ALARM SYSTEMS IN BUILDINGS

(1)

VŠB - Technical University of Ostrava

17. listopadu 15, 708 33 Ostrava-Poruba Czech Republic

Faculty of Safety Engineering

Department of Fire Protection

ALARM SYSTEMS IN BUILDINGS

Researches of effectiveness depending on the type of building and the user profile

Doctoral thesis

for acquiring the academic title “doctor”, abbreviated as “Ph.D.”

Author: Dipl.-Ing. Gero Gerber

Supervisor: Professor Dr. Ing. Aleš Dudáček Study Program: Fire Protection and Industrial Safety Field of Study: Fire Protection and Safety

Ostrava, 14 August 2019

(2)

(3)

Abstract

Gerber, Gero. 2019. Alarm Systems in Buildings - Researches of effectiveness depending on the type of building and the user profile. Dissertation Thesis: VŠB Technical University of Ostrava / CZ, Faculty of Safety Engineering. Supervision: Professor Dr Ing Aleš Dudáček, Ostrava

Alarm systems are used in large buildings to inform all persons quickly and effectively about an existing danger situation. They also intend to trigger a certain behaviour. In practice, there are many reports that show that people often do not behave correctly during an alert.

In a first step this paper investigates, what fire protection planners and building authorities expect from alarm systems and which criteria can be used to evaluate the effectiveness of the systems. In the second step, a large collection of case reports is used to analyse how the effectiveness of the alarm systems depends on the type of building, the user profile and the type of alarm system. Appearing deficits are examined for their causes.

As a result of the expert interviews, the criterion for an acceptable effectiveness for the further investigation was set at 90%. In 72 % of the collected case studies, this value and thus the protection goal of the plant were not achieved. The statistical evaluation showed that the effectiveness of the alarm system strongly depends on the type of building and the user profile, but only to a small extent on the type of alarm system. Since there are also large deviations within the building types and user profiles, further factors must have a strong influence. The further investigation resulted in the fact that above all missing instruction and drill of the users and a habituation to false alarms, lead to the fact that persons do not react in the case of alarm as intended.

The last part of this thesis contains general and building type-specific technical and organisational measures, with which a sufficient effectiveness of the alarm systems can be achieved.

Key words: Alarm systems; Danger; Alert; Buildings; Building Type, User Profile

(4)

Abstrakt

Gerber, Gero. 2019. Alarm Systems in Buildings - Researches of effectiveness depending on the type of building and the user profile. Dissertation Thesis: VŠB Technical University of Ostrava / CZ, Faculty of Safety Engineering. Supervision: ProfessorDr Ing.

Aleš Dudáček, Ostrava

Poplašné systémy jsou ve velkých budovách používány k rychlému a účinnému informování všech osob o stávající nebezpečné situaci a k výzvě k určitému jednání. V praxi existuje mnoho zpráv, které ukazují, že se lidé při poplachu nechovají určeným způsobem.

V prvním kroku této práce je zkoumáno, jaká očekávání mají projektanti protipožární ochrany a stavební úřady v oblasti poplašných systémů a podle jakých kritérií může být hodnocena jejich účinnost. Ve druhém kroku je na základě velké sbírky případových studií zkoumáno, jak závisí účinnost poplašných systémů na typu budovy, profilu uživatele a typu poplašného systému. Existující deficity jsou analyzovány z hlediska jejich příčin.

V návaznosti na rozhovory s odborníky bylo stanoveno pro další šetření kritérium pro přijatelnou účinnost na 90%. V 72% případů nebyla tato hodnota a tím i ochranný cíl zařízení dosaženy. Statistické vyhodnocení ukázalo, že účinnost poplašného zařízení silně závisí na typu budovy a uživatelském profilu, ale jen v malé míře na typu poplašného zařízení. Protože existují také velké rozdíly v rámci typů budov a uživatelských profilů, musí mít významný vliv další faktory. Následný výzkum ukázal, že především chybějící instruktáž a cvičení uživatelů a návyk na falešné poplachy vedou k tomu, že osoby v případě poplachu nereagují správným způsobem.

Poslední část práce obsahuje obecná a stavebně specifická technická a organizační opatření, kterými může být dostatečná účinnost poplašných zařízení dosažena.

Klíčová slova: poplašné zařízení, nebezpečí, poplach, budovy, typ budovy, uživatelský profil

(5)

Kurzfassung:

Gerber, Gero. 2019. Alarm Systems in Buildings - Researches of effectiveness depending on the type of building and the user profile. Dissertation Thesis: VŠB Technical University of Ostrava / CZ, Faculty of Safety Engineering. Supervision: Professor Dr.-Ing.

Aleš Dudáček, Ostrava

Alarmierungsanlagen werden in großen Gebäuden genutzt, um alle Personen schnell und wirksam über eine bestehende Gefahrensituation zu informieren und zu einem bestimmten Verhalten aufzufordern. Aus der Praxis gibt es viele Berichte, die zeigen, dass sich Personen bei einer Alarmierung oft nicht bestimmungsgemäß verhalten.

Im ersten Schritt dieser Arbeit wird untersucht, welche Erwartungen die Brandschutzplaner und Baubehörden an Alarmierungsanlagen haben und anhand welcher Kriterien die Wirksamkeit der Systeme bewertet werden kann. Im zweiten Schritt wird anhand einer großen Sammlung von Fallstudien untersucht, wie die Wirksamkeit der Alarmsysteme von der Art des Gebäudes, dem Benutzerprofil und der Art der Alarmanlage abhängt. Bestehende Defizite werden auf ihre Ursachen hin untersucht.

