Protecting People, Property and the Planet
The common causes of fire false alarms and
strategies to reduce their occurrence
Raman Chagger
LPCB
19 June 2014
Part of the BRE Trust
Introduction
– False alarms cost ~£1 billion a year in the UK.
– In the UK for the period 2011-2012 a total of 584,500
fire and false alarms reported
– 53.4% of these were “False alarms”
– Leads to:
• Drain on FRS authorities
• Business disruptions
• Reduced confidence of the public
What is a false alarm?
– A false alarm is a fire alarm signal resulting from a fire
detection and alarm system that has responded to a
cause that is not a fire, such as:
• A fire-like phenomenon (e.g. bonfire)
• Accidental damage
• Inappropriate human action (malicious)
• Equipment false alarms (fault).
– A false alarm becomes an unwanted fire signal
(UWFS) when FRS attend.
Purpose of the research project
– Investigate the causes of false alarms and identify
approaches to reduce their occurrence
– Could changes in standards and codes of practice
reduce false alarms?
– Identifying potential contributors and obtaining false
alarm data was impossible
– Suggests that gathering this kind of data is not
something that anybody does
– However, two different contributors were identifiedKings College London and Buckinghamshire and
Milton Keynes Fire Authority with very different data
Kings College London- Background
– KCL has an estate comprising of 74 buildings
– Including lecture theatres, residential spaces, libraries,
laboratories, teaching rooms, offices, restaurants and a
chapel.
– In total the premises of KCL cover a floor area of over
400,000 m2
– In the academic year 2009-2010 KCL had 283 alarm
activations.
Kings College London- Introduction
– Since 2007 KCL’s Senior Fire Safety Officer has instigated and led
the adoption of a very proactive approach to reducing false alarms;
engaging on multiple levels to reduce their numbers. The
strategies adopted include:
• effective incident reporting;
• follow-up investigations by suitably qualified personnel;
• control of contractors;
• procedural guidance;
• suitable internal responses.
Kings College London- Results
– This has led to a marked reduction in the number of activations
over the years
Academic
Year
Academic
Buildings
Activations
Halls of
Residence
Activations
Total
activations
Percentage
Cumulative
change (year % change
on year)
(from 20092010)
2009-2010
148
135
283
-
-
2010-2011
147
131
278
-2%
-2%
2011-2012
151
84
235
-15%
-17%
2012-2013
148
38
186
-21%
-34%
Kings College London- Alarm Activations 2009-13
Kings College London- Logging alarm activations
– KCL generate yearly internal reports
– All alarm activations, time/date of event, campus, building, floor,
location, zone, device, cause (generic heading), cause code (in
more detail), whether or not LFB attended, more (detailed)
information and the current status of each incident.
Cause
Cause Code
(Breakdown)
LFB Attended
Unwanted
Building Works
no
KCL maintenance contractors testing chimney & smoke set off alarm
Unwanted
Building Works
no
A contractor broke MCP by accident whilst removing rubbish
Unwanted
Unwanted various other
yes
Summer visitor mistakenly broke break glass whilst trying to open
magnetic locked door
Unwanted
Near Miss Fire incident
- cooking
yes
Cooking left unattended, oil in wok overheated producing large
quantities of smoke
Unwanted
Cooking
no
Toaster in crèche kitchen
Unwanted
Building Works
yes
Dust from building works
More Information
Kings College London- Further work with data
– Information detailing how each false alarm was addressed to
prevent future re-occurrence was not available.
– Therefore the raw data supplied was reviewed and analysed
– KCL provided raw data for 2010-2013 (699 incidents)
– Reduced to 432 valid false alarms
– Classified in to 110 activation categories
No.
Activation Category
Frequency
1
MCP accidentally triggered
27
2
General dust from building works
25
…
…
…
110
Water getting into the panel
1
Total
432
Kings College London- Proposed solutions
– The 110 activation categories were assessed for resolution through
6 physical interventions (summarised below)
– Each activation can have more than 1 solution
Solution
1
Proposed intervention action
No. of potential
causes
resolved
69.2%
3
Replace with multi-sensor
Use of appropriate approved detector/s correctly
located
Use of protective covers over approved MCPs with
adequate signage and CCTV where required
4
Use of EN 54-2 approved analogue addressable panel
10.2%
5
6
Better control of contractors
More rigorous maintenance of the system
9.7%
6.0%
2
43.5%
16.7%
Kings College London- Images of solutions
MULTISENSOR DETECTOR
MANUAL CALL POINT
WITH PROTECTIVE COVER
Images courtesy of Tyco Fire Protection Products
Kings College London- Multi-sensor detectors
– Multi-sensor detectors look for more than one fire signature (such
as heat and smoke)
– Few false alarm sources produce both smoke and sufficient heat.
– The cost for a optical smoke/heat multi-sensor are ~ £5-10 more
than an optical smoke detector.
– £1 billion cost for 312,000 false alarms the cost per false alarm
works out at ~£3.2k to businesses and FRS’s.
– ~ £300 has been the estimate used for costs to FRS’s
– Average cost to businesses is ~£2.9k per callout.
Bucks & MK Fire Authority- Intro
– Malcolm Brightman from Buckinghamshire and Milton Keynes Fire
Authority since 2006
– Unwanted Fire Signals Officer (only one?) in the UK
– To investigate UWFSs by making contact, to identify corrective
actions and to provide guidance when required.
– As a result Malcolm has reduced the number of fire alarm signals
by 45% over 7 years in Buckinghamshire and Milton Keynes.
