Bsi bip 2109 2013

The D esign, Installation,
C ommissioning and
Maintenance of F ire D etection
and F ire Alarm S ystems in
Non‑ domestic Premises

A Guide to B S 5 8 3 9 ‑ 1 : 2 01 3



The D esign, Installation,
C ommissioning and
Maintenance of F ire
D etection and F ire
Alarm S ystems in
Non‑ domestic Premises

A Guide to B S 58 3 9 ‑ 1 : 2 01 3

Co l i n

S.

To dd


First edition p ub lished b y C M P I nfo rm atio n Ltd in 2 0 0 3

Seco nd editio n pub lished b y B SI in 2 0 0 6
T hird editio n p ub lished in 2 0 0 8

T his new editio n first p ub lished in 2 0 1 3

by
B S I Standards Limited
3 8 9 C hiswick H igh Road
Londo n W4 4 AL

© T he B ritish S tandards I nstitution 2 0 1 3

All rights reserved. E xcep t as p ermitted under the
D e s i gn s

an d

Pa te n ts

Ac t 1 988,

Co p y ri gh t,

no p art of this p ub lication m ay b e

rep ro duced, sto red in a retrieval sys tem o r transm itted in any fo rm
or b y any m eans – electronic, p hotoco pying, recording o r o therwis e
– without p rior p erm issio n in writing fro m the p ub lisher.

Whilst every care has b een taken in develo p ing and com p iling this
p ub licatio n, B S I accep ts no liab ility fo r any lo ss o r dam age caus ed,
aris ing directly o r indirectly in connectio n with reliance o n its co ntents

except to the extent that such liab ility m ay no t b e excluded in law.

While every effort has b een made to trace all co p yright ho lders, anyone
claiming co p yright s ho uld get in to uch with the B S I at the ab o ve address.

B S I has no resp o nsib ility fo r the p ersistence or accuracy o f
URLs fo r external o r third- p arty internet web s ites referred to
in this b oo k, and do es no t guarantee that any co ntent on s uch
web s ites is, o r will remain, accurate o r ap p rop riate.

T he right o f C olin S. To dd to b e identified as the author o f this
Work has b een asserted b y him in acco rdance with sectio ns 7 7
and 7 8 o f the

Co p y ri gh t,

D e s i gn s

an d

Pa te n ts

A c t 1 9 88 .

T yp es et in C entury S cho o lb o o k b y M o no lith
P rinted in Great B ritain b y B erfo rts Gro up , www. b erfo rts. co . uk

B ri ti s h

L i b ra ry

Ca ta l o gu i n g

in

Pu b l i c a ti o n

D a ta

A catalogue reco rd fo r this b oo k is availab le fro m the B ritish Lib rary

I SB N 9 7 8 0 5 8 0 8 0 7 5 3 4


As in the previous edition, this book is dedicated to my children,
Keith, Jayne and Fiona, all of whom make me proud every
day.

And to Karen for her undying love and support.

(Plus a special mention for the cats of Hutton Roof. )

This current version is also dedicated to the memory of a good friend
and colleague of over 3 0 years, John Northey (1 93 8-201 0) , a member
of the technical committee responsible for all versions of B S 583 9-1
(and its predecessor, CP 1 01 9) since the 1 970s and chairman of the
committee for many years until the time of his death in 201 0.

I would also wish to include in my dedications those fourteen souls
who died in the tragic fire at Rosepark Care Home in January 2004,

and all of their loved ones.

Their deaths led to the amendment of

BS 5839-1 in 201 3 and, hence, to this current revision of this book.
The fire safety profession, through improvements in standards,
will strive to ensure that such a tragic event never occurs again.

v



Contents
About the author
Foreword

ix
xi

1.

Introduction

1

2.

History of fire alarm installation codes

5

3.

The format and layout of the code

29

4.

S cope of B S 58 39 ‑ 1 : 2 01 3

33

5.

C ontents of the code

39

6.

D efining the terms

57

7.

S electing a suitable fire alarm system

65

8.

The fire alarm contract and definition
of responsibilities

9.

73

The interface between the fire alarm system
and other systems

77

1 0.

The components of the system

81

11.

D esign of fire alarm circuits

85

12.

D etection z ones and alarm z ones

1 03

1 3.

C ommunication with the fire and rescue service

111

1 4.

Audible, visual and tactile fire alarm signals

119

1 5.

