Specifiers’ Guide

Section 1

A guide to fire alarm

systems design

www.apollo-fire.co.uk
INTRODUCTION TO THE GUIDE
Apollo Fire Detectors has created this guide as an aid for use by all who have responsibility for
specifying or designing systems incorporating automatic fire detectors.

It is intended as a guide to good practice but does not aim to be an exhaustive guide to fire design
standards or codes.

The guide is to assist in design and specification of fire systems.

Notes on the text are indicated by the convention of a superscript number. An example of this
might be ‘codes and standards2’. Notes appear at the foot of the relevant page.

For any queries concerning technical matters, please contact Apollo’s Product Support Department
(email: [email protected]).

All other comments and suggestions should be directed to the Marketing Department at Apollo
(email: [email protected]).

All the information is given in good faith but Apollo Fire Detectors cannot be held responsible
for any errors or omissions. We are grateful to readers who notify us in the event of any need for
corrections.

APOLLO FIRE DETECTORS

Apollo Fire Detectors has specialised in the design and manufacture of high quality fire detection
products since 1980. In that time, the company has broadened its capability from a straightforward
focus on conventional fire detectors to include the manufacture of sophisticated analogue addressable
detectors and interfaces for monitoring and controlling equipment in fire protection systems.

Apollo is an Open Protocol manufacturer and over the past 30 years has developed trusted
partnerships with over 70 Panel Manufacturers who supply panels incorporating Apollo’s open, digital
protocol, meaning the customer is free to choose different companies to service or maintain the
system. All Apollo products are forwards and backwards compatible, giving customers the added peace
of mind that they will be able to source fire detectors that are compatible with their existing devices.

Our Product Lifetime Guarantee provides a warranty on our products, which for detectors is10 years
(CO detectors, 5 years). The guarantee supports our recommended working life of the product and
endorses our commitment to providing reliable, quality fire detection products.

Apollo is part of the Halma group of companies. Halma is a FTSE top 250 listed PLC with over 40
subsidiaries worldwide, all engaged in specialist engineering activities.

Apollo Fire Detectors Limited Tel: +44 (0)23 9249 2412

36 Brookside Road Fax: +44 (0)23 9249 2754
Havant Email: marketing@apollo-fire.com
Hampshire PO9 1JR productsupport@apollo-fire.com
England www.apollo-fire.co.uk
 Specifiers’ Guide Section 1 Contents

CONTENTS
6

CONTENTS
LIST OF ABBREVIATIONS

OPEN, CLOSED AND DIGITAL PROTOCOLS – WHAT DOES IT ALL MEAN? 7

Section 1

1. DESIGNING A FIRE PROTECTION SYSTEM 10

Using a code of practice 10

Manual and automatic fire detection and alarm systems 10
Doing a risk assessment 10
Recommendations of BS 5839-1:2002 (as amended) 10
System components 11
About BS 5839-1 11
Fire alarm system projects 13
The design process
The installation process
Communication with the fire brigade
Limitation of false alarms
The commissioning process
User responsibilities
Extensions and alterations to existing systems
Servicing and maintenance

1.1 SYSTEM PLANNING AND CONSULTATION 14

1.2 SELECTING THE CORRECT CATEGORY OF SYSTEM AND SYSTEM TECHNOLOGY 15

Category of system 15
Proposing the system category 18
Prescriptive categories 20
Non-prescriptive categories 20
System technology 21
Non-addressable systems 21
Analogue addressable systems 22
Intelligent systems 23 3

1.3 SYSTEM ZONE CONFIGURATION 24

Detection zones 24
Alarm zones 28
Non addressable versus analogue addressable zones 30
Visual indicators on detectors 30
Remote visual indicators 30
 Specifiers’ Guide Section 1 Contents

CONTENTS
CONTENTS

1.4 SELECTING THE CORRECT DETECTORS 32

Types of fire detector 32

Choosing the right detector 33
Smoke detectors 33
Heat detectors 35
Flame detectors 37
Carbon monoxide (CO) fire detectors 37
Multisensor detectors 38
General detector choices 39

1.5 LOCATION AND COVERAGE OF DETECTORS 41

Detector spacings 45
Flame detection 50
Duct detection 50
Optical beam smoke detection 50

1.6 MANUAL CALL POINTS 52

1.7 AUDIBLE AND VISUAL ALARM REQUIREMENTS 54

Audible alarm devices 54

Calculating sound pressure level 57
Visual alarm devices 61

1.8 CONTROL AND INDICATING EQUIPMENT/CONTROL PANEL EQUIPMENT 62

1.9 POWER SUPPLY REQUIREMENTS 63

Standby times 63
Standby times for Category P systems 63
Standby times for Category L systems 63
Calculating required battery capacity 64

1.10 CIRCUIT DESIGN 65

Detection circuits 65
Upgrading a non-addressable system 67
4 Alarm device circuits 68

2. INSTALLATION AND CABLES 71

Installation 71
Variations from BS 5839-1 71
Variations to contract 71
Cables and wiring 71
Use of 4-core cable 73

3. COMMUNICATION WITH THE ALARM RECEIVING CENTRE 75

 Specifiers’ Guide Section 1 Contents

CONTENTS
76

CONTENTS
4. FALSE ALARMS
Choosing the detector to avoid false alarms 79

5. THE COMMISSIONING PROCESS 80

Handover 81

6. USER RESPONSIBILITIES (THE RESPONSIBLE PERSON) 83

Keeping records in the log book 84

7. EXTENSIONS AND ALTERATIONS TO EXISTING SYSTEMS 85

8. SERVICING AND MAINTENANCE 86

Daily 87
Weekly 87
Monthly 87
Quarterly 87
Periodically 88
Annually 88
Handover for each service visit 89

LIST OF OTHER STANDARDS 90

UK Standards and Codes of Practice 90

Fire Safety Risk Assessment Guides issued by HM Gov for Eng & Wales 91
Scottish guidance on the Fire (Scotland) Act and the Fire Safety (Scotland) regulations 92
Healthcare 92
Standards and codes in other countries 92
Regulations 92

COMPLIANCE WITH THE EU CONSTRUCTION PRODUCTS DIRECTIVE 89/106EEC (CPD) 93

FIA APPLICATION GUIDELINES FOR CARBON MONOXIDE (CO) FIRE DETECTORS 95

1. Introduction 95
2. General guidance on the application of CO fire detectors - Important note 95
3. Risks where CO detectors can provide a means of early fire detection 96
4. Siting of CO fire detectors 97 5
5. Risks where CO fire detectors are not recommended as the prime means of detection 97
6. Lifetime factors including testing, servicing, maintenance and replacement 98
 Specifiers’ Guide List of Abbreviations

LIST OF ABBREVIATIONS
CONTENTS

AFD Automatic Fire Detection

ARC Alarm Receiving Centre

ASD Aspirating Smoke Detection

ATEX ATmospheres EXplosive

CIE or cie Control and Indicating Equipment

CO Carbon Monoxide (detector)

CoP Code of Practice

CP Competent Person

CPC Circuit Protective Conductor

CSP Critical Signal Path

dB(A) decibel, (A-weighted)

FD&A Fire Detection and Alarms

F&RS Fire and Rescue Service

MCP Manual Call Point

PII Professional Indemnity Insurance

RoR Rate of Rise (heat detector)

RP Responsible Person

6
 Specifiers’ Guide Open, Closed and Digital protocols - what does it all mean?

OPEN, CLOSED AND DIGITAL PROTOCOLS

– WHAT DOES IT ALL MEAN?
What are Protocols?
The term ‘protocol’, when used with reference to electronic products, refers to the way in which the
products have a communication ‘language’ which is termed ‘protocol’.

Protocols are often referred to as ‘open’, ‘closed/managed’, ‘digital’ and ‘analogue’. It is important to be
sure what each term means when comparing different types of fire detection system.

Building Services Protocols

With the development of more and more products that need to communicate with each other, in
particular products used for building services in sophisticated modern buildings, the need has arisen for
protocols to be agreed across a whole range of manufacturers or even entire industries. For instance, the
electrical trade has systems for switching large numbers of current-consuming devices, such as lights,
by using simple loop wiring and a data transmission protocol, rather than miles and miles of cables for
individual circuits. Examples of such protocols are ‘LonWorks’ and ‘EIB’.

Because such protocols are available for any manufacturer to use, they are often referred to as ‘open’.
The fire detection industry does not currently use such protocols and the term ‘open’ has come to mean
something different in this particular industry.

Fire Industry Protocols

In the fire detection industry addressable systems use control panels and detectors (and, of course,
devices such as interfaces) which communicate with each other by means of a protocol. Some
manufacturers offer both panels and detectors. These companies have no need to disclose the nature
of their protocol to anyone, since they offer all the elements needed to provide a complete addressable
system. No equipment supplied by other manufacturers is expected to be compatible with such systems,
so the protocol used is said to be ‘closed/managed’.
A number of manufacturers of detectors, including Apollo, make no control panels; they have built up
partnerships with independent panel manufacturers and, in some cases, companies who offer special
equipment, such as aspirating detection systems. The detector manufacturer determines the protocol
used by the detectors and produces the information and technical data required to panel manufacturers
7
in order to design panels that will drive the detectors. Since all details of the protocol must be disclosed,
it is referred to as an ‘open’ protocol. Apollo has written agreements with many panel manufacturers to
allow them to use the information which remains the intellectual property of Apollo.

Closed/Managed Protocols
Manufacturers of equipment using closed/managed protocols claim that all elements (detectors, panels,
call points, interfaces, special detectors such as beam detectors) will work harmoniously with each
other, since it is all designed and made by the same company. The percieved implication is that a system
comprising detectors and interfaces from one manufacturer and panels from another cannot work as well
with each other.
 Specifiers’ Guide Open, Closed and Digital protocols - what does it all mean?

Open Protocols
The manufacturers of the components of a system with an open protocol would reply that products from
different manufacturers of fire products work just as well with each other as does, for example, a McLaren
racing car with a Mercedes-Benz engine. Indeed, it may be said that there is an advantage in having different
specialist manufacturers concentrating on their own skill areas.

After-sales service
Whatever the arguments for either system may be, one point is indisputable: the owner of a fire protection
system with a closed/managed protocol is dependent on just one supplier for all spare parts, servicing,
modification and upgrade of the system, since no other manufacturer’s products will be compatible.
The owner of a system using an open protocol can freely choose a different company to service the system
or to supply different upgraded equipment, as well as make alterations to the system – for example, change
the text of a detector or modify the cause and effects.

Analogue and digital protocols

The term ‘analogue’ is used to describe a signal which goes up and down steplessly. See Fig 1.
Signals that record phenomena such as the increase of smoke or heat are necessarily analogue at source
and this is why fire detectors are described as analogue. For example, a heat detector will record stepless
increases in temperature from a typical starting point of 20-21oC (comfortable room temperature) to an
alarm level of 55oC, however fast the increase.

Each point on the analogue signal indicates a particular value. The problem with analogue signals, however,
is that, if, during transmission, electrical corruption affects the signal, a ‘2’ might appear as a ‘3’, for example.

The description of the communications protocol as “analogue” should not be confused with the description
of a fire alarm system as “analogue” or “analogue addressable”. An addressable fire alarm system is one in
which signals from manual call points and detectors are individually identified at the control and indicating
equipment. In practice, nearly all addressable systems are of the “analogue” type, meaning that signals from
each detector are individually processed to enable a representation of its amount of sensed phenomenon
(i.e. heat, smoke etc) to be transmitted to the control and indicating equipment. The method of transmission
of this value may be digital or analogue, but is normally digital.

Digital signals
The word ‘digital’ describes a signal that consists of a series of ‘0s’ and ‘1s’ or ‘offs’ and ‘ons’ which go to make
up a message in binary arithmetic. The advantage of this system is that there is a much lower risk of the
8
signal being poorly transmitted and hence giving wrong information.
Fig 2 shows the simple levels of 21, 22 and 23 (illustrated by the curve in Fig 1) as digital signals.
Each degree in the example of a heat detector given in the previous section can be expressed in digital form
(ie, binary arithmetic). ‘30’ degrees Celsius would then be ‘11110’ and this is what Apollo detectors would
transmit. Apollo fire detectors have always used a digital protocol which has remained basically unchanged
since its inception in 1986. It has been extended – in two steps, once for XP95 and a second time for
Discovery – but never modified.
Most detector manufacturers have now adopted digital transmission protocols.
In summary, an open protocol allows freedom of choice by the specifier, the installer and the end user of
the fire detection system.
A digital protocol is much less susceptible to corruption than the analogue protocol and is to be preferred
in a system which is life-critical.
 Specifiers’ Guide Open, Closed and Digital protocols - what does it all mean?

Apollo has a digital, open protocol

23

22

21

Fig 1 Analogue signals

1 2 4 8 16 32 64 1 2 4 8 16 32 64 1 2 4 8 16 32 64

20

10

0
9

Fig 2 Digital signals

 Specifiers’ Guide Section 1 Designing a fire protection system

1. DESIGNING A FIRE PROTECTION SYSTEM

Using a code of practice
Section 1

In many buildings, automatic fire detection is a key component of the overall strategy to protect
the building, its occupants and its contents from fire. To meet demands for fast, reliable detection,
manufacturers like Apollo Fire Detectors have developed more sophisticated, more intelligent products
- products that balance optimum performance with increased immunity to false alarms. But detectors
form only part of a fire detection and alarm system. System technology has also advanced greatly in recent
years. Many analogue addressable systems incorporate powerful features aimed at ensuring fast, efficient
response to fire. For a system to give maximum benefit, however, it must be suitable for its intended
application, and it should meet the recommendations of fire alarm standards and codes.
In this Specifiers’ Guide, we aim to explain the process of designing a fire detection and alarm system and
help you in specifying suitable products and components. The design of a system must usually satisfy the
criteria laid down in standards and codes. These vary and it is important that you comply with standards
and codes that apply in your country. However, to explain the impact that standards can have on the design
process, we have included a guide to the British Standard for Fire Detection and Alarm Systems, BS 5839-11.
This is presented in Part 1 of this guide.

Manual and automatic fire detection and alarm (FD&A) systems

It is not possible to prevent all fires. Occasionally a fire will occur. If someone notices it, they can warn
others by operating a manual call point (MCP). If a fire occurs in an unoccupied room, the fire could
develop, cause damage and threaten lives before it is noticed. Automatic fire detection (AFD) with smoke
and heat detectors can be used to compensate for this risk. For example, AFD in a boiler room compensates
for the risk of the boiler becoming faulty and causing a fire, or AFD in a bedroom compensates for the
possibility of a fire occurring whilst the room is unoccupied or the occupant is asleep.

Doing a risk assessment

The Regulatory Reform (Fire Safety) order 2005 requires that the employers and other persons having
control of premises must carry out, or have a consultant carry out on their behalf, a fire risk assessment.
The findings of that risk assessment may require that an FD&A system be installed to give adequate warning
of fire. Risk assessments are usually quite straightforward, for ordinary premises, and need not introduce
significant cost.

Recommendations of BS 5839-1:2002
10
This part of the guide aims to take you through the recommendations of BS 5839-11, the British Standard
code of practice for the design, installation, commissioning and maintenance of fire detection and alarm
systems in the UK. An understanding of its recommendations is fundamental to good system design. It is
also essential to know the code’s recommendations when specifying products and systems. Users too will
benefit, as the code contains many recommendations that apply to them.

In a short guide such as this, we cannot explain every clause of BS 5839-1. Rather, we aim to focus on the
key clauses that govern the process of system design and equipment selection, as well as those that relate
to the management of a system, once it has been installed and handed over. To fully understand all of its
recommendations, we would recommend you read through BS 5839-1 in conjunction with this guide.

BS 5839-1 is not the only standard relating to fire detection and alarm systems in buildings. Depending upon
the application, you may find that it is necessary to comply with other standards or codes. We have listed
a number of these later in this guide. Also listed are some of the standards that apply to system design in
other countries.

1 BS 5839-1 Fire detection and fire alarm systems for buildings - Code of practice
for system design, installation, commissioning and maintenance.
 Specifiers’ Guide Section 1 Designing a fire protection system

System components
Most of the components used in a fire alarm system, such as the control panel, smoke and heat detectors,
alarm sounders, interfaces and ancillary items have their own standard. In Europe the base standard is EN
54 and this has separate parts for each type of component, for example smoke detectors are covered by EN

Section 1
54 part 7. The control panel is split into two parts, the control and indicating equipment is covered by EN 54
part 2 and the power supply by EN 54 part 4. The control panel usually houses both items, so for simplicity
in this guide we will refer mainly to the control panel.

The various parts of EN 54 have been adopted as British Standards, then becoming BS EN 54 and it is these
that are referred to in BS 5839-1.

About BS 5839-1
Although preceded by other British Standards on fire alarm systems, BS 5839-1 was first published in 1980. It
was revised in 1988 to take account of changes in fire alarm technology and, in particular, the introduction
of analogue addressable systems. The current version was published in 2002. It again recognises new
developments in fire alarm technology such as carbon monoxide fire detectors, technical changes (for
example, regarding the fire resistance of cables) and new concepts such as the introduction of categories of
false alarms. The present standard is in a format showing

• “commentary”, in italics, for information and background knowledge; and

• “recommendations”, in regular font, as the parts that should be followed if compliance with BS 5839-1 is
being claimed.

The recommendations state, for example, “manual call points ‘should’ be located at all exits to the open
air”. The use of the word ‘should’ shows that if you are claiming compliance to BS 5839-1, then you would
comply with that recommendation. Recommendations do not use the words “shall” or “must”, those being
used in product standards such as the parts in BS EN 54, and in various government laws and regulations.

BS 5839-1 contains recommendations that apply to a wide range of different buildings. Although it covers
all types of building (other than dwellings), its recommendations are general in nature. For some buildings,
such as hospitals, more specific recommendations are required, and these are to be found in other codes
and standards2.

It is also important to remember that BS 5839-1 does not recommend whether or not a fire detection and
alarm system should be installed in the first place. The need for a system in any given premises will be
dependent on the type of premises. For example, it should be fairly obvious that, in a hotel where people
sleep, automatic fire detection (AFD) in all rooms would detect a fire and electrical alarms could then be
operated to wake the occupants so that they can evacuate to a place of safety. In a much simpler building, 11
in which all rooms are occupied and where no-one sleeps, if fire does start, it would be quickly seen by
someone and they could warn others by shouting a warning, ringing a hand bell or operating a MCP. In
this case AFD may not be necessary. This process of looking at a building, judging its size, the number
of people, whether they are awake or asleep, whether they would quickly notice a fire, or hear an alarm
warning is all part of a fire safety risk assessment. Legislation in the UK states that a fire risk assessment
MUST be done, and any organisation responsible for business premises that fails to do the risk assessment,
and take appropriate action, would be in breach of the law.

AFD can also be required by property insurers in order to safeguard valuable buildings and contents. In
new buildings, fire detection is often an essential element of a fire engineering strategy.

2 A list of other codes and standards that relate to fire detection and alarm
systems can be found on page 90 of this guide.
 Specifiers’ Guide Section 1 Designing a fire protection system

BS 5839-1 is a code of practice and not a specification. It contains recommendations, not requirements.
As such, there is scope to vary the recommendations of the code, particularly if the recommendations
are unsuitable or would lead to a system that would be difficult to install. However, there must be sound
reasons to deviate from the code’s recommendations, and any variations of this kind should be agreed
Section 1

with all of the interested parties. It may also be the case that BS 5839-1 is specified in a contract. In this case,
compliance becomes a contractual obligation.

BS 5839-1 is divided into seven sections:

1. General
This covers planning and consultation, and in particular, the need to specify the category of system.

2. Design considerations
This covers the selection of equipment, location and siting of devices such as detectors, interconnection of
system components and the methods of raising the alarm.

3. Limitation of false alarms

This highlights the problem of false alarms, including what causes false alarms and how they should be
avoided through careful design.

