by the European Commission – DG GROW.
This report contains the final report from the accomplished Task 0 and Task 1 and a progress report from the ongoing Task 2. Task 0 focussed on assembling all the knowledge of the consortium members regarding fire statistics, such as where fire data comes from, who is responsible for them, how the data sets are defined, how they are collected, analysed and what is the output of these fire statistics.
From there, issues and limitations have been identified, as well as the best practices in the investigated countries. The findings of this task will be used as preliminary groundwork for all the discussions that occur during the project and as an input for all the following tasks.
Task 1 investigated in depth the semantic differences used in the fire safety fields and the recording approaches for the information gathered.
Task 1 focuses on the terminology and data collection methodology of the 27 EU Member States as well as other European and non-European countries. Task 2 entailed developing a questionnaire that focused on fire data needed to provide meaningful datasets and allowing legislative and other policy decisions for the fire safety. The project has been communicated through different channels. A website dedicated to the project has been established (https://eufirestat-efectis.com/).
Due to the COVID-19 pandemic it has not been possible to organise physical meetings, however several web-based meetings have been held within the consortium and with different stakeholders.
More web-based meetings will be scheduled as the project advances in 2021. A list of challenges, a risk analysis, and a Comment Handling Document have been developed.
These documents are updated continuously during the course of the project. Finally, the consortium would like to thank all the interviewed persons and the stakeholders for their valuable input, support and commitment to the project.
This report provides an analysis and evaluation of the current state of fire statistics and data collection in European countries and selected non-European countries of interest. For each country, the project team created a detailed diagnostic sheet describing important components of fire statistics practice, with a focus on terminology, collection methods, interpretation issues and the state of existing data.
The information was gathered by the consortium members by researching in public datasets, literature review, and networks of contacts. For some countries, there are missing data, those will be added in the following tasks of the project.
The review of the literature shows that fire data collection systems have been instrumental in informing firefighting strategies, building codes, educational programs, and technical innovations, to cite just a few applications.
However, there is substantial agreement in the literature that differences between fire data collection systems in different countries complicate the ability to make comparisons that could be useful in evidence-based planning and prevention efforts. They are currently most useful in describing the global fire safety situation and certain fire-related trends.
The amount and quality of information in different data collection systems appear to be influenced in part by whether they include information from sources outside the fire service, such as insurers or medical authorities, through data linkage or other means.
Data collection systems that collect too little or imprecise information may not produce useful data. At the same time, it was also observed that overly detailed data collection systems may overwhelm data collectors and thereby compromise data quality.
In general, it appears that the fire data collection systems in most countries are presumed to provide an accurate representation of their respective experiences with fire incidents. However, information gathered through this initial project task suggests that those involved in data collection efforts may be unaware of important limitations of their data due to missing information, differences in the way terms are defined or interpreted, and other identified issues.
No European country appears to employ a methodology for dealing with missing data, although some countries do acknowledge that missing data is a potential problem that compromises data quality.
Additionally, none of the consulted reports included uncertainty estimations.
To provide relevant information regarding the national fire safety situation (number of fires, fire fatalities, fire injuries, fire losses, etc.), fire statistics will have to be improved through common terminology, common methodology, and common training and qualification of persons in charge of filling in the fire report, including uncertainty estimation methods.
The findings of this task will be used as preliminary groundwork for all the discussions that will occur during this project and as an output for all the following tasks.
This first task of the project is established to assemble all the knowledge of the consortium members regarding fire statistics. A special focus was set upon sharing feedback on practices of each country in Europe and at least in Canada, New Zealand and USA, about where fire data comes from, who is responsible for what, how the data sets are defined, how they are collected, analysed and what is the output of these fire statistics.
From there, common issues and limitations can be identified, as well as the best practices in the investigated countries.
As most of the organisations within the consortium are based in western and northern Europe, though covering a substantial part of European countries, naturally more information is available in these countries, whereas less information was accessible or available from the rest of Europe.
This has set an important challenge of gathering information through professional networks and through publicly available publication, despite the language barrier. In this report, a literature review of work on fire statistics is presented; those are separated into studies focusing on national fire statistics and studies focusing on international comparison between practices.
Publications of international organisations such as CTIF, ISO and WHO are also reviewed and considered as baseline of available international fire statistics. Finally, our findings about the investigated countries are presented in the different chapters of this report.
Some countries present deeper descriptions than others, this is either due to the fact that in some of them more substantial studies or feedback relevant to the project are available.
The findings of this task will be used as preliminary groundwork for all the discussions that will occur during this project and as an input for all the following tasks.
As part of the project’s initial work, the team undertook a comprehensive review of research into fire incidents and literature on fire data collection, as well as documents relating to fire data collection systems in member countries of the European Union, Australia, Canada, New Zealand, and the United States. In accordance with the project scope, our review of the existing literature focused as much as possible upon research and documents related to building fires.
Several reports from the World Fire Statistics Centre (WFSC) in Geneva provided background for the project team on basic recent indices of the fire experience at the international level [1]. These reports also offered important insight into how currently available data on fire incidents are used, as well as its potential limitations.
The WFSC’s information bulletin from 2011 reported findings from the United Nations that based on percentage of gross domestic product (GDP) calculations, direct losses from fire for the 2006-2008 period were generally stable, with most countries experiencing a slight decrease or holding steady [1].
Minor increases were observed in the United States, Finland, France, the Netherlands, and Poland.
Western and Central European countries were reported to compare well against the average situation in Eastern Europe and central Asia/Eurasian countries in the number of fire deaths per 100,000 population. Variance observed in the cost of fire service organizations between similar countries was attributed to different ratios of public, private, and volunteer organizations.
Notable differences were observed in the costs of fire protection for buildings between countries, reflecting varying requirements for different types and sizes of buildings. The report also noted that high fire death rates in Russia, Ukraine, Belarus, Moldova and the Baltic states were in decline, though still very high.
The high variance in fire deaths was attributed to inadequate fire protection services, to poor building construction and maintenance and to high level of cigarette and alcohol consumption.
Some of the Eastern and Eurasian countries were not members of the European Union and had not benefited from the possible advisory regulatory harmonization. The report emphasized that substantial differences in how fire data are collected and interpreted posed a critical challenge in using world fire statistics. Findings from the United Nations for the 2007-2009 period were reported in the WFSC bulletin on world fire statistics in 2012, with fire deaths in Eastern Europe and Eurasia reported for 2001-2009 [2].
Direct losses due to fire showed that most countries had very minor decreases or increases in their losses as a percentage of GDP.
Italy was an exception and experienced the largest increase.
Proportional costs for funding fire service organizations remained largely stable, with only minor variances over the previous year’s report, and absolute cost also saw relatively minor changes which generally took the form of small increases. Many countries continued to experience improved long-term trend in fire deaths, while Singapore showed the lowest proportion of fire deaths. Germany and Sweden continued to see slight increases in death figures for the three-year period of the report.
The fire mortality rates in Eastern Europe and Eurasia showed improvement since 2000, but remained significantly higher compared to the Western and Central European states.
In Estonia, Latvia, Lithuania and the Russian Federation, the reductions in mortality rates throughout the entire reporting period were dramatic, ranging from a low of twenty percent in the Russian Federation to a high of forty-two percent in Estonia. The WFSC’s 2014 report of fire analysis findings from the United Nations covered the period from 2008 through 2010 [3]. GDP calculations for direct losses were again reported to be stable or show a slight decrease, while showing decreasing costs in absolute figures.
Scandinavian countries suffered above average fire losses, seen as possibly a result of the harsh climate and higher percentage of buildings containing wood.
Most countries experienced small to noticeable decreases in per capita mortality due to fire for the reporting period, coinciding with decreases in deaths due to fire in 2010.
The proportional costs of funding fire service organizations were again reported to be generally stable, with minor variances from the prior year’s report, with the greatest variance being a noticeable decrease in costs in Japan. Absolute costs also saw relatively minor changes and frequently involved comparatively small increases, with the exception of the United States, which showed a noticeable cost increase in 2010.
Cost of fire protection showed significantly higher cost estimates for various types of buildings than earlier reports and raised important questions about appropriate methods of calculations.
The wide variation in building fire protection costs to some extent reflects differences in estimation assumptions and methods, but also differences in rates of construction activity within the larger economy.
Advocates of data collection emphasize that it is not an activity that is pursued for its own sake but should be guided by the practical goal of collecting useful information that can be applied to action. In the case of fire data collection, a number of studies were identified in the literature review that highlight the vital importance of information on fire incidents for efforts seeking to improve fire safety and guide interventions that reduce the human and economic cost of fire.
National fire incident databases were used in the majority of the research, but some of the studies relied upon other sources of fire-related research.
Fires occurring between 12 p.m. and 6 a.m., usually on early Sunday morning, accounted for most fire deaths. There was no working smoke alarm in eighty-two percent of fatal fires.
Research from the United States in 2006 used data from the National Fire Incident Reporting System (NFIRS) to examine fatalities and injuries in building fires during 1993 [5].
The research found that the number of injuries and deaths and proportion of deaths depend mainly on the extent of fire damage, area of fire origin, material ignited, and ignition factor, while the absolute number of fatalities depended upon the material ignited and form of the heat of ignition. High fatalities were found to be significantly influenced by the victim’s condition, location and activity at the time of ignition.
A report by Swedish researchers from the SP Technical Research Institute of Sweden in 2009 used the country’s fire data collection system to study arson fires between 1991 and 2007 [6].
In addition, detailed information from the municipal insurance company for the city of Gothenburg was used in studies of larger arson fires, including the role of technical equipment. Information from insurers was also used in identifying the costs of arson fires.
The research found that arson fires most often targeted cars and waste containers, while arson in buildings most often occurred in apartment buildings. School buildings experienced fewer arson incidents, but arson nevertheless accounted for approximately half of all school fires.
Insurance data indicated a higher number of school arson fires in Gothenburg than the national fire database, which the researchers attributed to the likelihood that the insurance company was likely to include minor fire incidents that did not require involvement of the fire service and that the fire service was likely to designate the cause of a fire as unknown rather than arson if there was uncertainty about the cause.
Insurance company investigations found that wood façades on school buildings were critical to fire development.
Single-story buildings with roof extensions beyond the façade also were found to constitute a risk, and inferior compartmentation of a building was associated with greater damage. Major arson fires could be reduced through restrictions on building access. A subsequent study from Sweden in 2012 elaborated an approach to understanding deficiencies in fire protection by drawing upon both statistical data and more qualitative data from case studies of school fire incidents in Sweden [7].
The study used Swedish fire statistics on fire cause, extent of the fire, and room of fire origin in order to identify the types of fires that produced the greatest damage in Swedish school buildings.
Data indicated that the most destructive school fires were those that were deliberately set during evening and night-time hours.
Fire investigation reports of these fires showed that an absence of fire detection and insufficient fire separation contributed to the ability of fire to spread along the façades and into the attic space of school buildings. All case reports showed that a fire was very difficult to extinguish without causing substantial damage to the building once it spread into the attic space.
A study from Poland in 2014 examined residential fires in order to assess safety levels for the country as a whole and for the city of Warsaw for the years from 2000 to 2012 [8].
The researchers geocoded and mapped residential building fire incidents on a 25-kilometer cartography grid for national incidents and a one-kilometre grid for the city of Warsaw. Administrative Districts were ranked to show the highest and lowest number of residential fires.
Fire causes were divided into three categories: human factor (caused directly or indirectly by humans), technical (caused by faulty devices, domestic appliances or other building structure defects) and unknown (unspecified). The research estimated that human factors accounted for sixty percent of residential building fires at the national level and seventy-three percent in Warsaw were caused by the human factor.
Technical factors were attributed to nineteen percent of national incidents and eleven percent of incidents in Warsaw, while twenty-one percent of national incidents and sixteen percent of Warsaw incidents were due to other factors.
A short communication analysed in 2017 the data collected by the Laboratoire central de la Police Prefecture in Paris (LCPP) concerning fires which caused casualties deceased on the spot for the years 2012, 2013 & 2014 [9].
The geographic sector includes Paris and its surrounding counties, constituting more than 10% of the French population. As this database is filled by trained staff that is called systematically when a fire causes one or several casualties, this data constitute one of the most reliable analyses in France, despite being limited to Paris and its suburbs. It was found that the origin of the fire was intentional for 19 deaths out of 124 (i.e. 15% of cases).
Out of the 108 fires that were analysed, the origin of the fire could be determined in 90 cases out of 108 (i.e. 83%). In residential buildings, for 88 fires having caused 100 deaths, 26 fires caused 37 deaths, i.e. 35% of the recorded deaths.
In four of the fires having caused 7 deaths, the flashover also reached the building staircase.
In 43 fires having caused 44 deaths, i.e. 41% of the total number of deaths registered, the fire was limited to the room where the fire started: most often in a bedroom (20 fires having caused 20 deaths), then the living room and the lounge (10 fires having caused 10 deaths), last in the kitchen (9 fires having caused 9 deaths).
The origin of the alarm was registered for 92% of the fires analysed: the alert was given by someone who neither was a victim nor lived on the premises in 90% of the cases registered; it was given by a victim or a resident in 7% of cases; smoke detectors gave the alarm in only 3% of the cases.
A significant number of death occurs during the day, when people are not usually asleep, with 44% of death from 9am to 9pm.
Knowing that 94% of the victims were found in their homes and that the home occupation rate is much lower in the daytime, this goes against general opinion that fires kill in majority when people are asleep. The gender of the deceased is unknown for 8 victims. Males represent 58% of the deceased of known gender, females 42%. Research from Australia in 2015 by Xiong and co-authors examined risk factors related to surviving or dying in residential fires [10].
The study used The Victoria University Fire Fatality Coronial Database, which includes fatalities from the states of Victoria, New South Wales, and Queensland, to identify fire deaths.
Survivors of residential fires were drawn from the Victoria University Residential Fire Survivor Database, which includes data on fire survivors from the state of Victoria.
Researchers removed deaths due to suicide, murder, deliberate fires, undetermined cause, fires in non-residential premises, or incidents with more than one victim, creating a database of 177 single-fatality fire deaths. The database of survivors included 183 survivors following the removal of missing data.
The researchers found that the leading risk factors associated with fatal fires were psychotropic and sedative drug intake, discarded cigarettes, living alone, being over seventy years of age, being asleep, location in the room of fire origin at time of ignition, and alcohol intake.
Risk factors most significantly associated with surviving fires included cooking fires, electrical fires, involvement of stove in ignition, and fire occurrence in a one- or two-family dwelling.
Fire survivors were more likely to wake up to nonsmoke alarm cues, such as smoke or breaking glass, than a cue from a smoke alarm.
The research utilized data on the causes of residential fires and deaths, fire death trends, fire death rates by age group, and victim location to compare risk indices. The United States was found to have much higher fire death rates than the United Kingdom and Japan in almost every age group.
Japan experienced the highest death rates among those 80 years or older, and it was also found that older or handicapped victims tended to be closer to the area of fire origin, suggesting that early detection was insufficient for such victims as evacuation probably had a key role. Common patterns for fire risk were shown in the United Kingdom and Japan. Further study involving international collaboration was identified as an important means for gaining a better understanding of fire risk.
A more recent study by the United States Fire Administration from 2011 examined fire death rate trends in the United States and twenty-three other industrialized countries [12].
The research found that although the USA had made substantial progress in reducing fire deaths since the 1970s, its fire death rate was the tenth highest among the 24 industrialized nations in 2007.
Fire death rates per million population consistently fell throughout the industrialized world from 1979 to 2007 but fell faster in North America and Eastern European regions than in other regions. From 1979 to 2007, the fire death rate in the United States declined by 66 percent. Fire death rates in Japan, a leader in fire safety, showed a slight worsening over the study period.
The study identified a variety of factors that might contribute to differences in fire death rates, including differences in fire prevention practices and education, building practices and regulations, differences in lifestyles and cultural attitudes, and the proportion of senior citizens in the population.
The authors note that one of the difficulties in determining the cause of different fire death rates in different countries is a lack of available data.
Increased cooperation is identified as an important step for learning more about the causes of fire deaths and developing prevention measures. Still more recently, a 2016 study used fire statistics from the United States, the United Kingdom, Japan and Finland for the 2002–2012 period to compare the number of deaths and injuries from structure fires started by a small open flame ignition source to those from fires started by smoking materials [13].
The research found that Japan had substantially more fires ignited by smoking materials than the other countries, but not more fatalities from these fires, while the problem of deaths from fires started by smoking materials was greatest in the United Kingdom.
A more detailed quantitative analysis compared the relative contributions of low-energy ignition sources (match, lighter, space heater, etc.) to ignitions by smoking materials ignition to the USA fire problem. The analysis found that fires ignited by smoking materials more commonly resulted in fatalities than low-energy ignition fires, while the overall volume of low-energy ignition fires and corresponding losses and injuries are greater.
Finally, a logistic regression model indicated that older persons were at greater risk of death from fires ignited by smoking materials than by low-energy ignition sources.
A 2016 study from Lund University in Sweden sought to understand why more people were killed in residential fires per population in Finland than in Sweden [14].
Using reports and fire databases in the two countries, the research found that fire victims in Finland were more likely to be male and that they were often single. A higher share of victims in Finland were under the influence of alcohol or drugs than in Sweden.
A research of possible risk factors for fire death found that alcohol consumption, tobacco use, and the proportion of the population living alone were all higher in Finland than in Sweden and were seen as potentially contributing to disparities in the death rates in building fires between the two countries.
Another 2016 study from Sweden reviewed fire fatality data from 42 countries, using data from World Fire Statistics reports and from the International Association of Fire and Rescue Services [15].
The research found that reported rates of fire fatality experienced a decrease in most countries, but were particularly significant in Estonia, Germany, Great Britain, Latvia, Russia, and the United States. Fire safety interventions that contributed to the decreases in deaths included increased use of smoke alarms and sprinklers systems and fire regulations for children’s sleepwear and for furniture and furnishings.
The research noted that differences in how data is collected or how deaths due to fire are defined could explain differences in death rates between countries, but the focus of the study was on trends within countries rather than a direct comparison between them.
Comparative data from thirteen countries was used in a 2020 Swedish study of arson in schools and preschools in order to better understand the apparently high rates of school fires in Sweden and to assess school arson trends in the country over the preceding 20-year period [16]. The data indicated that Sweden experiences a higher frequency of school fires than most other comparable countries, with only New Zealand showing more fires per million inhabitants.
However, the study concluded that inter-country comparisons on fires are complicated by substantial differences in the way they are collected and stored, as well as how fires or data elements are defined.
The report notes, for instance, that Sweden reports all incidents which receive a fire service response, even if the fire is already extinguished, while qualifying events in other countries may be more restrictive.
The report also points out that differences may also distinguish the classifications of different levels or types of schools from one country to another.
Due to the low level of consistency in how different countries record and obtain fire statistics, the Department for Communities and Local Government in the United Kingdom commissioned a study in 2011 with the European Union Fire Safety Network to examine and better understand fire data in member states of the European Union [17].
The research team relied upon an extensive literature review, telephone interviews with key stakeholders, and an on-line survey of national organizations collecting fire statistics in European Union states, as well as Norway and Iceland.
The research found fire data to be collected at national and state levels in order to track and inform policy measures, raise awareness about fire risk, evaluate the number of fires and deaths, and develop interventions.
However, the research also confirmed substantial differences between countries in numerous aspects of fire data collection. The research noted that countries used different ways of recording in-scope fires and fire-related deaths and injuries. While most countries recorded such key points as the number of fires or deaths, the recording of other critical factors, such as type of injury, information about fire safety systems, and size of the fire, was much more mixed and infrequent.
Such differences were recognized as complicating the ability to make comparisons in a number of key variables of interest.
The research determined that reconciling data between countries in its current form would be a significant task, but noted that respondents expressed support for developing comparable datasets.
The researchers emphasized that common datasets would facilitate comparisons that could be used to inform the development and effectiveness of fire safety interventions. Data collection of fire incidents in Canada at the national level has been inactive in recent years and two reports from the Canadian Centre for Justice Statistics published in 2017 review the state of the country’s fire data collection [18]-[19].
One report summarized preliminary observations from the National Fire Information Database, a pilot project designed to: collect ten years of microdata information on fire incidents and fire losses from Fire Marshals and Fire Commissioners Offices across Canada, create a standardized national system for the collection of fire statistics, and link the database with other relevant socio-economic data to assist in the development of new, relevant, evidence-based research related to fire incidents, public safety and security.
Seven of Canada’s provinces and territories participated in the project. The report found that this participation covered seventy-two percent of the Canadian population.
Total reported fires declined by twenty-five percent between 2005 and 2014.
The share of structural fires in the database rose from half of all fire incidents to more than sixty percent of fire incidents over that period, but the number of structural fires declined by twenty-six percent. Three-quarters of structure fires were in residential structures. Cooking equipment and smoking materials were the leading cause of structural fires. Fire-related deaths rose between 2004 and 2008 before declining by thirty-two percent between 2008 and 2014.
Most fire-related deaths occurred in structural fires (87%), followed by vehicle fires (11%) and outdoor fires (4%). Smoke inhalation was the most common cause of death.
The second report from Canada evaluated the pilot National Fire Information Database project, examining the data collection system in relation to: the development of content (taxonomy), data collection, standardization of data, geo-coding activities and the addition of social domain data, and aggregation of jurisdictional files [8].
The report pointed out that although seven of thirteen provinces participating in the project represented seventy-two percent of the Canadian population, the database should ideally represent full coverage of every jurisdiction. In addition, the content of reporting data elements was found to vary between fire service participants.
Some jurisdictions had limited reporting capacity due to financial or other constraints.
Certain data elements were argued to require updating in order to reflect developments in technology (i.e., cell phones, laptops) or language conventions. A variety of data quality issues and data limitations due to underreporting were identified as issues that required resolution. The report argued that harmonization of the contents of reports and greater standardization of data was needed, and longer and more flexible data collection periods were proposed as one option for improving participation in reporting.
In order to improve the capacity to provide linkage to other data sets, standardization of data elements for victims and geographic identifiers is recommended.
The report concluded that the sustainability of any forthcoming national fire data collection system would require interest and support of stakeholders and funding support.
Several studies were identified which examined data quality issues with the US National Fire Incident Reporting System (NFIRS) [20][23].
A 2012 study by Butry and Thomas from the National Institute for Standards and Technology evaluated the ability of NFIRS data to depict fire activity in non-reporting US cities, to test differences between those cities which reported fire incident data and those which did not, and to discuss how findings could be used to weight NFIRS-based statistics in order to produce more accurate national estimates [20]. The report analyses data from NFIRS from 2002 to 2009.
In addition, the report used socioeconomic data from the Census of Population and Housing (income, sex, race and housing items, such as the status of a housing unit, median age of units, and median value).
Data from the National Fire Department Census were used to include information about fire departments listed with the US Fire Administration. The research found that there were differences in socioeconomic and fire department characteristics between those cities that submit reports to NFIRS and those that do not.
The authors argued that if those factors also affected the risk of fire, then generalizations made about fire safety and risk based only on NFIRS data will not apply to non-reporting areas of the United States.
A second study of NFIRS data by the National Association of State Fire Marshals examined the problem of unknown data to better understand why NFIRS data elements for area of origin and causal information were frequently reported as undetermined, unreported, or inappropriately coded as “none” [21].
The researchers conducted qualitative interviews, collected and analysed fire department data collection policies, and conducted an on-line survey of fire departments in seeking to understand the problem of unknown data. They found that information containing causal information from fire investigations was often not included in NFIRS incident reports because the reports were not updated after the information became available.
Liability concerns also influenced decisions about how much information to report in fire incidents.
The researchers suggested that another factor could be insufficient training of firefighters about the importance of data collection, while the reporting system itself was seen to be overly complex and not to be user-friendly.
They concluded program managers at the state level could be valuable resources for data quality by interacting with fire departments, supporting fire department participation, and providing quality control and feedback. “Data champions” within fire departments who recognize the value of data were also seen to be important resources for data quality efforts. The authors also argue that future iterations of the NFIRS system should strive to improve user-friendliness and reduce confusion about appropriate codes.
A 2016 study of NFIRS data by Kinsey and Ahrens examined the NFIRS three-digit coding system for types of fire incidents [22].
In this coding structure, the first digit of an incident type code is a broad category, the second is somewhat narrower, and the third provides the most detail.
The researchers compared NFIRS narrative information from three large municipal fire departments with corresponding codes for type of incident, property use, and actions taken by the fire department. The authors independently reviewed the information provided and assigned their own incident type codes to the records. The authors found that they were most likely to agree at the broadest code level, but results were poor at the three-digit level.
Agreement between at least one author and the incident report was about the same as the agreement between authors. More than seventy-five percent of incidents used just ten percent of the codes, suggesting that many codes get forgotten.
The authors pointed out that long lists reduce accuracy and that too many choices can lead to no decision or a default choice.
A standard research procedure is to measure broad categories first, then use follow-up questions for details. The research concluded that the NFIRS coding structure was based upon the researcher preferences rather than taking account of the way firefighters think and process information. A more recent study of NFIRS data quality by the United States Fire Administration from 2017 assessed data quality and data usability for required data elements in NFIRS reports [23].
Data quality for each data element was indexed on the total proportion of valid entries for that element. Data usability for each data element was indexed on the proportion of valid known entries for that element.
Both metrics were assigned a maximum value of 100.
The report found that required data elements (Incident Type, Property Use, Incident Date, Alarm Time, Deaths, Injuries, Property Loss, Contents Loss, Area of Fire Origin, Heat Source, Item First Ignited, Cause of Ignition, Factors Contributing to Ignition, Equipment Involved in Ignition, Fire Spread, Presence of Detectors, Detector Operation, Detector Effectiveness, Presence of Automatic Extinguishing Equipment) had either or close to a perfect value of 100 for the overall data quality index.
Lower scores were recorded for outside fires, fires with injuries, and the presence of detectors and automatic extinguishing systems in fires other than buildings and mobile property structures.
The data quality index for fires with injuries was in the low 90s due to the number of injuries that occurred in outside and other fires where alternate reporting methods were allowed, and reporting requirements were less stringent. Fire injuries occurred frequently and across all types of fires.
The lower data quality index for the presence of detectors and automatic extinguishing systems in fires other than buildings and mobile property structures was largely a result of deaths and injuries in vehicle incidents and outside fires.
Research on fire prevention measures can be useful in informing policy decisions and educational outreach. Three studies reviewed by the research team used fire incident data in order to examine and evaluate prevention-related interventions. A study by the University of Surrey in 2005 argued that detailed fire statistics in the United Kingdom facilitated the ability to evaluate the potential benefits of fire safety design features of consumer products or residential fire safety equipment [24].
The study used fire data to show that prior to the 1990s, the use and effectiveness of residential smoke alarms in the United Kingdom was low and referenced findings from the United States indicating that a decline in the country’s fire deaths was strongly correlated with the increased use of residential smoke alarms.
The study went on to conclude that it was important for Europe to improve its fire data collection practices in order to establish high levels of health, safety, and consumer protection. A 2009 study conducted by researchers from the Netherlands Institute for Safety (IFV) examined fire statistics in European countries and research on fire causes and development and found that domestic fires were more often caused by human behaviour than by technical failure [25].
The most common causes of fires involved smoking materials or cooking, while the presence of upholstered furniture or textiles were also contributing factors to fire development.
The research also emphasized the need to exercise caution in drawing comparisons between countries on the basis of fire statistics due to the absence of European standards for gathering, analysing, and publicizing data on fire statistics.
The report also cautioned that national fire data only include information reported to the responsible data bodies for data collection and are likely to substantially undercount the actual number of domestic fires. Smoke alarms have received substantial attention as a fire safety intervention that can save lives by alerting occupants to fires and providing opportunities to extinguish a fire before it grows or additional time to escape.
Research in the United States by Ahrens used data from the National Fire Reporting System and the NFPA annual fire experience survey to examine the presence and performance of smoke alarms in home fires [26].
Information about unreported fires was obtained from the US Consumer Product Safety Commission’s 2009 report, 2004-2005 National Sample Survey of Unreported Residential Fires.
The research found that the death rate in reported home fires with working smoke alarms was half that of fires without smoke alarms or with nonworking smoke alarms. Two of every five home fire deaths resulted from fires with no smoke alarms at all. Nine percent were caused by fires in which smoke alarms failed to operate, typically due to missing or dead batteries or disconnected power. Hardwired smoke alarms were more likely to have operated than those powered by batteries.
The thirty-seven percent of fatal fire victims with working smoke alarms were more likely to have been at least 65 years old, been in the area of origin when the fire began and unable to act, to have been engaged in fighting the fire than were victims without working smoke alarms. Victims with working smoke alarms were less likely to have been sleeping at the time of the fire.
The death rate per one thousand reported home fires was lowest in fires in which hardwired smoke detection and sprinklers were both present. A 2019 study by Shokouhi and co-authors conducted a systematic review of 30 studies of residential building fires, risk factors, and prevention measures in multiple countries, including the USA, Canada, the U.K., Denmark, Scotland, the Netherlands, Taiwan, Australia, New Zealand, and India [27].
The authors found that older adults, people with physical or mental disabilities, tobacco users, and people impaired by alcohol and drugs were at particular risk for fire-related injury or death.
Single-parent households, low-income families, and areas with large numbers of young children were also identified as being at higher risk for fire-related injuries or death.
Key prevention measures were identified as environmental modifications, behaviour change, improvements in emergency medical services, and education.
More recently, a multi-national group of researchers reviewed a number of emerging trends that present challenges for fire safety and highlighted the critical role of data in driving fire safety interventions in response to these challenges [28]. The report identifies climate change, resiliency, sustainability, population growth, urbanization, and globalization as key developments likely to influence - and potentially exacerbate – the threat of fire.
Hazards related to these developments include the need for quickly-constructed buildings to address rapid population growth, adequate fire protection, and the threat of wildland fires and fires in the wildland-urban interface as a result of climate change.
The report calls for greater collaboration across disciplines and between countries on fire data collection in order to facilitate applied research and formulate interventions to improve fire safety. The paper calls for a common database of information in order to promote the ability to learn from experience and promote effective collaboration.
The fragmentation of fire statistics, lack of common terminology and uneven access to data in different countries are seen to be major challenges to meeting the emerging challenges.
A review of the literature provides strong evidence that fire data collection systems have been instrumental in reducing building fires and their associated deaths, injuries, and economic damage. The utility of information about these fires is apparent in the design of many fire safety interventions and policy initiatives. Data on fire incidents can inform firefighting strategies, building codes, educational and training programs, and technical innovations, to cite just a few applications.
It is logical to assume that safety efforts can benefit from strategies that have worked in other places.
However, there is substantial agreement in the literature that differences between fire data collection systems in different countries complicate the ability to make comparisons that could be useful in evidence-based planning and prevention efforts.
While national fire data collection systems are likely to share certain core features and to gather some fire incident data in common, there appears to be considerable variation in the type and scope of information collected, the way that data elements are defined and levels of detail they seek, as well as the types of training and resources dedicated to collection efforts. In addition, literature suggests that fire data are influenced by differences between data collection procedures and practices.
Some data collection systems appear to provide opportunities to update information that may not be available at the time an incident record is first created, such as the cause of a fire or deaths that occur some time after the incident.
The amount and quality of information in different data collection systems also appear to be influenced by whether they include information from sources outside the fire service, such as insurers or medical authorities, through data linkage or other means. Literature suggests that the issue of how much information to collect is an important area for consideration in the design of fire data collection systems.
Data collection systems that collect too little or the wrong kind of information may not produce data that are useful, while overly detailed data collection systems may overwhelm data collectors, and thereby compromise data quality, as suggested by studies from the United States.
In many respects, the issue of how much information to collect appears to be driven by available resources, as well as the capacities of data collectors, who mainly are fire service personnel, to collect and record information. Concise data collection records will require less support and fewer resources than those that are more complex.
To that end, recent literature on fire data collection in Canada emphasizes that such factors as funding, resources, personnel, and stakeholder acceptance are critical considerations in the design and sustainability of national fire data collection systems.
Whatever model of fire data collection system is employed in host countries, available literature suggests that closing the gap between the data needed for drawing comparisons and information that is practically accessible will remain a challenge, but one with substantial public benefits.
The World Health Organisation (WHO) maintains a wide range of data collections related to global health and well-being, as mandated by its member countries. These databases contain not only information on diseases, such as malaria, influenza, tuberculosis, cholera, but also data on nutrition, health expenditure, road safety, etc. [29][30], and they have some utility for examining fire-related injury outcomes. Data for “burn deaths” offer one potential source of information.
The most recent fact sheet with respect to burn deaths was published in 2018 [31], while the most recent official publication and evaluation is a 2008 report outlining a plan for burn prevention and care [32].
It is important to note and observe that the databases for burn deaths includes all type of burn injuries and is defined as an injury to the skin or other organic tissue primarily caused by heat or due to radiation, radioactivity, electricity, friction or contact with chemicals [31]. “Thermal (heat) burns occur when some or all of the cells in the skin or other tissues are destroyed by:
SDRs are an age-standardized death rate calculated using the direct method and standard European population structure, while CDRs are calculated as a simple ratio: the number of registered deaths/ mid-year population (per 100000).
The mortality rates have been calculated by the WHO Regional Office for Europe using the data on deaths by cause/age/sex and mid-year population by age/sex, annually reported to WHO by European Member States. It should be noted that mortality rates for some countries may be biased due to the under-registration of death [32] [34].
The parameters include deaths from exposure to smoke, fire and flames and are consequently a broader classification than exposure to fire. Following groups are available on the WHO website:
2 shows the averages available for a variety of country classifications. The MDB has several visualisation tools available on its website [32].
Figure 1: SDR data, Death per 100 000 population based on European WHO database for a selection
of European countries (retrieved from [33]).
Figure 2: SDR Death per 100 000 population based on European WHO database with averages for
group of countries (retrieved from [33]) Information on burn death trends for the same selection of European countries is available from Gapminder, an independent educational non-profit Swedish foundation which uses data from the World Health Organization and other sources [33], and is shown in Figure 3 below.
Figure 3: Trend of Burn Deaths per 100.000 for a selection of European countries and the USA
retrieved from [35] The figures from both WHO Europe and Gapminder show peaks influenced by specific large fires, including the 1998 Gothenburg discotheque fire in Sweden and the 1994 Switel fire in Belgium). The figures nevertheless show that burn deaths have followed a declining trend for the last two decades followed by a stabilisation. It is important to acknowledge that the databases currently have limited utility in examining fire-related outcomes because they lack sufficient detailed fire data.
However, the power for analysing data can be noted, contingent upon data quality. More information can be found in the paragraph 8.2.2. on ICD codes.
As fire is a very complex phenomenon involving many different parameters, achieving a complete and precise understanding – including causes and circumstances of the fire, casualties, and damages, as well as firefighting issues – needs relevant, efficient, and complementary indicators. These indicators may be found in fire statistics. In many countries, provinces, municipalities, and cities, fire departments regularly collect important fire-related statistics.
These statistics are frequently published and increasingly available on official websites.
Although comparisons from one country to another are frequently complicated by a lack of common definitions and methodology, they can still be useful to describe the global fire safety situation and trends for groups of countries, as well as the specific fire safety situation in a given country.
Data collection in some countries is performed by private bodies, such as insurance companies, and these data are generally not publicly available. Several studies have used available data to make inter-country comparisons, such as the evaluation of fire death rate trends [29]. In this chapter, data from the “World Fire Statistics” reports [37], published annually by the International Association of Fire and Rescue Service (CTIF), were compiled and analysed for the years between 2009 and 2018.
CTIF collects (annual numbers) data from each country on a voluntary basis. Data are usually provided by fire and rescue services or firefighter associations.
In some situations, data are provided by two institutions, such as firefighters and national institutes. In these cases, data may differ for the same country.
In general, it was observed that the differences reported by the two main institutions range between 5% and 10%. There are also situations where countries do not provide data and we sought in such cases to fill the gaps where possible. We should also note that some countries only provide structure fires, while others provide structure fires, vehicle fires and wildfires.