S taged fire alarms

1 35

1 6.

Manual call points

1 41

1 7.

Which type of detector?

1 47

vii


D e s i gn ,

i n s ta l l a ti o n ,

co m m is s io n i n g

an d

m a i n te n a n c e

1 8.

Spacing and siting of automatic fire detectors

1 61

1 9.

C ontrol and indicating equipment

1 79

2 0.

Networked systems

1 85

21 .

Power supplies

1 89

22.

C ables and wiring

2 05

2 3.

Radio‑linked systems

21 7

2 4.

E lectrical safety and electromagnetic compatibility

223

2 5.

False alarms and their limitation

227

2 6.

Installation work

2 57

2 7.

Inspection and testing

2 63

2 8.

C ommissioning

2 67

2 9.

Documentation

2 73

30.

C ertification

2 79

31 .

Acceptance by the user or purchaser

2 83

32 .

Verification

2 87

33.

Routine testing

2 91

34.

Servicing

2 95

35.

Repair and non‑routine attention

303

36.

User responsibilities

309

viii


About the author
C olin
FInstP,

Todd

MS c,

C. E ng,

F IFireE ,

FIE T

F B E ng,

MIRM,

MS FPE ,

C. Phys,

graduated from E dinburgh University with an

honours degree in Physics. He then undertook a one year Master’ s degree
in Fire Safety E ngineering, developing a specific interest in quantitative
assessment of risk, mathematical modelling and systems engineering.
In 1 975, he j oined the captive insurance company of Unilever Ltd.
As a member of the risk management section, he carried out regular
fire surveys of Unilever premises and was responsible for providing
in-house advice on loss prevention matters. He later j oined the technical
department of the Fire Offices’ Committee (FO C) , which dealt with the
preparation of codes and standards on fire protection and approvals of
fire protection equipment. With the FOC he specializ ed in electrical
matters, and was responsible for assessing the suitability of fire alarm
equipment for FO C approval. During this time, he represented the FOC
on national committees including those of B SI. (The FOC was later
incorporated into the Loss Prevention Council and, subsequently, the
Building Research E stablishment. )
Colin Todd is a chartered engineer and a Fellow of the Institution
of E ngineering and Technology (formerly the Institution of E lectrical
E ngineers) ,

the

Institute

of

Physics,

the

Association

E ngineers and the Institution of Fire E ngineers.

of

Building

He is a corporate

member of the Institute of Risk Management and the Society of Fire
Protection E ngineers. He is also a standards associate of the British
Standards Society.
As the

final President of the

Protection

E ngineers,

he

was

UK Chapter of the

instrumental

in

the

Society of Fire
merger

between

that organiz ation and the Society of Fire Safety E ngineers to form the
Institute of Fire Safety, which subsequently became the E ngineering
Council Division of the Institution of Fire E ngineers (IFE ) (now the
Registrants’

Group of the Institution) .

He is a previous member of

ix


D e s i gn ,

i n s ta l l a ti o n ,

co m m is s io n i n g

an d

m a i n te n a n c e

the Board of the Division and the Division’s Membership Committee,
which for many years he chaired. The Division was empowered to award
engineering qualifications, including chartered engineer, to suitably
qualified fire safety engineers. He is also a previous member of the
IFE Board, where he held the responsibility for technical issues. He
serves on a number of British Standards Committees, including those
concerned with fire detection and fire alarm systems. He also represents
the Confederation of British Industry on an expert group of the Loss
Prevention Certification Board that is responsible for the development
of approval schemes for all aspects of automatic fire alarm equipment.
Colin is also experienced in assessment of fire alarm contractors’
competence in fire alarm work, acting on behalf of certification bodies.
He also acts as a technical expert for the United Kingdom Accreditation
Service (UKAS), in the accreditation of certification bodies that certificate
fire alarm contractors under the relevant industry certification schemes.
Colin’s consulting practice, C.S. Todd & Associates Ltd, is independently
certificated by the National Security Inspectorate (NSI) for the design
and verification of fire detection and alarm systems under the British
Approvals for Fire Equipment (BAFE) SP203 Scheme for fire alarm
specialists. The practice was the first independent consulting practice
to receive this certification.