4. Installation
Site work and practices associated with installing the system are covered.

5. Commissioning and handover

Recommendations relating to the testing, commissioning and certification of the system are outlined here.

6. Maintenance
This covers testing, services, repairs and modification.

7. User responsibilities
This addresses the need for the user to appoint a responsible person to manage the system once installed
and the need to keep records relating to the system in a log book.

It is not our intention to follow the sequence of clauses and recommendations as they are set out in the
code. Instead, we will highlight key recommendations as we take you through the process that needs to
be followed in designing a fire detection and alarm system.

12
 Specifiers’ Guide Section 1 Designing a fire protection system

Fire alarm system projects

This guide has a number of sections, dealing with the important steps involved in the progress of a fire
detection and alarm system project.

Section 1
1. The design process
The process of designing a fire detection and alarm system involves a number of elements as follows:

1.1 System planning and consultation

1.2 Selecting the correct category of system and choosing the system technology

1.3 System zone configuration

1.4 Selecting the correct detectors

1.5 Location and coverage of detectors

1.6 Manual call points (MCPs)

1.7 Audible and visual alarm requirements

1.8 Control panel equipment

1.9 Power supply requirements

1.10 Circuit design

We will consider each one of these in turn in this guide.

2. The installation process

In this section of the guide requirements for cables will also be discussed.

3. Communication with the fire and rescue service

Automatic communications, usually via an alarm receiving centre.

4. Limitation of false alarms

The responsibility for limiting false alarms is discussed in this section along with information on causes of
false alarms and how to avoid them.

5. The commissioning process

The process of commissioning includes handover to the user or purchaser and training of staff.

6. User responsibilities 13

All systems require management after installation and user responsibilities are explained in this section.

7. Extensions and alterations to existing systems

Design and documentation for modifications to systems.

8. Servicing and maintenance

This section includes routine and non-routine attention.
 Specifiers’ Guide Section 1 System planning and consultation

1.1 SYSTEM PLANNING AND

CONSULTATION
Section 1

The purpose of a fire detection and alarm system is to support the fire safety strategy for the building in
which it is to be installed. Therefore, the system design must support the fire evacuation procedures that
are to be followed.

The system requirements must be considered at an early stage. This is part of the consultation that
BS 5839-1 (Clause 6) recommends should take place between the user or purchaser and other interested
parties. Interested parties can include:

- the local fire and rescue authority and the relevant building control body
- the Health and Safety Executive (the body charged with enforcing workplace health and safety in
the UK)
- the property insurer.

It is important that the system designer ascertains the requirements for the system by consultation with the
user or purchaser of the system. There may also be a need to consult others such as architects, mechanical
and electrical consultants and fire engineering consultants.

Outside the UK different legislation and building codes apply. It is essential that the local building control
and/or fire service officials or any other authority having jurisdiction is consulted.

Any proposal to vary the recommendations of BS 5839-1 should be agreed with the relevant interested
parties. Variations should be documented on the design, installation or commissioning certificate
as appropriate.

Consultation is a fundamental part of the planning of a system because it enables the purpose of the system
to be determined, and allows the designer to establish the requirements for the new system. From this
point, it is possible to go on to specify the category of system.

14
 Specifiers’ Guide Section 1 Selecting the correct category of system and choosing the system technology

1.2 SELECTING THE CORRECT CATEGORY

OF SYSTEM AND CHOOSING THE
SYSTEM TECHNOLOGY

Section 1
Category of system
BS 5839-1 (Clause 5) divides fire alarm systems into a number of different categories. These relate to the
purpose of the system, principally whether it is to protect life (categories L1, L2, L3, L4, L5 and M) or to
protect property (categories P1 and P2).

There are six different life protection categories.

L1 Category L1 systems should have AFD throughout the building. The only possible exceptions are
toilets, toilet lobbies, stairway lobbies, and small cupboards of less than approximately 1sq m,
provided in each case, the room in question is of low fire risk (i.e. virtually no ignition sources or
combustible material and no likelihood of fire spread). Detectors should be optical detectors on
escape routes. In rooms, the detection would be chosen to suit the risk, without causing
false alarms.

L2 Category L2 systems should be as L3, but with additional AFD in rooms of high risk, even if they do
not open onto an escape route. For example, rooms with a high probability of ignition: plant rooms,
hot-work rooms etc. Rooms or areas, in which there would be serious risk to people if a fire started
in the room, should also be protected with smoke detectors. All these rooms should be listed or
identified so the installer knows where the additional detectors should be fitted.

L3 Category L3 systems should be as L4, but with AFD installed in all rooms or areas opening onto escape
routes. The detectors in these rooms need not necessarily be smoke detectors, but should be chosen
to suit the risk, for example, in a kitchen, heat detection would be appropriate to avoid false alarms.

L4 Category L4 systems should have MCPs and sounders as for category M with automatic fire
detection (AFD), using optical smoke detectors, installed on the escape route corridors and
stairways, so that occupants can be warned of fire if it reaches an escape route. Optical smoke
detectors are recommended for escape routes in BS 5839-1 because escape routes themselves
should not contain equipment, substances, fixtures or materials that could start a fire or add
significantly to a fire, so any smoke will have travelled some distance and the smoke particles will
15
have aged and coalesced into larger particles. These larger particles are best detected by optical
smoke detectors (see a later section for a description of detector types).

L5 Category L5 systems do not need to include MCPs, but sounders are recommended throughout the
building. AFD could be designed to meet a specific fire safety objective, which may not necessitate
installing one of the other categories. One simple example could be a single detector in a high risk
area (e.g Boiler Room) where, if a fire occurred, it could threaten the escape route for occupants.
A more complex example would be for a large building with many occupants, where the fire safety
objectives can be met without following a recognised code of practice. For category L5 a specification
or plan should be produced so the installer knows the type of AFD and where it should be fitted.

M Category M systems should have manual call points (MCPs) at each final exit to the open air and at
each storey exit near stairs, with sounders throughout the building, but no AFD.
For Category L5, M is not included as standard, so if if MCP are required, then this needs to be
made clear: L5/M.
 Specifiers’ Guide Section 1 Selecting the correct category of system and choosing the system technology

There are two different property protection categories.

P1 Category P1 systems should have AFD throughout the building covering all rooms and areas. A
means of contacting the fire and rescue service (F&RS) would need to be included. This could be
automatic transmission equipment or 24 hour cover by security personnel who could telephone
Section 1

the F&RS.

P2 Category P2 systems would have AFD installed only in defined parts of the building, sometimes as
specified by the insurers, again with a means of contacting the F&RS.

M For categories P1 and P2, M is not included as standard, so if MCPs are required, then this needs to
be made clear: P1/M; or P2/M.

Examples of categories of system that might be appropriate to particular buildings would be3:

Hotels L1 or L2

Hospitals (refer to HTM 05-03 Part B) L1 (with possible variations)

Offices M

Large offices, high value M/P1 or M/P2

Shopping complexes L1

Factories/warehouses M/P1 or M/P2

For both category L2 and category P2 systems, it will be necessary for someone to define the areas in
which detectors are to be installed. This should normally be the purchaser, but if you are the designer you
are responsible for this, you should ensure that the areas selected are agreed by the interested parties.
Reference to other standards and codes may be required in order to define these areas4. For example, for
a category L2 system, it would be relatively simple to make a list of boiler rooms, kitchens, plant rooms etc,
to be protected in addition to a category L3 system, but less obvious could be day-lounges where people
could be asleep.

16

3 Examples of appropriate categories of system for typical premises are given in Annex A of BS 5839-1.
4 See page 90 of this guide
 Specifiers’ Guide Section 1 Selecting the correct category of system and choosing the system technology

Where, in any category 1 system, there is a need to protect a room or area, voids deeper than 800mm within
the room or area would, in general, need to be protected (Fig 3a). For example a category L4 would be
protected with smoke detectors located in the escape route corridors and stairways. A void deeper than
800mm above the escape route would need to be protected with fire detection, but a void elsewhere would

Section 1
not need protection. However, if a void bridges across the escape route and other areas then the whole of
that void would need fire detectors, see figure 3b and 3c.

> 800mm

Fig 3a Voids deeper than 800mm generally need to be protected

Detectors not required

in the rooms, nor in the
voids above the rooms

17
Esc
ap
Roo Rou e
m te
s

Roo > 800mm

m s
Detectors recommended in escape routes
and in void immediately above only

Fig 3b Voids above rooms

 Specifiers’ Guide Section 1 Selecting the correct category of system and choosing the system technology

Detectors
throughout
the void
Section 1

Roo Esc
det m s ap
Rou e
e
req ctors
uire not
d fo
r L4
te

Roo
det
em s > 800mm
req ctors
uire not
d fo
r L4

Detectors recommended in escape routes and

in voids if a fire could spread from one area to
threaten the escape route

Fig 3c Voids bridging an escape route

Where only part of the building is protected by AFD, account should be taken of the potential for a fire
to start in an unprotected area and spread to the area covered by detectors. In category P2 systems, there
should normally be fire resistant construction separating the protected from the unprotected areas, but
there is no guidance on what standard of fire resistance would be appropriate for this construction. The
insurers may require a 1 hour fire resistant construction, or otherwise category P1 may be necessary. The
insurer should be consulted for advice on property protection.

Remember that BS 5839-1 does not recommend which category of system should be used in any particular
building. It would be part of the risk assessment procedure to determine the appropriate category.

Establishing the category of system is a key step in the design process. This will determine the areas in
18
which automatic fire detection will need to be provided. It also has implications for detector choice, siting
and spacing5.

Proposing the system category

BS 5839-1 states that the specifier or purchaser should ideally specify the category of the system required,
but it goes on to state that in the absence of the category being specified, the fire alarm designer may
propose a category. While it may not be the designer’s job to specify the category, it will expedite the work
if some assistance is given to the owner or manager of the premises. However, in this case, the category
designed should be made clear to the purchaser, so that he can seek professional advice from a fire safety
specialist (if he wishes) regarding the suitability of the category for the premises in question.

5 See page 41 of this guide (LOCATION AND COVERAGE OF DETECTORS)

 Specifiers’ Guide Section 1 Selecting the correct category of system and choosing the system technology

Section 1
L4

L3

L2

L1

Fig 4 Typical areas covered by AFD in L Categories

The examples below show how decisions could be made on choosing the category of fire protection:

• Category M. An office building divided into a number of rooms with people occupying most rooms, or
circulating between rooms regularly. If a fire were to start someone would quickly be aware of it and
could operate a manual call point (MCP). With alarms throughout the building, all the other occupants
would quickly be warned and they would then be able to evacuate.

• Category L4. A larger single or multi-storey building with more than one escape route, with many
unoccupied areas and good fire resistant construction with self-closing fire doors to limit fire spread.
Optical smoke detectors would be fitted on all escape route corridors and stairways so that if a fire were
to start the occupants would be warned when it reached one of the escape routes. Even if an escape
route became blocked by smoke, people would be able to escape by an alternative route.

• Category L3. A building with a number of unoccupied rooms, or a number of different tenants and
perhaps with only one escape route (e.g. one stairway). In this case, with fire detectors (not necessarily
smoke detectors) in all rooms opening onto the escape route(s), a fire would be detected and the 19

sounders would operate before smoke reached the escape routes, enabling people to evacuate, possibly
walking past the door to the room of fire origin. This would offer a better standard of detection than
category L4, because a fire would quickly be detected, in the room of origin, rather than waiting until
smoke reached the smoke detectors in the escape route.

• Category L2. A similar building to the category L3 above but with higher risk rooms such as boiler rooms,
plant rooms, kitchens etc. These rooms would be separately listed and protected with appropriate fire
detectors, even if they did not open onto an escape route. This category is usually appropriate for hotels.

• Category L1. Sleeping accommodation such as a hotel or care home, where the intention is to protect the
lives of all the occupants. Smoke detectors in all bedrooms and day rooms would protect the occupants,
so that if a fire started they would be woken by the alarms and could evacuate, or be assisted to evacuate.
Non-bedrooms (where no-one would be asleep) would be protected with a detector to suit the risk, such
as a heat detector in a kitchen.
 Specifiers’ Guide Section 1 Selecting the correct category of system and choosing the system technology

• Category P1. A valuable building or valuable contents, of any size or design, that the insurers or owners
would not want extensively damaged. In this case category P1 would be appropriate. With detectors in
all rooms and areas where a fire could start, it would be detected within a few minutes. The fire & rescue
service (F&RS) could then be contacted either by on-site personnel (e.g. security), or by using a radio
Section 1

and/or telephone land line link to an Alarm Receiving Centre (ARC). The ARC then contacts the F&RS.
Sounders would only be required to alert staff. The likelihood is that the building would be occupied at
some time of the day, in which case the appropriate category would be M/P1.

For a small unmanned building, only occasionally visited by one or perhaps two building engineers, MCPs
would not normally be necessary, unless their working procedures were to disable the AFD whilst working
to avoid false alarms and if a fire started, or if the engineers accidentally started a fire, to operate an MCP
and so send a signal immediately to the ARC.

Prescriptive categories. All the above categories L4, L3, L2, L1, P1 may be regarded as prescriptive categories
because the recommendations for each are clearly specified in BS 5839-1. For strict accuracy, the high
risk parts of category L2 would be listed separately and therefore those parts could be regarded as non-
prescriptive.

Non-prescriptive categories. The categories P2 and L5 shown below may be regarded as non-prescriptive
categories because there is no unique specification for the areas that need to be protected, these are
selected on the basis of risk.

• Category P2. A partly valuable building or contents where the insurers or owner identify certain areas for
protection. Normally there would be fire resistant construction separating the protected area(s).

• Simple category L5. A building that would not need AFD, other than for just one area, such as a boiler
room, where a heat detector would be installed to operate the sounders, allowing people to escape
before the fire threatened their escape route. This one detector, with sounders throughout the building
would constitute a simple category L5 system. Normally the fire alarm system designer would be able to
take on the responsibility for this decision and proceed with the design.

• Complex category L5. A more complex building for which someone has carried out a full fire risk
assessment could be the subject of a fire engineered solution, taking into account constructional and
procedural details. A category L5 system could be part of this solution with the extent of protection, siting
of detectors and plans to specify the system given to the fire alarm system designer. It can be seen that
this work would require considerable skill and time involving many fire safety disciplines. Information on
the areas to be protected should always be given to the fire alarm system designer.

• Categories X/M. BS 5839-1 specifies MCPs as prescriptive recommendations in categories L1, L2, L3, L4,
20
but if, for categories P1, P2 or L5, MCPs were required, then the category becomes M/P1, M/P2 or M/L5.
 Specifiers’ Guide Section 1 Selecting the correct category of system and choosing the system technology

System technology
Fire detection systems are available as non-addressable (also known as “conventional” systems) or as
analogue addressable. There is no definitive size or type of building that would be suitable for analogue
addressable rather than non-addressable systems. Often the benefits of the analogue addressable will

Section 1
bias the choice towards this technology, but for buildings with only a few zones a non-addressable system
would meet the recommendations of BS 5839-1 and provide perfectly adequate fire detection.

Non-addressable systems
Non-addressable detectors and MCPs are usually connected in parallel across a two-wire radial circuit
and terminated at an end-of-line device (EOL). The EOL provides monitoring of open and short circuit
faults at the fire alarm control panel. Each detector is a two-state device capable of indicating one of two
states - normal (non-fire), or alarm (fire). Each circuit would constitute a zone of the system6 and would
therefore be wired so as to connect the detectors in the area covered by that zone7. The control panel can
then indicate if there is an open or short circuit fault on any zone, or if a fire has been detected or an MCP
operated in a zone. Non-addressable systems are widely available and provide an economic and perfectly
satisfactory means of fire detection in many small to medium-sized buildings (Fig 5a).

ZONE 1

ZONE 2

ZONE 3

21

Conventional Manual Addressable End of Line

Detector Call Point Sounder device

Fig 5a Circuit arrangement on a conventional system

6 See page 24 of this guide (SYSTEM ZONE CONFIGURATION)

7 See page 28 of this guide
 Specifiers’ Guide Section 1 Selecting the correct category of system and choosing the system technology

ZONE 1
Section 1

ZONE 2

ZONE 3

Analogue Addressable Addressable Manual Addressable

Detector Call Point Sounder

Fig 5b Circuit arrangement on an analogue addressable system

Analogue addressable systems

In analogue addressable systems, detector circuits act as data communication highways that allow detectors
to communicate much more information about their status to the control panel, using digital signalling
techniques (Fig 5b). In contrast with a conventional system, the control panel can display the address,
showing which detector or MCP has been activated and a description of its exact location, level, room etc.,
and also whether more than one device has activated on the same circuit. Multiple detector activations and
MCP operations during the spread of a fire could then be observed and dealt with more effectively. The
control panel for analogue addressable systems usually has an on-board printer so that past events can be
printed out and appropriate action taken, and if necessary any false alarms dealt with.

Because an addressable system provides the exact location of any fire that is automatically detected, it
enables more rapid action to be taken to deal with the fire and ensure the occupants in this vicinity are
evacuated. This can for, example, be important in a care home, in which some time that is needed to assist
residents with evacuation can be lost in searching for a fire if a conventional system is installed.

The analogue signal represents the level of the phenomenon sensed (e.g. heat, smoke, etc) at that detector
22 and this can be processed and used to select day/night settings, for example to make detectors more
sensitive out of working hours to give faster protection for property (while maintaining EN 54 compliance).
Some designs of analogue addressable systems can sense the difference between a very slow build up of
contamination and a fast changing signal due to a genuine fire. This also helps to reduce false alarms.

Service and maintenance can also be improved using analogue addressable systems, because any
contamination in the detector can be displayed or printed out and the detector replaced if necessary, to
ensure correct operation and reduce false alarms.

Analogue addressable systems are more sophisticated than conventional systems and in the past they
tended to be used only in large buildings. However, they are now in more widespread use and can often be
as economically viable in small to medium-sized buildings as conventional systems.

In order to maximise the potential benefits in terms of false alarm reduction, BS 5839-1 recommends that
systems incorporating a large number of detectors (more than 100 is given as an example) should be of the
analogue addressable type.
 Specifiers’ Guide Section 1 Selecting the correct category of system and choosing the system technology

Intelligent systems
Fire alarm control panels, in conjunction with analogue addressable detectors can have a certain amount of
intelligence. As an example, this can be used for drift compensation, so that if a detector slowly becomes
affected by contamination, the control panel adjusts for this, within certain limits, and the resultant

Section 1
sensitivity to fire remains unaffected. This can also be read via the control panel for service levels.

Some intelligence, such as drift compensation, can be built into the detector rather than the control panel,
the terminology for this being “distributed intelligence” (see Fig 6). Sensitivity modes can also be set
from the control panel at commissioning, to select the most suitable speed of detection, according to the
application in the area, and avoid false alarms. The Apollo Discovery range of detectors includes distributed
intelligence with high sensitivity mode 1 to lower sensitivity mode 5, all modes being within the relevant
EN 54 standard.

COMPENSATION IN ACTION

Alarm threshold

Raw analogue value

Compensated analogue value

COMPENSATION LIMIT
Limit of compensation

Alarm threshold

Raw analogue value

Compensated analogue value

23

NORMALISATION

Alarm threshold

Raw analogue value

Compensated analogue value

Fig 6 Drift diagrams

 Specifiers’ Guide Section 1 System zone configuration

1.3 SYSTEM ZONE CONFIGURATION

Zoning is necessary in order to provide suitable indication of the area of the building in which a detector
Section 1

has operated or a manual call point has been activated. Such zones are referred to as ‘detection zones’.
However, particularly in buildings in which evacuation is phased, or it is intended to operate a two stage
alarm arrangement, the term ‘zone’ may also be used to describe an area in which either the ‘Evacuate’
signal or the ‘Alert’ signal is given8. These zones are referred to as ‘alarm zones’.