– Martin Duggan (FIA) and I interviewed Malcolm- article is presented
in Fire Magazine, Oct 2012 (pgs. 39-41).
Bucks & MK Fire Authority- Process
– Malcolm reviews the data recorded in the IRS database
– Identify the worst offenders
– Engages with ~5 premises a week. Phone call or email is often
enough
– If not then a site visit is arranged with the responsible person (RP)
– The corrective measures and long term result are not recorded
electronically so statistical data is not available to try and identify
the frequency and type of successful interventions.
– Anecdotal accounts reviewed
– Data generated from the IRS for BMKFA reviewed
Bucks & MK Fire Authority- IRS
– In October 2007 the first version of the Incident Recording System
(IRS) database was released.
– web-enabled IRS with the intention to modernise the collection and
subsequent statistical handling and publication
– collecting, validating and transmitting data to DCLG on all incidents
attended by the FRS.
Bucks & MK Fire Authority- Causal Factors
– Fire crews choose one of 53 causal factors which are generic
descriptions that can lead to a loss of accuracy
– Subjective nature of completing forms and technical competence of
those completing the forms
Bucks & MK Fire Authority- Categorisation
– The categorisation of false alarms from the IRS database differs
from BS 5839-1:2013
– Both have Malicious False Alarm and Good Intent false alarm.
– IRS has “Fire alarm due to Apparatus”
– BS 5839-1:20139 (Clause 3.17) has:
• unwanted alarms, in which a system has responded, either as
designed or as the technology may reasonably be expected to
respond;
• equipment false alarms, in which the false alarm has resulted from
a fault in the system;
– RP completes the record of the false alarm incident to BS 5839-1
which is different from IRS database.
Bucks & MK Fire Authority
– The top 6 causal factors for UWFSs from completed IRS entries for
BMKFA from June 2009 to April 2013 (6612 independent events)
Causal Factors
Unknown
System: smoke alarm
Faulty
Human Accidentally/
carelessly set off
Contaminants Dust
Human Cooking/burnt
toast
System: other Faulty
Number
(% of total)
1351 (20.4%)
893 (13.5%)
778 (11.8%)
734 (11.1%)
620 (9.4%)
615 (9.3%)
Further explanation
Undetermined cause
Faulty smoke alarm suspected but not
validated
This includes all MCPs, smoke detectors etc.
accidentally set off by humans
Dust in detector but no details of what
proportion from smoke, beam etc.
This constitutes all events including misuse
e.g. toaster used in office
Faulty system- no detail of panel, device or
cabling etc. is provided
Bucks & MK Fire Authority- Strategies implemented
Observed cause of UWFS
Implemented action
Smoking under smoke detectors
Changed to heat detectors
Activating MCPs to call warden
MCP’s removed and replaced with correctly
located smoke detectors.
Isolated the zone prior to works.
Contractors producing dust or paint
fumes setting off smoke detectors.
Dust settles on smoke detector cover
during works. Once removed dust
activates the smoke alarm.
Cooking triggering smoke detectors
Hitting MCP button to open doors
rather than exit button.
MCPs maliciously activated
Educated contractors on the need to clean
smoke detector covers some time prior to
their removal
Specified the correct detection (heat
detectors in kitchen). Automatic extractor fans
linked to ovens that come on
MCP covers used to prevent incorrect
operation
CCTV in entrances for security purposes
deter malicious activations
Bucks & MK Fire Authority – Guidance for RPs
– BMKFA report provides the following guidance for RPs/managers:
• Consult with your fire alarm engineers to ensure the correct type of
detection is installed in the most suitable location.
• Have detection and alarm systems regularly serviced.
• Ensure there are appropriate ‘call filtering’ measures in place such
as a ‘Double Knock’ procedure, to enable safe investigation
• Ensure Fire Marshalls are appointed and have received appropriate
training.
• If system is linked to a ‘Call monitoring’ centre make sure they are
delivering the service you require.
• Ensure out-of-hours key holder can respond to alarm activations
speedily
Bucks & MK Fire Authority- Statistics
– In 2005/6 the total numbers from domestic and commercial
premises were 3308 and in 2012/13 were down to 1815.
Bucks & MK Fire Authority
– It has been estimated that the associated annual savings for
BMKFA are in the region of around £450k per year.
– Confirms that the strategies used by BMKFA are the most direct
and effective means for reducing UWFSs
– BMKFA have reduced false alarms by 45% over 7 years
– If the same methods used by all other FRS’s in the UK with the
same reduction FRS’s in the UK this could save £42m per year.
– This reduction could save UK businesses ~£408m per year
Conclusions
– False alarms can be reduced!
• Logging every false alarm and investigating
• Controlling how contractors work
• Demanding high quality service from fire alarm service providers
• Regular servicing of fire alarm system
• Use of appropriate approved detector/s located correctly
• Technology can help (e.g. use of multi-sensors)
• Educating users
• Common sense (applying protective covers over MCPs)
• Use of investigative periods especially during pre-alarm stages
Further work
– A paper reporting the findings is available
for free from the BRE website:
http://www.bre.co.uk/podpage.jsp?id=1752
– The Scottish Fire and Rescue Service
Board are very keen to continue this
research with BRE and other parties.
– This research will utilise a false alarm
investigator that will be on standby to
attend live incidents and accurately
diagnose false alarm causes.
– Subscribe to receive monthly newsletters:
http://www.redbooklive.com/register.jsp
Acknowledgements
– Thanks to Kings College London and Buckinghamshire and
Milton Keynes Fire Authority for their contributions.
Questions