Technical ReportPDF Available

FIRESAFE II Detection systems in open ro-ro and weather decks

Authors:

Abstract and Figures

One of the main issues with regard to fire safety of open ro-ro spaces and weather decks is that detection systems may not be as efficient as in closed ro-ro spaces. Several recent total losses of ro-ro ships have stressed the need for investigating more efficient fire detection solutions. This study evaluated available and emerging fire detection technologies for use in open ro-ro spaces and on weather decks. A review of relevant regulations was performed as well as an evaluation of the expected efficiency of the identified alternative detection technologies, considering detection time and sensitivity to weather conditions, loading conditions and deck configuration, as well as cost. Fibre optic linear heat detection and thermal imaging camera detection were selected for fire tests in open ro-ro space and on weather deck, respectively, onboard a commercial RoPax vessel. Both systems were found functional and suitable for the relevant ro-ro space environments. The risk reduction potentials of the systems were quantified and a cost-effectiveness assessment was performed. Thermal imaging camera detection was found cost-effective for all types of RoPax (Existing ships and Newbuildings), and fibre optic linear heat detection system was found cost-effective for Standard and Ferry RoPax (Existing ships and Newbuildings).
Content may be subject to copyright.
<a
FIRESAFE II
Detection systems in open ro-ro
and weather decks
Version 2.2 December 2018
2/91
FIRESAFE II | Bureau Veritas RISE Stena
Contract No.:
2017/EMSA/OP/17/2017
Project Name:
Second study investigating cost-efficient measures for reducing the risk from
fires on ro-ro passenger ships (FIRESAFE II)
WP:
Part 4: Detection systems in open ro-ro and weather decks
Project Partners:
Bureau Veritas Marine & Offshore
RISE Research Institutes of Sweden, Fire Research
Stena Rederi
WP Leader:
RISE Research Institutes of Sweden
Authors:
Pierrick Mindykowski, RISE Research Institutes of Sweden
Jérome Leroux, Bureau Veritas Marine & Offshore
Ola Willstrand, RISE Research Institutes of Sweden
Blandine Vicard, Bureau Veritas Marine & Offshore
Franz Evegren, RISE Research Institutes of Sweden
Mattias Frösing, Stena Rederi
Lisa Gustin, Stena Rederi
Disclaimer:
The information and views set out in this study are those of the author(s) and do not necessarily
reflect the official opinion of EMSA. EMSA does not guarantee the accuracy of the data included in
this study. Neither EMSA nor any person acting on EMSA’s behalf may be held responsible for the
use which may be made of the information contained therein.
Rev.
Date
Author(s)
Reason for issue
0
08/12/2017
See list of authors
Draft version sent to EMSA for consideration
0.1
15/12/2017
Interim Report Deliverable
1.0
07/11/2018
Addition of alternative systems selection
tests results- risk reductions
2.0
08/11/2018
Final draft for internal review
2.1
16/11/2018
Final report sent to EMSA
2.2
19/12/2018
Revisions after receiving comments from
EMSA
Bureau Veritas RISE Stena | FIRESAFE II
3/91
1 ABSTRACT
One of the main issues with regard to fire safety of open ro-ro spaces and weather decks is that detection
systems may not be as efficient as in closed ro-ro spaces. Several recent total losses of ro-ro ships have
stressed the need for investigating more efficient fire detection solutions.
This study evaluated available and emerging fire detection technologies for use in open ro-ro spaces and on
weather decks. A review of relevant regulations was performed as well as an evaluation of the expected
efficiency of the identified alternative detection technologies, considering detection time and sensitivity to
weather conditions, loading conditions and deck configuration, as well as cost.
Fibre optic linear heat detection and thermal imaging camera detection were selected for fire tests in open
ro-ro space and on weather deck, respectively, onboard a commercial RoPax vessel. Both systems were
found functional and suitable for the relevant ro-ro space environments. The risk reduction potentials of the
systems were quantified and a cost-effectiveness assessment was performed. Thermal imaging camera
detection was found cost-effective for all types of RoPax (Existing ships and Newbuildings), and fibre optic
linear heat detection system was found cost-effective for Standard and Ferry RoPax (Existing ships and
Newbuildings).
4/91
FIRESAFE II | Bureau Veritas RISE Stena
2 EXECUTIVE SUMMARY
The main objective of Part 4 of the FIRESAFE II study was to investigate the possibilities and effectiveness
of installation of new fixed fire detection systems for open ro-ro spaces and weather decks of ro-ro passenger
ships (existing or newbuildings), with an aim to discuss specific proposals for rule-making.
In order to perform this investigation, a review of regulations concerning fire detection systems was
conducted. This review gave an overview of fire detection requirements applicable for ro-ro spaces of ro-ro
passenger ships, with a specific focus on weather decks and open ro-ro spaces, to provide a base for the
discussion.
Through consultation with detection specialists, both within and outside the maritime industry, available and
emerging fire detection technologies relevant for open ro-ro spaces and weather decks were identified. The
basic principles of each of the technologies are described in the report.
The relevance, benefits and limitations of the identified systems were evaluated during a workshop with all
project partners. A conclusion from the identification and first evaluation of the new fire detection systems
was that none of the identified systems seemed optimal for both types of ro-ro spaces.
For open ro-ro spaces, the following systems were judged to have the highest potential and were short-listed
for further evaluation:
Fibre optic linear heat detection;
Aspirating smoke detection (ASD); and
Gas detection, only in combination with ASD.
For weather decks, the systems judged to have the highest potential were:
Video detection: Smoke [or combined smoke] and flame] video detection
Video detection: Thermal imaging camera
Video detection: Flame video detection
Flame detection
With a view to assess the performance and adequacy of the short-listed systems for the different types of
ro-ro spaces, a further specific evaluation was carried out. Based on the results of a workshop and an
extensive literature review of fire detection tests, it considered two main aspects:
Relevance and applicability for open ro-ro spaces and weather decks;
Qualitative evaluation of the activation time and its sensitivity to cargo configuration depending of
the deck configuration.
In order to have a first assessment of the cost-effectiveness of the systems, costs of each of the systems
(for components and installation for a given ship) were also estimated through quotations from relevant
detection system manufacturers. While the cost of a conventional point smoke and heat detection system
for open ro-ro spaces was estimated to € 55 000, the estimated total costs for the short-listed new systems
with judged high potential varied from € 50 000 to € 105 000. For the systems considered for weather decks,
the costs varied from € 65 000 to € 150 000.
Based on the evaluation of the short-listed systems, a selection process was carried out based on a decision-
support matrix, to select the two alternative detection systems to be tested. The results of this selection were
unanimous among the different partners of the project, with the Fibre Optic Linear Heat system selected for
the open ro-ro space and the thermal imaging camera system selected for the weather deck.
Both selected systems were evaluated in fire tests onboard a commercial RoPax vessel with a weather deck
and an open ro-ro space (the latter with a conventional heat and smoke detection system). A Liquefied
Petroleum Gas burner was used to reproduce a cargo fire and several fire scenarios were evaluated to
challenge both detection systems.
Both systems were found functional and suitable for the relevant ro-ro space environments. On the weather
deck, the thermal imaging camera detected a relatively small fire (80 kW) at a distance of around 50 meters.
It was also capable of detecting a fire when the gas burner was fully obstructed (after around 3 minutes for
Bureau Veritas RISE Stena | FIRESAFE II
5/91
a fire of 300 kW) or half obstructed (after half a minute for a fire of 80 kW). Even when heavy rain was
simulated between the fire source and the thermal imaging camera, fire was detected (after 4 minutes for a
fire of 400 kW). For the open ro-ro space, the fibre optic linear heat detection system showed capacities to
detect a fire faster than a conventional point heat detection system. The improved performance was judged
to be mainly attributed to the used detection criterion, based on a rate of temperature rise instead of a given
critical temperature. Furthermore, the new system also provided improved coverage and thus a shorter
required traveling distance of the hot gases, contributing to further shortened detection times.
The second part of the onboard tests was to test the false alarm rate of each alternative systems. The
systems were therefore left onboard the test vessel for one month. The vessel did not record any fire or give
any alarm from the conventional heat and smoke system nor from the fibre optic linear heat detection system.
The thermal imaging camera system recorded many alarms during period, but only during cargo loading and
unloading.
Based on the onboard fire detection tests as well as simulation studies performed in the FIRESAFE II study,
the risk reduction potential was assessed for each selected detection system. For the fibre optic linear heat
detection system, the detection fault tree developed within FIRESAFE II for open ro-ro spaces was used to
quantify the effects on each failure node probability. As no fixed fire detection system is required for weather
deck, a new fault tree was developed to assess the risk reduction from the thermal imaging camera fire
detection system.
The costs for the implementation of the selected fire detection systems were estimated in further detail.
Technical items available on the market were as far as possible quantified by offers from system
manufacturers. In addition, the cost estimations were contributed by previous cost assessments from internal
projects, specifications, reconstructions, etc. A cost-effectiveness assessment was performed, and both of
the selected systems were found cost-effective assuming a Gross Cost of Averting a Fatality (GCAF) of
€7 M.
The findings of the cost-effectiveness assessment are summarised in the below table.