Im Ergebnis der Experteninterviews wurde das Kriterium für eine akzeptable Wirksamkeit für die weitergehende Untersuchung auf 90% festgelegt. In 72 % der gesammelten Fallbeispiele wurden dieser Wert und damit das Schutzziel der Anlage nicht erreicht. Die statistische Auswertung ergab, dass die Wirksamkeit der Alarmierungsanlage stark vom Gebäudetyp und vom Nutzerprofil aber nur in geringem Maße von der Art der Alarmierungsanlage abhängt. Da es auch innerhalb der Gebäudetypen und Nutzerprofile große Abweichungen gibt, müssen weitere Faktoren einen starken Einfluss haben. Die weitere Untersuchung ergab, dass vor allem fehlende Unterweisung und Übung der Nutzer und eine Gewöhnung an Falschalarme, dazu führen, dass Personen im Alarmfall nicht bestimmungsgemäß reagieren.

Der letzte Teil der Arbeiten enthält allgemeine und gebäudetypspezifische technische und organisatorische Maßnahmen, mit denen eine ausreichende Wirksamkeit der Alarmanlagen erreicht werden kann.

Schlüsselworte: Alarmierungsanlage; Gefahr; Gebäude; Gebäudetyp; Nutzerprofil

(6)

Statutory declaration and consent to publication

I hereby declare that the entire thesis has been elaborated by myself according to the instructions of the supervisor with the use of literature stated in the list of bibliographic references and in compliance with Study Regulations. In accordance with Section 47b of Act No. 111/1998 Coll., as amended, I agree with the publication of the entire doctoral thesis by means of the remote access information system of VŠB – Technical University of Ostrava.

I am informed that Act No. 121/2000 Coll., on Copyright and Rights Related to Copyright and on Amendment to Certain Acts (the Copyright Act), as amended, Art. 60 – School Work applies to my doctoral thesis.

I note that VŠB – Technical University of Ostrava does not infringe upon my copyright by utilizing my doctoral thesis for the internal needs of VŠB – TU Ostrava (Art. 35, Par. (3) of Act No. 121/2000 Coll., as amended). If I utilize my doctoral thesis, or provide a license to utilize it, I am aware of the duty to inform VŠB – Technical University of Ostrava about this fact; in such a case, VŠB – Technical University of Ostrava is entitled to claim an adequate contribution from me to cover the cost expended by VŠB – Technical University of Ostrava for producing the work up to its real amount (Art. 60, Par. (3) of Act No. 121/2000 Coll., as amended).

Ostrava, 14 August 2019

………....

Dipl.-Ing. Gero Gerber

(7)

Preface (motivation)

Fires and other dangerous situations in buildings with many users pose a particular challenge for operators and rescue services. The risk of personal injury can be reduced by structural and technical measures. A particularly important means is the rapid information of users about an identified hazardous situation combined with the request for self-rescue.

In large and complex buildings this is only possible with the help of alarm systems.

The technical inspection of such plants belongs to the professional tasks of the author.

He and many professional colleagues have experienced that people react very differently to the danger signals and often do not behave as intended.

The present paper examines how the effectiveness of alarming depends on the type of alarming system, the building type and the user profile and how the effectiveness of alarm systems can be increased. If all users of a building who are capable of doing so behave as intended when receiving a danger signal, there is a high chance that personal injury and deaths will be avoided even in the event of critical events.

Ultimately, the application of the research results in practice should improve the safety of people in large buildings.

I would like to take this opportunity to thank my supervisor Professor Dr Ing Aleš Dudáček for his competent guidance and Professor Dr Tilmann Betsch from the University of Erfurt for their professional support.

My special thanks go to the twelve experts who gave me time for an interview and to the numerous participants in the survey. I would also like to thank my wife Ina and my colleagues for their understanding when professional and family tasks were neglected due to the research work.

(8)

List of tables

Table 2.1: Types of special buildings ... 4

Table 2.2: Distribution of hazard potentials among building types ... 6

Table 2.3: User profiles ... 12

Table 2.4: Hazard parameters and sensory perceptions ... 15

Table 2.5: Response times for different alarm signals ... 22

Table 2.6: Response times for different alarm signals and user experiences ... 23

Table 2.7: Suitable danger signals ... 28

Table 2.8: Meaning of Speech Transmission Index ... 34

Table 2.9: Alarm types in various special buildings ... 38

Table 2.10: Normative requirements ... 39

Table 2.11: Types of hazard signals ... 40

Table 3.1: Qualifications and areas of experience of the interviewed experts ... 46

Table 3.2: Variables for Statistical Evaluation ... 51

Table 3.3: Signal variations and combinations ... 57

Table 4.1: Requirements for the perceptibility of hazard signals ... 61

Table 4.2: Alarm systems preferred by the experts in various building types ... 66

Table 4.3: Duration of alarm in case of evacuation alarm (expert opinion) ... 68

Table 4.4: Professional qualification of the survey participants ... 71

Table 4.5: Frequencies of building types in the case reports ... 72

Table 4.6: Frequencies of the user profiles in the case reports ... 74

Table 4.7: Frequencies of alarm types in the case reports ... 75

Table 4.8: Frequencies of the alarm causes in the case reports ... 76

Table 4.9: Frequencies of the different alarm technology in the case reports ... 77

Table 4.10: Frequencies of the alarm trigger in the case reports ... 78

Table 4.11: Frequencies of alarm transmission types in the case reports ... 78

Table 4.12: Comparison of the expert recommendation with the survey result ... 81

Table 4.13: Results of the Chi-Square-Test ... 84

Table 4.14: Tests of intermediate subject effects ... 86

Table 4.15: Influence of user profile and type of alarm system on effectiveness ... 88

Table 4.16: Influence of type of building and type of alarm system on effectiveness ... 90

(12)

Table 4.17: Causes of poor perceptibility ... 94

Table 4.18: Causes of poor effectiveness ... 95

Table 5.1: Proposal for an urgency-optimised alarm system ... 119

Table 6.1: Chi square test B-A: calculation ... 1

Table 6.2: Chi-square-test B-A: results ... 2

Table 6.3: Chi square test B-U: calculation ... 3

Table 6.4: Chi square test B-U: results ... 4

Table 6.5: Chi-square-test U-A: calculation ... 4

Table 6.6: Chi-Square-test U-A: results ... 5

(13)