CTIF recommends analysing data by decades (i.e. sequential average over 10 year period) and not by year in order to capture proper trends and to adequately consider potential outliers that may be present in the data.
If the data is analysed by year, then every big or exceptional fire event would cause some spikes in the trends, therefore the comparison year by year might be a bit more challenging.
This study draws upon CTIF data from European countries and select non-European countries included in the research pool, including Australia, Canada, USA and New Zealand. A total of 37 countries is included in the review.
In this analysis, the number of fires, fire deaths and fire injuries were divided by the average population for each country. The averaged population was calculated from the population numbers reported by CTIF over the period from 2009 to 2018.
Figure 4 presents the increase percentage of population from years (2009-2013) to years (2014-2018) and
shows that the difference is less than 5% for most of the countries. Only in Ukraine, Ireland, New Zealand and Sweden, the variation is between 6% and 8%.
Figure 4. Variation percentage of population from (2009-2013) to (2014-2018).
USA
Average number of fires per 100 000 inhabitants. Numbers of fires are shown in the countries whose data were available in CTIF reports [37].. Data for Luxembourg and Germany are derived from annual reports of these countries. In Luxembourg, the data found in the reports are significantly below the number which was given by CTIF [2]. This disparity could be due to a reliance upon different sources of information. However, CTIF had data for Luxembourg covering only one year (2018).
Figure 5 gives an overview about the average number of fires between 2009 and 2018 per 100 000 inhabitants (inh.). Large disparities in the numbers of fires can be seen between countries.
Those vary from around 50 fires per 100 000 inh. to more than 800 fires per 100 000 inh. in the case of Moldova and Cyprus, respectively.
Figure 5. Average number of fires reported per 100 000 inh. (between 2009 and 2018).
Averaged number of fires per 100 000 inh. (2009-2018) #### Number of fires per 100 000 inh.
Based on the averaged numbers over ten years, presented in Figure 5, the countries are divided into three categories:
| Category | NF per 100 000 inh. | Countries |
|---|---|---|
| 1 | > 400 | Austria, Bulgaria, Cyprus, Estonia, France, Ireland, Latvia, Lithuania, New Zealand, USA |
| 2 | 200-400 | Croatia, Finland, Germany, Great Britain, Greece, Hungary, Italy, Luxembourg, New Zealand, Poland, Portugal, Serbia, Slovakia, Spain, Sweden |
| 3 | <200 | Belarus, Czech Rep., Moldova, Norway, Romania, Russia, Switzerland, Ukraine |
Figure 6, Figure 7 and Figure 8 present the average number of fires per 100 000 inh., for the period (2009- 2018). In general, it can be observed that the NF is relatively constant or slightly decreasing. Most spikes in the graphs are related to an extraordinary event.
Figure 6. NF per 100 000 inh. for the period (between 2009 and 2018) – Category 1
Figure 7. NF per 100 000 inh. for the period (between 2009 and 2018) - Category 2
Figure 8. NF per 100 000 inh. for the period (between 2009 and 2018) – Category 3
Average number of fire deaths per 100 000 inhabitants. Figure 9 presents the average NFD per 100 000 inh. for the period (2009-2018).
Figure 9. Average number of fire deaths per 100 000 inh. (2009-2018).
Based on the results presented in Figure 9, the countries are divided in four categories:
| Category | NFD per 100 000 inh. | Countries |
|---|---|---|
| 1 | > 2 | Belarus, Estonia, Latvia, Lithuania, Moldova, Russia, Ukraine |
| 2 | 1 - 2 | Bulgaria, Czech Republic, Denmark, Finland, Hungary, Poland, Romania, Serbia, Sweden, USA |
| 3 | 0.5 - 1 | Croatia, Cyprus, France, Great Britain, Greece, Ireland, Norway, Slovakia |
| 4 | < 0.5 | Austria, Canada, Germany, Iceland, Italy, Netherlands, New Zealand, Slovenia, Spain, Liechtenstein, Luxemburg, Portugal |
A surprising observation from the data is, that for some countries with fewer fires (see Table 1) including Belarus, Czech Rep., Moldova, Norway, Romania, Russia and Ukraine, also have more fatalities. This raises a question about how fires and fire deaths are defined and counted. Figure 10, Figure 11, Figure 12 and Figure 13 present the NFD per 100 000 inh. for each year between (2009- 2018), for the four defined categories, i.e., Category 1, 2, 3 and 4, respectively.
In general, it can be observed that the NFD is decreasing for the Category 1 countries and remaining relatively constant for the Category 2 and 3 countries. All the spikes in the graphs suggest that an extraordinary event (larger scale) has occurred, such as the 2015 Colectiv nightclub fire in Romania (Figure 11). These spikes are mostly noticeable for countries with smaller population.
Figure 10. NFD per 100 000 inh. for the period (between 2009 and 2018) – Category 1
Figure 11. NFD per 100 000 inh. for the period (between 2009 and 2018) – Category 2
Figure 12. NFD per 100 000 inh. for the period (between 2009 and 2018) – Category 3
Figure 13. NFD per 100 000 inh. for the period (between 2009 and 2018) – Category 4
All figures dealing with fire fatalities must be considered carefully since the exact definition of a fire death is rarely specified in documents on fire statistics. Countries have different approaches in how to consider the time elapsed after a fire to record a fire fatality. Many of them, such as Australia, France, Russia, the UK, and the USA, do not fix any time limit for recording a fire death.
Canada counts a fire fatality as a result of injuries leading to death within one year and one day after the fire incident. In other countries, “death within 48 hours after fire” is regarded as a fire death. Additionally, some countries do not include fire deaths resulting from arson. France and Switzerland only report fire fatalities at the location of the fire – either those discovered by rescuers, or those declared dead after unsuccessful resuscitation attempts.
Therefore, the French and Swiss official database does not consider fire fatalities that occurred at the hospital or during transport of casualties to the hospital.
For instance, in France, the number of fire deaths was 6 deaths per million inhabitants in 2008, while figures derived from the French Institut National de la Santé et de la Recherche Médicale (INSERM) data are higher (9 deaths per million inhabitants in 2008) as they take into account not only the victims deceased on the spot, but also those who died in hospital. This is not the case in other countries like Germany, UK and Spain.
Therefore, it is possible that a significant number of people with fire injuries died in hospital or during transport, and so may not have been accounted for in the final statement, skewing the final result. It is surprising that the number of fires and the categories do not correlate with the number of fire deaths.
For instance, Russia, Belarus and Ukraine belong to the category 3 for number of fires per 100 000 inh. with less than 200 fires per 100 000 inh. but are in category 1 for fire deaths with more than 2 fire deaths per 100 000 inh. That suggests that fires in these countries are less frequent but deadlier – or it could be a result of different definitions of fires as explained previously. This will be further examined in Task 1 of this project.
Average number of fire injuries per 100 000 inh. Figure 14 presents the average NFI per 100 000 inh., for the period (2009-2018). One must notice that fewer countries reported the number of fire injuries than the number of fires and number of fire deaths.
Figure 14. Average number of fire injuries per 100 000 inh. (between 2009 and 2018).
Based on the average NFI presented in Figure 14, the countries are divided into three categories:
| Category | NFI per 100 000 inh. | Countries |
|---|---|---|
| 1 | > 10 | Czech Rep, Finland, France, Great Britain, Latvia, Lithuania, Portugal, Slovenia, Sweden |
| 2 | 6 – 10 | Estonia, Hungary, New Zealand, Norway, Russia |
| 3 | < 6 | Belarus, Bulgaria, Croatia, Greece, Italy, Moldova, Romania, Serbia, Slovakia, Ukraine, USA |
The term of “fire injury” is very difficult to define, and the differences between “minor injury”, “moderate injury” and “severe injury” are numerous between countries – indeed, much more so than for “fire death”. Having highlighted these caveats, the current fire statistics show important differences in trends between countries. Figure 15, Figure 16 and Figure 17 present the NFI per 100 000 inh. for each year between (2009-2018), for the three defined categories.
In general, it can be observed that the NFI are irregular for most countries. In our opinion, this can be due to a lack of definition of an “injury” or due to a change in the methodology, but can also be due to others parameters that we did not identify.
Figure 15. NFI per 100 000 inh. for the period (between 2009 and 2018) – Category 1.
For example, in the case of France, we can notice a steep decline in the number of fire injuries between 2015 and 2016. When investigating the French annual reports [38], we notice that the data for fire injuries prior to 2016 correspond to the number of injuries of “absolute emergency” and “relative emergency”, the former meaning that the injured persons needed medical care whereas the latter means that the injured person did not need medical care.
After 2017, the number of injuries reported in the CTIF reports by French officials correspond to the number of absolute emergency only, corresponding to the injuries that needed medical care.
Figure 16. NFI per 100000 inh. for the period (between 2009 and 2018) – Category 2
Figure 17. NFI per 100000 inh. for the period (between 2009 and 2018) – Category 3.
The data collection in each country is dependent on the national definitions and collection methodology. Therefore, general trends can give indications, however the direct comparison between countries is not recommended.
Here are a few of the most evident aspects, which should be taken into consideration:
This analysis raises questions on the sampling size of countries with small population and a reduced numbers of fires for which comparison with countries with larger datasets can create issues considering the difference in the statistical data populations examined. Indeed, a reflexion should take place on how to analyse data in these situations and how to properly use their outcome for decision making.
Finally, with the existing fire statistics, it remains difficult to derive useful conclusions from them, especially when wishing to compare the situation in different countries.
To provide relevant information regarding the national fire safety situation (number of fires, fire fatalities, fire injuries, fire losses), fire statistics will have to be internationally improved through common terminology, common methodology, and common training and qualification of persons in charge of filling in the fire report, including uncertainty estimation methods.
The work of ISO/TC 92 — WG13 (Fire safety — Statistical data collection) started in 2015 and resulted in two publications: an overview of national fire statistics practices [39] and a document proposing 85 harmonized definitions of terms commonly used in fire statistical data [40], further developed in Task 1.
The WG13 roadmap identifies upcoming ISO 17755 work: Part 3 on collection methodology, Part 4 on data interpretation, and an ongoing work item on intentional fires.
ISO TR 17755:2014 Fire safety — Overview of national fire statistics practices [39] assembles data on national fire statistics practices. The introduction states that, in the absence of international standards on such practices, ISO/TC 92 needs an overview of existing practices and their implications for fire statistical data, describing the worldwide situation on standardization of fire statistics.
The document concentrates on existing fire statistics and fire data collections. A general call to nations participating in ISO TC 92 was issued; ten countries (in alphabetical order) completed a survey distributed via ISO TC 92 WG 8: Australia, Canada, China, France, Japan, Kenya, Republic of Korea, Russia, United Kingdom, and USA. The survey appears in Annex A of ISO TR 17755.
The TR is organized into four parts:
Fire deaths (Clause 3). Countries differ in use of fire department vs medical records and in efforts to coordinate both into one database with consistent definitions. They differ on how long after injury a death counts as a fire death; delayed deaths may not reach fire authorities or may not be linked to fire by medical authorities. Treatment of fatalities from combined fire and non-fire harm (e.g. collision then fire) also varies.
Fire injuries (Clause 4). Similar cross-source coordination issues apply. Countries differ in treatment of injuries from incidents involving both fire and non-fire harm.
Victim characteristics (Clause 5). Most countries collect victim age and gender; many collect other characteristics.
Property damage (Clause 6). Reporting is generally limited to fires with fire department response. Some countries use minimum-loss thresholds; others recognize under-reporting of very small fires. Some permit simplified reporting for certain small fires. Monetary damage may come from fire departments, insurers, or both (e.g. USA). Non-monetary damage measures (counts of objects, area, confinement scale) are also common.
Other losses (Clause 7). Nearly all countries collect firefighter deaths and injuries from acute fire effects; most collect other on-duty firefighter casualties. Kenya alone reported chronic illness data for firefighters and collection of indirect property, environmental, and cultural-heritage damage.
Location of fire (Clause 8). Nearly all countries distinguish buildings, non-building structures, vehicles, commercial vegetation, other commercial outdoor locations, non-commercial vegetation, and other outdoor locations without commercial value.
Type of construction (Clause 9). Most countries collect construction type information.
Deliberately set fires (Clause 11). Most countries collect deliberately set fires using varying estimation approaches (e.g. “suspicious” cause, proportional allocation for unknown cause, trained arson investigators, inclusion/exclusion of non-malicious “playing with fire”). Several countries identify riot-related fires; some separate juvenile firesetting. Only the USA reported detailed arson motives (voluntary NFIRS Arson Module).
The UK incorporated homicide/suicide categories for deliberate fires; vital-records coding captures fire homicides and suicides not attended by fire services.
Natural cause (Clause 12). Most countries collect natural-cause fires; Australia and the USA reported sub-types.
Exposure fires (Clause 13). Most countries collect exposure fires; Australia, Canada, and the USA reported sub-types (Canada’s detailed reporting may be legacy).
Smoking materials and open flame (Clause 14). Most countries collect smoking or open-flame fires, often with partially specified categories that hinder separate estimates. Open-flame categories vary (torches, lamps, embers, novelty lighters, etc.).
Equipment involved in ignition (Clauses 15–19). Many countries collect heating/HVAC, cooking/kitchen, clothes dryer, entertainment, and office equipment fires. Office-equipment coding often lags technology (e.g. smartphones not distinguished).
Other ignition characteristics (Clause 20). Grouped into electrical fires, other appliances, item first ignited (form/function and material), ignition factors, first major fuel package, and fire-growth factors. Reporting depth varies strongly by country (China, Japan, Korea, Kenya, France, Russia as noted in the TR).
Mitigation (Clauses 21–25). Most countries collect automatic extinguishing and detection/alarm data. Manual extinguishers and smoke management are less common (Australia, Canada, UK report with detail). Compartmentation/construction information is widely collected.
Figures not included in this markdown edition: The published Task 0 report includes Figures 18–23 illustrating ISO TR 17755-1 survey themes. See the March 2021 progress report PDF for graphics.
The information gathered in Task 0 is the result of cooperation among consortium members who described fire statistics in several countries based on prior research, literature, public datasets, and professional networks. For each country, a detailed diagnostic sheet was completed (see Annex I); information from those sheets is summarized here and in the following sections. Sources and references for each country are typically listed in the corresponding diagnostic sheet.
Countries are grouped into EU Member States, other European (non-EU) countries, and international countries. Among EU Member States, Belgium, Cyprus, Estonia, Finland, Lithuania, Malta, Portugal, Romania, and Slovenia were not covered in Task 0 and were addressed in Task 1. Other European countries in Task 0 include Norway, Russia, Switzerland, Turkey, and the UK; international countries include Australia, Canada, New Zealand, and the USA.
The analysis covers collection methodology and fields recorded. Task 1 extended the country tables with definitions, methodology, fields covered, recording systems, and semantic analysis of building-related variables.
Task 0 overview tables below summarize, per country: responsible bodies, publication frequency and language, years covered, and main applications of the statistics. Fire statistics are usually published in the national language; in some countries partial datasets or reports are also available in English.
| Country | Language | Issuing body | Type of data collected | Years covered | Publication frequency | Definitions | Data origin | Collection methodology | Data usage |
|---|---|---|---|---|---|---|---|---|---|
| Austria | German | Sicherheitsinformationszentrum SIZ / Brandverhütungsstelle für Oberösterreich BVS / fire brigades | No systematic fire data | 2008-2019 | yearly | Yes, partly | Fire brigades, insurers, fire prevention associations | Ministry, fire brigades, others | |
| Belgium | Dutch | Home Affairs (Federal Government Service) | Fatalities and injured, date, response time, intervention characteristics, object of fire, building and human characteristics | 2014-2015 | Unknown | Yes | |||
| Bulgaria | Bulgarian | Ministry of the Interior - Directorate-General "fire safety and protection of the population" statistical and graphic information about the activities of gdpbzn-mbp | Number of accidents and fire and rescue equipment from 2015 to 2019. Fires with material losses, distributed by reasons of occurrence. Fires with material losses, distributed by industries. Statistics on fires, deaths and injuries. | 2015-2019 | Yearly | ||||
| Croatia | Croatian and English | Ministry of the Interior | Fire and explosions, killed, seriously injured, slightly injured, material damage (in local currency - Kn) | 2000-present | Monthly | ||||
| Czech Republic | Czech | Ministry of the Interior - General Directorate of the Fire and Rescue Service of the CR | Total fires, Number of fires per 1,000 people, Cause, type of activity, Direct damage in thousands of CZK, Protected values in thousands of CZK, Killed persons, Injured persons | 2010-present | Yearly (N-1) | Yes | National Fire Service of Czech Republic | ||
| Denmark | Danish | DEMA (Danish Emergency Management Agency) | Fatalities and injured, type of municipality, dates, response time, intervention characteristics, cause/room of fire, building and human characteristics, smoke detectors | 2011-present | Yearly | Yes | Fire brigades, DEMA | Fire brigade giving input to a database | Used to assess the fire brigade |
| Estonia | Estonian | Siseministeerium | Fatalities and injured, type of municipality, dates, response time, intervention characteristics, cause/room of fire, building and human characteristics, smoke detectors | 2013-2017 | Yes | ||||
| Finland | Finnish | Finnish Rescue Services | Fatalities and injured, date of reporting, response time, intervention characteristics, cause/object/room of fire, building and human characteristics, smoke detectors | 2010-2016 in different forms | Yes | Data on emergency calls | |||
| France | French | Ministry of Interior | Number of fire interventions, fire deaths, fire injuries, number of people involved for different types of buildings, firefighter fatalities | 2002-2018 | Yearly (N-2) | No | Fire departments | Data extrapolated for counties that did not collect data or partial data. | |
| Germany | German | Countries Ministries of Interior, DFV | fire interventions, fire deaths, technical and medical help, equipment used, size of fire, fire injuries firefighter fatalities | 1990-2017 | Yearly (N-3) | No | Fire departments | operation reports | Fire departments |
| Country | Language | Issuing body | Type of data collected | Years covered | Publication frequency | Definitions | Data origin | Collection methodology | Data usage |
|---|---|---|---|---|---|---|---|---|---|
| Greece | Greek | Hellenic Fire Corps | Type of incident (fire, assistance, etc), start time – ending time, duration of the incident, kind – type of staff participating in (staff of Hellenic Fire Service, military, volunteers, staff of foreign Fire Services e.g. earthquakes - Athens since 1999, etc.), vehicles, air means, kind of forest area land (e.g. forest, marsh, agricultural area, etc.), injured persons (personnel, citizens) | 2000-present | Yearly | Fire departments, General Secretariat for Civil Protection, Forest Offices per region, National Observatory of Athens | The Hellenic Fire Corps collects all information. Besides, General Secretariat for Civil Protection and Forest Offices per region collect specific information, which is required by their sections. | ||
| Hungary | Hungarian | National Directorate General for Disaster Management (NDGDM) - not publicly available | Number of fires, Number of victims (deaths, injured, rescued persons, missing persons), Fire causes, Fires by fire objects (buildings types, sectors of industry, etc.) | 1990-2019 | Yearly | There's a guide how to provide data from incidents | Local fire brigades | ||
| Ireland | English and Gaelic | Department of Housing, Planning and Local Government | Fire service, fire prevention, fire brigade activities, location of fire, causes of fires, fatalities from fires | Fire service from 2013 to 2015 (data also from 1996); fire prevention/causes of fires/fatalities from fires from 2000 to 2018; fire brigade activities from 2000 to 2017; location of fires from 2000 to 2016 | Yearly | No | Fire authorities | Every year the Department produces statistics about fire and other emergency calls dealt with by local authority fire brigades during that year. We also produce the fire death statistics for the year. The statistics are based on information supplied by fire authorities | Fire departments, governments, industries, public and private use |
| Italy | Italian | Corpo Nazionale Vigili del Fuoco (C.N.VV.F) | Fire incidents, location, causes, material ignited, response time, property type | Data from 2000 to 2018 | Yearly | Yes | Fire departments | STAT-RI - STATistica e Rapporto di Intervento. | Fire departments, governments, industries, public and private use |
| Latvia | Latvian | State Fire and Rescue Service - Valsts ugunsdzēsības un glābšanas dienests | Cause of fire, Number of fire, fires Victims, Destroyed buildings | 2013-2019 | Fire departments |
| Country | Language | Issuing body | Type of data collected | Years covered | Publication frequency | Definitions | Data origin | Collection methodology | Data usage |
|---|---|---|---|---|---|---|---|---|---|
| Lithuania | Lithuanian | Fire and Rescue Department under the Ministry of the Interior analysis of fire and rescue statistics | Deaths in fires, Number of fires and deaths by location, Statistics of children killed and injured in fires, Causes and places of deaths of people, Distribution of deaths in fires by age and sex, Number of fires and their dynamics, Main causes of fires, Number of people rescued in fires, Use of respiratory protection equipment, Performance of fires, Rescue operations | 2014-2019 | Yearly | Fire departments | |||
| Netherlands | Dutch | Fire Service Academy (IFV) | Fatal residential fire, cause, smoke alarms, materials, response time, location, rooms, characteristics of building/property/victim/fire and fire service actions. Report from 2016 regarding fatal fires is in English. | Data from 2008 till now | Yearly report, dashboard every quarter | Yes | Fire brigades of the 25 safety regions | Survey data from questionnaire sent to fire brigades, data quality check by researcher, nearly 100% response | Fire departments, government researchers, public reports from IFV, media, policy and educational |
| English | Central Bureau of Statistics (CBS) | Fire and non-fire incidents | 2016 till now | Access to public database | Yes | Fire brigades of the 25 safety regions | Extraction from emergency registration, not checked | Public | |
| English/Dutch | Central Bureau of Statistics (CBS) | Fire and non-fire incidents, number of alarms, response time, damage by fire, fires by cause and object, indoor (and chimney) fires, outdoor fires, victims, fire equipment, costs | From 1985-2013 | Yearly + access to public database | Yes | Fire brigades of the 25 safety regions | Survey data. Data extrapolated for counties that did not collect data or partial data. | Fire departments, government and public | |
| Poland | Polish | This information is found in literature | Fatalities, cause of fire and death, room of fire and victim, age, gender, building type, accidental fires | 2003-2011 | Yearly | Yes | |||
| Slovenia | Slovenian | Firefighter's Association of Slovenia | Fatalities, number of injured people, date of intervention, response time of firefighters, building type, cause of the fire, rough estimation of material damage, size of the fire, technical and medical assistance, used equipment for intervention, number of fire brigades required for intervention. | 1985-present | Yearly | Yes | Fire brigades are entering the data within the information system that is defined by the government. | Fire brigades are entering the data within the information system that is defined by the government. | Partially available to the public; however, some information is only accessible to the professional audience. |
| Country | Language | Issuing body | Type of data collected | Years covered | Publication frequency | Definitions | Data origin | Collection methodology | Data usage |
|---|---|---|---|---|---|---|---|---|---|
| Spain | Spanish | Fundación MAPFRE and Asociacion Profesional de Tecnicos de Bomberos | Number of interventions due to fires and explosions; number of fire and explosion victims. Distribution of fatalities by: Age, Gender, Month, Day, Hour, Region, Type of building | 2007-present | Yearly (N-1) | Yes | Fire departments | Working groups within fire departments, then ratification of all input then verification with Institutes of Legal Medicine | |
| Sweden | Swedish | Swedish Civil Contingencies Agency (MSB) | Fire incidents, location, causes, material ignited, response time, property type and much more | Database from 1996, statistics published in web portal https://ida.msb.se from 1998. | Yearly and continuously | Yes | Fire Departments | National data checked and quality controlled by MSB experts | Fire departments, governments, industries, public, private use and researchers |
| Country | Language | Issuing body | Type of data collected | Years covered | Publication frequency | Definitions | Data origin | Collection methodology | Data usage |
|---|---|---|---|---|---|---|---|---|---|
| Norway | Norwegian | DSB (The Norwegian Directorate for Civil Protection) | Fire and non-fire incidents, number of alarms, response time, damage by fire, fires by cause and object, indoor (and chimney) fires, outdoor fires, victims, etc. | Data from 2016-present | 70% registered automatically - 30% manual work, but published at least every 14 days | Unknown | Fire departments | BRIS | Fire departments, governments, industries, public and private use |
| Russia | Russian and English | State (federal) fire service EMERCOM of Russia | 100 parameters | <2008-present | yearly | Yes | Fire departments | Harmonised across the country | |
| Switzerland | French/German | the Association of Cantonal Insurance Institutions - Kantonale Gebäudeversicherung / Interkantonaler Rückversicherungsverband IRV | Number of fire incidents by day, months, and type of cases, and damage. Gender of fire deaths. Evolution of fire damage rates and the share of damaged buildings in CHF or in ‰. | 1990-present | Yearly and every 10 years | Yes, building type only | Insurances | Data is recorded on site by the damage estimators and communicated to the PIRE claims service. They are entered and managed electronically by the PIREs and transmitted once a year to the APIRE in a standardized format for the preparation of the "Statistics of APIRE damage". Data is from 19 of the 26 cantons covering around 80% of Swiss buildings. | For insurances, Government, public and private use |
| French/German | Unknown | Unknown | Yearly | Unknown | Fire departments | Fire departments | |||
| Turkey | Turkish | Istanbul Metropolitan Municipality | Fire incident types, structural/nonstructural fire types, building/property types and usage areas, response times, causes of fire. | Data from 2008 to 2018 | Every 5 years | Yes | Istanbul fire departments | Fire departments, governments, industries, public, private use and researchers | |
| Turkish | Civil Defence - Not available anymore | Types of fire, Number of fires, Loss of life (public, personnel, animal), real injury (in TL), cause of fire | 1990-2004 | Every 5 years | Unknown | Fire departments |
| Country | Language | Issuing body | Type of data collected | Years covered | Publication frequency | Definitions | Data origin | Collection methodology | Data usage |
|---|---|---|---|---|---|---|---|---|---|
| UK - England | English | Home Office | fire incidents, causes, consequences, fatalities/casualties, response time, alarms and automatic extinguishing systems | Most updated datasets from 2010/11 to 2019/20 | Yearly | Yes | Fire departments | Incident recording system (IRS). The Home Office filters the data removing potential errors and inconsistencies before publishing the datasets. | Fire departments, governments, industries, public and private use |
| UK - Northern Ireland | Northern Ireland Fire and Rescue Service | No publicly available datasets | Yearly | Fire departments | |||||
| UK - Scotland | Scottish Fire and Rescue Service | Incident type, property type, fatalities and casualties, fire stations and workforce | Data from 2009 to 2019 | Yearly | Yes | Fire departments | Incident recording system (IRS). The Home Office manages the IRS, though SFRS has access to the Scottish data. | Fire departments, governments, industries, public and private use | |
| UK - Wales | Welsh Government | Fire incident, location, cause, motive, casualties, response time, smoke alarms | Data from 2015 to 2019 | Yearly (April-March) | Yes | Welsh Fire and Rescue Services collect data. Welsh Government publishes the datasets | The Welsh Government compiles the statistics in this bulletin from reports submitted by FRAs to the Home Office. | Fire departments, governments, industries, public and private use |
| Country | Language | Issuing body | Type of data collected | Years covered | Publication frequency | Definitions | Data origin | Collection methodology | Data usage |
|---|---|---|---|---|---|---|---|---|---|
| Australia - Queensland | English | Queensland Fire and Emergency Service | Fire incident, fire causes, attendance time of fire brigades, type of building | from 2016 to 2019 | Yearly | Yes | Queensland Fire and Emergency Service | Information gathered by fire service | Fire departments, governments, industries, public and private use |
| Canada | English | Statistics Canada | Fire incidents, property type and casualty are available. The other fields are not publicly available. | Data from 2005 to 2014 | Yearly | Yes | The data were collected by the Canadian Centre for Justice Statistics (CCJS) in collaboration with provincial/territorial Fire Marshals and Fire Commissioners Offices across Canada. | National Fire Information Database | Fire departments, governments, industries, public and private use |
| New Zealand | English | Fire and Emergency New Zealand | Fire incident, location, duration | Last 7 days. A full incident report can be provided on request under the Official Information Act | Every day | No | Fire departments | Fire Awareness & Intervention Programme (FAIP) Survey Data | Fire departments, governments, industries, public and private use |
| USA | English | US Fire Administration, NFPA | Fire incidents, causes, associated losses and injuries by type of property; firefighter deaths and injuries | 1977-present | yearly for previous calendar year | Yes | Fire departments | incident reporting from fire departments; stratified random sampling of fire departments for fires, losses, injuries; census of firefighter deaths | Fire departments, governments, general public, industries |
Most fire services in the investigated countries centralise their data into one unique database. That is the case for at least Bulgaria, Croatia, Czech Republic, Denmark, France, Greece, Hungary, Italy, Luxembourg, Norway, Poland, Romania, Russia, Slovakia and Spain.
However, in France, Spain and other countries, due to the lack of official definitions and national fire statistical collection, it is possible that differences exist in the terminology adopted by the various fire departments when the data are gathered after attending fire incidents. More details are provided in the diagnostic sheets (see annexe I for each country).
Other investigated countries display differences with respect to how data is collected and maintained, which data elements are collected, and how variables are coded in different regions, states or even within a state.
Those are Australia, Austria, Canada, Germany and the Netherlands. Regarding the Netherlands, the regional registration methods have also changed for a number of regions between 2013 and 2014. This has had a particular impact on the 2013 figures for some regions. In Sweden, since different fire services report to Swedish Civil Contingencies agency (MSB) there will be some differences.
However, MSB has developed documentation and an online education to support individual fire and rescue services and reporters in the reporting [41]. Prior to 2018 there were three systems (Alarmos, Core or Daedalos) used by the fire and rescue services when reporting incidents.
Nowadays all fire and rescue services in Sweden report their incidents into the same system. In the UK, fire safety data are separate for England, Northern Ireland, Scotland and Wales.
English statistics appears to provide the highest number of fields publicly available after a fire incident while Scottish fire statistics do not have data on the quantification of damage and presence of alarms or automatic extinguishing systems. Welsh fire statistics have additional fire safety data also on fire causes and motive and only on smoke alarms. Finally, Northern Ireland Fire and Rescue Service statistics do not use the Incident Recording System (IRS) and do not publish comparable statistics.
In the USA, Incident reports and data elements are standardized and can be found online [42] and will be presented in Task 1.
However completeness of data entered into incident reports may vary by locality, creating differences in interpretation. For example, “Burnt food” may be considered a fire, excessive heat, a smoke scare, or a false alarm.
Similarly, many non-fatal civilian injuries are not captured by the fire service [43]. It is estimated that in the USA 21,174 of 48.202 civilian non-arson fire injuries resulted from residential or consumer-product fires attended by fire departments from July 1, 2002 to June 30, 2003. NFPA estimated totals of 18,425 civilian injuries in 2002, and 18,125 in 2003, including injuries caused by arson.
This means that some of the injured may have left the scene before the fire department arrived or been transported by a non-fire department organization.
Last but not least, in Switzerland, in addition to the Fire service database, the Public Insurance Companies for Real Estate (PIRE) collect fire data using a specific codification. Data are then gathered by the insurance association (APIRE) who analyses said data.
The latter collects data from 19 out of 26 Cantons; this covers 80% of the country’s buildings. Private insurances that cover buildings in the other seven cantons do not use this code.
In general, there are four main sources of fire databases in most countries, those are originated from fire services, medical field, insurance and police departments. Regarding the fire service data, in some countries like France, Germany and Norway, due to the current lack of official definitions of terms and expressions for fire statistics, it is most likely that differences and contradictions with the aforementioned domains exist.
While in other countries such as Germany and Sweden, it is clear that the fire service, police and insurers organise their data very differently, hence an overall picture cannot be built.
Particularly, classifications in fire statistics from insurance companies do not correspond to those used by the fire and rescue services. For example, one fire incident according to the fire services could result in multiple fire claims from the insurance.
Finally, there are discrepancies between fire service data and the medical data, which is usually based on ICD-10 coding of death certificates, detailed in the next section (8.2.2).
The World Health Organization’s (WHO) 10th Edition of the International Classification of Disease (ICD) is used by health care systems and coroners for medical records, billing, and death certificates. ICD is a global classification system and tool for different diagnoses. The primary purpose of ICD is to enable classification and statistical description of diseases and other health problems that result in human death or contact with the health care system in a country.
In addition to traditional diagnoses, the classification includes a wide range of symptoms, abnormal findings, ailments and social conditions.
The ICD system was first launched in the late 1800s and the World Health Organization (WHO) has been responsible for the maintenance of the system since 1948.
The current version of ICD is ICD-10; however, according to the WHO [44], the 11th edition (ICD-11) was adopted by the World Health Assembly in May 2019 and is scheduled to come into effect on January 1, 2022. The coding for lethality due to exposure to smoke, fire and flames is included in Chapter 20 in the 2016 edition of ICD-10 [45] and is expressed with the codes X00-X09 (see Table 11). External causes of injury codes on death certificates can be used to obtain fire death data.
Separate categorizations are used for unintentional, intentional self-harm, assault, and undetermined intent.
It is important to separate this coding from the section X10-X19 which is related to contact with heat and hot substances that are not related to smoke, fire and flames.
It is also important to note that Chapter 20 in ICD-10 is intended to be used as secondary to a code from another chapter, indicating the nature of the condition. Most often, the condition will be classifiable to according to Chapter 19, “Injury, poisoning and certain other consequences of external causes” [46].
| Code | Description |
|---|---|
| X00 | Exposure to uncontrolled fire in building or structure |
| X01 | Exposure to uncontrolled fire, not in building or structure |
| X02 | Exposure to controlled fire in building or structure |
| X03 | Exposure to controlled fire, not in building or structure |
| X04 | Exposure to ignition of highly flammable material |
| X05 | Exposure to ignition or melting of nightwear |
| X06 | Exposure to ignition or melting of other clothing and apparel |
| X08 | Exposure to other specified smoke, fire and flames |
| X09 | Exposure to unspecified smoke, fire and flames |
| X76 | Intentional self-harm by smoke, fire, and flames |
| X97 | Assault by smoke, fire, and flames |
| Y26 | Exposure to smoke, fire, and flames, undetermined intent |
Except for X01 and X03, it is not easy to separate building fires from non-building fires, since X04-X09 do not distinguish if the exposure occurs indoors or outdoors. An additional code addresses wartime fires. Y36.3, “War operations involving fires, conflagrations and hot substances.” Codes for different methods of terrorism were added to ICD-10 after the attacks in the US on September 11, 2001.
The fire-related code is U01.3 "Terrorism involving fires, conflagration, and hot substances.” Codes T20-32 “Burns and corrosions” identify burns on different parts or percentages of the body. Code T58 “Toxic effect of carbon monoxide“ and T59- “Toxic effect of other gases, fumes, and vapors” may be combined with external cause of injury codes indicating fire to obtain the share of fires with deaths from burns only, smoke inhalation only, or both.
In the USA, the Centers for Disease Control and Prevention (CDC) [47] have two query tools that access fatal injury data. WISQARS™ — Web-based Injury Statistics Query and Reporting System [48] is the simpler of the two. Users may select intent, and fire and burn injuries, fire injuries only, or residential fatal injuries. Although not shown on the query page, residential includes unintentional only. CDC’s Wonder [49] is harder to use but allows more flexibility.
Users may select multiple causes of death such as fire, with burns and or smoke inhalation.
The CDC provides additional information about what should be included in this category [50]:
In a Swedish study [51], three different sources of information on fire fatalities were compared, and one of these sources, the cause of death register, uses the ICD-10 coding system.
The study showed that the cause of death registry underestimates the number of fire fatalities by about 25%, and the authors argue that none of the single sources are sufficient to assess how many people actually die in fires. Since the introduction of the ICD system, classifications have been changed several times and this needs to be accounted for when comparing data over longer periods. Jonson et al. [52] have used the ICD data for Sweden to conduct this type of study of temporal trends.
