New Technologies for Suppression on Board Naval Craft (FiST) Final Report

Tommy Hertzberg, SP Boras, Sweden John A. Hiltz, DRDC – Atlantic Research Centre Rogier van der Wal, TNO, The Netherlands Michael Rahm, SP Boras, Sweden

Defence Research and Development Canada Scientific Report DRDC-RDDC-2015-R224 September 2015

IMPORTANT INFORMATIVE STATEMENTS

This work was carried out under the Canada/Netherlands/Sweden Cooperative Science and Technology Memorandum of Understanding (dated May 2003) as Project Arrangement Number 2010-06.

Template in use: SR Advanced_Oct_Release_EN_2015-08-14.dotm

© Her Majesty the Queen in Right of Canada, as represented by the Minister of National Defence, 2015 © Sa Majesté la Reine (en droit du Canada), telle que représentée par le ministre de la Défense nationale, 2015

Abstract

This report summarizes the results of a Project Arrangement entitled “New Fire Suppression Technologies on Board Naval Ships (FiST)” carried out under the Canada/Netherlands/Sweden Memorandum of Understanding on Cooperative Science and Technology. The FiST project had three main areas of focus; 1) fixed fire suppression systems, 2) portable, manually operated fire suppression systems, and 3) on submarines. For fixed fire suppression systems, large scale fire suppression testing was carried out to determine the effectiveness of low pressure water mist systems under well ventilated conditions, the effectiveness of high pressure water mist systems in a damaged condition, and the effectiveness of using water mist in conjunction with deluge systems for the protection of ammunition storage spaces. For the damaged high pressure system, damage was simulated by reducing the number of operational nozzles, by reducing system pressure and by introducing damaged water delivery pipe segments. The results of large scale fire suppression testing of a dual agent (water mist / Novec™ 1230) system are reported. The use of water mist in conjunction with Novec™ 1230 was found to significantly reduce the levels of acid gas in the test space. The effectiveness of gaseous fire suppression agents, including Novec™ 1230 (a perfluorinated ketone), carbon dioxide and nitrogen, in suppressing or extinguishing electrical cabinet was investigated. This study included analysis of acid gases produced by Novec™ 1230. For portable, manually operated systems, testing focused on the evaluation of the effectiveness, toxicity and corrosiveness of hand-held aerosol fire suppression agents. Other technologies, including compressed air foam systems, the use of additives in water mist systems and cool gas generator technology for use in fire suppression systems, are also reviewed and discussed.

Significance to Defence and Security

The full scale fire suppression testing results for low and high pressure water mist systems provide Royal Canadian Navy and engineering support personnel with information relevant to the selection and design of water-based fire suppression systems for new build naval vessels. The results are discussed with respect to the strengths and weaknesses of the systems and how, if they are installed on ships, they can be used to ensure optimum . The testing of damaged systems provides knowledge of the residual capacity of these systems and how this can be optimized. The evaluation of Novec™ 1230 provides information on, in addition to its effectiveness, the potential problems arising from the thermal degradation products (especially hydrogen fluoride (HF)) of this gaseous agent and approaches to reducing hazards associated with its use. The dual water mist / Novec™ 1230 fire suppression system testing indicated that HF levels can be significantly reduced if used in conjunction with a water-based suppression system. The electrical cabinet fire suppression testing showed that ventilation can be used to reduce HF levels in a space where the agent have been used. This work also indicates that re-entry procedures to spaces where ‘new’ fire suppression agents, such as Novec™ 1230, are used should be reviewed and changed to minimize hazards to the ship’s crew. The results of the testing of hand-held aerosol fire suppression units provides new information on the effectiveness of these agents as a first responder device and hazards associated with their use. These include toxicity and the corrosivity of the aerosols when in contact with metal surfaces and computer circuit boards.

DRDC-RDDC-2015-R224 i

Résumé

Le présent rapport résume les résultats d’une entente de projet intitulée « Nouvelles technologies d’extinction d’incendie à bord des navires militaires (FiST) » s’inscrivant dans le protocole d’entente Canada/Pays Bas/Suède sur la recherche coopérative en matière de science et technologie. Le projet FiST comporte trois domaines d’intérêt particulier : 1) systèmes fixes d’extinction d’incendie, 2) systèmes d’extinction d’incendie manuels portatifs et 3) systèmes de lutte contre l’incendie à bord des sous-marins. Pour ce qui est des systèmes fixes d’extinction d’incendie, des essais à grande échelle ont été réalisés afin de déterminer l’efficacité des systèmes à brouillard d’eau à basse pression dans un endroit bien aéré, celle des systèmes à brouillard d’eau à haute pression en mauvais état et celle de l’utilisation d’un brouillard d’eau combiné à des systèmes de type déluge pour protéger les aires de stockage des munitions. Dans le cas d’un système à haute pression en mauvais état, on a simulé les dommages en réduisant le nombre de lances d’incendie opérationnelles, en réduisant la pression du système et en ajoutant des segments de conduites d’eau endommagées. On présente les résultats des essais d’extinction d’incendie à grande échelle sur un système utilisant deux agents (brouillard d’eau / NovecMC 1230). L’utilisation d’un brouillard d’eau conjuguée à du NovecMC 1230 permet de réduire considérablement le niveau des gaz acides dans l’espace d’essai. L’efficacité des agents d’extinction d’incendie gazeux, incluant le NovecMC 1230 (cétone perfluorée), le dioxyde de carbone et l’azote, pour la suppression ou l’extinction d’incendie dans une armoire électrique, a été étudiée. L’étude comportait une analyse des gaz acides produits par le NovecMC 1230. Dans le cas des systèmes manuels portatifs, les essais ont porté essentiellement sur l’évaluation de l’efficacité, de la toxicité et de la corrosivité des agents d’extinction d’incendie en aérosol. D’autres technologies, notamment les systèmes à mousse à air comprimé, les additifs utilisés dans les systèmes à brouillard d’eau et les générateurs de gaz à froid employés dans les systèmes d’extinction d’incendie, sont également examinées.

Importance pour la défense et la sécurité

Les résultats de l’essai des systèmes à brouillard d’eau à basse pression et à haute pression pour l’extinction d’incendie à grande échelle a fourni à la Marine royale du Canada et au personnel de soutien en ingénierie l’information pertinente sur le choix et la conception des systèmes d’extinction d’incendie à base d’eau pour les navires militaires nouvellement construits. On a examiné les résultats en tenant compte des points forts et des faiblesses des systèmes et de la manière dont on peut les utiliser pour assurer une protection optimale contre les incendies, lorsqu’ils sont installés à bord des navires. L’essai des systèmes endommagés permet de comprendre la capacité résiduelle de ces systèmes et la façon dont on peut l’optimiser. L’évaluation du NovecMC 1230 nous renseigne non seulement sur son efficacité, mais également sur les problèmes potentiels découlant des produits de dégradation thermique [particulièrement le fluorure d’hydrogène (HF)] de cet agent gazeux et des démarches visant à réduire les risques associés à son utilisation. L’essai du système d’extinction d’incendie utilisant deux agents (brouillard d’eau et NovecMC 1230) a indiqué que l’utilisation d’un brouillard d’eau combiné à du NovecMC 1230 permet de réduire considérablement le niveau de HF. L’essai portant sur l’extinction d’incendie dans une armoire électrique a montré que la ventilation permet de réduire ii DRDC-RDDC-2015-R224

le niveau de HF là où l’agent a été utilisé. Les travaux montrent également que les procédures de réintégration là où on a utilisé de « nouveaux » agents d’extinction d’incendie, comme le NovecMC 1230, devraient être examinées et modifiées afin de minimiser les risques pour l’équipage du navire. Les résultats des essais sur les appareils portatifs contenant des agents d’extinction d’incendie en aérosol nous informent sur l’efficacité de ces agents comme outils pour les premiers intervenants et sur les risques associés à leur utilisation. Ceux-ci comprennent la toxicité et la corrosivité des aérosols, lorsqu’ils sont en contact avec des surfaces métalliques et des cartes de circuits imprimés d’ordinateur.

DRDC-RDDC-2015-R224 iii

This page intentionally left blank.

iv DRDC-RDDC-2015-R224

Table of Contents

Abstract ...... i Significance to Defence and Security ...... i Résumé ...... ii Importance pour la défense et la sécurité ...... ii Table of Contents ...... v List of Figures ...... vii List of Tables ...... ix 1 Summary ...... 1 2 Introduction ...... 4 2.1 Objectives ...... 4 3 Fixed Systems ...... 6 3.1 Water Mist and Influence on Fire ...... 6 3.2 Water Mist Suppression Effectiveness in Battle-Damaged Conditions ..... 6 3.2.1 Threat Definition ...... 6 3.2.2 Effectiveness of Low Pressure Water Mist Systems ...... 7 3.2.3 Effectiveness of Damaged High Pressure Water Mist Systems ..... 10 3.2.4 Effectiveness in Breached Compartments ...... 14 3.2.5 Ruggedizing Methods ...... 15 3.2.6 Design: High Pressure Systems Versus Low Pressure Water Mist Systems . 15 3.2.6.1 Differences Between High and Low Pressure Systems ..... 15 3.2.7 Survivability Aspects ...... 17 3.2.7.1 System Integration ...... 17 3.2.7.2 Mass ...... 17 3.2.7.3 Vulnerable Area ...... 17 3.2.7.4 Material Use ...... 17 3.2.8 Pressure and Pressure Drop ...... 18 3.2.9 Experimental Blast Resistance ...... 18 3.2.10 Finnish Testing on Damaged Water Mist System ...... 19 3.2.11 High Versus Low Pressure Systems ...... 20 3.3 Special Spaces ...... 20 3.4 Compressed Air Foam Systems (CAFS) ...... 21 3.5 Water Spray Systems in Ammunition Stores ...... 22 3.6 Models for Traditional Low Pressure Sprinkler Systems ...... 26 3.7 Dual Agent Systems ...... 28 3.8 Water Mist Additives ...... 32 3.9 Cool Gas Generator for Use in CAFS ...... 33 3.9.1 Compressor System ...... 34

DRDC-RDDC-2015-R224 v

3.9.2 Cool Gas Generator System ...... 34 3.9.3 Pressurized Bottle System ...... 34 3.9.4 Conclusions ...... 37 4 Portable, Manually-Operated Systems ...... 38 4.1 Portable Water Mist Fire Extinguishers ...... 38 4.2 Validation of Models for Current Fire Extinguishers ...... 40 4.3 Compressed Air Foam Systems (CAFS) ...... 40 4.3.1 Portable Independent Systems ...... 40 4.3.2 Portable Dependent Systems ...... 41 4.4 Aerosols ...... 41 5 New Fire Attack Strategies and Doctrines ...... 46 6 Firefighting on Board Submarines ...... 47 6.1 Extinguishing Systems on Board Submarines ...... 47 6.1.1 Water Spray Systems ...... 47 6.1.2 Foam and Foam-Water Spray Systems ...... 48 6.1.3 Compressed Air Foam Systems (CAFS) ...... 48 6.1.4 Conventional Water Mist (or Fine Water Spray) Systems ...... 48 6.1.5 Water Mist Systems Combining Water and Inert Gas ...... 48 6.1.6 NanoMist® ...... 48 6.1.7 High Expansion Foam Systems (Inside Air Systems) ...... 49 6.2 Electrical Cabinet Fire Suppression Systems...... 49 7 Conclusions ...... 52 8 FiST Reports, Memos and Conference Proceedings ...... 53 8.1 Reports and Memos ...... 53 8.2 Conference Presentations and Proceedings ...... 53 8.3 Journal Paper ...... 54 References ...... 56

vi DRDC-RDDC-2015-R224

List of Figures

Figure 1: Plots of Average Compartment Temperature Versus Time for Fire Suppression Tests Using the Low Pressure Water Spray System (10 bar and 5 bar) on 50% Obstructed Fires. The Results for a Free-burning Test with No Suppression System (Aborted 3 min After Ignition) are Also Shown. Extinguishment is Marked with a Thick Dot...... 9 Figure 2: Plots of Average Compartment Temperature Versus Time for Fire Suppression Tests Using the Low Pressure Water Spray System (10 bar) and 4 or 9 Operational Nozzles on 50% Obstructed Fires. The results for a Free-burning Test with No Suppression System (Aborted 3 min after Ignition) are Also Shown. Extinguishment is Marked with a Thick Dot...... 9 Figure 3: Section of Damaged Pipe (Left) and Spray Pattern from Pipe when Installed in Water Mist System (Right) for Damage Scenario 1. Damage Scenarios are Described in Appendix 1 of Reference 6...... 10 Figure 4: Plots of Average Compartment Temperatures as a Function of Time for Fire Suppression Tests Run at Full System Pressure (100 bar), 75% System Pressure, 50% System Pressure, 25% System Pressure and 5 bar System Pressure. The Fire was 50% Obstructed. Extinguishment is Marked with a Thick Dot...... 11 Figure 5: Plots of Average Compartment Temperatures as a Function of Time for Fire Suppression Tests Using the High Pressure Water Mist System with 4 Nozzles and 2 Nozzles Operational. The Fire was Not Obstructed. Extinguishment is Marked with a Thick Dot...... 12 Figure 6: Compartment Temperatures Versus Time in Tests with Damaged Pipes and Pressures Between 25% and 50% of Normal Operating Pressure. Damage Scenarios (Tests 19, 20, 24 and 27) are Compared with Intact System with Reduced Pressure (Tests 3 and 4). Damage Scenarios are Described in Appendix 1 of Reference 6...... 14 Figure 7: Overview of the Bunker with Spray Head Arrays Left and Right Against the Walls and Explosive Charge Hanging from Ropes [9]...... 19 Figure 8: Schematic Showing the Main Components of a Compressed Air Foam System (CAFS)...... 21 Figure 9: Dummy Torpedo Positioned Above the Fuel Pan. Thermocouple Positions are Shown as White X-es...... 23 Figure 10: Torpedo Dummy (Left) Above the Fire and (Right) Next to the Obstructed Fire Prior to Activation of Water Spray System...... 23 Figure 11: Plots of Peak Surface Temperatures Versus Time During Fire Testing of Dummy Torpedo in Position 1 (Above the Fire). Comparison of Results for the Free-burning Test (Aborted 3 Min after Ignition) and Different Water Spray Systems. Extinguishment is Marked with a Dot...... 25

DRDC-RDDC-2015-R224 vii

Figure 12: Plots of Peak Surface Temperatures Versus Time During Fire Testing of Dummy Torpedo in Position 2 (to the Side of the Fire). Comparison of Results for the Free-Burning Test (Aborted 3 Min after Ignition) and Different Water Spray Systems. Extinguishment is Marked with a Thick Dot...... 25 Figure 13: Schematic of the Test Compartment Used in the Novec™ 1230 / Water Mist Fire Suppression Testing...... 29 Figure 14: Mass of a Dependent CAF Back Pack System (with Different Gas Supplies) Necessary to Supply CAF for a Given Time. Dark Blue for Compressor, Light Blue for Pressurized Tank and Green for Cool Gas Generators...... 35 Figure 15: Volume of a Dependent CAF Back Pack System (with Different Gas Supplies) Necessary to Supply CAF for the Time Shown. Dark Red for Compressor, Orange for Pressurized Tank and Purple for Cool Gas Generators...... 35 Figure 16: From Left to Right, the Intelagard Macaw, Trimax Mini-CAF and NAFFCO CAF BP10l Portable CAFS Systems...... 41 Figure 17: Left: StatX First Responder, and Right: DSPA 5-4 Manual . ... 43

viii DRDC-RDDC-2015-R224

List of Tables

Table 1: Summary of Results for Damaged Low Pressure Water Spray Systems. ... 8 Table 2: Summary of Results for High Pressure Water Mist System Tests...... 11 Table 3: Summary of Results for Water Mist Fire Suppression Tests Using Damaged Pipe Segments. Damage Scenarios are Described in Appendix 1 of Reference 6...... 13 Table 4: Comparisons of Parameters for High and Low Pressure Water Mist Systems. . 16 Table 5: Criterion Analysis of Low and High Pressure Water Mist Systems from a Survivability Viewpoint...... 20 Table 6: Summary of Results for Fire Testing of Sand Filled Dummy Torpedo. .... 24 Table 7: Results of the Novec™ 1230 / Water Spray Fire Suppression Tests...... 31 Table 8: Criteria Analysis of CGG System Versus Compressor and Pressurised Systems...... 36 Table 9: Properties of Portable CAF Systems...... 41 Table 10: Characteristics and Specifications of the StatX FR and DSPA 5-4 Aerosol Extinguishers...... 44 Table 11: Summary of Results for Unobstructed Diesel Fires...... 44 Table 12: Summary of Results for Obstructed Diesel Fires. Burn Room Door Closed after the Activated Aerosol Unit was Placed in the Space...... 45 Table 13: Summary of Test 3a and 3b Results for Obstructed Diesel Bilge Fires. .... 45 Table 14: Extinguishment and Reignition Times for Electrical Cabinet Tests...... 50 Table 15: Measured HF Concentrations for Electrical Cabinet Fires Tests...... 50

DRDC-RDDC-2015-R224 ix

This page intentionally left blank.

x DRDC-RDDC-2015-R224

1 Summary

The design of fire suppression systems for naval vessels can either be based on prescriptive regulations or be performance-based. Prescriptive regulations arise from experience as they are created from established practices and knowledge from disastrous incidents. Many prescriptive regulations are based on either non-validated assumptions or on validations that have been lost. Frequently the origin and rationale for the regulation is not traceable and may have become obsolete due to the development of new methods, equipment and materials. Therefore it is often not possible to assure the applicability of the regulation. This can lead to installation of costly, poorly dimensioned and ineffective systems.