x

Foreword
This book follows on from two previous guides to B S 5839-1 : 2002 written
by Colin Todd and published in 2003, 2006 and 2008. Since publication
of these earlier guides, some amendments have been made to the code,
while fire safety legislation in the United Kingdom has been subj ect to
radical change.
The advice provided by B S 5839-1 : 201 3, the British Standard code of
practice for fire detection and fire alarm systems, is presented in the
form of explanatory information followed by specific recommendations.
The explanatory information helps readers to understand the rationale
behind the recommendations. This book serves to further that understanding. It is not a replacement for the code.
This

guide

not only explains

the

code,

it also

provides

extensive

information regarding issues that have to be considered when designing,
installing and commissioning a fire detection and fire alarm system. It

could be considered as a course textbook for all those concerned with
fire detection and fire alarm systems, including users, regulators and
those who approve fire detection and fire alarm systems.
The guide benefits from the fact that Colin Todd was contracted by
B SI to produce the draft for the 2 002 version of B S 5839-1 . In preparing
that draft, Colin consulted various interested parties. It also benefits
from the fact that Colin is a member of the B SI technical subcommittee
(FSH/1 2/1 ) ,

responsible

for

the

code,

and

of its

parent

committee

FSH/1 2. He was involved in the committee discussions that took place
before the 2002 version of the code was published and in those before
each of the amendments to the code were made, including those leading
to the latest (201 3 ) edition of the code.
Colin was particularly instrumental in proposing and drafting the

amendments in 2 01 3 that were made as a result of the findings of the
Fatal Accident Inquiry into the tragic fire at Rosepark Care Home
in Scotland in 2004, which resulted in the deaths of 1 4 elderly and

xi


D e s i gn ,

infirm

p eo p le.

i n s ta l l a ti o n ,

C o lin

was

a

co m m is s io n i n g

C ro wn

exp ert

an d

m a i n te n a n c e

witness

at

the

I nquiry,

and

reco mm endatio ns he made to the Inquiry fo r amendments to B S 5 8 3 9 - 1
were co mm ended in the findings o f the I nquiry fo r co nsideratio n.
B SI

technical

co mm ittee

resp o nsib le

fo r

BS

5 839-1

had

no

T he

hesitatio n

in co nsidering and ado p ting C o lin’ s reco mm endatio ns in the ho p e that
they wo uld help to p revent such a tragedy in the future.

J.

Na a r

Ch a i rm a n ,

xii

B S I Te c h n i c a l

Co m m i tte e

FS H/1 2 /1


1 .

I n trod u cti on

The aim of this book is to provide guidance on the current edition
of B S 5839-1 , which was published in March 201 3. The code, which,
as one part of the B S 5839 suite of codes and standards, bears the

generic heading ‘ Fi re
is

entitled

an d

‘ Co de

of

m a i n te n a n c e

of

de te c ti o n
p ra c ti c e

s y s te m s

an d

fo r
in

fi re

a l a rm

de s i gn ,

s y s te m s

i n s ta l l a ti o n ,

n o n - do m e s ti c

p re m i s e s

fo r

b u i l di n gs

’,

com m is s ioning

’ . As such, the

code represents the universal ‘ Bible’ for those involved in the design,
installation, commissioning, maintenance and use of fire detection and
fire alarm systems in the United Kingdom.
Compliance with the code is very commonly required by building
control

bodies

throughout

the

United

Kingdom,

fire

and

rescue

authorities and other authorities who may enforce provisions for fire
safety in certain occupancies (e. g. housing authorities in the case of
houses

in multiple

occupation,

or Health

and Safety E xecutive

for

construction sites) . Frequently, property insurers also require automatic
fire detection systems complying with the recommendations of the code,
or are prepared to take account of such systems in their underwriting
considerations, provided the system complies with the code.
Consulting engineers also commonly demand compliance with the

code, often as part of the electrical specification for a building. However,
following the 2002 version of B S 5839-1 , a simple reference to the code,
along with siting of devices on associated drawings, does not, by itself,
constitute design of the system, as that term is defined in the code.
Whereas, under previous versions of the code, there could be some
ambiguity as to whether some aspects of design were the responsibility
of the specifier, the supplier of the equipment, or the installer, the scope
and duties of the ‘ designer’ since the 2002 version of the code are much
clearer, thereby, hopefully, resulting in less contract disputes if systems
are found to fall short of compliance with the code. E xperience shows,
nevertheless, that there is still an educational process required before