The codes for unintentional fire fatalities from ICD-6 to ICD-10 are presented in Table 12.
| Years | ICD version | Codes for unintentional fire fatalities |
|---|---|---|
| 1952-1967 | ICD-6 | E894-E899 |
| 1968-1978 | ICD-8 | E890-E899 |
| 1979-1998 | ICD-9 | E890-E899 |
| 1999-2001 | ICD-10 | X00-X09 |
| 2002-present | ICD-10 | X00-X09 |
Finally, there are discrepancies between fire service data and the ICD-10 coding of death certificates. For instance, NFPA counted the fatalities in the World Trade Center and Pentagon attacks as fire deaths, while the death certificates called them terrorism. It can also be unclear to determine when vehicle fire deaths should be counted as transportation events rather than fires in the ICD-10 codes.
There are other cases where ICD- 10 codes do not show if they are due to fire or not, such as defenestration, suicide and collapses or collisions by a falling object. While ICD codes have their issues and limitations, they are widely used and usually managed by national statistic institutes. A way to determine the reliability of fire service data is to conform said data with those of ICD.
Fire departments are responsible for entering data with key details from the incidents for which they are dispatched, which in some countries include not only fires, but also emergency medical services, severe weather and natural disasters, and other incidents. Reports are filed for incidents and, if applicable, casualties (fatal and non-fatal). Firefighter casualties and civilian casualties generally utilize separate reports.
Fire departments are in most cases composed of civilian professional firefighters and civilian volunteer firefighters. In some cases, military professional firefighter units can be responsible for cities or areas (e.g. Paris and Marseille).
Volunteer, professional, military and mix usually participate in filling the fire response reports. In particular, they are filled and signed by the firefighter in charge of the operation.
The information is then collected by the fire department then is sent to the regional or national body (e.g. Ministry of interior), where all data are compiled. Since data is collected at local levels - which vary by resources, staffing, and leadership – there are substantial opportunities for disparities between jurisdictions related to the completeness and accuracy of data. Moreover, different fire departments provide different levels of quality control.
For instance, in the Netherlands, fire investigators report for deadly fires and the report is then reviewed, whereas, other fires are reported by firefighters.
It is certain that the difference in the level of training for reporting will result in different data quality.
That does not mean that firefighters need more training, but that as the main goal of firefighters is to rescue, tackle fires and other activities, they will have less time to spend on filling sheets during their shifts. Another striking example is for Austria where, the Austrian Fire Prevention Associations collect data from the police stations and insurers for each federal state and publish them yearly.
Missing data is a serious issue which compromises the quality and completeness of fire incident data. In the United States, the National Fire Protection Association (NFPA) employs a “national estimates approach” to correct for data that goes unreported in data collected by the National Fire Incident Reporting System (NFIRS), the national database of fire incidents sponsored by the United States Fire Administration (USFA).
No other countries appear to employ a methodology for dealing with missing data, although some countries do acknowledge that missing data is a potential problem that compromises data quality.
Several issues were identified in the current data reporting systems and publications; those are presented for each country listed below.
Not all countries have been included in this list due to a lack of information for certain countries.
There is a lack of definitions for collected terms, statistics fields, as well as a lack of training for the firefighters in charge of the fire response report.
The database from the fire brigade does not take into account the fire casualties occurring at the hospital or during their transportation to the hospital by emergency medical services (EMS).
Lack of training for the firefighters on how to fill information in the fire response report.
The police are responsible for the fire investigation and there is no feedback-loop into the database from their investigation.
The database from the fire departments is not taking into account the fire casualties, which are not reported or are reported in a separate database.
The lack of definitions for fields collected and expressions.
The lack of methodology to fill the gaps where information is missing.
The lack of training for the firefighters in charge of the fire response report.
The database from the fire departments does not take into account the fire casualties occurring at the hospital or during their transportation to the hospital by EMS #### Germany
No uniform fire statistic has been enforced in Germany. Hence, disparity between the practices in different regions makes it difficult to obtain an overview of the country’s statistics.
All fire service interventions and statistics are obtained with different criteria in Germany. Due to this situation, there is a lack of statistical information on extensive fire service interventions and reasonable statistical findings on fire service intervention, on the fire phenomenon and on the effectiveness of fire protection measures.
Much detailed information (building / first burning item / smoke detector us etc.) is missing.
The link between fire causes and fire consequences is missing.
The link between the different sources of data is missing. For instance, it is not possible to link data from insurer, police or fire service to gain information.
There have been issues with missing data when different systems (Alarmos, Core or Daedalos) were used at different fire and rescue services. There have previously been some double counting of incidents when several fire and rescue services from different municipalities/regions are involved. This has been corrected in the yearly quality control done by MSB. The routines at joint incidents have been improved in 2020 and MSB expect this to more or less eliminate problems with double counting.
Due to incident report content revisions and changes in local routines for recording data, there is a potential to over-interpret discontinuities in some time series.
It appears that a significant amount of fires is not recorded and not all Australian fire services contribute to the database.
Not all the fields are required or need to be completed and the compulsory fields are based on the nature of the call.
A number of tables contain a relatively high proportion of unknown values. Although these counts are removed from the calculation of proportions for other categories in the table, the proportion of known values is artificially inflated.
National database is incomplete, this is due to the fact that not all jurisdictions provide data for national consolidation and not all local fire departments provide data to the office of the fire commissioner or fire marshal.
Some jurisdictions refuse to report dollar loss. Some have policies that require causal information to be reported as undetermined when fires are referred for investigation.
To make it easier for firefighters, information about causal factors and details on fire protection are not required for six types of bulding fires, collectively called confined fires. These include confined cooking fires, confined chimney or flue fires, confined fuel burner or boiler fires (mostly oil burner blowbacks), confined compactor fires, confined incinerator fires, and rubbish trash fires in or on a building that did not extend to the building or other contents.
Some data elements were left optional. “None” is a choice in some data elements such as factor contributing to ignition and equipment involved in ignition that can seem like an easy out.
The United States Fire Administration (USFA) places a heavy reliance upon states as cooperative partners in administering the NFIRS program but much of the cost burden for NFIRS is carried by the states. No federal funding is provided to states for personnel, and USFA provides no guidelines for the staffing of state NFIRS programs.
Consequently, the levels and form of staffing and the resources available to NFIRS programs varies from state to state. Funding and resource limitations can undermine support for data collection, including participation in training and access to computer and software support.
The reliance on firefighters as primary data collectors is a recognized barrier in as much as firefighters are not trained researchers and have substantial responsibilities that can diminish attention to data collection and reporting. Liability concerns can also discourage complete reporting of information.
Little attention has been paid to the reliability of the data. It is clear that many firefighters who are doing their best to complete the reports disagree about how the reports should be coded [53].
With respect to the completeness and accuracy of reporting is that NFIRS codes are seen to be overly complex, resulting in frustration that can deter reporting. Because the list of code choices is so long for many data elements, many fire departments use cheat sheets (short set of notes) with the most commonly used code choices. Rarely used codes may be completely forgotten.
Many differences exist among who is interpreting the fire statistics, depending on the purpose. First of all, how many organizations within a country are interpreting it? To which purpose? It is notable that in smaller countries (in terms of number of buildings and population) the interpretation of the collected statistics is done by a single institute or organization. Examples of countries in which one or two organizations interpret the statistics are: the Netherlands, Switzerland, Denmark and Hungary.
In Austria, the purpose is to provide data for research and scientifically reasons as well as for performance-based fire prevention measures.
For the USA, both forms of interpretation are present, due to the differences between states. States with fewer resources may rely upon a single person.
Some state programs include full-time research analysts, while others rely upon administrators, information technology staff, or investigators to run their programs, often on a part-time basis. The UK (Wales, Scotland, England and Northern Ireland) has different organizations that interpret as well. Other countries that have several organisations interpreting the statistics are Sweden and Russia. Secondly, almost all countries differ in which organization/who is interpreting the data.
Organizations that have been mentioned as responsible for interpreting the statistics are: insurance associations, governmental organizations, universities, research institutes, individual municipalities (represented through the local fire and rescue service), departments of State fire service, scientific and educational organizations, ministries, General Inspectorate for Emergency Situations, Directorate for Civil Protection, the fire corps, first responders and technical experts.
A remarkable practice regarding who is interpreting the data was seen in Sweden and in Russia. Everyone can interpret it because fire statistics are presented in an open access database.
Having so many different organizations interpreting the data makes it difficult to exchange fire statistics.
Not only because different organizations have different interests and objectives but also because different sorts of organizations interpret based on different languages, definitions and terminologies.
The purpose of collecting fire statistics is the same despite all the differences between countries. When comparing all the diagnostic sheets for all countries, we find that there are eleven main purposes that are important for most countries (listed in no particular order):
When discussing issues with analysing data, the most important issue mentioned is dealing with missing data. This could be due to the lack of reported cases of fire or different databases exist that do not correspond or complement each other. Different databases from insurances, police, fire service, hospitals that are not possible to link with each other makes it very difficult to gain complete information about a fire and its corresponding damage or victims.
An illustrating example of this issue is when fire deaths at the hospital or in the ambulance are not accounted for.
If an organization that interprets the data has only small parts of databases available and therefore lacking the whole picture of a fire incident, it is an issue to learn from fire statistics. Unfortunately, many other issues arise when analysing data. The following are mostly mentioned:
Some countries use, at least to some extent, a method to follow up on the collected data. This method mostly consist on reviewing and correcting the data, ensuring the data is complete and if possible crosschecking one database with another. Giving fire services the chance to comment or reconsider data is also current practice. All countries have in common that updates are carried out regularly (often once a year), depending on the needs of the Fire Rescue Service.
This first task of the project is established to assemble all the knowledge of the consortium members regarding fire statistics in European countries and other non-European countries of interest. The review of the literature shows that fire data collection systems have been instrumental in reducing building fires and their associated deaths, injuries, and economic damage. The utility of information about these fires is apparent in the design of many fire safety interventions and policy initiatives.
Data on fire incidents can inform firefighting strategies, building codes, educational and training programs, and technical innovations, to cite just a few applications.
For example, with populations aging more than ever before, we might expect higher death rates among senior citizens, despite early fire detection.
It is logical to assume that safety efforts can benefit from strategies that have worked in other places. However, there is substantial agreement in the literature that differences between fire data collection systems in different countries complicate the ability to make comparisons that could be useful in evidence-based planning and prevention efforts.
While national fire data collection systems are likely to share certain core features and to gather some fire incident data in common, there appears to be considerable variation in the type and scope of information collected, the way that data elements are defined and levels of detail they seek, as well as the types of training and resources dedicated to collection efforts. In addition, literature suggests that fire data are influenced by differences between data collection procedures and practices.
Some data collection systems appear to provide opportunities to update information that may not be available at the time an incident record is first created, such as the cause of a fire or deaths that occur sometime after the incident. The amount and quality of information in different data collection systems also appear to be influenced by whether they include information from sources outside the fire service, such as insurers or medical authorities, through data linkage or other means.
Literature suggests that the issue of how much information to collect is an important area for consideration in the design of fire data collection systems.
Data collection systems that collect too little or wrong kind of information may not produce data that are useful, while overly detailed data collection systems may overwhelm data collectors, and thereby compromise data quality, as suggested by studies from the United States. In many respects, the issue of how much information to collect appears to be driven by available resources, as well as the capacities of data collectors, who mainly are fire service personnel, to collect and record information.
Concise data collection records will require less support and fewer resources than those that are more complex.
To that end, recent literature on fire data collection in Canada emphasizes that such factors as funding, resources, personnel, and stakeholder acceptance are critical considerations in the design and sustainability of national fire data collection systems.
In general, it appears that the fire data collection systems in most countries are presumed to provide an accurate representation of their respective experiences with fire incidents. However, information gathered through the initial phase of research suggest that they may be unaware of important limitations of their data due to missing information, differences in the way terms are defined or interpreted, and other identified issues.
We identified significant issues with fire data from Australia, Bulgaria, Canada, Denmark, France, and Germany which complicate confidence in the data, particularly for their use in inter-country comparisons.
Most of the issues stem from the lack of definitions for collected terms, lack of training, dispersed data, missing information and low coverage.
USA, Italy, and the Netherlands have very different systems while having each separate advantages and drawbacks. The fire data collection system in the USA has an existing terminology, includes a large number of data fields, and has vast experience in this field, but also appears to have a significant problem with missing fire incident data. However, because the EU is in a comparable situation to the USA, there are many lessons from the experience of the USA that can be directly applied to the EU.
Italy has adopted a quality control system to ensure the integrity of all data treated but is missing important fire data.
The approach of the Netherlands has been to reduce the problems posed by uncertainties by focussing data collection efforts on fatal residential fires.
We estimate that Austria, Russia, Sweden and the UK (in particular England, Wales and Scotland) provide data with high confidence level due to the existing definitions, important covered areas and collected terms and existing quality safeguards. For the other countries that were not mentioned, we do not have enough details to evaluate the level of confidence in their data. It is important to note that none of the consulted reports included uncertainty estimations.
However, it will be important to introduce uncertainty estimates to be able to analyse the relevance of the collected data and their trends.
Due to the lack of terminologies and precise collection methodologies and other issues identified, it is clear that current fire statistics cannot be compared from one country to another (with a few exceptions).
They can only be useful to describe the global fire safety situation and trends to some extent for a group of countries, or the specific fire safety situation.
To provide relevant information regarding the national fire safety situation (number of fires, fire fatalities, fire injuries, fire losses), fire statistics will have to be internationally improved through common terminology, common methodology, and common training and qualification of persons in charge of filling in the fire report, including uncertainty estimation methods.
The findings of this task will be used as preliminary groundwork for all the discussions that will occur during this project and as an output for all the following tasks.
[1] Geneva Association. World Fire Statistics, No 27. Information Bulletin of the World Fire Statistics Centre. October 2011. [2] Geneva Association. World Fire Statistics, No 28. Information Bulletin of the World Fire Statistics Centre. October 2012. [3] Geneva Association. World Fire Statistics, No 29. Information Bulletin of the World Fire Statistics Centre. April 2014. [4] National Safety Council. (2003). An analysis of fire deaths in Ireland 2001-2002. Conference Presentation. October 2003. [5] Hasofer, A.
M., & Thomas, I. Analysis of fatalities and injuries in building fire statistics.
Fire Safety Journal, 41(1), 2–14, 2006. https://doi.org/10.1016/j.firesaf.2005.07.006 [6] Blomqvist P., Johansson H. Brandstatistik - Vad vet vi om anlagd brand (Fire statstics – What do we know about arson), SP Technical Research Institute of Sweden, 2009, report 2008:48.
[7] Johansson, N. and van Hees P., Combining Statistics and Case Studies to Identify and Understand Deficiencies in Fire Protection, Fire Technology 48: 945–960, 2012. Doi: 10.1007/s10694-012-0255-z. [8] Assessment of Safety Levels in the Context of 2000-2012 Statistics. A statistical analysis of residential building fire causes at the national and city level. BiTP 35: 47-59. 2014. Doi:10.12845/bitp.35.3.2014.5. [9] Carlotti, P., Parisse, D., & Risler, N. (2017).
Statistical analysis of intervention reports for fires resulting in casualties deceased on the spot in Paris area. Fire Safety Journal, 92, 77–79.
https://doi.org/10.1016/j.firesaf.2017.05.017
[10] Xiong, L., Dorothy B., & Michelle B. Comparative investigation of ‘survival’ and fatality factors in accidental residential fires. Fire Safety Journal 73: 37-47, 2015. [11] Sekizawa, A., International comparison analysis on fire risk among the United States, The United Kingdom, and Japan. Fire Safety Science 4: 961-969, 1994. Doi:10.3801/IAFSS.FSS.4-961 [12] United States Fire Administration. Fire Death Rate Trends: An International Perspective. Topical Report Series, Vol. 12, 2011.
[13] Anderson A., Janssens M., A multi-national survey of low-energy and smoking materials ignition fires. Fire Technology 52:1709–1735, 2016. Doi: 10.1007/s10694-015-0550-6.
[14] Östman, L., Varför omkommer det fler personer i bostadsbränder i Finland än i Sverige?
["Why are more people killed in residential fires in Finland than in Sweden?"; In Swedish], Division of Fire Safety Engineering, Lund University, 2016. [15] Winberg, D., International Fire Death Rate Trends, SP Technical Research Institute of Sweden, 2016. [16] Johansson, N., McNamee, M. and van Hees P., Anlagd brand i skolor och förskolor," ["Arson in schools and pre-schools" in Swedish]. Division of Fire Safety Engineering, Lund University, 2020. [17] Department for Communities and Local Government.
Comparison of European Fire Statistics. Final report for the Department for Communities and Local Government. Fire Research Report 1/2012, 2011. ISBN: 978-1-4098-3135-8.
[18] Canadian Centre for Justice Statistics. Fire Statistics in Canada, Selected Observations from the National Fire Incident Database 2005 to 2014., September 2017. [19] Canadian Centre for Justice Statistics.
Sustainability of the National Fire Information Database – Next Steps. September 2017. [20] Butry, D.T., Thomas, D.S. Evaluating potential bias in non-randomly reported fire incident fata. National Institute for Standards and Technology. Technical Note 1770, 2012. [21] National Association of State Fire Marshals. Fire Research and Education Foundation. Conquering the Unknowns. 2014. [22] Kinsey, K. and Ahrens, M. NFIRS incident types: Why aren’t they telling a clearer story? Quincy, MA: NFPA, 2016.
[23] United States Fire Administration. Review and assessment of data quality in the National Fire Incident Reporting System. 2017. [24] Emsley, A., Lim, L., Stevens, G.
International fire statistics and the potential benefits of fire countermeasures. Polymer Research Centre and Peter Williams, Department of Mathematics and Statistics, University of Surrey, Guildford, Surrey GU2 7XH, UK. 2005.
[25] Kobes, M. and Groenewegen, K. Consumer fire safety: European statistics and potential fire safety measures, Netherlands Institute for Safety. 2009 [26] Ahrens, M., (2011). Smoke alarm presence and performance in U.S. home fires. Fire Technology 47: 699–720, 2011. https://doi.org/10.1007/s10694-010-0185-6 [27] Shokouhi, M., Nasiriani, K., Cheraghi, Z., Ardalan, A., Khankeh, H., Fallahzadeh, H., & Khorasani- Zavareh, D. (2019).
Preventive measures for fire-related injuries and their risk factors in residential buildings: a systematic review. Journal of Injury and Violence Research, 11:1-14, 2019. [28] McNamee M., Meacham B., van Hees P., et al. IAFSS Agenda 2030 for a fire safe world.
Fire Safety Journal 110, 2019. [29] Data collections - WHO. World Health Organisation. Retrieved December 22, 2020, from
https://www.who.int/data/collections
[30] The Global Health Observatory. Retrieved December 22, 2020, from https://www.who.int/data/gho [31] Burns. (2018, March 6). World Health Organisation. https://www.who.int/news-room/factsheets/detail/burns [32] World Health Organisation (WHO), A WHO plan for burn prevention and care, Geneva 2008, ISBN 978 92 4 159629 9. [33] European mortality database (MDB). (2020). World Health Organisation.
https://gateway.euro.who.int/en/datasets/european-mortality-database/
[34] SDR, Exposure to smoke, fire and flames, per 100 000. (2018). World Health Organisation.
https://gateway.euro.who.int/en/indicators/hfamdb_361-sdr-exposure-to-smoke-fire-and-flames-per-100-
000/ [35] Gapminder. Retrieved December 22, 2020, from https://www.gapminder.org [36] Winberg. D., International Fire Death Rate Trends. SP rapport 2016:32, 2016. [37] Brushlinsky N.N., Ahrens M., Sokolov S.V., Wagner P.. World Fire Statistics. Center of Fire Statistics, International Association of Fire and Rescue Services. Issue no. 23, 2018. [38] Les statistiques des services d'incendie et de secours - “General Direction of Civil Security and Crises Management” (DGSCGC) of the French Ministry of Interior.
Official web site:
https://www.interieur.gouv.fr/Publications/Statistiques/Securite-civile
[39] ISO TR 17755:2104(E) Fire safety — Statistical data collection — Part 1: Overview of national fire statistics practices, Geneva. [40] ISO/TS 17755-2:2020(E) Fire safety — Statistical data collection — Part 2: Definition of terms, Geneva. [41] The basics of event report. (2015). Myndigheten För Samhällsskydd Och Beredskap (MSB).
http://cursnet.srv.se/fortb/hr/start0/
[42] NFIRS Definition Tables. (2013). US Fire Administration.
https://www.usfa.fema.gov/downloads/xls/NFIRS_Spec_Tables_2013.xls
[43] Miller, D. (2005). Estimates of Fire Injuries Treated in Hospital Emergency Departments (July 2002- June 2003). U.S. Consumer Product Safety Commission. https://www.cpsc.gov/s3fs-public/pdfs/neissfire.pdf [44] International Statistical Classification of Diseases and Related Health Problems (ICD). (n.d.). World Health Organisation. Retrieved December 21, 2020, from https://www.who.int/classifications/classification-ofdiseases [45] World Health Organisation. (2016).
International Statistical Classification of Diseases and Related Health Problems 10th Revision ICD-10 Version:2016 Chapter XIX Injury, poisoning and certain other consequences of external causes (S00-T98). https://icd.who.int/browse10/2016/en#/XIX [46] World Health Organisation. (2016).
International Statistical Classification of Diseases and Related Health Problems 10th Revision ICD-10 Version:2016 Chapter XX External causes of morbidity and mortality (V01- Y98). https://icd.who.int/browse10/2016/en#/XX [47] U.S. Department of Health & Human Services. (n.d.). Center for Disease Control and Prevention. Retrieved December 21, 2020, from https://www.cdc.gov/ [48] U.S. Department of Health & Human Services. (n.d.-b). WISQARS (Web-based Injury Statistics Query and Reporting System).
Center for Disease Control and Prevention. Retrieved December 21, 2020, from
https://www.cdc.gov/injury/wisqars/index.html
[49] U.S. Department of Health & Human Services. (n.d.-c). WONDER. Center for Disease Control and Prevention. Retrieved December 21, 2020, from https://wonder.cdc.gov/ [50] U.S. Department of Health & Human Services. (n.d.-b). ICD - Classification of Death and Injury Resulting from Terrorism - Appendix I. Center for Disease Control and Prevention.
Retrieved December 21, 2020, from https://www.cdc.gov/nchs/icd/terrorism_code_appendix1.htm [51] Jonsson, A., Bergqvist, A. and Andersson, R., Assessing the number of fire fatalities in a defined population, Journal of Safety Research, 55, pp. 99-103, 2015. https://doi.org/10.1016/j.jsr.2015.10.001 [52] Jonsson, A., Runefors, M., Särdqvist, S. and Nilsson, F., Fire-Related Mortality in Sweden: Temporal Trends 1952 to 2013, Fire Technology, 52, pp. 1697–1707, 2016. https://doi.org/10.1007/s10694-015- 0551-5 [53] U.S.
Fire Administration.
(2011, October). NFIRS Representativeness Study.
https://www.reginfo.gov/public/do/DownloadDocument?objectID=53926600
The information gathered in Task 0 is the result of the collaboration of the consortium members who were able to provide the description of the fire statistics in several countries based on their previous research, experiences, and studies investigated in the literature review, public datasets and through a network of contacts. For some countries there are missing data, which will be complemented in the following tasks of the project, when it is necessary.
For each country, a detailed diagnostic sheet was filled, based on the gathered information. The standard structure of the diagnostic sheet is presented hereafter: #### 1. Terminology issues
References of existing database/studies
https://www.beredskabsstyrelsen.dk/viden/odin/vejledninger/definitioner/Pages/default.aspx
Summaries of existing database
At national level: Danish national fire statistics is published every year. It published by Danish Emergency Management Agency – (DEMA) www.brs.dk
Existing definitions
In the database there exists the following definitions:
Capacities (information regarding the municipal fire department) Fire stations: (Name, address, category, type (full time/part time/volunteer) Personal: (Name, full time/part time/volunteer, education) Vehicles (ID, type, station) Response activities Alarm:
Information about the incident address:
Information about used vehicles:
Information of personal:
Task:
Fire incidents:
Fire alarm system:
Are there differences within the same country?
No both the local (municipal) fire department and the state fire department use the same program/database
Are there differences and contradictions with other domains?
No, not known. Health authorities make there own statistics about people insured or dead in fires / smoke intoxication etc.
Identification of missing information
Fire´s the fire brigade don´t know about (no official fire alarm) is not stored in the database-system. The insurance companies have more data / databases but they are not public available. #### 2. Statistics collection issues
Fire department responsibilities
The local fire departments are responsible to deliver the data to the database.
Fire response organisation
The local fire department is owned by the municipal, and is always involved in a incident. If it is a large scale incident then the state can assist with materiel and men.
Who collects data?
The database is driven by the State (DEMA) but it is the local fire department, that delivers the data. The police department and Health authorities (ambulance) do not deliver any data. The insurance company don’t deliver any data
Who issues the data?
DEMA
Are there different levels of collection?
The local fire department delivers the data from their alarms.
The national firebrigade delivers the data from their alarms
Identify disparities in data feedback
The education in filling out the report (database) is rather weak, which increase the risk of different interpretations of a incident.
Where is the data stored?
DEMA (in the database ODIN)
Who is interpreting the statistics
DEMA
Purpose for which data is collected
The main purposes are: 1. To elaborate and develop the fire departments 2. To optimize fire response time 3. To define the volume of personnel and equipment 4. To detect any trends in incidents
What are the methods used to fill the gaps where information is missing?
Unknown
Is there follow up to data collected?
DEMA is controlling that all reports are made and contact the municipal if any is missing
Determining the level of confidence
Medium, mostly due to the lack of education. Pinpointing issues and limitations
The lack of training for the firefighters in charge of the fire response report.
The police is responsible for the fire investigation. There is no feedback-loop into the database from their investigation.
The database from the fire departments is not taking into account the fire casualties, which are not reported or reported in a separate database.
Examples
In 2019, the fire activity has accounted for 54% of the total activity and it represented the main activity.
In 2019, the majority of fire interventions was for single-family homes (43%), followed by fire interventions in multi-storey dwellings (33%), townhouse and others (8%), farmhouse (agricultural) (8%) and others (8%).
The source of the data is https://brs.dk/globalassets/brs---beredskabsstyrelsen/dokumenter/forskning-statistikog-analyse/2020/-redningsberedskabet-i-tal-2019-2.pdf
References of existing database/studies
DSB (The Norwegian Directorate for Civil Protection): https://www.brannstatistikk.no/brus-ui/ Some older studies of fatal fires (2005-2014) by a independent, state owned Swedish institute (RISE) Link: https://risefr.no/media/publikasjoner/upload/2017/a17-20176-1-analyse-av-dodsbranner-i-norge-iperioden-2005-2014.pdf
Summaries of existing database
It is an overview over all the reported fires by the fire department. (for public use) You can get a more detail information from the fire department in specific cases. (BRIS Rapport) There is also published a yearly report.
Existing definitions
Unknown.
The fire brigade officers, regardless if they are fulltime / halftime / professionals or volunteers – they fill in information in partly “drop down field” database.
More detailed information require most likely interviews with DSB-officers (task 1)
Are there differences within the same country?
No, Norway only have this national system for reporting.
Are there differences and contradictions with other domains?
Most likely / because no definitions / free text / only 30% manual input by fire brigades, other institutions (health authorities, police etc.).
Identification of missing information
Cause of fire is only a best guess from the fire department... The police are only reporting about 25-30% of their findings about the fire, so there is lack of correct information.
Insurance are investigating fires, but the data is not collected – There would probably more accurate information here. Generally it also depends on the firefighter / officer who put´s in the information (lack of definitions). #### 2. Statistics collection issues
Fire department responsibilities
Fires and other incidents – according to the department for justice and emergency. See also: https://lovdata.no/dokument/SF/forskrift/2002-06-26-729/KAPITTEL_3#KAPITTEL_3
Fire response organisation
Professionals and volunteers – according to the Norwegian laws : See also : https://lovdata.no/dokument/SF/forskrift/2002-06-26- 729?q=dimensjonering%20av%20brannvesenet
Who collects data?
Fire department collects all aspects of the fire. Police collects only 25-30% data from the fire that they are working with.
Who issues the data?
DSB (The Norwegian Directorate for Civil Protection)
Are there different levels of collection?
Only National / DSB
Identify disparities in data feedback
Lack of definitions which can lead to misunderstanding – 30% detailed information / 70 % automated-
Where is the data stored?
DSB (The Norwegian Directorate for Civil Protection)
Who is interpreting the statistics
DSB (The Norwegian Directorate for Civil Protection) / respectively the ministry of justice / emergency
Purpose for which data is collected
Fire preventive actions – and probably also other purposes (response time, etc.) What are the methods used to fill the gaps where information is missing? 30 % of the fires are reported more detailed where police and other departments are involved afterwards – and contribute to the database.
Is there follow up to data collected?
We don´t know that (yet)
Analyse potential cause and consequences in trends
For example: https://www.dsb.no/reportasjearkiv/brannstatistikk-2018/ - a link to a yearly report – see diagrams.
But statistics include all kind of fires (not only buildingfires) and are naturally also focused on fatalities and other general data (not very detailed). #### 4. Analyse existing data
Determining the level of confidence
Lack of definitions / automatic input versus manual (70/30)
Pinpointing issues and limitations
The statistics will not be accurate is long as the data inn is not correct.. it will be difficult to make preventing actions when they only make best guess in cause of fire. Police is only reporting inn 25-30% of the fires they are working with. Insurance, who investigate most fires are not reporting to the DSB (The Norwegian Directorate for Civil Protection). So a lot of data is missing in the official reports.
Prior to 2018 there were three systems (Alarmos, Core or Daedalos) used by the fire and rescue services when reporting incidents. However, since 2018-01-01 all fire and rescue services in Sweden report their incidents in Daedalos.
The contents of the incident report have been revised in 2005 and 2016. It took the slowest municipal brigade two years to introduce the latest version, called “Händelserapporten”.
We are currently working on the first revision of händelserapporten, to be implemented by 1st January 2022. The IDA system can be accessed by anyone at ida.msb.se. Some basic tables can be accessed without login, one can create one’s own tables by logging in. The quality and reliability of the statistics is assessed yearly by MSB1.
For e.g. researchers it is also possible to get more detailed data/analysis by making a request to the statisticians at MSB as not all fields in the incident report are available to the public due to e.g personal integrity. The database contains all type of incidents/accident that the rescue service responds to, not only fires.
Statistics Sweden https://www.scb.se/en/ keeps statistics on much data in Sweden (economy, number of people in households etc.). Data can be retrieved from there by anyone at
https://www.statistikdatabasen.scb.se/pxweb/en/ssd/.
Statistics on fatalities is kept by the National Board of Health and Welfare (Socialstyrelsen)
https://www.socialstyrelsen.se/statistik-och-data/register/alla-register/dodsorsaksregistret/. Socialstyrelsen
provides information on all the data in the database (definitions), death cause is organised according to ICD- 10. Anybody can get statistics from the database on https://sdb.socialstyrelsen.se/if_dor/val.aspx. Socialstyrelsen also have a comprehensive database covering all in-patient treatments in Swedish hospitals. This is an important source for researchers studying fire injuries. MSB maintain a database on fatal fires and fire victims, and publish statistics on IDA.
Data is combined from the relevant authorities so that fire fatalities can be associated with a specific fire incident.
In addition, there are cases when rescue services do not respond to a fatal fire, this can e.g. be when someone is living in remote areas and the fatal fire is discovered a couple of weeks after the fire actually occurred.
The statistics for the period 1999 – 2015 are considered to be of very high quality due to the use of data from the National Forensic Centre. Unfortunately MSB has not been able to access forensic data from 2016 and until the Swedish government clarify the legal situation, the statistics onwards are considered preliminary
https://ida.msb.se/ida2#page=e3d46ba0-8f87-4ab7-b28d-7b950cd8a43. A description of the fatal fires
database is provided by Johnson et al.2 Insurance companies also collects statistics; however, their statistics is not very detailed, and they can in many cases not differ between a fire and a thunder incident. The statistics is available at
Brandskyddsföreningen collects data on fires from media. This data is not available. 1 MSB, Kvalitet i MSB:s insatsstatistik 2019,
https://ida.msb.se/dokument/insatsstatistik/kvalitet2019/Kvalitetsdeklaration2019.pdf
2 Jonsson A., Bergqvist A., Andersson R. ”Assessing the number of fire fatalities in a defined population”, Journal of Safety Research, Vol 55 December 2015, pp 99-103 There has been and collection of more detailed residential fire statistics in the project “Lärande från bostadbrand” initiated by MSB. Only a few fire and rescue services have participated in this project.
Summaries of existing database
The database is online available at MSBs website.
Existing definitions
Complete definitions of the different fields in the incident reporting system is available, however it is all in Swedish. Some definitions/terms for residential fires that might be useful are translated below
Year (År)
Name of Rescue service (Räddningstjänst), in many cases similar to municipality
County (Län)
Municipality (Kommun)
Type of Municipality (kommungrupp) Suburban to big city, Suburban to larger cities, Rural Municipality, Sparsely populated region, Densely populated region
Commuter municipality, Big city, Larger city, Tourism city, Production city
Building (Byggnad) type of building (gas station, student house, vacation house, hotel, jail, defence building, school, industry, farm, dwelling, outside, hospital, etc.)
Building type (Byggnadsgrupp) (public building, other, dwelling, outside, industry or unknown)
Month (Månad)
Day of the month (Dag)
Day of the week (Veckodag)
Date (datum)
Cause of fire (Brandorsak)
Size of fire upon arrival (Omfattning vid ankomst)
Total size of fire (Brandens totala omfattning)
Smoke detector present (brandvarnarförekomst)
Start room (startutrymme)
Start item (startföremål) MSB plan to publish a homepage with an English translation of the entire contents of the revised händelserapport by the end of 2020. Terminology is also discussed in appendix 2 of Andersson et. al.3
Are there differences within the same country?
Since it is different fire and rescue services and individuals that report to MSB there will be some differences. However, MSB has developed documentation and an online education (http://cursnet.srv.se/fortb/hr/start0/) to support individual fire and rescue services and reporters in the reporting. Prior to 2018 there were three systems (Alarmos, Core or Daedalos) used by the fire and rescue services when reporting incidents.
However, since 2018-01-01 all fire and rescue services in Sweden report their incidents into the same system.
Are there differences and contradictions with other domains?
Classifications in fire statistics from insurance companies do not correspond to those used by the fire and rescue service. Identification of missing information 3 Andersson P, Johansson N, Strömgren M. ”Characteristics of fatal residential fires in Sweden” SP-report 2015:53 The quality and reliability of the statistics is assessed yearly by MSB4. There have been issues with missing data when different systems (Alarmos, Core or Daedalos) where used at different fire and rescue services.
There has previously been some double counting of incidents when serval fire and rescue services from different municipalities/regions are involved. This has been corrected in the yearly quality control done by MSB. The routines at joint incidents have been improved in 2020 and MSB expect this to more or less eliminate problems with double counting. Due to incident report content revisions and changes in local routines for recording data, there is a potential to over-interpret discontinuities in some time series.
MSB has conducted some minor analysis of this. However, it is something that the individual user needs to be aware of and asses if it can be a problem in each individual case.
Fire department responsibilities
After each incident, the fire and rescue service documents what has happened and which measures that have been taken. Each rescue service has appointed a statistics coordinator, who coordinates their organization's reporting to national statistics. After an accident it is, according to law5, the responsibility of the municipality to a reasonable extent clarify the causes of the accident, the course of the accident and how the operation has been carried out.
MSB continuously monitors the inflow of reports from the respective rescue services and, if necessary, contact the relevant statistics coordinator to inquire about the cause to late reporting or inform about detected errors in the organization reports.
Fire response organisation
The municipality is responsible for the local rescue service according to law6.
The Swedish state is responsible for rescues service in the areas of aviation, sea and mountains.
Who collects data?
The rescue service fills in the incident reports and MSB collects it. The police conduct fire investigations, often in close cooperation with the local fire and rescue service. Police investigations are not freely available, at least until criminal charges are made. The National Forensics Centre have detailed information from post mortems for nearly all fire victims. The Socialstyrelsen collects data on deaths.
Who issues the data?