Prescriptive regulations have advantages. They are relatively simple to use, incorporate common experiences into standard protocols and are agreed upon as providing an adequate level of safety. From a liability point of view, prescriptive regulations free the design authority from responsibility, to some extent. A disadvantage of prescriptive regulations is that they have a tendency to favor existing technologies, making innovation and technological development more challenging.

The need and desire for innovation is an important reason behind the philosophy of ‘goal-based ship constructions’ that has been introduced in the International Maritime Organization (IMO) Safety of Life at Sea (SOLAS) code in 2002. As naval vessels are likely to have specific requirements for operability and threat management, performance based safety design is generally the only approach available.

A performance based design approach involves the establishment of agreed upon goals and objectives, deterministic and/or probabilistic analysis of fire scenarios and quantitative assessment of how alternative designs/systems meet the fire safety goals and objectives. To do so, accepted engineering tools, experimental methods and performance criteria are used.

Performance based design of firefighting systems may require more work than would be required if prescriptive regulations were used. However, performance-based design has the potential to provide a more cost effective system that is tailored to a particular vessel or to a space on that vessel. From the perspective of naval firefighting applications, there is also the potential for enhanced operability, effectiveness and survivability of the vessel. Performance based fire regulations have been adopted in SOLAS and the Naval Ship Code (NSC).

Research organisations from Canada, The Netherlands and Sweden have applied the performance based approach to a number of firefighting issues for naval vessels in the joint project New Fire Suppression Technologies on Board Naval Ships (FiST). Parties involved are DRDC in Canada, TNO and Royal Netherlands Navy in The Netherlands and SP and FOI in Sweden.

There are no prescriptive requirements regulating residual capacity of fixed fire-fighting systems after weapon induced damage. Such a requirement is an example of where a performance based approach is desirable. The results of our research illustrate what residual capacity might be achieved from a damaged system and can be used to select redundancy approaches to reach the required performance.

DRDC-RDDC-2015-R224 1

Test results indicate that a damaged low pressure water spray system (reduced number of nozzles and pressure reduced to 5 bar) still meets the performance criteria: the fire is quickly suppressed and the compartment temperature is reduced to below 60°C within seconds. Performance criteria are met when redundant water feed pipes each supply 50% of the nozzles in a compartment and the system is fed by the ship’s fire main.

Tests on a high pressure water mist system indicate that after reducing the number of nozzles and operational pressure to 25 bar the system is still compliant with the performance criteria. The compartment temperature did not rise above 150°C and was reduced to below 60°C within a few minutes. In such conditions, flashover did not occur and the ship divisions did not lose integrity. Even at fire main pressure, the system kept the compartment temperature below 350°C. This could prevent flashover and postpone or prevent fire breakthrough to adjacent spaces. When pump capacity is such that the operational pressure can be maintained at 25 bar, even in case of damage to the fire suppression system, the fire can be controlled and prevented from spreading.

There are prescriptive requirements regarding water discharge densities to be applied in weapon storage spaces ranging up to 32 L/m2min. It would be beneficial to reduce discharge densities as the installation of systems for handling and storing large volumes of water can be challenging. A performance-based approach was again used to identify suitable protection systems for a defined design fire scenario.

The performance criteria were based on the assumption that 200°C is the critical temperature for a fast heating phase and 150°C for a slow but prolonged heating of a weapon. The criteria were based on weapon cook-off tests performed at TNO. The testing showed that a system with a flow rate of 10 L/m2min met the performance criteria for our specific scenario. Prescriptive requirements in this case lead to an over-dimensioned and costly system.

At their best, prescriptive regulations can provide good quality fire safety design. At their worst, prescriptive regulations can result in expensive and poorly adapted systems for managing fire threats on naval vessels. Designers, researchers and regulatory bodies must challenge the prescriptive requirements and try to find better, safer solutions when this is financially and technically feasible.

In the FiST project we investigated performance criteria for cases where no regulations were prescribed, for example on damaged water mist systems. There have been suggestions for technical innovations, such as the use of bulkhead nozzles (rather than ceiling nozzles), more robust layout and local protection, residual capacity in the pump and the ability to connect water mist systems to the fire main. These innovations might have been inhibited by prescriptive regulations.

Furthermore, testing of prescriptive regulations for water spray systems for ammunition storage spaces indicated that these can lead to poorly designed systems. Poor design not only affects the system, but also influences ship construction. In the past designers have had to cope with the free surface effects of enormous water flows into magazine spaces during discharge of the system.

Canada was able to verify specific design options with respect to the use of Novec™ 1230. Knowledge on combining Novec™ 1230 with water mist or other water-based suppression systems may be implemented in future ship designs or mid-life upgrades. In Sweden the

2 DRDC-RDDC-2015-R224

knowledge has been applied in design of firefighting systems for submarine engine compartments. In the Netherlands the work from FiST will be input for selection and layout of water mist systems on new shipbuilding programs with high degree of automation. We are currently investigating opportunities to develop practical methods, rules of thumb and tools for ship designers to employ performance based approaches to fire suppression themselves.

DRDC-RDDC-2015-R224 3

2 Introduction

Over the past decades, naval damage control philosophies have changed significantly. There are a number of important reasons for this:  Recent changes in ship design are, among other things, driven by the fact that western navies are seeking to reduce manning. Manpower is expensive, so reduction in crew numbers may reduce the through-life costs of ships. There is also the concern that the interest in navy careers is declining. Reduced manning puts a strain on damage control (DC) effectiveness since DC is traditionally a manpower-intensive activity [1].  New legislation in the countries of all the Participants, aimed at preserving the environment, prohibits the use of ozone-depleting substances for fire-fighting agents. Therefore, DC in general, and fire-fighting in particular, must be achieved through solutions that require fewer crew members utilising healthy and environmentally-sound products.  Furthermore, new construction materials for naval craft, such as polymer composites, drive the need for new, effective fire safety measures to compensate for the combustibility of these construction materials.  Submarines have specific limitations with respect to fire-fighting agents due to their limited and confined living atmosphere.

Solutions for these issues can be found, e.g., in damage-tolerant automated and autonomous fixed fire-fighting systems, innovative extinguishing equipment and the development of doctrines, procedures, and methods dedicated to these new technologies. These solutions and developments must be suitable for the naval environment and fulfil naval requirements, including those for adequate performance when a fire suppression system is damaged.

2.1 Objectives

This project had several objectives. One was to improve the collective understanding of fire-related phenomena. A second was to more effectively fight fire on board naval craft in hostile environments using technologies and methodologies that are new to naval craft. Another was to develop practical guidelines and tools for the design and specification of new fire extinguishing technologies on naval craft. This project produced new information on damage-tolerant systems, and design guidelines and tools for their use on naval craft.

The early phase of the Project focussed on the assessment of the knowledge of the participants on the topics of interest. This was achieved by reviewing literature on the topics, discussions with international contacts and interaction with manufacturers. Industry was explicitly involved in the project, although the main focus of the participants was on consolidation and development of knowledge for design or modification of systems to perform adequately when damaged.

4 DRDC-RDDC-2015-R224

Technologies were identified and work packages were established to address existing knowledge gaps, which were identified by national knowledge surveys of the participants. These work packages were focused on three basic areas:

1. Fixed Systems (Chapter 3)

2. Portable, Manually Operated Systems (Chapter 4)

3. Fire-fighting on board submarines (Chapter 6)

Technologies within the first two areas were further developed within the third to suit the specific conditions on board submarines. The structure of this Scientific Report follows the work breakdown elements delineated in Project Arrangement Number 2010-06 under the Canada/Netherlands/Sweden Cooperative Science and Technology Memorandum of Understanding (dated May 2003).

DRDC-RDDC-2015-R224 5

3 Fixed Systems

3.1 Water Mist and Influence on Fire

Water mist refers to fine water sprays in which 99% of the droplets are less than 1000 microns in diameter. The water mist droplet size distributions are defined in the National Fire Protection Association (NFPA) Standard 750 as Class1 (90% of the volume of spray with diameters of 200 microns or less), Class 2 (90% of the volume of spray with diameters of 400 microns or less), and Class 3 (90% of the volume of spray with diameters greater than 400 microns). Water mist suppresses a fire through cooling, oxygen displacement, radiant heat attenuation, and the kinetic effect of water mist on flames.

Water mist can be generated in using different nozzles and pressures. Impingement nozzles have a large diameter orifice and a deflector and operate at low (12.0 bar or less) and intermediate (12.0 to 43.0 bar) pressures. Pressure jet nozzles have small diameter orifices (0.2 mm to 0.3 mm) or swirl chambers. Their operating pressures can range between low (5.1 bar) and high (272 bar). Twin fluid nozzles operate with a compressed gas (usually air) and water and consist of a water inlet, a compressed gas inlet and an internal mixing chamber. The pressures of the gas and water inlets are controlled separately and are in the low pressure region (between 3 bar and 12 bar).

Physical and theoretical descriptions of water droplet dynamics in a fire environment have been studied and reported. Droplet size distribution was analyzed. It was shown that droplet size distribution depends on nozzle type, water pressure and collisions between droplets. Further, the dynamics of water droplets were analyzed considering initial droplet velocity, retardation due to drag and estimation of terminal falling velocities when drag and gravity are in equilibrium. Heat transfer to water droplets with emphasis on vaporization of water droplets was discussed and applied to show fire extinguishing effects. Finally the absorption of heat radiation in water mist was analyzed [2].

3.2 Water Mist Suppression Effectiveness in Battle-Damaged Conditions

3.2.1 Threat Definition

Primary objective of the FiST project is to assess the residual performance of firefighting systems and equipment in case of damage from weapon effects following a weapon hit. In order to determine the damage to the systems and infrastructure typical threats to warships must be identified and engineering parameters must be quantified [3]. Using the engineering parameters, damage to the systems and post detonation fire loads (e.g., from residual weapon propellants) can be predicted. Typical expected damage scenarios are:  Perforations from bullets, shaped charge jets, direct fragments, spall.  Deformation up to rupture (of piping) from direct blast or from deformed structure.

6 DRDC-RDDC-2015-R224

 Malfunction or loss of function by blast induced shock motions from blast loaded structure.  Burning of combustible parts exposed to heat load from blast and fire.

Ambition levels for post incident capabilities are commonly quantified in three categories: fight, move and float. One could argue that for the fight and move ambition levels a water mist system should have adequate residual capacity after a hit. If the ship is only required to float after a specific incident, the water mist system may not be required if evacuation is arranged properly. How this relates to actual weapons is dependent on ship size and operational theater and such information supersedes the classification level of this report.

3.2.2 Effectiveness of Low Pressure Water Mist Systems

A number of full scale tests were run at SP Technical Research Institute of Sweden during late December 2011 and January–February 2012 [4], [5]. The main goals of the tests were to investigate possible redundancy solutions for active firefighting on board navy ships and to evaluate the effectiveness / residual capacity of a low pressure water mist system following damage.

The tests were conducted in a 55 m³ compartment with a 1.6 m² door opening. The walls of the test compartment consisted of wood studs with calcium silica panels. The water spray system consisted of nine open full cone nozzles with a K-factor of 5.93, mounted in a 3 x 3 pattern with a spacing of 1.75 m and operating at 10 bar. The performance criteria for the low pressure system were:  fast extinguishment (≤ 1 min) of a non-obstructed diesel pool fire in a well-ventilated environment; and  fast suppression (≤ 1 min) of an obstructed diesel pool fire in a well-ventilated environment where suppression is defined as:  reduction of incident radiation to 20% of that in the free burning condition; and  reduction of the difference in plate thermometer (PT) temperature and ambient temperature to 20% of that for free burning condition.

To comply with the performance criteria the water flow was adjusted to suppress and extinguish a fire in an open space, i.e., a space imitating a situation where decks and bulkheads were gone due to an explosion. This meant that no enclosure effects existed, e.g., suppression due to lowered oxygen content and the presence of combustion gases that play a significant role in closed spaces were not involved or were minimized in the tests. The amount of water used therefore was somewhat higher than the IMO standard (6 L/m2min).

A 1.08 m² steel pan filled with 16 litres of diesel fuel was used for the pool fire testing. The diesel was normally allowed to burn for 30 seconds prior to activation of the suppression system. The maximum heat release rate of a diesel pool fire this size is 1.3 MW. During the tests the fuel surface was either fully exposed to the water spray or obstructed by pipes or calcium silica boards. Two damage scenarios were tested, one where the number of operational nozzles was reduced and the other where the water pressure in the fire suppression system was reduced.

DRDC-RDDC-2015-R224 7

Reduced number of nozzles: These tests were performed based on the assumption that the water mist system in a compartment was fed by two pipes and that one of the pipes was damaged. Therefore only half of the nozzles were operational and they had a larger spacing than in the undamaged system. For these tests, 4 nozzles (instead of 9) with a spacing of 3.5 m (instead of 1.75 m) were used.

Reduced pressure: These tests were performed based on the assumption that 1) the piping delivering water to the nozzles was damaged or 2) the fire main was used to supply water to the system. Both of these scenarios result in reduced system water pressure. A system pressure of 5 bar (instead of 10 bar) was used for these tests.

Compartment temperatures presented in Figure 1 and Figure 2 are an average of the temperatures from two thermocouple trees. The time to extinguishment and the time until the temperature was reduced below 60°C are shown in Table 1 and the compartment temperatures as a function of time are plotted in Figure 1 and Figure 2.

Table 1: Summary of Results for Damaged Low Pressure Water Spray Systems.

Water discharge Time (after Time (after activation) Description of setup into the space activation) to to temp <60°C [s] [l/min] extinguishment [s] 9 nozzles, 50% obstruction 168.8 202 8 Full pressure (10 bar) 4 nozzles, 50% obstruction, 75.0 Did not extinguish 8 Full pressure (10 bar) 9 nozzles, 50% obstruction, 119.3 Did not extinguish 25 Reduced pressure (5 bar)

In the second part of the test series, damaged pipe segments were introduced to the water mist system and the ability of the damaged system to suppress the fire evaluated. The damage was inflicted by the explosion of an energetic weapon after which the ship should be able to fight (ambition level from Section 3.2.1), i.e., the water mist system must be able to fulfil the task it was designed for. In each of the damage scenarios a different damaged pipe segment or segments was used. One of these is shown in Figure 3 along with spray pattern from the pipe when installed in the water mist system.

8 DRDC-RDDC-2015-R224

600

Free-burning 500 10 bar 5 bar

400

C] °

300 Temperature [ Temperature 200

100

0 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 Time [min] Figure 1: Plots of Average Compartment Temperature Versus Time for Fire Suppression Tests Using the Low Pressure Water Spray System (10 bar and 5 bar) on 50% Obstructed Fires. The Results for a Free-burning Test with No Suppression System (Aborted 3 min After Ignition) are Also Shown. Extinguishment is Marked with a Thick Dot.

600

Free-burning 500 9 nozzles 4 nozzles

400

C] °

300 Temperature [ Temperature 200

100

0 0 0.5 1 1.5 2 2.5 3 3.5 4 4.5 5 Time [min] Figure 2: Plots of Average Compartment Temperature Versus Time for Fire Suppression Tests Using the Low Pressure Water Spray System (10 bar) and 4 or 9 Operational Nozzles on 50% Obstructed Fires. The results for a Free-burning Test with No Suppression System (Aborted 3 min after Ignition) are Also Shown. Extinguishment is Marked with a Thick Dot.