List of figures

Figure 1.1: Reactions of persons to alarm signals ... 1

Figure 2.1: Danger zones: abstract representation ... 8

Figure 2.2: Danger zones: example of a room fire in an office floor ... 8

Figure 2.3: Intended reactions in case of alarm ... 9

Figure 2.4: Computational simulation of an evacuation ... 10

Figure 2.5: Stimulus and response based models of alarm ... 14

Figure 2.6: Examples for human hearing threshold ... 17

Figure 2.7: Sound level at different sound events ... 18

Figure 2.8: Human field of vision ... 19

Figure 2.9: Decision-making ... 21

Figure 2.10: Alerting by random information transfer ... 25

Figure 2.11: non electrical an electrical alarm devices ... 25

Figure 2.12: Components of an alarm system ... 26

Figure 2.13: Danger signals according to ISO 8201 and DIN 33404-3 ... 29

Figure 2.14: Sound reflections in a room ... 30

Figure 2.15: Reduction of the sound pressure level due to distance and obstacles ... 32

Figure 2.16: Signal transmission in voice alarm systems ... 33

Figure 2.17: Size of safety signs ... 35

Figure 3.1: Research Plan ... 44

Figure 3.2: Visual representation of the ANOVA on sample data ... 53

Figure 3.3: Psychological Laboratory of the University of Erfurt, preparation room and experimental cubicles ... 56

Figure 4.1: Professional qualifications of the survey participants ... 72

Figure 4.2: Frequencies of building types in the case reports ... 73

Figure 4.3: Frequencies of the user profiles in the case reports ... 74

Figure 4.4: Frequencies of alarm types in the case reports ... 75

Figure 4.5: Frequencies of the alarm causes in the case reports ... 76

Figure 4.6: Frequencies of the different alarm technology in the case reports ... 77

Figure 4.7: Frequencies of alarm triggers in the case reports ... 78

(14)

Figure 4.8: Frequencies of alarm transmission types in the case reports ... 79

Figure 4.9: Distribution of alarm systems in buildings ... 80

Figure 4.10: Distribution of the alarm technologies in the building types ... 83

Figure 4.11: Distribution of the user profiles in the building types ... 84

.Figure 4.12: Application of alarm systems for different user profiles. ... 85

Figure 4.13: Effectiveness of alarm systems for various user profiles ... 87

Figure 4.14: Influence of user profile and type of alarm system on effectiveness ... 89

Figure 4.15: Effectiveness of alarm systems for various types of building ... 89

Figure 4.16: Influence of building types and of alarm system types on effectiveness ... 91

Figure 4.17: Effectiveness for various types of alarm systems (all cases) ... 92

Figure 4.18: Effectiveness for various types of alarm systems (perceptibility >80%) ... 93

Figure 4.19: Response time to test signals ... 103

.Figure 5.1: Inclusion of disabled people ... 117

Figure 5.2: Recommendation for levels of urgency ... 120

(15)

Glossary

Technical terms can have different meanings depending on the application. The following overview explains the meaning of the terms in this paper.

Alarm system Technical equipment for warning people in a hazardous situation Alert Warning procedure for persons in a hazardous situation

Building type The building type describes the function or type of use of the building.

e.g. residential house, hospital, school

Effectiveness Relationship between the result achieved and the target set

Special building Special constructions are installations and rooms of a special kind or use which fulfil certain requirements.

e.g. high-rises, schools, meeting places

Normal residential buildings are not special buildings.

User profile The user profile describes the features relevant for alerting and self- rescue that connects the majority of building users.

e.g. “mainly local user” or “many people in need of assistance”

(16)

List of abbreviations

ANOVA Analysis of variance

DIN Deutsches Institut für Normung German Institute for Standardisation

DKE Deutsche Kommission Elektrotechnik Elektronik Informationstechnik German Commission for Electrical Engineering Electronics Information Technology

E1, E2…. IDs of the interviewed experts

EES Electroacoustic emergency warning system FAS Fire Alarm System

IOI Inter-Onset-Intervall (property of the hazard signal) IPI Inter-Pulse-Intervall (property of the hazard signal) NFPA National Fire Protection Association (USA)

SPL Sound Pressure Level STI Speech Transmission Index VAS Voice Alarm System

VDE Verband der Elektrotechnik Elektronik Informationstechnik e.V.

Association of Electrical Engineering Electronics Information Technology e.V.

VdS Service company of the Gesamtverband der deutschen

Versicherungswirtschaft (Association of the German Insurance Industry)

(17)

1 Introduction

1.1 Occasion of the work

Fire safety plays a central role in safety concepts for buildings. For simple residential buildings, the focus is on structural fire protection. For special buildings - these are buildings of a special kind and use - there are additional requirements for plant-technical and organisational fire protection. The primary goal is to protect the lives and health of the building users. In the event of a fire, as many people as possible should be able to save themselves.

This requires, on one hand, that the building has suitable escape routes, and on the other hand, that people are informed quickly and effectively in case of danger. In special buildings, such as skyscrapers, department stores, hotels and meeting places, alerting all those present quickly is possible only with technical aids. German building law requires installing an alarm system for these buildings in most cases. The alarm system must be installed by a specialist company, and in many cases tested by experts recognised by the building authorities.

During inspections, after evacuation exercises and in real danger cases, it often happens that many people do not behave as intended even though the alarm system is installed correctly and functional. The spectrum of reactions of the alarmed persons ranges from disinterest, short attention, and long hesitation to correct behaviour. (Figure 1.1)

Figure 1.1: Reactions of persons to alarm signals

(18)

Example:

In 2016, the central station of a medium-sized city in southern Germany was renovated. An alarm system with sounders was installed. The testing of the system by the expert followed and was to be carried out under lifelike conditions. There were numerous passengers, railway employees and a police patrol in the station concourse and on the platforms. No one was informed about the planned test. A fire alarm was simulated and the danger signal sounded through the building.

All passengers and employees should have left the building as planned. The test engineer was afraid that there might be unrest and an interruption of rail traffic. But what happened?