Newbuildings
Existing Ships
Detection system
Cargo
RoPax
Standard
RoPax
Ferry
RoPax
Cargo
RoPax
Standard
RoPax
Ferry
RoPax
Fibre optic linear heat
detection
Not cost-
effective
Cost-
effective
Cost-
effective
Not cost-
effective
Cost-
effective
Cost-
effective
Thermal Imaging
Camera
Cost-
effective
Cost-
effective
Cost-
effective
Cost-
effective
Cost-
effective
Cost-
effective
6/91
FIRESAFE II | Bureau Veritas RISE Stena
3 CONTENTS
1 ABSTRACT 3
2 EXECUTIVE SUMMARY 4
3 CONTENTS 6
4 TABLE OF FIGURES 9
5 TABLE OF TABLES 11
6 INTRODUCTION 12
6.1 Scope and Objectives 12
6.2 Background 12
6.3 Methodology 12
7 REVIEW OF REGULATIONS 14
7.1 Reference documents 14
7.2 Definitions 15
7.3 Requirements 16
7.4 Fixed fire detection systems for outside areas 20
8 IDENTIFICATION OF RELEVANT ALTERNATIVE
DETECTION SYSTEMS 22
8.1 Presentation of possible relevant fire detection
systems 22
8.2 Selection of alternative fire detection system 26
8.3 Conclusion of the selection of relevant alternative
fire detection systems 28
9 EVALUATION OF RELEVANCE/APPLICABILITY OF
SYSTEMS IN OPEN RO-RO SPACE AND WEATHER DECK 29
9.1 Open ro-ro space 29
9.2 Weather deck 30
Bureau Veritas RISE Stena | FIRESAFE II
7/91
10 EVALUATION OF ACTIVATION TIME 32
10.1 Literature on activation time of different fire
detection systems 32
10.2 Deck loading configuration and its influence on
activation time 36
11 EVALUATION OF COSTS 39
11.1 Description of the decks used for the cost
evaluation 39
11.2 Cost estimations of alternative fire detection
systems 41
12 SELECTION OF ALTERNATIVE DETECTION SYSTEMS
FOR TESTING 50
12.1 Selection criteria 50
12.2 System selected for the weather deck 51
12.3 System selected for the open ro-ro space 52
13 TESTING OF ALTERNATIVE FIRE DETECTION SYSTEMS
53
13.1 The test vessel - Stena Scandinavica 53
13.2 Detection time test setup 54
13.3 False alarm test setup 61
13.4 Detection time test results 61
13.5 False alarm test results 64
13.6 Discussion of fire detection tests 65
13.7 Fire detection test acknowledgements 68
14 RISK REDUCTION FROM SELECTED ALTERNATIVE
SYSTEMS 69
14.1 Translation of tests in term of detection failure
probabilities 69
14.2 Estimation of Risk Reduction by the implementation
of the selected systems 75
8/91
FIRESAFE II | Bureau Veritas RISE Stena
15 COST-EFFECTIVENESS ASSESSMENT OF THE
SELECTED SYSTEMS 79
15.1 Cost-effectiveness assessment background 79
15.2 Estimation of costs 79
15.3 GCAF ratio for the selected systems 81
16 PROPOSALS FOR RULE-MAKING / DISCUSSION 84
16.1 Analysis of regulations for fire detection system for
weather deck 84
16.2 Analysis of regulations for alternative fire detection
system 84
16.3 Analysis of regulations for new criterion for heat
type fire detectors 85
17 CONCLUSION 87
18 REFERENCES 89
19 LIST OF ABBREVIATIONS 90
A1 ANNEXES 91
A1.1 List of the participants of the workshop for the
selection of relevant alternative fire detection
systems 91
Bureau Veritas RISE Stena | FIRESAFE II
9/91
4 TABLE OF FIGURES
Figure 1. Possible spacing of accumulators (sampling points) fulfilling the regulations. ............................. 30
Figure 2. Typical fire scenario with respect to fire detection (or prevention with respect to gas). ............... 32
Figure 3. Example of video blockage due to high cargo. ............................................................................. 37
Figure 4. Example of smoke-detector traveling time for two deck heights. .................................................. 38
Figure 5. Example of accelerated air flow in case of high height cargo. ...................................................... 38
Figure 6. Picture of the Standard RoPax. ..................................................................................................... 39
Figure 7. General arrangement of the open ro-ro space (deck 4) of the Standard RoPax. ......................... 40
Figure 8. Picture of the Cargo RoPax. ......................................................................................................... 40
Figure 9. View of the weather deck of the Cargo RoPax. ............................................................................ 41
Figure 10. General arrangement of deck of the Cargo RoPax. .................................................................... 41
Figure 11. The aft of the test vessel showing the weather deck of deck 7 and the aft opening of the open ro-
ro space. .......................................................................................................................................... 53
Figure 12. General arrangement of deck 7 of the test vessel. ..................................................................... 53
Figure 13. Side openings of the open ro-ro space. ...................................................................................... 54
Figure 14. Measured and specified HRR ..................................................................................................... 55
Figure 15. Position of the thermal imaging camera on the test vessel. ....................................................... 55
Figure 16. Fire test arrangement for the thermal imaging camera on the weather deck. ............................ 56
Figure 17. Positions of the fibre optic cable and existing conventional detectors in the ceiling of the open ro-
ro space. .......................................................................................................................................... 57
Figure 18. Illustration of the detection zones of the fibre optic linear heat detection system and existing
detectors. ......................................................................................................................................... 57
Figure 19. Position of the fibre optic cable in relation to existing conventional detectors. ........................... 58
Figure 20. Existing detector (circled in red) and fibre optic cable positioned at a side opening. ................. 58
Figure 21. Positions of the fire in relation to existing detectors and wind influence. .................................... 59
Figure 22. Position of the fibre optic cable on the floor. ............................................................................... 59
Figure 23. Picture of fire detection for test 2 Thermal imaging camera. ................................................... 61
Figure 24. Illustration of heavy rain from the fire monitor. ............................................................................ 62
Figure 25. Screen shot at the detection time for test 4. ............................................................................... 62
Figure 26. Screen shot for an alarm during loading of cargo. ...................................................................... 64
Figure 27. Fire positions, conventional detectors positions and fire zone of the fibre optic system. ........... 66
Figure 28. Illustration of the smoke (circled in red) produced by the burner. ............................................... 68
Figure 29. Updated Main Fire Risk Model for the Standard RoPax Newbuilding (Open ro-ro spaces part) 70
Figure 30. Detection fault tree (closed and open ro-ro spaces). .................................................................. 71
Figure 31. Sub-tree for System detection failure (Closed and open ro-ro spaces). ..................................... 71
Figure 32. Sub-tree for Late/no manual detection (closed and open ro-ro spaces). .................................... 72
Figure 33. Sub-tree for System detection failure for weather deck. ............................................................. 74
10/91
FIRESAFE II | Bureau Veritas RISE Stena
Figure 34. Relative Risk Reduction of the two selected systems on Newbuildings (lighter blue and red
corresponds to Existing ships) ......................................................................................................... 76
Figure 35. Relative Risk Reduction of the two selected systems on Newbuildings ..................................... 76
Figure 36. Relative Risk Reduction of the two selected systems on Existing ships .................................... 77
Figure 37. Relative risk reduction induced by the thermal imaging camera with and without considering the
side effects of its implementation (on the other parts of the main fire risk model) .......................... 78
Bureau Veritas RISE Stena | FIRESAFE II
11/91
5 TABLE OF TABLES
Table 1. List of documents used for the review of regulations of fire detection requirements applicable in ro-
ro spaces of ro-ro passenger ships ................................................................................................. 14
Table 2. Spacing of detectors (FSS Ch. 9 Table 9.1) .................................................................................. 18
Table 3. Pros and cons for fibre optic linear heat detection ......................................................................... 23
Table 4. Pros and cons for aspirating smoke detection ............................................................................... 24
Table 5. Pros and cons for gas detection (flammable gases and fire products) .......................................... 24
Table 6. Pros and cons for flame detection .................................................................................................. 25
Table 7. Pros and cons for video detection .................................................................................................. 25
Table 8. Pros and cons for light beam linear smoke detection .................................................................... 26
Table 9. Detector performance in Törnskogstunneln, Stockholm. ............................................................... 33
Table 10. Detector performance in Northern Link tunnel, Stockholm. ......................................................... 33
Table 11. Detector performance in bus toilet compartment. The response times are given in seconds after
ignition. ............................................................................................................................................. 34
Table 12. Video detector performance in a road tunnel [10]. ....................................................................... 35
Table 13. Video detector performance in a road tunnel [11]. ....................................................................... 35
Table 14. Summary table of the evaluation of costs of alternative fire detection systems........................... 49
Table 15. Illustration of the selection process by a decision-support matrix. ............................................... 50
Table 16. Decision-support matrix for the selection of the detection systems. ............................................ 51
Table 17. Results of the selection process for the alternative detection system for the weather deck. ....... 52
Table 18. Results of the selection process for the alternative detection system for the open ro-ro space. . 52
Table 19. Fire growth parameters ................................................................................................................ 54
Table 20. Fire tests procedure ...................................................................................................................... 60
Table 21. Results of the fire tests for the thermal imaging camera. ............................................................. 61
Table 22. Results of the fire tests for fibre optic linear heat detection system. ............................................ 63
Table 23. Reduction of the failure probabilities for the nodes impacted by the criterion based on temperature
gradient and use of the linear fibre optic heat detection system for closed (CRS) and open (ORS) ro-
ro spaces. ........................................................................................................................................ 73
Table 24. ΔRisk, ΔCosts, GCAF and GCAF Factor values for the selected systems on reference vessels
(Newbuildings) ................................................................................................................................. 81
Table 25. Extended cost-effectiveness assessment: ΔRisk, ΔCosts, GCAF and GCAF Factor values for the
selected systems on generic ships (Newbuildings) ......................................................................... 81
Table 26. ΔRisk, ΔCosts, GCAF and GCAF Factor values for the selected systems on reference vessels
(Existing ships) ................................................................................................................................ 82
Table 27. Extended cost-effectiveness assessment: ΔRisk, ΔCosts, GCAF and GCAF Factor values for the
selected systems on generic ships (Existing ships) ........................................................................ 82
Table 28. Response time limits for the rate of temperature rise of different classes of heat type fire detectors
......................................................................................................................................................... 86
12/91
FIRESAFE II | Bureau Veritas RISE Stena
6 INTRODUCTION
6.1 Scope and Objectives
The main objective of FIRESAFE II was to improve the fire safety of ro-ro passenger ships by cost-efficient
safety measures reducing the risk of ro-ro space fires, with an aim to discuss specific proposals for rule-
making. In part 4 of the study, reported here, the objective is to investigate the possibilities and effectiveness
of alternative fixed fire detection systems in open ro-ro decks and on weather decks of passenger ships
(newbuildings and existing).
6.2 Background
In 2016, EMSA initiated the FIRESAFE study in order to investigate cost-efficient measures for reducing the
risk from fires on ro-ro passenger ships with a focus on Electrical Fire as ignition source as well as Fire
Extinguishing Failure. These areas were considered the greatest risk contributors by the EMSA Group of
Experts on fires on ro-ro decks.
The study also produced a coarse risk model covering the various stages of a fire incident on a ro-ro
passenger ship, namely: ignition, detection/decision, extinguishment, containment and evacuation.
In 2017, EMSA initiated the FIRESAFE II study to investigate risk control options for mitigating the risk from
fires on ro-ro decks in relation to Detection and Decision (Part 1) as well as Containment and Evacuation
(Part 2), which were not specifically addressed in FIRESAFE.
Early detection is often cited as key to preventing loss of life, ship and extensive cargo damage. However,
one of the main issues with open ro-ro spaces and weather decks is that detection systems may not be as
efficient as in enclosed ro-ro spaces. To address this hazard, EMSA launched a specific part focusing
specifically on detection systems in open ro-ro and weather decks, running in parallel with the two
aforementioned parts.
Except the lack of a required detection system on weather deck, the main challenge for open ro-ro spaces
and weather decks concerns the potentially significant ventilation, which can delay and delocalize detection.
There are also challenges with regard to the potential fire scenario. The fire will be well-ventilated and for
open ro-ro spaces the steel deck directly above the cargo reflects heat and accumulates smoke, which
promotes fire spread. There are also notable challenges with regard to escape ways, location of live-saving
appliances and air intakes to the engine room and emergency generator space, which can be contaminated
and damaged by smoke.
These challenges have materialized in recent years through several total losses of ro-ro ships with open ro-
ro spaces (e.g. Norman Atlantic, Sorrento, Lisco Gloria, Und Adriyatik).
6.3 Methodology
In order to achieve the objective described in section 6, a five step methodology was followed. Details of the
steps are provided below:
1st step: Desk study to evaluate the efficiency of available and emerging fire detection technologies
for use in open ro-ro and on weather decks;
2nd step: Selection of the system expected to have the best performance in combination with a
feasible cost;
3rd step: Testing of the selected system in order to measure the expected risk reduction in relation
to a conventionally expected detection times;
4th step: Cost-effectiveness assessment for the selected systems; and
5th step: If relevant, development of specific proposals for rule-making.
Bureau Veritas RISE Stena | FIRESAFE II
13/91
In this report, alternative systems refer to available and emerging fire detection systems for use in open ro-
ro and on weather decks, which are different from the systems commonly used
1
.
Regarding open ro-ro spaces, it should be noted that the alternative systems may also be studied as an add-
on to the current systems.
1
In practice, in open ro-ro spaces, the most common type of detection system relies on smoke detectors and there is no requirement
for detection system on weather decks
14/91
FIRESAFE II | Bureau Veritas RISE Stena
7 REVIEW OF REGULATIONS
The present review aims to give an overview of fire detection requirements applicable for ro-ro spaces of ro-
ro passenger ships, with a specific focus on weather decks and open ro-ro spaces.
7.1 Reference documents
Requirements related to fire detection are mainly covered by IMO regulations and in a few IACS texts. As a
general remark, there are few specific requirements related to fire detection in rules from Classification
Societies and in national regulations and interpretations. Therefore, this review is mainly based on the IMO
and IACS documents presented in the Table 1.
Table 1. List of documents used for the review of regulations of fire detection requirements applicable in ro-ro
spaces of ro-ro passenger ships
IMO
Documents
Safety of Life at Sea (SOLAS) Convention, as amended in 2017
Fire Safety Systems (FSS) Code, as amended in 2017
MSC/Circ.1035 Guidelines for the use and installation of detectors equivalent to smoke
detectors
MSC.1/Circ.1242 Guidelines for approval of fixed fire detection and fire alarm systems
for cabin balconies
MSC.1/Circ.1369 Interim explanatory notes for the assessment of passenger ship
systems’ capabilities after a fire or flooding casualty
MSC.1/Circ.1430 Revised guidelines for the design and approval of fixed water-based
fire-fighting systems for ro-ro spaces and special category spaces, May 31, 2012
MSC.1/Circ.1437 Unified interpretation of SOLAS II-2/21.4
IACS
Documents
UI SC35 rev.3 July 2013 “Fixed Fire Detection and Fire Alarm System”
UI SC73 rev.2 Nov. 2005 “Fire protection of weather decks”
UI SC117 rev.2 Nov. 2005 “Fire detection system with remotely and individually
identifiable detectors”
UR E22 rev.2 June 2016 “On Board Use and Application of Computer based systems”
Classification
Rules
BV Rules for Steel Ship (NR467), as amended in January 2018
BV NR598 “Implementation of Safe Return to Port and Orderly Evacuation” dd. January
2016
DNVGL Rules for the Classification of Ships, January 2017
LR Rules and Regulations for the Classification of Ships, July 2016
NKK Rules for the Survey and construction of Steel Ships, June 2016
MMF (French Flag Administration) Division 221 “Passenger ships engaged in international
voyages and cargo ships of more than 500 gross tonnage”, 04/08/17 edition
Bureau Veritas RISE Stena | FIRESAFE II
15/91
Flag
Administration
Regulations
US Coast Guard Code of Federal Regulations (CFR) 46, 2017 online edition
Swedish Transport Agency (Swedish Flag Administration) “Comments and interpretations
by the Swedish Transport Agency regarding IMO Conventions”, version 03 dd.15/05/2017
MCA (UK Flag Administration) Guidance on SOLAS Ch.II-2
Other
documents
Cruise Lines International Association (CLIA) industry policy on fire detection on covered
mooring decks as exposed on
https://www.cruising.org/about-the-
industry/regulatory/industry-policies/fire-protection/fire-protection-measures-for-covered-
mooring-decks
MSC 96/6/1 submitted by CLIA at IMO MSC 96 (2016) “Fire protection in category "A"
machinery spaces and on covered mooring decks”
It should be noted that the present review is based on the currently applicable regulations. Therefore, some
of the requirements detailed below may not be applicable on old ships. As an indication, FSS Code Chapter
9, dedicated to fixed fire detection systems has been fully reviewed through MSC.311(88) and applies to
ships of which the keel was laid after 01/07/2012.