DRDC-RDDC-2015-R224 9

Figure 3: Section of Damaged Pipe (Left) and Spray Pattern from Pipe when Installed in Water Mist System (Right) for Damage Scenario 1. Damage Scenarios are Described in Appendix 1 of Reference 6.

3.2.3 Effectiveness of Damaged High Pressure Water Mist Systems

These tests are fully described in [6]. The goal of these tests was to study the residual capacity of a high pressure water mist system following damage. The tests were conducted in a 135 m³ compartment (7.5 m x 7.5 m x 2.4 m) with a 1.6 m² door opening. A high pressure (100 bar) water mist system with Ultrafog 202-2,09-O nozzles (K-factor of 2.09 l/min*bar1/2) complying with IMO, Maritime Safety Committee (MSC) Circ. 1165, 10 June 2005—Revised Guidelines for the Approval of Equivalent Water-Based Fire-Extinguishing Systems for Machinery Spaces and Cargo Pump-Rooms’ was used. The fire and obstruction for the 50% obstructed fires were the same as those used in the low pressure system tests.

In the first part of the test series two damage scenarios were tested. The first used reduced pressures to the nozzles (75 bar, 50 bar, 25 bar and 5 bar) to simulate pressure drops from a damaged pump or the use of the fire main for water supply (5 bar) and the second used two nozzles instead of four. Time to extinguishment and time until the average temperature in the test compartment was reduced below 60°C were used to measure the effectiveness of the damaged system. The results are shown in Table 2. The average compartment temperatures as a function of time are plotted in Figure 4 and Figure 5.

10 DRDC-RDDC-2015-R224

Table 2: Summary of Results for High Pressure Water Mist System Tests.

Time (after activation) Description of Discharge rate into Time (after activation) to to extinguishment setup compartment [l/min] temp <60°C [min:s] [min:s] 100% pressure 83.6 Did not extinguish 6:01 unobstructed 100% pressure 83.6 Did not extinguish 4:40 50% obstructed 75% pressure 72.4 Did not extinguish 8:55 50% obstructed 50% pressure 59.1 2:50 3:09 50% obstructed 25% pressure 41.8 3:20 5:08 50% obstructed 5 bar 18.7 Did not extinguish - 50% obstructed 100% pressure 2 nozzles (50%) 41.8 10:36 11:07 Unobstructed

Figure 4: Plots of Average Compartment Temperatures as a Function of Time for Fire Suppression Tests Run at Full System Pressure (100 bar), 75% System Pressure, 50% System Pressure, 25% System Pressure and 5 bar System Pressure. The Fire was 50% Obstructed. Extinguishment is Marked with a Thick Dot.

DRDC-RDDC-2015-R224 11

Figure 5: Plots of Average Compartment Temperatures as a Function of Time for Fire Suppression Tests Using the High Pressure Water Mist System with 4 Nozzles and 2 Nozzles Operational. The Fire was Not Obstructed. Extinguishment is Marked with a Thick Dot.

The tests showed that a high pressure water mist system (qualified according to MSC/Circ. 1165) was capable of reducing the compartment temperatures to levels where human lives could be saved even when the system pressure was reduced to 25% of normal operating pressure. The reduced temperatures would also prevent division (bulkhead) failure and eliminate the risk of flashover so that the fire could most likely be contained in the compartment of origin. Even at pressures as low as 5 bar, i.e., a typical fire main pressure, the compartment temperatures were reduced such that the probability of flashover was significantly reduced and a division failure would be delayed or avoided.

The tests with a reduced number of nozzles indicated that the average temperature in the space peaked at 150°C–170°C and decreased to about 80°C after 10 minutes. This implies that the suggested redundancy design using two separate water pump systems, each supplying 50% of the nozzles in an enclosure with water, would be an efficient fire risk control measure.

Extinguishment was found to be very scenario dependent for the conditions used in this test series. For instance, the fires were not extinguished at 100% and 75% of the full system pressure although the fire was suppressed. Changing test conditions, such as pre-burning time, ventilation and compartment size might result in varying performance of the water mist system.

The results for the damaged pipe segment tests are shown in Table 3. Plots of the compartment temperature against time for tests 19, 20, 24 and 27 are shown in Figure 6. The pressures at the water mist nozzles for these tests were between 29% and 35% of the system design pressure. For comparison the results for the ‘undamaged’ system at 25% and 50% pressure are also included in the plot. The effectiveness of the system with a damaged pipe segment was similar to that for the intact system at comparable pressures. For tests 19, 20, 24 and 27, the time for the temperature to drop below 60oC varied. The ability of the damaged system to extinguish the fire was dependent

12 DRDC-RDDC-2015-R224

on the damage to the pipe segment. In Test 19 the fire was not extinguished before the fuel was consumed. In Test 20 the fire was extinguished 2:00 min after activation, in Test 24 4:50 min after activation and in Test 27 10:59 min after activation of the water mist system.

Table 3: Summary of Results for Water Mist Fire Suppression Tests Using Damaged Pipe Segments. Damage Scenarios are Described in Appendix 1 of Reference 6. Average Water Time (after Time (after system discharge Total water activation) activation) Time (after activation) Test Damage pressure rate discharge** to temp to temp to extinguishment no scenario [% of design nozzles* [l/min] <80°C <60°C [min:s] pressure] [l/min] [min:s] [min:s] 18 scenario 1 68.9 69.4 103.4 1:19 1:40 1:30 19 scenario 2 34.3 49.0 104.6 1:52 6:55 Did not extinguish 20 scenario 3 28.8 44.9 104.9 1:57 2:11 2:00 21 scenario 4 22.0 39.2 105.0 1:22 2:00 2:49 22 scenario 5 16.8 34.3 105.5 Never > 80°C 3:09 6:43 23 scenario 6 64.8 67.3 103.4 0:57 1:14 1:25 24 scenario 7 35.8 50.0 105.0 1:25 4:49 4:50 25 scenario 8 2.9 14.2 106.5 - - Did not extinguish 26 scenario 9 13.7 30.9 105.8 1:40 5:28 Did not extinguish 27 scenario 10 29.7 45.6 105.0 5:44 11:09 10:59

*Flow through nozzles (flow through pipe puncture(s) excluded). **Flow through nozzles plus pipe puncture(s).

DRDC-RDDC-2015-R224 13

160 Test 19, Scenario 2, 34% pressure, comp temperature Test 20, Scenario 3, 29% pressure, comp temperature 140 Test 24, Scenario 7, 36% pressure, comp temperature Test 27, Scenario 10, 30% pressure, comp temperature 120 Test 3, 50% pressure, comp temperature Test 4, 25% pressure, comp temperature

C] 100 °

80

Temperature [ Temperature 60

40

20

0 0 2 4 6 8 10 12 14 Time [min]

Figure 6: Compartment Temperatures Versus Time in Tests with Damaged Pipes and Pressures Between 25% and 50% of Normal Operating Pressure. Damage Scenarios (Tests 19, 20, 24 and 27) are Compared with Intact System with Reduced Pressure (Tests 3 and 4). Damage Scenarios are Described in Appendix 1 of Reference 6.

Comparison of the results for tests using damaged pipe sections with the results of tests with intact (undamaged) piping but at reduced operating pressure indicated that the spray from the punctures in the damaged pipe generally improved the performance of the system as measured by average compartment temperatures, the mixing of gases as measured by the temperature uniformity factor in the test space and time to extinguishment. The temperature uniformity factor is described in Reference 6 and is a measure of mixing of the hot gases higher in the space with the cooler gases lower in the space.

This implied that a damaged system generally performs at least as well as an intact system with reduced operating pressure. This knowledge can be an important factor in decision making progress on whether or not to shut off a damaged system.

3.2.4 Effectiveness in Breached Compartments

The large scale tests described in Sections 3.2.2 and 3.2.3 [5], [6] were carried out in well ventilated spaces. This mimicked a breached compartment. Both the low pressure and high pressure water mist systems evaluated were capable of suppressing and, in most instances, extinguishing the fires in the space. The ‘damaged’ systems, that is those with reduced number of nozzles, reduced pressures, and damaged piping, also exhibited varying degrees of effectiveness in controlling fires in the space.

14 DRDC-RDDC-2015-R224

3.2.5 Ruggedizing Methods

The results of the large scale tests suggested approaches to reducing the vulnerability of water mists systems to damage. One is to install two separate systems in a space. In this configuration one half of the nozzles are fed by one pump and the other half of the nozzles by a second pump. If one system is damaged, the other will provide some fire suppression capabilities to the space. Another approach is to design the system so that if the water mist pump fails or is damaged that the system can be supplied by the fire main system on the ship. Even at the fire main pressure (5 bar to 10 bar), the water mist system provides some fire suppression capability. A third approach is to install a system with pumps capable of producing larger pressures than those required to operate in the undamaged condition. If piping is damaged, the pressure in the water mist system can be increased. This will increase the pressure at the nozzles and therefore the efficacy of the damaged system.

The use of water spray nozzles mounted to blast bulkheads was investigated [4], [5]. These bulkheads are less likely to sustain damage from blast and therefore the nozzles mounted on them can provide fire protection even if the primary fire protection system is destroyed by the blast. This approach suggests a more radical way to design the firefighting system on board ships to provide a higher degree of redundancy-safety. The higher degree of redundancy arises from the fact that the system is fed from another compartment that might remain intact after an explosion has destroyed the primary system. The tests showed that a water discharge density of 12–14 L/m2min is sufficient to extinguish an obstructed diesel pool fire in a ventilated compartment provided that the nozzle configuration distributes the water spray over the entire protected area.

3.2.6 Design: High Pressure Systems Versus Low Pressure Water Mist Systems

3.2.6.1 Differences Between High and Low Pressure Systems

High pressure and low pressure systems are tested and approved to the same IMO standards (MSC/Circ. 668/728/1165/1172), and systems being approved in accordance with the test standards do also, in general, have comparable firefighting capability within the type of spaces that the approval standards cover [6]. The difference between high and low pressure water mist systems is primarily to be found in the “nozzle” designs where different means of atomising water require different water pressures. The different water mist nozzle designs and the required water pressures affect requirements for water/gas supplies, pumps, controls, and piping for low pressure and high pressure systems. Comparisons of typical high and low pressure systems parameters are given in Table 4 [7].

DRDC-RDDC-2015-R224 15

Table 4: Comparisons of Parameters for High and Low Pressure Water Mist Systems.

Low Pressure Systems High Pressure Systems Approval test standards and The same as for high pressure systems The same as for low pressure systems acceptance criteria Typical water Working pressures: 3–12 bar Working pressures: 60–200 bar pressures Standby pressures: 3–12 bar Standby pressures: 5–20 bar Typical water droplet Dv90: 50– 200 μm Dv90: 200–350 μm sizes (Twin-fluid: Dv90: 50–200 μm) For systems with water delivery with pumps, the power requirement is proportional Power requirements to water pressure × water flow. Because of the much lower water Because of the much higher water pressure requirement of low pressure pressure requirement for high pressure systems, the power requirement also systems, the power requirement becomes

becomes less than that of high pressure greater than that of low pressure water water mist systems for the same total mist systems for the same total discharge discharge rate. rate. The required water densities depend on system design, system activation time, Typical water densities protected occupancy, ventilation, fuel type and arrangement, etc. Low pressure water mist nozzles in High pressure water mist nozzles in general have larger orifices than those general have smaller orifices than those Nozzle bores and sizes of high pressure water mist nozzles. of low pressure water mist nozzles. of the water ways. Orifice size depends on nozzle type Orifice size depends on nozzle type and and brand. brand. Because of larger nozzle orifices, low Because of smaller nozzle orifices, high Water filtration and pressure systems have less severe pressure systems have more severe water qualities. requirements for water filtration and requirements for water filtration and quality than high pressure systems quality than low pressure systems. Cylinder systems Centrifugal pump systems (sprinkler Cylinder systems Pumping systems pumps) Positive displacement pump systems Town mains Low pressure water mist systems operate in the same water pressure High pressure systems require that range as traditional sprinkler systems. piping and fittings are corrosion-resistant However the water mist nozzles have and suitable for high water pressures. smaller orifices than those of sprinkler Pipe systems Piping and fittings are typically made of nozzles. Therefore, filtration is stainless steel. The pipe sizes are typically required. The piping and typically smaller than those used in low fittings may be made of stainless steel, pressure systems. copper and plastics, depending on specific applications. Automatic systems often have a Automatic systems often have a lower System standby standby pressure, which is the same as standby pressure than the system pressures the system working pressure. working pressure. Nozzle spacing and Nozzle spacing and installation height depend on manufacturer and the protected height space, not whether the system is a high pressure or a low pressure system.

16 DRDC-RDDC-2015-R224

3.2.7 Survivability Aspects

3.2.7.1 System Integration

Low pressure systems are highly beneficial when it comes to system integration. They will have their separate infrastructure, but in case of an emergency a small jockey pump on the fire main can be used to generate the required pressure. High pressure systems require a separate infrastructure and pump capacity altogether. However, it has been shown that high pressure systems also have some residual firefighting capacity when operating at very low pressures, for example, at fire main pressures.

3.2.7.2 Mass

Marco Pesaloa of Eusebi Impianti argued that high pressure systems are generally more resistant to shock and blast because of their lower mass and smaller area of the piping [8]. Lengths of piping must be regarded as drag objects rather than pressure objects when subjected to blast from an explosion. This is true when the objects are smaller than the wave length of the blast. Blast is therefore the major contributor to the quasi static pressure (QSP), a longer term effect of the explosion. The QSP exists around the pipe and will therefore exert almost equal pressure from all directions.

There are two modes of pipe failure:

1. Buckling between supports by drag

2. Local buckling due to the blast wave

As both modes of failure are independent of mass, mass does not appear to be a crucial factor for survivability of pipe systems when regarding blast.

3.2.7.3 Vulnerable Area

In survivability, design vulnerable area is an important measure. It is the presented area of a component or object in the direction of explosion. The smaller the vulnerable area, the smaller the blast load exerted on it and the smaller the hit probability for fragments. Vulnerable area of a pipe is directly proportional to the diameter, so the smaller the diameter of piping in the system the better, for instance, the hit probability of a 50 mm tube is twice as high as that of a 25 mm tube. Standard Marioff HI-FOG® tubes are: 12 mm - branch, 16 mm - distribution, 25 mm - main distribution, 30 mm - main riser, 38 mm - main riser, 60 mm - mega system riser. Wall thickness of the 30 mm main riser is 2 mm. Tubing used in low pressure systems usually has a larger diameter when compared to high pressure systems. From this perspective, high pressure systems are intrinsically less vulnerable to fragment impacts.

3.2.7.4 Material Use

Marioff uses AISI 316L stainless steel for their tubing in their HI-FOG® high pressure systems. In some land-based low pressure systems regular carbon steel (ST-37) is used. For ballistic

DRDC-RDDC-2015-R224 17

performance against fragments, the resistance against plugging/shear is important. This is proportional to the yield strength. The yield strength, work hardening and thermal conductivity of AISI 316L stainless steel and regular carbon steel (ST-37) are comparable. For that reason there will be no significant difference in ballistic resistance of AISI 316L stainless steel compared to regular carbon steel. In that respect there is no preference for either a low pressure or high pressure system. However, because of the sea water environment, stainless steel tubing is recommended at all times.

3.2.8 Pressure and Pressure Drop

Carsten Palle, VID ApS founder, mentioned that the relative pressure drop of a low pressure system from the pump to the nozzle is generally larger (typically from say 10 bar at the pump to 4 bar at the nozzle) than that for a high pressure system (typically 100 bar at the pump and 70 bar at the nozzle). This might suggest that a low pressure system is more sensitive to pressure drops due to fragment holes or other damage. If one assumes that there’s a 20% loss in pressure that cannot be compensated for, for example, due to excessive water loss through a hole, this will result in the following scenarios:  For the low pressure system, the pressure will go from 10 bar to 8 bar. If there is a 6 bar loss in the piping (due to friction and height differences, this doesn’t necessarily have to scale down proportionally), then there will be 2 bar remaining at the nozzle. This is 50% of the design pressure.  For the high pressure system, the pressure will go from 100 to 80 bar. If there is a 30 bar loss in the piping, then there will be 50 bar remaining at the nozzle. This is 62.5% of the design pressure.

The work from Section 3.2.2 however suggests high efficiency of the low pressure water mist system down to pressures as low as the fire main (5 bar). So while the relative effect of pressure drop on low pressure systems may be greater, the actual performance of low pressure systems in controlling the fire may be similar to that of high pressure systems.

As flow (q) is equal to a hole size parameter (K) times the square root of pressure (√p), the amount of water lost from a hole of a particular size will increase with system pressure.