Detection zones
The process of configuring the detection zones involves sub-dividing the building into a number of
separate areas which are sufficiently small and which relate to the layout of the building so as to provide an
unambiguous indication of the location of the fire. A zone plan should be installed near the control panel
so that the zonal indication can be related to the location in the building. The zone plan should show the
basic building layout, the zones, all the escape routes, circulation areas, and all entrances. The intention
is to assist those responding to a fire, such as the F&RS by reducing the search time to locate a fire and
helping to identify the source of a fire quickly. It should be noted that BS 58 39-1 recommends that this plan
should always be provided. A simple list of zones does not comply with the British Standards.

Detection zones should be determined on the following basis9:

• Zone floor area is 2000sq m maximum (see Figure 7a)

• Except in the case of small buildings (those with a total floor area of 300sq m or less),
no zone should cover more than one storey.

For the avoidance of doubt, multi-storey zones are not permitted (except as above), even if each floor is
substantially less than 2000sq m, see Figure 7b.

ZON
E7

ZON
24
E4
ZON
ZON E5 ZO
E1 NE
ZON 8
ZON E6
E2
ZON
Each storey, for example 4500sq m would
need to be split into 3 zones
E3

Fig 7a Large storeys need to be split into zones

8 See page 54 of this guide (AUDIBLE AND VISUAL ALARM REQUIREMENTS)

9 The criteria for determining zoning are given in clause 13 and 14 of BS 5839-1.
 Specifiers’ Guide Section 1 System zone configuration

Section 1
ZON
E3

ZON
E2 ZON
E4

ZON
E1

ZON
E5

ZON
E4
ZON ZON
E3 E6 25

ZON
E2
ZON
E1

Fig 7b Zones are restricted to one storey, even if substantially less than 2000sq m
 Specifiers’ Guide Section 1 System zone configuration

• The ‘search distance’, the distance that needs to be travelled after entering the zone in order to
determine visually where the fire is, should not exceed 60m for non-addressable systems (Fig 7c).
Section 1

60m MAX

Fig 7c Search distance for a fire should be no more than 60m in a zone

• Any enclosed flue or chimney-like structures, such as stairways and lift shafts, should be separate
zones, see Figure 7d.

Lift shaft
(ZONE 7)

Hoist
(ZONE 8)
ZON
E5

ZON
E4
ZON ZO
E3 NE
26
6
ZON
E2
ZON
E1

Each storey is less than 2000sq m

Fig 7d Stairways and lift shafts are separate zones

 Specifiers’ Guide Section 1 System zone configuration

• In multi-storey buildings the MCPs near the storey exit should display fire on the associated storey
zone. This applies whether the MCP is sited within the stairway landing or in the accommodation area
(e.g. corridor or office).

• Where an MCP is sited in a stairway near a final exit to open air at the bottom of the stairs, then that

Section 1
MCP may either be incorporated in the stairway zone (see Figure 7e), or it can be incorporated within
the zone that serves the floor on which the exit is located (e.g. the ground floor selection zone).

ZO
NE
5

Z4

Z3
ZON
E4 Z4

Z5 ZON ZO
Z2 E3 NE
6

ZON
Z1 E2 Z3 Z2

ZON
E1
Z6
MCPs located on the stairway, configured
or wired to show at the control panel in the
associated storey zone
Z1

Fig 7e MCPs on stairways show fires in related zones

Similar zoning criteria apply to systems comprising only manual call points (category M), but for an open
27
area such as a large warehouse, someone may see a fire and travel some distance before operating an MCP.
In this case the 2000sq m zone size limitation can be increased to 10,000sq m.

If the warehouse also has AFD then it would be necessary to assign 2000sq m areas for the AFD,
superimposed over the MCP zone. For example, for a 10,000sq m warehouse, one zone could be assigned
to the MCPs and five zones assigned to the AFD, making 6 zones total.
 Specifiers’ Guide Section 1 System zone configuration

Z1 Z1 Z1
Section 1

AFD AFD AFD AFD AFD

66m Z2 Z3 Z4 Z5 Z6

Z1 Z1 Z1

150m

Fig 8 AFD zones superimposed on an MCP zone

These zoning criteria apply to both analogue addressable and non-addressable (conventional) systems,
the 60m search distance being the exception that applies only to non addressable systems.

Alarm zones
Alarm zones may be required in a building with:

• a staged alarm (e.g. where areas threatened by fire would have the full evacuate signal, but areas
remote from the fire may have an alert pulsing signal, often one second on, one second off); or
• a phased evacuation system where, for example, the building is evacuated typically two floors at a time,
starting with the floor of activation, and the floor immediately above.

The operation of alarm zones would be documented in a cause and effect sequence, programmed into
the control panel and fully tested by the commissioning engineer.

The boundaries of alarm zones should comprise fire resisting construction (unless they are external walls).
While an alarm zone can incorporate more than one detection zone, a detection zone cannot cover more
than one alarm zone. Boundaries of alarm zones should coincide with boundaries of relevant detection
zones (Figs 7a, 7b and 7d). There is no area or size limitation for alarm zones given in BS 5839-1.
28
 Specifiers’ Guide Section 1 System zone configuration

Section 1
Detection Detection Detection
ZONE 1 ZONE 2 ZONE 3

Fig 9a Alarm zone containing 3 detection zones

All in one detection zone

29

Fig 9b Alarm zone containing just one detection zone

 Specifiers’ Guide Section 1 System zone configuration

Non-addressable zones versus analogue addressable zones

The time taken to locate a fire would clearly be reduced if its exact location were known. An advantage of
the analogue system is that the exact location is displayed at the control panel, whereas in a conventional
system, the indication only narrows the location down to “somewhere in that zone”, which could be as
Section 1

large as 2000sq m (or within the 60m search distance). The 60m maximum search distance recommended
applies to non-addressable systems, but for analogue addressable systems, provided that the text
description of the location of the activation is clearly displayed and can be readily interpreted by those
unfamiliar with the building without any manual operation of controls on the control panel, the search
distance criterion does not apply.

In some buildings, with complex search routes, the resulting non-addressable zones may have to be
much less than 2000sq m, because of the limitation imposed by the 60m search distance. In the same
building with an analogue addressable system, the zones could be the full 2000sq m area.

The importance of establishing the zoning of a system at an early stage in the design cannot be over
emphasised. With a conventional non-addressable system, each zone corresponds to a separate circuit.
Re-configuring the zoning at a later stage can therefore result in costly wiring changes.

Detection zoning on an analogue addressable system is a function of software, and a single circuit,
usually in the form of a loop with detectors, MCPs etc., will often have several zones. Changes at a later
stage in a project can therefore be more readily accommodated. However, alarm zones sometimes
comprise discrete circuits of sounders and re-configuring of these zones can also involve costly wiring
changes. Note that different criteria govern circuit design in analogue addressable systems10.

Visual indicators on detectors

The visual indicators on detectors should be orientated to be readily seen by a person entering the room
where the detector is located. This may be less important where detectors have two indicators.

Remote visual indicators

Remote visual indicators can be particularly useful for showing if a fire is located in a void or a locked
room. Wiring should be of the same fire resistant type as the detection circuit. However, the wiring to
remote visual indicators is not normally monitored and it is good working practice to limit the cable
run from detector to indicator to just a few metres. The operation of the indicator should be checked
annually with the associated detector.

BS 5839-1 does not actually state in any recommendation that remote indicators should be fitted, but it
30
does mention in the commentary that they would be advisable, especially for non-addressable systems.
It is therefore good working practice to allow for remote indicators, where necessary, and it is worth
reading the job specifications carefully because some state that for voids remote visual indicators be
fitted as part of the contract.

10 See page 65 of this guide (CIRCUIT DESIGN)

 Specifiers’ Guide Section 1 System zone configuration

Remote visual
indicator
Corridor
Smoke detector

Section 1
with LED
Same
cable as
detection
circuit

Typical Typical Locked

Room Room Room

Fig 10a Visual and remote indicators

Protected Void

Remote visual
indicator

Protected Area

Fig 10b Remote visual indicator for a void

31
 Specifiers’ Guide Section 1 Selecting the correct detectors

1.4 SELECTING THE CORRECT DETECTORS

Types of fire detector (BS 5839 pt1 Clause 21)
Section 1

Fire detectors are designed to detect one or more of the four key characteristics of fire:
• Smoke
• Heat
• Infra-red or ultra-violet radiation (flames)
• Combustion gas (e.g. carbon monoxide).

The most widely used detectors are those that respond to either smoke or heat.

Flame detectors that look for infra-red, ultra-violet or combined infra-red and ultra-violet radiation tend
to be used for more specialist applications and, in particular, where a fire involving flammable liquids
(such as alcohol) is expected.

Combustion gas detectors, in particular carbon monoxide (CO) fire detectors, respond well to slow,
smouldering fires and may be advantageous when used to protect bedrooms. Slow burning fire such
as most bedding fires produce CO gas as a product of combustion, whereas fast burning flaming fires
produce carbon dioxide (CO2) with the smoke and other products of combustion. CO detectors are
useful in avoiding certain false alarms. For example, if a resident, in a hotel bedroom with en suite
shower room, leaves the shower room door open, a smoke detector may see the shower “steam” as
smoke and be activated, giving rise to a false alarm. But with a CO detector, even if it is fitted near to
the shower room door, a false alarm due to “steam” would be avoided. CO detectors therefore have
applications in areas where smoke detectors would produce false alarms, but where heat detectors
would be too slow to operate.

Multisensor detectors contain more than one sensor within the same detector. They are able to monitor
more than one of the characteristic phenomena of fire (for example, heat and smoke). Use of these
detectors can potentially result in a significant reduction in false alarms.

Detectors can be either ‘point’ type, where the relevant characteristic of fire is detected at a defined
point within the protected area, or ‘line’ type, where the detector senses along a defined line in the
protected area. Most smoke detectors that are used are of the point type, and these are satisfactory
32 for most applications. However, optical beam smoke detectors are available, which are effectively line
type smoke detectors. These can be particularly beneficial and cost effective in certain applications, for
example, large open spaces with high ceilings (e.g. warehouses) or where access for maintenance of
point detectors would present difficulties.

Aspirating smoke detectors are also available, in which samples of air are drawn, by a pump or fan,
through holes in pipework within the protected area, to a central detector. The sensor within the
detector usually operates by means of optical principles. These often utilise sensors of very high
sensitivity, and are thus particularly beneficial in the clean environment of a computer room, where any
smoke is diluted by the rapid air movement within the room. However, they are also used:
 Specifiers’ Guide Section 1 Selecting the correct detectors

• For reasons of aesthetics, the pipework could be concealed above a ceiling with capillary tubes to small
sampling points through which samples of air are drawn, this being the only visible part.
• Where access for maintenance may be difficult, such as for high ceilings or in some voids. Pipework and
sample holes once installed do not always require attention during routine servicing and can often be

Section 1
cleaned using vacuum techniques without needing access to the pipework in the protected area.
• Where the temperature inside the room is too low for point detectors, for example in cold stores,
where the aspirating detector can be located outside the room.

Heat detectors can also take the form of either point type or line type detectors and are designed to either
respond when a fixed temperature is reached (fixed temperature heat detectors) or when the rate of
change of temperature within the room is abnormal (rate-of-rise heat detectors11). In the case of the latter
the heat detector should still respond when the fixed temperature limit is reached.

Choosing the right detector

No one type of detector will be suitable for all applications. The choice of detector will largely be
governed by three key considerations:

• The nature of the fire hazard.

• The speed of response required.
• The need to minimise false alarms.

Other factors such as cost, suitability for the working environment and ease of maintenance will also need
to be considered.

Smoke detectors
Smoke detectors respond quickly to most fires and generally much faster than heat detectors. They are
usually the first choice in most applications, unless false alarms would be a problem or the fire load would
not give off smoke (for example in an alcohol bottling plant).

Smoke detectors work on one of two principles:

1. Ionisation smoke detectors. The sensing part of the detectors consists of two chambers - an open,
outer chamber and a semi-sealed reference chamber within. Mounted in the reference chamber is a
low activity radioactive foil of Americium 241, which enables current to flow between the inner and
outer chambers when the detector is powered up. As smoke enters the detector, ions become attached
to the particles, causing a reduction in current flow in the outer chamber and hence an increase in
33
voltage measured at the junction between the two chambers. The voltage increase is monitored by the
electronic circuitry, which triggers the detector into the alarm state at a pre-set threshold. An externally
visible red LED lights up when the detector changes to alarm state.

Ionisation chamber smoke detectors respond well to small particles. These are usually found in the smoke
from fast burning fires. Use these detectors particularly where rapid, open flaming fires are likely. However,
although modern ionisation detectors have an extremely low level of active material, they are subject to the
Ionising Radiation Regulations, which control storage, transport and disposal. Consequently they tend to
be used for protecting particular risks, where other detectors would not be suitable.

Integrating ionisation smoke detectors work on the same principle as the ionisation smoke detector,
but has modified signal processing circuitry, which allows an alarm threshold to be present for up to
20 seconds without initiating an alarm. This type of detector is suitable for use in areas where transient
high levels of smoke – like pollutants – may be expected.

11 Rate-of-rise heat detectors also respond when a fixed temperature is reached

 Specifiers’ Guide Section 1 Selecting the correct detectors

10 volts on
foil holder

Radioactive
foil
Section 1

Reference
chamber

Sensing
electrodes

Sensing
chamber

0 volts
In clean air With smoke

Positively Negatively Smoke

charged charged particle
ion ion

Fig 11 Ionisation smoke chamber

2. Optical detectors utilise a small light source and a light sensor in a chamber that excludes external light.
In clean air conditions, the light sensor is arranged to receive little or no light from the source. If smoke
particles enter the chamber the light is scattered, some falls on the sensor and the signal from this is
processed to sense fire.

However, smoke detectors are also more likely to give rise to false alarms than other types of detector.
They may not therefore be suitable in kitchens or workshops where steam, cooking fumes or dust
would be in the normal environment.

Although ionisation chamber and optical smoke detectors have a wide range of response, there are
differences in sensitivity between the two types. This may make one type more suitable than the other
for your application.
34
Optical smoke detectors respond well to visible smoke. If you can see the smoke, the optical detector
will detect it. Optically dense smoke is formed with large particles, such as smoke that has aged
before it reaches the detector. An example of aged smoke would be where a fire has started in a room
opening onto an escape route and the smoke has travelled a long way, or “aged” in passing through
any gap around the door, on its way to the smoke detector in the escape route. In the ageing process,
the smaller particles of smoke merge together to make larger more visible particles. For this reason BS
5839-1 recommends the use of optical smoke detection for escape route corridors and stairways. It is
important to remember that most fires are mixed and produce both large and small particles of smoke
and an optical smoke detector is most likely to detect the fire.
 Specifiers’ Guide Section 1 Selecting the correct detectors

Photo-diode

Section 1
Infra-red LED

In clean air With smoke

Light Diffused Smoke

beam light beam particle

Fig 12 Optical smoke chamber

It should be understood that whether a smoke detector takes two seconds or two minutes to detect
a fire, it is not normally critical in the initial stages of a fire, because fire doors and fire resistant
construction should contain the fire long enough for people to escape to a place of safety.

Heat detectors
Heat detectors are generally less sensitive to most fires than other types of fire detector, although they
may detect the heat from certain clean burning fires. For fires involving flammable liquids such as alcohol
(which produce very little smoke), a heat detector would operate before a smoke or CO detector.

In a category L3 system, although smoke detectors in rooms opening onto escape routes would give the
earliest response, BS 5839-1 permits the use of heat detectors. The objective is to give warning, while the
fire is still restricted to the room of origin, before the escape route becomes impassable for the other
occupants. In these situations, a heat detector has been found to respond fast enough12.

Heat detectors are unlikely to respond to smouldering fires. This lack of sensitivity to certain fires will
35
make them unsuitable for situations where warning of the presence of smoke is required or where
unacceptable damage would result from a small fire. However, heat detectors can often withstand
environmental conditions that would adversely affect other types of detector or give rise to false alarms.
They also require little maintenance.

Rate of rise heat detectors would be used in, for example, a kitchenette with microwave, kettle and
toaster, where smoke detectors may give rise to false alarms.

Fixed temperature heat detectors would be used in a kitchen or boiler room, where smoke detectors and
also rate of rise detectors would give rise to false alarms.

Line type heat detectors can be used with analogue addressable and conventional systems, using suitable
interfaces for the fixed temperature and rate of rise technologies.

12 Heat detectors may not be suitable for all adjoining rooms. Smoke detectors should always be provided if the
room is a dormitory with several people sleeping or is a bedroom used by disabled people.
 Specifiers’ Guide Section 1 Selecting the correct detectors

Dual thermistor heat detection

Normal conditions Rate-of-rise response Fixed temperature response

Section 1

Thermistor Thermistor Fire detected on fast increase Fire detected on slow increase
partially exposed of ambient temperature of ambient temperature
sealed from to air
surrounding
air

Note: Analogue addressable heat detectors use only a single thermistor

Fig 13a Dual Thermistor Heat Detection

Address buttons Thermistor bead Case moulding

PCB Lid moulding

LED LED

Heat shrink sleeving

Fig 13b Single Thermistor

36
 Specifiers’ Guide Section 1 Selecting the correct detectors

Flame detectors
Unlike smoke and heat detectors, flame detectors do not rely on convection currents to transport the fire
products to the detector. Flame detectors are designed to sense the radiation emitted by flames during
combustion. They provide a very fast response to fires involving flammable liquids or gases, and will often

Section 1
be the most suitable type in these applications. They can be used to protect very large open areas without
reducing the speed of response, where other types of detector would not be effective. To ensure that full
coverage is maintained, flame detectors require a clear line of sight within the area to be protected.

Height

Length Width

Fig 14 Flame detectors require clear line of sight

Carbon monoxide (CO) fire detectors

CO fire detectors are immune to many of the environmental influences that would cause false alarms
37
with smoke detectors, such as dust, steam and cigarette smoke. They respond well to smouldering fires,
where there is incomplete combustion, but they will not respond early enough to fast, free-burning
flaming fires, which have sufficient oxygen for complete combustion, producing CO2 (rather than CO).

Care is needed to ensure that CO fire detectors are suitable for their intended application.
 Specifiers’ Guide Section 1 Selecting the correct detectors

CO

Filter
Section 1

Sensing electrode

Wick
(contains electrolyte)

Counter electrode

In clean air With CO

Sensing 2CO + 2H20 2CO2 + 4H+ + 4e-

Counter 4H+ + 4e- + O2 2H20

Fig 15 Carbon monoxide fire detectors

Multisensor detectors
The combination of different principles of detection in a multisensor detector means that the overall
performance of the detector in responding to different fires is enhanced. It also introduces the potential
to significantly reduce false alarms. A combined optical/heat multisensor detector has been found to
respond well not only to smouldering fires, but also to the fast burning fires favoured by ionisation
smoke detectors. At the same time, it is possible to reduce the sensitivity of the detector to the sources of
false alarms, such as steam, that normally affect optical smoke detectors

Photo-diode

38

Infra-red LED Thermistor exposed to air

Fig 16 Multisensor detector

 Specifiers’ Guide Section 1 Selecting the correct detectors

General detector choices

Whether a detector will respond quickly enough depends on whether the response will satisfy the fire
safety objective for the system. Smoke detectors must be used in escape routes for life safety systems
so as to provide sufficient early warning; heat detectors would be unsuitable for this purpose. Smoke

Section 1
detectors are preferred for property protection, but this will depend on whether false alarms that might
result from the normal environmental conditions in the protected area can be avoided.

In selecting the detector, the designer must take account of the environment to which the detector will
be exposed and the potential for false alarms. There are many ways of managing the false alarm issue
and these are discussed later13. However, the most fundamental step the designer can take is to make the
correct choice of detector.

The relative sensitivity for Apollo detectors is shown in figure 17.