7.2 Definitions
7.2.1 Ro-ro space, vehicle space and special category space
Cargo spaces are spaces used for cargo, cargo oil tanks, tanks for other liquid cargo and trunks to such
spaces. (SOLAS II-2/3.8)
Ro-ro spaces are a type of cargo spaces, defined accordingly (SOLAS II-2/3.41):
Ro-ro spaces are spaces not normally subdivided in any way and normally extending to either a substantial
length or the entire length of the ship in which motor vehicles with fuel in their tanks for their own propulsion
and/or goods (packaged or in bulk, in or on rail or road cars, vehicles (including road or rail tankers), trailers,
containers, pallets, demountable tanks or in or on similar stowage units or other receptacles) can be loaded
and unloaded normally in a horizontal direction
2
.
Vehicle spaces are cargo spaces intended for carriage of motor vehicles with fuel in their tanks for their own
propulsion. (SOLAS II-2/3.49)
Special category spaces are those enclosed vehicle spaces above and below the bulkhead deck, into and
from which vehicles can be driven and to which passengers have access. Special category spaces may be
accommodated on more than one deck provided that the total overall clear height for vehicles does not
exceed 10 m. (SOLAS II-2/3.46)
Special category spaces are the most frequent type of closed ro-ro spaces on ro-ro passenger ships.
7.2.2 Closed, open and weather deck
One of the most important definitions for the current study is the definition of closed and open ro-ro space
and weather deck. Ro-ro spaces can be divided in these three categories depending on how they are
enclosed:
Weather deck is a deck which is completely exposed to the weather from above and from at least
two sides. (SOLAS II-2/3.50)
2
In other words, ro-ro spaces are vehicle spaces into which vehicles can be driven. It is to be noted however that, for the purpose of
the application of SOLAS II-2/19, the following interpretation can be found in MSC.1/Circ.1120 and IACS UI SC 85: Ro-ro spaces
include special category spaces and vehicle spaces”
16/91
FIRESAFE II | Bureau Veritas RISE Stena
Open ro-ro spaces are those ro-ro spaces which are either open at both ends or have an opening
at one end and are provided with adequate natural ventilation effective over their entire length
through permanent openings distributed in the side plating or deckhead or from above, having a total
area of at least 10% of the total area of the space sides. (SOLAS II-2/3.35)
Closed ro-ro spaces are ro-ro spaces which are neither open ro-ro spaces nor weather decks.
(SOLAS II-2/3.12)
SOLAS states that a weather deck is a deck which is completely exposed to weather from above and from
at least two sides. IACS UI SC 86 additionally details that: “For the purposes of Reg. II-2/19 a ro-ro space
fully open above and with full openings in both ends may be treated as a weather deck.” For practical
purposes, fixed fire-extinguishing systems cannot be fitted on weather decks due to the absence of
deckhead. Therefore, this criterion is often used for a practical definition of weather decks.
It can be noted that ro-ro spaces with less than 10% openings are considered closed, even though such a
deck can have significant openings. Furthermore, one deck can include several categories of ro-ro spaces
and the borders between for example weather deck and closed deck can be vague.
As a reference criterion, it can be considered that a vehicle space that needs mechanical ventilation is a
closed vehicle space.
7.3 Requirements
Below are presented different requirements for systems, their performance, the system arrangement,
including required fire detectors and electrical arrangements.
7.3.1 Type of systems, spaces to be covered
7.3.1.1 General requirement
SOLAS II-2/20.4.1 requires a fixed fire detection and fire alarm system to be fitted in all ro-ro spaces.
It is widely accepted however that no fixed fire detection and fire alarm system is required on weather decks
used for the carriage of vehicles with fuel in their tanks, as per IACS interpretation UI SC73.
It is to be noted that fire detection is required in open ro-ro spaces (although some discussion on this point
regularly arises at the shipbuilding phase).
7.3.1.2 Special category spaces
In special category spaces, SOLAS II-2/20.4.3.1 allows that a fixed fire detection and fire alarm system is
replaced by an efficient fire patrol system, maintained by a continuous fire watch at all times during the
voyage.
It is to be noted that some Flag States require a fixed fire detection system, independently of the existence
of continuous fire watch (e.g. French Flag).
7.3.1.3 Type of fixed fire detection system
SOLAS II-2/20.4.1 requires a standard fixed fire detection and alarm system in line with the FSS Code
requirements. For practical purposes, it can be noted that sample extraction smoke detection systems are
not allowed on ro-ro passenger ships since SOLAS II-2/20.4.2 prohibits such systems
3
in open ro-ro spaces,
open vehicle spaces and special category spaces”. Therefore, this regulation overview will focus on fixed
fire detection and fire alarm systems as described in Chapter 9 of the FSS Code.
In addition, on passenger ships constructed on or after 1 July 2010, the system is to be addressable i.e.
capable of identifying remotely and individually each detector and manually operated call point (FSS Code
3
Sample extraction smoke systems have been prohibited in SOLAS 1989 amendments (MSC.13(57)), applicable to ships constructed
on or after 1 February 1992. As far as BV knows, this was a consequence of the bad service conditions observed on ro-ro ships for
such systems (pipe ageing and corrosion) which usually had a common steel piping with the gas fire-extinguishing system.
Bureau Veritas RISE Stena | FIRESAFE II
17/91
Ch 9 §2.1.7). Before 2010, the fixed fire detection system was required to be divided in sections, and to be
able to indicate in which section a detector has been activated.
7.3.1.4 Fire patrol
Efficient fire patrols are required as per SOLAS II-2/7.8 and SOLAS II-2/20.4.3.1. On passenger ships
carrying more than 36 passengers, it is made clear that each member of the fire patrol is to be provided with
a two-way portable radiotelephone apparatus, properly trained and familiar with the ship.
7.3.2 Performance
7.3.2.1 General
SOLAS II-2/20.4.1 sets the following general performance requirements:
“The fixed fire detection system shall be capable of rapidly detecting the onset of fire”
“After being installed, the system shall be tested under normal ventilation conditions and shall give
an overall response time to the satisfaction of the Administration”
Common practice as per Bureau Veritas field experience is to test the system using a smoke generator. A
usual criterion is that the fire detection system is to be activated within 3 minutes.
A similar criterion can be found in French Flag Regulations (div 221-II-2/7.4) and Bureau Veritas Rules
(NR467 Pt F, Ch 3, Sec 1 [3.2.15]) for fire detection in unattended machinery spaces.
FSS Code Ch.9 §2.1.2 lists the following main functionalities for the fire detection system:
“control and monitor input signals from all connected fire and smoke detectors and manual call
points;
provide output signals to the navigation bridge, continuously manned central control station or on-
board safety centre to notify the crew of fire and fault conditions;
monitor power supplies and circuits necessary for the operation of the system for loss of power and
fault conditions; and
the system may be arranged with output signals to other fire safety systems” (communication, alarm
and public address systems, ventilation, fire doors and fire dampers, fire extinguishing and systems
supporting evacuation such as Low Location Lighting (LLL))
7.3.2.2 Maintenance
In-service testing and proper maintenance are required in FSS Ch 9 §2.5.2, SOLAS II-2/7.3 & SOLAS II-
2/14.2.2.
7.3.3 Prescriptive detection system arrangement
7.3.3.1 Location of detectors
SOLAS II-2/20.4.1 clarifies that the “spacing and location [of the detectors] shall [… take] into account the
effects of ventilation and other relevant factors”. Further detail is provided in FSS Ch 9 §2.4.2, together with
a table summarizing the maximum spacing between detectors:
Detectors shall be located for optimum performance. Positions near beams and ventilation ducts, or other
positions where patterns of air flow could adversely affect performance, and positions where impact or
physical damage is likely, shall be avoided. Detectors shall be located on the overhead at a minimum
distance of 0.5 m away from bulkheads, except in corridors, lockers and stairways.
18/91
FIRESAFE II | Bureau Veritas RISE Stena
Table 2. Spacing of detectors (FSS Ch. 9 Table 9.1)
Type of
detector
Maximum floor area per
detector (m2)
Maximum distance apart
between centres (m)
Maximum distance away
from bulkheads (m)
Heat
37
9
4.5
Smoke
74
11
5.5
It can be noted that FSS Code requirements for detector location are applicable for all kinds of spaces; they
are not specific for ro-ro spaces. Furthermore, in case the fixed fire extinguishing system is a manual deluge
system, automatic deluge system or pre-action system, MSC.1/Circ.1430 makes it clear that:
only smoke or heat detectors are allowed below hoistable ramps; and
reduced spacing is to be considered for spot-type heat detectors where beams project more than
100 mm below the deck.
7.3.3.2 Section arrangement
Fire detection sections are not allowed to cover more than one MVZ (FSS Ch 9 §2.4.1.4). In addition, a fire
detection section covering a ro-ro space is to be separated from (FSS Ch 9 §2.4.1.2):
Control station
Service spaces
Accommodation spaces
For practical purposes, this means that ro-ro spaces are to be provided with dedicated fire detection sections,
since ro-ro spaces generally are located in a dedicated Main Horizontal Zone. Only machinery spaces other
than category A located in the same horizontal zone can be covered by the same detection section.
In addition, in case the fixed fire extinguishing system is a manual deluge system, automatic deluge system
or pre-action system, MSC.1/Circ.1430 requires that fire detection sections be the same as the zones of the
fixed fire-extinguishing system: “The area of coverage of the detection system sections should correspond
to the area of coverage of the extinguishing system sections.”
For practical purposes, on addressable fire detection and fire alarm systems, several sections may be
arranged in series on the same electrical cable and separated by suitably located isolators.
7.3.4 Fire detectors
7.3.4.1 General
The fire detection system is to include fire detectors and manually operated call points.
FSS Code Ch 9 §2.1.5: All components to be qualified for operation in marine environment (standard
requirements for electrical equipment onboard ships). In addition fire detectors located in hazardous areas
4
are to be adequate for such use (FSS Ch 9 §2.3.1.8).
7.3.4.2 Type of detectors
The FSS Code allows Detectors […] operated by heat, smoke or other products of combustion, flame, or
any combination of these factors.” (FSS Ch 9 §2.3.1.1)
As a complement, in case the fixed fire extinguishing system is a manual deluge system, automatic deluge
system or pre-action system, MSC.1/Circ.1430 requires that two types of fire detectors are combined.
In addition, it may be noted that several Flag States and classification societies require smoke detectors
exclusively or in combination with other detectors in ro-ro spaces. BV Rules require that smoke detectors
4
For practical purposes, fire detectors installed in ro-ro spaces below the bulkhead deck are in Zone 1, others are in Zone 2, since fire
detectors are fitted on the deckhead.
Bureau Veritas RISE Stena | FIRESAFE II
19/91
are installed in ro-ro spaces (NR467 Pt C, Ch 4, Sec 12 [3.1.1]). Similar requirements are given by the US
Coast Guard and the Swedish Flag. The MCA (UK Flag Administration) requires smoke detectors exclusively
or a combination of smoke and flame detectors.
The requirement to have at least smoke detection in ro-ro spaces is based on the fact that smoke detection
is considered as more reliable than standard flame or heat detectors. Standard heat or flame detectors are
also considered less efficient in ro-ro spaces since:
Heat sensors located on garage space deckhead were expected to result into quite long activation
times due to deck height
Flame detectors were expected to lead to a number of false alarms due to reflections etc.
7.3.4.3 Qualification and performance standards
In general, fire detectors are to be qualified according to EN 54:2001 and IEC 60092:504 (FSS Ch 9 §2.3.1
and MSC/Circ.1035). Usual performance requirements are:
For smoke detectors: Activation for 2% obscuration/m smoke density ≤ 12.5% obscuration/m
“Smoke detectors […] shall be certified to operate before the smoke density exceeds 12.5%
obscuration per metre, but not until the smoke density exceeds 2% obscuration per metre”
Heat detectors: Activation when 54°C temperature 78°C
“Heat detectors shall be certified to operate before the temperature exceeds 78ºC but not until the
temperature exceeds 54ºC, when the temperature is raised to those limits at a rate less than 1ºC
per minute.