The passengers stayed in the concourse or went to the trains. Newspapers and coffee were sold at the kiosks. The railway staff continued their work and the police patrolled unimpressed.

The result of the test was disappointing. An expensive and correctly installed alarm system had not led to the intended behaviour. The alarm system was functional but ineffective.

Discussions with experts, maintenance companies and building operators showed that they had made similar observations. This research project investigates the factors on which the effectiveness of alarm systems depends, and which technical and organisational measures can be used to improve their effectiveness.

The investigation was carried out in Germany. The findings can be transferred to other countries.

(19)

1.2 Research objectives

The task of alarm systems in buildings is to make people aware of an existing dangerous situation. To get to safety, people must understand the danger signal and act accordingly.

The research paper has the following main goal:

Improving the effectiveness of alarm systems in buildings

Partial goals leading to the main goal are:

• Investigation of the effectiveness of alarm systems in buildings depending on the type of alarm system, building type and user profile

• Determination of deficits

• Determination of technical and organisational measures that enable effective alerting

The requirements and evaluation criteria were defined in the results of the expert interviews. The data for the main investigation were collected in a survey among experts from fire protection and safety engineering. For the final compilation of suitable measures to improve the effectiveness of alarm systems, both the expert interviews and the case studies from the survey were evaluated qualitatively.

The investigation was carried out in Germany.

(20)

2 State of Knowledge and technology

2.1 Buildings: Refuge and source of danger 2.1.1 Hazards in buildings

Very early in the history of mankind, people sought shelter in caves, tents and later in houses. Houses provide shelter from rain, cold, wild animals and evil neighbours. But life in the house holds dangers as well. In the past, when cooking was done on an open flame, kitchen fires often destroyed the whole house. Modern buildings are safer, but they harbour new dangers. The risk potential increases with the size of the building and the number of users.

In addition to normal residential buildings, there are numerous buildings of special kind and uses, the so-called ‘special buildings’. Typical examples are shown in Table 2.1.

The application criteria in column 2 refer to the model building regulations of the Federal Republic of Germany. (Fachkommission Bauaufsicht der Bauministerkonferenz 13.05.2016)

Table 2.1: Types of special buildings Kind of special

building Criteria according to German building law

Examples

Residential institutions No criteria Student dormitory

Flying buildings No criteria Circus, marquees

Garages

medium and large size

Medium: 100 – 1.000 m² Large: >1.000 m²

Underground car parks, multi- storey car parks

High-bay warehouses Shelf warehouse with a storage

height of >7.5 m Logistics centres

High-rise buildings

Height of the floor on the top floor with lounges > 22 m above ground level

Residential skyscrapers, office skyscrapers, observation towers Hospitals

Buildings in which medical services are provided

Hospitals, medical centres, nursing homes

Industrial constructions

Buildings for the manufacture, treatment, distribution or storage of goods

Factories, logistics warehouses, power plants

(21)

Kind of special

building Criteria according to German building law

Examples

Penitentiaries No criteria Prisons, remand centre

Places of

accommodation Number of guest beds >12 Hotel, youth hostel, bed and breakfast

Places of assembly

Meeting rooms >200 visitors Outdoor meeting places >1000 visitors,

Sports facilities >5000 visitors

Theatre, cinema, concert hall, stadium

Pubs and restaurants

Number of inside guest seats >40,

Number of outdoor seats >1000 Restaurant, canteen, beer garden Sales rooms Showrooms with a total area

>2.000 m²

Department stores, large supermarkets, exhibition halls

Schools

General and vocational schools, except schools where only adults are taught

Primary school, secondary school, grammar school

Hazards for people in buildings can be divided into two groups. On one hand, there are circumstances that lead to health damage only after prolonged exposure. Examples are moulds, electric smog or toxic building materials such as certain wood preservatives. These dangers do not require an immediate reaction, no alarm system and are, therefore, not relevant for this paper.

On the other hand, there are dangers, such as fire or gas leakage, which pose an immediate threat to users. People must react immediately to get to safety. This is not always easy in large and complex buildings. The following factors can lead to increased risk potential in special buildings:

• High number of persons

• Lack of local knowledge

• Long or unclear escape routes

• Increased risk of fire formation or fire propagation

• Increased risk of accidents

• Difficult access for rescue forces

• Limited mobility or responsiveness

• Many people in need of supervision and guidance

(22)

Table 2.2 shows which factors are to be expected in which building types.

Table 2.2: Distribution of hazard potentials among building types

Risk due to

Type of building High number of persons Lack of local knowledge Long or unclear escape routes Increased risk of fire formation or fire propagation Increased risk of accidents Difficult access for rescue forces Limited mobility or responsiveness Many people in need of supervision and guidance Residential

institutions Flying buildings Garages, medium and large size High-bay warehouses

1)

High-rise buildings

2)

Hospitals ³⁾

Industrial constructions

1)

Penitentiaries Places of accommodation

2)

Places of assembly

2)

Pubs and restaurants

2)

Sales rooms ²⁾

Schools

(23)

legend:

Typical hazard ¹⁾ for temporary employees

Possible hazard ²⁾ für customer and guests

Unlikely hazard ³⁾ for patients and visitors

As can be seen in the table, in most special buildings there are several factors that lead to a risk potential that is higher than that in normal residential buildings.

One danger that has not been mentioned so far is amok situations. The probability of occurrence for a certain building is indeed low. However, the concrete danger for the users is extremely high. A fast and effective alarm system can help to bring endangered persons to safety.

2.1.2 Degree of hazard

In order to understand the task of an alarm system, it is important to understand the different degrees of hazard and their spatial layout.

Many dangerous situations arise locally in one place or in one room and then spread throughout the building. Initially, only the persons in the vicinity of the danger point are at risk. If the event can spread, there is a potential danger to people in adjacent rooms and areas.

The danger increases with the proximity to the point of danger and with the speed at which the point of danger spreads. Persons staying in rooms or areas where the danger cannot penetrate under normal circumstances are not at risk. The normal circumstances may also include intervention measures such as deployment of the works fire brigade.