1


D e s i gn ,

i n s ta l l a ti o n ,

co m m is s io n i n g

an d

m a i n te n a n c e

the identity and role of the ‘ the designer’ is fully understood by all
relevant parties.
This guide is less detailed than the equivalent guide to the 1 988
version of the code. The reason for this is that the code now contains
substantial

explanatory

text

that

provides

background

information

on the reason for the maj ority of the recommendations made within
the code. Moreover, the ‘ unbundling’ of the explanatory text and the
recommendations within the code, effectively makes the explanatory
text,

referred to

in the

code as ‘ Commentary’ ,

a form

of guide to

the recommendations of the code. Although, inevitably, some of the
contents of this guide will virtually repeat parts of the commentary

within the code, the intention of this guide is not merely to pull together
such explanatory text into a single document. Instead, the guide is
intended to provide readers with guidance on practical application of
the recommendations in a variety of situations, taking into account the
insight to the recommendations provided by the commentary. Thus,
this guide is not a substitute for the code itself, but should be read in
conj unction with the code.
The guide should not be regarded as offering any final authoritative
interpretation

on any recommendations

of the

code,

although it is

hoped that the opinions expressed are an accurate reflection of the
relevant B SI technical committee’ s intent when the code was published.
If it is essential to obtain a definitive interpretation, such as in the
event of a dispute, advice may be sought from B SI, which will refer the
matter to the relevant committee. Ultimately, the final arbiter in such a
dispute can, however, only be the courts. Hopefully, the current edition
of B S 5839-1 is sufficiently less ambiguous and clearer than previous
versions that the need for such interpretations, and the occurrence of
disputes, are not common. To further assist users in interpretation, B SI
has published a series of interpretations given in response to queries
put to the technical committee (rather like FAQs on internet websites) .
This information is given in PD 6531 , the latest version of which was

published in 201 0.
In 2006, there were radical changes to fire safety legislation in Great
Britain, with greater responsibility placed on those who employ people
to work in, and/or have control over, buildings to ensure the adequacy
of fire precautions in the buildings for which they are responsible. More
specifically,

suitable and sufficient fire risk assessments need to be

carried out by all such dutyholders. E quivalent changes to legislation in
Northern Ireland were implemented in 201 0.
There is a compelling logic in this approach to fire safety, and it has
already led to a much more flexible and pragmatic approach to the

2


Introduction
formulation of appropriate ‘ recipes’ for fire safety in buildings. There
is no doubt that this has led to enhanced recognition of automatic fire
detection as one of the ingredients in the recipe. Whereas, traditionally,
different components of fire protection were thought to be watertight
compartments, each considered separately and independently of the
others, the modern and more holistic approach to fire safety recogniz es
the influence that the presence of automatic fire detection can have
on the

level of safety afforded

occupants

of a building,

sometimes

resulting in the possibility to relax requirements in respect of other
fire

precautions.

This

recognition

has

become

more

explicit

since

the publication of a new comprehensive guide to the design of fire
precautions in buildings, in the form of B S 9999. For example, B S 9999
permits slightly longer distances of travel to fire exits, and a slight
reduction in the width of exits, if a suitably designed fire warning
system is provided.
This modern approach to fire safety is clearly acknowledged and

well recogniz ed within B S 583 9-1 (as amended) . O n this basis, it is
reasonable to assert that the 201 3 edition of the code constitutes an
important member of the suite of codes and standards on which fire
safety in the twenty-first century is based. The 201 3 edition of the
code was the culmination of three amendments to the 2002 version,
necessitating this fourth edition of this guide. These factors, coupled
with the rate at which fire detection technology advances,

make it

unlikely that the current code and this current edition of the guide will
have the 1 4-year life of its predecessor.

3



2.

H i story of fi re al arm i n stal l ati on cod es

G en eral

The earliest codes of practice for the installation of automatic fire
detection and alarm systems were those produced by the then Fire
Offices’ Committee (FOC) . The ‘ Committee’ comprised representatives
of most of the maj or fire insurers in the United Kingdom, and the
‘ rules’ of the FOC were associated with fire insurance underwriting
considerations.
The FO C was constituted to represent fire insurers’ interests, after

insurers suffered numerous severe fire losses in the latter part of the
nineteenth century,

particularly in London wharfs and warehouses.