OPTICAL
IONISATION

MULTISENSOR
SENSITIVITY

CO

HEAT

OVERHEATING SMOULDERING GLOWING FLAMING - EARLY FLAMING - LATE

Fig 17 Relative detector sensitivity

An example of a growing fire is shown by the orange shaded portion and lines show the possible 39
sensitivity of the various detectors.

13 See page 76 of this guide (FALSE ALARMS)

 Specifiers’ Guide Section 1 Selecting the correct detectors

Relative performance of detectors in test fires

Optical Multisensor Ionisation CO/Heat CO Heat

Section 1

Overheating/thermal
Very good Very good Poor Very poor Very poor Very poor
decomposition
Smouldering/glowing Moderate/
Very good Very good Excellent Excellent Very poor
combustion good

Flaming combustion Good Good Very good Moderate Poor Poor

Flaming with high Moderate/

Good Very good Very good Very good Poor
heat output good
Flaming clean Moderate/ Moderate/ Moderate/
Very poor Poor Very poor
burning good good good

40
 Specifiers’ Guide Section 1 Location and coverage of detectors

1.5 LOCATION AND COVERAGE

OF DETECTORS

Section 1
Heat and smoke from a fire will rise and collect at the highest point in the room. It is there that heat
and smoke detectors should be sited. However, as smoke and hot gases rise from the fire, they become
diluted with cool, clean air. As a result, the size of fire required to operate smoke and heat detectors
increases as the height of the ceiling above the fire increases. The use of more sensitive detectors can,
to some extent, counter this effect, but it is still necessary to limit the height at which detectors can be
installed. Optical beam smoke detectors are less sensitive to this effect than point detectors, and, as a
result, they can be used at greater heights.

BS 5839-1 gives the maximum height for mounting detectors14 as:

Heat detector, rate of rise 9m (13.5m)

Heat detector, fixed temperature 7.5m (12m)

Point smoke detector 10.5m (15m)

Point CO fire detector 10.5m (15m)

Optical beam smoke detector 25m (40m)

Figures in brackets apply to category P systems if there is a “rapid response” fire brigade attendance time
of not more than five minutes from being called.

It is possible, particularly in rooms with high ceilings, for smoke to cool to such an extent that it stops
rising before reaching the ceiling. This is known as stratification. The smoke may form a layer a few
metres below the ceiling and, as a result, ceiling mounted detectors may not operate. It is difficult, in
practice, to predict where stratification will occur, because it may be caused by time related effects such
as the sun shining on the roof, or overhead heaters causing a build up of heat. Eventually, as the fire
grows, the smoke will have sufficient buoyancy to overcome and break through any stratification to reach
the ceiling and thereby operate the ceiling mounted detectors. As a result, it is always necessary to mount
detectors on ceilings, even if additional detectors are also mounted below.

Alternative techniques for consideration could be to use a number of vertical aspirating pipes in addition 41
to horizontal pipes near the ceiling, or point detectors on the ceiling.

When locating detectors, the designer should take account of the following:

• Detectors should be positioned at the highest part of the protected area, though additional detectors
may be lower than this (see earlier comments on stratification).
• Detectors should not be sited more than 150mm below the ceiling in the case of heat detectors
(Fig 19a), and no more than 600mm in the case of smoke detectors (Fig 19b).

14 Limit on the heights of ceilings are given in Tables 3 and 4 of BS 5839-1.

 Specifiers’ Guide Section 1 Location and coverage of detectors

<25mm X
Section 1

>150mm X

Thermistor

HEAT DETECTOR

Fig 19a Heat detectors must not be sited more than 150mm below the ceiling

<25mm X
>600mm X

Optical chamber

SMOKE DETECTOR

Fig 19b Smoke detectors must not be sited more than 600mm below the ceiling

• For vertical chimney-like or flue like shafts (e.g. lift shafts or open risers) a detector should be located
in the shaft, at the top, and also, in the accommodation area within about 1.5m, at each level where
there is a door or access hatch to the shaft (e.g. lift doors) (see Fig 20).

42

LIFT LIFT
SHAFT SHAFT

1.5m

Fig 20 Detectors should be located within 1.5m of vertical shafts

 Specifiers’ Guide Section 1 Location and coverage of detectors

• Detectors should not be mounted closer than 500mm to any walls or ceiling beams (over 250mm deep)
(Fig 21).
• If a partition or storage rack reaches to within 300mm of the ceiling, treat it like a wall dividing the
room in two and provide a detector each side (Fig 21).

Section 1
BEAM
X <300mm X >10%
Ceiling
<500mm <500mm height
PARTITION

Fig 21 Detectors should be kept away from partitions and beams

• Keep detectors at least 1m away from any air vents blowing air into a room (Fig 22).

<1m >1m

AIR VENT/
AIR CON

43
Fig 22 Detectors should not be within 1m of air vents

• Where the air circulation is through a perforated ceiling, an area equal to a circle of 600mm radius
around each detector should be blanked off to make it imperforate.

• Keep a clear space of 500mm around and below each detector (cupboards and small spaces would not
need a variation if this cannot be achieved).
• Ceiling obstructions, such as structural beams, deeper than 10% of the ceiling height should be treated
as walls (Fig 21).
 Specifiers’ Guide Section 1 Location and coverage of detectors

• Where there is a honeycomb (coffered or cellular) ceiling or closely-spaced structural beams15 (Fig 23),
BS 5839-1 gives particular guidance and spacings may have to be reduced to compensate for the cells
or coffers obstructing the movement of smoke and heat. Common dimensions for the coffer structures
is 600mm square with a downstand of 300mm deep. For these dimensions, with a floor to ceiling height
Section 1

into the coffer of 3m or more, then detectors should be placed on the downstand at normal spacings
for the detector as discussed previously.

Width Depth

Height from floor

or likely height of
fire to ceiling slab

Fig 23 Siting detectors on beams

• Site detectors away from shallow beams, ductwork or other isolated attachments, such as light fittings,
that might interrupt the flow of smoke to the detector – keep them clear by at least a distance of twice
the depth of the obstruction (Fig 24). If the obstruction is more than 250mm deep, keep the detector at
least 500mm away.

44 LIGHT FITTING
OBSTRUCTION

D
250mm
(max)

X
<2D

Fig 24 Never mount devices closer than twice the depth of fittings or obstructions

15 The spacing and siting of detector on honeycomb type ceilings and where
there are closely-spaced beams are given in Table 1 and 2 of BS 5839-1.
 Specifiers’ Guide Section 1 Location and coverage of detectors

• Where in an area that needs provision of detectors for the category of system in question, there is
a suspended ceiling with a void above, the voids should be provided with detectors unless they are
shallow (less than 800mm deep) and will not give rise to extensive fire spread before detection occurs.
See section 1.2 for more information on voids.

Section 1
• Detectors should normally be mounted on the ceiling.
• Wall mounting of detectors is only permitted in rooms opening onto escape routes for category L3
systems. If these detectors (heat, CO or smoke) are wall mounted, then they should be near the door,
above the level of the top of the door to the escape route and within a band between 150mm to 300mm
from the ceiling.
However, it is preferable to mount the detector on the ceiling where it will operate with the sensitivity
intended, keep clean and dust free for maximum time, and also be suitable for upgrade to category L1
at a later date by simply upgrading if it is a heat detector to a smoke detector.

Unlike smoke and heat detectors, flame detectors do not rely on convection to transport the fire
products to the detector. Instead, they ‘see’ the radiation emitted from the flames. There is therefore no
need to mount them on ceilings, and they are not affected by stratification. However, their sensitivity to a
fire does decrease as the distance from the fire increases.

Carbon monoxide fire detectors are less affected by stratification, as the carbon monoxide reaches the
detector not only a means of convection, but also by diffusion. However, the ability of the detector to
detect a fire depends on the extent to which the carbon monoxide has been diluted before reaching the
detector. Therefore, in a large space it may be advisable to use a combination of detection technologies
to be sure of quickly detecting a fire.

Detector spacings
To take account of these effects, standards for fire detection and alarm systems incorporate limitations
on the spacing and siting of detectors16. For example, the horizontal distance from any point in the room
to the nearest detector should not exceed 7.5m for smoke detectors and 5.3m for heat detectors (Figs 25a
and 25b). This can be regarded as a “radius of detection” around each detector.

7.5m
radius

45

5.3m
from wall

10.6m
spacing

5.3m
from wall

Fig 25a Siting of smoke detectors with 7.5m radius

16 Siting and spacing of detectors is covered by clause 10 of BS 5839-1.

 Specifiers’ Guide Section 1 Location and coverage of detectors

5.3m
radius
Section 1

3.8m
from wall

7.5m
spacing

3.8m
from wall

Fig 25b Siting of heat detectors with 5.3m radius

BS 5839-1 offers a convenient concession for escape routes up to 2m wide, in which smoke detectors
may be sited no more than 15m apart and no more than 7.5m from the end wall (derived from the 7.5m
radius), see Figure 26. For corridors that are not escape routes and where heat detectors are to be used,
BS 5839-1 allows spacings for heat detectors up to 10.6m apart and no more than 5.3m from the end wall
(derived from the 5.3m radius).

7.5m
radius

7.5m 15m 7.5m

46

Fig 26 Siting of smoke detectors in escape routes up to 2m wide

These figures are based on open areas under a flat horizontal ceiling. If the ceiling is sloped or is
sub-divided by beams or other obstructions to the flow of smoke and heat, different spacings will apply.
 Specifiers’ Guide Section 1 Location and coverage of detectors

Important note.
Multisensor detectors can be treated like smoke detectors, and the 7.5m radius used but, if you intend
to use them in ‘heat only’ mode, you must reduce to the 5.3m radius, as for heat detectors.

Section 1
For a sloping ceiling or roof, the radius of detection may be increased by 1% per degree of slope up to a
maximum of 25%. For example, if the roof slopes by 10 degrees, the radius of detection for detectors may
be increased by 10%. For smoke detectors, this equates to 1.1 times the 7.5m radius making 8.25m. The same
increase for sloping roofs can be applied for heat detectors, based on the 5.3m radius of detection.

With a sloping ceiling or pitched roof, detectors should always be located near the apex, the highest point.
However, BS 5839-1 allows smoke detectors to be mounted up to 600mm down from even a flat ceiling, so
if the ceiling is only gently sloping it may not be necessary to locate them exactly at the apex (Figure 27a).
For example, Figure 27c shows for a 5 degree roof smoke detectors could be as much as 6.86m from the true
apex and still qualify for the extra 5% on the radius of detection. Remember that, to protect a pitched roof,
smoke detectors should be within 600mm of the apex (150mm for heat detectors), so always start near the
apex and, if necessary, add detectors towards the eaves, as in Figure 27d.

<600mm

47

Fig 27a Siting of detectors in roof rise of 600mm or less

Figures 27a to 27d show smoke detector sitings with a symmetrical roof, but for an unsymmetrical or
one-sided (lean-to) roof the same principles apply.
 Specifiers’ Guide Section 1 Location and coverage of detectors
Section 1

600mm

>600mm

Fig 27b Siting of detectors in roof rise of more than 600mm

48

600mm
Shallow angle
e.g. 5 degrees

Fig 27c Siting of detectors in shallow angled roof

 Specifiers’ Guide Section 1 Location and coverage of detectors

Section 1
600mm 600mm

Detector(s) should be sited within

600mm of the apex, so in this example
a 20m wide roof without detector(s)
near the apex is not protected

Fig 27d Siting of detectors within 600mm of apex

The area coverage for sloping roofs also increases in proportion to the roof angle up to the maximum of 25
degrees. Figure 27e shows the square areas for a smoke detector with the coverage superimposed for various
roof angles. In Figure 27e, the dimensions of the square are calculated by adding to the basic dimension of
10.6m for smoke detectors the percentage increase allowed because of the roof angle for apex detectors.
Thus for a 10 degree roof angle the increase is 10% and the square becomes 10.6m + 10% = 11.66m.

14
13 25˚ Roof
13.25m X 13.25m
12
=176sq m
11
10 20˚ Roof
9 12.72m X 12.72m
8 =162sq m
METRES

7 10˚ Roof
6 11.66m X 11.66m 49
5 =136sq m
4
Flat Ceiling
3 10.6m X 10.6m
2 =112sq m
1

1 2 3 4 5 6 7 8 9 10 11 12 13 14
METRES
Fig 27e Affect of roof angles on smoke detector coverage

Different spacing applies to detectors sited in environments with rapid air movement such as computer
rooms17. Here it is usually necessary to reduce the area covered by each detector to compensate for
dilution of the smoke.

17 This is covered in BS 6266:2002, Code of practice for fire protection for electronic equipment installations
 Specifiers’ Guide Section 1 Location and coverage of detectors

BS 5839-1 states that the same recommendations for the siting and spacing of smoke detectors should
apply to CO detectors. Guidance on the application of CO detectors is also given in a Fire Industry
Association (FIA) document, which is reproduced from page 96 of this document.

Flame detection
Section 1

Flame detectors must have a clear line of sight across the area being protected. However, there is no
specific guidance in BS 5839-1 on their coverage. The designer is therefore recommended to consult the
manufacturer, and Apollo can provide guidance to designers on the coverage of their detectors.

Duct detection
Where smoke detectors are to be installed in ventilation ducts (air extraction systems), it is essential that
the manufacturer’s guidance be followed. Optical smoke detectors are not sensitive to air movement, so
use them rather than ionisation detectors. The optical detector is normally used in combination with an
inlet probe, which is inserted in the air path within the duct. The probe should cover at least two thirds
of the widest dimension of the duct, and to avoid turbulence, should not be placed too close to bends,
corners or junctions. Duct probes should be located at a distance of at least three times the widest
dimension of the duct away from the bend.

The probe works by pressure differences due to the airflow in the duct, so gravity is irrelevant and the
probe can be mounted on any face of the duct (top, sides or bottom). A remote indicator is not required
in BS 5839-1, but it would be good working practice to provide one, unless the detector indicator is
directly visible through the clear cover of the detector probe housing.

Optical beam smoke detection

The area covered by any one optical beam smoke detector should not exceed 2000sq m.

The width of cover is 7.5m either side of the beam, making a total width covered of 15m, see Figure 27f.
There are two types of optical beam smoke detector:

a) Separate transmitter and receiver, where the transmitter generates a beam of infra-red light to be
sensed by the receiver; and
b) Combined transmitter and receiver with a reflector, where the beam of light is sent to the reflector and
back to the receiver.

50
7.5m

7.5m

Fig 27f A single optical beam smoke detector covering 15m

In both types, obscuration of the beam of light by more than about 25% would be detected as a fire,
and the area and width of cover are the same as given above. As already mentioned, BS 5839-1 limits
the area for a single beam detector to 2000sq m, but does not limit the beam length, leaving that to the
manufacturer to specify. Most manufacturers specify a maximum beam length of about 100m.
 Specifiers’ Guide Section 1 Location and coverage of detectors

Optical beam smoke detectors, should be installed within 600mm of the ceiling, and when used near the
apex of a pitched roof, the width of cover increases in proportion to the roof angle, up to a maximum
of 25%, see the table below. If the manufacturer declares a length of cover of, for example 120m, then
care should be taken to observe the area limit, because when using the increased cover at the apex, the

Section 1
2000sq m could be inadvertently exceeded, see Table 28 below.

As with other detectors, when covering a pitched roof, remember that an optical beam smoke detector
should be within 600mm of the apex. If this one detector does not cover the whole width of the roof,
detectors towards the eaves would be required.

Optical beam smoke detector mounted within 600mm of the ceiling or the apex of a pitched roof

Based on length of cover, Based on length of cover.

Width each Total width
Roof Angle L=100m, area covered = L L=120m, area covered = L
side of cover = W
xW xW

Flat 7.5m 15m 1500sq m 1800sq m

5 degrees 7.87m 15.75m 1575sq m 1890sq m

10 degrees 8.25m 16.5m 1650sq m 1980sq m

15 degrees 8.62m 17.25m 1725sq m *2070sq m

20 degrees 9m 18m 1800sq m *2160sq m

25 degrees 9.37m 18.75m 1875sq m *2250sq m

*A variation would need to be agreed, or more beam detectors should be used, where the area exceeds 2000sq m.

Table 28 showing coverage for optical beam smoke detectors

51
 Specifiers’ Guide Section 1 Manual Call Points (MCPs)

1.6 MANUAL CALL POINTS (MCPs)

Section 1

Manual call points are an important part of any fire detection and alarm system. They allow people in the
building to raise the alarm, and people are sometimes better placed to notice a fire than automatic detectors.

MCPs should be sited so that someone discovering a fire can always operate a call point before leaving the
building. This means that manual call points should be sited at all exits to open air.

MCPs should ‘stand out’ from the background. Where they will be viewed from the side (e.g. in corridors),
they should not be flush mounted, although they may be semi-recessed provided that they stand out by at
least 15mm.

Although not a specific recommendation of BS 5839-1, it is sensible to mount manual call points such that
the test key can be inserted without obstruction from nearby doorframe or wall.

If installing an MCP on a red wall, to make it a more visible contrast with the background, mounting on
a white flange would be a simple solution.

In buildings with more than one storey, people should be able to operate an MCP before leaving the floor
they are on. This means locating MCPs at the storey exits, either by exits to stairways or on stairway landings
at each level.

In the case of buildings with a staged alarm for phased evacuation, MCPs should not be located on stairway
landings, otherwise the evacuation may start from the wrong floor, because there can be no guarantee that
a person, when travelling down a stairway, will not operate an MCP below the floor on which the fire is
located. For alarm zones, it is essential that MCPs be located by all exits from an alarm zone; this is intended
to avoid someone operating a call point remote from the alarm zone, which could result in evacuation of
the wrong part of the building.

BS 5839-1 recommends that 45m18 should be the maximum distance anyone should have to travel to reach
an MCP 9 (Fig 29a), taking into account the layout of walls, partitions and fittings. If, at the design stage,
the final layout is unknown, the maximum straight line distance between any point in the building and the
nearest MCP should not exceed 30m. These figures are further reduced to 25m and 16m respectively in,
for example, plants handling flammable liquids where there are processes in the area that could result in
52
rapid fire development, or where a significant proportion of occupants have limited mobility and they can
reasonably be expected to be the first to operate an MCP.

45m (max)

Fig 29a Maximum distance to reach an MCP

18 The recommendations on manual call point siting and

spacing are addressed in clause 20 of BS 5839-1.
 Specifiers’ Guide Section 1 Manual Call Points (MCPs)

MCPs should normally be mounted at a height of 1.4m above floor level (Fig 29b), but BS 5839-1 allows
a tolerance of plus or minus 0.2m, so any height from 1.2m to 1.6m could be used without writing a
variation. There may be a good reason to vary this height, in particular for wheelchair users, in which
case, a lower height might be appropriate. Approved Document M (ADM)18, which gives guidance on the

Section 1
requirement of the Building Regulations in England and Wales suggests 0.9m to 1.2m as a suitable height
for switches (ADM does not specifically mention MCPs, but they could be regarded as a type of switch)
for wheelchair users, so 1.2m would seem to be a good compromise, complying with the ADM and
without requiring a variation from BS 5839-1. Remember that not all disabled people use a wheelchair
and if they use a frame or sticks, an MCP mounted too low may be out of reach for them.

1.4m
(+/- 200mm)

Fig 29b MCPs should be 1.4m above floor level

It is preferable for all MCPs in a building to be of the same make and type. If this is not the case, it is
nevertheless essential that the method of operation is the same, so if a system has mainly modern MCPs,
53
but there are still some of the old “hammer and chain” types, then they should be updated.

BS 5839-1 states that if one or more detectors are unplugged, then it should not affect the operation of
any MCP. For further information see section 1.10 CIRCUIT DESIGN, page 65.