Example: A fire caused by a candle or a defective electrical device initially only affects this room. Initially, only the persons in the fire compartment are at risk. If the fire cannot be extinguished immediately in the development phase, the spread of smoke and fire in the building must be expected. Persons in the same usage unit are already potentially at risk.

There is still no danger to persons living in areas separated by fire protection technology (e.g. in other fire compartments or other floors).

Strictly speaking, the degree of exposure is not a discrete quantity. Nevertheless, it is possible to define five distinctive levels. In many cases, the spatial demarcation between the steps is formed by components such as walls or ceilings. Figure 2.1 show the degree of

(24)

danger in an abstract representation and for the example of a fire. The zone with higher hazard is a subset of the adjacent zone with lower hazard. Figure 2.2 shows an example of an office unit burning.

Figure 2.1: Danger zones: abstract representation

Figure 2.2: Danger zones: example of a room fire in an office floor

The identification of danger zones always represents a snapshot. The zones can shift, grow or shrink over time.

2.1.3 Typical actions in dangerous situations

There are three typical reactions to hazardous situations in buildings. The choice of reaction depends on the type of hazard. If there is a danger in the building, the users must leave the building. Typical examples are fires or gas leaks in buildings.

If the danger exists outdoors (i. e., storm or chemical accident), people should remain in the building or seek protection in the building. In amok situations, people should go to sheltered rooms and barricade themselves in. (Figure 2.3)

Damage zone

The stay in this area leads to physical damage.

Acute danger zone

Personal injury can occur at any time.

Potential danger zone

Area that will soon become a danger zone.

Possible danger zone

Area that can become a danger zone if the danger is not ended.

Safe Zone

Area up to which the danger zone is unlikely to penetrate.

Stairs

corridor corridor

(25)

Figure 2.3: Intended reactions in case of alarm

2.1.4 Flow of people in dangerous situations

Alarms are used to instruct people to leave a danger zone. This means that many people move simultaneously to seek shelter outside the building or in safe rooms. The escape routes in the buildings must be planned and built in such a way that they can accommodate these flows of people. Critical points where traffic jams can occur are doors and stairs.

For normal buildings there are specifications for the permissible length of escape routes and the width of doors, corridors and stairs. For large and complex buildings that are used by many people, the standard specifications are not sufficient. In these cases, the escape times can be calculated using engineering methods. An individual calculation is also required if specifications from the building regulations are not complied with, e.g. if the escape routes are too long or too narrow.

The calculation must prove that, in the event of a fire, the duration of the evacuation plus a safety margin is less than the duration of the usability of the escape routes. The following quantities are included in the calculation:

• Type of alarm

• Identification of the dangerous situation ( discovered by a person by chance or automatically via a fire alarm system)

• Alarm system or mutual warning of people

• Automatic or manual activation of the alarm

• Comprehensibility and perceptibility of the danger signal

• Previous knowledge of users (trained staff, or anonymous public)

Leave the building! Stay in the building! Barricade yourself!

i.e. fire or gas in the building i.e. hurrican or toxic air outside i.e. amok shootimg

(26)

• Emergency routes, geometry of building

• Accessibility/availability

• Configuration

• Potential contamination of escape routes with smoke

• Orientation

• Signposting

• Illumination of indication signs

• Escape lightning

• Manual direction signs

• Security staff

• Trained staff

• Firemen

The duration of the evacuation consists of the time needed for the detection of fire and the alerting, as well as the response time of the people and the actual time required for leaving the dangerous zone. (Wilk 2016)

t

evacuation

= t

detection

+ t

alarm

+ t

response

+ t

escape

The calculation can be carried out manually, e.g. by using the dynamic-hydraulic model developed by Predtetschenski/Millinski in the 1960s or with the help of computer programs e.g. with the method of cellular automatons, like shown in Figure 2.4.

Figure 2.4: Computational simulation of an evacuation

(27)

In buildings in which compliance with the permissible evacuation time is not or only very tightly observed, fire protection planners like to prescribe the installation of an alarm system. This is to ensure that all persons at risk are informed at the same time. However, the calculations assume, that all persons react immediately. Personal experience shows that this is does not always happen. Whether these are exceptions or a general problem is to be examined in the present paper.

If the investigation reveals that this is a common problem, many mathematical proofs of the evacuation times must be questioned. This would mean that theoretically calculated safety would not exist in practice in these buildings.

2.2 User profiles

In normal residential buildings, people of all ages usually live over a longer period of time. The inhabitants have the best local knowledge. Children are supervised by parents.

People in need of help are looked after by relatives or have technical aids for communication and mobility. Normally, alarm systems are not needed to ensure fire protection.

The situation in special buildings can be completely different. There is a high level of fluctuation in sales outlets, meeting places and railway stations. Visitors and customers are often not familiar with the location. The non-local people cannot be trained in fire protection measures. An intended reaction in the event of an alarm cannot be assumed.

In certain buildings, such as schools and kindergartens, there are particularly many people in need of supervision. A training for the alarm case is possible. The responsibility for the intended reaction lies with the supervisory personnel (teachers, educators).

In nursing homes and hospitals there is a high proportion of people in need of assistance. Fire protection instruction for residents and patients is only possible to a very limited extent. The bigger problem, however, is that many people in need of assistanc are physically or mentally unable to react as intended without external help.

Another special group of people are prison inmates. A special feature here is that these persons are prohibited by law from reacting independently in the event of danger. The cell doors and emergency exits must always be released by the security personnel. The hazard alarms must therefore be transmitted quickly and effectively to the guards and not to the inmates. Prison inmates occupy a special position and have not been taken into account in this research.

(28)

Four typical user profiles are defined for the study. The decision for a user profile does not mean exclusivity of the groups of persons. Of course, children (persons in need of supervision) or persons in need of assistance are also allowed to stay in a building with predominantly local people. It is also a matter of course that employees with local knowledge also work in a sales outlet with predominantly unfamiliar customers.

In order to investigate the influence of the user profile on the effectiveness of the alarm system, it is important to know which groups of people predominantly use the building.