Although the FO C became well known throughout the world as a fire
protection approvals and code producing body, its principal raison d’ être
was to set minimum rates or ‘ tariffs’ for different classes of trade.
Insurance premium discounts could be offered to clients who installed
adequate fire protection measures, whereas premium penalties could be
imposed for risks that had adverse features.
In order to ensure that the fire protection systems for which premium
discounts could be offered were reliable and effective, it was necessary
for the FO C to enter the business of approving equipment; the first heat
detector (known then as an ‘ approved thermostat’ ) was approved in the
early 20th century. It was, however, also necessary to develop ‘ rules’
for the installation of the fire protection equipment. Thus, the earliest
automatic fire detection codes of practice in the United Kingdom were
associated with protection of property, rather than safety of life.
Although

simple,

electrical,

manual

fire

alarm

systems

became

well recogniz ed as essential for protection of occupants of buildings,
it was to be many decades before automatic fire detection became

5


D e s i gn ,

i n s ta l l a ti o n ,

co m m is s io n i n g

an d

m a i n te n a n c e

recogniz ed in the same light. Automatic detection tended to be regarded
as

insufficiently

reliable,

or

unnecessary,

for

protection

of life.

As

recently as the 1 970s, guidance that supported the (now repealed) Fire
Precautions Act 1 971 , in respect of certification of hotels and boarding
houses under the Act, advised that a manual fire alarm system was
sufficient and that automatic fire detection might only be necessary to
compensate for shortcomings in structural fire protection measures,
such as means of escape.

(E ven today, it is, thankfully very rarely,

possible to find a hotel or boarding house that has very little automatic
fire detection, having been certificated under the Fire Precautions Act
in the 1 970s.

However, such a situation would not meet the standards

now applicable in E ngland and Wales under the Regulatory Reform (Fire
Safety) O rder 2005. * Shortcomings in fire detection should be identified
by the fire risk assessment required by this legislation and should be
acted upon by the relevant dutyholder (normally the employer) . )

1 95 1

cod e: C P 3 27. 4 0 4 /4 02 /5 01

The British Standards Institution (B SI) first produced a code of practice
(CP 3 27. 404/402/501 ) in 1 951 . This code of practice was based on the
FOC rules and, indeed, reference to these rules, and more particularly,
the FOC list of approved equipment, was frequently made in situations
in which the role of the automatic fire detection system did relate to life
safety, rather than the property protection obj ective for which the rules
and the approved list were published.

1 9 72 cod e: C P 1 01 9

The FO C continued to publish their rules, and both these rules and
the B SI code appeared to stand the test of time, in that a new version
of the B SI code did not appear until 1 972, when it was published as
CP 1 01 9. Arguably, it was only then that the specialist nature of fire
alarm systems was recogniz ed within B SI, in that, whereas CP 32 7
comprised a series of codes of practice for telecommunications facilities,
sound distribution, clock systems and fire alarms, CP 1 01 9 was a code
of practice dedicated to the installation and servicing of electrical fire
alarm systems.

*

6

and equivalent legislation in Scotland and Northern Ireland.

Hi s to ry

Systems

o f fi re

a l a rm

i n s ta l l a ti o n

c o de s

designed and installed in accordance

with CP

1 01 9

still

exist today, but, as they will be over 30 years old, they will be coming
to the end of their natural life, particularly if they incorporate automatic detectors. Although these systems might continue to satisfy the
requirements of legislation, it is likely that they would fall significantly
short of satisfying the recommendations of any subsequent version
of

BS

5 83 9- 1 .

For

example,

CP

1 01 9

contained

no

quantitative

recommendations regarding sound pressure levels of alarm signals;
it was merely required that there be at least two sounders inside the
building and,

in the

case

of a system

incorporating automatic fire

detectors, an additional sounder outside the building. With regard to
audibility, the recommendation was that the type, number and location
of alarm sounders should be such that the alarm was distinct from the

background noise in every part of the premises.
None of the wiring used in a CP 1 01 9 system needed to be fire resisting,
and the wiring to fire alarm sounders did not need to be monitored.
For small manual systems, a single power supply was satisfactory, and
it is only with the publication of the 2002 version of B S 5839-1 that
this recommendation, which continued to appear within the earlier
versions of B S 5839-1 , was withdrawn. However, such systems do not
meet current legislation, as they contravene the requirements of the
Health and Safety (Safety Signs and Signals) Regulations 1 996* * (see
Chapter 21 ) .
The fact that CP 1 01 9 had a lifetime of only eight years before it was
completely revised and published as B S 583 9-1 , compared with the
21 -year life of its predecessor, is a reflection of the speed with which
developments were occurring in the field of fire detection and alarm
systems (and electronics generally) . E qually, a number of concepts that,
today, we regard as ‘ modern’ were addressed in CP 1 01 9.
For example, the forerunner of the modern ‘ voice alarm system’ , which
1