18a Switches, outlets and controls, section 4.25 – 4.3

 Specifiers’ Guide Section 1 Audible and visual alarm requirements

1.7 AUDIBLE AND VISUAL ALARM

REQUIREMENTS
Section 1

Audible alarm devices

Audible alarm signals need to be provided to warn the occupants of the building that there is a fire.
However, this depends to some extent on the purpose of the system. It may be that not all of the
occupants need to be warned. For example, in health care areas, where patients need assistance, it is only
staff who need to be alerted, and not patients. For further guidance on health care premises in England
and Wales, refer to HTM 05-03. Systems intended for property protection may only need to alert key
personnel in order to ensure that fire fighting action is taken.

Devices used to generate audible alarm signals are usually bells or electronic sounders, connected to
the control panel on dedicated alarm device circuits as discrete devices. However, the sounders can
be incorporated within the base of a detector, either wired on a separate circuit or on the same pair of
wires as the detectors and MCPs. If wired on the same pair of wires on a non-addressable circuit or on an
analogue addressable loop circuit, considerable savings in cable can be made. BS 5839-1 does allow the
use of a 4-core cable, where 2 cores would be for the detection circuit and the other 2 cores would be a
circuit for the discrete alarm devices or sounder bases. This would be more labour intensive to install, but
once installed, it would work perfectly well and comply with BS 5839-1, provided that the detection zone
areas and loop areas etc., previously described, were followed.

Sounders connected onto an analogue addressable loop circuit can often be programmed to operate
individually or as part of a group.

In some buildings, it is more appropriate to use spoken messages to warn the occupants. In these cases,
the audible alarm signal should be given by a voice alarm system. These are specially engineered public
address systems that should meet the recommendations of a different part of BS 5839, namely Part 819.
Voice sounders may be applicable in some situations; these have pre-programmed messages and can be
powered directly from the control panel.

Where it is necessary to alert all occupants of the building regardless of their location, fire detection
standards20 normally specify a minimum sound pressure level that must be achieved. This will depend
54
on whether the people are awake or asleep. BS 5839-1 recommends, for rooms and areas more than
60sq m, a minimum of 65dB(A) or 5dB(A) above any other noise likely to persist for longer than 30
seconds, if this works out to be more than 65dB(A)21 (Fig 30). For example, if the background noise was
63dB(A), then adding 5dB(A) to it to make 68dB(A) gives the sound pressure that should be achieved in
that area. However, a few concessions written into BS 5839-1 allow 60dB(A) for stairways, for small rooms
less than 60sq m in area, and for points of limited extent (small areas probably at the discretion of the
commissioning engineer). All the dB(A) figures apply to the core area of a room, the outer border of 0.5m
around a room being not critical and need not be measured in commissioning.

19 BS 5839-8. Fire detection and alarm systems for buildings - Code of practice for the design, installation and servicing of voice alarm systems.
20 This is covered by clause 16 of BS 5839-1.
21 The noise created by running water in bathrooms and shower rooms can be ignored.
 Specifiers’ Guide Section 1 Audible and visual alarm requirements

Section 1
65dB(A) +5dB(A)
@ 500Hz to 1,000Hz @ 500Hz to 1,000Hz

NO WITH
BACKGROUND NOISE BACKGROUND NOISE

Fig 30 Minimum sound levels for areas more than 60sq m

Where people sleep, the sound level should be at least 75dB(A) at the bedhead if the system is intended
to rouse them from sleep (Fig 31).

75dB(A)
55

Fig 31 75dB(A) is required to rouse people from sleep

 Specifiers’ Guide Section 1 Audible and visual alarm requirements

The audibility of the fire alarm signal in an area will depend upon the output from the sounder, the
distance from the sounder to the person, the extent to which doors and any other constructional
features attenuate the sound and the effect of furniture and fittings in absorbing sound. If these factors
are difficult to determine at the design stage, alarm sound pressure levels can be checked, in the
Section 1

relevant areas, prior to commissioning and modified as necessary. This should ensure the system passes
audibility tests at commissioning.

However, experience has shown that there are two tips that a designer should consider:

In rooms where people sleep, provide a sounder in every bedroom, as 75dB(A) is unlikely to be
achieved unless there is a sounder within the room.

If there is more than one door between any point and the nearest sounder, 65dB(A) is unlikely to be
achieved. Where this is the case, allow for extra sounders.

The concessions discussed of 60dB(A), points of limited extent and measuring the core area only,
probably means that the commissioning engineer has some flexibility about the acceptability of figures.
However it does not mean that the designer should assume this and all calculations should strive to
achieve the full dB(A) as appropriate for the area.

The frequency range of sounders is usually specified in standards. BS 5839-1 recommends that the
frequency should lie between 500Hz and 1000Hz, unless the frequency range of background noise would
mask these frequencies.

On large sites with many buildings or in the case of a large building with many entrances it may be
appropriate for a sounder or visual alarm to be provided externally. The purpose of this sounder is two-fold:

• to draw attention of passers-by to the fact that the alarm system has been activated so that they will call
the fire and rescue service; and
• to assist the fire and rescue service by guiding them to the premises.

However, to minimise noise pollution, such sounders should be silenced automatically after 30 minutes
unless the premises are manned continuously and the occupants can silence the alarm manually. An
alternative method would be to use an external visual alarm, which causes no noise nuisance, but is still
apparent to the fire and rescue service, so the timer would not be required.

In the case of a staged alarm it will be necessary for the audible alarm signal to be different for each
stage of alarm. Typically there will be two stages – ‘Alert’ and ‘Evacuate’. The evacuate signal is usually a
56
continuous sounding of the bells or sounders, whereas the alert signal is intermittent.

Make sure that sounders are not positioned immediately by the control panel, because it distracts the
fire warden or fire and rescue service when trying to assess the situation.

Someone will need to telephone the fire and rescue service, so they will need a reasonably quiet area
to make the emergency call, or alternatively you could consider a timed silencing push button for that
area (auto timed to return to normal after, for example, 2 minutes).

Electronic sounders and bells should not normally be used on the same system. It is important that the
sound characteristics of fire alarm sounders within the same building are similar. Where voice alarm
systems or voice sounders are used in part of the building and conventional sounders used elsewhere,
the attention drawing signal that precedes voice announcements should be the same as the sound from
the fire alarm sounders.
 Specifiers’ Guide Section 1 Audible and visual alarm requirements

Calculating sound pressure level

Most audible alarm devices are characterised with sound pressure output in dB(A) specified at 1m from
the device and the manufacturer will usually publish a tolerance of ±3dB(A). Often an adjustment facility
is provided so that the output can be reduced if required. For example, a sounder could be rated with a

Section 1
sound output of 100dB(A) ±3dB(A) at 1m. This means that at 1m from the device you should be able to
measure at least 97dB(A), but it might be as high as 103dB(A).

In calculations, always assume the worst and work on the minimum, usually -3dB(A).

Also, it is advisable to subtract a further -3dBsafety factor, to allow for unknown sound attenuation.

The sound pressure from an audible alarm reduces with distance away from the device, so as you move
away the sound pressure becomes less by 6dB every time you double the distance. Starting from the
1m and doubling it to 2m results in a loss of 6dB and then doubling the 2m to 4m results in another 6dB
totalling 12dB. Doubling again to 8m and the loss becomes 18dB.

Calculation example

Using the 100dB(A) ±3dB(A) sounder, the sound 8m away would be:

100 -3(manufacturing tolerance) -3(safety factor) -18(for the 8m distance) = 76dB(A)

With a good degree of confidence, we can therefore say that we should achieve at least 76dB(A), 8m from
the sounder, but it might be more if the manufacturing tolerance is up and the safety factor was unnecessary.
We would not be concerned if the measured sound pressure was a little higher than required, but we would
not want it to be less, because the system may be “snagged” at commissioning and it could cause delays and
expense to correct the situation at a late stage in a project.

Tables are shown below and on the next page with the loss shown at increasing distance from the sounder.

dB loss for a sounder rated at 1m

Distance 1m 2m 3m 4m 5m 6m 7m

dB 0dB -6dB -9.5dB -12dB -13.9dB -15.5dB -16.9dB

Distance 8m 9m 10m 11m 12m 13m 14m

57
dB -18dB -19dB -20dB -20.8dB -21.5dB -22.2dB -22.9dB

Distance 15m 16m 17m 18m 19m 20m 25m

dB -23.5dB -24dB -24.6dB -25.1dB -25.5dB -26dB -27.9dB

Distance 30m 35m 40m 50m 60m 80m 100m

dB -29.5dB -30.8dB -32dB 33.9dB 35.5dB -38dB -40dB

Table 32a FIA sound pressure dB(A) table – 1m

 Specifiers’ Guide Section 1 Audible and visual alarm requirements

dB loss for a sounder rated at 3m

Distance 3m 4m 5m 6m 7m 8m 9m
Section 1

dB 0dB -2.4dB -4.4dB -6dB -7.3dB -8.5dB -9.5dB

Distance 10m 11m 12m 13m 14m 15m 16m

dB -10.4dB -11.2dB -12dB -12.7dB -13.3dB -13.9dB -14.5dB

Distance 17m 18m 19m 20m 25m 30m 35m

dB -15dB -15.5dB -16dB -16.4dB -18.4dB -20dB -21.3dB

Distance 40m 50m 60m 80m 100m 150m 200m

dB -22.4dB -24.4dB -26dB -28.5dB -30.4dB -33.9dB -36.4dB

Distance 250m 300m

dB -38.4dB -40dB

Table 32b FIA sound pressure dB(A) table – 3m

Table 32a above shows that increase in distance from 2m to 6m results in a loss of about 10dB, but the
increase in distance from 16m to 20m results in a loss of only 2dB. Consequently, it is much less critical
to do accurate distance measurements above about 20m; a measurement to the nearest 2m would be
acceptable. Thus we could say at 30m to 34m the loss would be 30dB.

The table suits any sounder, so for a 90dB(A) ±3dB(A) at 1m sounder base, the sound at 7m would be:

90 -3(manufacturing tolerance) -3(safety factor) -17(for the 7m distance) = 67dB(A)

(the 16.9dB was approximated to 17dB because it would be impossible to notice a difference of 0.1dB(A))

58
 Specifiers’ Guide Section 1 Audible and visual alarm requirements

Adding two equal sounds:

To calculate the audible sum of two sounders in an area giving equal sounds at a point, add 3dB(A) to the
sound pressure from one of the sounders at that point.

Section 1
To calculate how this would apply, consider a corridor with detectors and sounder bases producing
90dB(A) at 1m positioned 7.5m from each end and 15m apart.

90dB(A) ± 3dB(A) 90dB(A) ± 3dB(A)

7.5m 7.5m 7.5m 7.5m

Fig 33 Adding two equal sounds

The sound pressure at each end would be: 90-3-3-18 = 66dB(A)

The sound pressure in the centre between the sounder bases would be:

90-3-3-18(for 7.5m, assume it is 8m)+3(for adding two equal sounds) = 69dB(A)

Notice the positioning of the sounder bases very conveniently coincide with detector spacings in the
same 2m wide corridor as discussed previously.

Distances to achieve target dB(A) levels

Sounder rating Target Target Target

(at 1 metre) 85dB(A) 74dB(A) 65dB(A)

85dB(A) - 1.5m 5m

90dB(A) - 3m 9m

95dB(A) 2m 5m 16m

100dB(A) 3m 9m 30m
59

105dB(A) 5m 16m 50m

Note that the 3dB(A) manufacturing tolerance and 3dB(A) safety factor have already been allowed
for in this table

Table 34 Distances to achieve target dB(A) levels

 Specifiers’ Guide Section 1 Audible and visual alarm requirements

Sound loss through a door:

Sound loses a substantial amount through a door. A standard door would normally reduce sound
pressure by about 20dB(A) and a half hour fire door, being heavier, would cause about 30dB(A) of loss (Fig
35). To calculate the effect of this, consider a 90dB(A) sounder base just outside a room with a 30dB(A) fire
Section 1

door, the sounder being 1m from the door.

-20dB(A) -30dB(A)

Fire door

Fig 35 Decibel loss through doors

Just outside the door, 1m from the sounder the sound pressure will be 90dB(A) (it is rated at 1m).

Just inside the door, the sound pressure will be 30dB(A) less = 60dB(A).

We can see that the sound pressure is not sufficient, because as soon we move into the room by, for
example 2m, the sound pressure would be less than 60dB(A), which is the minimum for a room less than
60sq m. A bedroom would need 75dB(A) near the bedhead, so it would be deficient by a considerable
margin.

Remember, when calculating dB(A) all doors should be shut, and the same applies for measurements at
commissioning.

Calculating sound pressure in a bedroom:

60
The recommended sound pressure to awake someone is 75dB(A).

Referring to Table 34 (page 59) an 85dB(A) sounder within 1.5m of the bedhead would be sufficient.

Alternatively, a 90dB(A) sounder could be 3m from the bedhead and still achieve the 75dB(A).

The distances used in calculations are the horizontal distance as seen on a plan view. In reality the
distances would be lengthened because of the diagonal. However, for a standard ceiling height of about
3m, sound reflections give a benefit and experience has shown that the diagonals can normally be
neglected. If in doubt do sample measurements using a dB(A) meter in a typical room.

Sounders in high warehouses:

If the warehouse ceiling is, for example, 10m high, the sound loss will be 18 to 20dB(A). It is therefore
advisable to install sounders, about 4m high, on the wall or on pillars to achieve acceptable sound coverage.
 Specifiers’ Guide Section 1 Audible and visual alarm requirements

Visual alarm devices

Where ambient noise levels exceed 90dB(A) visual alarm devices should be provided. These are necessary
to supplement the sounders, for example in a noisy plant room or where hearing protection is likely to
be used in normal circumstances. Notice, that visual alarms should be supplementary to audible alarm

Section 1
devices, normally 65dB(A).

Visual alarm devices may also be necessary in areas occupied by deaf or hard of hearing people. The
Equality Act is intended to ensure no disabled person is placed under undue risk. Consequently in some
areas of buildings, where people may be deaf or hard of hearing, such as bedrooms for disabled people,
visual alarms will be necessary.

Visual alarm devices may be ceiling mounted, or if not on the ceiling, at a minimum height of 2.1m (Fig 36).

Visual alarms are available as standalone devices, combined with sounders, or incorporated into a
detector base possibly with a sounder base. Some visual alarms use LED technology, which is economical
on current consumption.

>2.1m

61

Fig 36 Visual alarms should be mounted at least 2.1m from floor level
 Specifiers’ Guide Section 1 Control panel equipment

1.8 CONTROL PANEL EQUIPMENT

Section 1

The control panel, referred to in BS 5839-1 as the control and indicating equipment (CIE), performs three
principal functions:

• Gives indication of fire and fault signals and their location.

• Allows manual control of the system so that operators can take actions such as testing, disablement of
detectors, activation of audible fire signals, silencing and resetting the system.
• Automatically monitors and controls both the detectors and other interfaces controlling field
equipment (such as sounders) and their circuits, and the power supplies to the system.

In small, simple systems, the control panel will normally be a single enclosure complete with its power
supply and battery. However, in some large systems, the control equipment can be distributed - for
example, there may be a main control panel in one location, a repeat panel in a separate location, and a
separate power supply and battery unit serving both. In very large systems, there may be several control
panels networked together.

Requirements for the design of control panel, and the functions and indications that they must provide,
are included in product standards22. These include the time taken for the control equipment to respond
to: the operation of a detector; to an MCP; to the presence of a fault; or the removal of a device such as a
detector from its circuit.

CIEs should indicate the zone in which the detector has operated usually by means of LEDs or similar
devices. BS 5839-1 recommends that, on or adjacent to the control panel, there should be a diagrammatic
representation or zone plan of the building, which should at least show the building’s entrances,
circulation areas and escape routes. It should also show how the building has been divided into zones.
The alternative would be to provide an illuminated mimic panel, or use a display on a VDU, but note that
this should have a back-up supply to operate in mains failure condition. A VDU also needs a back up
display. Note that, in each case, a diagram is necessary. A simple list of zones is not adequate.

In general, control panels should be sited on the ground floor of the building and in the immediate
vicinity of the entrance most likely to be used by the fire and rescue service. Where there are a number
of entrances that might be used, repeat panels may need to be provided.
62
The area in which a control panel is sited should be of low fire risk, reception or the entrance usually
being suitable, so it would be bad practice if the control panel or its power supply were sited in a plant
room or kitchen. To provide some protection for the control panel and power supply, detector(s) should
normally be in the same room, so that if the system is connected to an ARC, at least the fire signal will be
transmitted before the control panel or power supply is affected by fire.

22 The product standard for control panels is BS EN 54-2

 Specifiers’ Guide Section 1 Power supply requirements

1.9 POWER SUPPLY REQUIREMENTS

Section 1
The normal power supply for a fire detection and alarm system is derived from the low voltage mains
supply to the building. The supply should be reliable and be capable of supplying the largest load
that would be placed on it under normal, fire and fault conditions. This can be checked in practice
by disconnecting the standby battery, simulating a fire or evacuate situation and checking that the
power supply output holds up the load at about nominal voltage and the alarms operate correctly. The
Standards23 state that warning systems, such as fire alarm systems, should have a back-up or standby
supply to support the system in the event of a mains failure. Standby supplies normally take the form
of rechargeable batteries. The automatic transfer between the mains supply and the battery would be
designed into the control panel and should not affect the operation of the fire detection alarm system.

Standby times
The period for which the standby power supply should maintain the system in operation is given in
BS 5839-1. This period depends upon whether the system is for life protection or property protection
(categories L or P) and the level of supervision in the building.

Standby times for Category P systems

If the system is supervised (i.e. the panel is checked at intervals not exceeding 6 hours), or a mains failure
fault is signalled to an ARC, the standby battery for category P systems should be capable of maintaining
the system in operation for at least 24 hours, after which time there should still be sufficient capacity to
provide a full evacuation signal for 30 minutes. In the case of unsupervised systems, where the building is
at times unoccupied and mains failure is not monitored at an alarm receiving centre, the standby supply
should be sufficient to provide protection for a period of at least the unoccupied period plus 24 hours (up
to a maximum of 72 hours in total) and, after that time, still be capable of giving the evacuation signal for
30 minutes. Note a power failure signal is a fault signal and would be in addition to the fire signal, so two
channels would be required in the transmitter. If the transmitter is separately powered, then it should have
the same standby back-up capability as the fire alarm control panel, usually 24h.

Standby times for Category L systems

In the case of category L systems the standby battery should be capable of maintaining the system in
operation for at least 24 hours. Again, there should then still be sufficient capacity to sound the alarm for
63
30 minutes.

An emergency generator may be used as a partial replacement for standby batteries in a category
L system, provided that the generator starts automatically on failure of the normal supply and it is
inspected monthly and run for 1h and any fuel oil and water replaced as necessary. For these conditions,
the standby duration may be reduced to no less than 6h. If these conditions are not achievable, then the
normal standby of 24h minimum applies. This (6h) relaxation does not apply to a system intended for
property protection (category P), nor is it normally adopted in hospitals.

23 Power supplies are covered in clause 25 of BS 5839-1.

 Specifiers’ Guide Section 1 Power supply requirements

Calculating required battery capacity

The following formula can be used to calculate the capacity of standby batteries in amp-hours (A-h).
This formula incorporates de-rating factors to take account of both the ageing of the batteries and the
performance of a battery when higher current is drawn under alarm conditions.
Section 1

C = 1.25 x ( [ Tq x Iq ] + [ D x Ia x Ta ] )

Where

C = battery capacity in A-h (amp-hours)

1.25 = the de-rating factor to allow for the ageing of the battery over its 4 year life
Tq = required standby period in hours (usually 24h for category L systems)
Iq = quiescent current in amps, supplied by the battery with the mains failed (non-fire condition, no
alarms)
D = de-rating factor (typically 1.75)
Ia = alarm load in amps, supplied by the battery with the mains failed in fire conditions
Ta = alarm time, which is always 0.5h

Inserting D = 1.75 and Ta = 0.5h, and then placing the 1.25 at the end for simplicity, the formula becomes:
C = ( [ Tq x Iq ] + [ Ia x 0.875 ] ) x 1.25

Thus, the required battery capacity can be calculated, by:

a) deciding on the standby time, Tq (usually 24hr);

b) predicting and/or measuring Iq and Ia;
c) using the formula given immediately above; and
d) using a simple calculator to arrive at the battery capacity, C in A-h. The battery chosen should be no
less than this figure to achieve the duty required.