Table 2.3 shows the user profiles defined for the study with a description of the properties and examples of buildings where these user profiles can be expected. The user profile is determined individually by the survey participants for each case report.

Table 2.3: User profiles

User profile Properties Example

Mainly local users

Most users are regularly or frequently in the buildings.

Escape routes and meeting places are known.

Regular instructions and drills are possible.

Administration buildings workplaces Mainly non-local

users

Most users are rarely or only once in the building.

Most users only know the main routes but no additional escape routes.

Regular instruction and drill is only possible for a small number of users.

Sales points meeting places railway stations airports places of accommodation Many people in

need of supervision

A very large proportion of users are underage and require adult supervision.

Escape routes and meeting places are known, but are not used independently.

Regular instructions and drills are possible.

Schools day-cares for children

Many people in need of assistance

Due to physical or mental limitations, a large proportion of users are unable to react to dangerous situations without assistance.

Most users only know the main routes but no additional escape routes.

Regular instruction and drill is only possible for a small number of users.

Nursing homes hospitals

(29)

2.3 Human factor

2.3.1 Human behaviour in hazardous situations

Human behaviour patterns in a danger situation are ancient and essential for survival.

The individual recognises a threat and has to decide what to do in a very short time. The two main options are flight or fight. In social communities of humans, but also of animals, it has proved successful to warn the other individuals of the group as well. The warning can be an invitation to flight or a call for support in the fight.

The term ‘alarm’ has many meanings in psychology, technology and colloquial language. Psychologist Dr Neville Stanton writes about this in the introductory article of his anthology ‘Human Factors in ALARM DESIGN’:

‘A frequently given definition of an alarm is ´a significant attractor of attention!

However a dictionary (Collins, 1986) gives nine definition of the word ´alarm´. These are:

• To fill with apprehension, anxiety or fear;

• To warn about danger: alert;

• Fear or terror aroused by awareness of danger: fright

• A noise signal, etc., warning of danger;

• Any device in the alarm clock that triggers off the bell or buzzer;

• A call to arms

• A warning or challenge.’

(Stanton 1994, S. 2)

The above descriptions always focus on just a few characteristics of the alarm. From a psychological point of view, two issues are mixed here.

‘The term may be used to define both, the stimulus and the response on different occasions. In the stimulus-based model an alarm exist in the environment and its presence has some effect on the individual, whereas in the response-bases model, the stimulus causes an alarm state in the individual.’ (Stanton 1994, S. 2)

Figure 2.5 shows example for both models.

The stimulus-based model assumes a largely homogeneous perception by all persons involved. If, for example, a fire breaks out in a seminar room, all persons present will perceive the hazardous situation in a similar way.

(30)

The response-based model assumes that different people evaluate the situation (the stimulus) differently.

Figure 2.5: Stimulus and response based models of alarm

Even if a human being devotes himself/herself very intensively and with concentration to a certain activity, the brain simultaneously carries out a permanent scan of the environment via the sensory organs. Even with very concentrated work on the PC, we notice when the coffee cup is too hot or the earth suddenly quakes. To a limited extent, these monitoring routines also work during sleep. If the unconscious environmental scan detects

Stimulus-based model

Response-based model

ALARM Cause for alarm

Response to alarm ALARM

(31)

an abnormality that produces a sufficiently strong stimulus, the attention changes from the previous activity to the investigation and evaluation of the unusual situation.

This behaviour pattern is used for alerting in case of danger. By a sufficiently strong mostly acoustic stimulus people are induced to interrupt their activity and to occupy themselves with the new situation. If they themselves do not recognize any danger or do not receive any further instructions or are not trained, their attention will quickly drop and no adequate reaction will take place.

2.3.2 Perception of danger

The basic prerequisite for the correct reaction in a hazardous situation is that the person recognises the hazardous situation or is informed about the hazardous situation. Humans have many senses and normally recognise dangers in their vicinity very quickly. They perceive their immediate surroundings through the close-up senses of smell, taste and touch.

Events that take place at a greater distance are captured via the longer-distance senses of vision and hearing. The Table 2.4 shows examples of which hazard parameters are perceived with which senses.

Table 2.4: Hazard parameters and sensory perceptions

Sense Fire Gas Earthquake Amok

Vision Flame, smoke,

fleeing people x Wobbly or

falling parts

Perpetrators, victims, fleeing people Hearing Crackling, collapsing

components x Falling parts Shots fired,

screams

Touch Vibrations x Vibrations Vibrations

Smell Smell of burning Smell of gas x x

Taste Smoky taste x x x

This does not apply or applies only with restrictions in the following cases:

• Sleeping people

• People with physical or mental disabilities

• Persons under the influence of alcohol, medication or narcotics

• Young children

(32)

In the potential and possible danger zones, perception of the danger by the people present is not ensured. If it is to be expected that if the danger zone spreads, the per-sons in the adjacent areas will not be able to get to safety without damage to their health, these persons must be warned.

To warn people in dangerous situations, acoustic signals are used in most cases. The main reason for this is the fact that the ear is always ready to receive and much less selective than the eye with regard to the direction of reception. In addition or as a replacement for acoustic signals, optical signals are used for alarms as well. In special cases, e.g. for hearing- impaired persons, the alarm is triggered with tactile signal transmitters.

2.3.3 Perception of acoustic signals

In order to understand the effect of acoustic signals on humans, the psychoacoustic and cognitive aspects have to be considered. Psychoacoustics investigates the relationship between the physical world of sound waves and subjective perception. Two important characteristics of the signal are audibility and urgency. (Malter und Guski 2001, S. 5–10)

The following questions are important for the person receiving the signal:

• Detection: Can the person hear anything?

• Discrimination: Can the person differentiate between two signals?

• Identification: Is it possible for the person to identify and recognise the stimulus?

• Categorisation: Is the person able to identify the stimulus as belonging to a specific class?

• Scaling: Is it possible for the person to gauge the degree of an auditory or psychological dimension (for example: loudness or urgency)?