3.2.9 Experimental Blast Resistance

In 2008 TNO performed a series of six high explosive blast tests in their 390 m³ blast bunker at the Laboratory of Ballistic Research; see Figure 7 [9]. The Marioff HI-FOG® system survived the four subsequent internal explosions (in the order of 25 kg TNT equivalent bare charge) without visual damage to the tubing and nozzles. The Marioff HI-FOG® water mist system consisted of 20 spray heads of type 4S 1MC 8MC 1000 on 30 mm diameter tubing. The peak of the blast wave reached values as high as 6 MPa, the blast impulse (until 5 ms) 2.5 kPa∙s and the quasi static pressure 210 kPa. This suggests that tubing and nozzles are very damage tolerant against the direct blast effects of the internal explosion of a small anti-ship missile.

18 DRDC-RDDC-2015-R224

Figure 7: Overview of the Bunker with Spray Head Arrays Left and Right Against the Walls and Explosive Charge Hanging from Ropes [9].

3.2.10 Finnish Testing on Damaged Water Mist System

This testing [10] was conducted with a prefragmented warhead penetrating into an aluminum ship superstructure. According to Turkka Jäppinen, Finnish Naval Research Institute: “After the test firing, the on-board HI-FOG® system was inspected and several fragment hits on the piping network were found. Only one hit had been energetic enough to penetrate the 30-S (30 mm OD, 2.5 mm wall thickness) stainless steel main line just below the impact zone. The stainless steel nozzles didn’t show any signs of damage. The effect of such a hole in the piping is about equivalent of 2–3 extra water mist nozzles (the puncture’s K-factor is about 5.5). A pump unit can compensate for such extra flows up to the point where its maximum flow rate is reached. After that the discharge pressure will start to drop, and if the drop is significant enough it will have an effect on the fire protection performance of the water mist system. These kinds of systems are sectioned and should have ring mains and isolation valves so that the flow can be diverted around damaged sections if necessary. Actually the considerably high strength of stainless steel tubes and high wall thicknesses of high pressure water mist systems are beneficial from the survivability point of view when compared to low pressure water mist and/or conventional fire main systems which can be damaged (bent, ruptured, deformed) even by blast (pressure) effects only. The HI-FOG® system was remotely discharged immediately after the impact and functioned as designed (no fires aboard when the first response team boarded the ship) despite the damage to the piping.”

DRDC-RDDC-2015-R224 19

3.2.11 High Versus Low Pressure Systems

A criteria analysis of low and high pressure water mist systems is shown in Table 5. With the proper weighting factors for each of the criteria the recommendation for low pressure or high pressure systems can be made. The weighting factors may be different from one vessel to another.

Table 5: Criterion Analysis of Low and High Pressure Water Mist Systems from a Survivability Viewpoint. System Mass Vulnerable Material use Effectiveness integration area on pressure drop Low pressure systems + +/– – +/– +/– High pressure systems – +/– + +/– +

3.3 Special Spaces

Previous and current IMO requirements for the protection of roll on-roll off (ro-ro) cargo spaces, their background and development and relevant research have been reviewed [11]. The review indicated that the previous design and installation requirements, dating back to 1967, for water spray systems for ro-ro cargo and special category spaces are considered to be out-of-date. With the introduction of IMO MSC.1/Circ. 1430, the efficiency of “prescriptive-based” systems, i.e., traditional wet pipe, dry pipe, pre-action or deluge systems, can be considered sufficient for ro-ro cargo decks. Concerns have been raised about the expected efficiency of “performance-based” systems, which typically are water mist systems that are tested to a standard (IMO MSC.1/Circ. 1272) that does not reflect the potential magnitude of a fire on a ro-ro cargo deck.

Model scale fire tests and calculations have shown that sectioning/dividing of a cargo space into smaller volumes can be an efficient way to limit the size of a fire. This may be achieved with fire-resistant textiles. A water spray or water mist may be used to provide the cooling within the smaller volumes in the space.

The design and installation requirements of IMO MSC.1/Circ. 1430 can be considered appropriate for protection of spaces such as ro-ro decks and hangars on naval ships unless the fire hazards exceed those relevant for freight truck trailers. If flammable liquid fires are a concern, use of a foam additive in the water spray or mist system is recommended.

The ability of water sprays and mists to mitigate the effect of thermal radiation was reviewed. Several studies showed that water droplets in the form of mists and fogs are more effective in blocking thermal radiation than larger droplets found in water sprays. Theoretically the optimum droplet diameter is ≈ 1 µm, as droplets of this size absorb thermal radiation most effectively. However, droplets with diameters of 1 µm can be difficult to create in practice. Experiments have also indicated that water in itself is so effective in absorbing thermal radiation that it is difficult to increase its effectiveness through the addition of reasonable quantities of additives.

Research also indicated that a water spray curtain impinging on the outer surfaces of an object to be protected results in a higher level of shielding compared to a water curtain that does not impinge on the object. The attenuation of thermal radiation can be as high as 90%. This is an

20 DRDC-RDDC-2015-R224

important consideration when designing systems that reduce the transfer of thermal radiation from a fire to an object to be protected.

3.4 Compressed Air Foam Systems (CAFS)

A compressed air foam system (CAFS) is a water pumping system that has an entry point where compressed air can be added to a foam concentrate / water solution to generate foam. Typical components of a CAFS include a centrifugal pump, a water source, foam concentrate tanks, a rotary air compressor, a direct-injection foam proportioning system on the discharge side of the pump, a mixing chamber or device, and control systems to ensure the desired mixture of concentrate, water, and air are achieved. A schematic of a CAFS is shown in Figure 8. The compressed air also provides energy which propels compressed air foam farther than foam from aspirated or standard water nozzles.

Figure 8: Schematic Showing the Main Components of a Compressed Air Foam System (CAFS).

Compressed air foams (CAFs) suppress fires by reducing the availability of oxygen, absorbing heat from the fire, and shielding the fuel source from the radiant energy being produced by the fire.

CAFs have a number of properties that make them attractive fire suppressants compared to aspirated foams. CAFs have higher momentum. This allows the foam to penetrate the fire plume and reach the seat of the fire. They have improved stability due to smaller distribution of bubble sizes. The expansion ratios of foam can be adjusted resulting in improved flexibility for the fire attack team as the finished foam qualities, that is, water content and foam drain times, can be varied. Wet, fluid and dry foam consistencies are possible. These factors result in improved foam

DRDC-RDDC-2015-R224 21

effectiveness and a reduction in the amount of water and foam forming agent necessary for fire suppression. This in turn reduces clean up after a fire.

Foam formulations have been developed for both Class A and Class B fires. Class A fires are defined as those involving any solid combustible materials that are not metallic. The foam concentrates used on Class A fires are a mixture of foaming and wetting agents. When mixed in correct proportions with water, they change two properties of the water. Firstly, they increase the wetting effectiveness of the water. This allows for greater penetration of water into Class A fuels. Secondly, they result in the production of foam that can cling to vertical and horizontal surfaces. The increase in contact time allows the water in the foam to absorb more heat from those surfaces.

Class B fires are defined as those involving any non-metals that are in a liquid state or flammable gases. The foam concentrates used for Class B fires are designed to suppress flammable liquid and combustible gas fires. The type of foam used on a liquid fire will depend on the nature of the flammable liquid. For instance, alcohol resistant foams have been developed for suppression of alcohol fires or those involving certain polar solvents. Class B foams have two major subtypes; synthetic foams and protein foams.

Synthetic foams are based on synthetic surfactants and provide better flow and faster knockdown of flames, but give limited post-fire security. Aqueous film forming foams (AFFF) are water-based and frequently contain hydrocarbon-based surfactants, such as sodium alkyl sulfate, and fluorosurfactants. They have the ability to spread over the surface of hydrocarbon-based liquids. Alcohol-resistant aqueous film forming foams (AR-AFFF) are foams resistant to the action of alcohols and are able to form a protective film when alcohols are present.

Protein foams contain natural proteins as the foaming agent and are bio-degradable. Protein foams flow and spread slower than synthetic foams, but provide a foam blanket that is more heat resistant and more durable than synthetic foams. Protein foams include regular protein foam (P), fluoroprotein foam (FP), film forming fluoroprotein (FFFP), alcohol resistant fluoroprotein foam (AR-FP), and alcohol-resistant film forming fluoroprotein (AR-FFFP).

3.5 Water Spray Systems in Ammunition Stores

Alternative designs of weapon storage fire suppression systems were investigated [4], [5]. The goal of these tests was to investigate the prescribed requirements for fire protection of an ammunition storage space. For instance, NATO ANEP 77 (Naval Ship Code) [12] requires a system that operates at an application rate of at least 24 L/m2 per minute for an ammunition storage space.

The tests were conducted in the same compartment as the low pressure water spray system tests (Section 3.2.2). The fire was identical to that used for the water spray and water mist system tests. A dummy torpedo (3 mm thick steel cylinder, 2 m long and 0.35 m diameter) was used to evaluate the performance of the fire protection system alternatives evaluated in this series of tests. The dummy torpedo was filled with sand and the total weight of the steel tube and sand was 371 kg. The dummy torpedo and the positions of some of the thermocouples are shown in Figure 9. The dummy torpedo was either placed above or beside the fuel pan. When placed beside

22 DRDC-RDDC-2015-R224

the fuel pan, a calcium silica board was used to obstruct the fire. The dummy torpedo during fire suppression testing is shown in Figure 10.

Figure 9: Dummy Torpedo Positioned Above the Fuel Pan. Thermocouple Positions are Shown as White X-es.

Figure 10: Torpedo Dummy (Left) Above the Fire and (Right) Next to the Obstructed Fire Prior to Activation of Water Spray System.

DRDC-RDDC-2015-R224 23

The fires were suppressed or extinguished using a drencher system, a low pressure water spray system (WSS) or a combination of the drencher and water mist system. The drencher system consisted of 4 nozzles spaced 1.0 m apart and in-line with the dummy torpedo. It was run at discharge densities of 32 L/m2min, 10 L/m2min and 5 L/m2min. The water mist system was run at a discharge density of 6 L/m2min.

The performance criteria used to evaluate the performance of the candidate systems were:  maximum surface temperature of dummy torpedo must not exceed 200°C; and  60 seconds after activation of the system, the maximum surface temperature of the dummy torpedo must not exceed 150°C.

Peak surface temperature of the dummy torpedo and peak surface temperature of the dummy torpedo 1 minute after activation of the fire suppression system are shown in Table 6. Plots of the surface temperature of the dummy torpedo as a function of time are shown in Figure 11 and Figure 12.

Table 6: Summary of Results for Fire Testing of Sand Filled Dummy Torpedo.

Description of setup Peak surface Peak surface Time (after temperature temperature > 1 min activation) to [°C] after activation [°C] extinguishment [s] Free burning 550* - - Fuel pan position 1 WSS 6 L/m²min 203 203 Did not extinguish Fuel pan position 1 Drencher 32 L/m²min 128 39 24 Fuel pan position 1 Drencher 10 L/m²min 138 89 45 Fuel pan position 1 WSS 6 L/m²min + Drencher 5 L/m²min 148 92 53 Fuel pan position 1 WSS 6 L/m²min + Drencher 5 L/m²min 150 81 Did not extinguish Fuel pan position 2 Drencher 32 L/m²min 113 35 27 Fuel pan position 2 Drencher 10 L/m²min 162 82 Did not extinguish Fuel pan position 2 WSS 6 L/m²min + Drencher 5 L/m²min 151 105 Did not extinguish Fuel pan position 2 Free burning 604* - - Fuel pan position 2 * When the test was stopped, the temperature was still rising. Position 1: dummy torpedo above the fire; Position 2: dummy torpedo to the side of the fire.

24 DRDC-RDDC-2015-R224

600

Free-burning 500 Drencher 32 mm/min Drencher 5 mm/min + WSS

400 Drencher 10 mm/min C]

° Water spray (WSS) 6 mm/min

300 Temperature [ Temperature 200

100

0 0 1 2 3 4 5 6 7 8 9 10 Time [min] Figure 11: Plots of Peak Surface Temperatures Versus Time During Fire Testing of Dummy Torpedo in Position 1 (Above the Fire). Comparison of Results for the Free-burning Test (Aborted 3 Min after Ignition) and Different Water Spray Systems. Extinguishment is Marked with a Dot.

700

Free-burning 600 Drencher 32 mm/min

500 Drencher 5 mm/min + WSS

Drencher 10 mm/min C] ° 400

300 Temprature [ Temprature

200

100

0 0 1 2 3 4 5 6 7 8 9 10 Time [min] Figure 12: Plots of Peak Surface Temperatures Versus Time During Fire Testing of Dummy Torpedo in Position 2 (to the Side of the Fire). Comparison of Results for the Free-Burning Test (Aborted 3 Min after Ignition) and Different Water Spray Systems. Extinguishment is Marked with a Thick Dot.

DRDC-RDDC-2015-R224 25

The results of the tests indicated that, for the fire scenario studied, a much lower volume flow rate of water than stated in specifications was sufficient to keep the “torpedo dummy” below critical temperatures. If the design fire is a diesel pool fire and if the temperature criteria used in these tests are applicable, a drencher system providing a water discharge density of 10 L/m2min was found to be sufficient for ammunition store protection.

3.6 Models for Traditional Low Pressure Sprinkler Systems

Literature [13]–[16] reveals several reports on how to quantify the effect of sprinklers on the heat release rate of a wood crib fire. Although it may not be entirely representative for other fires, it was decided to apply this model since it seems justified from a physical viewpoint. Moreover, modification is possible by adjusting the constants. Evans [13] developed a correlation for the exponential decay time constant () of the fire heat release rate from the value at sprinkler actuation:

tt  act    Qf (t  t act )  Qf (t act ) e (1) in which  Qf = heat release rate of the fire (kW) t = time (s) tact = time of sprinkler actuation (s) τ = exponential decay time constant (s) τ = 435 (s) when 0.001 ≤ w " ≤ 0.07 (mms-1) τ = 3 -1.85 when ≥ 0.07 (mms-1) = spray density (mms-1)

The spray density is determined by taking the water flow (L/s) divided by the area covered by the sprinkler or sprinklers (m2). When the spray areas of adjacent sprinklers overlap, the total spray density can be obtained from summing the individual spray densities.

The value 0.001 (mms-1) as a lower bound for was chosen arbitrarily to prevent the unlikely situation of the modelled fire being extinguished with a sprinkler spray density of 0 (mms-1). Equation (1) implies that in the end any fire can be extinguished, even with limited quantities of water. This results from the fact that the suppression factor goes to zero for large t. This clearly is not a realistic result. To overcome this deficiency, an addition to the model is proposed based on the maximum cooling effect of the applied water. Because the energy absorbing capabilities of water are well quantified, they can be used as a basis to calculate the theoretical maximum fire size that can be extinguished. The energy needed to evaporate a certain quantity of water by heating from ambient temperature (or sea water at roughly 15°C) can be determined from:

26 DRDC-RDDC-2015-R224

373 c dT  4.2103  (373  288)  0.357 (kJ  g 1 )  p,l 288 -1 ∆Hv,373 = 2.26 (kJg ) -1 ∆Hg = 0.357 + 2.26 = 2.62 kJ/g = 2.62 (MJkg ) where -1 -1 cp,l = specific heat (kJkg K ) T = temperature (K) -1 ∆Hv = heat of vaporisation (kJkg ) -1 ∆Hg = heat of gasification (kJkg )

This means that 2.62 MJ is needed to change 1 kg of water into steam. Unfortunately, water must be applied at 10 to 100 times the theoretical rate to control and extinguish the fire [16] due to the fact that only part of the water will evaporate.

The maximum quenching capacity of a sprinkler system can now be determined using

n   Qs,max  Hg Vi (2) i1 where  Qs,max = maximum quenching effect (kW) n = number of sprinklers (-) β = efficiency factor (-) -1 ∆Hg = heat of gasification (kJkg ) ρ = density (kgm-3)  th 3 -1 Vi = (water) flow of the i sprinkler (m s )

Typically, the flow of a sprinkler will be in the order of 1 litre per second. For the efficiency factor we will assume 0.1, since it is likely that sprinklers are the most efficient of the extinguishers based on the principle of cooling, yielding a value of roughly 260 kW for the maximum quenching effect. Therefore, depending on the spray flow we get the maximum heat release rate of a fire that can be extinguished with one sprinkler. It is assumed that the effect of individual sprinklers may be summed.

During the FiST project it was concluded that sprinkler systems are unlikely to be used on naval platforms in the future. Hence there was no experimental work carried out to validate the above model.