Carbon monoxide detectors: Alarm threshold set at 40ppm, sensitivity settings to be adjusted
considering the fire hazard, likely source and risk of false alarm.
In addition, provisions are given for in-service function testing (FSS Ch 9 §2.3.1.6).
7.3.5 Electrical arrangement
7.3.5.1 System architecture
The system is to be organized into sections as per FSS Code Ch 9 §2.1.4 and 2.4.1.1.
The first initiated fire alarm is not to prevent any other detector from initiating further fire alarms as per FSS
Code Ch 9 §2.1.6.3, applicable to addressable systems.
7.3.5.2 Components
The control panel is to be tested according to standards EN 54-2:1997, EN 54-4:1997 and IEC
60092-504:2001 (FSS Ch 9 §2.3.2)
Cables are to be flame retardant as per IEC 60332-1 (FSS Ch 9 §2.3.3)
Cables routed through MVZ that they do not serve and cables to control panels in an unattended fire
control station are to be fire resisting as per IEC 60331 (FSS Ch 9 §2.3.3)
7.3.5.3 Sources of power
7.3.5.3.1 Continuous fire detection capability
The fixed fire detection and fire alarm system is to be fed from two sources of power with separate feeders,
including an emergency source of power (FSS Code Ch 9 §2.2.1). An emergency source of power has to
comply with the requirements of SOLAS II-1/42 and 42-1 regarding location and autonomy. Especially, it has
to be able to supply the fire detection system for 36 hours, after which it has to be capable of operating the
fire alarm for 30min (FSS Ch 9 §2.2.4). It is either the ship emergency generator (+ transitional source of
emergency power) or dedicated accumulator batteries (FSS Ch 9 §2.2.4 & 2.2.5).
20/91
FIRESAFE II | Bureau Veritas RISE Stena
An automatic change-over switch is to be provided to manage the transition between the main and
emergency source of power, and a fault should not lead to the loss of both power supplies.
No temporary loss of the fire detection capability due to this change-over switch is accepted. In addition, a
transitional battery may be required if the temporary loss of power can damage the fire detection system as
per FSS Ch 9 §2.2.2.Although the alarm sounder is not formally required to be part of the fire detection
system, IACS UI SC35 makes it clear that it is to be powered from a main and emergency source of power
and from the transitional source of emergency power where required.
7.3.5.3.2 Sizing of the source of power
The power supply is to be sufficient for operation with 100 detectors activated, or all detectors provided
onboard if this number is lower than 100 (FSS Ch 9 §2.2.3).
7.3.5.4 Consequences of a fault
After an electrical fault or electrical failure:
Identification capability is to be kept in the whole section, except for the faulty detector (FSS Code
Ch 9 §2.1.6.1, applicable to addressable systems)
The initial configuration is to be restored (FSS Code Ch 9 §2.1.6.2, applicable to addressable
systems)
7.3.5.5 Temporary disconnection
FSS Code Ch 9 §2.1.1 allows temporary disconnection of the fire detectors in ro-ro spaces during loading
and off-loading, provided:
Detectors in other spaces remain operational
Fire patrol is maintained in the ro-ro space while the detectors are disconnected
The detectors are automatically re-connected after a pre-set duration
MCA (UK Flag Administration) clarify in their guidance that:
Manual call points and manual release mechanisms may not be disconnected
The duration of the timer is to be adapted to the time of loading/unloading
The central unit is to indicate whether the detector sections are disconnected or not
7.4 Fixed fire detection systems for outside areas
As detailed in 7.3 above, fixed fire detection systems shall be provided in enclosed and open ro-ro spaces
of ro-ro passenger ships, but not on weather decks.
This section lists existing regulations or documentation that could be used in order to propose solutions that
might be relevant for open ro-ro decks and weather decks.
7.4.1 Fire detection for cabin balconies
Fire detection is required on passenger ship cabin balconies unless the furniture and furnishings are of
restricted fire risk as a result of SOLAS
5
2006 amendments (See SOLAS II-2/7.10). Since they are installed
in non-enclosed spaces, these systems can be a useful reference with respect to fire detection in open ro-
ro space and on weather decks.
Fire detection systems on cabin balconies are to be in line with MSC.1/Circ.1242. Basically, these guidelines
are very close to chapter 9 of the FSS Code, as detailed in paragraph 7.3. The following requirements can
be stressed, since they are related to adaptation to an outdoor location:
5
See Res. MSC.216(82).
Bureau Veritas RISE Stena | FIRESAFE II
21/91
2.3 The system should be capable of fire detection on cabin balconies with expected wind conditions while
the vessel is underway
2.4 […] External components should additionally be designed to withstand sun irradiation, ultraviolet
exposure, water ingress and corrosion normally encountered on open deck areas.
2.6 The location and spacing of the detectors should be within the limits tested.
For practical purposes, fire detection systems installed on cabin balconies rely on flame detection
technology, since it is not affected by wind as heat or smoke detectors would.
7.4.2 Fire detection on covered mooring decks
Following a fire on the aft mooring deck of the Grandeur of Seas, CLIA recommends to its members to
implement a fixed fire detection system and a fixed fire extinguishing system on covered mooring decks.
Flame and smoke sensor technology integrated into CCTV systems are deemed effective as fire detection
systems for covered mooring decks, and more adequate than conventional fire/smoke detection systems.
22/91
FIRESAFE II | Bureau Veritas RISE Stena
8 IDENTIFICATION OF RELEVANT ALTERNATIVE
DETECTION SYSTEMS
In order to identify relevant alternative fire detection systems, a workshop was conducted. The participants
represented a wide field of expertise, a list of which is provided in annex A1.1.
The workshop was organized in three steps. The first step consisted in establishing the current means for
detection (by a regulatory review) as well as the challenges for detection in open ro-ro spaces and weather
decks. Thereafter, a list of possible alternative fire detection systems was created, along with their
advantages and disadvantages. The third step entailed discussions on the applicability of each system for
the different types of ro-ro decks. Based on the previous steps, a list of relevant alternative fire detection
system was determined.
The preliminary list of possible relevant alternative fire detection systems included the following technologies:
Fibre optic linear heat detection
Aspirating smoke detection (Sample extraction smoke detection)
Gas detection
Flame detection
Video detection
Light beam linear smoke detection
Acoustic detection
8.1 Presentation of possible relevant fire detection systems
8.1.1 Fibre optic linear heat detection
Fibre optic linear heat detection is an optomechanical type of linear heat detection. The cable, which basically
is the same as used in telecommunication, is routed back and forth for example in the ceiling of the protected
area. For a single detection system, consisting of an optical fibre and a detector unit, the cable may be
several kilometres in length. The detector unit emits a light pulse into the optical fibre and detects reflected
light returning through the fibre optic cable. The intensity or wavelength of the reflected light together with
the time between emitting the light pulse and detecting the reflected signal provides information about the
temperature along the fibre optic cable. It is possible to monitor continuously temperature distribution along
the entire cable and to follow temperature variations during fire.
There are some different technologies within fibre optic heat detection. The most common techniques for
distributed temperature sensing are Raman scattering, Brillouin scattering and fibre Bragg gratings (FBG),
of which Raman scattering is the most common [1]. Both Raman and Brillouin scattering use the temperature
dependence of light scattering due to molecular vibrations within the glass core of the optical fibre. Fibre
Bragg gratings are modifications of the fibre which enable reflection of a temperature depending specific
wavelength from the grating (the modification of the fibre core). The gratings (temperature sensing points)
can for example be spaced every meter along the cable.
Fibre optic heat detection is used in a wide range of applications, including tunnels, conveyer belts, pipelines,
wildlands, aircrafts and hazardous or EMI (electromagnetic interference) intense environments.
Some pros and cons identified and discussed during the workshop are presented in Table 3.
Bureau Veritas RISE Stena | FIRESAFE II
23/91
Table 3. Pros and cons for fibre optic linear heat detection
Pros
Cons
One detector can cover a large area
(several km of fibre optic cable per
detector)
Fast activation time compared to other
heat detection technologies
Robustness against EMI, humidity, dirt
and dust
Continuous monitoring of temperature
even during fire
Low false alarm rate (depending on
setting, but generally lower for heat
detection)
Easy service (one detector unit)
Sustain high temperatures (exact
temperature depends on coatings and
type of fibre)
Redundancy is easy to implement (loop of
the fibre cable or adding of a second
detector unit)
Heat detection is considered slow
compared to other technologies
Convective heat transport from fire to
detector is needed. Area height and
airflow can be crucial
Hard to detect smouldering fires (low heat)
Sampling system (time delay between
light pulses). However, not an issue since
order of seconds
More expensive than other linear heat
detection systems
8.1.2 Aspirating smoke detection
Aspirating smoke detection (ASD), or sample extraction smoke detection as it is named at IMO, consists of
a pipe network with small holes (sampling points) where smoke and air is sucked into the pipe network and
transported to a detector unit. The detector unit consist of a smoke chamber usually much more advanced
and more sensitive than conventional point smoke detectors.
Several sampling points will dilute the smoke if the smoke is only present at one sampling point. ASD systems
are therefore classified based on the sensitivity at one sampling point when clean air is sampled from the
other holes. Class C means that the sensitivity is comparable to conventional point smoke detectors, while
Class A and B means a more sensitive detector. ASD systems are often used where it is important to have
a very early warning, especially in combination with high airflow. That is the reason why these systems are
common in e.g. data centres and air ducts. An experimental study on fire detection in buses showed that
among several tested systems, ASD systems are the least affected by high airflow at the position of the
detector/sampling hole [2]. ASD systems are used in many other applications as well, e.g. in applications
where one aspirating system can replace many point smoke detectors.
Some pros and cons identified and discussed during the workshop are presented in Table 4.
24/91
FIRESAFE II | Bureau Veritas RISE Stena
Table 4. Pros and cons for aspirating smoke detection
Pros
Cons
One detector can replace many
conventional point smoke detectors
Less sensitive to airflows at the sampling
point compared to conventional point
smoke detectors
ASD is considered as very early warning
for smouldering fires
Detector unit can be protected, hidden and
located in a clean environment (easy
service access)
Sampling pipe network can be used also
for gas detection if needed (interesting for
Alternatively Fuelled Vehicles)
Space needed for sampling pipe network
Smoke transport from fire to sampling
points needed. Affected by area height
and airflow.
Transport delay time for long pipes and
dilution of smoke if many sampling points
are used (often compensated by a more
sensitive detector)
Sometimes sampling between different
pipes (sampling delay time)
Ageing problem reported for metal pipes
Addressability
8.1.3 Gas detection
Gas detection includes many types of technologies and is used to detect flammable or toxic gases as well
as fire products, such as CO and CO2. CO point detectors are common fire detectors in buildings and are
often used as a complement to smoke detectors, for example to avoid false alarms. Detection of flammable
gases is most often associated with detection of different hydrocarbons. In general, flammable gas detection
needs less sensitive detectors compared to toxic gas detection since the levels of toxic gases that may be
harmful for humans are often much lower than the LEL (lower explosive limit) of flammable gases. Different
principles and technologies available for gas detection include for example catalytic sensors, infrared (IR)
technology, semiconductors, electrochemical sensors and thermal conductivity sensors. Some technologies
are more suitable for toxic gases and other for flammable gases.
Some pros and cons identified and discussed during the workshop are presented in Table 5.
Table 5. Pros and cons for gas detection (flammable gases and fire products)
Pros
Cons
Early warning (fire or explosion risk)
Possible to detect gas leakages from
alternatively fuelled vehicles
Can be combined with an aspirating
smoke detection system
Sensitive to airflow (however, with high
airflow there is also less risk of
accumulation of flammable gases)
Easily contaminated and consumed by
time (catalytic and electrochemical
sensors). IR gas detectors are more
robust to a higher initial cost.
May require frequent calibrations
8.1.4 Flame detection
Flame detectors detect electromagnetic radiation emitted by flames, which radiate in the visible, the
ultraviolet (UV) and the infrared (IR) spectrum. Soot particles radiate almost as black bodies, which means
that there will be a wide radiation spectrum, similar as for the sun, mainly in the infrared spectrum (heat
radiation). Hot soot particles also cover the visible spectrum (the characteristic yellow light from a flame), but
particles with lower temperature radiate only in the infrared spectrum (smoke). Different molecules in the
flames radiate at specific narrow bands either in the UV or in the IR spectrum, e.g. CO2 is typically detected
at 4.3 µm. Radiation in the ultraviolet spectrum is due to electron transitions and radiation in the infrared
spectrum is due to molecular vibrations. Flame detectors are usually constructed to detect radiation at one
or several of these narrow bands to be able to distinguish a flame from other sources of black body radiation,
e.g. hot metal surfaces or sunlight. Most common today are multi-spectrum IR detectors using several
wavelength bands or combined IR/UV detectors using both spectrums.