(33)

For a signal to be audible, the sound pressure level at the ear must be above the hearing threshold. The hearing threshold depends on the frequency. The audible frequency spectrum is between 20 Hz and 16 kHz. The highest sensitivity is in the 2-3 kHz range. (Malter und Guski 2001, S. 22) .

The transmission of speech and signals takes place in the range of 300 Hz to 3 kHz.

Hearing thresholds rise with age and due to noise overload as shown in Figure 2.6.

Figure 2.6: Examples for human hearing threshold

Danger signals must be audible. This means that they must stand out clearly from the background noise level. However, they must not be too loud, so that the persons do not suffer hearing damage. It should also be borne in mind that persons alerted should still be able to communicate verbally. (DIN EN ISO 7731, S. 7) Sound pressure levels above 90 dB(A) are already very unpleasant and can lead to chronic hearing damage from prolonged exposure.

From about 120 dB(A), acute hearing damage must be expected even from short exposure.

Optimum speech intelligibility is achieved at approx. 80 dB(A). Typical sound events are shown in Figure 2.7. (Simon 2018, S. 13–14)

In the past, various methods were presented for determining the listening threshold.

An empirical procedure was described by Zwicker and Scholl in 1963. EN 457 of 1992 describes a procedure for estimating the listening threshold on the basis of octave band levels of background noise. In 1982, Patterson presented a model in which monitoring thresholds up to a noise level of 95 dB(A) can be predicted. In 1996, Moore and Glasberg further developed the Zwicker method. Laroche studied the influence of hearing impairment at the

-10 dB 0 dB 10 dB 20 dB 30 dB 40 dB 50 dB

10 Hz 100 Hz 1.000 Hz 10.000 Hz

frequency

sound pressure level

Young adult

elderly person with slight restriction in the 2 kHz band

(34)

University of Montreal in 1991. ISO 7731 describes a calculation method for determining the listening threshold on the basis of the measured background noise level. (Malter und Guski 2001, S. 35–49)

Figure 2.7: Sound level at different sound events

People are able to recognise sound patterns that are quieter than their surroundings.

Otherwise one would not even notice the softer instruments in the symphony orchestra.

However, danger signals must be detected quickly and clearly. Therefore the sound pressure level must be above the background noise level. Standards require a difference from 10 to 15 dB and a minimum of 65 dB (A).

140 130 120 110 100 90 80 70 60 50 40 30 20 10

0 Hearing limit

Windless forest Bedroom Conversation Office

Medium traffic Heavy traffic Jackhammer Pop concert Start jet aircraft, distance 100 m Start jet aircraft, distance 25 m

(35)

The hazard signal must be different from other signals and sounds and its meaning must be unmistakable.

Processing and storage of sound signals takes place in the human brain similar to speech signals. Both stimulus types are analysed in the dictionary memory and compared with formal characteristics stored by the person based on previous experience. (Malter und Guski 2001, 53 ff.)

The brain now tries to assign a meaning to the extracted traits. This is only possible if a corresponding conditioning has taken place in the past. The connection between the signal and the meaning must be learned.

2.3.4 Perception of optical signals

While acoustic signals can be heard largely independently of the position of the hearer, optical signals must be in the visual field of the observer. The visual field describes the area a person sees when looking straight ahead with his or her head upright. It covers the entire area that is imaged on the retina of the eye. This includes the central area of sharp vision and the peripheral areas, where changes are only perceived blurred. For adults, the monocular visual field in the horizontal direction is approx. 120°. The binocular visual field is approx.

180° horizontally. In the vertical direction it has an extension of approx. 70° upwards and downwards (Missfeldt 2019). A schematic representation is shown in Figure 2.8.

Figure 2.8: Human field of vision

The human eye reacts best to changes in the field of vision. This effect is used in the advertising industry. (Moving ads, short films, changing posters). It can also be used for

120°

180°

140°

(36)

optical alerting. For optical signals to be perceived well, they must be adapted to the ambient conditions. Important factors are brightness, colour spectrum and contrast.

2.3.5 Decision-making process

The preceding sections have dealt with the perception of danger signals. However, the perception of the signal is only the first prerequisite for the intended behaviour in the event of an alarm. Further preconditions are:

• Knowledge of the meaning of the danger signal

• Willingness to act

If the danger signal is a tone sequence or a light signal, its meaning must be learned.

The willingness to act depends on several factors. Persons who receive a danger signal may ask themselves:

• How seriously do I take the information?

• What experiences have I had in similar situations?

• What ‘price’ do I pay for intended behaviour? (e.g. that urgent work is not done)

• What is the ‘price’ for the refusal? (e.g. criticism from the boss)

• What advantage does a quick reaction give me? (e.g. a break from an unpleasant activity)

Figure 2.9 shows possible ways of decision-making.

Psychologists use the model of subjectively expected utility to simulate such processes. The model is based on the assumption that the decision-maker chooses the variant with the greatest expected benefit. The expected benefit of each variant is the sum of the individual benefit values of the possible consequences multiplied by their probability of occurrence. The higher and more serious the risk of an option is assessed, the more likely it is that the person will decide to avoid that risk. The perceived risk can be described as the product of the severity of the consequences and the perceived probability of occurrence of an event. (Frost 2019, S. 9–10). In a dangerous situation, relevant knowledge from one's own experience is activated. These experiences can derive from real danger situations or from danger exercises.

(37)

Figure 2.9: Decision-making

Depending on the nature of the experience, people can react very quickly or try to ignore a hazard signal. The latter occurs above all when people are very often alerted without

Initial state

Start of alarm

Is the signal perceptible?

Is the meaning of the signal known?

Are there any further indications of a hazardous

situation?

Am I willing to follow an unverified instruction?

Intended behaviour No reaction

Yes No

Yes Yes Yes

No

(38)

there being any real danger. Psychologists speak of ‘Alarm Fatigue’. Alarm Fatigue can be explained by the cry-wolf-effect.