Example: Tq = 24h; Iq = 0.1A; Ia = 1.8A.

C = ( [24hx0.1A] +1.6A ) x 1.25 = 5A-h. A battery capacity of 6A-h or 7A-h would suffice.

64
 Specifiers’ Guide Section 1 Circuit design

1.10 CIRCUIT DESIGN

Section 1
It is important to recognise the potential for faults on circuits to prevent the system from detecting a
fire and giving an alarm signal. It is also necessary to consider the effect that work (e.g. extensions or
modifications, repairs or routine attention) will have on the ability of the system to operate when required.

System design should therefore take into account the extent to which faults or impairments of the
system due to works may affect the system’s ability to perform as required in the event of fire. As a result,
standards24 normally make recommendations on circuit design and the extent to which there should be
monitoring for faults.

Detection circuits
In the case of non-addressable systems, while a fault on a radial detector circuit might prevent the
operation of the other detectors on the same zone circuit, no other zone circuits will be affected. More
detectors can be connected to a single circuit in an analogue addressable system, so it is important that
the effect of faults will be no worse. It therefore means that any fault should not remove protection from
an area greater than the maximum area that would normally be covered by one zone. BS 5839-1 therefore
states that, in the case of a single fault, protection should be lost from an area no greater than 2000sq m or
from more than one floor of a building25.

As wiring in analogue addressable systems is usually run in a loop, an open-circuit fault usually has no
effect on the operation of the detectors, because it is usually possible to communicate with detectors
in both directions around the loop. However, where detection radial circuits are used on an analogue
addressable system, the area limit is 2000q m as given above.

Two simultaneous faults should not remove protection from an area greater than 10,000sq m. This
effectively is the maximum area that a single loop should cover.

Isolators in analogue addressable systems operate by sensing the overload current caused by a short
circuit and then switching off.

Isolators are usually employed to limit the extent to which short-circuit faults can affect detector or call
point circuits on an analogue addressable system.

This has the effect of isolating the devices (detectors and MCPs etc.) between the two nearest isolators
on the section of the circuit where the fault has occurred. 65

Detectors and MCPs disabled by the operation of short circuit protection should again be limited in
extent to no more than 2000sq m or one floor of a building. For addressable systems only, BS 5839-1
allows a concession in the positioning of short circuit isolators, stating that 5 devices (MCPs, detectors
etc) above a floor and 5 devices below a floor may be included between two isolators, provided that
the 2000sq m area is not exceeded. It is advisable that designers do not use all of these 5 devices, but
leave most of them for the installer to use, if necessary, to cover a void or other feature not known to the
designer.

The more isolators you have on a circuit, the better the integrity will be. The ultimate integrity, if
required, would be an isolator in every device so that a single short circuit would not affect the operation
of any device.

24 Circuit design parameters are given in clause 12 of BS 5839-1

25 Up to 5 devices on the floor above and below can also be disabled at the same time.
 Specifiers’ Guide Section 1 Circuit design

It is important that the removal of a detector from a circuit should not result in the disablement of any
manual call point. This would not be an issue for analogue addressable systems, but for non-addressable
systems the installer can check this is working correctly by removing all the detectors on a circuit and
operating the MCP furthest from the control panel and checking that the alarms operate as intended.
Section 1

If not, remedial work to check wiring polarity in each detector will have to be done. Alternatively, the
Apollo Orbis range offers easy checks for circuit polarity to speed up this checking.

Notes

Spare loop capacity

It is good working practice to leave some spare capacity (10% spare would be reasonable) on each loop,
if possible, to allow for future expansion. Consider a typical loop with a capacity of 126 devices, then
10% would be about 12 devices, so making the initial utilisation of about 114 devices on the loop and
leaving space if, for example, a room is split into two or a ceiling void is created.

Apollo loop calculator

When designing the capacity of loops, it is quite time consuming to calculate whether a loop with a
certain length cable run, and certain loading, would operate correctly. In the calculation, the number
of detectors, MCPs and alarm devices on the loop would have to be taken into account, with all alarms
operating and with a single open circuit fault at the most heavily loaded end of the loop.

The Apollo loop calculator, available from www.apollo-fire.co.uk/loopcalc, does all this in a few seconds.
You just have to enter the manufacturer of the fire alarm control panel, the quantity of each device
(including isolators), the length of the loop, the distance to the first device and the cable size. The loop
calculator has all the current consumption figures for each device built in and it then instantly calculates
the maximum cable run allowed and the voltage drop, also displaying the maximum voltage drop
allowed. This can be displayed and printed out for the documentation records.

Detection Circuits

Zones Non-addressable Analogue addressable

Zone Area 2000sq m 2000sq m

66 No limit if first activation shows with

Search distance within a zone 60m maximum no manual intervention and text
display is clear and explicit

Faults Non-addressable Analogue addressable

Single open circuit 2000 sq m 2000sq m

Single short circuit 2000 sq m 2000sq m

Two faults on a circuit Not relevant for radial circuits 10,000sq m

Not relevant due to inherent

Detector removal No MCP affected
addressable operation.
 Specifiers’ Guide Section 1 Circuit design

Upgrading a non-addressable system

It is possible to connect a circuit of conventional detectors to an analogue addressable control panel.
This can be particularly useful when an existing system is being upgraded by the replacement of the
control equipment only, and it is the intention to re-use the existing radial circuits and non-addressable

Section 1
detectors and MCPs. Special interface units with zone circuits and EOL units with isolators between each
zone for fault protection as shown in Figure 37a must be provided to connect the circuits to a loop of the
analogue addressable system. The loop could be quite short, with the interfaces installed close to the
control panel.

5
1 to 5 are non-addressable zone circuits with detectors and MCPs, terminating in an EOL

Analogue
Short circuit End of Line Interface
addressable
isolator device Unit
detector

Fig 37a Special interface units and EOL units connecting circuits to an analogue addressable system loop

67
Another way of upgrading a system, to retain the original radial circuits, but replacing all the non-
addressable detectors and MCPs with new analogue addressable devices, would be as shown in Figure
37b. To preserve the integrity, each radial circuit would need to be protected with a short circuit isolator
either side. Integrity would not be as good as a true loop, in which a single open circuit would not lose
any communication, but the integrity would be at least as good as the radial non-addressable system
being upgraded.
 Specifiers’ Guide Section 1 Circuit design

1
Section 1

5
1 to 5 are analogue addressable zone circuits with detectors and MCPs

Analogue
Short circuit
addressable
isolator
detector

Fig 37b Method of upgrading a system to retain original radial circuits

In all cases a single open or short circuit fault should affect no more than 2000sq m.

Alarm device circuits

Circuits containing sounders should be designed such that, should there be a failure of the circuit due to
a short circuit condition caused by the fire, a minimum of one alarm sounder will continue to sound. This
can be achieved by installing sounders on two separate circuits.

However, it does not necessarily imply that the two circuits must be run together throughout the
building, with the sounders interspersed on the two circuits. Indeed, the recommendations of the
standard could be met by connecting all but one of the sounders on one circuit, with the remaining
68 sounder connected on the other circuit. The sounder on this latter circuit would normally be located
near the control equipment where its operation would be obvious to those organising the evacuation
and subsequent re-entry of the building (Figure 38a). Normally, in a large building several alarm device
circuits would need to be used, to comply with current ratings or voltage drop limitations (for example,
one circuit per floor), so the requirement for two circuits would easily be met. However, what BS 5839-1 is
stating is that even in a small building two alarm circuits are recommended.

Despite the above, there are circumstances in which two sounder circuits should be provided and
evenly interleaved in an area. This applies to public spaces, such as shopping malls and airport terminals,
where the space is greater than 4000sq m or is designed to hold more than 500 members of the public.
 Specifiers’ Guide Section 1 Circuit design

Sounder circuit 1 with sounder

located to be audible at the control panel

Section 1
Sounder 2 circuit covering the rest of the building

Sounder End of Line device

Fig 38a Circuit arrangement with dedicated sounder circuits

Notes

The two sounder circuits normally provided to satisfy the recommendations of BS 5839-1 should not
be run in the same cable, as failure of the cable could affect the operation of both circuits. However,
this does not preclude running both detector and sounder circuits together in the same cable. This is
established practice where base mounted sounders are used.

If addressable sounders are used on the same circuit as detectors, this does not mean that it is always
necessary to provide two circuits. Provided that the circuit is arranged as a loop and there is sufficient
short-circuit protection to ensure that not all sounders will be isolated by a fault, a single loop could
suffice (Figures 38b and 38c).

It is good practice not to place a sounder too close to the control panel, because the fire warden or
fire fighters may not be able to concentrate clearly on the situation, and may be tempted to silence the
alarms before knowing whether it is a genuine alarm or a false alarm.

69

Sounder protected by isolators,

to be audible at the control panel

Analogue Short
Manual Addressable
addressable circuit
Call Point sounder
detector isolator

Fig 38b Circuit arrangement with addressable sounders

 Specifiers’ Guide Section 1 Circuit design
Section 1

Sounder protected by isolators,

to be audible at the control panel

Analogue Short
Manual Addressable
addressable circuit
Call Point sounder
detector isolator

Fig 38c Circuit arrangement with addressable sounders

70
 Specifiers’ Guide Section 1 Installation and cables

2. INSTALLATION AND CABLES

Section 1
Installation
The guidance of the manufacturer must always be followed when installing the components of a fire
detection and alarm system. Good installation practice is essential if the system is to work as designed
and as required by standards. In particular, cables must always be securely fixed at regular intervals
unless run in metal conduit or metal trunking, and fixed with non-combustible clips to the surface.
Plastic tie wraps can be used when cables are run on top of cable trays, but metal ties should always be
used to support cables fixed below trays or on vertical runs. In all cases, consider what would happen
in a fire and ensure that the fire alarm cable would not fall away from the surface and be damaged. For
cables run in plastic mini-trunking, cable clips should be fixed through the trunking to the surface so
that even if the mini-trunking melts and falls away in a fire, the cable will still be held in place. Similarly,
for plastic conduit, metal saddles should be used to support the cable inside in the event of the conduit
melting. Fixing centres for all these clips should be according to the manufacturer’s instructions or, if in
doubt, to the distances given in the IET On-Site Guide to BS 7671.

Variations from BS 5839-1

It is possible too that the installer may encounter circumstances that were not known by the designer of
the system. For example, ceiling features may be present which were not apparent to the designer. This
may result in the need to vary detector siting and spacing. Where this is likely to result in a departure
from the recommendations of BS 5839-1, it is essential that interested parties be consulted and agree
with the changes.

Variations to contract
The designer will try to complete the design according to the information, drawings and any information
gathered from site visits, However, when the installer is working on site, items of which the designer was
not aware, will often become apparent. A few examples could be:

• A new final exit door, added since the design was done, will need an MCP;
• A void or riser, not shown on the site drawings, needs AFD protection; or
• A noisy machine, recently installed, needs a visual alarm or increased dB(A).

These examples require more equipment to be fitted to compensate for the site as it is. There may be 71
cost implications. These items would be the subject of a variation to contract, in which commercial
negotiations would need to be followed through to continue compliance with BS 5839-1.

Cables and wiring

In the event of a fire it is important that all cables used within a fire alarm system are able to withstand
attack by the fire. All cables, including those carrying the mains supply to the control panel, should
be fire resistant. BS 5839-126 recognises that there may be circumstances in which a higher level of fire
resistance is required which only certain types of cable can provide. This is referred to as ‘enhanced’ fire
resistance. It is particularly necessary for tall buildings (over 30m in height) and buildings in which there
is phased or delayed evacuation. However, the enhanced grade is not necessary in these situations if the
building has sprinkler protection.

26 Cables and wiring are covered by clause 26 of BS 5839-1

 Specifiers’ Guide Section 1 Installation and cables

Not only should fire detection and alarm system cables be able to resist fire, they should also be
sufficiently strong to avoid mechanical damage. For some types of cable, this means that they must be
provided with additional mechanical protection. However, this is not necessary in the case of steel wire
armoured (SWA) and mineral insulated (MI) copper sheath cables. All cables for the fire alarm system
Section 1

should have conductor cross sectional areas of no less than 1sq mm.

PVC cables (e.g. twin and earth) are not recommended for any of the fire alarm system circuits, such as
detector, MCP, alarm and network circuits. Fire resistant cables should always be used on new systems,
system upgrades and system additions (even the short length of cable for just one additional detector).
An example of equipment that could be wired in PVC cable would be fail-safe equipment such as to a fire
door retainer electro-magnet, because if the cable is damaged by fire, the magnet will release and the
door will close to the safe position.

Except for SWA and MI, cables should be provided with additional mechanical protection in the
following conditions:

• where they are less than 2m above the floor in non-benign environments, such as warehouses
(see Fig 39);
• where physical damage to the cable is likely, for example in a corridor where wheelchairs or
trolleys pass;
• where cables feed through a floor they should be sleeved for 300mm above the floor (steel conduit
with an end bush or gland would be suitable);
• where rodent attack is likely.

72

2m

Fig 39 All cabling should be mechanically protected up to a height of 2m

 Specifiers’ Guide Section 1 Installation and cables

Suitable means of providing mechanical protection would include enclosing cables in conduit, mini-
trunking or ducts, or alternatively chasing the cable into plaster finishes in walls.

Standards normally require that there is suitable segregation between fire detection and alarm system
cables and cables of other services. The purpose of segregation is to minimise any potential for the other

Section 1
circuits to give rise to malfunction of the fire detection and alarm system as a result of:

- breakdown of the cable insulation of the other circuit and/or the fire alarm system circuits.
- a fire caused by a fault on the other circuit.
- damage resulting from the need for other circuits to be installed in, or removed from, conduit, ducts
or trunking containing the fire detection and alarm system circuits.
- electromagnetic interference to the fire detection and alarm system circuits as a result of the proximity
of another circuit.

Cables should be segregated and separated from other services by means of one of the following:

• installation in conduit, ducting or trunking reserved solely for fire detection and alarm wiring;
• installation in a compartment of trunking separated by a rigid, non-combustible continuous partition
from any other services;
• separating from conductors of other systems (the distance required varies depending on the nature
of the cable and the manufacturer of the system). This separation is not normally critical for cables
bunching together to go through a hole, or crossing at right angles, but separation by, for example
300mm for long parallel runs could be important to avoid interference and cross coupling.

Use of 4-core cable

BS 5839-1 does not generally preclude the use of 4-core cable, but it does state that if a fault occurs to a
4-core cable then it should be assumed that all four cores are affected. A single fault should not affect
more than 2000sq m of AFD protection, so in the case of damage to a 4-core cable, for example all four
cores being shorted together, this 2000sq m limit should be observed.

For example, a spur (or tee) extension on a loop could be considered using a 4-core cable, as shown
in Fig 40. In the figure, a junction box has been fitted to the loop and a 4-core cable taken from that to
another junction box on the spur, where two separate 2-core cable are used in a loop extension. It is
essential that integrity is retained, so the area protected between the isolators “X-Y” as shown does not
exceed 2000sq m.
73
 Specifiers’ Guide Section 1 Installation and cables

Y
4-core cable
Section 1

2-core cable
In this instance 4-core cable
would be acceptable because
the whole area between X and Y
is less than 2000sq m Less than 2000sq m

Detectors Manual Call Point Isolators Junction box

Fig 40 All Spur extension using 4-core cable

For alternative circuit wiring methods, please see the CIRCUIT DESIGN section, paragraph titled
“Upgrading a non-addressable system” (page 67).

However, the following are instances where 4-core is definitely not recommended in BS 5839-1:

a) Where the minimum of two alarm device circuits are used, they should not both be in a 4-core cable,
because a single fault would remove all alarms;
b) Do not use 4-core cables with 2 cores for fire alarms and 2 cores for emergency lighting, always use
dedicated and separate cables for different services; and
c) For public areas, where BS 5839-1 recommends the use of two separate alarm device circuits to provide
alternate or interleaved sounder circuits, do not use 4-core cables, but separate cables for
each circuit.

74
 Specifiers’ Guide Section 1 Communication with the Alarm Receiving Centre

3. COMMUNICATION WITH THE ALARM

RECEIVING CENTRE

Section 1
For the fire detection and alarm system to give the maximum benefit, any alarm should be reported
to the fire and rescue service as soon as possible. In many cases, people in the building will call the
fire and rescue service by telephone. However, this method will not always be sufficiently reliable.
Automatic transmission of fire and fault signals to an ARC can overcome the disadvantages of relying
upon the building occupants to summon assistance. In the case of category P systems BS 5839-1
specifically recommends that there be a means of automatic transmission of alarms unless the premises
are continuously occupied, for example by security personnel. It is also advisable to provide automatic
transmission facilities in care homes (and in small hospitals without a continuously manned switchboard
from where the fire and rescue service will be summoned)

ARCs should have reliable means of informing the fire and rescue service in the area in which the
protected premises are located. They should comply with recommendations of BS 597927.

There are several methods for automatic transmission of fire and fault signals. These include

• digital communicators which use the public switched telephone network

• systems using public or private cellular radio communication
• carrier systems which transmit signals over normal telephone lines (e.g. RedCARE).
Some large sites, such as hospitals, have their own fully manned alarm receiving centre, from where the
fire and rescue service is contacted by telephone.

Equipment for transmitting signals to the alarm receiving centre (ARC) is sometimes incorporated within
the control panel and operates from the same battery as the control panel. Battery capacity calculations
would therefore take into account the loading of the transmitter. Where the transmitter is housed
separately and has an independent power supply, it is essential that the associated battery has the same
standby duration as the control panel.

The connection from the control panel to the transmitter for the ARC should be monitored for open
circuit and short circuit faults, in the same way as any other part of the critical signal path. Any fault in this
connection should display at the main fire alarm control panel. However if the transmitter is in the same
housing as the fire alarm control panel, or, for example, within about 300mm and the interconnecting 75

cable is mechanically protected, then it can be regarded as all in one housing and monitoring would not
be necessary.

Even if there is an automated means of transmission to an ARC, fire procedures for all buildings must
require that, in the event of fire when the building is occupied, people in the building must call the fire
and rescue service without delay. This is just in case the automatic transmission system fails.

27 BS 5979 Code of practice for remote centres receiving signals from security systems.
 Specifiers’ Guide Section 1 False alarms

4. FALSE ALARMS
False alarms cause a great deal of disruption to the normal activities that take place in a building. This can
Section 1

include lost production and lost revenue. More worryingly, they can lead to complacency on the part of
the occupants, who may fail to react correctly when the alarm is genuine.

BS 5839-1 categorises false alarms as follows:

• Equipment false alarms, arising from faulty equipment. With good quality equipment properly installed
and well maintained, these false alarms should be very uncommon.
• Unwanted alarms arising from environmental influences, fire-like phenomena (someone using a
toaster in an office area with smoke detectors), unthinking action by people in the building (making
a lot of dust without regard to smoke detectors) and accidental damage (sweeping up and the broom
handle accidentally breaks the element of an MCP).
• Malicious false alarms (school pupils operating an MCP for some malevolent reason).
• False alarms with good intent (human error, someone smelt burning and genuinely believing there was
a fire, operated an MCP, but was mistaken).

There are many causes of false alarms, these include:

• Steam (particularly from bathrooms and shower rooms).
• Cigarette smoke (now generally banned, but people still surreptitiously smoke in rooms where there
could be a smoke detector and, for example, smoking bedrooms are still allowed in hotels).
• Cooking fumes (leaving the kitchen door open, or not switching on the extractor, or possibly not
cleaning the filter on the extractor).
• Thrips in country areas can cause problems at certain times of year, with some of the older designs of
optical smoke detector. Modern designs are unlikely to be affected by thrips.
• Dust (builders cutting bricks without thinking of the fire detectors).
• Electromagnetic interference (from old equipment not designed to the Electro-Magnetic Compatibility
Regulations).
• High humidity (causing condensation in detectors and other fire alarm equipment).