MSB 4 MSB, Kvalitet i MSB:s insatsstatistik 2019,
https://ida.msb.se/dokument/insatsstatistik/kvalitet2019/Kvalitetsdeklaration2019.pdf
5 Lag (2003:778) om skydd mot olyckor 6 Lag (2003:778) om skydd mot olyckor
Are there different levels of collection?
All data in the IDA database is national, but it is possible to break down the data on municipality level. It is also possible to break data down into the small areas defined by Statistics Sweden for all detailed local and regional statistics.
Identify disparities in data feedback
Data is used by governmental bodies and research.
Where is the data stored?
In databases hosted by each relevant local or national authority. #### 3. Statistics interpretation issues
Who is interpreting the statistics
Since the data is open everyone can interpret the data. There are a lot of different publications that has been made using the statistics over the years. These have been issued by MSB and by universities like Lund and Karlstad Universities, as well as research institutes such as RISE. Individual municipalities (often represented through the local fire and rescue service) frequently analyse statistical data as a basis for the local planning.
Purpose for which data is collected
MSB:s statistics are intended to shed light on which accidents the municipal rescue service respond to and what measures are taken in the event of these accidents.
The overall purpose is to streamline society's work to protect against accidents through increased knowledge about accidents and measures taken.
Society is constantly changing and the need to be able to follow the development of accidents over time is considered particularly important.
What are the methods used to fill the gaps where information is missing?
MSB reviews the quality of the statistics yearly and reports if data is missing.
Is there follow up to data collected?
The fire and rescue services can update and resubmit the incident report (händelserapport) if they become aware of new information which makes a revision necessary. The statistics published by MSB include all revisions up to the publishing date.
For example, if a fire and rescue service update a report from 2018, then the revision will be included in that year’s statistics on the next annual release of new statistics.
Analyse potential cause and consequences in trends
It is easy to identify trends with statistical tools.
However, there would appear to be quite a lot of random variation and many “trends” that seem to exist in time series are not significant when tested rigorously. It is often extremely difficult to identify what has caused any trend. #### 4. Analyse existing data
Determining the level of confidence
High in the data collected by authorities. The insurance company data is limited.
Pinpointing issues and limitations
Regarding data in the IDA database:
Slightly more than 1% of the residential fires result in at least one fatality
A fatal residential fire is often large when the rescue service arrives at the scene
Fatal residential fires occur more commonly in the late night/early morning
Usually only one person is present in the fire compartment when the rescue service arrives to a fatal residential fire
There is only marginal differences regarding the presence of smoke detectors in residential fires and fatal residential fires.
Both fatal residential fires and residential fires are less common in and around larger cities
Fires starting in beds, sofas, armrest chairs or clothing results more often in fatalities than other start items while fires starting on the stove, in the fireplace or the chimney seldom results in fatalities compared to the number of these fires.
Fires starting in the living room or bedroom result more often in fatalities than fires starting in other rooms.
Smoking is a common cause in fatal residential fires while lightning, forgotten stove, chimney fire, technical malfunction and re-ignition seldom results in fatalities compared to the number of this type of fires.
The likelihood of dying in a residential fire might be somewhat higher for a person living in an apartment than for a person living in a single-family house. However, the differences are small and given the uncertainties in the number of people living in different types of dwellings it might be that there is no difference.
The number of floors in a building does not seem to differ between fatal and nonfatal fires.
The fire start floor in an apartment building seems to be slightly higher in fatal fires than in non-fatal fires. For more information about the study and the conclusions drawn please view the report2. This report is also the basis, together with other Swedish studies on fatal fires, for recommendations on measures to take in order to decrease the number of fatalities in fires8. Measures recommended includes:
Continued work to increase the use of smoke detector in homes, both in terms of number of smoke detectors and better smoke detectors
Information by e.g., increased awareness of informative webpages like “Din säkerhet (your safety)” 7 Andersson, P., Johansson, N. & Strömgren, M. (2015) Characteristics of fatal residential fires in Sweden, SP report 2015:53, Borås, Sweden. 8 Andersson, P., Arias, S., Arvidson, M., Frantzich, H., Larsson, I., Vermina Lundström, F., Nilsson, D., Runefors, M., (2018) Riskreducerande åtgärder för dödsbränder i bostäder, RISE report 2018 :37, report in swedish
Increased fire safety for people under home care.
Fire performance requirements for furniture
Continued and improved follow up on fatal fires
More emphasis on fire safety in updates of the building regulations even if it cannot be proven that the measures are cost effective
Continued research on fire safety for people e.g., on better smoke detectors, tools for evaluating the efficiency of different measures, and on fires with unknown cause. School fires In a recent report9 based on publicly available data from MSB (the IDA database, see section 0) the following conclusions were drawn about school fires:
The number of fires in schools building have varied between 300 to 500 during the last 20 years.
Arson is the most frequent cause (around 50% in average) of school fires.
There is a steep increase of school and preschool fires in Sweden during after 2015. It is also clear that this increase is due more fire caused by arson.
There is a clear difference of the room of origin between school fires and preschool fires. This indicates that there are some differences in the characteristics of these fires. It is also clear that fires in “Bathroom/toilet” have increase substantially during the last three years in schools.
The number of fires during school hours have increased during the last couple of years. There is also an indication that the number of school fires during the night and early morning is decreasing.
The analysis presented shows that the severe and costly fires have not increased, rather it is the number of small and more moderate fires that have increased. These fires occur during school hours in secluded spaces like toilets and corridors. For more information about the study and the conclusions drawn please view the report3. Examples of fire trends in Sweden based on national fire statistics The following graphs have been developed to illustrate the type of data available in the national fire statistics in Sweden. All the data have been retrieved from the IDA which is maintained by MSB.
Fatal fires The database on fatal fires includes a total of 1,871 fire incidents (including 2,046 fatalities) from 1999 to 2015.
The following data is a selection of the data from the fatal fires database. Note that data from 2016-2019 is also available in a preliminary dataset (353 fire incidents including 363 fatalities). 9 Johansson, N., McNamee, M. & van Hees, P.
(2020) ANLAGD BRAND i skolor och förskolor, report 3230, Lund University, Sweden. (report in Swedish)
Figure 30 : Smoke alarms presence in fatal fires.
School fires The database on building fires includes a total of > 230,000 fire incidents in buildings from 1998 to 2019. The following data is a selection of the data from the building fires database.
Figure 36 : Damage estimate for “outside building” school fires.
Information from ISO 17755-1 & -2
The United Kingdom’s Incident Reporting System (IRS) is based on separate reports on each incident requiring a response by a fire brigade. There is a national standard for coding of incidents. All fire brigades are participants, and all are required to report on all incidents; therefore, the design is a census and there is no adjustment for missing data. Analysis is by counting only.
Most reports are completed by firefighters who lack extensive training in fire investigation, but an estimated 10% of reports are completed by personnel with extensive training in fire investigation.
The U.K. also conducts periodic household surveys, which provide regular estimates of the percentage of all home fires reported to fire brigades. The U.K. estimates that brigades are called to 1/5 of home fires.
Fires subject to reporting
(ISO 17755:2014, page 5) A fire is an incident, attended by a local authority, of uncontrolled burning involving flames and/or heat and/or smoke. An unknown number of departments employ truncated/reporting thresholds. These thresholds are determined on a department by department basis. Fire does not include the following except when they cause fire or occur as a consequence of fire:
Fires are categorized for analysis and reporting purposes according to major incident type in the following way:
Fire deaths subject to reporting
(ISO 17755:2014, page 7) For United-Kingdom, the definition of a fire fatality is given in the document named “Incident Recording System (IRS) Help and Guidance – version 2.3” question 3.5: “Killed/ Fatality - a person who has died as a direct or indirect result of injuries received at the incident” and specified at question 9.21”: “in general, fire-related deaths are those that would not have otherwise occurred had there not been a fire”.
Fire injuries subject to reporting
(ISO 17755:2014, page 10) For United-Kingdom, the definition of a fire injury is given in the document named “Incident Recording System (IRS) Help and Guidance – version 2.3” question 3.5: “Injured/Non fatal - a person injured as a direct result of the incident (but not fatally injured) who required first aid (provided by anyone) at the scene or more medical treatment than could be given at the accident.
This includes any person advised to attend hospital or see a doctor, whether or not they actually follow up the advice”.
All should be recorded. Completeness is believed to be good including all physical injuries, not just burns and smoke.
Victim characteristics
(ISO 17755:2014, page 16) Age Gender Ethnicity Where was the victim when the fire started?
Room, cabin or compartment of origin
Different room, cabin or compartment on floor of origin
Floor above origin (includes mezzanine above floor of origin)
Two or more floors above origin
One floor below origin – includes stairway leading down from floor
Two or more floors below origin
Outside building, vehicle etc. of origin
Seat of fire unknown or multi-seated (and above non applicable)
Location of person unknown
Not applicable
Other location Where was the victim found?
Room, cabin or compartment of origin
Different room, cabin or compartment on floor of origin
Floor above origin
Two or more floors above origin
One floor below origin
Two or more floors below origin
Outside building, vehicle etc. of origin
Seat of fire unknown or multi-seated (and above non applicable)
Location of person unknown
Not applicable
Other location What role did the victim play in the incident?
Firefighter on duty
Other emergency service personnel
Resident/occupant
Passer by
Driver
Passenger
Visitor
Employee in workplace
Customer in shop
Other FRS personnel on duty
Other Was victim rescued?
Yes
No If rescued, where was the victim rescued from?
Room, cabin or compartment of origin
Different room, cabin or compartment on floor of origin
Floor above origin (includes mezzanine above floor of origin)
Two or more floors above origin
One floor below origin – includes stairway leading down from floor
Two or more floors below origin
Roof
Outside building of origin
Not applicable
Other Circumstances of fatal casualty
Thought to be already dead when firefighter arrived
Unable to resuscitate, confirmed dead at scene
Unable to resuscitate, confirmed dead at hospital
Alive on leaving scene, but died later
Not known Has the casualty been reconciled against the appropriate death certificate? Was the death/injury fire related?
Yes (in general, fire related deaths are those that would not have otherwise occurred had there not been a fire)
No
Don’t know What is your understanding of the cause of the death?
Overcome by gas, smoke or toxic fumes; asphyxiation
Burns – severe
Combination of burns and overcome by gas/smoke
Shock/anaphylactic shock
Other medical condition
Fracture
Other physical injury
Cuts/lacerations
Impalement
Drowning
Hypothermia
Heat exhaustion
Back/neck injury (spinal)
Head injury
Chest/abdominal injury
Chest pain/Heart condition/Cardiac arrest
Other
Unknown What were the circumstances of the victim?
Bedridden
Chair-ridden
Other immobility
Suspected under influence of alcohol
Suspected under influence of drugs
Discovering fire
Fell onto fire
Fighting fire (including attempts)
Trapped by fire because unaware (e.g. asleep)
Trapped by fire other than unaware
Trapped by collapse of structure
Trapped by smoke
Injured escaping
Injured rescuing person
Injured rescuing property or animals
Injured being rescued
Injured by blast
Return to fire
Intentionally sustained at start of fire (e.g. suicides and attempts)
Injuryaccidentallysustainedatstartoffir e
Notapplicabl e
Other
(ISO 17755:2014, page 27) From the document named “Incident Recording System (IRS) Help and Guidance – version 2.3” part “On attendance –Damage, the damages are the total loss to the structure and contents, including contents damaged by fire, heat, smoke, water.
Monetary damages are not included in national analyses, which defer to insurance companies. Measures used at the national level are:
Indoor area damaged (for example, square meters in a building):
Should be reported as required in the document named “Incident Recording System (IRS) Help and Guidance – version 2.3” question 8.20, “the estimated flame and/or damage on arrival”.
Should be reported as required in the document named “Incident Recording System (IRS) Help and Guidance – version 2.3” question 8.24, “the horizontal area damaged by flame and/or heat at stop”.
Should be reported as required in the document named “Incident Recording System (IRS) Help and Guidance – version 2.3” question 8.25, “the horizontal area damaged by flame and/or heat and/or smoke and/or water at stop”.
Outdoor area damaged (for example, acres in a wildfire):
Should be reported as required in the document named “Incident Recording System (IRS) Help and Guidance – version 2.3” questions 8.35 and 5.16a, “the estimated outdoor fire damage by flame and/or heat and/or smoke”.
Number of rooms damaged:
Should be reported as required in the document named “Incident Recording System (IRS) Help and Guidance – version 2.3” question 8.20, “the estimated extent of flame and/ or heat damage on arrival”.
Should be reported as required in the document named “Incident Recording System (IRS) Help and Guidance – version 2.3” question 8.22, “the extent of flame and/or heat damage at stop”.
Number of floors damaged:
Should be reported as required in the document named “Incident Recording System (IRS) Help and Guidance – version 2.3” question 8.20, “the estimated extent of flame and/ or heat damage on arrival”.
Should be reported as required in the document named “Incident Recording System (IRS) Help and Guidance – version 2.3” question 8.22, “the extent of flame and/or heat damage at stop”. Spread from one primary property to another primary property (or a secondary property) will only show the damage for the first property damaged. Only the distance, in meters, between the incident location and the neighbouring property(s) will be shown.
Spread from ‘secondary’ (e.g. some outdoor structure types) property to a primary property (including all vehicles & buildings that are not derelict) will only show the damage to the primary property.
Other lossessubject to reporting
(ISO 17755:2014, page 29)
Locations of fires based on surveyresponses
(ISO 17755:2014, page 32) The U.K. has 295 categories, including 21 dwelling categories (private residential), 18 other residential, and 160 non-residential buildings. The complete list can be downloaded from
https://www.gov.uk/government/publications/incident-recording-system-for-fire-and- rescue-authorities.
Specific types ofbuildingsand other structures
(ISO 17755:2014, page 33) The UK is not present in this section.
Specifictypesofroomsandoth (ISO 17755:2014, page 46) For Dwellings
Reporting of building height and other characteristics
(ISO 17755:2014, page 54)
Reportingandestimationofdeliberatelys
et fires
(ISO 17755:2014, page 57) Yes (under Question 5.15 – Cause/Motive) reported as “Deliberate – own property”, “Deliberate – other’s property” and “Deliberate – unknown owner”.
Also (under Question 8.1 – Cause of the fire), each of three types of deliberate fire is subdivided as to
Other information relevant to motive (under Question 8.3 – Caused by) can be used to isolate juvenile firesetters:
ag e 65 ormor e)
(ISO 17755:2014, page 61) Yes, reported as “Accidental/Natural occurrence” under Question 8.1, Cause of the fire, and as “Natural occurrence” underQuestio n 8.4, Mainsourceofignitio n.
Nodetailedbreakdownsreporte d.
Reportingandestimationofexposu
re fires
(ISO 17755:2014, page 62) Yes, reported as “Spread from secondary fire” under Question 8.4, Main source of ignition. No detailed breakdowns reported.
Reportingandestimationofsmokingmaterialandopenfla
me fires
(ISO 17755:2014, page 65) Yes, can be reported under Main Source of Ignition:
sourceforsomecountriesbutnotother s)
(ISO 17755:2014, page 68) Yes, can be reported under Main Source of Ignition:
Reportingandestimationofcookingandkitchenequipme
nt fires
(ISO 17755:2014, page 72) Yes, can be reported under Main Source of Ignition:
Reportingandestimationofclothesdry
er fires
(ISO 17755:2014, page 75) Yes, can be reported under Main Source of Ignition:
Reportingandestimationofentertainmentequipme
nt fires
(ISO 17755:2014, page 77) Yes, can be reported under Main Source of Ignition:
Reportingandestimationofofficeequipmen (ISO 17755:2014, page 78) Yes, can be reported under Main Source of Ignition:
Reporting of electricalandelectricaldistributionorlightingequipment
(ISO 17755:2014, page 82) Yes, can be reported under Main Source of Ignition:
Reporting of otherapplianceandequipment
(ISO 17755:2014, page 86) Yes, can be reported under Main Source of Ignition:
Reporting of item first ignited in terms of form and function
(ISO 17755:2014, page 97) Yes, can be reported under Item First Ignited:
hed
These are the factors under the Accidental section of “What was the cause?” excluding those already cited for intentional and fireplay fires:
alcoholasacontributoryfactor Reporting of factorsinignition
(ISO 17755:2014, page 106) Yes.
These are the factors under the Accidental section of “What was the cause?” excluding those already cited for intentional and fireplay fires:
dthereisaYe s/ Nodataelementonsuspecteddrug s/ alcoholasacontributoryfactor
Reporting of factors in fire growth e growth
(ISO 17755:2014, page 108) Yes.
Item Mainly Responsible for Spread of Fire uses same choices as Item First Ignited (see Table 20D, Reporting of item first ignited in terms of material composition). Also some dat aelementsonexplosion s, dangeroussubstancesinvolve d.
Presenceandtypeofsprinklerorotherextinguishingequipment
(ISO 17755:2014, page 114) Yes, included in reporting under “Active Firefighting Systems present in vicinity of fire (origin of fire).” Question 7.11 is whether any active safety system was present, yes or no.
Type of system; multiple types of systems can be checked:
tofir e:
Inroomoforigi n
Onsamefloo r (ISO 17755:2014, page 116) Whether the system operated, answered for each system:
No
Yes but did not raise alarm
Yes and raised alarm Number of sprinkler heads operating (asked only for sprinklers). Answers are 0,1,2,3,4,5, more than 5, and unknown. System’s impact upon fire:
Extinguished
Contained or controlled
Did not contain or control
Unknown Main reason why system did not function as intended
System not set up correctly or not installed correctly
System damaged by fire
Fault in system (such as defective system, lack of maintenance, heads painted over)
Sy
stemturnedoff
Fireinareanotcoveredbysyste m
Other
(ISO 17755:2014, page 119) Yes, included in reporting.
How the fire was discovered.
tionrelativetofir e:
Inroomoforigi n
Onsamefloo r (ISO 17755:2014, page 123) Did the system operate?
Answered for each system:
o –
underconstructio n [ andnothabitabl e] (ISO 17755:2014, page 126) Yes, included in reporting under main action taken by “the general public” prior to arrival [of fire brigade] – clarified to mean the main method of firefighting.
Fire extinguishers and use of hose reels are two of the coding options; these are the others:
Removal from heat source
Fuel supply disconnected
Smothering
Water from bucket/container
Water from garden hose
Beaten out
Work team [clarified to mean a trained private team and not just a fire warden]
Other Not known Possibly relevant are the choices on fixed firefighting facilities present:
Dry risers
Wet risers
Firefighting lift
Firefighting shaft
Foam makers/drenchers/downcomers
Smoke extraction/ventilation
Other For each type checked above, a separate data element asks whether it was used, Yes or No. Another data element asks for reasons i fitwasnotworkin g:
Poormaintenanc e
Vandalis m (ISO 17755:2014, page 127) Yes, “smoke extraction or ventilation” is one of seven choices included in reporting under fixed firefighting facilities. One data element can be used to indicate presence of the equipment, a second data element can be used to report whether it was used, and a third data element can be used to record reasons if it was not workin g:
Poormaintenanc e
Vandalis m
DamagedbyfireReportingonfiredoor s,
firewallsand other compartmentation
(ISO 17755:2014, page 129) Yes, included in reporting.
Compartmentation performance
References of existing database/studies
DATABASES: England, Home Office Website: https://www.gov.uk/government/collections/fire-statistics Dwellings fire statistics dataset [1] Other building fire statistics datasets [2] Northern Ireland, Northern Ireland Fire and Rescue Service: Website: https://www.nifrs.org/?s=fire+statistics Scotland, Scottish Fire and Rescue Service [3] Website: https://www.firescotland.gov.uk/about-us/fire-and-rescue-statistics.aspx Wales, Welsh Government [4] Website: https://gov.wales/fire-and-rescue-incident-statistics STUDIES:
Manes, M., & Rush, D. (2018). A Critical Evaluation of BS PD 7974-7 Structural Fire Response Data Based on USA Fire Statistics. Fire Technology, 55(4), 1243–1293. https://doi.org/10.1007/s10694- 018-0775-2
Manes, M., & Rush, D. (2020). Assessing fire frequency and structural fire behaviour of England statistics according to BS PD 7974-7. Fire Safety Journal.
https://doi.org/10.1016/j.firesaf.2020.103030
Ramachandran, G., & Charters, D. (2011). Quantitative risk assessment in fire safety. Routledge. https://doi.org/10.4324/9780203937693
Rutstein, R. (1979a). The estimation of the fire hazard in different occupancies types. Fire Surveyor, 8(2), 21–25.
Rutstein, R. (1979b). The probability of fire in different sectors of industry. Fire Surveyor, 8(1), 20–23.
Rutstein, R., & Cooke, R. A. (1979). The value of fire protection in buildings. Home Office, Scientific Reserach and Development Branch.
Summaries of existing database
Data is subdivided into England, Northern Ireland, Scotland and Wales. England fire statistics of the Home Office presents fire safety data related to fire incidents, causes, consequences, fatalities/casualties, response time, alarms and automatic extinguishing systems. Fire statistics in Northern Ireland is not publicly available.
The fire statistics in Scotland of the Scottish Fire and Rescue Service contains information related to Incident type, property type, fatalities and casualties, fire stations and workforce.
In Wales, the fire statistics of the Welsh Government publicly available covers fire incident, location, cause, motive, casualties, response time, smoke alarms.
Existing definitions
The complete list of definitions is available in the following documents: England [1], [2]; Scotland [3] and Wales [4]
Are there differences within the same country?
Yes, a unique national database is not available, and the fire safety data are subdivided into England, Northern Ireland, Scotland and Wales. Northern Ireland Fire and Rescue Service statistics does not use the IRS system and does not publish comparable statistics.
Are there differences and contradictions with other domains?
Yes, there are differences in the fields recorded in the various statistics within the same nation.
Identification of missing information
England statistics appears to provide the highest number of fields publicly available after a fire incident while Scottish fire statistics does not have data on the quantification of damage and presence of alarms or automatic extinguishing systems.
Finally, Welsh fire statistics has fire safety data also on fire causes and motive and only on smoke alarms.
Fire department responsibilities
England: When the Fire and Rescue Service attends an incident, the details are uploaded to the Incident Recording System. Scotland: When the Scottish Fire and Rescue Service attends an incident, the details are uploaded to the Incident Recording System run by the Home Office. Wales: When the Fire and Rescue Authorities in Wales attend an incident, the details are submitted to the Home Office.
Fire response organisation
The Fire Service in the United Kingdom operates and follows different administrative legislations in England, Northern Ireland, Scotland and Wales.
Who collects data?
England: fire safety data are collected by the Fire and Rescue Service in the aftermath of an incident.
Scotland: fire safety data are collected by the Scottish Fire and Rescue Service in the aftermath of an incident. Wales: fire safety data are collected by the Fire and Rescue Authorities in Wales in the aftermath of an incident.
Who issues the data?
England: the vast majority of statistics for England are produced by the Home Office. Scotland: “Prior to October 2015, these statistics were produced by the Scottish Government and accredited by the United Kingdom Statistics Authority as National Statistics (signifying full compliance with the Code of Practice for Statistics). Subsequently they have been produced by SFRS in voluntary compliance with the Code of Practice for Statistics.
However, as SFRS are not currently named in legislation as producers of Official Statistics, it was no longer possible for these statistics to be designated as National Statistics.
SFRS are continuing to work towards becoming named as producers of Official Statistics with a view to regaining accreditation as National Statistics for this series. SFRS has agreed to being named as producers of Official Statistics in upcoming legislation” [3]. Wales: the Welsh Government compiles the statistics in the bulletin based on reports submitted by FRAs to the Home Office.
Are there different levels of collection?
Currently no national data is collected, and they are published according to their origin (England, Scotland and Wales)
Where is the data stored?
England: databases are stored in the archive of the Home Office.
Scotland: databases are stored in the archive of the Scottish Fire and Rescue Service and accessible through the Scottish Governments ScotStat service. Wales: databases are stored in the archive of the Welsh Government. #### 2. Statistics collection issues
Who is interpreting the statistics
England: “Data received by the Home Office undergo a quality assurance process to ensure the data is fitfor-purpose and published to the highest possible standard. Any data quality issues are flagged and subsequently resolved with FRSs” [5]. Scotland: “The SFRS Data Services team run exception checking scripts on the IRS database to identify potential errors which are subsequently manually checked” [3]. Wales: Welsh Government applies further validation and verification to the data extract from the IRS.
Purpose for which data is collected
In England, the uses of fire statistics are:
Informing the general public;
Policy making and monitoring;
Fire and rescue service;
Third parties;
Informing public marketing campaigns;
The Office for National Statistics;
Inspections and auditing;
and National and international comparisons. Users are:
Ministers;
Members of Parliament;
Fire and rescue authorities and services;
Other colleagues within the Home Office;
Other government departments HMICFRS;
Trade Unions;
Journalists;
Chartered Institute of Public Finance and Accountancy;
Local Government Association;
Individual citizens and private companies;
Students, academics and universities;
and Charities [5]. In Scotland: all the users who may be interested, use a range of fire statistics or are involved in the production of fire and rescue related statistics. In Wales: the uses of fire statistics are:
to advice to Ministers;
to measure government targets and key performance indicators;
to provide context and evidence for the Welsh Government’s policies;
fire service comparisons and benchmarking;
to compare fires and the fire service in Wales with other countries;
to inform the debate in the National Assembly for Wales and beyond;
to assist in fire research and analysis;
and to provide information on FRSs’ performance and activities to citizens and communities in Wales. Users are:
Ministers;
Assembly Members and the Members Research Service in the National Assembly for Wales;
The Office for National Statistics;
Department for Communities and Local Government;
Chartered Institute of Public Finance and Accountancy;
Fire and Rescue Authorities and Services;
Welsh Local Government Association;
Students, academics and universities;
Other colleagues within the Welsh Government;
Other government departments;
Individual citizens and private companies;
and Charities [6]
What are the methods used to fill the gaps where information is missing?
In England: “Information about the incident attended is input by a member of the attending fire crew and then quality assured by their line manager. The IRS consists of up to 175 questions, not all of which are asked for every incident. The IRS has on-line data entry with in-built validation rules which ensures that basic validation errors are avoided. However, there is more detail on the QA process in the previous section.
Following the transition to the online IRS in 2009, the main types of errors in the data is thought to relate to recording and classification errors.
The level of missing data on fields is very low, with such missing data reported as unknown and therefore no grossing, imputation or other estimation methods are used” [5].
In Scotland: “The officer in charge of an incident enters data to the IRS using software forms with inbuilt validation rules. The SFRS Data Services team then run exception checking scripts on the IRS database to identify potential errors which are subsequently manually checked. Further quality assurance is completed by cross checking incidents details with control room logs” [3]. In Wales: “due to time and resource constraints, we sometimes either record data as missing or we seek to impute the data.
If we impute or change any data, then we inform the data provider of this and give them a chance to comment or challenge this.
This is a compromise to resolve validation issues and does not impact significantly on the usability of the dataset” [6].
Is there follow up to data collected?
In England: “In order to ensure the IRS data is complete, Home Office statisticians carry out monthly monitoring of the number of incidents submitted as discussed in the quality assurance section” [5]. In Scotland: “After the most severe fire incidents the SFRS fire investigation team complete a report on the details of the incident. Following these reports the IRS record is amended to reflect the findings where necessary. SFRS Data Services use these reports to crosscheck the IRS records” [3].
In Wales: “If we impute or change any data then we inform the data provider of this and give them a chance to comment or challenge this” [6].
Analyse potential cause and consequences in trends
In England, the Home Office investigate how causes and consequences vary over the years.
In Scotland, the Scottish Fire and Rescue Service publishes reports investigating how causes and consequences vary over the years. In Wales, the Welsh Government publishes reports investigating how causes and consequences vary over the years. #### 3. Statistics interpretation issues
Determining the level of confidence
In England: “Fire Statistics team run some specific checks before publishing the data. Accuracy can be broken down into sampling and non-sampling error. The data requested and provided by FRSs are not required under legislation, but we aim to achieve 100 per cent response for all fire and rescue collections, therefore reducing sampling error to the minimum.
In order to ensure the IRS data is complete, Home Office statisticians carry out monthly monitoring of the number of incidents submitted as discussed in the quality assurance section. Non-sampling error includes areas such as coverage error, non-response error, measurement error, processing error.
We aim to reduce non-sampling error through the provision of guidance about the data collections and the definitions of the data items.
There are validation checks within the IRS to ensure that data is of good quality and fit-for-purpose” [5]. In Scotland: “the high level totals for primary fires, secondary fires, chimney fires, false alarms and non- fire incidents have a very low margin of error (below 1%) as categorisation issues would largely effect subcategories” [3]. In Wales, “the production of Welsh fire statistics follows a quality strategy”. Moreover, “accuracy can be broken down into sampling and non-sampling error.
Non-sampling error includes areas such as coverage error, nonresponse error, measurement error, processing error.
The fire data requested and provided by fire and rescue services are not required under legislation but we aim to achieve 100 per cent response for all fire collections. There will only be sampling error where there is missing data.
We aim to reduce non-sampling error through the provision of significant guidance about the data collections and the definitions that should be adhered to. We put the data through multiple validation checks to ensure that data is of good quality and fit for purpose” [6].
Pinpointing issues and limitations
In England: despite the review process, “there are likely to be some inaccuracies in the data due to reporting or keying errors, such as misclassification or missing cases” [5].
In Scotland: “As quality assurance is an ongoing process and we currently have no automated means to cross check the IRS logs with control room logs, it is difficult to accurately estimate the ‘true’ error margin, either at the time of publication or following revision, of the subcategories of incidents or casualties.
There may be some miscategorisation which has yet to be addressed, or is not possible to address without access to another data source” [3]. In Wales, “due to time and resource constraints, we sometimes either record data as missing or we seek to impute the data. If we impute or change any data then we inform the data provider of this and give them a chance to comment or challenge this. This is a compromise to resolve validation issues and does not impact significantly on the usability of the dataset” [6].
Examples
ENGLAND Period: April 2019 to March 2020 Figure 39 : Fatality rate (fatalities per million people) for all ages and selected age bands, England;
2009/10 to 2019/20 [7]
Figure 40 : Percentage of fires in accidental dwelling fires by cause of fire, England; 2019/20 [7]
SCOTLAND
Figure 43 : Long-term trend in fatal fire casualties. Note that the series changed from calendar year to
financial year after 1993 [8]
Figure 44 : Long-term trend in non-fatal fire casualties. Note that the series changed from calendar
year to financial year after 1993 [8] WALES
Figure 47 : Percentage of fatal casualties by cause of death [4]
[5] Home Office, “Fire and rescue incident statistics. Methodology and quality report.,” 2020. [6] Welsh Government, “Quality Report for Welsh Fire Statistics,” 2012.
[7] Home Office, “Detailed analysis of fires attended by fire and rescue services. England, April 2018 to March 2019,” 2020.
[8] Scottish Fire and Rescue Service, “Fire and Rescue Incident Statistics (Scotland) 2018-2019,” 2019.
Information from ISO 17755-1 & -2
The National Fire Incident Reporting System (NFIRS) is based on separate reports on each incident requiring a response by a fire department and each casualty associated with a reported incident. There is a national manual for coding of incidents, overseen by the U.S. Fire Administration, which administers NFIRS.
Fire department participation is voluntary, which means a significant fraction of fire departments do not participate, and some participating fire departments do not report every year.
Therefore, NFIRS data is projected using a second database, the NFPA fire experience survey, which is based on summary information reported from a stratified random sample of fire departments. The methods used by most analysts to combine these databases for analysis are documented, but there is no national standard for analysis.
Most reports (60-89%) are completed by firefighters who lack extensive training in fire investigation, but some (11-40%) are completed by professionals with extensive training. Fires subject to reporting (ISO 17755-1, page 5) All fires that result in a fire department response should be reported. An unknown number of departments employ truncated/reporting thresholds. These thresholds are determined on a department by department basis.
Fires are categorized for analysis and reporting purposes according to major incident type in the following way:
Independent of fire incident data collection, a fire related death will be captured when exposure to fire, fire products, or explosion was the underlying cause of death or was a contributing factor in the chain of events leading to death, as reported on the death certificate through vital records reporting channels.
Examples of circumstances that can lead to exclusion of a death, include automobile collision resulting in fire, in which the fire caused death may not be identified as fire deaths.
Deaths captured through coroners/vital records reporting channels are dependent upon recording personnel ability to determine original cause of a fire-related condition that contributes to death. Examples of deaths that may be missed include deaths that occur after the fire incident report is completed – not necessarily just due to extended time lag. Closing the loop between medical and fire reporting systems requires a degree of coordination that does not happen in an unknown number of instances.
Although not common, there may be some fire departments that do not report fire deaths on NFIRS records.
Governmental and Non-Governmental Organizations (NGOs) may choose specific ICD-10 codes for inclusion in analysis of vital records data, depending upon the scope of the study.
United States Fire Administration (USFA) uses ICD-10 codes F63.1, W39-W40, X00-X09, X75- 76, X96-97, Y25-26, and Y35.1 to define fire deaths. Fire injuries subject to reporting (ISO 17755-1, page 10) A fire injury is a person who is injured as a result of a fire, including injuries from natural or accidental causes sustained while involved in the activities of fire control, attempting rescue, or escaping from the dangers of the fire.
Injuries are also captured by a sample survey of hospital emergency rooms and reported to the Consumer Product Safety Commission (National Electronic Injury Surveillance System, NEISS).
Fire injuries that occur in a combination of injuries from an overarching event, such as earthquake or automobile accident, may not be recorded as fire injuries, and non-fire injuries in such circumstances may be recorded as fire injuries when they were not fire injuries.
Other injuries that may be missed include injuries discovered after fire department has closed report, injuries masked by pre-existing conditions, and injuries noticed and treated only by the victim. Also, some fire departments may not report fire injuries on NFIRS records, and far more fire injuries occur in (typically small) fires not reported to fire departments than in reported fires.
Reporting on victim characteristics (ISO 17755-1, pages 20-25) Gender Age Race Ethnicity Severity
Measures used are:
Other losses subject to reporting (ISO 17755-1, page 31)
Note: there are other federal agencies and non-governmental organizations that collect and report various disparate fire loss data, that go beyond the scope of this survey.
Locations of fires based on survey responses (ISO 17755-1, page 32)
Separate reporting of single family dwellings, buildings with multiple private housing units, and commercial residential properties, such as hotels, dormitories
All residential properties reported together but distinguished from other buildings
All buildings with private housing units reported together but distinguished from other buildings. Specific types of buildings and other structures (residential only) (ISO 17755-1, page 40) Residential
One- or two-family dwelling, including manufactured home
Multi-family dwelling
Boarding or rooming house
Hotel or motel
Residential board and care
Dormitory
Sorority or fraternity house
Barracks
Other residential Specific types of rooms and other areas (ISO 17755-1, page 49-51) Means of Egress
Hallway or corridor
Exterior stairway
Interior stairway
Escalator
Lobby
Other egress/exit Assembly or Sales Area
Assembly area with fixed seats for 100 or more people
Assembly area without fixed seats for 100 or more people
Assembly area for less than 100 people
Common room, den, family room, living room, lounge, sitting room
Sales area or showroom
Art gallery, exhibit hall, library
Swimming pool
Other assembly or sales area Function Area
Bedroom for fewer than five people
Bedroom for five or more people
Dining room, bar, cafeteria
Kitchen or cooking area
Bathroom, checkroom, lavatory, locker room
Laundry area
Office
Personal service area
Other function area Technical Processing Area
Laboratory
Photography area
First-aid area
Operating room
Computer room
Performance or stage area
Projection room
Processing or manufacturing area
Other technical processing area Storage Area
Storage room, area, tank or bin
Closet
Tool or supply storage
Records storage
Shipping or receiving area
Trash chute
Garage
Other storage area Service Area
Elevator shaft
Conduit, pipe, utility, or ventilation shaft
Light shaft
Laundry or mail chute
Duct
Display window
Conveyor
Other service area Service or Equipment Area
Machinery room or area
Heating room or area
Switchgear area or transformer vault
Incinerator area
Maintenance shop or area
Test cell
Pressurized air enclosure
Enclosure with enriched oxygen atmosphere
Other service or equipment area Structural Area
Crawl space or substructure area
Exterior balcony
Ceiling/floor assembly or space between stories
Attic or concealed roof/ceiling space
Wall assembly or concealed wall space
Exterior wall surface
Exterior roof surface
Awning
Other structural area Reporting of type of construction (ISO 17755-1, page 53) No, not included in reporting since 1999. Type of construction was included in fire incident reporting during 1980-1998 with these codes.