This effect refers to the fable ‘The shepherd boy and the wolf’ by the Greek narrator Aesop. The boy called ‘Wolf’ out of boredom. All the villagers rushed to help. When later actually a wolf came, nobody took the call for help seriously. The wolf ate the boy and the herd. Just as the villagers ignored the repeated call for help, people also get used to danger signals that have no consequences. In psychology, it's called ‘habituation’ and means that the repeated setting of a stimulus (danger signal) leads to desensitisation.

2.3.6 Response time

The reaction time is the period of time from the reception of the danger signal to the beginning of the required action. The response time depends on the type of alarm signal and the level of experience of the building users. In the 1990s there was a study on this in England. The results were published in the British Standard BS DD 240 in 1997. The relevant results are summarised in Table 2.5 and Table 2.6:

Table 2.5: Response times for different alarm signals

Type of building

Response time Live

announcement

Announcement from a voice

memory

Alarm bell or siren Office building, trading- and industrial area

Users are vigilant and familiar with the building

< 1 min 3 min > 4 min

Shops and event buildings

Users are vigilant and familiar with the building

(39)

Table 2.6: Response times for different alarm signals and user experiences

Evacuation drills Regularly --- Never

Warning system Good

response time Mean

response time Bad response time

Alarm sirens < 4 min 7 min > 10 min

Alarm sirens

with rising and falling tones < 3 min 5 min > 7 min

Alarm sirens

with rising and falling tones and additional information message

< 2min 3,5 min > 5 min

Alarm sirens

with rising and falling tone and additional optical information system (screen)

(British Standard DD 240, 1997)

The determined values give interesting hints for the investigation of effectiveness of alarm systems. A good response time does not automatically lead to good effectiveness.

Good effectiveness can only be achieved if the required action is completed by most people.

If all people react quickly but change their decision later, the effectiveness of the alarm can be very poor. In the other case, a good effectiveness can also be achieved with long reaction times. Nevertheless, the values obtained allow the following conclusions to be drawn:

• Alarmed people react much better to live spoken situation-related announcements than to stored voice messages.

• Any kind of voice announcement is better for the response time than a simple acoustic signal.

• In buildings with predominantly local users, shorter response times are achieved than in buildings with many sporadic users.

• Regular exercises lead to a significant improvement of the response time.

• The more and more precise information the alarmed persons receive, the faster they react.

(40)

2.4 Alarm systems

2.4.1 Tasks of alarm systems

The task of an alarm system is to inform all potentially affected persons quickly and effectively about a hazardous situation. Alarm systems may be required in the following hazardous situations:

• Fire

• Technical incident

• Environmental hazards

• Burglary or robbery

• Terrorist act

• Amok situation

The alarm can be directed to different groups of people:

• Persons deployed to combat the danger or provide assistance

• Persons who are to save themselves

These groups of people can be alerted together or separately. The intended reaction of the people depends on the nature of the danger. The signal can mean:

• Leave the building (i. e., fire, bomb threat)

• Stay in the building (i.e. nat. disaster, chemical incident)

• Look for protection and barricade yourself in your room (i.e. robbery, amok) The main goal of alerting is a successful self-rescue. (Brandschutz-Wegweiser 2016).

Warning signals should draw attention to a danger without frightening the alarmed persons.

2.4.2 Possibilities of alarming

In small and medium-sized buildings, it is often sufficient that the present individuals alert each other in case of danger. (See Figure 2.10). In most cases this happens randomly.

Often it is not guaranteed that the information reaches everyone.

(41)

Figure 2.10: Alerting by random information transfer

If the building has a fixed user group, the alert sequence can be specified on an organisational level. Alarm chains can be formed. Fire safety officers can check, that everyone has left the area.

If the buildings are larger or there are many people in them, the mutual alerting takes too long. In order to alert everyone quickly, technical aids must be used. The simplest solutions are manually operated fire bells, sirens, and compressed-air fanfares, as shown in the left side of Figure 2.11.

Figure 2.11: non electrical an electrical alarm devices

In order to create a longer-range electrical alarm, devices like motor sirens or mega- phones can be used as shown in in the right side of Figure 2.11

Advantages of these systems are low acquisition and operating costs. Their disadvantage is the short range of signal transmission.

In very large buildings like shopping malls or skyscrapers, these tools are insufficient to alert everyone fast enough. In those cases alarm systems are needed. An alarm system consists of a control unit with a power supply, one or more alarm buttons and the signal generators. Those could be sounders, flashing lights or speakers as shown in Figure 2.12.

(42)

Figure 2.12: Components of an alarm system

In simple systems, the alarm is generated with acoustic and, if necessary, optical signalling devices. Acoustic and optical signals attract attention. However, they do not contain generally understandable information about the nature of the danger and do not provide instructions on how to behave. Simple alarm systems with sounders can be autonomous systems or they can be a part of the fire detection system.

The best, but also the most expensive alarm system is a voice alarm system. These systems can also be used for other operational purposes such as voice output, for example company announcements or music playback. In case of danger, an attention signal and text information about the type of danger and the required behaviour are alternately emitted. The announcements can be multilingual. When the alarm is activated by a fire alarm system, the electronically stored announcements are first broadcast. Later individual announcements of the fire brigade or the object management can follow. (Brandschutz-Wegweiser 2016)

Nevertheless, those options have to be checked for suitability as each facility requires special provisions. Taking the example of a nursing home, one has to consider that a loud audible alarm would lead to unease and panic. Therefore, it is reasonable to carry out only a silent alarm for the staff.

ALARM SYSTEMS IN BUILDINGS

VŠB - Technical University of Ostrava

Faculty of Safety Engineering

ALARM SYSTEMS IN BUILDINGS

Researches of effectiveness depending on the type of building and the user profile

Doctoral thesis

for acquiring the academic title “doctor”, abbreviated as “Ph.D.”

Abstract

Abstrakt

Kurzfassung:

Statutory declaration and consent to publication

Preface (motivation)

Table of contents

List of tables

List of figures

Glossary

List of abbreviations

1 Introduction

2 State of Knowledge and technology

t

= t

+ t

+ t

+ t