Bureau Veritas RISE Stena | FIRESAFE II
25/91
Some pros and cons identified and discussed during the workshop are presented in Table 6.
Table 6. Pros and cons for flame detection
Pros
Cons
One detector can cover a large area (field
of view)
Volume detection (no need for heat and
smoke transport to the detector)
Very fast detection in case of flaming fire
Not much affected by airflow and weather
Some new flame detectors have possibility
to mask certain area of their field of view
(hot spots with high risk of false alarms)
Cannot detect smouldering fires
Historically there have been a high risk of
false alarms (especially UV detectors and
single wavelength detectors)
Field of view affected by obstructions
Sensitive for contamination of detector
lens/window
Some flame detectors cannot detect slow
growing fires
8.1.5 Video detection
Video detection uses a camera and software for image analysis. The camera provides either a visual or a
thermal image and detection of flames or smoke is achieved by image analysis. The analysis can consider
colours, motion, shape, transparency, flicker, energy, or boundary disorder information in the video or a
combination thereof combined by different algorithms [3].
Video detection is a fairly new technology which mainly is used in special applications. Flame video detection
is an easier and more mature technology than smoke video detection. However, smoke video detection has
great advantage in early detection of slow growing fires and smouldering fires. For already existing
surveillance camera systems (CCTV), fire detection algorithms may be added to the software.
Some pros and cons identified and discussed during the workshop are presented in Table 7.
Table 7. Pros and cons for video detection
Pros
Cons
Both smoke and flame detection are
possible (some systems use only smoke
OR flame detection)
Can be combined with surveillance
One detector can cover a large area (field
of view)
Volume detection (no need for heat and
smoke transport to the detector)
Fast detection
Alarm events are easy to locate and
monitor
Possible hydrocarbons and H2 detection
(thermal imaging camera)
Apprehension to use it due to the relative
immaturity of the technology
False alarm rate (e.g. due to reflections,
but depending on technology and settings)
Slower than flame detection because of
image processing (can be faster if smoke
detection is used)
Field of view affected by obstructions
Hard to detect smoke in poor light
conditions
Probably more affected by weather
conditions compared to flame detectors
(visual cameras)
Sensitive for contamination of detector
lens/window
8.1.6 Light beam linear smoke detection
Light beam linear smoke detection consists of a light source and a photocell separated in space. The light
source emits a light beam which is detected by the photocell and any smoke interfering with the light beam
will be detected. There is usually tens and sometimes hundreds of meters between the light emitter and the
detector. These systems are often used in for example high ceiling buildings (atriums, warehouses etc.),
where the smoke does not reach the ceiling due to smoke stratification. They are also used in outdoor
applications, such as power plants or oil rigs.
26/91
FIRESAFE II | Bureau Veritas RISE Stena
Some systems use a reflector, such that the beam travels back and forth, and has the photocell located
together with the light emitter. There could also be two light beams at different wavelengths, which are not
attenuated equally by smoke. By comparing the signals from the two beams one can distinguish smoke from
other beam interfering false alarm sources.
Some pros and cons identified and discussed during the workshop are presented in Table 8.
Table 8. Pros and cons for light beam linear smoke detection
Pros
Cons
Possible to cover long distances with a
single detector
Linear detectors provide better coverage
than point detectors
Can be positioned lower than ceiling
height or outdoors (no ceiling)
Smoke detection means early warning for
most fires (especially smouldering and
slow growing fires)
Beam alignment important. Ship deflection
can be a problem
Smoke concentration and transport is
affected by airflow
Can be sensitive to other light sources,
e.g. sunlight reflections
May give false alarms to anything blocking
the beam (dust, fumes, objects, etc.)
Lack of attachment points for the system
along the sides of weather decks
8.1.7 Acoustic detection
Ultrasonic detection is available for high pressure (typical over 10 bars) flammable gas leakages. Acoustic
detection of gases compared to conventional gas detection has advantages in coverage (volume detection)
and is not affected by airflow and transportation of gases to the detector. However, the pressure must be
high such that detection of ultrasonic tones is possible which may be only applicable to AFV, e.g. using
compressed gas (CNG) or compressed hydrogen for fuel cells. In conclusion, for fire detection this solution
does not seem to fit for the current study.
8.2 Selection of alternative fire detection system
The selection of alternative fire detection systems was based on the pros and cons of each system,
presented above, as well as the challenges for detection in open ro-ro spaces and on weather decks. The
main points made in the discussion during the workshop are summarized below.
8.2.1 Fibre optic linear heat detection
Fibre optic linear heat detection was the only heat detection technology discussed. Heat detection is
generally considered as slow compared to smoke detection but has other benefits such as low false alarm
rate and better monitoring of fire development and fire spread. However, since response time is very
important fibre optic linear heat detection was considered as the best choice. Linear detection provides better
coverage than point detection with shorter response time in case of non-optimal fire locations. Furthermore,
heat detectors responding on temperature rate-of-rise in addition to a fixed alarm temperature have generally
better response times. Fibre optic heat detection combine linear detection, temperature rate-of-rise detection
and good addressability of temperature location along the sensor and seems to have the ability of continuous
monitoring of temperatures also during fire.
The question of the effect of the ship deflection and vibrations on the capacity of the detector has been
raised. This question has been tackled with an explanation of the technology used by the system; it relies on
molecular vibrations which cannot be affected by ship deflection and vibrations.
Bureau Veritas RISE Stena | FIRESAFE II
27/91
One of the main issues with fibre optic linear heat detection is the difficulty to use it on a weather deck.
Indeed, the detector (fibre cable) needs to be positioned relatively close to a possible fire (source of heat i.e.
radiation or convection). Therefore, this solution is envisaged to be positioned in open ro-ro space only
6
.
8.2.2 Aspirating smoke detection
It should be noted that even if SOLAS II-2/20.4.2 had forbidden its use in open ro-ro spaces, open vehicles
spaces and special category spaces, this system is considered here as a possible alternative fire detection
system.
This system is reputed to have an earlier alarm than smoke detection. However, one important issue is the
addressability of the system. To address the fire location, one would need to have separate sampling pipes
that cover specific areas. There are systems that allow complex pipe networks to be connected to one
detection unit, giving multiple addressable zones for one detection system. It is also possible to address the
fire location by changing the pipe flow direction at an alarm. After purging the pipes, the flow direction is
again changed and by measuring the time until smoke is again detected the system can identify where the
smoke enters the pipes.
As well as the previous alternative system, the ASD system cannot be installed on weather deck.
8.2.3 Gas detection
Only detection of flammable gases was considered during the workshop. Detection of flammable gases at
an early stage may prevent a fire or explosion scenario and is interesting for ro-ro spaces with increasing
number of alternative fuel vehicles. Gases that are relevant to be able to detect are e.g. methane (CNG and
LNG vehicles), propane/butane (LPG vehicles), hydrogen (fuel cell vehicles) and combustible gases from
battery ventilation (electric and hybrid vehicles).
As this type of detector can be combined with an aspirating smoke detection system, this alternative system
is considered primarily in that combination, then envisaged to be positioned in open ro-ro space only. Note
that LPG is heavier than air and would be difficult to detect with such a system.
8.2.4 Flame detection
Flame detection has been used for long time in different industry applications (Oil and Gas industry),
especially where liquid or gas fires are expected to occur (high risk areas). As long as the flames are within
the detector’s field of view most flame detectors activate an alarm within seconds or sometimes within
milliseconds. In addition, flame detection is also often used for outdoor applications due to difficulty of smoke
and heat detection. Flame detectors may face difficulties with sun and light reflections causing false alarms.
However, considering visibility (field of view when installed between cargo and deckhead) this system is not
suitable to be installed in open ro-ro space but nevertheless considered for weather deck.
8.2.5 Video detection
In the case of video detection, as this system is a fairly new technology, the discussion has been based on
own experiences. Three system technologies can be distinguished:
Smoke or combined smoke and flame video detection: this system is based on video analysis of
CCTV camera feed installed on board. Cameras must have a quite good resolution. The system is
only applicable in illuminated areas and the system seems to be really sensitive to exterior lights
(reflections and moving shadows) causing false alarms. Therefore, this system could possibly be
considered for weather deck.
Video detection using thermal imaging camera: these cameras are generally more expensive
compared to conventional CCTV cameras but have the ability to detect thermal events that might
result in a fire. False alarms can be an issue, but this type of video detection is reputed less sensitive
to exterior lights as targeting infrared light spectrum. It is considered for weather deck.
6
This detection system was also considered for use in closed ro-ro spaces in an extended cost effectiveness
assessment later in the study.
28/91
FIRESAFE II | Bureau Veritas RISE Stena
Flame video detection (visual or near infrared): this system can be integrated with existing CCTV
cameras or use special designed cameras. Video detection of flames is generally less sensitive to
false alarms compared to smoke video detection. Some systems have been extensively used in oil
and gas industry. Experiences from this industry show a robust system against dusts, salt, moisture,
etc., and having a low ratio of false alarms. From those experiences, this system is primarily
considered for weather deck.
8.2.6 Light beam linear smoke detection
Due to ship deflections this system can be questioned if positioned in longitudinal direction and if positioned
transversely the lack of appropriate attachment points along the sides of most weather decks, it was decided
to not consider light beam linear smoke detection for further evaluations. However, a supplier of this type of
system claims that tolerances of the system are enough to handle ship deflection. Light beam linear smoke
detection is together with smoke video detection the only realistic approaches for smoke detection on
weather deck. Though not considered further in this report, the system might be relevant if side structures
are available for system attachment.
8.2.7 Acoustic detection
This system does not fit for the current study and the project group did not know of any commercial acoustic
detection systems for fire detection. In conclusion, this technology was not selected for further evaluation.
8.3 Conclusion of the selection of relevant alternative fire detection systems
The main conclusion is that no system is optimal for both types of ro-ro spaces. The alternative systems
discussed in the previous subchapter were therefore divided for further evaluation with regard to the specific
type of ro-ro spaces where they are most relevant, as presented below:
Fibre optic linear heat detection Open ro-ro space
Aspirating smoke detection Open ro-ro space
Gas detection, only in combination with ASD Open ro-ro space
Video detection: Smoke or combined smoke and flame detection Weather deck
Video detection: Thermal imaging camera Weather deck
Video detection: Flame video detection Weather deck
Flame detection Weather deck
Light beam linear smoke detection Not further analysed, however, could be relevant for weather
decks if solid side structures above cargo height are available
Acoustic detection Not selected
Bureau Veritas RISE Stena | FIRESAFE II
29/91
9 EVALUATION OF RELEVANCE/APPLICABILITY OF
SYSTEMS IN OPEN RO-RO SPACE AND WEATHER DECK
A conclusion of the above review of relevant alternative fire detection systems was that no system seems to
be optimal for both open ro-ro space and weather decks. Indeed, according to the definitions in paragraph
7.2.2, a weather deck is completely exposed to weather from above and from at least two sides, which
implies that installation of a detection system requiring detectors above to the fire is impossible.
In this context, the evaluation of the applicability of alternative fire detection systems for open ro-ro spaces
and weather decks will be split between those two types of decks.
9.1 Open ro-ro space
Fibre optic linear heat detection requires a ceiling structure for attachment of the optical fibre cable and it is
hence not relevant for weather decks. Heat transport is affected a lot by high airflows and the system is
probably most suitable for closed or open ro-ro spaces with the forward end closed. However, comparing
different heat detection systems, the fibre optic system has the lowest response time. In addition, a fibre
optic cable can easily be routed with smaller spacing without significantly increasing costs, which is good for
improved activation time. Furthermore, the fibre optic linear heat detection is reputed insensitive to harsh
environment with moisture, dirt, salt, etc. and does not need to be covered during drencher testing.
A fibre optic heat detection system can be installed together with a smoke detection system, if combined
smoke and heat detection is preferred. Heat detection complements smoke detection with better monitoring
of fire development and fire spread and it can also stay activated during loading and discharging of the deck.
The fibre optic cable can easily be routed together with, for example, the pipe network of an ASD system.
Concerning open ro-ro spaces and smoke being ventilated away through side openings, it was discussed
within the project group whether heat detection systems might have a better possibility to detect fire than
smoke detection systems. Heat detection systems are generally not affected by heat radiation but only by
heat convection, which means that earlier detection than with smoke is unlikely. Another safety measure
may be to have video detection based on smoke or combined smoke and flame detection, covering the side
openings (and fore and aft), which means that video detection could be suitable for open ro-ro spaces for
complementary detection.