Definitions are made more specific by reference to the categories of construction defined by model building codes of the time:
Fire resistive
Heavy timber
Protected noncombustible or some type of limited combustible
Unprotected noncombustible or the other types of limited combustible
Protected ordinary
Unprotected ordinary
Protected wood frame
Unprotected wood frame
Unclassified
Unknown Reporting of building height and other characteristics (ISO 17755-1, page 54)
Height of building. Stories at or above grade / Stories below grade
Level or floor where fire began
Structure status, such as vacant, under construction, or under demolition. Reporting and estimation of deliberately-set fires (ISO 17755-1, pages 58-59) (under Cause of Ignition) reported as “Intentional”.
All four loss measures reported as well as fire department deaths and injuries Classification as deliberate – some by trained arson investigators, some by fire officers on the scene with no arson training, some by police or other law enforcement personnel (through separate database, Federal Bureau of Investigation’s Uniform Crime Reports), some by insurance investigators or other insurance personnel Statistical analysis of deliberate fires includes some fires with unknown cause or cause still under investigation Reporting of fireplay (under Factor Contributing to Ignition) and fires can be categorized as both intentional and fireplay.
Under Human Factor Contributing to Ignition, can check “Age was a factor” and then age of person can be entered but is often left blank.
Optional Arson Module offers data elements to report on:
Case status (for example, open, closed, inactive)
Availability of material first ignited (refers to whether fire-starting materials were available at scene or transported to scene)
Suspected motivation factors (see details below)
Apparent group involvement (including gangs, organized crime, hate groups)
Entry method (how the firesetter entered the property)
Extent of fire involvement on arrival of fire department
Type of incendiary device used, if any
Other investigative information (including code violations, other indicators of vulnerability of property, possible motives for fraud, evidence of other criminal activity on site)
Property ownership
Initial observations (including status of doors and windows)
Laboratory used (for analysis of evidence)
Characteristics of subject (for juvenile firesetters), including age, gender, race, ethnicity, family type, motivation (see below), disposition of case motivation factors (not limited to juvenile firesetters)
Extortion
Labor unrest
Insurance fraud
Intimidation
Void contract or lease
Personal
Hate crime
Institutional
Societal
Protest
Civil unrest
Fireplay or curiosity
Vanity or recognition
Thrills
Attention or sympathy
Sexual excitement
Homicide
Suicide
Domestic violence
Burglary
Homicide concealment
Burglary concealment
Automobile theft concealment
Destroy records or evidence
Other suspected motivation
Unknown motivation Suspected motivation or risk factors (limited to juvenile firesetters)
Mild curiosity about fire
Moderate curiosity about fire
Extreme curiosity about fire
Diagnosed or suspected attention deficit (hyperactivity) disorder
History of trouble outside school
History of stealing or shoplifting
History of physically assaulting others
History of fireplay or firesetting
Transiency
Other
Unknown Reporting and estimation of natural cause fires (ISO 17755-1, page 61) Yes, recorded as “Act of nature” under Cause of Ignition. Also, recorded as any of seven choices under Factor Contributing to Ignition:
High wind
Storm
High water, including floods
Earthquake
Volcanic action
Animal
“Other” (unclassified or unknown-type) natural condition Additional details may be provided in the Chemical or Natural Heat Source section under Heat Source:
Sunlight
Spontaneous combustion or chemical reaction
Lightning discharge
Other static discharge
Other (unclassified or unknown-type) chemical or natural heat source “Sunlight” and “lightning discharge” are clearly natural causes and are not identified under any other data element. “Spontaneous combustion or chemical reaction” and “other static discharge” can arise from natural or other causes and so would not be sufficient by themselves to designate a fire as natural, although they would provide additional detail for a fire designated as natural under Cause or Factor Contributing to Ignition. Reporting and estimation of exposure fires (ISO 17755-1, page 62) Yes, recorded in three places. First, can be reported as Exposure Number greater than zero.
Second, can be reported with any of five choices for mechanism of heat transfer from another fire under Heat Source:
Direct flame or convection currents
Radiated heat
Flying brand, ember or spark
Conducted heat
Other (unclassified or unknown-type) Also, recorded as one of the six choices in the Fire Spread or Control section of Factor Contributing to Ignition, but this is difficult to interpret because another of the six choices in the same section is “Other (unclassified or unknown type) fire spread or control” which could mean some exposure fires and some other fires. Reporting and estimation of smoking material and open flame fires (ISO 17755-1, page 65) Yes, can be reported with any of five choices for mechanism of heat transfer from another fire in the Other Open Flame or Smoking Materials part of Heat Source:
Cigarette
Pipe or cigar
Undetermined smoking material
Match
Lighter
Candle
Warning or road flare
Backfire from internal combustion engine
Flame or torch used for lighting
Other (unclassified or unknown type) open flame or smoking material Also, “hot ember or ash” is a choice under the Hot or Smoldering Object part of Heat Source, as are several unrelated categories, all linked to “other (unclassified or unknown type) hot or smoldering object”. Torches are now choices under Equipment Involved in Ignition, as are “cigarette lighter,” “charcoal or utility lighter” and “novelty lighter”.
There are four choices for torches:
Reporting and estimation of cooking and kitchen equipment fires (ISO 17755-1, page 73) Yes, can be reported with any of 28 choices under Equipment Involved in Ignition:
Reporting and estimation of clothes dryer fires (ISO 17755-1, page 75) Yes, can be reported with any of 3 choices under Equipment Involved in Ignition, which are linked along with many other choices to a partially specified choice (other (unclassified or unknown type) personal or household equipment):
Clothes dryer
Washer/dryer combination
Washing machine Other data elements record fuel or power source and portability. Reporting and estimation of entertainment equipment fires (ISO 17755-1, page 77) Yes, can be reported with any of 21 choices under Equipment Involved in Ignition, which are linked along with many other choices to a partially specified choice (other (unclassified or unknown type) electronic or other electrical equipment):
Guitar
Piano or organ
Musical synthesizer or keyboard
Other (unclassified or unknown type) musical instrument
Audio CD (compact disc) player
Laser disk player, including DVD player or recorder
Radio, excluding two-way radios
Two-way radio
Record player, phonograph, turntable
Audio speaker as separate components
Stereo equipment, including receivers, amplifiers, and equalizers
Tape recorder or player
Other (unclassified or unknown type) sound recording or receiving equipment
Cable converter box
Film, slider or overhead projector
Television (TV)
VCR (video cassette recorder) or VCR-TV combination
Electronic video game
Camcorder or video camera
Photographic camera and equipment
Other (unclassified or unknown type) video equipment Other data elements record fuel or power source and portability. Reporting and estimation of office equipment fires (ISO 17755-1, page 79) Yes, can be reported with any of 16 choices under Equipment Involved in Ignition, which are linked along with many other choices to a partially specified choice (other (unclassified or unknown type) electronic or other electrical equipment):
Computer
External computer storage device, including tape or disk drive
External computer modem
Computer monitor
Computer printer
Computer projection device, LCD panel, or projector
Other (unclassified or unknown type) computer device
Adding machine or calculator
Telephone or answering machine
Cash register
Copier
Fax machine
Paper shredder
Postage or shipping meter equipment
Typewriter
Other (unclassified or unknown type) office equipment Other data elements record fuel or power source and portability. Reporting of electrical and electrical distribution or lighting equipment fires (ISO 17755-1, pages 83- 84) Yes, electrical fires can be reported with any of 8 choices in the Electrical Failure or Malfunction section of Factor Contributing to Ignition:
Water-caused short circuit arc
Short circuit arc from mechanical damage
Short circuit arc from defective or worn insulation
Unspecified short circuit arc
Arc from faulty contact or broken conductor
Arc or spark from operating equipment
Fluorescent light ballast
Other (unclassified or unknown type) electrical failure or malfunction The nature of the electrical failure can be inferred to some degree from entries under Heat Source. The following are the eight most frequently reported Heat Source entries for fires reported under Electrical Failure or Malfunction as electrical-failure fires, in U.S. homes:
Electrical arcing
Unclassified heat from powered equipment
Unclassified heat source
Radiated or conducted heat from operating equipment
Spark, ember or flame from operating equipment
Unclassified hot or smoldering object
Heat or spark from friction
Molten or hot material Yes, electrical distribution or lighting equipment fires can be reported with any of 40 choices under Equipment Involved in Ignition:
Electrical power (utility) line
Electrical service supply wires, from utility pole to meter box
Electrical meter or meter box
Electrical wiring from meter box to circuit breaker or fuse box or panel board
Panel board or switchboard
Electrical branch circuit wiring or cable
Outlet or receptacle
Wall switch
Ground fault interrupter, portable or plug-in
Other (unclassified or unknown type) electrical wiring
Distribution-type transformer
Overcurrent disconnect equipment
Low-voltage transformer
Generator
Inverter
Uninterrupted power supply
Surge protector
Battery charger or rectifier
Battery
Table, floor or desk lamp
Lantern or flashlight
Incandescent light fixture
Fluorescent light fixture or ballast
Halogen light fixture or lamp
Sodium or mercury vapor light fixture or lamp
Portable work or trouble light
Light bulb
Other (unclassified or unknown type) lamp or lighting
Night light
Decorative light on line voltage
Decorative or landscape lighting on low voltage
Sign
Electric fence
Traffic control device
Lightning rod
Detachable power cord or plug
Permanently attached appliance power cord or plug
Extension cord
Other (unclassified or unknown type) cord or plug
Other (unclassified or unknown type) electrical distribution, lighting or power transfer Equipment Reporting of other appliance and equipment fires (ISO 17755-1, pages 87-91) Yes, can be reported with many choices under Equipment Involved in Ignition, including the following choices that have not been listed under prior entries in this report:
Power saw
Power lathe
Power shaper, router, joiner, planer
Power cutting tool
Power drill or screwdriver
Power sander, grinder, buffer, polisher
Power hammer
Power nail gun or stapler
Other (unclassified or unknown type) power tool
Paint dipper
Paint flow coating machine
Paint mixing machine
Paint sprayer
Coating machine
Other (unclassified or unknown type) painting tool
Air compressor
Gas compressor
Atomizing equipment
Pump
Wet/dry vacuum
Hoist, lift or crane
Powered jacking equipment
Drilling machinery or equipment
Other (unclassified or unknown type) hydraulic equipment
Heat-treating equipment
Incinerator (also can be reported separately as a type of confined fire)
Industrial furnace, oven or kiln
Tar pot or tar kettle
Casting, molding or forging equipment
Distilling equipment
Digester or reactor
Extractor or waste recovery machine
Conveyor
Power transfer equipment
Power takeoff
Powered valves
Bearing or brake
Picking, carding, or weaving machine
Testing equipment
Gas regulator
Separate motor
Internal combustion engine
Printing press
Car washing equipment
Other (unclassified or unknown type) shop tool or industrial equipment
Dental, medical or other powered bed or chair
Other dental equipment
Dialysis equipment
Medical imaging equipment
Medical monitoring equipment
Oxygen administration equipment
Radiological equipment
Medical sterilizer
Therapeutic equipment
Other (unclassified or unknown type) medical equipment
Transmitter
Telephone switching gear
Television monitor array
Studio-type television camera
Studio-type sound recording or modulating equipment
Radar equipment
Amusement ride equipment
Ski lift
Elevator or lift
Escalator
Microfilm or microfiche viewing equipment
Photo processing equipment
Vending machine
Non-video arcade game
Water fountain or water cooler
Telescope
Electron microscope
Other laboratory equipment
Other (unclassified or unknown type) commercial or medical equipment
Combine or threshing machine
Hay processing equipment
Farm elevator or conveyor
Silo loader, unloader, screw/sweep auger
Feed grinder, mixer, blender
Milking machine
Pasteurizer
Cream separator
Farm or garden sprayer
Chain saw
Weed burner
Lawn mower
Lawn or landscape trimmer or edger
Lawn vacuum
Leaf blower
Mulcher, grinder or chipper
Snow blower or thrower
Log splitter
Post hole auger
Post driver or pile driver
Tiller or cultivator
Other (unclassified or unknown type) garden tool or agricultural equipment
Trash compactor (also can be reported separately as a type of confined fire)
Hot tub or whirlpool
Swimming pool equipment
Other (unclassified or unknown type) floor care equipment
Electric broom
Carpet cleaning equipment
Floor buffer, waxer or cleaner
Vacuum cleaner
Comb or hair brush
Curling iron
Electrolysis equipment
Hair curler warmer
Hair dryer
Lighted makeup mirror
Electric razor or shaver
Sunlamp or suntan equipment
Electric toothbrush
Other (unclassified or unknown type) portable appliance designed to produce heat
Baby bottle warmer
Electric blanket
Heating pad
Clothes steamer
Clothes iron
Automatic door opener
Burglar alarm
Garage door opener
Gas detector
Intercom
Smoke or heat detector or fire alarm
Thermostat
Ashtray
Fire extinguishing equipment
Insect trap
Timer
Model vehicle
Powered toy
Woodburning kit
Clock
Gun
Jewelry-cleaning machine
Scissors
Sewing machine
Shoe polisher
Non-medical sterilizer
Other (unclassified or unknown type) personal or household equipment
Other (unclassified or unknown) equipment
No equipment
Undetermined Other data elements record fuel or power source and portability. Fires can be reported as any of six specific types of confined fires – for which much less detailed reporting is permitted – and two of the six refer to these other types of equipment (incinerator, commercial compactor).
Reporting of item first ignited in terms of form and function (ISO 17755-1, pages 99-101) Yes, can be reported with many choices under Item First Ignited:
Exterior roof covering
Exterior sidewall covering
Exterior trim or appurtenance including door or porch
Floor covering
Interior wall covering
Interior ceiling covering
Structural member or framing
Thermal or acoustical insulation
Other (unclassified or unknown type) structural component or finish
Upholstered furniture
Non-upholstered chair or bench
Cabinetry
Ironing board
Appliance housing or casing
Household utensil
Other (unclassified or unknown type) furniture or utensil
Mattress or pillow
Bedding
Linen other than bedding
Clothing not on a person
Clothing on a person
Curtain, blind, drapery or tapestry
Fabrics and yard goods
Luggage
Other (unclassified or unknown type) soft goods or clothing
Christmas tree
Decoration
Sign
Chips
Toy or game
Awning or canopy
Tarpaulin or tent
Other (unclassified or unknown type) adornment or recreational material
Box or bag
Raw materials being used to make a product
Empty pallet or skid
Cord, rope, twine or yarn
Packing or wrapping material
Baled goods or material
Bulk storage
Palletized material
Rolled or wound material
Other (unclassified or unknown type) storage supplies
Atomized or vaporized liquid
Flammable liquid or gas in or escaping from combustion engines
Flammable liquid or gas in or escaping from final container or pipe before engine or burner
Flammable liquid or gas in or escaping from container or pipe
Uncontained flammable liquid or gas
Pipe, duct, conduit or hose
Pipe, duct, conduit or hose covering
Filter
Other (unclassified or unknown type) liquid, piping or filter
Agricultural crop
Light vegetation excluding crop
Heavy vegetation excluding crop
Animal, living or dead
Human, living or dead
Cooking materials
Feathers or fur not on a bird or animal but not processed into a product
Other (unclassified or unknown type) organic material
Electrical wire or cable insulation
Transformer or transformer fluid
Conveyor, drive or V-belt
Tire
Railroad tie
Fence or pole
Fertilizer
Pyrotechnics or explosives
Book
Magazine, newspaper, writing paper
Adhesive
Dust, fiber or lint
Film or residue
Rubbish, trash or waste
Oily rags
Multiple items first ignited
Other (unclassified or unknown type) item first ignited
Undetermined Reporting of item first ignited in terms of material composition (ISO 17755-1, pages 102-103) Yes, can be reported with any of many choices under Type of Material First Ignited (reporting not required for organic or general materials):
Natural gas
Liquefied petroleum (LP) gas
Anaesthetic gas
Acetylene gas
Hydrogen
Other (unclassified or unknown type) flammable gas
Class 1A flammable liquid
Class 1B flammable liquid excluding gasoline
Class II combustible liquid
Gasoline
Class 1C flammable liquid
Class IIIA combustible liquid
Class IIIB combustible liquid
Other (unclassified or unknown type) flammable or combustible liquid
Fat or grease
Non-food grease
Polish, paraffin or wax
Adhesive, resin, tar or glue
Applied paint or varnish
Combustible metal
Solid chemical
Radioactive material
Other (unclassified or unknown type) volatile solid or chemical
Wood chips, sawdust or wood shavings
Round timber
Sawn wood
Plastic
Rubber excluding synthetic rubber
Cork
Leather
Hay or straw
Grain or natural fiber
Coal or coke
Food or starch
Tobacco
Other (unclassified or unknown type) natural product
Plywood
Fiberboard or particleboard
Wood pulp
Paper
Cardboard
Other (unclassified or unknown type) wood or paper
Fabric or fiber excluding fur and silk
Fur or silk
Wig
Human hair
Plastic-coated fabric
Other (unclassified or unknown type) fabric, textile or fur
Linoleum
Oil cloth
Asphalt-treated material
Other (unclassified or unknown type) material compounded with oil
Multiple types of material
Other (unclassified or unknown type) type of material
Undetermined Reporting of factors in ignition (ISO 17755-1, page 107) Yes, these are the choices under Factor Contributing to Ignition, excluding those cited earlier:
Abandoned or discarded material or product
Heat source too close to combustibles
Cutting or welding too close to combustibles
Flammable liquid or gas spilled
Improper fuelling technique
Flammable liquid used to kindle fire
Washing or painting part or material with flammable liquid
Improper container
Other (unclassified or unknown type) misuse of product or material (also linked to “playing with fire”)
Automatic control failure
Manual control failure
Leak or break
Worn out
Backfire
Improper fuel used
Other (unclassified or unknown type) mechanical failure or malfunction
Design deficiency
Construction deficiency
Installation deficiency
Manufacturing deficiency
Other (unclassified or unknown type) design, manufacturing or installation deficiency
Collision, knock down, run over, turn over.
Accidentally turned on or not turned off
Equipment unattended
Equipment overloaded
Failure to clean
Improper start-up or shutdown procedure
Equipment not used for intended purpose
Equipment not operated properly
Other (unclassified or unknown type) operational deficiency
Rekindle
Outside or open fire for debris or waste disposal
Outside or open fire for warming or cooking
Agriculture or land management burn
Other (unclassified or unknown type) fire spread or control (also linked to exposure fire)
Other (unclassified or unknown type) factor contributing to ignition Reporting of factors in fire growth (ISO 17755-1, pages 108-111) Yes, a data element on Primary Item Contributing to Fire Spread is available with the same choices shown for Item First Ignited. Also can be reported with any of the many choices under Fire Suppression Factors, shown below.
Roof collapse
Roof assembly combustible
Ceiling collapse
Holes or openings in walls or ceilings
Wall collapse
Difficult to ventilate
Combustible interior finish
Balloon construction
Internal arrangement of partitions
Internal arrangement of stock or contents
Floor collapse
Lack of fire barrier walls or doors
Transoms
Attic undivided
Insulation combustible
Stairwell not enclosed
Elevator shaft
Dumbwaiter
Vertical duct
Rubbish or laundry chute
Supports unprotected
Composite plywood I-beam construction
Composite roof/floor sheathing construction
Wood truss construction
Metal truss construction
Bars on windows or other fixed burglar protection assemblies
Quick release failure of bars on windows or doors
Previously damaged by fire
Other (unclassified or unknown type) building construction or design
Door left open or outside door unsecured
Fire door blocked or did not close properly
Violation of applicable or locally adopted fire, building or life safety code
Illegal or clandestine drug operation
Intoxication by alcohol or other drugs
Riot or civil disturbance
Person interfered with operation
Accelerant used
Other (unclassified or unknown type) act or omission
Aisle blocked or improper width
Significant and unusual fuel load from structure components
Significant and unusual fuel load from contents of structure
Significant and unusual fuel load outside from natural environment conditions
Significant and unusual fuel load from man-made condition
Improper storage
Radiological hazard on-site
Biological hazard on-site
Cryogenic hazard on-site
Hazardous chemical, corrosive material or oxidizer
Flammable or combustible liquid hazard
Explosives hazard present
Decorations
Natural or other lighter-than-air gas present
Liquefied petroleum (LP) or other heavier-than-air gas present
Combustible storage extending more than 12 feet above ground
High rack storage
Other (unclassified or unknown type) on-site materials
Delayed detection of fire
Delayed reporting of fire
Alarm system malfunction
Alarm system shut off for valid reason
Alarm system inappropriately shut off
Unable to contact fire department
Information incomplete or incorrect
Communications problem
Blocked or obstructed roadway
Poor or no access for fire department apparatus
Traffic delay
Trouble finding location
Size, height or other building characteristic delayed access to fire
Power lines down or arcing
Poor access for firefighters
Secured area
Guard dog
Aggressive animal excluding guard dog
Suppression delayed due to evaluation of hazardous or unknown materials at incident scene
Locked or jammed door
Apparatus failure before arrival at incident
Hydrant inoperative
Air space restriction
Military activity
Closest apparatus unavailable
Other (unclassified or unknown cause) delay
Automatic fire suppression system problem
Automatic sprinkler or standpipe connection problem
Water supply (private) inadequate
Water supply (public) inadequate
Electrical power outage
Delayed reporting of fire
Alarm system malfunction
Alarm system shut off for valid reason
Alarm system inappropriately shut off
Unable to contact fire department
Information incomplete or incorrect
Communications problem
Blocked or obstructed roadway
Poor or no access for fire department apparatus
Traffic delay
Trouble finding location
Size, height or other building characteristic delayed access to fire
Power lines down or arcing
Poor access for firefighters
Secured area
Guard dog
Aggressive animal excluding guard dog
Suppression delayed due to evaluation of hazardous or unknown materials at incident scene
Locked or jammed door
Apparatus failure before arrival at incident
Hydrant inoperative
Air space restriction
Military activity
Closest apparatus unavailable
Other (unclassified or unknown cause) delay
Automatic fire suppression system problem
Automatic sprinkler or standpipe connection problem
Water supply (private) inadequate
Water supply (public) inadequate
Electrical power outage Presence and type of sprinkler or other extinguishing equipment (ISO 17755-1, page 114) Presence; only one system can be reported.
Present
Partial
None present
Undetermined Type of system; choose the system present in the area of fire origin, and if there are multiple systems in the area, choose the system designed to protect the hazard where the fire started:
Wet-pipe sprinkler system
Dry-pipe sprinkler system
Other sprinkler system (including deluge or pre-action sprinkler system)
Dry chemical system
Foam system
Halogen-type system (including nonhalogenated systems that operate on same principle)
Carbon dioxide system
Other special hazard system
Undetermined Performance of sprinkler or other extinguishing equipment (ISO 17755-1, page 116) Operation of system
Operated and was effective
Operated and was not effective
Fire too small to activate system
System did not operate
Other operation
Undetermined Number of sprinkler heads operating. Reason for system failure [or ineffectiveness]
System shut off
Not enough agent discharged to control the fire
Agent discharged, but did not reach the fire
Inappropriate system for the type of fire
Fire not in area protected by the system
System components damaged
Lack of maintenance, including corrosion or heads painted
Manual intervention defeated the system
Other reason system not effective
Undetermined Presence and type of detection or alarm equipment (ISO 17755-1, page 120) Yes, included in reporting. Presence within designed range of fire; only one system can be reported.
Present
None present
Undetermined Type of detector; choose the system present in the area of fire origin:
Smoke detector
Heat detector
Combination smoke and heat detector in a single unit
Sprinkler, water flow detection
More than one type present
Other detector type
Undetermined Detector power supply:
Battery only
Hardwire only
Plug-in
Hardwire with battery backup
Plug-in with battery backup
Mechanical, including spring-wound, stored pressure source
Multiple detectors and power supplies
Other detector power supply
Undetermined Performance of detection or alarm equipment (ISO 17755-1, page 124) Operation
Fire too small to activate detector
Detector operated
Detector failed to operate
Undetermined Effectiveness.
Detector alerted occupants, occupants responded
Detector alerted occupants, occupants failed to respond
There were no occupants
Detector failed to alert occupants
Undetermined For confined fires, where reduced reporting is permitted, there is a mandatory question that covers some of this information
Detector alerted occupants
Detector failed to alert occupants
Unknown Reason for system failure [or ineffectiveness]
Power failure or hardwired detector shut off or disconnected
Improper installation or placement of detector
Defective detector
Lack of maintenance, including not cleaning
Battery missing or disconnected
Battery discharged or dead
Other detector failure reason
Undetermined Presence of extinguishers or other manual extinguishing equipment (ISO 17755-1, page 126) No, not included in reporting. Presence of smoke management or control equipment by country (ISO 17755-1, page 127) No, not included in reporting. Reporting on fire doors, fire walls and other compartmentation (ISO 17755-1, page 129) No, not included in reporting since 1999. During 1980-1983, the USA used codes similar to those still used in Canada. Some of those codes were used until 1998. References of existing database/studies “Conquering the Unknowns” (2014).
Report by the National Association of State Fire Marshals Fire Research and Education Foundation. Butry, DT, Thomas, DS. (2012). Evaluating Potential Bias in Non-Randomly Reported Fire Incident Data. National Institute for Standards and Technology.
Technical Note 1770.
https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=912210
Anderson, A, Ezekoye O. (2017). Exploration of NFIRS Protected Populations Using Geocoded Fire Incidents. Fire Technology. https://i-psdi.org/assets/research/exploration-of-nfirs-protected-populationsusing-geocoded-fire-incidents.pdf United States Fire Administration. (2017). Review and Assessment of Data Quality in the National Fire Incident Reporting System.
https://www.usfa.fema.gov/downloads/pdf/publications/nfirs_data_quality_report.pdf
NFPA 901, Standard Classifications for Incident Reporting and Fire Protection Data www.nfpa.org/901
Summaries of existing database
The National Fire Incident Reporting System (NFIRS) relies upon local fire departments to collect and submit data to state fire protection agencies or through their server partitions. Data is then processed by the state agencies and submitted or released to the National Fire Data Center. NFIRS was created following the publication of America Burning in 1973 by the National Commission on Fire Prevention and Control.
This report culminated in the passage of the Federal Fire Prevention and Control Act of 1974, which authorized the USFA to gather and analyze information on the magnitude of the nation's fire problem, as well as its detailed characteristics and trends. NFIRS has been through several iterations since its initial release.
USFA has developed a standard data collection package that includes separate forms for incidents and casualties (fatal and non-fatal, firefighters and civilians), a coding structure for data processing purposes, reference guides, computer software and procedures, documentation and a National Fire Academy training course for utilizing the system.
While reporting is voluntary at the federal level, different states have different reporting requirements, ranging from mandatory for all incidents to completely voluntary.
Existing definitions
A data dictionary with a list of definitions is available here.
Are there differences within the same country?
Incident reports and data elements are standardized. Completeness of data entered into incident reports may vary by locality.
There are differences in interpretation. “Burnt food” may be considered a fire, excessive heat, a smoke scare, or a false alarm. The data dictionary is used in software but lacks detail found in the is used in software but lacks detail found in the NFIRS 5.0 Complete Reference Guide See
https://www.cpsc.gov/s3fs-public/pdfs/neissfire.pdf . He found that an estimated 21,174 of 48.202 civilian
non-arson fire injuries resulted from residential or consumer-product fires attended by fire departments from July 1, 2002 to June 30, 2003. NFPA estimated totals of 18,425 civilian injuries in 2002, and 18,125 in 2003, including injuries caused by arson. Some of the injured may have left the scene before the fire department arrived or been transported by a non-fire department organization.
Are there differences and contradictions with other domains?
There are no differences within NFIRS domains. However, there are some differences with the ICD-10 coding of death certificates.
For instance, NFPA counted the fatalities in the World Trade Center and Pentagon attacks as fire deaths, The death certificates called them terrorism. It can also be unclear to determine when vehicle fire deaths should be counted as transportation events rather than fires in the ICD- 10 codes.
Small fires that did not require suppression but result in death – such as fatal burns caused by smoking while on oxygen -- may be considered EMS and not fire deaths based on the responder. , who may be emergency medical crews not affiliated with the fire department. NFPA’s estimates of fire deaths derived from our fire department experience survey seem close to fire death data.
See pages 7-8 of NFPA’s 2019 report, US Fire Death Rates by State at: http://www.nfpa.org//- /media/Files/News-and-Research/Fire-statistics-and-reports/US-Fire-Problem/osstate.pdf
Identification of missing information
Missing data is a long-established problem with NFIRS.
Undetermined or missing information on fire cause is a particular concern and is seen to be especially prevalent in the case of more serious fires, as indicated in a 2014 report by the National Association of State Fire Marshals Fire Research and Education Foundation. Some jurisdictions refuse to report dollar loss. Some have policies that require causal information to be reported as undetermined when fires are referred for investigation.
To make it easier for firefighters, information about causal factors and details on fire protection are not required for six types of structure fires, collectively called confined fires.
These include: confined cooking fires, confined chimney or flue fires, confined fuel burner or boiler fires (mostly oil burner blowbacks), confined compactor fires, confined incinerator fires, and rubbish trash fires in or on a structure that did not extend to the structure or other contents. The type of material first ignited is not required for certain types of items first ignited, such as cooking materials, including food, dust, electrical wire, etc.
Some data elements were left optional. “None” is a choice in some data elements such as factor contributing to ignition and equipment involved in ignition that can seem like an easy out.
NFPA’s Fire Experience Survey uses population and survey data to get big picture estimates that include fires reported to local fire departments but not in NFIRS.
USFA has been exploring the possibility of using more in-depth analysis of NFIRS to identify and compensate for unreported fires. See the 2017 white paper, National Fire Estimation Using NFIRS Data at https://www.hsdl.org/?view&did=828558.
Fire department responsibilities
Fire departments are responsible for entering data with key details from the incidents for which they are dispatched, which include not only fires, but also emergency medical services, severe weather and natural disasters, and other incidents. Separate reports are filed for incidents and, if applicable, casualties (fatal and non-fatal). Firefighter casualties and civilian casualties utilize separate reports.
Fire response organisation
Municipal fire departments (volunteer, career, and mix) participate. Department of Defense also participates, but their data is not publicly released.
NFIRS and NFPA estimated generally do not include data from federal or state firefighting organizations or industrial fire brigades.
Who collects data?
Fire departments are responsible for collecting data on calls that they respond to.
In addition to fires, they include reports for emergency medical services, vehicle accidents, weather or natural disasters, and hazardous conditions. Incident reports of firefighter injuries include injuries during incident responses as well as non-incident events.
Police provide some data on arson and arson arrests to Federal Bureau of Investigation’s (FBI’s) Uniform Crime Reporting System. https://www.fbi.gov/services/cjis/ucr Fire investigators can provide data to the Alcohol, Tobacco and Fire Arms’ (ATf’s) Bomb Arson Tracking System and use its case management features. See https://www.atf.gov/explosives/bomb-arson-trackingsystem-bats. ATF also participates in some investigations.
The Chemical Safety Board investigates incidents involving chemicals.
See https://www.csb.gov/about-thecsb/ The US Consumer Product Safety Commission’s (CPSC’s) National Electronic injury Surveillance System (NEISS) collects data from a sample of hospital emergency departments about injuries caused by consumer products, and records if fire was involved and if the fire department attended. See
https://www.cpsc.gov/Research--Statistics/NEISS-Injury-Data NEISS collects less information is collected when a consumer product was not involved. Fires or burns are grouped together are grouped together in a database of non-fatal injuries. See: https://www.cdc.gov/injury/wisqars/nonfatal.html
The National Center for Health Statistics collects vital statistics, including death certificates. These may be accessed at https://www.cdc.gov/injury/wisqars/fatal.html. NFPA collects summary data from a subset of local fire departments to estimate fires, civilian deaths and injuries, firefighter injuries and direct property damage from fire by broad incident type and property class through its fire experience survey (FES).
Surveys are sent to all departments protecting at least 5,000 and a random sample of departments from less populated areas.
Who issues the data?
State programs at the state level. The United States Fire Administration provides the national database to interested parties.
Narratives are not included in the public data release. NFPA, USFA, and CPSC all release some type of national estimates of the fire problem with CPSC focusing on residential fire losses.
Are there different levels of collection?
Data at the state level includes data submitted by local fire departments. National data includes data on all incidents submitted by state agencies. Local fire departments may analyse their own data. NFPA, CPSC, USFA, consulting firms, students, etc. analyse the data
Identify disparities in data feedback
Since data is collected at local levels -- which vary by resources, staffing, and leadership – there are substantial opportunities for disparities between jurisdictions related to the completeness and accuracy of data. Different fire departments and states provide different levels of quality control.
Where is the data stored?
State agencies house information submitted by local fire departments on their systems or in partitions on the federal server, or some combination of the two.
The National Fire Data Center is the custodian of data at the national level. #### 2. Statistics collection issues Who is interpreting the statistics? Each organization determines their approach to handling unknowns (and deciding what should be classified as unknown) and deciding how to group things.
Purpose for which data is collected
NFIRS data may be used by fire prevention and public education programs to target at- risk groups or to address critical problems, fire officials use the data in decision making that affects the allocation of firefighting resources, and consumer groups and litigators may use the data to assess product fire incidence. The NFIRS report may stand as the public record of an event. Engineers use it for modeling.
Policymakers and administrators may use data to justify budgets and or support legislation related to fire issues. Regulators may use it in the evaluation of performance-based designs.
Media uses it for context,
What are the methods used to fill the gaps where information is missing?
USFA seeks to accommodate missing information by including an “undetermined” option for many data elements and encouraging fire departments to update reports if and when a determination is made. In practice, however, information appears to be rarely updated once the undetermined selection is made. USFA does not make post-hoc statistical adjustments to its data in order to account for data gaps.
The National Fire Protection Association, which uses NFIRS data in its own analyses, does apply a “scaling factor” in order to compensate for fires reported to fire departments but not to NFIRS, and allocate unknown values proportionally among known values. USFA uses a cause hierarchy of multiple data elements. If a fire does not “fit” anywhere, it is considered unknown.
While NFPA allocates unknowns separately for each data element, USFA allocates only those that make it through the hierarchy. NFPA also allocates unknowns when sufficient data are provided for optional data elements.
Is there follow up to data collected?
State program managers are responsible for cleaning data and reconciling inconsistencies prior to submission to the national level. Although causal information for incidents under investigation can be updated once an investigation is complete, this is frequently not done.
While local records can be updated at any time, once the federal database is released, updates will not be incorporated
Analyse potential cause and consequences in trends
Data indicate that fires and losses in the U.S. have dropped dramatically since the early 1970s.
Undoubtedly a number of factors play a role in this decline, but education, the adoption of more stringent fire codes, and increased use of fire detection are key factors. The recognition of fires as a serious problem in need of intervention and the introduction of NFIRS in the 1970s as a way to document where problems lay and solutions might be targeted must certainly be recognized as a vehicle for reducing the threat of fire. #### 3. Statistics interpretation issues
Determining the level of confidence
USFA reports that approximately 75 percent of fires reported in the U.S. are included in the NFIRS database. More than 24,000 fire departments participated in NFIRS in 2018. In addition, 38 fire departments representing protected populations of more than 500,000 residents participate in NFIRS. Because NFIRS was not designed as a statistical sample, it is hard to determine levels of confidence. In addition, tensions exist between fire authority needs and NFIRS guidelines.
For example, NFIRS has advised fire departments not to use NFIRS to report fatal fires if the fire department did not respond to them.
See: usfa.fema.gov/data/nfirs/support/nfirsgrams/nfirsgram_reporting_deaths_no_response.html. However, a state may wish to have all of the fire deaths in their NFIRS database for analysis. Little attention has been paid to the reliability of the data.