DRDC-RDDC-2015-R224 27

3.7 Dual Agent Systems

Many naval vessels have dry chemical/foam dual agent systems on flight decks and aircraft hangers. The dry chemical is used for rapid knockdown of a fire and the foam is used to cool the fire and produce a foam layer on flammable liquids on the flight deck or in the hanger.

In this section, results of the evaluation of a dual agent system consisting of a water mist system and a gaseous fire suppression agent (Novec™ 1230) system will be discussed. The research is described in detail in [17].

Many legacy fire protection systems on naval vessels utilize Halon 1301 as a main fire suppression agent. However, because of restriction on the use of Halons, navies are actively seeking environmentally friendly and economic alternatives to use in onboard fire suppression systems. Novec™ 1230 (3M) is being investigated as one of these alternatives. Novec™ 1230 is a perfluorinated ketone (perfluoro(2-methyl-3-pentanone)) with a very short atmospheric life-time (ozone depleting potential of 0) and a global warming potential of one. However, it is a halocarbon and when exposed to open flame and high temperatures can produce thermal decomposition products (TDP). Hydrogen fluoride (HF) and carbonyl fluoride (COF2) are common decomposition by-products of halocarbons containing fluorine, such as Novec™ 1230, with HF being the primary by-product by a factor of two to three over COF2.

The concentration of TDP generated by gaseous fire suppressants is influenced by a number of factors including the temperatures to which they are exposed and the time of exposure to temperatures where thermal degradation can occur. Therefore, the generation of TDP from a suppressant can be reduced by ensuring that the concentration of the suppressant necessary for extinguishment is achieved in a very short time or the intensity of the fire is reduced (the space is cooled) prior to the release of the suppressant gas.

This study investigated the hazards associated with the use of Novec™ 1230 for fire suppression in a submarine machinery space and how these might be lessened by using a water mist system in conjunction with the Novec™ 1230 system. Full-scale fire tests were carried out in a mock-up of a submarine machinery space (7.32 m x 4.88 m x 4.2 m high, volume 148.5 m3) on large heptane fuel fires (1.49 m2, ~3 MW). A schematic of the test compartment with the positioning of water mist and Novec™ 1230 nozzles, thermocouples, pressure sensors, gas sampling ports and the fire pan is shown in Figure 13.

The fire suppression systems used in the tests were a FireFlex (SERVO) 3M Novec™ 1230 system and a low pressure (5.0 bar or 85 psi) water mist system with 36 L/min, FogJet 3/4 7G5 nozzles (Spraying Systems Co.).

Several configurations of the nozzles were used in the test series. Twelve nozzles, numbered 1 through 12, were installed on a 1.83 m by 1.52 m grid, as shown in Figure 13. To vary the water flow rate in the test compartment, the number of water mist nozzles utilized in the test was changed, by removing some water mist nozzles and plugging the outlets with pipe caps. For the 8 water mist nozzle configuration, a 1.83 m by 3.083 m grid was used with nozzles 1, 2, 3, 4, 9, 10, 11, and 12 operational. For the 6-nozzle configuration, a 1.52 m by 3.66 m grid was used with nozzles 2, 4, 5, 7, 10, and 12 operational. For the 4 water mist nozzle configuration, a 5.49 m by 3.083 m grid was used with nozzles 1, 4, 9, and 12 operational.

28 DRDC-RDDC-2015-R224

In the test series, a heptane pool fire was used to evaluate the effectiveness of the water mist / Novec™ 1230 dual fire suppression system in extinguishing the fire and to investigate the amount of thermal decomposition products produced by the system as well as the effectiveness of water mist in scrubbing HF. A rectangular pan, measuring 1.22 m by 1.22 m and 0.3 m high, was placed on the floor at the center of the test room. A metal hood was fitted over the pan such that both ends were open with an opening size of 1.22 m by 0.3 m. This provided a shielded fire, representing realistic fire scenarios in a submarine’s machinery space. Also, this shielded fire created more challenging conditions for extinguishment by the water mist system. In the pan, 20 litres of heptane was poured on top of 25 litres of water, providing 5 minutes of free-burn (13s/L of fuel consumed). The heptane pool fire was estimated to be approximately 3MW in intensity.

Figure 13: Schematic of the Test Compartment Used in the Novec™ 1230 / Water Mist Fire Suppression Testing.

Two Industrial MKS 2030 Multi-Gas FTIR spectrometers were used to measure gas concentrations in the test room at elevations 1.83 m from the floor and 10 cm below the ceiling. A heated sampling system, consisting of two heated lines kept at 180°C and a heated pump at 150°C, drew the gas samples to the FTIR spectrometers. The sampled gases flowed through a 225 ml gas cell (5.11 m path length) at a rate of 4 l/min. The gas cell was kept at 190°C. The FTIR spectrometer scanned the range from 400–6000 cm-1 with a 0.5 cm-1 resolution and acquired a full spectrum every second. A proprietary soft-ware calculated gas concentrations via a multivariate algorithm using the Classical Least Square fit and a library of calibration curves

DRDC-RDDC-2015-R224 29

generated at 190˚C. Gas concentrations were corrected for temperature and pressure effects by measuring the temperature and pressure inside the gas cell as a function of time.

The test results showed that water mist alone did not extinguish the shielded heptane pool fires while Novec™ 1230 did. However, the tests indicated that the use of Novec™ 1230 generates a large amount of HF and COF2. Although shielding the heptane pool fire reduced the levels of HF generated compared to tests with an unshielded fires, the concentration of HF in the test compartment was too high for the safety of personnel in the compartment. The results of the tests are shown in Table 7.

Activation of the water mist system prior to the activation of Novec™ 1230 system resulted in a substantial reduction in the production of HF. The water mist spray reduced the size of fire in the test compartment and consequently the maximum HF concentration in the test compartment. For example, in Test 12 (Novec™ 1230 extinguishment without water mist), the maximum HF concentration in the test compartment was approximately 750 ppm. By comparison, in Test 5 (water mist spray turned on for 30 s (from 4 nozzles with total water flow rate of 144 L/min) prior to the activation of Novec™ 1230 system), the maximum HF concentration dropped to 468 ppm.

HF levels in the test compartment were also dependent on the period of time the water mist system was activated prior to the activation of Novec™ 1230. For example, in Test 4 (30 s of water mist spray from 8 water mist nozzles prior to Novec™ 1230 discharge), the maximum HF concentration measured in the test compartment was 251 ppm. However, in Test 13 (65 s of water mist spray from 8 water mist nozzles prior to Novec™ 1230 discharge), the maximum HF concentration measured in the test compartment dropped to 95 ppm.

The levels of HF produced were reduced significantly if the water mist system was operated continuously during the test. That is, if the Novec™ 1230 system was activated while water mist system was operating. For example, in Test 14 (no water mist), the maximum concentrations of HF and COF2 in the test compartment were 1810 ppm and 950 ppm respectively. But in Test 15 (water mist system started 30 s prior to the activation of Novec™ 1230 system and left on for the duration of the test), the maximum concentration of HF in the test compartment was deduced to 58 ppm.

Water mist was also effective in scrubbing gaseous HF from a room. For instance in Test 14, activation of the water mist spray for 5 minutes following extinguishment reduced the HF concentration in the test space from 1810 ppm to less than 10 ppm. The pH of the water flowing out of the test compartment increased from less than 1 to approximately 7 after 5 minutes of water mist application.

The results of the test series showed that Novec™ 1230 would extinguish large liquid fuel fires in a submarine machinery space. However, it would also generate high concentrations of HF. The use of Novec™ 1230 in combination with a water mist system reduces the concentration of HF in the space. This reduction is realized when the water mist is activated prior to or during discharge of the Novec™ 1230 system. Because of the toxicity and corrosivity of HF, installation of a dual water mist / Novec™ 1230 fire suppression system will require measures to ensure that HF levels are reduced to safe levels prior to crew re-entering a space and that the space is washed down to remove HF from surfaces in the space. The water mist system should also be operated for an extended period (10 min or more) beyond the fire extinguishment in the machinery space in order to bring the pH level of the condensed water mist back to normal (pH 7).

30 DRDC-RDDC-2015-R224

Table 7: Results of the Novec™ 1230 / Water Spray Fire Suppression Tests.

DRDC-RDDC-2015-R224 31

3.8 Water Mist Additives

There has been a considerable amount of research in the use of additives for water mist systems. The majority of this research has investigated how (or if) the additives enhance the effectiveness of water mist fire suppression. There was one example of the use of additives to decrease the freezing point of water. In that instance, the water mist system was to be used at temperatures below 0oC. A review of the research referenced below is found in [18] and includes references to the original research papers.

The UK Ministry of Defence has carried out an extensive program that investigated the use of Aqueous Film forming Foam (AFFF) (MIL Specification MIL-F-24385 and Def Stan 42/40-1) and wetting agents to enhance the performance of a low pressure (7 bar) fine water spray fire suppression system. The results indicated that the time to extinguishment and burnback resistance of the water mist with AFFF (94/6) improved over that for water alone. The volume of water required for fire extinguishment decreased significantly when water with AFFF was used. It was also observed that when the water/AFFF (6% concentrate) ratio was reduced from 94/6 to 99.5/0.5 that the fire extinguishment times decreased. However, the burnback times also decreased. A 99/1 water/AFFF ratio was selected to optimize fire extinguishment and burnback properties.

Quad-Ex (primarily a mixture of potassium carbonate (K2CO3) in water) has been evaluated as a water mist additive in both high and low pressure systems. High pressure testing (~100 bar) was carried out in a 3.0 m x 3.0 m x 2.4 m box against eight small (1 kW) heptane pan fires. Fire suppression times were found to decrease as the concentration of Quad-Ex was increased to 50 wt%. However, small fires located 1.2 m and above the floor of the compartment were not extinguished. In a larger compartment (9.1 m x 9.1 m x 4.6 m) and using a low pressure system (5.5 bar), 12 wt% and 24 wt% solutions of Quad-Ex extinguished a heptane spray in 20 and 30 seconds respectively. Without the additive the system did not extinguish the fire. No studies of this additive’s toxicity were carried out. It was concluded that the use of additives, such as Quad-Ex, would only be considered if the optimized water mist system could not extinguish fires in the space it was protecting. It was noted that the water mist system used in the low pressure tests was not optimized.

The effectiveness of MC Additive as a water mist additive was evaluated on diesel and alcohol pool fires and wood crib fires in a 3.0 m x 3.0 m x 3.0 m compartment. Fire extinguishment times for the pool fires decreased as the concentration of additive was increased to 0.2 wt% and then began to increase again as the concentration of the additive was increased. For the wood crib fires, extinguishment times decreased as the concentration of additive was increased to 0.8 wt%. The enhanced performance of water mist systems using this additive have been attributed to both chemical and physical extinguishing mechanisms of the constituents of the additive.

Potassium acetate, added to lower the freezing point of water, was found to enhance the flame extinction of water mist in a cup burner apparatus. A 3.2 wt% solution of potassium acetate in water was approximately 2.3 times as effective as water alone. The enhancement did not increase linearly with potassium acetate concentration. This was attributed to two phenomena, a thermodynamic limitation of the concentration of catalytic species in the vicinity of the flame and a limitation of the catalytic species to lower the flame propagation radicals to their equilibrium concentration. This additive has not been tested in a larger scale fire suppression test.

32 DRDC-RDDC-2015-R224

Forafac® WM is a fluorinated surfactant that has been evaluated as a water mist additive in 9.1 m³ and 56 m³ compartments using an AM4 and Yulian nozzles respectively. Testing indicated that water containing 2 vol% of this additive was more effective as a fire suppressant than water alone. This additive also imparted burnback resistance to the water mist and improved the ability of the system to extinguish obstructed fires. The authors of the report stated that the additive was non-toxic and testing indicated the additive did not increase corrosion of metals over that observed for water alone.

Small scale tests (0.70 m x 0.46 m x 0.45 m compartment) indicated that 3% solutions (mass/volume) of alkali metal salts (NaCl, KCl, and KHCO3) in water significantly decreased extinction times for heptane fires over water alone. Solutions containing 3% (mass/volume) cobalt, zinc and manganese chlorides and sucrose were all more effective fire suppressants than water too, but were less effective than the alkali metal salts. The effectiveness of the alkali metal salts was attributed to the radical scavenging ability of the halide ions and the radical quenching properties of the alkali metal ions. Both disrupt the chemical reactions required for the propagation of the fire. These additives were not evaluated in large scale fire tests. NaCl and KCl would both promote corrosion of metallic components and solutions of them used in a water mist fire suppression system would require clean up after use on a ship. The additives containing Cl could produce hydrogen chloride (HCl) and this would have an effect on materials and re-entry times and/or procedures for the space in which they were used.

3.9 Cool Gas Generator for Use in CAFS

First responders in the damage control team can use the central water supply of a ship to provide the water for CAF. This water supply line runs throughout the ship and can be accessed at many locations. The pressure on this line is 7 to 12 bar, enough to produce CAF.

The proposed system consists of a gas supply, a foaming liquid tank and a CAF mixing chamber and nozzle which is carried on the back of a first responders. The water is supplied by a hose from the central water supply line. The system can be used as long as the gas and the foaming liquid supplies last. As the water is the major ingredient of the CAF and also the heaviest, such a system may remain active for a much longer time than an independent CAF system. The gas can be supplied in different ways, three have been studied here:

1. A compressor, powered by batteries or a petrol engine

2. A high pressure bottle

3. Cool Gas Generators

To assess the systems, a spreadsheet has been set up to investigate the size and volume of the systems in order to compare them. In the spreadsheet the mass and volumes of the systems can be computed as a function of the operation time. The following assumptions were made:  The central line supplies water at 7 bar or more in sufficient quantities.  The foaming liquid to water ratio is 2%.  The foaming liquid density is equal to that of water.

DRDC-RDDC-2015-R224 33

 Mixing chamber and nozzle are assumed to be the same for all systems and were therefore excluded from the assessment.  The water flow was assumed to be 23 litres per minute. This number was selected as it is in the same order as the flow of the independent system.

3.9.1 Compressor System

Information on the Airpress DC 12-170 compressor was found on the internet [19]. It is battery powered (Li-ion battery package in this case). The mass of the compressor is constant, only the number of batteries and the mass of the foaming liquid varies with the operation time. For each component (foaming liquid tank, compressor and battery pack) an estimate of the mass is made. For the batteries a performance of 150 Whr/kg is assumed and a volumetric performance of 300 Whr/L. The advantage of the compressor system is that its properties are not very dependent on time, that is, the mass and volume of the compressor are fixed. However, for short duration usage, this is a disadvantage.

3.9.2 Cool Gas Generator System

In this system, a cool gas generator (CGG) is used for the production of nitrogen gas. To buffer the gas expelled by the CGG, this system includes a small buffer tank. The system also has a foaming liquid tank similar to that used for the compressor system. The advantage of the CGG system is that for small operation times it is lighter than the Compressor System. However for longer operation times it is heavier. The storability, absence of maintenance and the low pressure of the CGG system are further advantages.

3.9.3 Pressurized Bottle System

The pressurized bottle system is similar to the CGG system, but instead of the buffer tank a regulator is included. The foaming liquid tank was also similar.

To estimate the mass and volume of the CAF systems, a statistical model was set up based on the Keesiedive website information on bottles [20]. In Figure 14 and Figure 15, the mass and volume for three pressure generating options are given as a function of time.

34 DRDC-RDDC-2015-R224

Mass versus operation time 50

40

30

20 Mass(kg)

10

0 0 5 10 15 20 25 30 35 Time (min)

Figure 14: Mass of a Dependent CAF Back Pack System (with Different Gas Supplies) Necessary to Supply CAF for a Given Time. Dark Blue for Compressor, Light Blue for Pressurized Tank and Green for Cool Gas Generators.

The dark blue line in Figure 14 represents the compressor system, the green line the CGG system and the light blue line the pressurized gas system. For operation times up to 10 minutes, the CGG system has the lowest mass, for longer times the compressor system is the lightest.

45 40

35

30 25 20

15 Volume(litres) 10 5 0 0 5 10 15 20 25 30 35 Time (min)

Figure 15: Volume of a Dependent CAF Back Pack System (with Different Gas Supplies) Necessary to Supply CAF for the Time Shown. Dark Red for Compressor, Orange for Pressurized Tank and Purple for Cool Gas Generators.