now warrants its own dedicated code of practice (B S 5839-8 ) warranted
a subclause in CP 1 01 9, which addressed the subj ect of ‘ audible alarms
provided by public address equipment’ . Similarly, two-stage alarms were
addressed, albeit that this subj ect warranted only a single sentence.
Moreover, some of the concepts, and even the detailed text, incorporated
within B S 5839-1 : 201 3 are identical to specific clauses of CP 1 01 9; an
example concerns the use of mains powered sounders to reinforce the
primary fire alarm sounders in areas with high ambient noise levels.

**
1

and equivalent legislation in Northern Ireland.
B S 5 839-8: 201 3 ,

fo r th e

de s i gn ,

Fi re

de te c ti o n

i n s ta l l a ti o n ,

an d

fi re

a l a rm

co m m i s s i o n in g

an d

s y s te m s

fo r b u i l di n gs

m a i n te n a n c e

o f v o ice

–

Co de

a l a rm

o f p ra c ti c e

s y s te m s

.

7


D e s i gn ,

i n s ta l l a ti o n ,

co m m is s io n i n g

an d

m a i n te n a n c e

There was an implication in CP 1 01 9 that manual fire alarm systems
were provided to satisfy legislation and that automatic fire detectors
were

purely

provided

for

property

protection.

This

principle

arose

from the attitudes to automatic fire detection described above and
the fact that CP

1 01 9 was based on the fire insurers’

FOC

rules.

Thus, CP 1 01 9 ‘ allowed’ manual call points to be incorporated within
automatic fire detection systems, but, if automatic fire detection was
installed, compliance with CP 1 01 9 demanded that every portion of the
building should be covered, other than small lavatories, which could

rely on detection within any common lobby serving them; this reflected
insurers’ attitudes that manual call points did not contribute to property
protection and that partial coverage by automatic fire detection was not
recogniz ed as a valid principle for fire insurance purposes.

BS 5 83 9-1 :1 980

B S 5 839-1 was first published in 1 980 as a revision of CP 1 01 9. For the
first time, the 1 980 code was produced by a fire standards committee
within

B SI,

as

opposed

to

a

more

general

electrical

engineering

committee. Although, on the face of it, this first version of B S 5839-1

had the same eight-year life as its predecessor, in fact the 1 980 code was
amended five times, in some cases quite fundamentally, between 1 980
and 1 984.
Unlike CP 1 01 9, B S 5839-1 drew a significant distinction between
systems intended for the protection of life and those for the protection
of property.

The

role

of automatic

fire

detection,

particularly

for

protection of escape routes, was acknowledged. For property protection,
the code still encouraged protection of all parts of the premises with
automatic detectors. However, for the first time, it was acknowledged
that a lower standard of protection, by installation of detectors in high
risk areas only, could still be worthwhile. However, a ‘ health warning’
was included to draw attention to the fact that such an installation
would be unlikely to satisfy the requirements of fire insurers.
Another maj or difference between B S 583 9-1 and CP 1 01 9 was that
the use of telephones for initiating a fire alarm signal within a building

was

no

longer

recommended;

such

an

arrangement,

whereby

the

fire alarm signal was given by dialling a predetermined number on a
telephone, was acceptable under CP 1 01 9.
Also,

other important changes

from

CP

1 01 9

(either in the

first

published version of B S 5839-1 or as it was ultimately amended by
1 984) included a distinction between cables that were permissible if

8


History of fire alarm installation codes
operation during a fire was required and cables permissible where
prolonged operation during a fire was not required.