It is clear that many firefighters who are doing their best to complete the reports disagree about how the reports should be coded. A study of representativeness relating NFIRS to NFPA’s FES was published in 2011, See: reginfo.gov/public/do/DownloadDocument?objectID=53926600 Pinpointing issues and limitations
Staffing for better-resourced state NFIRS programs may include multiple administrative, research, and support staff. Other states with fewer resources may rely upon a single person.
Some state programs include full-time research analysts, while others rely upon administrators, information technology staff, or investigators to run their programs, often on a part-time basis.
Resource disparities also influence the level and sophistication of the technology utilized by state NFIRS programs. USFA provides basic – but not particularly sophisticated – software to states at no cost, but individual fire departments or states can choose their own software. Some state programs provide licenses to a specific software vendor, and large fire departments may have software integrated with computer assisted dispatch in custom systems.
The variety of systems within and between states makes it difficult to introduce changes in the NFIRS system. In addition, because USFA is within the U.S.
Department of Homeland Security, security requirements also complicate the introduction of changes and makes them more expensive.
It is also important to recognize that funding at both federal and state levels is subject to the vicissitudes of both the policy environment and economic conditions, making the sustainability and consistency of support for NFIRS a potential ongoing challenge.
Another issue with respect to the completeness and accuracy of reporting is that NFIRS codes are seen to be overly complex, resulting in frustration that can deter reporting.
Efforts to improve quality of data with new relational edits have at times had unintended consequences. For example, requiring a valid equipment involved in ignition when heat source or factor contributing to ignition resulted in more unknowns in those fields.
This edit was later removed. Because the list of code choices is so long for many data elements, many fire departments use cheat sheets with the most commonly used code choices. More rarely used codes may be completely forgotten.
Examples
Key Findings from NFPA’s Fire Experience Survey (https://www.nfpa.org//-/media/Files/News-and- Research/Fire-statistics-and-reports/US-Fire-Problem/osFireLoss.pdf): In 2019, local fire departments responded to an estimated 1.3 million fires. These fires caused roughly 3,700 civilian fire deaths and 16,600 reported civilian fire injuries. Property damage was estimated at $14.8 billion.
Key Findings from NFPA Home Structure Fire Report: During 2014 - 2018, US fire departments responded to an estimated average of 353,100 home structure fires per year.
These fires caused an annual average of 2,620 civilian deaths, 11,030 civilian fire injuries, and $7.2 billion in direct property damage. Sixty-nine percent of reported home fires in 2014–2018 were in one- or two-family homes, including manufactured homes. These fires caused 85 percent of home fire deaths, 65 percent of home fire injuries, and 79 percent of the direct property damage from home fires.
Most home fires and fire casualties result from five causes: cooking, heating equipment, electrical distribution and lighting equipment, intentional fire setting, and smoking materials.
Over the five-year period of 2014–2018 as a whole, cooking was the leading cause of home fires and home fire injuries, while smoking was the leading cause of home fire deaths.
Leading causes of home structure fires: 2014–2018 Key Findings from NFPA Fires in Structures under Construction or undergoing Major Renovation report (https://www.nfpa.org/News-and-Research/Data-research-and-tools/Building-and-Life-Safety/Fires-in- Structures-Under-Construction-or-Renovation): Local fire departments responded to an estimated average of 3,840 fires in structures under construction and 2,580 fires in structures under major renovation per year in 2013-2017.
The fires in structures under construction caused an average of four civilian deaths, 49 civilian injuries, and $304 million in direct property damage annually, while those in structures under major renovation caused averages of eight civilian deaths, 52 civilian injuries, and $104 million in direct property damage annually.
Task 1 of the EU FireStat project is focused on understanding the terminology and data collection methodology used by the 27 EU Member States and 8 Other European and Non-European countries (Australia, Canada, New Zealand, Norway, Russia, Switzerland, UK and USA); a total of 35 countries were examined. The 8 Other European and Non-European countries have been chosen based on their structured and detailed fire statistics.
The aims of Task 1 are mainly focused on increasing the awareness of the data fields collected in the various countries fire statistics and understanding the definitions provided for the terminologies adopted in those statistics.
It is important to evaluate the variables collected related to the pre- and post-fire conditions of fire incidents that affect buildings.
In Task 1 summary tables were created for each country examined and these tables focused on major areas of investigation, such as: fire incident, building description, fire causes, fire consequences, fatalities, casualties, fire safety measures, fire response, fire costs, and fire prevention. The differences that may appear in the fields recorded in the summary tables of the countries investigated are due to the differences in the fire statistics and collection methodologies available in each of them.
The summary tables are subdivided into two parts: Part 1 relates to the definitions adopted in the fire statistics and Part 2 relates to the data collection methodology and available fields recorded in the fire statistics.
One aspect to consider is that not all the fire statistics of the countries examined were structured in a way that enabled the completion of Part 1 and Part 2 and thus there is a difference in the number of countries investigated for the definitions and fields recorded in the fire statistics. Abstracts of the summary tables are described in Section 2 for 24 EU and 8 Other European and Non- European countries, and a general description of the current practice of fire statistics in each country is provided.
This description highlights the responsible authorities and challenges faced by the consortium groups in gathering the information.
In addition, it gives a short description of Parts 1 and 2 of the summary table and examines key aspects and limitations encountered in the process.
Part 1 is focused on the definitions and the related tables for 15 EU and 8 Other European and Non-European countries are available in Appendix I and Appendix II, respectively. The references for each fire statistics have been specified in the tables provided in Appendix I and Appendix II where each definition is related to the specific recording system with the appropriate links, where possible.
In Section 3, the evaluation of the analogies and differences in the definitions provided in the fire statistics has been developed considering the major areas of investigation and comparing the terminologies used with those provided by ISO/TS 17755- 2:2020(E).
The fields recorded in the fire statistics for 21 EU and 8 Other European and Non-European countries are investigated in Part 2 of the summary table, which is considering the major areas of investigation, as discussed in Section 4.
It is suggested to link the considerations presented in the abstract of Section 2 with the definitions provided in Section 3 and the fields covered by the various fire statistics in Section 4. Overall, the research developed in Task 1 will contribute to subsequent tasks in the EU FireStat project as well as towards a better understanding of current practices, thus identifying optimization measures and providing insights towards harmonised fire statistics.
The data collected in the aftermath of real fire incidents in various building types can vary significantly according to the country considered, who collects the data, the specific data fields recorded, how each field is defined, and the collection methodology adopted. In Europe, a single unique fire statistics terminology and methodology system is not yet available and each Member State has its own recording system, fire statistics fields, and data elaboration.
Moreover, within a country, several databases of fire statistics, not always publicly available, could be present and managed by various organizations.
Data are usually collected by the fire brigades attending the fire scene in the aftermath of an event and inserted into an online database.
However, in some circumstances, data collection is voluntary and/or uses a paperbased collection system, and fire incident information could also be provided by police, private and state fire investigators, fire engineers and insurance companies. Even though similar terminology and factors of the fire statistics may be encountered in various recoding systems, their meanings and definitions can vary also based on the relevance that these fields assume in the various fire statistics.
Such inconsistencies often lead to challenges and have the potential for errors particularly when the data are compared between countries.
Based on the above, the aim of Task 1 of the EU FireStat project is to increase the awareness of the fire statistics available, deeply investigate the semantic differences used in the fire safety fields and the recording approaches for the information gathered.
Task 1 is focused on the terminologies and data collection methodologies of the 27 EU Member States and 8 Other European and Non-European countries (Australia, Canada, New Zealand, Norway, Russia, Switzerland, UK and USA) for a total of 35 countries examined. The two objectives of Task 1 mainly cover the: 1. Terminology
to create a complete inventory of the terminology adopted regarding pre and post-fire incident conditions in various building types. 2. Data collection methodology
to create an overview of which fire data are collected;
to create an overview of when they are collected;
to record who collects the information; and
to establish the quality assurance process adopted. Task 1 started on 22nd September 2020 and ended on 22nd February 2021. The following report presents the progress developed, main findings and specific conclusions obtained from the research. In Task 1, the research is focused on the analysis of the definitions and fields recorded in the fire statistics of EU, Other European and Non-European countries.
For each of them, in the abstracts provided, it is specified who collects the fire statistics, the recording system adopted, and the origin of the information gathered.
Instead of providing a unique list of reference at the end of the report, the references have been specifically addressed for each country investigated in the tables provided in Appendix I and Appendix II where each definition is related to the specific recording system. Moreover, in the references, where available, links to this information have been provided.
Furthermore, analysing the forms in and by themselves is somewhat beyond the scope of this task, particularly given the large number of languages involved. Our project has the goal to provide a clear understanding of the fire statistics related to buildings subjected to fire incidents and does not include the evaluation of “near misses” which are usually not collected in the recording systems examined.
For instance, in Scandinavia, reports can be created for fire spread in criminal cases, to judge how dangerous it could have been for human beings/property, if accidental circumstances had not prevented fire spread.
These evaluations represent a useful field of investigation to identify physical and societal hazards and support the creation of preventive measures.
The collection of “near misses” implies, in some cases, a detailed and challenging assessment able to determine benefits for user input. However, such reporting could also result in uncertainty in the data. Furthermore, it would also require a new system able to describe such investigation and could lead to a much higher need for resources to check the correctness and treat the data once a reporting system is in place.
The outputs generated in Task 1 intend to inform the European Commission and Parliament, National and Local authorities, regulators, policymakers, Fire and Rescue Services, International Fire Safety Community and the General Public.
Furthermore, the analyses will hopefully contribute to a thorough evaluation of current practices, including identification of missing fields that are relevant to fire safety.
The methodology of Task 1 is based on the collection of fire statistics through publicly available information and by contacting the responsible organizations in Europe and the other international countries. Direct contacts in each of the countries investigated have been established and considered at the base of the outputs generated. Elaborations are determined through an analysis of the summary tables. In detail, Task 1 is subdivided into 5 subtasks:
The summary table has been created considering and improving the information presented in Task 0 and is considered as a guide to follow where the differences that may appear in the structure of the countries examined are due to the differences in the fire statistics and collection methodology available in each of them.
The summary table is adapted to the information gathered and is always structured in two parts: Part 1 definitions adopted and Part 2 data collection methodologies and available fields recorded in the fire statistics.
Fire incident
Building description
Fire causes
Fire consequences
Fatalities
Casualties
Fire safety measures
Fire response
Fire financial costs
Fire prevention
Along with language barriers, confidentiality policies, private databases and lack of responses represent the most challenging aspects encountered while developing Task 1. The aforementioned risks have been overcome by establishing direct contact with the relevant authorities for the fire statistics of the countries, and by asking for their kind contribution to our project. The countries covered have been assigned to the consortium groups based on their location and existing interaction with relevant countries.
It is really important to acknowledge the precious collaboration and cooperation of the relevant authorities and fire brigades who kindly provided the information necessary to fill in the summary tables and allow the comparisons between countries.
The aspects of the fire statistics examined in each summary tables developed in Task 1 cover pre- and postfire conditions of fire incidents in buildings, and they have been classified according to 10 major areas of investigation: fire incident, building description, fire causes, fire consequences, fatalities, casualties, fire safety measures, fire response, fire costs, and fire prevention.
As shown in Table 1, many of the 10 major areas of investigation have various subcategories and the relevant authorities have been asked to indicate the fields covered in their fire statistics. Considering the major areas of investigation, the summary tables could partially bias the information received from the various countries. However, defining generic groups was necessary to have a method of comparison.
From the analysis developed, it appears difficult to evaluate the mandatory and optional fields collected by the various fire statistics. Within a specific country, fire statistics could be a voluntary system, differently managed at a local level or, considering a unique recording system (e.g., UK), only a number of fields are mandatory while others could be filled in only if specific fire conditions appear.
In the description provided by the abstracts, such differences have been highlighted to provide a clear overview of current practice in various countries. The consortium has created summary tables for a total of 35 countries. Unfortunately, no information has been received for Luxembourg, Malta and Portugal, and limited information for Lithuania and Spain. Therefore, the fire statistics of 24 EU countries and 8 Other European and Non-European countries have been investigated, thus 32 countries covered (Table 2).
For all the summary tables developed in Task 1, a short abstract is provided in Section 2.
The information collected has been summarised in Section 2 for EU, Other European and Non-European countries, based on the relevant information and the methodology adopted to provide descriptions about: - who collects the fire statistics; - if fire statistics are national or local; - the number of datasets available; - if definitions for each field of the fire statistics are provided, given by classification or not available; and - the various fields recorded in the fire statistics.
Part 1 of the summary tables related to the definitions is provided in Appendix I for the EU and Appendix II for the Other European and Non-European countries. Where no definition is available and the specific fields are included in dropdown menu, this is clearly stated. The possible responses in Appendix I and Appendix II have to be considered as follows: - “a”: fields available, - “b”: definitions not available, and - “c”: fields not clear to the relevant authority of the fire statistics.
A deep analysis of the definitions determining analogies and differences is described in Section 3 with a comparison with the terminology provided by the ISO/TS 17755-2:2020(E) (named ISO TS 17755-2 in this document). A semantic evaluation of the definitions is fundamental to understand what is covered by each term and allows correct comparisons amongst variables. Unfortunately, not all the countries considered in Task 1 have an available glossary or list of definitions for the fire statistics.
Therefore, the analysis is focused on the definitions available for fire incidents, fatalities and casualties, damage, safety measures, response time and the financial costs of fire for 23 countries (15 EU and 8 Other European countries). Part 2 of the summary table, focused on the fields recorded in the various fire statistics, is summarized in Section 4 based on the major areas of investigation highlighted in Table 1 for a total of 29 countries (21 EU and 8 Other European and Non-European countries).
| Country | Definitions | Statistics fields | Status | Comment |
|---|---|---|---|---|
| Austria | YES | YES | Completed | |
| Belgium | NO | YES | Completed | |
| Bulgaria | NO | YES | Completed | |
| Croatia | NO | YES | Completed | |
| Cyprus | NO | YES | Completed | |
| Czech Republic | YES | YES | Completed | Part 1 — definitions given by law and government but not specified |
| Denmark | YES | YES | Completed | |
| Estonia | YES | YES | Completed | |
| Finland | YES | YES | Completed | |
| France | YES | YES | Completed | |
| Germany | YES | YES | Completed | |
| Greece | NO | NO | Completed | |
| Hungary | YES | YES | Completed | |
| Ireland | YES | YES | Completed | |
| Italy | YES | YES | Completed | |
| Latvia | NO | YES | Completed | |
| Lithuania | NO | NO | Limited information | |
| Luxembourg | NO | NO | No information received | |
| Malta | NO | NO | No information received | |
| Netherlands | YES | YES | Completed | |
| Poland | NO | YES | Completed | |
| Portugal | NO | NO | No information received | |
| Romania | YES | YES | Completed | |
| Slovakia | YES | YES | Completed | |
| Slovenia | YES | YES | Completed | |
| Spain | NO | NO | Limited information | |
| Sweden | YES | YES | Completed |
| Country | Definitions | Statistics fields | Status |
|---|---|---|---|
| Australia | YES | YES | Completed |
| Norway | YES | YES | Completed |
| Canada | YES | YES | Completed |
| USA | YES | YES | Completed |
| New Zealand | YES | YES | Completed |
| UK — England | YES | YES | Completed |
| UK — Scotland | YES | YES | Completed |
| UK — Wales | YES | YES | Completed |
| UK — Northern Ireland | NO | NO | Completed |
| Russia | YES | YES | Completed |
| Switzerland | YES | YES | Completed |
The outputs of Task 1 of the EU FireStat project will be considered as inputs in Task 2, where data needed for decision making are evaluated, Task 3, focused on data collection methodologies, Task 4 where a unified terminology will be proposed. Finally, the results generated by Task 1 will increase awareness of the fire statistics collected in the countries examined.
The research developed will highlight not only the differences and similarities of the terms, data recorded and methodology adopted in the fire statistics of various countries, but also the missing aspects necessary for the prevention of fire frequency and limitation of fire consequences that could be improved in short- and long-term.
The outputs of Task 1 have to be considered as an extended map of fire statistics applied worldwide able to support future horizons of research and introduce a holistic approach based on the collaborations and cooperation amongst various nations in Europe and internationally.
Abstracts of the summary tables completed for 24 EU (Luxembourg, Malta and Portugal are not described) and 8 Other European and Non-European countries are presented in the following sections to provide an overview of the information gathered in Part 1 related to the definitions and Part 2 related to the fire safety fields recorded.
In the abstracts, the fire statistics of each country is described considering the national or local recording system, the authorities responsible for the data collection and management with a description of the recording system and the datasets available. Challenges faced by the consortium groups for the collection of the relevant information, if any, are also introduced.
A short description of Parts 1 and 2 of the summary table related to definitions and fields of the fire statistics available are present and key aspects and limitations encountered are examined. The information provided in the abstract of each country needs to be related to the available definitions provided in the tables of Appendix I and Appendix II.
DEMA provides some definitions for almost all of the 10 major areas of investigations with specific fields recorded.
In particular, there is no definition for accidental and deliberate fires, but there are definitions for false alarms. Fatalities, fire safety measures, fire response and fire prevention are provided/specified. There are no clear definitions for causalities and the building description, fire causes and fire spread are described by a dropdown menu with clear classes. No evaluation of direct or indirect costs of fire is present.
A national fire statistics is present.
Fire statistics in DEMA are present with some definitions and the fields available cover the major areas of investigations.
No evaluation of the economic impact of fire is available
Having a formal definition of what are false alarms differentiate Finland from many countries. Also, having most of the fire safety measures defined is remarkable.
The data is checked for its quality and therefore increases its reliability.
At present, the MSB do not have access to this data, and until the legal situation is clarified, there will be some under-reporting of fatal fires not attended by the fire service. The statistics are available at the IDA website. The number varies over time, from 80 to 130 per year.
In addition, the National Board of Health and Welfare (Socialstyrelsen) publishes cause of death statistics, covering all people dying in Sweden as well as residents of Sweden who die in foreign countries. The statistics use the international classification of diseases, so fire deaths can be identified. The numbers vary over time. In 2019, there were 47 fire deaths. The highest figure in recent years was 138 in 1998. Finally, the Swedish insurance companies collect data on insurance claims due to fires.
The latest figures are 27,794 in 2018 and 34,217 in 2019.
No special challenges have been encountered during the project except that it was sometimes difficult to translate specific terms and present them in the project’s predefined table structure.
Concerning the fire itself, data recorded includes item first ignited, ignition source and room of origin, as well as the fire spread expressed as contained in the item of origin, room of origin, fire compartmentation, initial building or spread to another building. The fatal fire database contains data on the fire and its victims. The cause of death register only contains the number of deaths per cause. The insurance companies have data on claims and the costs reimbursed for the fire. Further comments
DSB provides some definitions for almost all of the 10 major areas of investigations with specific fields recorded.
In particular, there is no definition for accidental and deliberate fires, but there are definitions for false alarms. Fatalities, fire safety measures, fire response and fire prevention are provided/specified. There are no clear definitions for causalities and the building description, fire causes and fire spread are described by a dropdown menu with clear classes. However, in Norway, the incident commander (Fire) must make an assessment of the cost of fire. This must be registered in BRIS.
The insurance companies make their own collection of data which one can apply for the access.
Fire incidents, building description, fire causes and consequences are present, as well as fatalities, fire safety measures, fire response and fire prevention. Further comments
A national fire statistics is available.
Fire statistics in DSB are present with some definitions and the fields available cover the major areas of investigations.
The assessment of the cost of the fire is done by the incident commander and is registered in the collection form.
The data collected by the insurance companies can be access with permission.
The data set includes separate modules for fire incidents and casualties (deaths and injuries for civilians and firefighters).
Basic module: State, incident data, location (street, city, etc.), type of incident (structure, vehicle, rubbish, etc.), alarm time, arrival time, controlled time, last unit cleared time, actions taken (extinguishment, overhaul, search and rescue, etc.), estimated dollar losses of property and contents, original value of property and contents, casualties, detectors, hazardous materials release, property use (assembly, educational, health care, residential, etc.), person or entity involved (name, phone, address, city, etc.) owner (name, phone address, city, etc.).
Fire module: State, date, property details (number of buildings, residential living units), on-site materials or products, area of fire origin, heat source, item first ignited, type of material first ignited, cause of ignition, factors contributing to ignition, human factors contributing to ignition, equipment involved in ignition, equipment type, equipment power source, equipment portability, mobile property involved.
Structure fire module: structure type (enclosed building, portable or mobile structure, air-supported structure, etc.), building status, building height, main floor size, story of fire origin, fire spread, number of stories damaged by fire, item contributing most to flame spread, type of material contributing most to flame spread, presence of detectors, detector type, detector power supply, detector operation, detector effectiveness, detector failure reason, presence of automatic extinguishing equipment, type of automatic extinguishing equipment, operation of automatic extinguishing equipment, number of sprinkler heads operating, reason for automatic extinguishing system failure.
Civilian fire casualty module (injury and fatality): state, date, fire department ID, injured person, number of casualties, age or date of birth, race, ethnicity, affiliation (civilian, non-fire department emergency medical services, police, other), date and time of injury, severity, cause, human factors contributing to injury, factors contributing to injury, activity when injured, specific location when injured, general location when injured, story at the start of the incident, story when injury occurred, primary apparent symptom, primary area of body injured, disposition (transferred to hospital, etc.).
Fire service casualty module (injury and fatality): state, date, fire department ID, victim name and ID, gender, affiliation, number of casualties, age, date of birth, date and time of injury, usual assignment, physical condition just prior to injury, severity, taken to, activity at time of injury, primary apparent symptom, primary body part, cause of injury, factor contributing to injury, object involved in injury, where injury occurred, story where injury occurred, specific location where injury occurred, equipment sequence number, protective equipment item, protective equipment problem, equipment manufacture, model, and serial number.
NFIRS is the world’s largest and most detailed collection of fire incident data. Data is collected and published on an annual basis. NFIRS includes incident and casualty forms, a coding structure for data processing, comprehensive manuals for users, and computer software and procedures. A basic web-based data entry program is available and used mostly by smaller fire departments. The National Fire Academy also sponsors a training program that details how to use the system. Local fire departments may use vendor software with state permission. Some states have purchased software licenses for all fire departments in the state.
Some fire departments have custom software that is integrated with other public safety functions.
Those who are tasked with entering data may not have enough time to complete all the modules or spend time identifying the best coding option, so they do the best they can with the time they have. This may be particularly true for volunteer fire departments. Consequently, data quality may vary. NFIRS coding options can be confusing and lead to inconsistent coding responses, even between experienced NFIRS researchers.
Many data elements had too many coding options or had options that were not clear, which can lead to no decision or poor data reliability. The authors also pointed out that memory is limited and that long lists of codes can frustrate those who are responsible for entering data.
Some states have dedicated program managers to perform quality control and provide feedback to fire departments, but many states lack the resources to perform this function.
States do not receive federal funding to support personnel engaged in NFIRS. NFIRS also records a high share of “unknown” responses to a number of data elements, possibly a result of being such a detailed reporting system. It is difficult to make changes to NFIRS because of the many different software packages in use.
While England, Scotland and Wales have adopted the Incident Recording System (IRS) for the collection of the fire incident data, Northern Ireland has its own system.
Therefore, four different fire statistics datasets are available with three of them based on the IRS.
However, the fire statistics of England, Scotland and Wales are published annually with a limited number of fields recorded. Information for Northern Ireland has been requested and its fire statistics description has not been inserted in the summary table due to delay in the response.
For the fire safety measures, definitions for alarms, automatic extinguishing systems, compartmentations and escape routes are described by a dropdown menu with clear classes.
No evaluation of direct or indirect costs of fire is present.
A national fire statistics is not present. In England, Scotland and Wales the statistics are based on the IRS while Northern Ireland has its own system.
Fire statistics in the IRS is very detailed with clear definitions and the fields available cover the major areas of investigations.
No evaluation of the economic impact of fire and economic impact are available.
Part 1 of the summary table involves the analysis of the definitions available in the fire statistics. For a total of 23 countries (15 EU and 8 Other European and Non-European countries), information has been received and reclassified according to the major areas of investigation highlighted in Table 1. The tables with the definitions present in the fire statistics of the countries examined are available in Appendix I for EU countries and Appendix II for Other European and Non-European countries.
In the following sections, the evaluation of the analogies and differences in the definitions provided has been developed and the terminology obtained compared with the one provided by the ISO TS 17755-2. FIRE INCIDENTS The analysis of fire incidents is subdivided into three main parts which are accidental fire, deliberate fire and false alarm.
Then ISO TS 17755-2 presents the following definitions:
“3.2 Accidental fire: fire for which the cause does not involve an intentional human act to ignite or spread the fire into an area where the fire should not be.
3.6 Arson: act of intentionally and maliciously starting a fire or causing an explosion
3.7 Arsonist: person who commits arson
3.24 Deliberate fire, incendiary fire, intentional fire, voluntary fire: fire intentionally ignited under circumstances in which the person knows that the fire should not be ignited.
3.35 False fire alarm: alarm for which no fire occurred or for which fire department response was unnecessary”. The investigation of the definitions adopted in the various fire statistics is explained below according to EU, Other European and Non-European countries. I. EU Countries Fire incidents
No definition is available for the description of the fire incident in Austria, Bulgaria, Denmark, France, and Germany while in the Netherland definitions are available but not provided.
In Croatia and the Czech Republic, the number of fire incidents is collected; however, no definition is available.
A unique definition for fire incident is present in the following countries: Romania: Fire - self-sustaining combustion, which takes place without control in time and space, which causes loss of life and/or material damage and requires an organized intervention in order to interrupt the burning process. Slovakia: A fire is any unwanted burning in which the lives or health of individuals or animals, property or the environment are immediately endangered, which results in damage to property, the environment or which results in the injury or death of a natural person or animal.
Slovenia: An incident is an event or a group of events that are caused by uncontrolled natural or other forces, which can endanger the life or health of people and animals.
It can also cause damage of property, cultural heritage and the environment in such extend, that it is required to use special measures, forces and resources to control the incident, since regular activities, forces and resources are not sufficient. Fire is a process of rapid burning that is spreading uncontrollably in time and space. The fire characteristics are energy release together with smoke, toxic gases and flames. The consequence of rapid burning is an explosion.
Sweden: Fire or fire incident: In Swedish fire is to some extent defined by language, there are separate words for unwanted and wanted fires (like a fire for heating your house).
In addition, a definition like "uncontrolled flame, glow or smoke that caused damage" for Brand (fire). Fire incident is defined as danger that a flame, glow, smoke or heating of flammable material might cause damage.
From these definitions is clear how fire incident is usually referred to an uncontrolled burning able to endanger the health of individuals and damage property as well as the environment. Accidental fires, referred to an unintended event causing the fire incident:
Estonia: Unexpected and unintended event, which causes damage to the person’s life, property or environment. Crucial elements of accident are suddenness, involuntary and damage.
Finland: Accidental fires include those where the motive for the fire was presumed to be either accidental, negligent, or not known (or unspecified).
In the Netherlands, definitions exist but are not specified. Deliberate fire, the fire incident is caused by an intentional cause defined as deliberate:
Estonia: Intentional activity, which purpose is to make harm to another person, using fire for it.
Finland: Deliberate fires include those where the motive for the fire was ‘thought to be’ or ‘suspected to be’ deliberate. This includes fires to an individual’s own property, others’ property, or property of an unknown owner. False alarm could involve the notification of a false alarm by an individual or automatic fire alarm system as presented in the following fire statistics:
Estonia: A false alarm of an automatic fire alarm system is an alarm caused by other factors than a fire.
Finland: No fire at the scene.
Romania: False alarm - Upon arrival of the crews at the scene it is found that it does not exist.
Moreover, in Hungary, only the definition for a false alarm is provided: There is no incident at the indicated location, no event requiring the intervention of firefighters, only an event deemed to be by a caller or fire alarm device (the notifier misjudged the event). The use of detection and surveillance tools does not mean intervention, e.g. lamp, thermal imager, gas sensor. But here counted chimney fires that do not require the intervention of firefighters and the fire does not spread from the chimney to its surroundings.
Finally, fire incidents can be classified as false alarms in Italy and as fire false alarm, malicious and good intent in Ireland. However, no definition is provided. II. Other European and Non-European countries Fire incident is provided where accidental, deliberate fires and false alarms are recorded:
Canada: Any instance of destructive or uncontrolled burning, including explosion of combustible solids, liquids, or gases. Accidental, deliberate and false alarms are recorded.
USA: Incident Types: Structure fire, Fire in Mobile Property Used as Fixed Structure; Building fire; Confined fire (Cooking fire without extension beyond cooking vessel, Chimney or Flue Fire confined to chimney or flue, Incinerator overload but no flame damage outside incinerator, Fuel burner; boiler without flame damage outside firebox, Commercial compactor confined to contents, Trash or rubbish fire in structure but no damage to structure of contents). Accidental, deliberate fire included under causes and false alarm included under type of incident.
In the UK, there are definitions for primary, secondary fires and chimney fires where primary fires are generally more serious fires that harm people or cause damage to property while secondary fires are generally small outdoor fires, not involving people or property. Moreover, a late fire is defined as are fires attended by a Fire and Rescue Service which were known to be extinguished when the call was made. Accidental fire is available in:
New Zealand: Fires where the proven cause does not involve an intentional human act to ignite or spread a fire into an area where the fire should not be.
Russia: Uncontrolled burning, causing material damage, harm to the life and health of citizens, the interests of society and the state.
UK: Accidental fires include those where the motive for the fire was presumed to be either accidental or not known (or unspecified). Deliberate fire:
New Zealand: Incendiary - An unlawful, deliberately-lit fire where, given the known information the fire is likely to be a result of malicious intent or reckless disregard of others and property, to cause unlawful damage.
Russia: Deliberately or inadvertently setting fire to objects in such a way that the fire is able to spread further spontaneously after removing the ignition means.
UK: Deliberate fires include those where the motive for the fire was ‘thought to be’ or ‘suspected to be’ deliberate. This includes fires to an individual’s own property, others’ property or property of an unknown owner. Despite deliberate fire records including arson, deliberate fires are not the same as arson. Arson is defined under the Criminal Damage Act of 1971 as ‘an act of attempting to destroy or damage property, and/or in doing so, to endanger life’.
In Switzerland, there is a classification for accidental and deliberate fires, but no definition is provided. False alarms could be found:
Norway: There are two types of false alarm. "ABA Feil i Bruk" og "ABA teknisk/ukjent" alarm: An alarm which is given accidental or in good faith without fire or risk of a fire or any damage which require or could require the fire brigade. A False alarm "Falsk ABA": An Alarm which is given intentionally and in bad faith without fire or risk of a fire or any damage which requires or could require the fire brigade or where there is no other damage.
Furthermore, in Australia, Russia and the UK, false alarms are recorded in the fire statistics, but a proper definition is not available. BUILDING DESCRIPTION In the building description there are generally two philosophies which are followed:
Buildings are distinguished between residential and non-residential buildings, and
Buildings are described with regard to the national building code, often with regard to height and use of the building. In several countries a drop-down list or a list of typical buildings is given from which the property type can be chosen. The ISO TS 17755-2:2020 presents several definitions regarding buildings:
3.10 building: permanent or semi-permanent walled and roofed structure that stands alone and separately from other structures, including those under construction, or any comparable structure
Note 1 to entry: See also built environment, ISO 13943:2017, 3.32.
Note 2 to entry: When buildings are used for automatic operations, this shall be specified.
3.11 building fire: fire involving any kind of building (3.10) such as residential, commercial, public building
3.52 height of a building: distance between the floor of the ground floor used by firefighters and fire engines and the floor of the highest level used by people of the building (3.10)
Note 1 to entry: This is at least the number of floors above the ground level of the building.
3.26 dwelling fire, home fire, residential fire: fire which occurs in a property that is also a place of residence, excluding hotels, hostels, and residential institutions. I. EU Countries Residential and Non-residential buildings:
In Switzerland, residential buildings are: Building whose major part of the horizontal area is used for residential purpose.
In Croatia, Fires by fire objects (buildings types, sectors of industry, etc.) are collected.
In Denmark, the building description is defined in a "pick list" which specifies which type of building is on fire.
In Estonia, it is referred to the Building Code: A building is a construction work that has an interior space that is separated from the external environment by the roof and other parts of the building envelope, Residential buildings/House, block of flats, dormitory, auxiliary buildings, Non-residential buildings, Public buildings, industrial buildings.
In Finland, several building characteristics are reported on separated Building form which are:
Residential buildings: Detached house, attached house or block of flats, free-time residents.
Non-residential buildings: Commercial building, office building, office building, transport and communication building, institutional and healthcare building, assembly building, educational building, industrial building, warehouse, building for fire services, agricultural building or other building.
In France, for residential buildings, there are definitions for different types of buildings in French regulations for dwellings (by type and height). These are also used in Fire statistics from firefighters. For non-residential buildings, there are definitions for different types of public buildings (by type and height). These are also used in Fire statistics from firefighters.
In Ireland, fires are recorded in Domestic buildings, Institutions, Industrial, Commercial, Service and Other. The definitions are the following:
Residential buildings: Domestic buildings: Chimney Fires in Houses; Other House Fires; Apartments, flats and bedsitters; Caravans/Mobile Homes.
Non-residential buildings: Institutions: Hospitals; Schools; Other institutions, Industrial: Factories, Chemical Plants; Storage Buildings/Warehouses, Commercial: Shops/Supermarkets; Offices; Hotels/Guesthouses/Boarding Houses, etc. Service: Places of Public Entertainment (Dance Halls, Discos, Cinemas, Theatres, Bingo Halls); Public Houses; Restaurants.
In Italy, residential buildings are Residential Places and Homes: Private flats and homes; Generic building; Others; Private parking; Gypsis camps; Temporary buildings. Non-residential buildings are Places for specific uses; Storages of solid combustibles; Commercial and sales stores; Agricultural and farming places; Traffic and parking areas; Mountain areas; Other places, waste storage rooms; Switchboard room.
In the Netherlands, residential buildings are buildings where people live for at least 6 months a year (houseboat and holiday home can be included) and non-residential buildings, all building types that are not residential.
In Slovakia, the residential part of a building is the part of the building that contains flats or a flat intended for long-term housing; the residential part of the building should have a separate entrance from the public space.
In Slovenia, specific requirements for buildings are stated: Article 17 (fire safety) says: (1) In order to reduce the risk to people in or near them and the environment, facilities must ensure fire safety and enable effective and safe action by firefighters and rescuers. A sufficient amount of water for extinguishing must be provided. (2) The load-bearing structure of a building must maintain the required load-bearing capacity for a certain period of time in the event of a fire.
To limit the rapid spread of fire throughout the building, building elements must be used that are difficult to ignite, emit small amounts of heat and smoke when ignited, and limit the rapid spread of fire over the surface.
(3) In order to limit the spread of fire throughout the building, the building must be divided into fire sectors. (4) Facilities must provide a sufficient number of properly carried out evacuation routes and exits at appropriate locations so that people can leave them quickly and safely.
To ensure the rapid and safe evacuation of people and the rapid intervention of firefighters and rescuers in the facility, fire alarm and alarm systems must be installed. (5) Unobstructed and safe access for firefighting and rescue must be provided in and around facilities. (6) Appropriate fire-fighting systems and devices and equipment must be installed or installed in the facilities.
(7) The external walls and roofs of buildings, partition walls, together with doors, windows and other penetrations, must reduce the risk of the fire spreading to neighbouring buildings.
Residential buildings are defined as buildings of which at least half of the usable floor area is used for residential purposes. In case less than half of the usable floor area is used for residential purposes, the building is categorised as non-residential, depending on the purpose of the building.