Manual call points can also give rise to the false alarms, usually as a result of accidental damage, but where
76 this is a problem, a hinged transparent cover can be fitted as protection, with a simple variation agreed.

By choosing the correct type of detector, it is possible to minimise many of these causes. Many others
can be minimised by adopting suitable management practices. The responsibility for limiting false alarms
initially rests with the designer of the system. It is essential that the designer is fully familiar with the use of
the building and with the activities that will take place in it. After handover, users must recognise that the
presence of fire detectors in a building will require a certain discipline on their part if inadvertent operation
of the system it is to be avoided.
 Specifiers’ Guide Section 1 False alarms

The rapidly advancing technology of fire detection and alarm systems offers the user many benefits in this
respect. Analogue addressable systems in particular can incorporate sophisticated filtering techniques to
help them determine whether an alarm is false or genuine. Multisensor fire detectors are also available, and
these can be used to give more reliable detection of fire.

Section 1
Some general guidance for avoiding false alarm in, for example hotel bedrooms, would be to fit smoke
detectors well away from the shower room doors. Convector heaters and uplighters may accumulate
some dust and insects during disuse in the summer then, when first used in the autumn, a puff of smoke
could be given off. It is advisable therefore to position smoke detectors well away from convector heaters
or uplighters.

HOTEL
BEDROOM

CONVECTOR
HEATER
SHOWER
ROOM

Fig 41 Siting of smoke detectors to avoid false alarms

As stated previously, it is not generally advisable to fit a smoke detector near the shower room door in a
77
hotel bedroom. However, if this is unavoidable, a multisensor detector, set to be less sensitive to “steam”,
could be considered to avoid false alarms.

BS 5839-128 also describes technology, which can be used to ‘manage’ a system in such a way that false
alarms are avoided. Special measures to filter out potential false alarms include the use of time-related
systems and a staff alarm arrangement. These are now commonly available in most systems.

28 False alarms are addressed in section 3 of BS 5839-1.

 Specifiers’ Guide Section 1 False alarms

Examples of time related systems include:

• Systems where it is possible to change the alarm thresholds (to decrease sensitivity of the system to
fire) during working hours, (when fumes are present) and automatically restore the system to normal
sensitivity outside of these times.
Section 1

• Multisensor systems in which smoke sensors are isolated during working hours but the heat sensors
remain operational (as mentioned previously, ensure that heat [only] sensors are spaced as for
heat detectors).
• A ‘staff alarm’ arrangement is commonly used in order to avoid unnecessary evacuation of the
occupants of a building arising from a false alarm. With this arrangement, the staff would be warned
that a detector has operated but the general alarm is not sounded at this stage. The staff then have
a pre-set time period in which to investigate and either cancel the alarm or evacuate the building. If,
however, a second detector were to operate, the general alarm would be sounded immediately (this is
called coincidence detection). This would also be the case if someone operated a manual call point or
the time delay period expired.
• Use of smoke detectors in rooms opening onto escape routes, while desirable, may give rise to an
unacceptably high rate of false alarms, if the rooms are used as bedsits or similar. If it is intended to
use smoke detectors under these circumstances, consider systems that can, initially, give only a local
alarm in the room of origin. This would prompt the occupant to take action and clear the cause if it is a
false alarm, but would not evacuate the whole building unnecessarily. In the event of a genuine fire the
system can be programmed to sound all alarms and evacuate after a set delay. For suitable products,
see www.apollo-fire.co.uk/productsearch.

Although all false alarms are undesirable, some will still occur, particularly with systems incorporating a
large number of smoke detectors. However, every effort needs to be made to keep this to a minimum.

The designer, or service engineer, as a Competent Person, should try and help the user reduce false
alarms. Very often the causes are not due to inadequacies in the fire alarm system, but due to building
management and outside contractors working. Recommendations to the user might then include
keeping building management procedures and equipment (door closer springs, extractors etc.) up-
to-date, and encouragement to make contractors produce method statements on how they are going
to avoid false alarms during works. There is no one solution to false alarms, but a gradual and vigilant
process of reduction and hopefully eventual elimination.

During building construction there will be a lot of dust. Apollo provides dust covers with all detectors.
78
These covers should be used during installation until the areas are clean and dust free, or during building
maintenance to protect detectors.

When fire alarm systems are serviced, the service technician should check the false alarm record, as
recorded by the user in the system log book. The service technician can then give suitable advice to the
user if necessary. This means that it is important for the user to record all false alarms in the log book
(see section 6).
 Specifiers’ Guide Section 1 False alarms

Choosing the detector to avoid false alarms

To help avoid false alarms, ensure that the type of detector is suitable for the environment in which it is
to be installed. Consider the following:

Section 1
• Rate-of-rise heat detectors will give rise to false alarms if there are sudden fluctuations in the ambient
temperature (e.g. in kitchens). In these circumstances use a fixed temperature type.
• The temperature near the ceiling in a kitchen or boiler room may often be substantially higher than
at first thought and even fixed temperature heat detectors may give rise to false alarms. To avoid false
alarms it is advisable to choose a fixed heat detector with a threshold about 15oC above the maximum
temperature expected.
• Optical smoke detectors may respond to steam. If it is necessary to site a fire detector close to a
shower room or another source of steam, consider a CO detector.
• Ionisation detectors are very sensitive to fumes from cooking, frying and making toast even if the
fumes are invisible, or to fumes from certain types of painting. They are also more prone to operate
if vehicle fumes are present, and can become unstable if subject to high air movements near the
chamber.
• Optical smoke detectors may be prone to dust from building work, because it looks like smoke and
can leave a permanent deposit in a detector, possibly causing false alarms later. False alarms might
come from cigarette smoke, but with this being mainly illegal now, the problem is much reduced.
Many years ago optical smoke detectors were sensitive to insects such as the tiny thrips or thunder
flies released from fields at harvest time, but modern detectors are designed to be immune to all but
the worst infestations. Optical detectors are not affected by air movement, so they are recommended
for use in duct probe detectors (see SELECTING THE CORRECT DETECTORS page 32).
• All smoke detectors can give rise to false alarms from toasters. In rooms designated for toasters, use
rate of rise heat detectors.

79
 Specifiers’ Guide Section 1 The commissioning process

5. THE COMMISSIONING PROCESS

Once installed, a system needs to be commissioned and set to work before it can be formally handed over
Section 1

to the purchaser. Commissioning is a process that involves thorough testing of the system in accordance
with BS 5839 part 1 Section 5. Every detector, manual call point, alarm device etc., must be activated in order
to establish that the system operates correctly and has been installed in accordance with the intended
category (L1, L3 etc.), to the design specification and recommendations of BS 5839-1. It is also necessary to
confirm that the wiring is satisfactory, and again this will involve testing.

Commissioning of a fire detection and alarm system will involve:

• Checking the insulation resistance of cables, and, in the case of addressable systems, testing to measure
other cable parameters such as loop resistance. This is usually best done during the installation phase
of the project, because if there were to be a problem with excessive cable resistance, the installer would
be on hand to take appropriate action. Also insulation resistance needs to be carried out with devices
disconnected.
• Checking that a suitable mains supply is provided.
• At each detector removing the dust cover and at each MCP fitting the frangible element.
• Testing the operation of every detector, call point, etc, to determine that each device operates correctly
and produces the correct response from the system. With addressable systems, it will be necessary to
confirm that the correct location is displayed on the panel.
• Checking for any obvious cause(s) of false alarms, such as a smoke detector fitted in a room now used as
kitchenette, or hot work room.
• Measuring sound pressure levels to confirm that the alarm audibility is satisfactory with all doors shut
and checking the operation of visual alarms.
• Checking that any link to an alarm receiving centre (ARC) works correctly, noting for the user’s benefit
and engineering uses any relevant unique ARC reference number for the site.
• Proving any site specific software and operational requirements (cause and effect logic, such as time
related sequences and staff alarms). This needs to be tested thoroughly and can take a considerable
amount of time on larger sites.
• Testing any other requirements specified by the designer or purchaser.
80 • Ensuring that any connections to ancillary services such as automatic door releases operate as required.
• Confirming that all control panel functions and controls are fully operational.
• Checking that circuit monitoring for all parts of the system operates correctly. As part of this, it is good
practice to create deliberate open circuit and short circuit fault conditions on each circuit in order to
check that the system responds as intended.
• Installing the correct capacity battery for the duty (please refer to the POWER SUPPLY REQUIREMENTS
section page 63) and marking indelibly with the “installed on” date, so the service engineer can replace
it in 4 years. The designer should have calculated the battery currents in non-fire and fire conditions and
specified the battery capacity required in A-hr, but these should be measured at commissioning.
• For non-addressable systems, testing that with detectors unplugged, MCPs operate correctly.
• For addressable systems, checking the operation of short circuit isolators by introducing short circuit
conditions on the loops.
 Specifiers’ Guide Section 1 The commissioning process

• Checking that any interfaces with EOL units show faults correctly at the control panel when open circuit
or short circuit faults are simulated.
• Finding out the service organisation’s name, telephone number etc., if possible, and displaying the
details near the control panel.

Section 1
• Checking that a suitable zone plain is displayed.

The commissioning process also provides another opportunity to check that the system meets the
recommendations of BS 5839-1 prior to handover to the purchaser. A thorough inspection is necessary.
This should include checking that no changes to the building have taken place since the system was first
designed that would rule out compliance with the standard. For example, a new final exit door added
without the knowledge of the fire alarm designer would require an MCP to be fitted. Any obvious non-
compliances in device siting should be identified.

To ensure that systems incorporating more than 50 detectors do not give rise to unforeseen false alarm
problems, a ‘soak test’ (this is an electrical run test to check for false alarms and general reliability) should
be carried out after commissioning. This should normally last at least a week. The test will be successful
if there are no false alarms during this period or suitable measures are taken to prevent re-occurrence of
any false alarms experienced.

Commissioning should always include a check of the standby battery capacity. The effect of changes
to the design of the system that can often occur during the course of the installation period can be
to render the standby power supply inadequate. Both the quiescent (normal, non-fire) and the alarm
currents should be measured under mains failure conditions. If the values are very close (within about
5%) of the currents calculated by the designer, then the designer’s battery calculations can be used
without further work. If not, the capacity of the battery would have to be recalculated using these values
and the appropriate battery fitted (see the battery capacity calculation method in the Power Supply
section page 63).

Handover
Handover involves acceptance of the system by the purchaser (or a representative of the purchaser such
as a consultant). Before accepting a system, the purchaser should check to see that the installation work
appears satisfactory and that all of the designer, installer and commissioning engineers have provided
the necessary documentation. It is usual for the purchaser to see a demonstration of the system and for
appropriate staff to be given instruction in its operation. Where a consultant or verification engineer is
involved, acceptance testing is likely to be more involved and may even include witness test operation of 81
every device on the system.

The purchaser may have employed a verification engineer to check each stage of the job, for example at
weekly intervals, and if all inspections prove acceptable, a verification certificate would be produced.

Verification can also be carried out as a single, final independent check post commissioning.

In handing over a completed fire detection and alarm system, the purchaser should be provided with the
following documentation:

• ‘As fitted’ drawings, which should show the locations of all items of equipment such as detectors, call
points, end of line units, interface modules, short circuit isolators, control panels and any junction
boxes. They should also indicate the type and size of cable used and the routes followed by the wiring.
This information is essential if future maintenance and alterations are to be carried out effectively.
 Specifiers’ Guide Section 1 The commissioning process

• Operating and maintenance manuals. For some small systems, these may simply take the form of a
set of instructions, but, for most systems, more comprehensive documents are necessary and should
include all site specific details (e.g. cause and effect logic programming and any ARC related unique
reference number (URN) for the site).
Section 1

• A log book, so as to enable a user to keep a record of all routine tests, false alarms and other events on
the system and to allow servicing and maintenance visits to be logged.
• Certificates (stating compliance with the Standard and drawing attention to any variations.):
- Design certificate, showing the category and extent of the fire alarm system, signed by the
design engineer;
- Installation certificate, signed by the installation engineer;
- Commissioning certificate if all is acceptable, signed by the commissioning engineer.

• If for some reason, an essential document, certificate or test is not available or not satisfactory, the
commissioning engineer would not issue the commissioning certificate, but set the system to work
and return later to complete the work and issue the certificate when all aspects are acceptable.
• Acceptance certificate for the purchaser or user with tick boxes for them to check off after a handover,
demonstration, and training for the system
All installation work appears to be satisfactory. The user may want to inspect the basic
workmanship, or may have employed a verification engineer to do this.
The system is capable of giving a fire alarm signal. The user will need a demonstration of MCP
operation and detector operation and possibly cause and effect (if applicable).
The facility for remote transmission of alarms to an Alarm Receiving Centre operates correctly (if
applicable). The user will need a demonstration of how to contact the ARC before weekly tests, to
avoid false calls.

• The following documents have been provided to the purchaser or user as part of the acceptance
and handover:
“As fitted” drawings.
Operating and maintenance instructions, including the ARC site URN.
Certificates of design, installation and commissioning.
A log book.

82
Sufficient representatives of the user have been properly instructed in the use of the system
including (at least): all means of triggering fire signals; daily inspections; weekly and monthly
routine testing; silencing and resetting the system; avoidance of false alarms; and contacting the
ARC before testing.
All relevant tests, defined in the purchasing specification have been witnessed (if applicable).

• Finally the user should be encouraged to keep all the above documentation, as handed over, available
in a file for safe keeping so that, if an engineer has to be called out for emergency repairs system, all
information is readily accessible.
 Specifiers’ Guide Section 1 User responsibilities (the responsible person)

6. USER RESPONSIBILITIES
(THE RESPONSIBLE PERSON)

Section 1
The responsible person29 is delegated to be responsible for the fire alarm system and the fire procedures.
Typically an office manager, building services engineer or security supervisor, this person will need the
necessary authority to authorise work needed to maintain a fully operational system. (This is not the
responsible person to which the Regulatory Reform (Fire Safety) Order makes reference; that “person” is
normally the employer).

For a fire detection and alarm system to be kept in efficient working order, it must be managed
effectively. The key to this is the appointment of a named person for this purpose. This person’s role is to
ensure that the system is tested and maintained, that appropriate records are kept and that any changes
to the building do not adversely affect the fire detection and alarm system. See the SERVICING AND
MAINTENANCE section, (page 86), of this document for the routine checks.

Good practice by the person responsible for the fire detection and alarm systems includes:

• Arranging for weekly testing and periodic services of the system.

• Establishing procedures for dealing with alarms, faults and other events occurring on the system.
• Being aware that dust causes false alarms and insisting that building contractors take steps such as
covering the detectors exposed to the dust with a bag and do their own fire patrols to compensate for
the increased risk. Each night when work stops, the covers or bags to be removed until work restarts.
• Disabling detectors in areas where “clean” work such as steam or fumes could cause false alarms, and
re-enabling detectors when work is complete.
• Being aware that if the work is not as clean as originally thought the disabled detectors may be
inadvertently contaminated with dust and may never work effectively in the future, or may give rise to
false alarms.
• Training staff in the operation of the system.
• Liaising with those responsible for alterations to the building in particular to ensure that any such
changes do not have a detrimental effect on call points, detector coverage and alarm audibility.
• Ensuring that manual call points are kept visible and unobstructed and that a clear space is maintained
around all detectors of at least 500mm. 83

• Ensuring that a record of the up-to-date ‘as fitted’ drawings are kept on file, accessible and ready to
assist an engineer if the system needs to be repaired at a later date.
• Keeping suitable spares, which, as a minimum, should include a number of spare manual call point
glasses or frangible elements.
• Ensuring the service organisation’s contact details are displayed near the control panel for easy
reference.

29 The role of the responsible person is covered in clause 47 of BS 5839-1

 Specifiers’ Guide Section 1 User responsibilities (the responsible person)

Keeping records in the log book

The person named as responsible for the fire detection and alarm system should record in the log book
routine inspections and tests, false alarms and other events relevant to the fire alarm system. The service
engineer should also be recording, in the log book, details of routine testing and servicing.
Section 1

For a full list of the inspections and tests that need to be arranged, please see the SERVICING AND
MAINTENANCE section, (page 86).

It is most important that whenever a false alarm occurs, the user should record as much detail as
possible on the circumstances of the alarm. This can be invaluable to the service engineer in seeking to
understand the problem and take action to avoid the situation in the future. The information that should
be recorded should include:

• date and time of occurrence;

• location of the device and its identity (for an analogue addressable system, address can usually be
obtained from a print-out);
• nature of the activity in the area (e.g. any smoke or dust);
• reason for the alarm (if known); and
• action taken following the alarm (e.g. improved working procedures to prevent dust).

84
 Specifiers’ Guide Section 1 Extensions and alterations to existing systems

7. EXTENSIONS AND ALTERATIONS TO

EXISTING SYSTEMS

Section 1
The design process for an extension to an existing fire detection and alarm system, although often a lot
simpler, should still involve the same steps as for a completely new system. However, there are some
additional considerations, not least the need to ensure compatibility between the new equipment and
the existing system and its components. New detectors may not be compatible with the control panel.
This is particularly so in the case of analogue addressable systems. Although some makes of control panel
support different signalling protocols, it is rarely possible to connect detectors using different signalling
protocols on the same system.

Without careful consideration and proper planning, adding to an existing system can have an adverse
effect on the performance of the system. For example, it is possible to exceed the capacity of the circuit
by increasing the number of detectors. Adding more sounders by extending sounder circuits may cause
excessive voltage drop, and as a consequence, the sounders may not work correctly.

A key element to consider when extending an existing system is the effect on the power supplies. It is
possible that the current load on the system has been increased, due particularly to additional sounders
or visual alarm devices having been added at some stage to such an extent that the rating and capability
of the power supply unit is exceeded. This can be checked by disconnecting the battery, simulating a
full evacuation with all alarm devices operating, and then checking that the power supply holds up at
nominal voltage and the alarm devices operate correctly. Alternatively, the current drawn under normal
(non-fire, mains failed) conditions may have been increased to such an extent that the standby duration is
reduced below the minimum period required. Battery current measurements, as described in the Section
5 on commissioning would need to be made.

Any works to extend or alter an existing system, however small, should still be subject to thorough
inspection and commissioning. It is also essential that a copy of the ‘as fitted ’ drawings for the system are
updated to reflect the changes, whether additions, deletions or modifications.

85
 Specifiers’ Guide Section 1 Servicing and maintenance

8. SERVICING AND MAINTENANCE

Servicing and maintenance of fire detection and alarm systems involves both routine attention, such
Section 1

as regular tests and scheduled service visits, and the non-routine servicing that must be carried out
in abnormal circumstances such as following a fire. It is essential for systems to receive this level of
attention in order that they remain fully operational and functionally effective in an alarm condition. BS
5839-1 section 6 contains recommendations to cover both routine attention and non-routine servicing30.

Circumstances requiring non-routine attention would include, for example:

• Checking the system following a fire and replacing any damaged parts
• Attending to an unacceptable rate of false alarms
• Repairing a fault
• Following long periods of disconnection

It may also be necessary to increase the frequency of routine servicing if the detectors are located in
environments that are dirty, dusty or involve corrosive atmospheres.

When a new servicing organisation is appointed, BS 5839-1 recommends that a “special inspection”
be carried out. This would involve checking acceptability of the previous servicing records. If the
records were found to be in order, then the new service contract would continue where the old one
was terminated. Where the records were unsatisfactory, it would be good practice for more thorough
checks to be carried out, testing all detectors, MCPs and alarm devices to the level recommended for the
annual tests, shown in the latter part of this section. Any discrepancies would be reported to the user or
responsible person.