In Figure 15 the volumes of the systems are shown as function of the operation time in minutes. Dark red represents the compressor system, purple the CGG system and orange the pressurized

DRDC-RDDC-2015-R224 35

bottle system. Up to operation times of 7 minutes the pressurized and CGG system are very close in volume and both have a lower volume than the compressor system for up to 20 minutes of operation. A comparison of the properties of the three systems is shown in Table 8. The operation time was taken as 5 minutes.

Table 8: Criteria Analysis of CGG System Versus Compressor and Pressurised Systems.

Criterion Compressor system CGG system Pressurised system Mass - ++ + Volume - + + Development effort 0 -- 0 Reload time (after use) + + 0 Maintenance effort -- ++ 0 Damage tolerance 0 + -- Effect of an impact 0 / - 0 -- Resistance to harsh 0 ++ + environments Total score -3 +7 -1

+ - positive factor. - - negative factor. 0 - neutral.

The trade-off analysis shows that the CGG system is the best, as it combines low mass and volume with low maintenance, good damage tolerance and ruggedness. Where pressurized tanks might explode when damaged by a high speed impact and broken compressor blades may be come projectiles, the CGG will just stop producing gas.

When the system is empty, the compressor system will have to be equipped with a new loaded battery pack and for the CGG system, the CGG must be replaced. For the pressurized system, a new pressurized bottle must be attached, which is more complicated due to the high pressure. CGG’s are inert during storage, where loaded battery packs and pressurized bottles might cause explosions and fire when hit.

On the maintenance side, the CGG also has an advantage: it does not need any maintenance for a period of ten years. The compressor system will need regular maintenance and service to keep it going due to the precision moving parts. A pressurized bottle needs regular inspection and maintenance as it is always pressurized during storage. The maintenance costs are significant and will surpass the original purchase price easily within five to ten years of service.

36 DRDC-RDDC-2015-R224

3.9.4 Conclusions

The following conclusions can be drawn from this study:

1. CAFS are an efficient method of fire suppression.

2. Several mobile and portable CAF systems have been identified and described.

3. Portable CAF systems have potential to be used as first responder equipment on board ships. Both independent (with their own water supply) and dependent systems (which need external water) can be made small enough to make them portable.

4. Options for independent backpack systems have been identified and analysed.

5. Cool Gas Generator technology could replace the high pressure vessel needed in independent systems, as CGG technology has several advantages.

6. A number of dependent backpack systems have been identified and analysed. The damage tolerance, lack of maintenance and low mass are advantages of a CGG system.

7. CGG may be cost effective as maintenance on board of ships is difficult and costly and the higher development costs of a CGG system may be more than compensated by the absence of maintenance.

DRDC-RDDC-2015-R224 37

4 Portable, Manually-Operated Systems

Portable fire extinguishers are an integral part of firefighting capabilities on board naval vessels. In many cases they allow crew to extinguish or suppress small fires before they can grow and become a significant threat to life and the vessel itself. There have been a number of developments in portable and hand-held extinguishers in the last twenty years. Many of these use technologies that have been developed to replace Halons in total flooding and local application or portable extinguishment systems. These technologies include water mist, aqueous film forming foams, compressed air foam systems, cutting extinguishers and propelled extinguishing agents technologies () or aerosols. However, the requirement that the system be hand-held or portable limits the volume of extinguishing agent that is available for fighting fires. The shipboard use of certain portable units will depend on whether or not the unit can be moved around the ship. This is the case for portable CAFS (non hand-held) and the cutting extinguisher system.

4.1 Portable Water Mist Fire Extinguishers

Södra Älvsborg Fire & Rescue Service (SERF) in collaboration with the SP Technical Research Institute of Sweden has conducted scientific studies of the Cutting Extinguishing Concept (CEC) and methodology in firefighting operations [21].

Firefighting inside burning buildings or spaces is, from a health and safety perspective, an activity with a very high level of risk. To reduce these risks, there is a requirement for new firefighting methods. In response to these requirements, The Swedish Rescue Services Agency (SRSA) now the Swedish Civil Contingencies Agency (MSB) initiated a program in 1996 which resulted in the development of the cutting extinguishing tool (COBRA) that enables a completely new methodology to be used for fighting fires.

The concept or system consists of fire detection using infrared technology, information and decision support, the COBRA cutting and extinguishing equipment for precision firefighting, and high-pressure ventilation to optimize the efficiency of the COBRA. The COBRA system is ready for use immediately upon arrival at the fire scene. The system can be integrated into normal fire appliances (vehicles) but is also available as part of the lighter quick response unit, the First Response Unit, developed by SRSA.

The COBRA was deployed in 675 firefighting operations during the period 2004–2008 in Sweden. The experiences have been compiled and scientific studies of the reported experiences underline the importance of the COBRA´s cutting capacity for quick access to a fire compartment or adjacent rooms and rapid fire suppression response action. The studies indicate that the COBRA was chosen for use to avoid the risk associated with the ignition of the accumulated fire gases in a space, its ability to attack the fire directly from outside the space, and its ability to slow the development of the fire.

As the COBRA produces a high pressure water mist, it works by cooling and inerting a space. The heat in the space turns the mist into steam which lowers the oxygen content to a point where it will not support combustion of the flammable gases in the space.

38 DRDC-RDDC-2015-R224

There are a number of conclusions in the report [21] concerning the Cutting Extinguishing Concept. It:  efficiently cools the fire gases in a space and stops the fire from developing even when the temperature in the space is low;  facilitates the use of high-pressure ventilation as it cools and dilutes flammable gases;  enables rapid fire suppression actions;  provides a method for extinguishing fires which are generally considered difficult to get access to, for example, fires in double flooring, roofs and attics;  increases choices/options on the tactical approaches to fire suppression as it combines IR detection technology, the COBRA and positive pressure ventilation;  decreases damage to property and the environment caused by conventional firefighting techniques that use large quantities of water; and  allows fire to be suppressed from outside the space and reduces risks to as they do not have to enter the space.

The report presents the results of studies of the ability of water and vaporized water drops to extinguish fires. Four case studies of fires in which the CEC has been implemented are presented in detail. Finally, proposals are made for future work and further development of the COBRA.

The COBRA is used actively for firefighting in different parts of Sweden. However, there is a need for more information on how fire suppression should be carried out and the effect of new approaches on fire suppression. The improved knowledge arising from this information would enhance and facilitate the exchange of experience and lessons learned within the fire and rescue services. This will accelerate the introduction of the new methodologies and technologies.

An education and training program for the Cutting Extinguishing Concept has been established in Sweden and forms part of the basic training of firefighters and intervention commanders. As part of the EU Project FIREFIGHT (http://www.eufirefight.com/documents.html), an e-learning package supplemented by a short practical training course for firefighters was developed. Partners in FIREFIGHT included fire training facilities in England, France, Spain and the Czech Republic with SRSA as the coordinator.

The study proposes that training facilities be adapted to allow for training in the complete CEC, i.e., IR technology, the COBRA and PPV (positive pressure ventilation). At present training establishments are not well set up for teaching CEC tactics, for example, cooling and inerting of the fire gas and air mixtures, particularly in large volume compartments.

Another conclusion is that the intervention reports clearly demonstrate a need for an improved and developed methodology for learning from the tactical response operations. The reports at present rarely contain an analysis of the appropriateness, efficiency, etc. of the implemented methodology. There is a need for the systematic evaluation of experiences with new methodologies and technologies to ensure that the lessons learned can be used by others.

DRDC-RDDC-2015-R224 39

4.2 Validation of Models for Current Fire Extinguishers

For this work package reference is made to Svensk Standard 5S 1192 [22]. The standard is based on the results of ongoing work within ISO on the preparation of an international standard for portable fire extinguishers and Nordtest project 82-77 (SP:RAPP 1981 :52) Portable fire extinguishers – Test methods.

A wood crib fire is ignited using propane and the propane is supplied to continue the burn for 3 minutes. The gas is then shut off and the table is rotated at 5 rpm. After a total burning time of 8 minutes, the extinguisher is discharged with the jet of the extinguishing medium aimed at the centre of the wood crib fire. For non-pressurized extinguishers the time for pressurizing them is set at 5 s. During the test, the operator may only raise the nozzle to compensate for the decrease in throw length due to the pressure drop in the extinguisher. The extinguisher must be discharged continuously without interruption and with the shut-off device fully open. After the test, the wood crib fire is observed for 5 min and the time to reignition, if this occurs, is noted.

Class C fire extinguishers are for electrically energized fires, i.e., Class ABC or Class BC, in North America. In European classification a Class E (electrical) fire would be handled with a dry chemical or carbon dioxide extinguisher (i.e., class BE or ABE). Reference [23] reports a 500 kW maximum capacity for a 6 kg powder extinguisher. Reference [24] reports a 650 kW for a 9 kg powder extinguisher. Based on recent developments in chemical recipe of the powders and classification of powder extinguishers, the maximum efficiency of a 12 kg powder extinguisher is around 1 MW.

Regarding equivalence of powder to carbon dioxide extinguishers, DNV Statutory Interpretations [25] states:  50 kg dry powder or 45 kg carbon dioxide is considered as equivalent to 135 l foam liquid; and  25 kg dry powder or 20 kg carbon dioxide is considered as equivalent to 45 l foam liquid.

British Steamship P&I suggest the following capacities may be taken as equivalents [26], as a rule of thumb:  9 litre fluid extinguisher (water or foam)  5 kg dry powder  5 kg carbon dioxide

4.3 Compressed Air Foam Systems (CAFS)

4.3.1 Portable Independent Systems

Three portable systems have been identified in this study, the NAFFCO CAF BP 10 l, the Intelagard Macaw and the Tri-Max 3 Mini-CAF systems. These are shown in Figure 16. They all carry their own water, foam solution supply and a pressurized air bottle. A comparison of the properties of these systems is shown in Table 9.

40 DRDC-RDDC-2015-R224

Table 9: Properties of Portable CAF Systems.

Type Intelagard Macaw Trimax 3 mini CAF NAFFCO CAF BP 10 l capacity (gal) 5 0003 capacity (litre) 19 11.4 10 Working pressure (bar) 6.9 8 8 Discharge duration dry (s) 38 30 42 Discharge duration wet (s) 80 liquid flow (litre/min) est 30 22.8 14.3 Gas flow (normal litre/min) 222 168.7 140 gas capacity (normal litre) 140.6 84.36 200–300 Weight (kg) 27.2 16.3 23.2

Figure 16: From Left to Right, the Intelagard Macaw, Trimax Mini-CAF and NAFFCO CAF BP10l Portable CAFS Systems.

4.3.2 Portable Dependent Systems

As a part of this study, the availability of dependable and portable fire suppression systems was investigated. Small compressors with the right delivery pressure and volumetric gas flow are available [19] that have a mass of below 10 kg. Also small pumps with masses under 10 kg and the right capacity and pressure are available [27]. Therefore it seems possible to design a petrol driven backpack system with a pump and compressor that weighs less than 30 kg and can deliver a near continuous flow of foam when connected to an outside water supply. However, no commercially available system was actually found.

The concept of a portable CAF system using the central water supply of the ship was conceived during this study (Section 3.9). In this system, a small backpack carried system provides air and a foaming liquid and the water is supplied at 7 bar from the ship fire suppression system.

DRDC-RDDC-2015-R224 41

4.4 Aerosols

Research focused on the efficacy of hand-held pyrotechnic aerosol generators and the toxicity and corrosivity of the aerosols and their residues is described in references [28]–[30].

Two variants of hand-held pyrotechnic aerosol extinguishers (StatX First Responder and DSPA 5-4 Manual Firefighter) were studied. These units are shown in Figure 17 and their specifications are listed in Table 10.

Aerosol fire suppressants of the type investigated in this study form micron-sized alkali metal salts such as potassium carbonate. Alkali metal salts, in particular sodium and potassium bicarbonate, have been used as fire suppressants since at least the early 1940s. The first NFPA standard (NFPA 2010 – Standard for fixed aerosol fire-extinguishing systems) was issued in 1955 and therefore their efficacy as fire suppressants has been known for many decades. Since the 1980s, attention and research efforts into aerosol fire suppression have focused on micron-sized alkali metal salts suspended and delivered to the fire environment in a gaseous medium, as the inverse relationship between particle size and suppression efficacy was better understood. Another reason for recent attention given to alkali metal salt aerosol extinguishers is that they are a relatively inexpensive technology and have been identified by several organizations such as the United States Environmental Protection Agency as safe and suitable substitutes for Halon 1301.

The primary objective of this research was to evaluate the fire suppression efficacy of StatX First Responder and DSPA 5-4 Manual Firefighter units against simulated, repeatable, marine fire scenarios. To achieve this objective an instrumented single fire compartment experiment with fuel loading and ventilation parameters set to produce a consistent and repeatable fully developed fire environment was set up and characterized. Live fire suppression tests were conducted to evaluate the impact of the aerosols on key suppression parameters such as upper gas layer cooling rate, thermal stratification, and total compartment cooling effect. A summary of the results of tests on the effectiveness of the two aerosols on unobstructed and obstructed diesel pool tests is given below. The effectiveness of the aerosols on wood crib fires was also studied and the results are given in references [28]–[29].

Secondary objectives of the research included the evaluation of the safety and safe storage of the pyrotechnically activated aerosol units and the impact of evolved gases and aerosol particulate deposition on equipment. These were addressed respectively through dedicated tests to evaluate the safe storage and incendiary potential of the units and via agent only (or cold compartment discharge) tests in the compartment. The details and results of these tests were reported in references [28]–[30].

While the testing was designed to evaluate hand-held aerosol extinguishers, the analysis and recommendations pertain to the broader subject of pyrotechnically generated aerosol agents in general, either as possible Halon 1301 replacement agents or as a technology that may be scaled to suit multiple marine applications.

42 DRDC-RDDC-2015-R224

Figure 17: Left: StatX First Responder, and Right: DSPA 5-4 Manual Firefighter.

Fire suppression testing was conducted in a 2.4 m wide x 3.5 m long x 2.4 m tall room (floor area 8.4 m2 and a compartment volume of 20.2 m3). The room had a single door 0.91 m wide x 1.75 m tall that was 0.05 m off the compartment floor to simulate the marine door to the machinery space. The dimensions and thermal properties of the burn room were designed to be similar to an ISO 9705:1993 ‘Fire tests – Full-scale room test for surface products’ test room.

For diesel characterization tests, the 0.82 m2 fuel pan was filled with 10 ℓ of diesel floated on 105 ℓ of water (reducing the freeboard to 1 cm). Both unobstructed and obstructed pan fires (0.82 m2) were run. The average heat release rate (HRR) for this fuel load was 903 kW. HRR is defined at www.iafss.org/publications/fss/9/1165/view. For the obstructed fires, an engine mock-up 1.4 m x 1.3 m x 0.46 m (width x length x height), constructed of sheet steel of nominal thickness 5 mm was placed directly over the fuel pan. It was suspended 40 cm above the burn room firebrick floor to simulate an engine enclosure obstructing the diesel pan fire.

Tests were conducted using a fully developed unobstructed fire (3 minute preburn and temperatures at the top of the burn room greater than 500oC) to compare the effect of oxygen starvation with the efficacy of the aerosols. The door of the burn room was left open 30 cm when using the aerosols. The results are shown in Table 11. They indicate that the aerosols result in similar cooling rates and total cooling effects as those for the fire extinguished by oxygen starvation.

DRDC-RDDC-2015-R224 43

Table 10: Characteristics and Specifications of the StatX FR and DSPA 5-4 Aerosol Extinguishers. Characteristics StatX FR DSPA 5-4 Active Substance 0.5 kg 0.9 kg Discharging Time 20 sec 25 sec Maximum Temperature at 50 cm Unknown 75oC Diameter 80.89 mm 165 mm Height 174.6 94 mm Weight 1.22 kg 2.0 kg Capacity (based on 50 g/m³) 20 m3 18 m3 Fuse Delay Timer 3.5–5 sec 6–10 sec Storage Life 10 years 5 years UL/ULC Listing for fixed systems Class A (or not hermetically sealed) 97 g/m3 96.4 g/m3 Class B 67 g/m3 32.1 g/m3

Table 11: Summary of Results for Unobstructed Diesel Fires. Peak Average Temp Total Cooling Effect Test Cooling Rate (oC/s)* [oC] [m·K] 1a: StatX FR 401 4.9 484 1b: StatX FR 385 5.4 518 1a: DSPA 5-4 504 5.2 558 1b: DSPA 5-4 402 16** 218 Closed Door 519 4.8 540 *Note 1: Cooling rates assessed from onset of suppression to 60 seconds after peak average temperature. **Note 2: Cooling rate is only for 9 seconds between peak average and lowest average temperature since the fire was not extinguished in this test.