In the former

case, which would apply to, for example, alarm sounder circuits, PVC
insulated cables in steel conduit were no longer acceptable without
additional fire protection (e. g. by chasing into walls or protection by
fire-resisting construction) . PVC insulated cables in rigid PVC conduits
did not require such additional protection, as it was considered that
greater thermal insulation would be provided to the cables.
Another
B S 5839-1

new
was

principle
a need

to

introduced

within

avoid

loss

total

the

of the

first
fire

version

alarm

of

signal

throughout a building in the event that fire results in a short circuit
of a sounder circuit at a single point in the building. The intent of this

recommendation, which is discussed further in Chapter 1 1 , was widely
misunderstood, to the extent that, even today, this issue is probably the
most misunderstood aspect of fire alarm design practice.
For the first time, quantitative guidance was given on sound pressure
levels within B S 5839-1 . It was no longer adequate to offer the fire
officer a tour of the building and hope that the fire alarm signal would
be sufficiently audible to him. The recommendation for a minimum
sound pressure level of 65 dB(A) , or 5 dB(A) above background noise,
and 75 dB(A) at the bedhead in sleeping risks, was introduced.
A very important development in the world of fire protection was also
reflected for the first time in B S 583 9-1 : 1 980, namely the self-contained
domestic smoke alarm. An amendment to the original 1 980 code gave
important advice on the use of domestic smoke alarms that is still valid
today.

Indeed,

recommendations for the installation of at least two

smoke alarms in a two-storey house, and for the interconnection of
smoke alarms, exceeded the first ‘ official’ recommendations produced by
the Home Office around eight years later. It is interesting to note that
it was then not until 1 992 that the original guidance was ‘ rediscovered’
or at least recirculated, in the guidance that supported the Building
Regulations 1 991 in E ngland and Wales, by which time, sadly, deaths
had occurred in two-storey dwellings ‘ protected’ by the single smoke
alarm that Home O ffice guidance suggested would be acceptable on the
basis that it was better than nothing.
In 1 982, during the early life of B S 5 839-1 : 1 980, the first addressable
systems,

in

which

each

detector

was

separately

identifiable

at

the

control and indicating equipment, appeared on the UK market. Since
these had not been anticipated when the 1 980 code was written, it did
not adequately cater for such systems. These ‘ new generation’ systems
gave rise to new perplexities in terms of compliance with the letter, or at
least the spirit, of the code. Particular issues included z onal indication;

9


D e s i gn ,

some

people

i n s ta l l a ti o n ,

claimed that,

co m m is s io n i n g

since

an d

m a i n te n a n c e

a text display of the identity and

location of each detector could be given at the control and indicating
equipment, conventional z one indicators were unnecessary.
Another issue was tolerance to fault conditions, particularly short
circuits.

The large number of devices that could now be connected

on a single circuit created the potential for areas larger than that of
the conventional z one to lose protection in the event of a single fault.
This led to the introduction of short-circuit isolators, but perplexity
regarding the number and location of isolators remained.
This situation was merely a reflection of the fact that codes of practice

do not lead technology, but merely endeavour to articulate good custom
and practice in the use of existing technology.

There is,

therefore,

always a ‘ phase angle’ between technology and the codes of practice
that describe recogniz ed good practice in its use. However, if, today, one
examines any addressable systems installed between 1 982 and 1 988, it
can well be the case that the systems do not comply with the 1 980 code
for reasons described above, while, equally, they do not comply with the
1 988 code, as the recommendations of that code could not have been
anticipated at the time of their installation.

BS 5 83 9-1 : 1 988

When the code was revised in 1 988, account was taken of addressable
systems, but all recommendations of the code could be applied to both
conventional and addressable systems. Indeed, generally, the view was
taken that the introduction of addressable systems, while providing
many benefits, should not significantly increase the vulnerability of
the system to faults. Hence, for example, the 1 988 code recommended
that, in the event of a single fault condition, the area throughout which
protection was disabled should not exceed the maximum permitted
for a single z one. This is, of course, inherently true in the case of a
conventional system, in which a circuit and a z one are synonymous, but
necessitated the provision of short circuit isolators at z one boundaries
(but not necessarily every z one boundary in the case of small z ones) in
the case of addressable systems.