Simple routine inspections and test, such as the weekly test of an MCP should be carried out by the user.
The more complex inspections and tests, such as the periodic tests, should be done by a competent
person, normally a fire alarm maintenance contractor.

The daily, weekly, monthly inspections for the user and the quarterly, periodic and annual inspections
and tests for a competent person are listed below and these should be done, where applicable, to ensure
continued compliance with BS 5839-1. If manufacturers recommend certain alternative or additional tests
for specialist equipment, such as aspirating detection, or flame detection, then this should be recorded
in the operation and maintenance manual and made part of the regular inspections and tests by a
86
competent person.

The person responsible for the system should ensure that the routine attention shown below is followed.
It will often be up to the commissioning engineer, and later in the life of the system it will be up to the
service engineer user to train the user in routine checking and keeping the log book up-to-date.

30 Servicing and maintenance is addressed in section 6 of BS 5839-1.

 Specifiers’ Guide Section 1 Servicing and maintenance

Daily - User
Check the control panel to ensure that the system is operating normally.

If a fault is indicated, make sure the servicing company is called to attend to it and record the fault in the

Section 1
system log book.

Check that any faults previously recorded in the log book have received attention.

If a false alarm occurs the user should enter in the log book the date and time, the detector, or MCP
concerned and any activities that could have caused the false alarm. This information will be of great
benefit to the servicing engineer in helping to reduce false alarms.

If a fire occurs, where practical, the user should record the event and the action taken in the log book,
and contact the service and maintenance organisation to make any repairs as necessary.

Weekly - User
Test one MCP to check that the system responds correctly and that the alarm is displayed on the control
panel. Test a different MCP each week and progressively, over a period of time, operate all MCPs in
rotation. It may be useful to make a list of MCPs to ensure all are tested regularly. Before the test,
remember to phone the ARC (if relevant) to take off watch for the test and then, when complete, to check
they received the signal and to go back on watch.

Monthly - User
If there are employees working out of normal hours, an extra MCP should be tested, similar to the
weekly test.

If a standby generator is being relied upon to supply the system, run the generator for an hour. This
would be applicable if the battery does not have sufficient capacity to achieve 24h standby (see the
section on CONTROL PANEL EQUIPMENT page 62). If the battery calculations and measurements show
the battery has sufficient capacity for 24h standby, the monthly generator test need not be done as part
of the fire alarm maintenance, but it will still need to be done (probably by others) at some stage for
building equipment testing.

Visually check any vented (wet) cells for condition and electrolyte level. This check is not applicable to
sealed batteries.

Quarterly - Competent Person

This only applies to sites with vented cells. Check the charge voltage and current are within the 87

specification and the general condition and electrolyte level of the cells is acceptable and top up
if necessary.
 Specifiers’ Guide Section 1 Servicing and maintenance

Periodically - Competent Person

The user should arrange for the system to be serviced by a competent person at periods not exceeding
6 months, but it may be more frequent if agreed as necessary. This should include an inspection of the
system to:
Section 1

• phone the ARC (if relevant) to take off watch;

• check the general condition of the control panel(s), that there is no apparent damage;
• check the log book and false alarm records and take action as necessary;
• check the battery condition and “installed on” date and replace if necessary (most sealed batteries
have a life expectancy of 4 years);
• simulate a full alarm load with battery disconnected, then with battery connected but mains
disconnected and check alarms operate correctly; confirm connections of battery remains adequate
(e.g. with a battery capacity meter);
• check the fault monitoring of all circuits is working correctly;
• test one detector or one MCP on each zone for non-addressable systems, or on each loop for analogue
addressable systems;
• check clearance around detectors is no less than 500mm;
• check MCPs are visible and not obstructed;
• carry out a visual inspection of the building to determine whether any changes to the layout or
occupancy, such as tenants with new work processes, have had a detrimental effect on the fire alarm
system;
• check printer legibility and replace consumables, if applicable;
• return all circuits and equipment to normal;
• check the ARC received the signals and put back on watch;
• ensure your Competent Person or organisation contact details are displayed near the control panel so
the user can quickly telephone if necessary.

Annually
Arrange with the user on site for a service visit, which should include all items covered in the periodic
visit and in addition:

• for analogue addressable systems, run a print out of all analogue values at the control panel, to check
88
that the values are within the range specified by the manufacturer;
• check signal strength for radio systems continues to be acceptable;
• functionally test every detector and MCP ensure that they are operating correctly in accordance with
the recommendations of the manufacturer;
• check any remote indicators show when the associated detectors are activated;
• check all audible, visual and any other alarm devices operate correctly;
• check all visible cables and devices are securely fixed;
• check the cause and effect has not been cancelled or deleted (on a large site, it probably would not
be practical to do the full check as in commissioning, so running the cause and effect through one
sequence should show it is still operating).
 Specifiers’ Guide Section 1 Servicing and maintenance

For a large site, it would take a considerable time to do the whole annual test once per year. BS 5839-1
makes it clear that the annual test may be over the 12 month period, so it may be agreed, for the servicing
organisation, to spread the testing of detectors over, for example, four periodic visits in the year and to
test 25% each time, keeping appropriate records to avoid duplications or omissions.

Section 1
Handover for each service visit
All tests and service visits should be recorded in the log book and a handover made to the person on
site who is responsible for the system making clear any action that may be needed to comply with BS
5839-1. The Competent Person should ensure that the person has understood any action recommended
and accepted responsibility by signing and clearly printing their name on the worksheet, report sheet or
maintenance certificate. In the case of ongoing site action necessary to comply with BS 5839-1, it should
be documented each visit and signed for as described above.

89
 Specifiers’ Guide Section 1 List of other standards and guides

LIST OF OTHER STANDARDS AND GUIDES

UK Standards and Codes of Practice
Section 1

BS 5839-1 Fire detection and fire alarm systems for buildings

- Code of practice for system design, installation, commissioning and maintenance.
BS 5839-3 Fire detection and alarm systems for buildings
- Specification for automatic release mechanisms for certain fire protection equipment.
BS 5839-6 Fire detection and fire alarm systems for buildings
- Code of practice for the design, installation and maintenance of fire detection and fire alarm systems
in dwellings.
BS 5839-8 Fire detection and fire alarm systems for buildings
- Code of practice for the design, installation, commissioning and maintenance of voice alarm systems.
BS 6266 Fire protection for electronic equipment installations. Code of practice.
BS 7273-1 Code of practice for the operation of fire protection measures
- Electrical actuation of gaseous total flooding extinguishing systems.
BS 7273-2 Code of practice for the operation of fire protection measures
- Mechanical actuation of gaseous total flooding and local application extinguishing systems.
BS 7273-3 Code of practice for the operation of fire protection measures
- Electrical actuation of pre-action sprinkler systems
BS 7273-4 Code of practice for the operation of fire protection measures
- Actuation of release mechanisms for doors
BS 7273-5 Code of practice for the operation of fire protection measures
- Electrical actuation of watermist systems (except pre-action systems)
BS 7671 Requirements for electrical installations. IET Wiring Regulations
- On-site guide.
BS 9991 Fire safety in the design, management use of residential buildings. Code of practice.
BS 9999 Code of practice for fire safety in the design, management and use of buildings.
BS EN 54-2 Fire detection and alarm systems
- Control and indicating equipment.
90
BS EN 54-3 Fire detection and fire alarm systems
- Fire alarm devices - Sounders.
BS EN 54-4 Fire detection and alarm systems
- Power supply equipment.
BS EN 54-5 Fire detection and alarm systems
- Heat detectors - Point detectors.
BS EN 54-7 Fire detection and alarm systems - Smoke detectors
- Point detectors using scattered light, transmitted light or ionisation.
BS EN 54-10 Fire detection and alarm systems
- Flame detectors.
BS EN 54-11 Fire detection and alarm systems
- Manual call points.
 Specifiers’ Guide Section 1 List of other standards and guides

BS EN 54-12 Fire detection and alarm systems - Smoke detectors

- Optical beam detectors.
BS EN 54-20:2006 Fire detection and fire alarm systems. Aspirating smoke detectors.
BS EN 54-25:2008 Fire detection and fire alarm systems. Components using radio links.

Section 1
BS EN 54-26 In course of preparation. Fire detection and fire alarm systems. Part 26. Point fire detectors
using carbon monoxide sensors.
Code of Practice for Design, Installation, Commissioning & Maintenance of Aspirating Smoke Detector
(ASD) Systems Fire Industry Association.

Fire Safety Risk Assessment Guides issued by HM Gov for Eng & Wales:
Regulatory Reform (Fire Safety) Order 2005 - A short guide to making your premises safe from fire
Fire Safety Risk Assessment - Offices and Shops
Fire Safety Risk Assessment - Factories and Warehouses
Fire Safety Risk Assessment - Sleeping Accommodation
Fire Safety Risk Assessment - Residential Care Premises
Fire Safety Risk Assessment - Educational Premises
Fire Safety Risk Assessment - Small and Medium Places of Assembly
Fire Safety Risk Assessment - Large Places of Assembly
Fire Safety Risk Assessment - Theatres, Cinemas and Similar Premises
Fire Safety Risk Assessment - Open Air Events and Venues
Fire Safety Risk Assessment - Healthcare premises
Fire Safety Risk Assessment - Transport Premises and Facilities
Fire Safety Risk Assessment - Means of Escape for Disabled People (Supplementary Guide)
Fire Safety Risk Assessment - Animal Premises and Stables
Website http://www.communities.gov.uk/fire/firesafety/firesafetylaw/

91
 Specifiers’ Guide Section 1 List of other standards and guides

Scottish guidance on the Fire (Scotland) Act and the Fire Safety (Scotland) regulations:
Care Homes Guide (Revised)
Offices, Shops & Similar Premises Guide (Revised)
Section 1

Factories & Storage Premises Guide (Revised)

Educational & Day Care for Children Premises Guide (Revised)
Small Premises Providing Sleeping Accommodation Guide (Revised)
Medium & Large Premises Providing Sleeping Accommodation (Revised)
Transport Premises Guides (Revised)
Places of Entertainment and Assembly Guide
Healthcare Premises Guide
Website http://www.firelawscotland.org

Healthcare
Firecode – Fire safety in the NHS. Health Technical Memorandum, HTM 05-03: Operational provisions
Part B: Fire detection and alarm systems
Scottish Health Technical Memorandum 82: Alarm and detection systems
Scottish Health Technical Memorandum 82: Supplement A: Automatic fire control systems and voice
alarm systems

Standards and codes in other countries

NFPA 72A. Standard for the installation, maintenance and use of local protective signalling systems.
NFPA 72B. Standard for the installation, maintenance and use of auxiliary protective signalling systems for
fire alarm service.
NFPA 72C. Standard for the installation, maintenance and use of remote station protective
signalling systems.
NFPA 72D. Standard for the installation, maintenance and use of proprietary protective signalling systems.
NFPA 72E. Standard on Automatic Fire Detectors.

Regulations
Ionising Radiation Regulations. Ionisation smoke detectors subject to the requirements of the

92
Environmental Permitting Regulations 2010 and to the Ionising Radiations Regulations 1999 made under
the provisions of the Health and Safety at Work Act 1974.

Health and Safety (Safety Signs and Signals) Regulations 1996.

 Specifiers’ Guide Section 1 FIA application guidelines for carbon monoxide (CO) fire detectors

FIA APPLICATION GUIDELINES FOR

CARBON MONOXIDE (CO) FIRE
DETECTORS

Section 1
1. Introduction
Carbon monoxide (CO) is a known-product of the combustion of carbon based material. It is an invisible
and odourless gas which, unlike smoke, cannot be detected by human beings. When present in the
atmosphere in sufficient quantity, carbon monoxide can seriously impair the ability of people to react
in a fire situation and can eventually lead to their death. In certain conditions where, due to oxygen
starvation, the fire develops slowly, a significant and dangerous concentration of carbon monoxide may
be present in the atmosphere before other known means of detection are able to operate.

Although there are many different techniques for sensing and detecting the presence of carbon
monoxide, currently available fire detection products use an electrochemical detection cell. This offers
clear advantages both in term of cost and power consumption. The electrochemical cell can be easily
configured into fire detection and fire alarm systems in the same way as other traditional fire detectors.
Electrochemical cells, however, have characteristics that can make them sensitive to other substances
thus creating a risk of unwanted alarms. Other substances, may adversely affect the response of the cell
by poisoning it and examples are listed later in these guidelines.

While these application guidelines are based upon currently available detectors, it is important to be
aware that advances in detection technology are likely, in future, to further improve the performance of
CO fire detectors.

It is very important to understand that there are significant differences in the performance characteristics of
Carbon Monoxide gas detectors for environmental monitoring, domestic Carbon Monoxide gas detectors
and Carbon Monoxide fire detectors. These differences include sensitivity levels, time of response and the
specification of the sensing element. Manufacturer’s information and recommendations should be used to
verify the correct application of each type of CO detector.

2. General guidance on the application of CO fire detectors – important note

CO fire detectors are not a general replacement for smoke detectors. Although their specific
93
characteristics can be of benefit in certain fire risk situations, it is important that, for each application, the
fire risk is evaluated before a choice of detector technology is made.

As with most types of detector, there are applications where CO detectors are appropriate and others
where they are inappropriate. The following list provides general guidance on these issues.

2.1 Where there is a slow smouldering fire, there will normally be reasonably high levels of Carbon
Monoxide. In addition, in the early development of most fires some degree of CO will be present.
2.2 Clean burning or rapid burning flaming fires, such as liquid fuel fires, produce low levels of CO gas
as a more complete combustion occurs. As the fire develops, the air supply may be insufficient to
adequately sustain the fire and oxygen starvation may begin to occur. Under these circumstances the
level of CO is likely to increase.
 Specifiers’ Guide Section 1 FIA application guidelines for carbon monoxide (CO) fire detectors

2.3 Like smoke detectors, carbon monoxide fire detectors will benefit from the convection currents
created by the heat at the source of combustion. These currents assist the CO to reach the sensing
element of the detector. However, as a gas, carbon monoxide will also diffuse within the protected
volume thus possibly enabling CO fire detectors to operate effectively in spaces where the presence
Section 1

of physical barriers may have restricted the spread of smoke. Examples of such physical barriers
include heavily beamed ceilings, ceiling voids, transportation of the gas to adjacent rooms and hot air
layers.
2.4 There are concerns that the application of CO fire detectors may sometimes result in early detection
being ignored because the location of the fire may not be readily identified even though high levels
of CO may be present. Those who are responsible for the safety of buildings and their occupants will
need to be instructed on this possibility.

Important note.
a) As for any detection technology, it is important that CO fire detectors are only used where the risk
assessment indicates that they are appropriate for detecting the types of fire that may occur.

b) Detectors that combine CO with one or more of the other detection technologies (multi-sensor
detectors) are becoming available. These may offer additional advantages over single sensor detectors.

c) Always refer to the manufacturer’s recommendations to determine the suitability of the detector for
the environmental conditions.

3. Risks where CO detectors can provide a means of early fire detection

3.1 Where there is a high risk of a slowly developing or smouldering fire. A CO fire detector may give an
earlier response where carbon monoxide is likely to be produced before smoke particles.
3.2 Where it would not have been possible to use smoke detectors because of the potential presence of
unwanted fire alarm sources. Examples include bathrooms and showers where steam may be present
or areas where particularly dusty processes take place or where simulated smoke may be present.
3.3 Where a smoke detector may be inappropriate, a CO detector may, in some applications, be a better
option than a heat detector. An example of this may be in an area where a specific type of fire risk can
be identified such as in sleeping accommodation.
3.4 Where the fire is starved of oxygen causing incomplete combustion. Examples include enclosed
spaces such as a laundry storage room or a cupboard where a CO detector sited outside the door may
94
detect a fire before smoke begins to spread outside the affected enclosure.
3.5 Where CO detection can supplement other forms of detection technology in applications where
movement of the smoke from the source of a fire may be restricted due to hot air layering effects
(stratification effect). In these circumstances the diffusion process of the CO gas may assist the
detection of a fire. In addition to diffusion through thermal barriers within a room, the diffusion of the
gas is also likely to permit penetration into, for example, roof spaces and voids.
 Specifiers’ Guide Section 1 FIA application guidelines for carbon monoxide (CO) fire detectors

4. Siting of CO fire detectors

It is recommended that the rules for the siting of conventional smoke detectors as defined in Clauses
12 and 13 of BS 5839 Part 1: 1988 are applied. These rules have been based on the convection process
for smoke transportation. In addition, the siting for the CO fire detectors may also benefit from the gas

Section 1
diffusion process.

It is however important that when siting any type of fire detector that all relevant data, especially in
relation to likely fires, the site conditions and the environment, is taken into account as part of the
selection process. It may be of benefit to use a combination of different types of fire detectors to
increase the overall level of detection performance and reliability.

5. Risks where CO fire detectors are not recommended as the prime means
of detection
CO detectors should NOT be used:
5.1 Where fire starts with rapid flaming and the heat generated ensures rapid and complete combustion.
A CO detector will respond only when such a fire becomes more developed and oxygen starvation
occurs.
5.2 Where there may be a risk of an electrical fire or over heating cables. Carbon monoxide may not be
produced in detectable quantities where pyrolysis of material rather than self-sustained combustion
is involved. Note: Pyrolysis is defined as the decomposition of a substance by heat, such as an
overheated cable or other overheated plastics material. i.e. not burning with a flame.
5.3 For the detection of liquid fuel fires e.g. heptane and other flammable liquids, producing low levels of
CO gas.
5.4 For duct fire detection applications if the main objective is prevention of the spread of smoke within a
building.
5.5 For applications where significant levels of carbon monoxide may be present under normal
conditions e.g. in areas where internal combustion engines are used (such as forklift trucks etc.) and
enclosed or semi-open car parking areas.
5.6 Where the atmosphere is abnormally polluted with substances that may affect the performance
of the CO detector e.g. alcohols (including ethanol and methanol), hydrogen (as found in battery
rooms) and ammonia or other substances of a similar nature. Chemicals used as propellants in sprays
can cause unwanted alarms. Silicon based materials, especially those included in sprays, may, over
a period of time, cause clogging of the membrane within the electrochemical cell thus requiring 95

preventive maintenance in order to obtain continued effective operation.

5.7 For some types of life protection applications, such as on escape routes, where smoke detectors are
used to trigger an alarm during a fire before smoke levels on the escape routes make them unusable.
In some circumstances, CO fire detectors may be slower to respond. A combination of the two
detection types may provide some advantage depending upon the findings of the risk assessment.
 Specifiers’ Guide Section 1 FIA application guidelines for carbon monoxide (CO) fire detectors

6. Lifetime factors including testing, servicing, maintenance and replacement:

The electrochemical cell within the CO detector has a loss of response with time. In certain environments
the degradation in detector sensitivity may occur at a faster rate than for smoke or heat detectors. It
is therefore important to always refer to, and follow, the manufacturer’s recommendations for testing,
Section 1

servicing, maintenance and replacement requirements.

Testing methods for CO fire detectors may include: a small can that delivers a measured amount of CO,
Joss sticks or a means of effectively mimicking CO gas. Manufacturers’ guidance on test methods should be
followed.

DISCLAIMER
The information set out in this document is believed to be correct in the light of information currently
available but it is not guaranteed and neither British Fire Protection Systems Association nor its
officers can accept any responsibility in respect of the contents or any events arising from use of the
information contained within this document.

BFPSA March 2001

96
Specifiers’ Guide Section 1 Notes

Section 1
97
 Specifiers’ Guide Section 1 Notes
Section 1

98
To request a printed copy of all sections of
the Specifiers’ Guide please contact
[email protected]
 PP2108/2013/ISSUE 3

© Apollo Fire Detectors Ltd 1999-2013

A HALMA COMPANY

36 Brookside Road, Havant, Tel: +44 (0)23 9249 2412 Email: sales@apollo-fire.com
Hampshire, PO9 1JR, UK Fax: +44 (0)23 9249 2754 Web: www.apollo-fire.co.uk