The aerosols had very little effect on the obstructed diesel fires. Temperatures in the test space dropped for a few seconds, but the fire was not extinguished and continued to burn. The fire test scenario was altered to determine if the aerosols would enhance the effect of oxygen starvation of the fire. This alteration involved closing the door of the burn room after the activated aerosol unit was placed in the space. The results are shown in Table 12.

Both aerosol units had a significant effect on the fire. The StatX FR unit increased the cooling rate by 21% and the total cooling effect by 43% over those for oxygen starvation alone. The DSPA 5-4 increased the cooling rate by 32% and the total cooling effect by 29% over those for oxygen starvation alone. These results show the benefit of employing aerosol agents during suppression of a fire, even when that fire can be confined and starved of oxygen. A 30–40% higher rate of cooling of the compartment in the same 60 seconds can significantly reduce the production of flammable vapour from unburned fuel. This in turn reduces the risk of migration of fuel vapour to other areas if gases later escape the compartment and the occurrence of more dangerous rapid fire growth phenomena should oxygen be re-introduced to the compartment. It also reduces heat transfer to other fuels within the compartment and to boundaries adjoining

44 DRDC-RDDC-2015-R224

adjacent compartments, lowering the probability and speed with which the original fire could spread past the room of origin.

Table 12: Summary of Results for Obstructed Diesel Fires. Burn Room Door Closed after the Activated Aerosol Unit was Placed in the Space.

Peak Average Temp Cooling Rate Total Cooling Effect Test (oC) (oC/s)* (m·K) 2c: StatX FR 255 2.3 235 2c: DSPA 5-4 221 2.5 211 Closed Door 220 1.9 164 *Note 1: Cooling rates assessed from onset of suppression to 60 seconds after peak average temperature.

Tests were also run where the aerosol unit was activated and placed in the water below the fuel in the burn pan. This was meant to mimic the unit being activated, thrown into a space and ending up in the bilge below an engine. In the 3a tests, the door of the burn room was closed after the aerosol unit was place in the bilge and in 3b tests the door was left open 30 cm after the aerosol unit was placed in the bilge. The results are shown in Table 13.

Table 13: Summary of Test 3a and 3b Results for Obstructed Diesel Bilge Fires. Total Cooling Test Peak Average Temp [oC] Cooling Rate (oC/s)* Effect [m·K] 3a: StatX FR 310 4.2 337 3a: DSPA 5-4 278 3.2 235 3b: StatX FR 377 5** 431 3b: DSPA 5-4 273 4.2 370 Closed Door 220 1.9 164 *Note 1: Cooling rates assessed from onset of suppression to 60 seconds after peak average temperature. **Note 2: Cooling rate is only for 27 seconds between the peak average and lowest average temperature since the fire was not extinguished in this test.

In the tests with the compartment door closed (3a), the StatX FR had a larger impact on the fire environment than the DSPA 5-4 unit, i.e., the StatX FR unit resulted in a cooling rate of 4.2 °C/s and total cooling of 337 m·K versus 3.2 °C/s cooling rate and 235 m-K total cooling for the DSPA 5-4 unit. When the door was open, however, the Stat-X FR did not suppress the fire, while the DSPA 5-4 performed in a fashion similar to the StatX FR in the test with door was closed, The DSPA 5-4 unit had a cooling rates of 3.2 °C/s and total cooling effect of 370 m·K. The addition of aerosol from either unit resulted in a significantly higher rate of cooling and higher total cooling effect than seen when the door was closed and no aerosol was added.

DRDC-RDDC-2015-R224 45

5 New Fire Attack Strategies and Doctrines

With the introduction of new fire suppression technologies on naval vessels, firefighting strategies and doctrines associated with their use should be evaluated. For instance, if a Halon 1301 system is replaced with a water mist system in a space, the procedures used prior to and after the activation of the suppression system to maximize effectiveness and minimize danger to the crew may vary considerably. The procedure for activation of a Halon 1301 system requires that the crew leave the space, the space be ‘closed’ prior to system activation to ensure the effective agent concentration is achieved, and the re-entry team wait for a time (15 minutes) to allow acid gas (HF) levels to decrease and the space to cool. As Halon 1301 does not cool a space to the degree that water would, the potential for reignition of flammable materials in the space after the effective suppression concentration of Halon 1301 is lost must also be considered in re-entry procedures. Also, the ventilation is not turned on at re-entry by the attack team. US Navy doctrine for activation of and re-entry to a space where a water mist system has replaced a Halon 1301 system are found in [31]. The water mist system can be activated prior to all crew exiting a space, ventilation can be reactivated to improve visibility/habitability for crew exiting the space, ventilation can be reactivated prior to attack team re-entry and re-entry can be made without concern for acid gases produced by the suppressant. As water mist cools the space, and the chance of reignition of flammables in the space is reduced.

The use of other gaseous fire suppressants, such as Novec™ 1230, will also have an effect on fire attack procedures and doctrine. Novec™ 1230 releases HF when exposed to heat and open flame and safe re-entry procedures will have to take this into consideration. As was noted for Halon 1301, Novec™ 1230 does not cool a space to the same degree as water mist, and reignition of flammable materials after its effective concentration is lost is a concern for firefighters.

Aerosols agents (propelled extinguishing agents) may be used in fixed firefighting systems in shipboard spaces. Their use may require that all crew exit a space prior to activation of the system, that an attack team wait before re-entering a space, result in reignition of flammable materials due to hot surfaces in the space, and affect the time before ventilation can be reactivated to improve visibility/habitability in the space for firefighters re-entering the space. The US Navy procedure for the use of aerosol fire suppression systems in diesel engine rooms and flammable storage rooms is given in Appendix D in [32].

On ships of the RNLN equipped with water mist installations the attack strategy is as follows. Regardless of an initiated water mist system, the crew will investigate any fire or smoke alarm using portable extinguishers. Should the investigators not be able to control the fire using the portable extinguishers, a second attack will be organized using breathing apparatuses (BA) and the one hose technique. A third attack is simultaneously organized where crew are equipped with BA and protective clothing. During these attacks the water mist installation can be running, this does not inhibit the crew attacks. Should the space be too hot to enter even in protective clothing, the water mist installation is kept running and crew actions focus on boundary cooling. The water mist will aid in cooling the compartment, hence boundary cooling has low priority.

46 DRDC-RDDC-2015-R224

6 Firefighting on Board Submarines

6.1 Extinguishing Systems on Board Submarines

A literature study was performed and the findings were compiled in a report [33] that provides a brief review of optional water based fire protection systems for the protection of machinery spaces on submarines. Relevant research as well as requirements and recommendations by IMO and/or NFPA were summarised. The compilation is not necessarily complete; however, the most familiar alternatives are discussed.

It should be emphasised that the selection of a fire protection system is dependent on a number of parameters, including (but not limited to):  the expected fire scenarios, e.g., type of fuels, type of fire, fire size, etc.;  the fire protection objectives, in terms of, for example, , fire suppression or fire extinguishment associated with a certain time limit;  the power demand;  considerations regarding system reliability and availability;  considerations regarding safety for personnel;  considerations regarding potential damage to equipment;  considerations regarding the consequence of unintentional system activation;  environmental considerations;  available space limitations;  maximum weight limitations;  system approval or certification requirement; and  cost.

Summaries of the benefits and disadvantages of the different alternatives discussed in the report are given below. No specific alternative or system solution is recommended, the ultimate solution must be made on a case by case basis.

6.1.1 Water Spray Systems

The benefits of water spray systems include sufficient fire control, fire suppression or fire extinguishment, given that the system is properly designed for the fire hazard. A system designed with a sufficiently high discharge density may suppress high flashpoint fuel fires in the order of a few minutes. System design, including the discharge densities and installation limitations of nozzles, are provided in internationally recognized standards. The drawbacks of water spray systems include relatively high discharge densities and limited protection against fire reignition.

DRDC-RDDC-2015-R224 47

6.1.2 Foam and Foam-Water Spray Systems

The benefits include fast fire suppression and fire extinguishment of flammable liquid pool fires and protection against fire re ignition. System design, including the discharge densities and foam solution supply times, are provided in internationally recognized recommendations. Nozzles and foam agents are tested and approved to internationally recognized standards. The drawbacks include environmental concerns associated with the specific foam agents and the clean up after a fire.

6.1.3 Compressed Air Foam Systems (CAFS)

The benefits of compressed air foam systems include fast fire suppression and fire extinguishment and the protection against fire reignition. These systems require up to four times less water and up to six times less foam concentrate than conventional foam-water spray systems for flammable liquid pool fire hazards. This reduces the required amount of water, foam concentrate and the environmental impact. System design, including the discharge densities and foam solution supply times, are provided in internationally recognized standards. The drawbacks include environmental concerns associated with the specific foam agents and the clean up after a fire.

6.1.4 Conventional Water Mist (or Fine Water Spray) Systems

The benefits of conventional water mist systems lays in their superior cooling ability, their (supposed) system simplicity and the obvious life safety and environmental benefits of water as a suppressant. Installation guidelines are given in internationally recognized standards. The drawbacks include long or unlimited fire extinguishment times for certain fire scenarios. Small fires relative to the volume of the protected compartment may not be extinguished or extinguishment times may be very long. The summary presented here does also reveal the drawbacks of the current IMO fire test procedures and their applicability to submarine fire hazards and geometries. Systems and systems designs certified to requirements of IMO are therefore not necessarily directly applicable.

6.1.5 Water Mist Systems Combining Water and Inert Gas

The benefits of systems combining water and inert gas include significantly lower water flow rates compared to conventional water mist systems. The use of inert gas will also improve the likelihood of extinguishment of small fires and shorten fire extinguishment times. There are several commercially available system technologies. A drawback is that the use of a gas will require over pressure relief venting from the protected compartment to an adjacent compartment. This may be difficult to achieve or undesirable on a submarine.

6.1.6 NanoMist®

NanoMist® appears to be a promising fire suppression system and explosion mitigation system for the protection of machinery spaces and electrical spaces on board submarines. The drawback may include the space filling time and the associated fire extinguishment time. There is only one manufacturer of the system technology and the commercial availability of the system is uncertain.

48 DRDC-RDDC-2015-R224

6.1.7 High Expansion Foam Systems (Inside Air Systems)

High expansion foam systems using foam generated by the air inside the protected compartment may be a viable alternative. The benefits include low water demands. Foam generators and foam agents are tested and approved to internationally recognized standards. The drawbacks include environmental concerns associated with the specific foam agents and the clean up after a fire.

6.2 Electrical Cabinet Fire Suppression Systems

The goal of this series of tests was to evaluate three gaseous fire suppressants / extinguishment systems with regards to their extinguishment efficiency and protection against reignition, the toxicity and corrosivity of their degradation products, and their cooling capacity [34]. The extinguishment systems evaluated were a fixed Novec™ 1230 system, a fixed nitrogen system and a portable carbon dioxide (CO2) extinguishing system. The CO2 was introduced to the electrical cabinet through a NATO coupling.

The Novec™ 1230 system was designed to deliver enough fire suppressant to reach a concentration of 5.9% in under 5 seconds. The nitrogen system was operated at 12.5 bar and the gas supplied to the space for 60 seconds. Full release of the 2 kg hand-held CO2 extinguisher took between 50 and 60 seconds.

The tests were performed in a cabinet with the dimensions 1.6 m x 0.4 m x 2.0 m (L x W x H). The cabinet was positioned in an air tight room where any gases escaping from the cabinet were collected and subsequently analysed. Two fire scenarios where evaluated:  A 5 kW heptane fire to represent a fast growing fire. The heptane pool was placed on a steel stand 1 m above floor level inside the cabinet.  An external radiation panel applying a thermal flux (~ 40 kW/m2) to a fuel source that contained an electrical cable bundle. This represented a short circuit fire where the fuel (cable sheathing) is heated until it ignites and the electrical short continues to provide energy to the fire.

Temperatures in the electrical cabinet and pressure in both the cabinet and the surrounding compartment were monitored. The extinguishment system was activated 60 seconds after ignition. The time to extinguishment and time to reignition, when it occurred, are shown in Table 14.

Steel, copper and zinc coupons were exposed to the environment in the cabinet to measure how corrosive the environment was during the tests. In the tests where Novec™ 1230 was used, HF concentrations in both the cabinet and the compartment were measured by passing gases from the spaces through an aqueous sodium hydroxide solution where any fluorides were collected and HF concentrations calculated. The results are shown in Table 15.

The effect of increasing the humidity in the compartment on HF-levels was evaluated in two tests. In another test, the flow of Novec™ 1230 into the electrical cabinet was reduced using a throttle washer. The objective of this test was to measure HF levels produced in a fire scenario where the system malfunctions.

DRDC-RDDC-2015-R224 49

Table 14: Extinguishment and Reignition Times for Electrical Cabinet Tests. Time from ext. Time from activation to ext. Scenario Ext. system Comment to reignition [min:s] [min:s]

Heptane Novec™ 1230 0:05 N/A Cable Novec™ 1230 High humidity (56%) 0:03 3:14

Heptane CO2 0:08 N/A

Cable CO2 0:09 7:27 Heptane Novec™ 1230 Flooding failure 0:55 N/A Cable Novec™ 1230 High humidity (80%) 0:02 - Cable Novec™ 1230 0:03 1:27

Heptane N2 0:59 N/A

Cable N2 0:10 - Cable Novec™ 1230 No vent. opening 0:02 13:58

Table 15: Measured HF Concentrations for Electrical Cabinet Fires Tests.

Scenario Sampling time (minutes: HF conc. cabinet [ppm] HF conc. room [ppm] seconds) 4:30 – 9:30 1030 33 Heptane fire 10:48 – 15:48 500 14 35% RH 17:30 – 22:30 470 12 12:30 – 17:30 160 22 Cable fire 23:00 – 28:00 1740 34 56% RH 38:00 – 43:00 370 20 Heptane fire 4:00 – 9:00 14300 67 Flooding failure 11:30 – 16:30 8550 147 37% RH 26:30 – 31:30 3540 61 13:44 – 18:44 260 23 Cable fire 23:44 – 28:44 110 30 80% RH 38:44 – 43:44 120 25 10:18 – 15:18 380 14 Cable fire 20:18 – 25:18 3400 74 39% RH 36:00 – 41:00 790 53

The CO2 system provided the most efficient cooling, resulting in average cabinet temperatures close to 0°C. Sub-zero temperatures occurred for short periods of time in areas close to the discharge nozzle. The results for the Novec™ 1230 and the N2 systems indicated that they cooled the cabinet effectively. However, the shorter discharge time of the Novec™ 1230 system resulted in a significantly faster cooling.

50 DRDC-RDDC-2015-R224

There was no obvious difference in corrosion damage to steel or copper for the three gaseous agents. The corrosion of steel and copper seemed to be more sensitive to the humidity in the compartment. The corrosion damage to zinc seemed to be more severe for tests with Novec™ 1230 than for CO2 and N2.

The over pressure in the cabinets resulting from discharge of the extinguishment systems was no higher than that caused by the fire during the pre-burn time. During and after discharge, the pressure in the cabinet became negative.

DRDC-RDDC-2015-R224 51

7 Conclusions

At their best, prescriptive regulations can provide good quality fire safety design. At their worst, prescriptive regulations can result in expensive and poorly adapted systems for managing fire threats on naval vessels. Designers, researchers and regulatory bodies must challenge the prescriptive requirements and try to find better, safer solutions when this is financially and technically feasible.

In the FiST project we investigated performance criteria for cases where no regulations were prescribed, for example on damaged water mist systems. There have been suggestions for technical innovations, such as the use of bulkhead nozzles (rather than ceiling nozzles), more robust layout and local protection, residual capacity in the pump and the ability to connect water mist systems to the fire main. These innovations might have been inhibited by prescriptive regulations.

Furthermore, testing of prescriptive regulations for water spray systems for ammunition storage spaces indicated that these can lead to poorly designed systems. Poor design not only affects the system, but also influences ship construction. In the past designers have had to cope with the free surface effects of enormous water flows into magazine spaces during discharge of the system.

Canada was able to verify specific design options with respect to the use of Novec™ 1230. Knowledge on combining Novec™ 1230 with water mist may be implemented in future ship designs or mid-life upgrades. In Sweden the knowledge has been applied in design of firefighting systems for submarine engine compartments. In the Netherlands the work from FiST will be input for selection and layout of water mist systems on new shipbuilding programs with high degree of automation. We are currently investigating opportunities to develop practical methods, rules of thumb and tools for ship designers to employ performance based approaches to fire suppression themselves.