In Sweden, a building is a permanent construction with a roof or roof and walls placed on ground or partly below ground or placed for a long period at a certain place in water constructed so that people can be in it. Residential and non-residential buildings are not defined. II. Other European and Non-European countries Residential and Non-residential buildings:
In Australia a code for buildings exists. Determining which major division, the fixed property falls within will assist in finding the correct subdivision. Three digits must be recorded for this code. For each of them property subtypes are available:
Public Assembly Property Division 1
Educational Property Division 2
Institutional Property Division 3
Residential Property Division 4
Shop/Store, Office Property Division 5
Primary Industry, Utility, Defence Property Division 6
Manufacturing Property Division 7
Storage Property Division 8
Special Property Division 9
Unclassified Division 0
In New Zealand, a building fire is referred to as a 'structure fire', either with or without damage. Residential building is a building where a person or persons normally live. Does not include hotels, temporary accommodation or hostels. Non-residential buildings: Nil - field breaks down into general property use. 'General Property Use' is defined as 'the board use of the location where the emergency has occurred.
In Russia, a building fire is a fire in aboveground construction with premises for living and (or) activities of people, location of production facilities, storage of products or keeping animals. Residential buildings are apartment buildings for permanent residence of people and dormitories for living during the period of work or study. Non-residential buildings are buildings which are not apartment buildings for permanent residence of people and dormitories for living during the period of work or study.
In the UK, building characteristics are reported: number of floors below and above the ground level. Residential buildings are Dwellings (residential homes and HMOs) and Other residential (hostel, B&Bs, Nursing homes, Students halls of residence, etc.) and Non-residential buildings are Offices, shops, factories, warehouses, restaurants, cinemas, public buildings, religious buildings, agricultural buildings, railway stations, sheds, etc.
In the US, a building is defined as a Property use, structure. Residential buildings include 9 code choices and non-residential buildings are defined as Assembly; Educational; Health care, detention and correction; Mercantile, business; Industrial, utility, defense, agriculture, mining; Manufacturing, processing; Storage.
In the other countries, the building is described according to terms used in the building regulations or with the implicit regard to building regulations but without a specific definition written down in the fire statistics. FIRE CAUSES Fire causes
The ISO TS 17755-2 defines the fire cause (3.16) as a predefined categorical class of the primary cause of the fire.
Most of the countries have no definitions but many of them have a dropdown menu from which it is possible to choose the fire cause.
The following countries have a definition: Italy, Romania, New Zealand, Russia.
The existing definitions are similar and can be summarized as the circumstances or conditions that cause the fire. Source of ignition
The existing definitions are similar to the ISO TS 17755-2 definition, where the source of ignition (3.74) refers to the energy that initiates the combustion.
Most of the countries have no definitions but many of them have a dropdown menu from which it is possible to choose the source of ignition, such as the heat source, material first ignited, and equipment involved in the ignition.
The following countries have a definition: Estonia, New Zealand, Russia.
The existing definitions are similar and can be summarized as the energy or source of energy that initiates ignition or combustion. Area of fire origin
The existing definitions are similar to the ISO TS 17755-2 definition, where the area of fire origin (3.5) refers to the localized area where the fire started.
Most of the countries have no definitions but many of them have a dropdown menu from which it is possible to choose the area of origins, such as kitchen, living room, others.
The following countries have a definition: Australia, New Zealand, Russia
The existing definitions are similar and can be summarized as the area/place where the fire originated. From the definitions provided by the fire statistics of the EU, Other European and Non-European countries, the following statements can be affirmed:
Drop Down menus with the limited amount of possibilities of causes, sources of ignition and areas of origins give a possibility to compare statistical data, once being unified. On the other hand, they limit the editor to given possibilities (a good asset would be to get the full drop-down menus from the countries to compare them with each other).
In the existent difference of definition/dropdown menus of fire causes and sources of ignition one can see possible different approaches of fire investigation - elimination process versus hypothesis building and verifying, even if the result will be in the most cases the same, the dropdown menu is limiting the number of possibilities (if there is not a free text box, the editor can fill in).
Differences in interpretation of the terms such as the cause of the fire and sources of ignition can be seen. Some countries report concrete possibilities of a cause of the fire, some also first fuel while others circumstances of the fire (what is also according to ISO) but perhaps difficult to compare statistically. Related to the source of ignition there is also a deviation in reporting. Some countries refer to heat transfer, some to kind of energy that leads to a fire, others also report first ignited material in this category of statistics.
FIRE CONSEQUENCES Only 5 EU countries and 7 Other European and Non-European countries provide definitions for the different generic groups dealing with fire consequences.
Those are Estonia, Finland, Netherlands, Romania, Slovenia, Australia, Canada, New Zealand, Norway, Russia, UK and USA while other countries (e.g. Hungary) record these fields and interpret them, but they do not have specific definitions. Fire spread
Fire spread or fire propagation (3.46) is defined according to ISO TS 17755-2 as the movement of fire from one place to another.
Estonia, Finland, Romania, New Zealand, UK and USA define fire spread generally as the evaluation of the extent of flame (and smoke) damage at firefighter arrival and at stop. The possible measurement is to report if the fire was confined to (spread to) room of origin, the floor of origin, building of origin, or beyond the building of origin.
While in Australia the focus is on determining the reason or the most important factor that allowed flame spread (or char) beyond the room or area of origin.
In Russia, fire spread is defined as an increase in the combustion zone and/or the likelihood of exposure to hazardous fire factors. Horizontal and vertical fire spread
According to ISO TS 17755-2, the extent of fire propagation (3.29) is defined as the horizontal and vertical dimension of fire spread.
Finland, New Zealand, and the UK measure the horizontal area damage in m2. New Zealand also measures the vertical area damage in m2. In contrast, the Netherlands does not necessarily separate horizontal and vertical fire spread but rather evaluate if and how the fire spread inside/outside the building or to another floor, which can be a combination of fire horizontal and vertical fire spread. Damage
For the definitions available in the ISO TS 17755-2:
"3.23 Damage: total loss caused by fire, including direct property damages (3.25) and indirect losses (3.54) such as business interruption, loss of future production and including loss of wildlife or watershed values in wildland fires.
3.25 Property damage: property damages: estimated monetary value of the damage to property and contents caused by fire and firefighting operations, including costs for demolition and decontamination as well as indirect losses (3.54) due to business interruption"
Slovenia, Canada, Russia, and ISO TS 17755-2 introduce different definitions for damage that are not specifically contradictory, but with different level of precisions.
Slovenia: Damage that is caused by natural or other accident.
Canada: Extent of Damage is the total extent of damage caused by actual burning or charring and includes damage caused by heat (browning, blistering, etc.), smoke, water and other extinguishing agents.
Russia: Direct material damage from a fire is understood as material values estimated in monetary terms, destroyed and (or) damaged as a result of exposure to hazardous fire factors and their associated manifestations.
Finally, Estonia has a mathematical equation to calculate the property damage as a function of different parameters such as building square metre value, burnt area, depreciation percentage, sanitary repairs, renovation, capital repairs and warranty repair.
Finland, New Zealand, and Romania estimate the damage caused by fire depending on the area affected by fire in m2. Fire
The definition in ISO TS 17755-2 is specific for reported fire (3.39):
“<fire statistics> fire that receives a fire department response regardless of loss, without any exception.
Note 1 to entry: In some countries, a fire can be reported to a fire station. In this case, it is specifically noted as such.
Note 2 to entry: Reported fire does not include the following, except where they cause fire or occur as a consequence of fire: explosion, lightning and electrical discharge.
Note 3 to entry: This definition differs from general definitions of fire given in ISO 13943:2017, 3.114, 3.115 and 3.116."
Estonia defines fire as “Combustion process, which is characterized by heat release, smoke and/or flames”.
In Russia it is “Uncontrolled burning, causing material damage, harm to the life and health of citizens, the interests of society and the state”.
In the UK, the definition is “The total horizontal area damaged by the flame and/or heat (in square metres) at the stop of the fire”.
While the first two definitions focus on the fire process and are complementary, the latter definition focuses on the fire damage.
All definitions are technically correct; however, they depend on the context and how they are expected to be used.
For this term, some fire statistics provide definitions for the fire incident while others for the fire damage caused by the incident. Flame
Estonia defines flame as a “combustion zone in the vapour phase which emits light” and Russia defines it as “Combustion process accompanied by flame or glow”.
Australia and New Zealand define flame damage in a similar way (but different wordings) covering “the extent of the area burned or charred by flame impingement”.
The UK only records if there were any heat and smoke damage, while the USA evaluates the number of stories where there is minor damage, significant damage, heavy damage and extreme damage.
Finally, Finland records fire damage in square meters and Euros. Smoke
The ISO TS 17755-2 does not include a definition just for smoke but for the extent of smoke propagation (3.30): horizontal and vertical dimension of smoke spread.
Estonia and Russia have similar definitions for smoke which covers “a visible suspension of solid and/or liquid particles (or aerosol) in gas formed by combustion or pyrolysis of materials”.
The definition in Australia is also similar but is worded differently: “The extent of the smoke and heat scorch or browned damage to the structure.” Overall, they don’t seem to contradict.
New Zealand records the approximate floor area of the structure that was affected by smoke. That is also the practice in Finland, in addition to the estimated cost of damage in Euros.
The UK only records if there were heat and smoke damage or not. Water
Finland records the area of damage by water in square meters and its cost in Euros, while New Zealand only records the approximate floor area of the structure that was affected by water. Similarly, Australia evaluates the extent of the damage to the building and contents caused by water or other extinguishing agents. However, there are no specific definitions. Total
Finland accounts for total damage in square meters and Euros, while New Zealand only considers all damage measured in square metres. In the UK, it is the total horizontal area damaged by the flame, heat, smoke and/or water (in square metres) at the stop of the fire. Finally, in the USA, the Estimated Dollar Losses cover Property and Contents. FIRE SAFETY MEASURES Members of the project team sought to assess whether countries included in the study collected data on a variety of fire safety measures that may have been present in fire incidents involving buildings.
Information was sought on eleven safety measures: Alarm, Type of Alarm, Automatic Extinguishing Systems, Type of Automatic Extinguishing Systems, Compartmentation, Fire Barriers, Safe Areas, Smoke Extractors, Fire Brigades on Site, Escape Routes, and Evacuation.
The project team sought to identify whether the various national data collection systems provided definitions of these safety measures, the type of information included in the definitions, and how differences and similarities between definitions in the respective data collections systems. “Alarm” is the only safety measure for which a definition is available in ISO TS 17755-2, limiting comparisons with an external standard:
For instance, it isn’t clear whether escape routes, safe areas, evacuation referred to building features that were specifically designed with fire safety in mind or, alternatively, whether occupants were simply able to find passages to safety, to shelter in safe areas, or to evacuate, and whether compartmentation referred to fire spread or to fire safety structural features.
There was also likely to be confusion whether fire brigades on site referred to whether or how soon fire departments reached the scene of a fire or to fire departments permanently stationed at the location of a fire, such as industrial fire brigades. The results of this research task are summarized below for three country categories: member states of the European Union, European nations that are not members of the European Union named Other European, and Non-European countries. I.
EU Countries Seventeen members of the European Union were included in the assessment of fire safety measures.
Our research found that the fire data collection systems in seven of these countries included a definition of at least two of the safety measures: Austria, Denmark, Estonia, Finland, Germany, Hungary, and Slovakia. However, no definitions were provided for Germany or Austria.
Definitions were classified as not available for four countries: France, Ireland, Italy, and the Netherlands. Definitions were not specified or unclear in six countries: Bulgaria, Croatia, the Czech Republic, Romania, Slovenia, and Sweden. Consequently, the review of fire safety measures in the European Union is limited to five countries for which at least some of the definitions for fire safety measures are available: Denmark, Estonia, Finland, Hungary, and Slovakia. Alarm:
Four countries (Denmark, Finland, Hungary, Slovakia) were found to capture information on “Alarm” in their data collection systems.
The information in each case referenced the physical presence of a fire alarm in a building, which differs from the ISO TS 17755-2 definition of alarm as “time to notification to fire service or other local service.” Type of Alarm:
Four countries (Denmark, Estonia, Finland, and Hungary) captured information on the type of alarm, but categories differed between them, such as “ion detector, optimal smoke detector, thermodetector, multifunction detector” vs. “autonomous fire alarm sensor, autonomous fire alarm system, automatic fire alarm.” Automatic Extinguishing System:
Information on the presence of automatic extinguishing systems is captured in four countries (Estonia, Finland, Hungary, Slovakia). Type of Automatic Extinguishing System:
Finland and Hungary capture information on the type of automatic extinguishing system.
Finland captures “type of sprinkler,” while Hungary captures whether the system is “water mist, gaseous, pressurized, drencher, foam, powder, or other.” Compartmentation:
Three data collection systems (Estonia, Finland, Hungary) capture compartmentation as a structural element of a building.
Information on compartmentation for a fourth (Denmark) references fire spread from the main space and the reason for the spread, such as open doors. Fire Barriers:
Two countries (Estonia and Slovakia) collect data on fire barriers and reference the barrier as a structural feature designed to stop the spread of fire.
Information on fire barriers in Finland is said to be included in compartmentation, but no additional information is provided. Safe Areas:
Safe area in Hungary is identified as a designated room or space in a building that is designed to ensure safety until rescue or escape.
Estonia identifies the safe area as a room where people are evacuated to but doesn’t indicate that this is the room designed as a safe area. Smoke Extractors:
Four countries (Estonia, Finland, Hungary, Slovakia) collect information on smoke extractors.
Finland and Slovakia distinguish between automatic and manual extraction systems.
Estonia references smoke extractors as a permanently installed building system. Fire Brigades on Site:
Two countries (Finland and Hungary) include information on fire brigades on-site, but reference different phenomena: Finland reports arrival time of fire brigades, while Hungary references fire brigades that are maintained by the facility of fire occurrence. Escape Routes:
Estonia, Finland, Hungary and Slovakia identify escape routes as a route used to facilitate evacuation. Two indicate or imply that the escape routes are a building design feature. Evacuation:
Estonia, Finland, Hungary, and Slovakia identify evacuation as the departure or removal of building occupants to safety. II. Other European and Non-European Countries Information was collected on fire safety measures for four European nations that are not member states of the European Union (Norway, Russia, UK, and Switzerland). Although the research indicated that the fire data collection systems for Switzerland included definitions on all of the fire safety measures within the scope of review, no definitions of those measures were identified.
Information on fire safety measures is therefore limited to Norway, Russia, and the United Kingdom. Information was also collected on fire safety measures for four non-European countries (Australia, Canada, New Zealand and the United States). Alarm:
All Other European and four Non-European countries indicate that information on alarm presence is captured in data collection systems. Type of Alarm:
Norway and the UK capture information on the type of alarm but utilize different categories.
Detector type in Norway is based upon the type of detection (ion detector, optical smoke detector, thermodetector, multifunction detector)
Categories in the UK are based upon alarm power source (Smoke alarm – 1-year battery; Smoke alarm - long-life battery; Smoke alarm – mains; Smoke alarm - mains and battery; Smoke alarm - Battery type not known; Mains security system including smoke alarm; Other; Not known).
Canada, New Zealand, and the USA capture information on the type of alarm.
Canada captures whether the alarm is: Single- or two-stage central alarm, whether the alarm is a voice alarm, and whether the alarm is connected to a remote monitoring agency single-stage central alarm with connection to the remote monitoring agency.
The USA identifies whether the alarm is a smoke alarm, heat alarm, combination smoke and heat alarm, sprinkler water flow detection alarm, or involves more than one type of alarm. Automatic Extinguishing System:
Russia and the UK capture information on the presence of automatic extinguishing systems.
All four Non-European countries have information on the presence of automatic extinguishing systems. Type of Automatic Extinguishing System:
Information is collected only by the UK, based on categories of Water mist; Gaseous system - halon; Gaseous system - other; Pressurization; Smoke ventilation; Drencher; Foam; Powder; Other.
The USA identifies whether the system is wet pipe, dry pipe, dry chemical, foam, halogen-type, carbon dioxide or other.
New Zealand identifies different types of system, but options are not available. Compartmentation:
Information on compartmentation was included for Norway and the UK.
The information for Norway refers to whether fire spread from the area of origin and whether spread was due to open doors, holes, etc.
The UK more directly references compartmentation as a building design feature, where the size of compartment is recorded and compartment categories include: Stopped/checked spread; Breached current building work; Breached - previous building work; Breached - fire doors left open or incorrectly fitted; Damage to compartmentation; Fire spread through gaps or voids in construction; No compartmentation in the building; Not applicable; Other.
No information on compartmentation as a building design safety feature in Non-European countries.
Australia collects information on the smallest compartment within which the fire is contained.
The USA collects information on the extent of fire spread. Fire Barriers:
Information on fire barriers for the UK is included in information on compartmentation.
The data collection system in Russia describes fire barrier as: A building structure with a standardized fire resistance limit and a structural fire hazard class of a structure, a volumetric element of a building or other engineering solution designed to prevent the spread of fire from one part of a building, structure, structure to another or between buildings, structures, structures, green spaces.
No information on the presence of fire barriers in Non-European countries. Safe Areas:
The data collection system in Russia is said to include safe areas as “An area where people are protected from the effects of fire hazards or where there are no fire hazards”.
The response for Norway indicates that safe area involves “Fire technical equipment and functionality (smoke alarm, fire blanket, fire technical installations”.
No information on smoke extractors as a feature of the building in Non-European countries.
New Zealand defines a “place of safety” as a place in the vicinity of a building from which people may safely disperse after escaping the effects of a fire. It may be a place such as a street, open space, public space or an adjacent building. Smoke Extractors:
Information on smoke extractors is only included for Russia, with this definition: The smoke exhaust system is a specialized complex of ventilation equipment designed for the prompt removal of combustion products from the premises, for removing smoke from the evacuation routes of people and contributing to the correct organization of measures to eliminate the fire.
No information on smoke extractors as a feature of the building for Non-European countries.
New Zealand records equipment adopted by fire brigades for the extraction of smoke. Fire Brigades on Site: No information. Escape Routes:
Russia defines an escape route as: An exit leading to the escape route directly outside or into a safe area.
The UK includes escape route information in several response options: Okay – no visible concerns; Exits locked; Exits blocked (e.g. Materials stored blocking exit); Exit route blocked by smoke/flames; Poor implementation e.g. doors swing the wrong way; Contents contributing to abnormal fire spread/smoke production; Not applicable; Other.
New Zealand records only if occupants could or could not escape safely. Evacuation:
Russia defines evacuation as: The process of organized independent movement of people directly outside or into a safe area from premises where there is a possibility of exposure of people to dangerous fire factors.
The UK includes evacuation information as: Yes (with data on people evacuated with or without assistance) and No Escape Routes.
New Zealand asks for evacuation status and whether the location of the emergency was fully evacuated during the emergency.
Fire Safety, Evacuation Procedures and Evacuation Schemes Regulations 2018 require buildings to have a 'means of escape from fire' either by "The owner of a building must have a procedure in place (evacuation procedure) for the safe, prompt, and efficient evacuation of the building’s occupants in the event of a fire emergency requiring evacuation, or an evacuation scheme to enable the safe, prompt and efficient evacuation of the building's occupants in the event of a fire emergency evacuation.
FIRE RESPONSE Fire response constitutes of two different entries, “Fire service time of response” and “Occupant fire response”.
These terms are part of ISO TS 17755-2 only as “response time (3.68) – time from the time of call to the arrival of the first fire engine”. Fire service time of response:
The examined fire statistics show that in many countries only a definition for “Fire service time of response” is provided. The fire service response time is also possible to be calculated in many cases based on different entries.
In Denmark and Norway, this can be constituted of several time entries like alarm time and date, departure, arrival at scene, leaving scene and back at the station.
In Hungary, the country has provided the requirements that the country has on the rescue service for response time (time between alert and the first vehicle starts moving) or what affects the response time. Estonia, Netherlands, Romania, Slovenia, Australia, Canada, New Zealand, Russia and UK have provided a similar definition as given by ISO TS 17755-2. Occupant fire response:
The term “Occupant fire response” is more ambiguous and few countries have provided a definition for this in the tables. In Finland, it is unclear if it, in fact, relates to the Fire service time of response.
Hungary explains the regulations on who should report a fire.
The Netherlands defines “Occupant fire response” as if someone tries to put out the fire.
New Zealand does not provide a definition but have a selection drop list on what equipment had been used to control the fire before the fire brigade arrived.
The UK has defined it as the time between ignition and discovery and between discovery and the call recorded. By studying the tables obtained in Part 2 of the summary table in the fire statistics, it is possible to gain further insight into what data different countries collect that can relate to “Occupant fire response”, this is in many cases related to what actions the occupants take to control the fire, but as the term is written now it is unclear what is meant by it.
FIRE FINANCIAL COSTS The direct financial costs related to the fire incidents are defined only in 4 EU countries, while the indirect financial costs only in Slovenia. When Other EU and Non-EU countries are considered, 4 countries determined the direct financial costs and only Russia the indirect financial costs.
In the ISO TS 17755-2, there are various definitions related to the direct and indirect fire costs:
“3.23 Damages: total loss caused by fire, including direct property damages (3.25) and indirect losses (3.54) such as business interruption, loss of future production and including loss of wildlife or watershed values in wildland fires
3.25 Direct property damages: damages (3.23) excluding indirect losses (3.54) Note 1 to entry: See also, damages (3.23) and indirect losses (3.54).
3.66 Property damages: estimated monetary value of the damage to property and contents caused by fire and firefighting operations, including costs for demolition and decontamination as well as indirect losses (3.54) due to business interruption Note 1 to entry: Property damages does not include land value. It can include indirect loss due to business interruption.
3.54 Indirect losses: amount of loss incurred as a result of being unable to use business property or equipment”. I. EU Countries The direct financial costs are evaluated as the amount of damage to the building caused by the fire in monetary terms:
In Estonia, damage to furniture or the environment is not taken into account.
In Finland, they are estimated by fire, smoke and water.
In Slovakia, the direct material loss is the sum of losses calculated from estimated residual value of tangible fixed assets, materials and other values destroyed by the fire.
In Slovenia, physical or legal person who intentionally caused an accident or because of negligence, the incident induced costs due to the emergency is required to cover the following: costs of the rescue intervention, costs of restoration to the previous condition, costs of compensation for physical and legal people. Funds for assessing the damage in the event of a natural or other incident is provided by the government The indirect financial costs:
In Slovenia, they are related to the cost of the firefighting operations. In particular, the owner or manager of the facility is obliged to cover the costs of performing firefighting operations. The costs of the intervention that arise due to the tasks performed by the fire brigade are covered by the municipality. Irrespective of the previous two sentences, the costs of the intervention are covered by: 1. The person responsible for the accident that was caused intentionally or due to negligence 2. Whoever is not taking precautions when transporting, storing or carrying out other tasks with toxic substances 3.
Whoever is not organising that the firefighters to be present at an event or activity according to the regulations 4. Whoever is on purpose contacting the fire unit without reason.
The costs of interventions that arise due to the interventions performed outside of the municipality of the fire unit are covered by the Republic of Slovenia if the intervention was performed on the basis of national protection. II. Other European and Non-European countries The direct financial costs in Other European and Non-European countries:
The fire direct financial costs are subdivided into dollar loss, property, contents and insurance in Australia:
Dollar loss: the estimated monetary value of the damage to property and contents caused by fire and firefighting operations. Do not include land value.
Property: the reporting officer's estimation of the value of the property. Do not include the value of the contents (property includes buildings, structures and mobile property).
Contents: the reporting officer's estimation of the value of the contents (includes crops).
Insurance: whether the contents or structure or both were insured.
In Canada, the direct financial costs are estimated in dollar loss but no methodology is provided.
In Russia, direct material damage from a fire is understood as material values estimated in monetary terms, destroyed and (or) damaged as a result of exposure to hazardous fire factors and their associated manifestations.
In USA, estimate of total property and contents dollar loss and pre-incident value of the property and contents are provided. For the indirect financial costs of fire:
In Russia, material losses due to violation of economic plans in the economy (e.g. a decline in production, a decline in trade and banking operations, a decrease in income, losses due to delays in the transport of goods). In the following Section, the analysis of fatalities and casualties, will be provided based on the definitions available in the ISO TS 17755-2 and the description of the current terminology adopted in EU, Other European and Non-European countries. FATALITIES AND CASUALTIES #### 3.8.1.
Description of the data status of ISO TS 17755-2 With regard to the current EU FireStat project, the ISO TS 17755-2 standard has been examined. Table 3 provides an overview of common terms and definitions.
| Code | Term | Definition |
|---|---|---|
| 3.03 | Age group of victims | Categorization by age of the victims (3.15) of fire. Note 1: this categorization may differ locally. Proposed categories: Newborn (child under 28 days); Child (28 days–9 years); Youth (10–17 years); Adult (18–64 years); Elderly (65 or more). |
| 3.15 | Casualty, Victim | Person killed or injured. |
| 3.18 | Cause of casualty | Phenomenon causing death of a person in a fire: smoke inhalation; burn; physical injury; other. |
| 3.21 | Condition of casualty | Predefined categorical classes of the circumstances of casualty (e.g. asleep at time of fire; impairment by alcohol or drugs; awake with no impairment). |
| 3.36 | Fatal fire | Fire with at least one fire fatality (3.37). See also multi fatal fire (3.62). |
| 3.37 | Fatal fire casualty / Fire fatality / Fire death | Person who has died as a result of injuries sustained during a fire incident (includes deaths during fire control, rescue, or escape; no time limit after the fire in this context). |
| 3.40 | Fire casualty / Fire victim | Person killed or injured as a direct effect of a fire without any limit of time following the date on which the injury was sustained (some countries use time limits). |
| 3.42 | Fire injury | Person injured (not fatally) as a result of a fire incident; includes injuries during fire control, rescue, or escape requiring first aid or further medical treatment. |
| 3.57 | Light injury / Moderate injury | Hospitalized 1–3 days or requires 1–3 weeks off work. |
| 3.61 | Minor injury | Hospitalized or off work for less than 1 day. |
| 3.63 | Nature / Severity of casualty | Stage of fire victim (3.40): e.g. minor, light, serious injury, life threatening, death. |
| 3.73 | Serious injury | Hospitalized 4+ days or more than 3 weeks off work. |
| 3.78 | Status of a victim | Role during a fire: civilian; firefighter; other rescuer; unknown. |
| 3.84 | Victim characteristics | Information collected about victims (3.15). |
The terms and definitions proposed by the ISO TS 17755-2 are inherently more than contradicting:
Thus, “fire casualty” logically means “fire injury” (person rescued from the fire) (see 3.42).
However, the words “casualty” and “injury” are misleading, especially since the definition of “...person who has died… during a fire incident” is used. Some inconsistencies are present and it would be more appropriate to adopt “fatal fire death”.
If a person is injured (i.e. saved alive, in a fire), this is to be classified as a "fire injury". If the person dies from the injuries, then this person is to be classified as "fire death". The term “fatal fire injury” is misleading because it is difficult to be associate with “non-fatal fire injury”. Regardless of this, the term “not fatally injured” (3.42) is used - also unclear.
No clear separation of the terms is described here: see 3.15. If one speaks of “fire casualty” (3.40) or “casualty” (3.15), and of “fire victim” (3.40) or “victim” (3.15) - nothing changes in terms of content: “victim” remains the generic term for “dead and injured”.
Thus, all other people involved are to be categorized either as “fire deaths” or simply as “affected by fire”.
The definitions for fatalities and casualties of the EU, Other European and Non-European countries were evaluated, where available.
Fatalities/Deaths
According to ISO TS 17755-2, a fire victim is a person killed or injured as a direct effect of a fire without any limit of time following the date on which the injury was sustained. Hungary has the same definition. In some countries, limits of time are used. For example, Finland, Denmark, Norway and Estonia use a limit of time of 30 days for the death after the fire. In Slovakia, it is 24 hours after the incident and in Sweden within 90 days of the fire.
Italy does have a limit of time, namely “the time of the intervention”.
Although these countries do not have a definition, the number of victims is collected in most of them.
In Canada, the classification of a fire death is: A person killed accidentally as a direct result of a fire or a person who dies from a fire injury within one year following the date on which the injury was sustained.
Type of fatality
Most of the countries do not have a definition for type of fatality or the term is not specified.
According to the tables provided in Appendix I and Appendix II, information on what an injured person is, and the types of injury could be obtained in Estonia, Finland, France, Hungary, Italy, Romania, Slovenia, Sweden, Australia, Canada, New Zealand, Russia, UK and USA.
To be recorded, the injury must be severe enough to require treatment by a medical practitioner, regardless of whether treatment is actually received, or the injury must result in at least one day of restricted activity immediately following the incident. A death is recorded if it is attributable to the incident or the action of handling the incident.
It can be stated that:
Regardless of this, it has been found that most countries still have definitions for the term “casualties”. The terms can be found in the internal service instructions of the responsible ministries but sometimes not publicly available.
Conclusions and recommendation:
The fields recorded in the fire statistics of the countries examined have been investigated in Part 2 of the summary table, as described in Section 1.2. A total of 21 EU and 8 Other European and Non-European countries have kindly provided information about their fire statistics. Due to the limited or no information received, Greece, Lithuania, Luxembourg, Malta, Portugal and Spain are not covered in this analysis.
The 8 Other European and Non-European countries have been selected based on their high accuracy in the fire statistics. They adopt a solid and complex systems for recording the information of the fire incidents and they have very extensive datasets.
These are the reasons why they appear to cover more fields than those recorded by EU Member States.
As described in Section 2 for each country, the data are mainly from the fire brigades, but they could belong to other sources such as insurance companies in Switzerland. Finally, as explained for Sweden, in some countries the recording system is composed of different datasets and the fields collected in the fire statistics could be covered by separated databases and not a unique one.
In Part 2 of the summary table, the fire statistics fields recorded by the relevant authorities have been classified according to the major areas of investigation described in Table 1 with a total of 98 subfields.
Each authority or organization has inserted the data available in their fire statistics and it is possible to affirm that between 0 and 49% and between 50 and 74% of the total fields examined are usually covered by the fire statistics of the EU countries, and Other European and Non-European countries, respectively (Figure 1). Therefore, despite the difference in the number of datasets evaluated, it appears that in Other European and Non-European countries more detailed fire statistics are available.
Figure 1: Data recorded in the fire statistics considering the total fields examined in Part 2 of the
summary table for EU, Other European and Non-European countries. In Figure 2, the fields related to the description of the fire incidents are introduced. The incident time, date and location are recorded for almost the total number of countries investigated despite their geographical location.
Furthermore, the false alarms are recorded in 14 EU and 4 Other European and Non-European countries while the distinction between deliberate and accidental fires is less frequent (recorded in more than 9 and 5 EU and Other European and Non-European countries, respectively).
Figure 2: Fields recorded for the description of fire incidents in EU, Other European and Non-
European countries. 0-49% 50-74% 75-100% Number of countries EU Other European and Non-European 14 15 6 5 7 8 8 Fire False alarm Deliberate Accidental Call Incident Incident time Incident date Incident location Number of countries EU Other European and Non-European For the description of the building involved in the fire incident, the classification of the building into residential and non-residential properties is usually available in 17 EU and 7 Other European and Non-European countries.
Moreover, in several fire statistics, the building type is usually further classified in various subproperty types.
The field related to the construction types is often available in Other European and Non- European countries (6) while it is present in only 4 EU countries.
The number of floors, origin floor and total building dimension are recorded for a maximum of 8 EU countries and almost the total Other European and Non-European countries investigated (Figure 3).
Figure 3: Fields recorded for the building description in EU, Other European and Non-European
countries. When the pre-fire conditions are analysed, fire causes are available in the fire statistics of 15 EU and 5 Other European and Non-European countries. For the factors contributing to ignition, source of ignition (9 & 6), fire room of origin (9 & 6), item first ignited (10 & 6), material first ignited (6 & 6) and material mainly responsible for the fire development (5 & 5) are available in the EU and Other European and Non-European countries as shown in
Figure 4. Figure 4: Fields recorded for the fire causes in EU, Other European and Non-European countries.
When the post-fire conditions are examined focused on the effects of the fire incident on the building, the analysis has been divided into the evaluation of fire spread, investigation and quantification of damage. According to Figure 5, the field related to the fire spread is usually present in 7 fire statistics of EU countries where 6 of them provide also a distinction between horizontal and vertical fire spread.
In Other European and Non-European countries, fire spread is generally referred to as horizontal spread even if the vertical spread is recorded in 6 countries. 18 17 7 8 8 7 6 5 6 Residential Non-residential Construction type Origin floor dimension Number of floor Total building dimension Number of countries EU Other European and Non-European 9 9 10 6 55 6 6 6 6 5 Fire cause Source of ignition Fire room of origin Item first ignited Material first ignited Material mainly responsible for the fire development Number of countries EU Other European and Non-European
Figure 5: Fields recorded for the fire consequences focused on the fire spread in EU, Other European
and Non-European countries. The distinction between fire and total damage appears in 7 EU countries and 3 Other European and Non- European countries while damage is usually linked only to fire damage. This is also supported by the damage types recorded which are generally only related to fire damage while in 2 Other European and Non-European countries, the subdivision of damage in fire, flame, smoke and water damage is available. Furthermore, in
Figure 6, it appears that fire damage is quantified based on m2 and the percentage of property damage in 5 &
3 EU countries while both quantifications in 3 Other European and Non-European countries.
Figure 6: Fields recorded for the fire consequences focused on the evaluation of damage in EU,
Other European and Non-European countries. The evaluation of the fire safety measures is an essential aspect that should be covered in the fire incident investigation. The analysis developed by this research has subdivided the fire safety measures into alarms, automatic extinguishing systems (Figure 7) and other safety measures such as compartmentation, smoke extractors and others (Figure 8).
Usually, the analysis of the presence and operation of alarms are recorded more often (11 & 9) than the type (7), effectiveness and failure (6) in EU countries.
For the automatic extinguishing systems, the presence in a building is the field generally recorded (8) followed by operation and effectiveness (6), type (5) and failure (4) in EU countries.
If the same analysis is repeated in Other European and Non-European countries, the fields recorded in a decreasing number of countries are operation and failure (8), type (7), presence (6) and effectiveness (5) for alarms while presence and type (7), operation and effectiveness (6), and failure (5) for automatic extinguishing systems.
As described in Figure 7, almost the majority of Other European and Non-European countries provide detailed information about the alarms and automatic extinguishing systems. 7 6 6 Rapid growth Fire spread Fire spread horizontally Fire spread vertically Number of countries EU Other European and Non-European 8 7 0 1 3 4 3 2 3 3 2 3 3 Fire damage Total damage Fire damage Flame damage Smoke damage Water damage Others m² % of property damaged Others Damage Type of damage Damage quantifcation Number of countries EU Other European and Non-European
Figure 7: Fields recorded for fire safety measures focused on alarms and automatic extinguishing
systems in EU, Other European and Non-European countries. When other fire safety measures are investigated in Figure 8, the evacuation (8 & 4), the presence of fire brigade on site (7 & 4), smoke extractors (6 & 2) and compartmentation (5 & 3) are the fields highly recorded in EU, and Other European and Non-European countries, respectively.
Fire barriers (4 & 1) and escape routes (3 & 2) are the fields recorded less often in EU, and Other European and Non-European countries while the field of safe areas is available only in 2 EU countries.
Figure 8: Fields recorded for other fire safety measures in EU, Other European and Non-European
countries. The response time to the fire incident has been classified according to the response of occupants and firefighters as illustrated in
Figure 9. The response time of occupants is divided into the time between ignition to discovery and between discovery to call and these two time intervals are available in the fire statistics of 4 and 7 EU countries and both time intervals in 2 Other European and Non-European countries. The response time of the fire brigades is referred to the time between the notification of the fire incident and the arrival of the fire brigade at the fire scene.
Moreover, the response time of the fire brigades is composed of 5 steps: notification, dispatch, preparation, travel time and set up time.
Based on the information received, between 8 and 12 EU countries and between 1 and 8 Other European and Non-European countries mainly collect the above-mentioned time steps where the notification time is the field usually available in the fire statistics.