Aspirating smoke detection systems require, as for the optical fibre system, a ceiling structure for attachment
of pipes and they are not relevant for weather decks. However, although it is forbidden on ro-ro passenger
ships (see 7.3.1.3), ASD systems may be considered relevant for open ro-ro spaces. The smoke
“accumulators” (mentioned in the FSS Code) are assumed to be the same as sampling holes. Referring to
Figure 1 it can be seen that the coverage area of an accumulator can be 288 m2, which is substantially more
than what is allowed for point smoke detectors (74 m2). Ageing and corrosion problems have been reported
for these systems when metal pipes common with extinguishing systems have been used. It is preferable to
use pipes dedicated only for detection of smoke and gases.
As mentioned in the previous chapter, ASD systems are often used where it is important to have a very early
warning, especially in combination with high airflow, since the ASD systems are little affected by airflow at
the position of the sampling points. Conventional point smoke detectors rely on diffusion of the smoke from
outside the detector to inside of the detector. With high airflow this process is slow and smoke will be difficult
to detect. Because of this, ASD systems seem to be more suitable for open ro-ro spaces compared to
conventional point smoke detectors. In addition, ASD systems are generally much more sensitive than
conventional point smoke detectors, which is beneficial in high airflow areas since the airflow will spread and
dilute the smoke produced. Furthermore, if the smoke is present at several sampling points, the total amount
of smoke will be easier to detect. However, current regulations are somewhat restrictive and allow no more
than four accumulators connected to the same detection unit.
30/91
FIRESAFE II | Bureau Veritas RISE Stena
Figure 1. Possible spacing of accumulators (sampling points) fulfilling the regulations.
9.2 Weather deck
Flame and video detectors are suitable for weather decks since they can overlook large areas from a
distance. A deckhead is not required for these types of detectors but it is rather a reason to why they are
generally not applicable for open ro-ro spaces. Indeed, in open ro-ro spaces, the field of view can be very
limited due to a small space between the cargo and the ceiling. However, these types of detectors may be
relevant for open ro-ro spaces if the ceiling height is well over the cargo height or if they are used as
complementary detectors to fixed smoke/heat detection systems.
Regarding robustness to weather conditions, many flame and video detection systems are used for outdoor
applications, on oil rigs for example, and should be sufficiently robust. The most sensitive part of the detector
is the detector lens/window, which can be affected by dirt or ice. To protect against ice formation, internal
heating of the window can be provided for the detector and it is common that these systems have internal
lens supervision, which means that a fault signal is provided if the lens/window is contaminated. In addition,
weather conditions such as rainfall, snowfall, fog and sun reflections may either cause false alarms or
blinding/shielding such that a fire might remain undetected.
Advanced flame detectors today are fairly robust. Using multi spectrum detection in combination with
advanced algorithms, for instance analysing the flickering of the radiation, should prevent a lot of potential
false alarms. However, sun blinding can be a problem, which means that combined radiation from a fire and
the sun might be interpreted as no fire. Other weather conditions should not affect detection by flame so
much. Furthermore, detection by flame radiation means rather slow detection in case of smouldering fires or
slow growing fires. Another challenge is that the field of view of the detector can be significantly affected by
high cargo, such as trucks.
For a discussion on video detection, reference is made to the categorization introduced in the paragraph
8.1.5, i.e. smoke or combined smoke/flame video detection (visual), video detection using thermal imaging
camera and flame video detection (visual or near infrared).
Smoke video detection has great potential for early detection of a fire. However, there is sometimes still a
trust issue with regard to false alarms for these systems. On weather decks specifically, there are a range
of potential issues that might cause problems, e.g. sun reflections, shadows, exhaust, fog, rain, water
splashes from sea, etc. For these systems to be effective, they probably need some training (software
learning from background effects) for the specific application.
Video detection using thermal imaging camera is similar to flame video detection, but the image analysis has
some different conditions. For example, in the visual spectrum one can use colour analysis that is not
possible for thermal images. However, a thermal imaging camera will be less affected by weather conditions
and light conditions. Poor light is especially a problem for smoke video detection.
Flame video detection is the most mature video detection technology. Flames are easier than smoke to
detect and there is less risk of false alarms. Flame video detection is very similar to conventional flame
detection in many aspects but could be more affected by weather conditions. For example, rainfall will affect
Bureau Veritas RISE Stena | FIRESAFE II
31/91
a camera image more than it affects flame radiation intensity reaching the detector. However, video detection
also adds value by monitoring and visualization of the alarm event.
32/91
FIRESAFE II | Bureau Veritas RISE Stena
10 EVALUATION OF ACTIVATION TIME
10.1 Literature on activation time of different fire detection systems
The activation time of a detector depends on many factors, for example the detector location relative to the
fire, the fire scenario, the detector technology, the airflow, and thermal inertia or delay times within the
system. Figure 2 presents a typical fire scenario with respect to detection technologies. Most fires have an
incipient stage and release smoke before any flames are visible, which could last for minutes up to several
hours. When flames are present the fire will normally develop fast with increased heat release. Electrical
fires, which have been identified as one of the main risk contributors on ro-ro decks in the previous
FIRESAFE study, typically start as smouldering fires. As seen in the figure, gas and smoke detection has
potentially faster activation times than flame and heat detection for these types of fires. However, gas and
smoke have to be transported to the location of the sensor (except for smoke video detection) with enough
concentration for activation of an alarm and this might only be possible for a larger fire. It should be noted
that gases could even be detected before fire occurs (e.g. in case of a leakage) but also that detectable
gases are not always produced before smoke. For a liquid or gas fire, which starts as a flaming fire, detection
response time will primarily depend on the transportation time of gas, smoke, heat and radiation from the fire
to the detector.
Figure 2. Typical fire scenario with respect to fire detection (or prevention with respect to gas).
To compare different fire detection systems with respect to activation time, all factors discussed above must
be taken into account. The fire scenario will affect the possibility of rapid response for different technologies,
but this could also to some degree be compensated by smaller detector spacing. Comparison within the
same type of technology is easier to achieve by focusing on for example sensitivity settings, internal delay
times, thermal inertia or airflow sensitivity.
For further evaluation of activation times, some different fire detection tests presented in literature are
summarized below.
Törnskogstunneln, Stockholm [4]
Nine different fire tests were conducted; seven with about 0.5 MW gasoline or heptane fire and two fires with
plastic components (test 8-9). Wind velocity was 3-6 m/s, with most tests carried out at 6 m/s. Smoke
detectors were placed 100 m and 200 m downstream the fire location. Flame detectors were placed 25 m
and 60 m upstream the fire location (FlameA and FlameB being different types). A summary of activation
times is presented in Table 9. Temperature measurements showed just a few degrees difference in the
tunnel ceiling for these fire tests, at most an increase of 2-3°C in about 5 min.
Bureau Veritas RISE Stena | FIRESAFE II
33/91
Table 9. Detector performance in Törnskogstunneln, Stockholm.
Test
Smoke 100 m
(min)
Smoke 200 m
(min)
FlameA 25 m
(min)
FlameA 60 m
(min)
FlameB 25 m
(min)
FlameB 60 m
(min)
1
2.2
2.6
0.1
0.1
-
0.2
2
3.0
3.8
0.1
0.1
3.6
0.2
3
1.3
1.7
0.1
0.1
1.7
0.2
4
3.0
3.7
0.2
0.1
-
0.4
5
2.2
3.0
0.6
0.2
-
0.4
6
1.4
2.9
0.1
0.1
-
0.2
7
2.9
3.3
0.1
0.0
1.6
0.1
8
-
-
-
-
-
3.9
9
-
-
-
1.0
-
2.3
Northern Link tunnel, Stockholm [4]
The 16 different tests are summarized in Table 10. The ceiling height of the tunnel is 6.6 m, and gasoline
was used as fuel. The “camera” was an incident detection camera system, which could also detect a fire
event, positioned 10 m upstream the fire location. The LHD system was a fibre optic linear heat detection
system and the smoke detectors were positioned 100 m (S1) and 190 m (S2) downstream the fire location.
Table 10. Detector performance in Northern Link tunnel, Stockholm.
Test
Wind
[m/s]
HRR [MW]
Time to detection (after ignition) [m:ss]
0.5-1.5
min
1.5-3.5
min
3.5-5
min
Camera
Pre-alarm
(LHD)
Alarm
(LHD)
Smoke S1
Smoke S2
1
3
0.54
0.79
0.75
00:50
-
-
1:41 (0:55)
3:34 (1:33)
2
3
0.75
1.21
1.47
00:55
05:36
-
1:40 (1:08)
2:47 (1:50
3
3
1.57
2.18
1.54
00:33
00:58
01:06
1:04 (0:55)
2:20 (1:54)
4
3
2.65
3.60
1.69
00:29
00:50
00:50
1:00 (0:56
2:01 (1:37)
5
3
1.35
2.01
1.55
00:34
01:05
01:21
1:07 (0:54)
2:17 (1:38)
6
3
0.69
1.17
1.48
00:59
05:00
-
1:43 (1:06)
3:10 (1:50)
7
3
0.31
0.52
0.60
01:09
-
-
2:38 (1:16)
4:05 (2:00)
8
6
0.36
0.48
0.62
02:44
-
-
4:02 (1:06)
5:08 (2:33)
9
6
0.88
1.37
1.31
01:13
04:38
-
2:49 (0:49)
3:16 (1:20)
10
6
1.45
2.26
1.89
01:03
01:50
01:54
1:58 (1:01)
2:27 (1:28)
11
7.7
1.68
2.68
1.03
00:55
00:56
01:20
1:30 (0:39)
2:09 (0:57)
12
7.7
0.88
1.53
1.14
02:06
-
-
2:47 (1:10)
3:20 (1:56)
13
7.7
0.58
0.95
0.19
02:35
-
-
2:16 (0:40)
2:46 (1:29)
14
1.5
0.40
0.54
0.61
01:11
02:00
-
2:21 (1:40)
4:22 (3:47)
15
1.5
0.88
1.24
1.31
00:42
00:55
01:03
1:41 (1:28)
3:46 (3:25)
16
1.5
0.87
1.21
1.45
00:36
01:07
01:15
1:31 (1:22)
3:13 (2:24)
Fire experiment in a cable tunnel [5]
Fire experiments were conducted in a small cable tunnel with ceiling height 2.2 m and with cable racks on
each side of the tunnel. Four systems were tested: two electrical heat sensing cables (one with fixed
temperature response and one with temperature rate-of-rise response), a Raman scattering fibre optic
sensing cable and a FBG fibre optic sensing cable. The detectors were routed together with the cables in
the racks. A small-scale test was conducted with a heated cable (no flames) where the detectors had to be
in contact with the heated cable to give a response. In this test the electrical heat sensing cables were faster
than the optical fibre sensors, basically due to different spatial resolutions for the different sensors.
In the large-scale fire test where an alcohol fire was located on the tunnel floor both the fibre optic systems
activated an alarm in about 30 seconds. The response time of the rate-of-rise electrical heat sensing cable
was about 50 seconds and 150 seconds (2 repetitions) and the fixed temperature sensing cable activated
after about 300 seconds.
34/91
FIRESAFE II | Bureau Veritas RISE Stena
Fire experiment in a real metro station [6]
Aspirating smoke detection systems were tested, and it may be interesting to compare activation times
versus smoke temperatures in the ceiling at that time. Three different fuels were used for the test fires;
smouldering paper, smouldering cotton wick and flaming plastics. For the smouldering fires the first warning
from the detection system was achieved after about 13 min (paper) and 3 min (cotton wick). At these times
the temperature difference between smoke and ambient was less than 0.5 °C. For the flaming fire the
warning was achieved in about 1 min and the temperature difference was then about 5 °C, which means that
a rate-of-rise heat detector could have worked as well in that scenario.
Response time comparison of spot smoke detection and ASD [7]
This is a comprehensive study of full-scale comparison tests of spot smoke detectors and aspirated smoke
detection systems in a telephone switch centre. The report is from 1999 and the detectors tested could have
been improved since then. The coverage of each spot detector was 19 m2 and 37 m2 (two separate systems)
and the aspirating detectors were installed according to manufacturer recommendations. Sampling points
were spaced similar to the 19 m2 spot detectors. The facility was 1670 m2. In total, 56 tests were conducted.
The 19 m2 spaced spot detector system and the aspirating system performed comparably, while the 37 m2
spaced spot detector system provided a decreased level of performance. The aspirating system responded
sooner in a majority of tests, however, the spot detector system detected a greater number of smoke sources.
It should be noted that the spot detector system used is designed for this application and has a lower
sensitivity compared to conventional point smoke detectors.
Fire detection tests in toilet compartment and driver sleeping compartment of buses [2]
Fire detection tests were conducted in mockups of a bus toilet compartment and a driver sleeping
compartment. Most tests were performed in the toilet compartment mockup and the response times are
presented in Table 11. The most interesting conclusion from the tests is that the aspirating systems were
less sensitive to high airflows compared to point detectors (mainly ref. to pos. 2 in the table where the highest
airflow was present).