The

slower

speed

of microprocessors

in

those

days

could

mean

something of a delay between manual call point or detector operation
and the operation of fire alarm sounders, particularly in systems in
which, to minimiz e false alarms, devices were polled several times to
confirm their alarm status before a fire alarm signal was given. While a
short delay between operation of a fire detector and the sounding of the

10


History of fire alarm installation codes
alarm was not significant (and, indeed, was already permitted by the
standards of the day) , concern was expressed regarding possible delays

between operation of a manual call point and the sounding of the fire
alarm. The concern did not relate to the overall evacuation time, but in
the confusion that could result if someone operated a manual call point
and the system appeared not to operate.
If, for example, a light switch were operated but the lights did not
come on for 1 0 seconds, long before that period had expired it would be
considered that the light was inoperative. Given that, in an emergency,
time appears to pass more slowly, on perception that a manual call
point

was

inoperative,

a

person

might

follow

some

inappropriate

course of action. Particular concern was expressed regarding certain
occupancies, such as hospitals. Whereas, in other occupancies, it would
be reasonable to assume that, having operated a manual call point, a
person would evacuate the building, in hospitals staff are trained to

raise the alarm and then begin movement of patients. If the fire alarm
system were considered to be inoperative, a nurse might well move
further from the patients who need assistance in order to raise the
alarm by other means.
After much debate of this subj ect, the technical committee responsible
for the code decided that, while not wishing to penaliz e new technology,
there was a need to limit the delay between operation of a manual
call point and the sounding of alarm devices in, at least, the z one of
origin (which would then be audible to anyone operating the manual
call point) . In order to minimiz e the effect of this recommendation on
systems already in the marketplace at the time, a maximum delay of
eight seconds was recommended in the 1 988 code; this figure was based
on the maximum time delay that was known to occur with systems
already on the market. However, the code gave notice that this period
would be reduced to three seconds from 1 January 1 990.
This three-second period remained somewhat controversial throughout
the 1 990s, at least within the E uropean Standards forum. Some E uropean
countries saw no good reason for imposing such a short period for the
maximum permissible delay. The result was that, when the E uropean
product standard for control and indicating equipment, B S E N 54-2, was
published in the UK in 1 998, a maximum time delay of 1 0 seconds was
permitted (a retrograde step in the opinion of the author) .
This presented a dilemma when the code was further revised in
2002. Should the code concede that a 1 0-second delay was permissible,
particularly as a fundamental principle of E uropean standardiz ation
is that barriers to trade between E uropean countries should not be
permitted by national codes (unless strictly necessary on the grounds of

11

D e s i gn ,

i n s ta l l a ti o n ,

co m m is s io n i n g

an d

m a i n te n a n c e

safety) ? Alternatively, should the maximum delay of three seconds be so
sacrosanct as to create a situation in which products complying with the
E uropean Standard adopted in the UK, and, therefore, possibly even
third-party certificated in accordance with the standard, could not be
used in installations complying with B S 5839-1 ?
Ultimately, it was agreed that the maximum period of three seconds
should

remain

within

BS

583 9-1 : 2002,

but

in

conj unction

with

‘a

health warning’ that B S E N 54-2 permits a delay of 1 0 seconds. As a
compromise, the code advises that a delay of between three seconds
and 1 0 seconds might be acceptable, but only with the agreement of
the relevant enforcing authority and the recording of the delay as a
variation in the completion certificate.
By 1 988, there was much greater use of automatic fire detection for
protection of life, particularly in premises in which people sleep. The
original principle for certification of hotels under the Fire Precautions
Act, that a manual fire alarm system was sufficient, had evolved into
a requirement that, in the event of a fire within, say, a hotel bedroom,
automatic fire detection should give a sufficiently early warning for
those beyond the room of fire origin to make their escape before smoke
from the fire made escape routes impassable. In practice, this ‘ escape
route protection’ translated into the installation of smoke detectors
within escape routes, such as corridors and staircases.
Around the mid 1 980s, however, the Home Office began to question
whether the obj ective described above was adequately achieved by the
installation of smoke detectors in escape routes only. This led to very
elegant research work by the then Fire Research Station, involving fullscale fire tests in a rig that simulated a hotel corridor with bedrooms
opening into it.
The research showed that, under certain conditions (but only certain
conditions) , it was possible that, when a fire occurred in a bedroom,
smoke could smoke log the corridor, precluding escape by those beyond

the room of fire origin, before detection by smoke detectors in the
corridor, spaced at the normally specified intervals for these detectors.
This research was taken into account in the 1 988 code, which introduced
the concept of the type L3 (‘ escape route protection’ ) system. In this
system, detectors are installed not only within the escape routes, but
also all rooms opening onto escape routes.
The purpose of the L3

system was to provide a specification for

automatic fire detection in circumstances in which the detection was
required to protect sleeping occupants under legislation, particularly
the (now repealed) Fire Precautions Act (i. e. in hotels and boarding
houses) .

12

It was

appreciated that the

L3

system

would represent a