The last two time intervals examined related to the response time are the occupant rescue and fire extinguishment that are present only in 4 and 5 EU countries and 1 and 2 Other European and Non-European countries, respectively. 6 6 5 6 6 6 7 8 8 7 7 6 6 5 Presence Type Operation Effectiveness Failure Presence Type Operation Effectiveness Failure Alarms Automatic extinguishing systems Number of countries EU Other European and Non-European 5 4 6 7 1 0 Compartmentation Fire barriers Safe areas Smoke extractors Fire brigades on site Escape ruotes Evacuation Number of countries EU Other European and Non-European
Figure 9: Fields recorded for the fire response time of occupants and fire brigades in EU, Other
European and Non-European countries. Two important areas of investigation highlighted in Table 1 are those of fatalities and casualties related to the life safety aspects derived by the fire incident. They have been investigated separately considering not only occupants but also firefighters.
Starting with the analysis of fatalities of Figure 10, the number of victims (20), age (14) and gender (10) are the highest recorded fields in EU countries followed by cause of fatality (8), type of fatality (6) and disability (7).
In Other European and Non-European countries, all the fields investigated for the description of fatalities are available for a minimum of 6 countries with the only exception of disability (4) and profession (2). The time between the incident and the fatality is available only in 6 EU and 3 Other European and Non-European countries.
If the victims were occupants or firefighters is determined in 10 and 5 EU and Other European and Non-European countries, respectively.
Figure 10: Fields recorded for fatalities in EU, Other European and Non-European countries.
The same analysis applied to casualties provides the outcome that casualties are recorded less often than fatalities. In particular, the fields available with a decreasing number of countries are the number of injured people (15), age (11), type of injury (8), gender (7) and cause of injury (5) in EU countries while the number of injured people (7), cause of injury, age, gender and disability (6) and type of injury (5) in Other European and Non-European countries.
The time between the incident and the injury is recorded only in 2 EU and Other European and Non-European countries.
The distinction if the injured people were occupants or firefighters is present in the fire statistics of 7 EU and 5 Other European and Non-European countries (Figure 11). 8 8 9 9 4 5 2 2 3 2 1 1 2 Ignition to discovery Discovery to call Notification Dispatch Preparation Travel time Set up time Occupant rescue Fire extinguishment Number of countries EU Other European and Non-European 3 4 7 6 6 6 6 6 Number of victims Cause of fatality Type of fatality Age Gender Ethnicity Disability Profession Other circumstances Time between incident and fatality Only occupants or firefighters?
Number of countries EU Other European and Non-European
Figure 11: Fields recorded for casualties in EU, Other European and Non-European countries.
Finally, the evaluation of the financial costs derived by the fire incidents is usually classified according to direct and indirect financial losses. As described in Figure 12, the methodology for the evaluation of direct costs caused by fire incidents is usually collected in the fire statistics of 5 EU and 4 Other European and Non- European countries. In particular, the direct costs are evaluated considering mainly property (9 & 5) and medical care (1 & 3) in EU, and Other European and Non-European countries.
For the indirect costs, temporal shelter and medical care are available in 1 EU (Denmark) and 1 Other European and Non-European countries (Norway).
Figure 12: Fields recorded for the fire financial costs in EU, Other European and Non-European
countries. Based on the major areas of investigation, several subfields have been determined considering fire incident description, building characteristics, life safety aspects and fire financial costs collected in the fire statistics of the countries examined.
Despite the difference in the number of countries investigated and geographical location (20 EU and 8 Other European and Non-European countries), the findings obtained show that the fire statistics of Other European and Non-European countries appear more detailed especially for what concerns the quantification of damage, the presence, operation and effectiveness of the fire safety measures and description of fatalities and casualties and, indeed, the fire statistics of Australia, Canada, New Zealand, Russia, UK and USA cover also the totality of the fields investigated. 4 4 4 77 6 5 6 6 6 Number of injured people Cause of injury Type of injury Age Gender Disability Profession Other circumstances Time between incident and injury Only occupants or firefighters?
Number of countries EU Other European and Non-European 1 1 1 1 0 1 1 1 0 1 1 Property Medical care Temporary shelter Medical care Direct costs Indirect costs Fire direct costs Indirect costs Methodology Insured and uninsured losses Cost incurred to insurance companies Number of countries EU Other European and Non-European
Task 1 of the EU FireStat project had the aim to provide a general overview on the fire statistics of the 27 EU and 8 Other European and Non-European countries. The 8 Other European and Non-European countries have been chosen based on their structured and complex fire statistics and accuracy in the data elaboration.
In Task 1, summary tables for the countries investigated have been created and subdivided into two parts: Part 1 related to the definitions and Part 2 related to the fields covered by each fire statistics. The summary table usually includes 10 major areas of investigation determining pre- and post-fire conditions in buildings subjected to fire incidents.
Considering the 35 countries examined, not all of them were able to provide a list of definitions and the fields recorded.
Therefore, the differences in the number of countries considered for each analysis developed have to be based on the methodology and current practice available in current fire statistics. In Section 2, short abstracts for the countries examined are presented with a clear description of the authority responsible for the fire statistics, the collection methodology, the number of datasets, how data are managed and explanations on the summary tables.
The research developed has been particularly focused on the available definitions to understand the meanings hidden behind the terminology used by each fire statistics.
It is essential to have a clear understanding of the fields considered to be able to appropriately compare the data from various countries.
The tables with the list of definitions are presented in Appendix I for EU countries and Appendix II for Other European and Non- European countries. Moreover, a semantic analysis of the definitions has been developed for the examined countries and compared to those present in the ISO TS 17755-2 in Section 3. Such analysis has highlighted differences and analogies in the terms adopted by the various fire statistics.
Finally, the fields collected have been investigated to understand the pre- and post-fire variables available in the fire statistics as described in Section 4 for EU and Other European and Non-European countries.
In the following Sections, specific and general conclusions of the work developed by Task 1 are provided.
Our review of national fire data collection systems found that although there was some variation between countries in data collection methodologies, administrative procedures, and information sources, these differences represented differences in national approaches rather than differences on the basis of European Union membership or between European and non-European countries. The collection of data on fire incidents is primarily the responsibility of fire services in all countries.
Data collection originates with the information provided by fire brigades on the fires to which they respond.
There are a small number of countries where public entities outside the fire service provide supplementary information to that collected by the fire service.
This may include the collection of data on deaths and injuries by health and welfare bodies or by police departments or data on building damage by insurance companies. The processing of data in most countries appears most often to be conducted by fire service authorities at the national level, although it may be performed at state or provincial levels in countries with more complex administrative systems.
In some countries with decentralized structures, processing of data is primarily a state or provincial responsibility and may or may not be routinely consolidated at the national level for further processing.
Responsibility for data processing at any level invariably resides with research units, regardless of whether they are part of emergency management bodies or broad statistical offices.
It was possible to identify fire data collection systems in which the reporting of fire incidents is mandatory such as in the Nordic countries and Estonia. However, even in well-established systems with strong participation at the local level (e.g. USA), compliance appears overwhelmingly to be based upon voluntary reporting of fire data. Within the individual national data collection systems, degrees of reporting compliance can be influenced by the support and resources available to encourage the reporting of data.
Anecdotal information also suggests that the reporting of fire data may be less reliable in rural areas.
However, there is generally little effort by fire data collection systems to assess how completely their methodologies capture the true number of fire incidents or the specific details for which information is sought.
To our knowledge, the United States is the only country which attempts to estimate and correct for fires that go unreported in the national data collection system.
The analysis of the definitions presented in Section 3 provides insights on the meanings related to each variable collected in the various fire statistics. It is fundamental to understand what is covered by each field recorded to be able to appropriately use the data and compared them with other countries. The definitions gathered has been grouped according to the major areas of investigations defined in Task 1 and compared to those available in the ISO TS 17755-2 where possible.
Specific conclusions for each area of investigation are presented below: Fire incidents:
In the ISO TS 17755-2, definitions for accidental and fire false alarm are present as well as for arson, deliberate fire, incendiary fire, intentional fire and voluntary fire.
Generally, in the EU countries, the definition of fire incident includes the various fire incident classifications while specific definitions for accidental, deliberate fire and fire false alarm are available in Estonia and Finland. In Romania and Hungary, only a definition for fire false alarm is present. No definition is provided in Italy and Ireland; however, fire incidents are recorded as false alarm in Italy and as false alarm, malicious and good intent in Ireland.
In Canada and the USA, a general definition for fire incident is provided where accidental, deliberate fires and false alarms are recorded. A specific definition for accidental and deliberate fires is available in New Zealand, Russia and the UK. False alarms are defined in Norway while in Australia, Russia and the UK, false alarms are recorded in the fire statistics, but a proper definition is not available. Building description:
The ISO TS 17755-2 presents definitions for building, building fire, height of the building, dwelling fire, home fire, residential fire.
Residential and non-residential buildings are recorded in EU, Other European and Non-European countries based on different classifications for the property types covered. Fire causes:
The ISO TS 17755-2 has definitions for fire causes, source of ignition and area of origin.
Various countries mention that definitions are available but not provided.
Several countries have no definition, but a dropdown menu is available.
Some countries report concrete possibilities of a cause of the fire, others also first fuel and some circumstances of the fire incident (as stated in the ISO TS 17755-2). However, the fields collected could be difficult to compare statistically. Related to the source of ignition, there is also a deviation in reporting. Some fire statistics refer to heat transfer while others to type of energy that leads to a fire. In a few fire statistics, the material first ignited is reported in this category. Fire consequences:
Fire consequences are collected by 5 EU countries and 7 Other European and Non-European countries. Some countries report what is collected while others provide the definitions.
Fire spread is similarly defined in Estonia, Finland, New Zealand, UK and USA and all these countries use the evaluation of fire spread “at the stop”. This expression is adopted in the UK and it is referred to the time when the fire is extinguished. In Sweden, there is an evaluation of the size of the fire at arrival and departure or extent.
In Australia, fire spread is mainly focused on the reason or factor that allowed the flame spread.
For the horizontal and vertical fire spread, some countries separate the two terms while in the Netherlands they are combined.
For the damage, there are some different definitions if the damage is caused by natural causes or accidents. The extent of damage is also evaluated in Canada while in Estonia, there is a mathematical equation to evaluate the damage.
Various fire damage definitions are available where some of them are more theoretical while others are referred to the evaluation and the same comments can be applicable to the flame damage. Fire safety measures:
In the ISO TS 17755-2, only alarms are defined, referred to as the time to the notification to fire service or other local services. This is different from any definitions provided by the countries.
Definitions are provided only in 6 EU countries and details are only available for 5 of them. For most of the EU countries, the terms are unclear or not reported. The field more consistently reported are alarm, type of alarms and information on extinguishing systems. This is also applicable to the Other European and Non-European countries.
There were probably some misunderstandings in the fields examined and inconsistencies in the interpretation of what these fire safety measures were referred to. For example, for the fire brigade on site, this was not always considered as the industrial fire brigade and not municipal fire brigade (in industrial buildings or factories there are private fire brigades on-site). Escape routes, evacuation and compartmentation, those measures are referred to fire safety design features but could also be interpreted if people were able to evacuate.
Limited information is available for fire barriers, safe areas, and smoke extractor.
Even in the USA and Canada, where the recording systems are very detailed, there are very few of these safety measures recorded. Fire response:
The fire response is composed of two parts: “occupant response” and “fire service response”.
Fire service response time is constituted of several times such as alarm time, departure, arrival, and departure time. Some countries have definitions similar to the one provided by the ISO TS 17755-2 while in others, the fire service response time can be calculated based on the different entries provided.
The term “occupant fire response” is more ambiguous. The UK has definitions for the time between ignition to discovery and discovery to call while other countries have definitions referred to the actions that the occupants take to control the fire. Fire financial costs:
In the ISO TS 17755-2, there are definitions for damage, direct losses specifying property damage and indirect losses.
Direct and indirect financial costs of fire incidents are seldom evaluated in EU, Other European and Non-European countries.
The evaluation of the direct financial costs of building fires are usually referred to the damage caused by the fire to the property and this could also include contents.
The indirect financial costs of fire, in Russia, are referred to the material losses due to violation of economic plans in the economy (e.g. a decline in production, a decline in trade and banking operations, a decrease in income, losses due to delays in the transport of goods) while in Slovenia, they are related to the cost of the firefighting operations.
It is also important to specify that there could be some data related to the costs provided by insurance and it is essential to relate the origin of the data to the source as described in the abstracts provided in Section 2. Fatalities and casualties:
A description of the definitions available in the ISO 17755-2 is provided. The terms casualty or victims are equally used and referred to a person killed or injured. This can cause misunderstandings in the following definitions. It would be suggested to adopt fatalities for deaths and casualties for injured people.
Fatalities:
Victims and type of fatality are examined. In the investigated countries, a clear distinction between fatalities and casualties is not available and the two terms are sometimes not distinguished. Moreover, in some countries, the term fatality is referred to occupants and in others to fire brigades. Most of the countries have definitions referred to the victims killed as a consequence of the fire.
Several countries establish a limit of time in the death of victim while in the ISO TS 17755- 2 “a fire victim is a person killed or injured as a direct effect of a fire without any limit of time”.
In Canada and the USA, a distinction between civilian and fire service victims is present.
The type of fatality is available only in the Netherlands, Romania and the UK.
Casualties:
While in some countries there is a clear distinction between people killed or injured in a fire addressed with two separated terms, in others, they are recorded in a unique field and defined with a unique definition. The understanding of terminology is essential, and it is therefore unfortunate that different countries have dissimilar meanings for the above-mentioned terms. It is sometimes clear how specific terms refer to different aspects.
From the evaluation of the information obtained, it is possible to confirm that the accuracy of the definitions available for specific fields of the fire statistics is supported by the high number of countries collecting those particular fields.
The highest is the number of countries in which a particular field is collected, the more accurate are the definitions available. This is also supported by the fields recorded in the various fire statistics examined in Section 4.
For example, the fire incident description is often collected, and this determines the precision of the definitions provided while when fire safety measures are considered, the few definitions available and recorded do not have consistency in their meanings in the various countries. It is, therefore, important to understand the meaning of the variables collected in the fire statistics to be able to provide adequate comparisons and correct evaluations of the aspects included.
The analysis developed in Section 4 is focused on the fields recorded in the various fire statistics examined. A total of 98 fields of investigation have been considered and subdivided according to major areas such as fire incidents, building description, fire causes, fire consequences, fire safety measures, fire financial costs and fatalities and casualties. The EU Member States that provided the list of fire statistical fields recorded have been compared to 8 Other European and Non-European countries.
It is important to affirm that the 8 Other European and Non-European countries have been chosen on the basis of their structured and detailed fire statistics which continuously undergo improvements and optimizations. Therefore, the comparisons provided for the various areas of investigation need to be interpreted considering the differences that inevitably arise.
Usually, fire incidents are described considering the incident time, date and location in a maximum of 16 and 8 while the distinction between accidental and deliberate fire is recorded in 9 and 5 EU and Other European and Non-European countries, respectively. The description of the property type subdivided into residential and non-residential buildings is available in the majority of countries examined while further building characteristics are seldom recorded.
Fire causes are available in the fire statistics of 15 EU and 5 Other European and Non- European countries and the other fields related to the source of ignition, item and material first ignited, material mainly responsible for the development of the fire and fire room of origin are determined in a maximum of 10 EU and 6 Other European and Non-European countries.
For the consequences of fire, fire spread is recorded in 7 EU countries and in the majority of Other European and Non-European countries. When fire damage is investigated, the number of EU countries decreases for what concerns how damage is recorded. In particular, it appears that a distinction between fire and total damage is available only in 7 EU countries and in 5 of them damage is recorded in m2.
In 2 Other European and Non-European countries, the fire damage is also classified according to types of damage such as fire, flame, smoke and water damage.
For the evaluation of the fire safety measures, the various fire statistics seem to record more often alarms and automatic extinguishing systems than compartmentation, smoke extractors, fire barriers, escape routes and others.
In particular, the fields related to the description of the alarms and automatic extinguishing systems mainly cover presence, operation, type and reasons for failure. Another important consideration is that the understandings of the other fire safety measures have not always been correctly evaluated and this is also supported by the description provided in Section 3.5 and by the limited number of countries which collect these fields.
The response time is usually referred as the time between the notification of the fire incidents to the fire brigade and the arrival of the fire brigade at the fire scene.
Only in 4 EU countries and 2 Other European and Non- European countries, the time between ignition to discovery and between discovery to call are available.
The fire financial costs are usually related to the direct losses of the property in 9 and 5 and to the medical care in 1 and 3 EU and, Other European and Non-European countries, respectively. The indirect financial costs are recorded for temporary shelter and medical care only in Denmark and Norway. Finally, the number of fatalities are recorded in 20 and 7 while the number of casualties in 15 and 7 EU and, Other European and Non-European countries, correspondingly.
The fields related to the number of deaths and injured people are mainly focused on the description of the person affected providing the age, gender and type of fatality or injury.
The fields recorded more often are those related to the description of the fire incidents, fatalities and casualties.
However, as already described in Section 5.2, the variables covered by these fields can be referred to as different interpretations in the various countries examined. Therefore, it is suggested to link the considerations presented for the analysis of the definitions with the elaborations of the fields recorded by the various statistics and to the information related to the collection methodologies described in the short abstract of each country.
Our review of fire data collection measures within and outside the European Union is critical for understanding the degree of commonality across the various systems and also for identifying opportunities and challenges in any efforts to create uniform measures that will facilitate comparisons in fire experience.
The ability to research variations in fire experience on the basis of comparable variables should facilitate the ability of policymakers and the fire safety community to identify opportunities for introducing improvements in fire safety practices in their own countries.
It is clear that meaningful comparison of fire incidents is influenced by similarities or differences in the way critical data elements are defined and measured.
In this review, we found considerable variation between national data collection systems both in relation to the number of data elements they included and in the way in which these elements were defined and measured. Indeed, such differences were found to extend in some cases to data collection systems by separate jurisdictions within the same country.
An additional complication identified in this review is that different institutions may serve as data providers in different countries, and different institutions may employ their own forms of measurement.
Although it was not possible to identify information on data collection measures from a number of countries, it appears that fire data collection systems in the European Union fall into different tiers with respect to the amount of information collected.
Some systems collect a limited amount of fairly basic information, such as information on the date, time, and location of the fire, type of fire (building, vehicle, etc.), type of building, fire cause, and number of deaths or injuries. Countries with more advanced data collection systems include to varying degrees a number of additional data elements.
These may determine information on the room where the fire originated, degree of fire spread, material contributing to fire spread, type of equipment involved in the fire, information on victim characteristics and involvement with the fire, types of fire safety measures, and other relevant information.
As our review indicates, there is also substantial variation in the amount and type of information sought by data elements that are common to different data collection instruments.
For instance, spread of fire may be recorded on the basis of number or rooms involved or some other physical measure, while automatic extinguishing equipment as to whether it was present or absent or involved specific types of equipment. It appears that some of the most detailed national fire data collection systems are represented by countries outside the European Union, as exemplified by New Zealand, Canada, and the United States.
Looking forward, comparability of fire data between systems will require the development and adoption of a core set of measures that have common categories or classifications.
In light of our findings, it does not seem realistic to expect that those countries with less advanced histories of fire data collection can simply adopt the more detailed data collection systems that have been built over time in countries with more extensive practices and traditions.
It may be useful as an initial step to explore the feasibility of identifying a set of major indicators with moderate levels of detail that could form a common content for a unified fire data collection instrument. Development of a limited set of common measures would not preclude individual countries with more mature data collection systems from collecting additional and more detailed information in areas of particular interest or concern.
Over time, participation in a common data collection system may encourage countries with less mature systems to enhance their data collection practices and to add data elements to the common core.
DENMARK DEFINITIONS 1/3 GENERIC GROUPS a. Yes, the definition exists b. No, the definition not available c. Term not specified or unclear
DEFINITIONS REFERENCE
FIRE INCIDENT
Accidental fire c Maybe police has a definition Deliberate fire c Maybe police has a definition False alarm a There are two types of false alarm.
"Blind" alarm: An alarm which is given accidental or in good faith without fire or risk of a fire or any damage which require or could require the fire brigade.
A False alarm: An Alarm which is given intentionally and in bad faith without fire or risk of a fire or any damage which requires or could require the fire brigade or where there are no other damage.
Beredskabsstyrel sen (DEMA)
http://www.brs.dk BUILDING DESCRIPTION Building fire a The building description is defined in a "pick list" which specifies which type of building, that is on fire Beredskabsstyrel sen (DEMA) http://www.brs.dk
Residential buildings a The building description is defined in a "pick list" which specifies which type of building, that is on fire Non-residential buildings a The building description is defined in a "pick list" which specifies which type of building, that is on fire
FIRE CAUSES
Fire causes a In the database the fire cause is described in a dropdown menu with suggestions of different fire causes Source of ignition a In the database the source of ignition is described in a dropdown menu with suggestions of different sources of ignition.
Heat source (open flame, spark, glow, lightning, etc.)
Material that was first ignited (paper, cardboard, plastic, textile, wood, etc.)
Equipment involved in the ignition (machines, combustion plants, electrical installations, etc.), including make, model and serial number Object that was subsequently ignited (building construction, furniture, storage, waste storage in container, etc.)
Beredskabsstyrel sen (DEMA)
Area of fire origin a In the database the place of origin is described in a dropdown menu with suggestions of different places.
Origin of fire (corridor area, kitchen, barn, etc.) is recorded.
Beredskabsstyrel sen (DEMA)
http://www.brs.dk FIRE CONSEQUENCE S IN TERMS OF UNIT MEASURES
Fire spread a In the database a possible fire spread is described in a dropdown menu with suggestions of different scenarios of fire spread.
Evaluation of the extent of fire at arrival a (smoke development, visible flames, etc.).
Beredskabsstyrel sen (DEMA)
Fire horizontal spread b Fire vertical spread b Damage b Fire b Flame b Smoke b Water b Total b
EU FireStat - Task 1 DENMARK DEFINITIONS 2/3 GENERIC GROUPS a. Yes, the definition exists b. No, the definition not available c. Term not specified or unclear
DEFINITIONS REFERENCE FATALITIES Victims a A fatality in a fire is defined as a person which is dead in a fire or within 30 days because of a fire, typically from smoke poison or burns.
A "Fatal fire" is a fire where one or more persons is killed in a fire.
"Accidental Fatal fire": when started because of an accidental human act or failure in the electric equipment or an external cause.
"Deliberate Fatal fire (dødsbrand)": when deliberately is started either by the victim or another person with the intention to start a fire.
Beredskabsstyrels en (DEMA)
Type of fatality c CASUALTIES Injured person c Type of injury c
FIRE SAFETY MEASURES Alarm a Detector number, Detector type, Cause of the alarm (tobacco smoking, steam, craftsman work, technical faults, etc.)
Beredskabsstyrels en (DEMA)
Type of alarms a Ion detector, optical smoke detector, thermodetector, multifunction detector Beredskabsstyrels en (DEMA)
Automatic extinguishing systemsa Beredskabsstyrels en (DEMA)
Type of automatic extinguishing systemsa Beredskabsstyrels en (DEMA)
Compartmentation a Fire spread from hearth spaces, including presumed cause of the fire spread (open standing doors, holes in floor separation, etc.)
Fire technical equipment and functionality (smoke alarm, fire blanket, fire technical installations lationer m.v.)
Beredskabsstyrels en (DEMA)
Fire barriers a Included in compartmentation.
Beredskabsstyrels en (DEMA)
Safe areas b Smoke extractors a Beredskabsstyrels en (DEMA)
Fire brigades on site b Escape routes b Evacuation b
EU FireStat - Task 1 DENMARK DEFINITIONS 3/3 GENERIC GROUPS a. Yes, the definition exists b. No, the definition not available c. Term not specified or unclear
DEFINITIONS REFERENCE FIRE RESPONSE Fire service time of response a Alarmed (date and time) Departure (date and time) Promotion at the scene of the accident (date and time) Released from place of injury (date and time) Ready at station (date and time) Beredskabsstyrels en (DEMA)
Occupant fire response c FIRE FINANCIAL COSTS Direct financial costs b Indirect financial costs b FIRE PREVENTION Fire regulations a If it is regulated byDEMA's prepareness act.
Beredskabsstyrels en (DEMA)
Trafik- bygge- og boligstyrelsen Fire prevention c
Useful links for Denmark:
Beredskabsstyrelsen DEMA (www.brs.dk): https://brs.dk/da/nyheder-og-publikationer/publikationer2/alle-publikationer/2018/dodsbrandsrapport-2018/ Beredskabsstyrelsen DEMA (www.brs.dk): https://brs.dk/globalassets/brs---beredskabsstyrelsen/dokumenter/forskning-statistik-og-analyse/odin-dokumenter/liste-foerste-melding-ordlyd-.pdf Beredskabsstyrelsen DEMA (www.brs.dk): https://www.retsinformation.dk/eli/retsinfo/2015/9111 Beredskabsstyrelsen DEMA (www.brs.dk): https://brs.dk/globalassets/brs---beredskabsstyrelsen/dokumenter/forskning-statistik-og-analyse/odin-dokumenter/notat_blinde_falske_reelle_alarmer-2.pdf Beredskabsstyrelsen DEMA (www.brs.dk): Trafik- bygge- og boligstyrelsen
EU FireStat - Task 1 FINLAND DEFINITIONS 1/2 GENERIC GROUPS a. Yes, the definition exists b. No, the definition not available c. Term not specified or unclear
DEFINITIONS REFERENCE
FIRE INCIDENT
Accidental fire a Accidental fires include those where the motive for the fire was presumed to be either accidental, negligent or not known (or unspecified).
Deliberate fire a Deliberate fires include those where the motive for the fire was ‘thought to be’ or ‘suspected to be’ deliberate. This includes fires to an individual’s own property, others’ property or property of an unknown owner.
False alarm a No fire at the scene BUILDING DESCRIPTION Building fire a Lots of building characteristics are reported on separated Building form.
Residential buildings a Detached house, attached house or block of flats, free-time residents Non-residential buildings a Commercial building, office building, office building, transport and communication building, institutional and healthcare building, assembly building, educational building, industrial building, warehouse, building for fire services, agricultural building or other building
FIRE CAUSES
Fire causes a Cause of fire, cause of ignition Source of ignition a Several sources of ignition are listed.
Area of fire origin a Room/compartment of origin is recorded. FIRE CONSEQUENCE S IN TERMS OF
UNIT MEASURES
Fire spread a Evaluation of the extent of flame and smoke damage at arrival, at largest and at stop.
Fire horizontal spread a Horizontal area damage in m2 by flame and smoke are recorded.
Fire vertical spread a Spreading to next or previous level recorded.
Damage a Square meters and Euros Fire a Square meters and Euros Flame a Square meters and Euros Smoke a Square meters and Euros Water a Square meters and Euros Total a Square meters and Euros FATALITIES Victims a Death caused by fire Type of fatality a immediately or within 30 days CASUALTIES Injured person a Injured by fire Type of injury a Serious or mild
FIRE SAFETY MEASURES Alarm a Alarm presence, type, location, operation and reason for failure are recorded.
Type of alarms a Type of smoke alarm Automatic extinguishing systems a Sprinkler or not Type of automatic extinguishing systems a Type of sprinkler Compartmentation a Type of compartment Fire barriers a Included in compartmentation.
Safe areas b Smoke extractors a Automatic or manual Fire brigades on site a Arrival time Escape routes a Were they, could they be used, were there Lightning Evacuation a data on people evacuated with or without assistance.
EU FireStat - Task 1 FINLAND DEFINITIONS 2/2 GENERIC GROUPS a. Yes, the definition exists b. No, the definition not available c. Term not specified or unclear
DEFINITIONS REFERENCE
FIRE RESPONSE Fire service time of response a Time of arrival https://prontonet.fi/
Occupant fire response a Times after call recorded. in case of human loss Time before call recorded.
FIRE FINANCIAL COSTS Direct financial costs a Estimated by fire and smoke and water Indirect financial costs b FIRE PREVENTION Fire regulations a if the fire safety regulations apply Fire prevention b EU FireStat - Task 1 SWEDEN DEFINITIONS 1/2 GENERIC GROUPS DEFINITIONS REFERENCE FIRE INCIDENT Accidental fire Fire or fire incident: In Swedish fire is to some extent defined by language, we have separate words for unwanted and wanted fires (like a fire for heating your house). In addition a definition like "uncontrolled flame, glow or smoke that caused damage" for Brand (fire). Fire incident is defined as danger that a flame, glow, smoke or heating of flammable material might cause damage Deliberate fire False alarm BUILDING DESCRIPTION Building fire Building is a permanent construction with a roof or roof and walls placed on gr gound or partly below ground or placed for a long period at a certain place in water constructed so that people can be in it. Residential buildings not defined Non-residential buildings Separate definition for each category, hospital, elderly home, industry etc. FIRE CAUSES Fire causes Source of ignition Definitions/clarification for some of the ignition sources Area of fire origin Definitions/clarification for some of the fire origins FIRE CONSEQUENCES IN TERMS OF UNIT MEASURES Fire spread A note that many buildings consist of only one fire compartmentation Fire horizontal spread Fire vertical spread Damage Fire Flame Smoke Water Total FATALITIES Victims The definition for fatality differs for the different sources. In the fire database, all people (except fire fighter) who died at scene are counted. In the fire death statistics it is people who dies within 90 days of the fire as a direct consequence of the fire, i.e. not when trying to flee the fire etc.
In the socialstyrelsen data it is only Swedish citizens that are counted.
Type of fatality CASUALTIES Injured person An estimate of the rescue service on site Type of injury EU FireStat - Task 1
SWEDEN DEFINITIONS 2/2 GENERIC GROUPS DEFINITIONS REFERENCE FIRE SAFETY MEASURES Alarm No special definitions Type of alarms Automatic extinguishing systems Type of automatic extinguishing systems Compartmentation Fire barriers Safe areas Smoke extractors Fire brigades on site Escape routes Evacuation FIRE RESPONSE Fire service time of response
Occupant fire response
FIRE FINANCIAL COSTS Direct financial costs Potentially provided by insurance company Indirect financial costs Potentially provided by insurance company FIRE PREVENTION Fire regulations Fire prevention EU FireStat - Task 1
USA DEFINITIONS 1/1 GENERIC GROUPS a. Yes, the definition exists b. No, the definition not available c. Term not specified or unclear
DEFINITIONS REFERENCE Fire incident a Incident Types: Structure fire, Fire in Mobile Property Used as Fixed Structure:; Building fire; Confined fire (Cooking fire without extension beyond cooking vessel, Chimney or Flue Fire confined to chimney or flue, Incinerator overload but no flame damage outside incinerator, Fuel burner; boiler without flame damage outside fire box, Commercial compactor confined to contents, Trash or rubbish fire in structure but no damage to structure of contents).
Incident Reporting System 5.0 Complete Reference Guide, January 2015; NFPA 901 Standard Classifications for Incident Reporting and Fire Protection Data. Damage a Fire Spread: Confined to room of origin, Confined to floor of origin, Confined to building of origin, Beyond building of origin.
Number of stories: minor damage, significant damage, heavy damage, extreme damage Estimated Dollar Losses: Property; Contents Alarm a Type Detector Type: Smoke, Heat, Combination Smoke and Heat, Sprinkler Water Flow Detection, More than One Type, Other Automatic extinguish system a Type Type of Automatic Extinguishing System: Wet pipe, Dry pipe, Other, Dry Chemical, Foam, Halogen-Type, Carbon Dioxide, Other Special Hazard System Compartmentation b Fire barriers b Safe areas b Smoke extractors b Fire brigades on site a Yes Escape routes b Evacuation b Fatalities a Civilian: Within 1 year.
Gender, Age, Race, Severity, Cause, Human Factors Contributing to Injury, Factors Contributing to Injury, Activity When Injured, Location at Time of Incident, General Location at Time of Incident, Story at Start of Incident, Story when Injury Occurred, Specific Location at Time of Injury, Primary Apparent Symptom, Primary Area of Body Injured.
Fire Service: Gender, Age, Career/Volunteer, Date and Time, Usual Assignment, Physical Condition Immediately Prior, Severity, Activity at Time of Injury, Primary Symptom, Cause, Object Involved, Primary Body Part, Factor Contributing, Where Injury Occurred, Specific Location, Vehicle Type, Protective Equipment and Type, Protective Equipment Problem, Casualties a Civilian and Fire Service. Note that NFIRS refers to "casualties," which can either be fatal or non-fatal.
Fatalities are captured under "Severity of Injury" field. "Severity of Injury" classifications:
Useful links for the USA:
NFIRS: https://www.usfa.fema.gov/downloads/pdf/nfirs/NFIRS_Complete_Reference_Guide_2015.pdf NFPA 901: https://www.nfpa.org/codes-and-standards/all-codes-and-standards/list-of-codes-and-standards/detail?code=901
Updated on 24 December 2020
TASK 0 Challenges in Task 0 were mainly related to the difficulty to establish connections in countries were the consortium are not present. Once contacts were established, we were able to gain enough information to fulfil the need of Task 0. TASK 1 The challenges of Task 1 are represented by the difficulties in creating summary tables and gathering the related information about terminology and data collection methodology of the fire statistics for the 27 EU Member States and 8 Non-EU countries.
In particular, the challenges for the collection of the fire statistics are represented by:
language barriers,
delay in receiving the response from relevant authorities,
confidentiality policies, and
private datasets. Furthermore, the difficulties and limits in the analysis of the data gathered are due to:
the lack of available glossary of terms for the fire statistics of specific countries;
the data fields with very different definitions, and the fact that the same term may have different meanings in different countries; and
the different level of detail and number of data fields recorded which can vary in the various countries examined. Some fire statistics collect only the fire incident date, time and location while others cover pre and post-conditions of the fire incidents in various property types. TASK 2 The approval of the questionnaire took more time than expected (3 weeks instead of 1 week) delaying the execution of Task 2.
Updated on 24 December 2020
Risk ID Task number Date Raised Description of Risk Consequences (resultant effect/impact) Likelihood (score 1 for lowest - 5 highest) Impact (score 1 for lowest
R1 0, 1, 2 17/08/2020 There is a risk that some
countries do not have fire data Some EU countries will not be covered in the analysis 1 2 2 Contact representatives and other public institutions
Closed R2 0, 1, 2 23/09/2020 Language barrier Some EU countries will
not be covered in the analysis 2 1 2 Find a person to translate information or verify if EC can provide translation help in certain languages
Closed R3 1, 2 23/09/2020 Delay in receiving the
response for the summary tables Some EU, other European and International countries will not be covered in the analysis 3 2 6 Establish direct contacts with the relevant organizations and present the whole project to provide a detailed description of the outputs and benefits generated.
Open R4 1 23/09/2020 Confidential policies and
private datasets Some EU, other European and International countries will not have public fire statistics datasets 2 2 4 Establish direct contacts with the relevant organizations and kindly ask for their contribution to the project providing the relevant information.
Open R5 1 23/09/2020 Lack of available glossary
of terms for the fire statistics of specific countries Some EU, other European and International countries will not have definitions or a glossary of terms for the fire statistics 3 2 6 Focus on high level and most important definitions available
Open R6 1 23/09/2020 The data fields can have
very different definitions and the same term may measure different aspects in two or more countries Some EU, other European and International countries will have different definitions for the same terms 3 2 6 Focus on the semantic analysis of the terms
Open
Risk ID Task number Date Raised Description of Risk Consequences (resultant effect/impact) Likelihood (score 1 for lowest - 5 highest) Impact (score 1 for lowest
R7 1 23/09/2020 Different level of detail and
number of fields recorded which could vary in the various countries examined. Some fire statistics collect only the fire incident date, time and location while others cover pre and post-conditions of the fire incidents in various property types. Some EU, other European and International countries will have different detail for the fire statistics recorded 3 2 6 Group the countries examined based on the level of detail provided for the fire statistics
Open R8 2 24/09/2020 Little or delayed response
from the stakeholders o the questionnaire Not enough information from the stakeholders 2 3 6 Establish contact as soon as possible, provide support letter from EC, and allow possibility to update the analysis later.
Open R9 2 24/09/2020 There is a risk that we will
receive unclear answers in surveys Unusable contribution from some EU countries 1 2 2 Make sure questions are explicit and provide example of answers. Send the questionnaire to different stakeholders of the same country. Follow-up unclear answers
Open