Table 11. Detector performance in bus toilet compartment. The response times are given in seconds after
ignition.
“ND” = No detection
“s” = smoke sensor
“h” = heat sensor
Toilet compartment
Cig.
Trash can
Heptane pool
Plastics & rubber
Low fan
Low fan
High fan
Low fan
High fan
Low fan
High fan
Detectors
Pos.
Test 1
Test 2
Test 3
Test 4
Test 5
Test 6
Test 7
Point smoke det.
1
ND
ND
ND
27
43
57
ND
2
ND
ND
ND
32
61
69
ND
4
ND
42
45
46
47
39
37
Point smoke/ heat det.
1 s
ND
ND
ND
41
30
32
ND
1 h
ND
ND
ND
82
56
ND
ND
2 s
ND
ND
ND
25
56
39
ND
2 h
ND
ND
ND
32
ND
ND
ND
Asp. smoke/ heat det.
2 s
ND
ND
ND
25
38
33
ND
2 h
ND
ND
ND
ND
ND
ND
ND
4 s
ND
21
21
40
36
12
21
4 h
ND
ND
ND
ND
ND
ND
ND
Asp. smoke det.
1+4
51
52
54
43
46
46
50
Flame detection tests [8], [9]
These two reports include flame detector tests with several different flame detectors. 36 different detectors
were tested in indoor bench tests, including false alarm tests. False alarm sources included arc welding,
different light sources and spark sources. Most detectors were immune to most false alarm sources, but arc
Bureau Veritas RISE Stena | FIRESAFE II
35/91
welding is a source of systematic false alarm when it is present at the close proximity of detectors. The most
sensitive detectors to false alarm sources were those with the shortest response times in fire tests.
12 different flame detectors were included in outdoor fire tests where the distance to different fire sources
was altered. The measured maximum detection distances for different fires (heptane, ethanol, cardboard,
methane and hydrogen) generally deviated a lot, at both longer and shorter distances, from the values
specified by the manufacturers. Some interesting conclusions are that an earlier detection occurs in windy
conditions and that detectors are not hindered by the presence of snow or frost.
Another test, focused on methanol fires, evaluated 10 different flame detectors. Both methanol and heptane
fires were used in combination with different obstructions and interfering sources. All detectors had a
response time of less than 10 seconds for most of the fire scenarios. An interesting aspect is that completely
obstructed fires can be detected due to reflections from the fire on metal surfaces, at least at short distances
(about 10 m).
Video detection tests [10], [11], [12]
First two reports present some tests in road tunnels. For the first report, the response times for different
distances and fire sizes are presented in Table 12. The article estimates that a conventional heat sensor
cable would need a fire size of 3-4 MW for activation in tunnels.
Table 12. Video detector performance in a road tunnel [10].
Video-based detector
Fire Size
HRR
25m
70m
95m
0.1 m2
0.25MW
40s
93s
N/D
0.2 m2
0.5 MW
37s
82s
N/D
0.5 m2
1 MW
32s
64s
121s
For the second report, the response times are presented in Table 13. Detection is based on both flame and
smoke analysing and the detector activates an alarm for open fires as well as obstructed fires.
Table 13. Video detector performance in a road tunnel [11].
Fire
scenario
Test
number
Fire source
(m)
Fuel type
HRR
(kW)
Air
velocity
(m/s)
Test
distance
(m)
Response
time (s)
Open fire
T-1
0.3 x 0.3
Gasoline
100 - 125
0
110
8
T-2
0.3 x 0.3
Gasoline
100 - 125
5.0
110
26
T-3
0.3 x 0.3
Gasoline
100 - 125
0
90
14
T-4
0.3 x 0.3
Gasoline
100 - 125
5.0
90
8
T-5
0.6 x 0.6
Gasoline
550 - 650
5.0
90
8
T-6
0.3 x 0.3
Gasoline
100 - 125
1.5
60
7
T-7
0.3 x 0.3
Gasoline
100 - 125
5.0
60
8
T-8
0.6 x 0.6
Gasoline
550 - 650
5.0
60
8
Fire in the
vehicle
T-9
0.6 x 0.6
Gasoline
550 - 650
0
50
25
T-10
0.6 x 0.6
Gasoline
550 - 650
0
50
6
Fire under
the vehicle
T-11
0.6 x 0.6
Gasoline
550 - 650
1.5
60
70
T-12
0.6 x 0.6
Gasoline
550 - 650
3.0
60
95
T-13
1.0 x 1.0
Gasoline
1500 -
1700
5.0
90
18
Fire
behind the
vehicle
T-14
0.6 x 0.6
Gasoline
550 - 650
5.0
50
100
T-15
1.0 x 1.0
Gasoline
1500 -
1700
5.0
50
17
T-16
1.0 x 1.0
Gasoline
1500 -
1700
3.0
50
13
Stationary
vehicle fire
T-17
Passenger
compartment
Wood and
Foam
1700
0
50
110
T-18
Passenger
compartment
Wood and
Foam
1700
3.0
50
80
Moving
vehicle fire
T-19
0.3 x 0.3
Gasoline
125
0
5
27
T-20
0.3 x 0.3
Gasoline
125
0
50
No Alarm
36/91
FIRESAFE II | Bureau Veritas RISE Stena
The third report presents some different fire tests with a smoke video detector and also some tests with
detector disturbances, such as illumination changes and moving people in front of the camera. The system
seems to be robust with respect to these tests and for the different fire scenarios the response time is typically
30-60 seconds from the time when smoke is visible.
Infrared Video Systems [13]
This report highlights the possibility of thermal imaging cameras to detect a hot spot or heating event, which
could result in a fire. To detect a possible fire event before happening can be considered as very early
warning.
Real fire incidents [14]
DNV-GL has studied 35 fires within ro-ro spaces between 2005 and 2016. Reliable data for detection was
available for 10 cases. The time to detect the fire is short for at least eight of these, and there are no
indications that there were significant delays for the two remaining cases. The fire was detected by the fire
detection system in seven cases, by fire patrol/crew in two cases, and essentially simultaneously by the fire
detection system and fire patrol in one case. In eight cases there was a fixed fire detection system (assumed
to be smoke detection), in one case a sample extraction smoke detection system and in one case no fixed
fire detection system (weather deck). Two cases provide some more details regarding activation time, one
stating that detection
7
was achieved after 4-5 minutes and one where the fire detection system activated an
alarm 2 minutes after the fire event was detected by crew passing by.
10.2 Deck loading configuration and its influence on activation time
According to the bibliography presented previously, the time detection has a strong dependency to the fire
scenario and situation context. In the case of the present study, a possible fire is also dependant on the
cargo and its loading configuration, as further discussed below.
10.2.1 Definition of deck loading configuration
The present study focuses on fire detection on weather deck and open ro-ro spaces (as described in 7.2.2).
The use of those kinds of decks is dedicated to the transport of cargo, mainly constituted by containers, cars,
buses, trucks or lorries. In addition, it should be noted that inside trucks or lorries, there are often all possible
kinds of items/goods.
From this assumption, a possible fire starting within cargo, and its detection, is influenced by several factors.
In fire safety engineering the fire load density, representing the available combustibles per square meter,
can be used to represent the potential development of fire. By coupling the fire load and the amount of
available fresh air (oxygen), a fire can be described as fuel controlled in case of a high amount of fresh air,
or ventilation controlled if the available fresh air limits the fire development. If the fire is driven by the
combustibles, the heat release rate of the fire can be assumed dependent on the nature and quantity of
fuels. Such scenarios are expected on weather deck and open ro-ro space.
For detection, the most important factor for a fire in open ro-ro space and weather deck is not cargo
influences on the fire development, but how the cargo influences the fire products (smoke and heat). Those
influences are in this case quite straightforward and can be presented in two groups: the geometric
configuration of the cargo and the fuel type of the cargo.
10.2.2 Influence of the type of cargo (fuel)
The type of cargo may affect detection for some special fire scenarios. For example, leakage of flammable
gases or liquids could cause a rapid flaming fire scenario and there could be need for gas detection or flame
detection. The density of different gases compared to air could affect optimum position of gas detectors.
Further, special liquids such as clean burning methanol may cause problems for smoke detection and visual
flame detection. However, these are special scenarios and regardless of type of cargo, a slow growing fire
7
This detection time is not based on the real fire ignition time but on the first sign of fire detected by CCTV during the investigation
phase.
Bureau Veritas RISE Stena | FIRESAFE II
37/91
(smouldering fire) should be expected initially for most cases. Gas detection can be a good complement
though, especially with an increasing number of Alternative Fuel Vehicles which could release gases before
a potential fire.
10.2.3 Influence of the geometric configuration of the cargo
The cargo geometric configuration has two main influences.
The first one is the cargo height, which will affect the field of view of flame and video detectors. Indeed, a
high cargo will block the view angle of flame or video detectors. A good example of this scenario is given in
Figure 3.
Figure 3. Example of video blockage due to high cargo.
Configurations with, for instance, dedicated areas for cars only would either give enough space between
cargo and ceiling such that flame or video detectors can be suitable, or the possibility to have dedicated
decks with lower ceiling height, which is beneficial for smoke and heat detectors with respect to activation
time, as described in Figure 4. The traveling time of heat and smoke and its concentration in air (dilution
effect) are reduced in case of a low ceiling height (blue arrow) compared to a high ceiling high (red arrow).
38/91
FIRESAFE II | Bureau Veritas RISE Stena
Figure 4. Example of smoke-detector traveling time for two deck heights.
Another aspect of the cargo height and cargo spacing is that high cargo (in relation to ceiling height) reduces
the cargo-ceiling space, which can accelerate airflow (as presented in the Figure 5). High airflow affects most
detection technologies, except flame and video detection (which are unsuitable due to field of view).
However, for instance ASD has been proven to be less sensitive to airflow compared to point smoke
detection.
Figure 5. Example of accelerated air flow in case of high height cargo.
Bureau Veritas RISE Stena | FIRESAFE II
39/91
11 EVALUATION OF COSTS
The last evaluation done for the desk study on the relevant alternative fire detection system is a cost study.
In order to perform the most accurate study, the evaluation of cost was performed with the ships chosen for
the FIRESAFE II study as basis (Interested readers can refer to section 7.5 of FIRESAFE II Part 1 report
[18] for a description of the generic vessels). All systems were evaluated as add-on systems covering the
relevant deck and bringing the signal to the main fire panel on the bridge. As concluded in paragraph 8.3, no
alternative fire detection system is optimal for both open and weather decks. Consequently, the evaluation
of cost is divided by the type of deck. And two ships with those particular deck types were chosen, the
Standard RoPax for open ro-ro spaces and the Cargo RoPax for weather decks.
Since it is expected that the equipment costs would be the same for newbuildings and existing ships
(installation costs would be slightly less), the cost evaluations were made only for existing ships.
In order to make this result generally applicable to the world fleet of ro-ro passenger ships, scalability
according the deck size is discussed.
Maintenance costs in the maritime context could not be retrieved. However, it was considered during the
selection process.
11.1 Description of the decks used for the cost evaluation
11.1.1 Ship with open ro-ro space
As described previously, in order to consider the costs of implementing the alternative fire detection systems,
an open ro-ro space of a real ship was selected, that of the Standard RoPax, illustrated in Figure 6.
Figure 6. Picture of the Standard RoPax.
Since Deck 4 of this ship was considered to represent well the open ro-ro spaces, this ship was selected.
This deck is:
enclosed by deckhead and side shell bulkheads;
naturally ventilated through large openings in the ship side and has an open aft (no mechanical
ventilation); and
largely exposed to the ambient environment (temperature, wind, humidity …)
40/91
FIRESAFE II | Bureau Veritas RISE Stena
All costs presented are based on installations and systems applied for this open ro-ro space, deck 4 of the
Standard RoPax. The deck is approximately 3 500 m2 over a length of 155 m. Further details can be found
in Figure 7.
Figure 7. General arrangement of the open ro-ro space (deck 4) of the Standard RoPax.
11.1.2 Ship with weather deck
Similar to above, a ship was also selected with a representative weather deck, in order to produce a cost
evaluation of alternative fire detection system. The selected ship was the Cargo RoPax, presented in Figure
8, with the weather deck shown in Figure 9.
Figure 8. Picture of the Cargo RoPax.
Bureau Veritas RISE Stena | FIRESAFE II
41/91
Figure 9. View of the weather deck of the Cargo RoPax.
The weather deck of the Cargo RoPax is approximately 1800 m2, with the dimensions 93 m long and 23.5
m wide. It should be noted that the most forward part of this deck is enclosed and not relevant for this
evaluation. Only the part with no deckhead (structure above) was considered, as illustrated (blue) in Figure
10.
Figure 10. General arrangement of deck of the Cargo RoPax.
11.2 Cost estimations of alternative fire detection systems
In this paragraph are presented the estimated costs of relevant alternative fire detection systems. As
described in paragraph 8.3, alternative fire detection systems relevant for open and weather decks were
identified, including the following technologies:
Fibre optic linear heat detection Open ro-ro space
Aspirating smoke detection Open ro-ro space
Gas detection, only in combination with ASD Open ro-ro space
Video detection: Smoke and combined smoke and flame detection Weather deck
Video detection: Thermal imaging camera Weather deck
Video detection: Flame video detection Weather deck
Flame detection Weather deck
Some of the presented costs are the result of two different quotations.
42/91
FIRESAFE II | Bureau Veritas RISE Stena
11.2.1 Systems for open ro-ro spaces
Systems judged as relevant for open ro-ro spaces are fibre optic linear heat detection, aspirating smoke
detection and gas detection in combination with ASD.
In order to have a reference in terms of cost evaluation, a common fire detection system (conventional point
smoke & heat detection system) was also studied.
11.2.1.1 Conventional point smoke & heat detection system as an add-on system
The present quote was based on addressable detectors with combined smoke and heat detection of IP55
class units. The deck will require about 50-60 detectors and the price was estimated based on 60 units.
System cost estimation
The estimated total cost is € 55 000.
Components cost
Components cost includes detectors and related equipment needed and is estimated to € 10 000.
Installation cost
Installation cost is estimated, based on ship owner experience, at € 40 000.
Commissioning cost
The commissioning cost, based on 5 days of system configuration and test work, is about € 5 000.
System Scaling
As mentioned above this deck will require between 50-60 detector units, subsequently a deck of half the size
will require 25-30 units. Estimated cost is 30 000 (30 units) for half the deck size.
11.2.1.2 Fibre optic linear heat detection
The below information is based on 2 quotes received from manufacturers.
Estimated total price range
The estimated total price ranges from 50 000 to 80 000. The total price includes, when available, the
component cost, the installation cost and the commissioning cost. Note however that it may be required to
have two controller units instead of the currently quoted one unit. It must be mentioned since this is significant
to the cost of the system. The total estimated price for such a system is estimated to € 70 000 – 125 000.
Component cost
Details on the component cost, ranging from 20 000 to 45 000 are detailed below. Both quotes received
included the following components:
Components
Units
Comments
Controller unit with 1 km range and
dual channels
1
Allow cable to be looped for redundancy
1 quote did not specify the cable range
Fibre optic cable - Plastic sensor
cable with accessories
650 -
700 m
Mounting material not included
Modbus TCP
1
Allow the system output to be integrated into an
existing control management system or PLC
Software package
1
Optional and not included in component cost
figure above.
Used for further analysis, storage and visualisation
Important differences in components costs have been identified but the technical specifications received
from the manufacturers did not allow determining whether there were any profound differences between the
Bureau Veritas RISE Stena | FIRESAFE II
43/91
systems. It may be so that the quote which did not specify the cable range for the controller unit considered
larger controller units which have brought up the price.
Installation cost
The installation cost is about € 31 000, based on ship owner estimation.
Commissioning cost
Commissioning cost is based on two days of the system configuration and test work and has been quoted
at € 2 000.
System Scaling
The system costs are governed by the cost of the controller unit. Scalability is based on information from
one of the suppliers which states their controller units are available for cable ranges from 1 to 10 km, ranging
in price from € 17 000 to 37 000. The quote received have been based on 650-700m of fibre optic cable for
which a 1 km range controller can be used.
A system for half the deck size would cost more or less the same as the same controller unit would be used.
Some cost could be saved of the required amount of cable, as only half the length would be necessary.
Hence, for half a deck, the system cost was estimated to about € 40 000, as opposed to € 50 000 for a full
deck.
Notable for this system is that it might become quite cost efficient if it is applied on larger or several decks.
For example, a 2 km dual channel system with 2 km cable would cost about € 25 000 (component price for
controller unit and cable) and could potentially cover at least three decks of the considered size.
11.2.1.3 Aspirating Smoke Detection
11.2.1.3.1 Aspirating Smoke Detection with NO addressability function
The aspirating smoke detector system considered here has no addressability function, which means that
parallel systems would be needed if addressability is desired. The below quote is based on an 8 zone system,
corresponding to the drencher zones. The detectors can be calibrated to different sensitivities (class A, B or
C), class C being equivalent to the sensitivity of a conventional smoke detector (as explained in the
paragraph 8.1.2).
Estimated total cost
The estimated total price is 115 000, including (when available) the component cost, the installation cost
and the commissioning cost.
Components cost
The component cost was estimated to a total of € 50 000, based on the required components detailed below.
44/91
FIRESAFE II | Bureau Veritas RISE Stena
Components
Number of units
Aspirating smoke detector units
8
Stainless IP66 detector enclosure
9
Power supply units
9
Sub-rack
2
Remote display
8
In-line filter
8
Water trap
8
Pipes, fittings and cables
Necessary amount for the installation
Installation cost
The installation cost is € 60 000, based on ship owner estimation.
Commissioning cost
The commissioning cost was estimated to 5 000. It includes commissioning of the system, including all
necessary tests and configurations, as well as crew training on the system functionality.
System Scaling
The system cost is highly dependent on addressability requirements, i.e. the number of detector zones. The
current cost estimate is based on 8 detector and drencher zones. Subsequently, a system of half the size
would use 4 detector units and the price would basically be half of that for a full deck system (likely a bit
more since commissioning and crew training will be similar for both installations). The cost of a half deck
was estimated to € 45 000.
11.2.1.3.2 Aspirating Smoke Detection with addressability function
The system cost is highly dependent on addressability requirements, i.e. the number of detector zones. The
current cost estimate is based on 8 detector and drencher zones. Subsequently, a system of half the size
would use 4 detector units and the price would basically be half of that for a full deck system (likely a bit
more since commissioning and crew training will be similar for both installations). The cost of a half deck
was estimated to € 45 000.
Estimated total cost
The estimated total price is about € 105 000 and includes, when available, the component cost, the
installation cost and the commissioning cost.
Component cost
Details on the component cost are detailed below, totalling a cost of € 45 000.
Bureau Veritas RISE Stena | FIRESAFE II
45/91
Components
Number of units
Aspirating smoke detector
2
Relay modules STAX
2
High level interface
1
IP66 enclosure
3
Power supply unit
3
Remote display
1
Pipes, fittings and cables
Necessary amount for the installation
Installation cost
The installation cost is 50 000, based on ship owner estimation.
Commissioning cost
The commissioning cost was estimated to € 12 000. It includes commissioning of the system, including all
necessary tests and configuration, as well as crew training on the system functionality.
System Scaling
The system cost is highly dependent on addressability requirements, i.e. the number of detector zones. The
current cost estimate is based on 38 detector zones, which corresponds approximately to a conventional
point detection system. The detector can monitor 40 zones in its basis version, which can be extended to 80
or 120 zones by adding one or two expansion modules. The micro bore pipe has a maximum length limited
to 100 meters and one detector has a maximum coverage area of 1 600 m2, which is why the quoted system
has two detector units. For a deck with an area less than 1600 m2, almost half of the current deck size, the
cost would thus be nearly half. This assumes the same zone density and results in a cost of approximately
€ 65 000.
11.2.1.4 Gas detection in combination with ASD
The cost evaluation of the present system, as explained previously i.e. paragraph 8.2.3, has been done as
an add-on of the ASD system without addressability function.
Estimated total cost
The estimated total price is around 82 000. The total price includes, when available, the component cost,
the installation cost and the commissioning cost.
Here, are presented the cost for the gas detection system only, meaning that the cost of the entire system
(gas detection system in combination with ASD) is around € 197 000.
Components cost
Details on the component cost are detailed below, for a cost of 74 000.
46/91
FIRESAFE II | Bureau Veritas RISE Stena
Components
Number of units
Gas detector (Methane)
16
Gas detector (Propane)
16
Gas detector (Hydrogen)
16
Pipes, fittings and cables
Necessary amount for the installation
Installation cost
The installation cost is € 5000, based on ship owner estimation.
Commissioning cost
The commissioning cost is 2500. It includes commissioning of the system including all necessary tests and
configuration, as well as crew training on the system functionality.
System Scaling
System cost is highly dependent on addressability requirements, i.e. number of detector zones. Current cost
estimate is based on 8 detector zones, same as the drencher zones. Subsequently a system of half the size
would use 8 detectors and the price would be basically half of full deck system, likely a bit more as
commissioning and crew training will have to be formed in similar fashion for both installations. Cost of half
deck is estimated at € 45 000.
11.2.2 Systems for weather deck
Systems judged as relevant for the weather deck are video detection: smoke or combined smoke and flame
detection, video detection: thermal imaging camera, video detection of flames only and conventional flame
detection.
11.2.2.1 Video detection: Smoke or combined smoke and flame detection
For the detection system based on combined smoke and flame detection, two manufacturers have answered
that this system is not suitable for open ro-ro space. Indeed, it seems that the system is not yet robust enough
for an application in open ro-ro space, due to ambient conditions (light, light reflexion, etc.) causing a lot of
false alarms.
11.2.2.2 Video flame detection: thermal imaging camera
The below information is based on 2 quotes received from manufacturers.
System cost estimation
The estimated total cost is € 95 000.
Component cost
Details on the component cost are detailed below for a cost of € 65 000.
Bureau Veritas RISE Stena | FIRESAFE II
47/91
Components
Number of units
thermal imaging
cameras
3
camera houses
1
computer
1
software
1
software camera
module
1
Installation cost
The installation cost was estimated to € 20 000, based on ship owner experience.
Commissioning cost
The commissioning cost was estimated to about € 8000, based on 3 days of configuration and test work.
System Scaling
The system is expandable with essentially an unlimited number of cameras, so changing the deck size only
change the number of cameras which accounts for the majority of the component cost. The cost for a system
covering half the deck size was estimated to € 50 000 using the same premises as for a full deck system.
11.2.2.3 Video flame detection
The system requires 6 cameras to cover the deck. Three of the cameras will have integrated CCTV output,
which could be used for closer identification of the fire once the system has given alarm. Otherwise these
detectors work as any conventional detector, with an alarm signal when a fire has been detected. The
addressability precision is therefore the range/area one camera covers.
System cost estimation
The estimated total cost of the system is € 65 000.
Component cost
The component cost is detailed below, resulting in a cost of € 27 000.
Components
Number of units
Cameras
3
Camera (with CCTV output)
3
Address unit
6
VPT4 unit
6
End of line Unit
6
Installation cost
The installation cost was estimated, based on ship owner experience, to about € 35 000.
48/91
FIRESAFE II | Bureau Veritas RISE Stena
Commissioning cost
The commissioning cost was estimated to about 4 000, based on configuration and test work, as well as
installation spot check.
System Scaling
The system cost is highly dependent on the number of cameras required. A deck of half the size would likely
require half the number of cameras, possibly some more since it is not necessarily easier to cover a smaller
deck. It would also be desirable to have two cameras with CCTV. This system would then likely consist of 4
cameras (two of each type). Commissioning cost would be more or less the same, reaching a total estimation
for half a deck of approximately 47 000.
11.2.2.4 Flame detection
The below information is based on quotes received from two manufacturers. The basic difference between
these quotes is the camera itself, while all other details are the same. The first system uses a triple frequency
camera which will detect and analyse three different infrared (IR) frequencies to detect a fire. The second
system uses a frequency band camera which will detect and analyse a band of IR frequencies. These
cameras are more expensive and should be more robust in their detection. Both camera types will output, in
addition to a detection signal, normal CCTV signal via NTSC or PAL.
System cost estimation
The estimated total cost ranges from € 95 000 to € 115 000
Component cost
The component cost is detailed below, resulting in a cost range between € 40 000 and € 60 000.
Components
Number of units
Cameras with fastening
brackets
14
It should be noted here that CCTV is not a standard feature on conventional flame detectors, and this may
affect component cost.
Installation cost
The installation cost was estimated, based on ship owner experience, to about € 45 000.
Commissioning cost
The installation cost was estimated, based on ship owner experience, to about 8 000 €.
System Scaling
The system cost is highly dependent on the number of cameras required. A deck of half the size would likely
require half the number of cameras, possibly a little bit more as it is not necessarily easier to cover a smaller
deck. The cost for a system installed on a deck with half the size was estimated to approximately € 55 000
to € 66 000, depending on the used camera type.
11.2.3 Conclusion on the cost evaluation of alternative fire detection systems
For an overview of the cost evaluation of the detection systems, Table 14 shows the deck type where each
system can be installed, as well as the total cost, component cost, installation cost, commissioning cost and
the total cost for a system installed on half the deck size.