52 DRDC-RDDC-2015-R224

8 FiST Reports, Memos and Conference Proceedings

8.1 Reports and Memos

• Hiltz, J.A., Additives for water mist fire suppression systems: A review, DRDC Atlantic TM 2012-236, Defence Research and Development Canada, November 2012.

• Sheehan, T., Topic, A., Weckman, E., Strong, A. and Hitchman, G., Maritime evaluation of aerosol fire knock down tools, University of Waterloo, Waterloo, Ontario, Canada, December 2012.

• Weckman, E., Topic, A., Sheehan, T. and Hitchman, G., Maritime evaluation of aerosol fire knock down tools, Part 2: toxicity and corrosion potential, DRDC-RDDC-2014-C32, February 2014.

• Rahm, M. and Lindstrom, J., Fire protection of weapon storage and water mist redundancy philosophies, SP Technical Research Institute of Sweden, DRDC Atlantic CR 2012-193, September 2012.

• Rahm, M., Claesson, A., Försth, M. and Ochoterena, R., Tests of fire suppression effectiveness of damaged water mist systems, DRDC-RDDC-2014-C70, May 2014.

• van der Wal, R., High Pressure versus Low Pressure, FiST memo, May 2, 2012.

• Sanders, B. and van der Wal, R., Compressed Air Foam for extinguishing fires on board of naval vessels, TNO 2014 R11866, December 2014.

• Rahm, M. and Leandersson, A., Electrical cabinet fire extinguishment testing, DRDC-RDDC-2014-C181, December 2014.

• Arvidson, M., Optional fire protection systems for machinery spaces on submarines, SP Report P900038, February 2013.

8.2 Conference Presentations and Proceedings

• van der Wal, R., RESIST simulations on damaged systems, April 12, 2011, FiST meeting Grindsjön.

• Lindström, J., Rahm, M., Hiltz, J., Boonacker, B., van der Wal, R. and Haara, M., Fire protection in ammunition storage spaces on board naval craft: An evaluation of the water application rate, 2012 International Water Mist Association Conference, Barcelona, Spain (DRDC Atlantic SL 2012-184).

• Boonacker, B., van der Wal, R., Lindström, J., Rahm, M. and Hiltz, J., Fire protection in ammunition storage spaces: An evaluation of the water application rate, International Water Mist Association 2012 Conference in Barcelona, Spain. Presentation.

DRDC-RDDC-2015-R224 53

• van der Wal, R., Rahm, M., Claesson, A., Hiltz, J. and Boonacker, B., Residual capacity of a damaged water mist system, 2013 MAST Europe Conference, June 2013, Gdansk, Poland. DRDC Atlantic SL 2013-110.

• van der Wal, R., Rahm, M., Claesson, A., Hiltz, J. and Boonacker, B., Residual capacity of a damaged water mist system, MAST Conference June 5, 2013, Gdansk, Poland - Presentation.

• Hiltz, J.A., Residual capacity of a damaged high pressure water mist system, 2013 Society of Fire Prevention Engineers Conference and Exposition, Austin Texas, October 28–29, 2013. DRDC Atlantic SL 2013-192.

• Rahm, M., Hiltz, J., van der Wal, R., Hertzberg, T. and Lindström, J., Testing of naval fire suppression systems to support a performance-based approach, In Proceedings of IMarEST INEC 2014 Conference, Amsterdam, The Netherlands, May 2014 (DRDC-RDDC-2014-P20).

• van der Wal, R., Rahm, M., Hiltz, J., Hertzberg, T. and Lindström, J., Testing of naval fire suppression systems to support a performance-based approach, INEC 2014 Presentation.

• van der Wal, R., Naval Aspects of Fire Safety, Naval Damage Control Conference, October 7, 2014, Presentation.

8.3 Journal Paper

• van der Wal, R., Hiltz, J. and Hertzberg, T., The benefit of performance based fire safety engineering, Naval Forces, VI/20I4.

54 DRDC-RDDC-2015-R224

This page intentionally left blank.

DRDC-RDDC-2015-R224 55

References

[1] Spindel, R., Laska, S., Cannon-Bowers, J., Cooper, D., Hegmann, K., Hogan, R., Hubbard, J., Johnson, J., Katz, D., Kohn, E., Roberts, K., Sheridan, T., Skalka, A. and Smith, J. Optimized Surface Ship Manning, (NRAC 00-1) Naval Research Advisory Committee, Arlington, Virginia, USA, 2000.

[2] Hertzberg, T., et al. Water mist: Theory, physics, simulation, SP Technical Research Institute of Sweden, SP Report 2004:15 [1].

[3] van der Wal, R. and Posthuma, A. Threat Description; 032.32507/01.02-1, Threat description V1.0.doc, 22 February 2011.

[4] Rahm, M. and Lindström, J. Fire protection of weapon storage and water mist redundancy philosophies, SP Technical Research Institute of Sweden, 2012.

[5] Rahm, M. and Lindstrom, J. Fire protection of weapon storage and water mist redundancy philosophies, SP Boras, Boras, Sweden, DRDC Atlantic CR 2012-193, September 2012.

[6] Rahm, M., Claesson, A., Försth, M. and Ochoterena, R. Tests of fire suppression effectiveness of damaged water mist systems, SP Boras, Boras, Sweden, DRDC-RDDC-2014-C70, May 2014.

[7] IWMA FAQ (http://iwma.net/about-us/faqs/#c57), (access date: 25 August 2015).

[8] Pesaola, M. and Impianti, E. Water Mist Systems Designed For Earthquake Conditions; International Water Mist Conference, Hamburg, October 12–13, 2011.

[9] Rhijnsburger, M. and van der Wal, R. Water mist can almost halve the impact of an explosion; Interview in TNO TIME magazine, May 2009.

[10] Jäppinen, T. Finnish Naval Research Institute; Private email correspondence, January 2012.

[11] Arvidson, M. Fire protection of special spaces, SP Technical Research Institute of Sweden, Borås, Sweden, 2013.

[12] NATO STANDARD ANEP77; Naval Ship Code: Edition E Version 1; Allied Naval Engineering Publication, January 2014.

[13] Evans, D.D. Sprinkler Fire Suppression Algorithm for HAZARD; NISTIR 5254, August 1993.

[14] Madrzykowski, D. and Vettori, R.L. A Sprinkler Fire Suppression Algorithm for the GSA Engineering Fire Assessment System; NISTIR 4833, 1992.

[15] Walton, W. D. Suppression of Wood Crib Fires with Sprinkler Sprays: Test Results; NISTIR 88-3696, 1988.

56 DRDC-RDDC-2015-R224

[16] Madrzykowski, D. and Stroup, D. W. Demonstration of Biodegradable, Environmentally Safe, Non-Toxic Fire Suppression Liquids; NISTIR 6191, 1998.

[17] Kim, A. and Crampton, G. Fire Protection of Submarine’s Machinery Space and Electronic Equipment Cabinets, National Research Council of Canada, Institute for Research in Construction Report B-4715.1, February 21, 2011.

[18] Hiltz, J. A. Additives for water mist fire suppression systems - A review, Defence Research and Development Canada, DRDC Atlantic TM 2012-236, January 2013.

[19] Airpress Compressoren DC 12-170, airpress website www.airpress.nl, (access date: 26 April 2012).

[20] www.keesiedive.nl/cilinders.htm, Cutting Extinguishing Concept – practical and operational use, MSB, SERF, SP, Sweden, (access date: 24 May 2012).

[21] Cutting Extinguishing Concept: practical and operational use, MSB, SERF, SP, Sweden.

[22] Svensk Standard 5S 1192, Version 6, SMS Reg 611.31, 15 November 1985.

[23] Stensaas, J. P. and Jacobsen, H. C. Testing of different portable fire extinguishers against fires in twin tyres, Sintef Report NBL10 A01159, 2001.

[24] van der Wal, R. and Smit, C. S. Active and passive fire retarding measures, TNO report DV2 2005-A43.

[25] Det Norske Veritas AS, Statutory Interpretations, February 2013.

[26] British Steamship P&I; Recommendation for the Safety of Cargo vessel of less than Convention Size Part IV / IV; BSM-RM-10/009/13, 29 November 2013.

[27] Calpeda pumps website, http://uk.calpeda.com/scheda_prodotto.php?id=69, (access date: 7 June 2012).

[28] Sheehan, T. D. Royal Canadian Navy Evaluation of Handheld Aerosol Extinguishers, Master of Applied Science in Mechanical Engineering Thesis, University of Waterloo, Waterloo, Ontario, Canada, 2013.

[29] Sheehan, T., Topic, A., Weckman, E., Strong, A. and Hitchman, G. Maritime Evaluation of Aerosol Fire Knock Down Tools, Fire Research Group, Department of Mechanical and Mechatronics Engineering, University of Waterloo, Waterloo, Ontario, 2012.

[30] Weckman, E., Topic, A., Sheehan, T. and Hitchman, G. Maritime Evaluation of Aerosol Fire Knock Down Tools Part 2: Toxicity and Corrosion Potential, DRDC-RDDC-2014-C32, 2014.

[31] Farley, J.P. and Williams, F.W. Water Mist Machinery Space Fire Doctrine, NRL/MR/6180--04-8740, Naval Research Laboratory, Washington, DC, 2004.

DRDC-RDDC-2015-R224 57

[32] Forssell, E. W., Scheffey, J. L., Dinaburg, B. J., Farley, J. P. and Whitehurst, C. L. Evaluation of Protection Options for Diesel Engine Rooms and Flammable Liquid Storage Rooms Onboard the U. S. Navy Landing Craft Utility (LCU), NRL Ltr 3900/Ser6180/0012, Naval Research Laboratory, Washington, DC, 11 January 2011.

[33] Arvidson, M. Optional fire protection systems for machinery spaces on submarines, SP Report P900038, February 2013.

[34] Rahm, M. and Leandersson, A., Electrical cabinet fire extinguishment testing, DRDC-RDDC-2014-C181, December 2014.

58 DRDC-RDDC-2015-R224

DOCUMENT CONTROL DATA (Security markings for the title, abstract and indexing annotation must be entered when the document is Classified or Designated) 1. ORIGINATOR (The name and address of the organization preparing the document. 2a. SECURITY MARKING Organizations for whom the document was prepared, e.g., Centre sponsoring a (Overall security marking of the document including contractor's report, or tasking agency, are entered in Section 8.) special supplemental markings if applicable.)

DRDC – Atlantic Research Centre UNCLASSIFIED Defence Research and Development Canada 9 Grove Street P.O. Box 1012 2b. CONTROLLED GOODS Dartmouth, Nova Scotia B2Y 3Z7 (NON-CONTROLLED GOODS) Canada DMC A REVIEW: GCEC DECEMBER 2012

3. TITLE (The complete document title as indicated on the title page. Its classification should be indicated by the appropriate abbreviation (S, C or U) in parentheses after the title.)

New Technologies for Fire Suppression on Board Naval Craft (FiST): Final Report

4. AUTHORS (last name, followed by initials – ranks, titles, etc., not to be used)

Hertzberg, T.; Hiltz, J.A.; van der Wal, R.; Rahm, M.

5. DATE OF PUBLICATION 6a. NO. OF PAGES 6b. NO. OF REFS (Month and year of publication of document.) (Total containing information, (Total cited in document.) including Annexes, Appendices, etc.) September 2015 72 34

7. DESCRIPTIVE NOTES (The category of the document, e.g., technical report, technical note or memorandum. If appropriate, enter the type of report, e.g., interim, progress, summary, annual or final. Give the inclusive dates when a specific reporting period is covered.)

Scientific Report

8. SPONSORING ACTIVITY (The name of the department project office or laboratory sponsoring the research and development – include address.)

DRDC – Atlantic Research Centre Defence Research and Development Canada 9 Grove Street P.O. Box 1012 Dartmouth, Nova Scotia B2Y 3Z7 Canada

9a. PROJECT OR GRANT NO. (If appropriate, the applicable research 9b. CONTRACT NO. (If appropriate, the applicable number under and development project or grant number under which the document which the document was written.) was written. Please specify whether project or grant.)

01ea

10a. ORIGINATOR’S DOCUMENT NUMBER (The official document 10b. OTHER DOCUMENT NO(s). (Any other numbers which may be number by which the document is identified by the originating assigned this document either by the originator or by the sponsor.) activity. This number must be unique to this document.)

DRDC-RDDC-2015-R224

11. DOCUMENT AVAILABILITY (Any limitations on further dissemination of the document, other than those imposed by security classification.)

Unlimited

12. DOCUMENT ANNOUNCEMENT (Any limitation to the bibliographic announcement of this document. This will normally correspond to the Document Availability (11). However, where further distribution (beyond the audience specified in (11) is possible, a wider announcement audience may be selected.))

Unlimited

13. ABSTRACT (A brief and factual summary of the document. It may also appear elsewhere in the body of the document itself. It is highly desirable that the abstract of classified documents be unclassified. Each paragraph of the abstract shall begin with an indication of the security classification of the information in the paragraph (unless the document itself is unclassified) represented as (S), (C), (R), or (U). It is not necessary to include here abstracts in both official languages unless the text is bilingual.)

This report summarizes the results of a Project Arrangement entitled “New Fire Suppression Technologies on Board Naval Ships (FiST)” carried out under the Canada/Netherlands/Sweden Memorandum of Understanding on Cooperative Science and Technology. The FiST project had three main areas of focus; 1) fixed fire suppression systems, 2) portable, manually operated fire suppression systems, and 3) firefighting on submarines. For fixed fire suppression systems, large scale fire suppression testing was carried out to determine the effectiveness of low pressure water mist systems under well ventilated conditions, the effectiveness of high pressure water mist systems in a damaged condition, and the effectiveness of using water mist in conjunction with deluge systems for the protection of ammunition storage spaces. For the damaged high pressure system, damage was simulated by reducing the number of operational nozzles, by reducing system pressure and by introducing damaged water delivery pipe segments. The results large scale fire suppression testing of a dual agent (water mist / Novec™ 1230) system are reported. The use of water mist in conjunction with Novec™ 1230 was found to reduce the levels of acid gas in the test space significantly. The effectiveness of gaseous fire suppression agents, including Novec™ 1230 (a perfluorinated ketone), carbon dioxide and nitrogen, in suppressing or extinguishing electrical cabinet fires was investigated. This study included analysis of acid gases produced by Novec™ 1230. For portable, manually operated systems, testing focused on the evaluation of the effectiveness, toxicity and corrosiveness of hand-held aerosol fire suppression agents. Other technologies, including compressed air foam systems, the use of additives in water mist systems and cool gas generator technology for use in fire suppression systems, are also reviewed and discussed. ------

Le présent rapport résume les résultats d’une entente de projet intitulée « Nouvelles technologies d’extinction d’incendie à bord des navires militaires (FiST) » s’inscrivant dans le protocole d’entente Canada/Pays Bas/Suède sur la recherche coopérative en matière de science et technologie. Le projet FiST comporte trois domaines d’intérêt particulier : 1) systèmes fixes d’extinction d’incendie, 2) systèmes d’extinction d’incendie manuels portatifs et 3) systèmes de lutte contre l’incendie à bord des sous-marins. Pour ce qui est des systèmes fixes d’extinction d’incendie, des essais à grande échelle ont été réalisés afin de déterminer l’efficacité des systèmes à brouillard d’eau à basse pression dans un endroit bien aéré, celle des systèmes à brouillard d’eau à haute pression en mauvais état et celle de l’utilisation d’un brouillard d’eau combiné à des systèmes de type déluge pour protéger les aires de stockage des munitions. Dans le cas d’un système à haute pression en mauvais état, on a simulé les dommages en réduisant le nombre de lances d’incendie opérationnelles, en réduisant la pression du système et en ajoutant des segments de conduites d’eau endommagées. On présente les résultats des essais d’extinction d’incendie à grande échelle sur un système utilisant deux agents (brouillard d’eau / NovecMC 1230). L’utilisation d’un brouillard d’eau conjuguée à du NovecMC 1230 permet de réduire considérablement le niveau des gaz acides dans l’espace d’essai. L’efficacité des agents d’extinction d’incendie gazeux, incluant le NovecMC 1230 (cétone perfluorée), le dioxyde de carbone et l’azote, pour la suppression ou l’extinction d’incendie dans une armoire électrique, a été étudiée. L’étude comportait une analyse des gaz acides produits par le NovecMC 1230. Dans le cas des systèmes manuels portatifs, les essais ont porté essentiellement sur l’évaluation de l’efficacité, de la toxicité et de la corrosivité des agents d’extinction d’incendie en aérosol. D’autres technologies, notamment les systèmes à mousse à air comprimé, les additifs utilisés dans les systèmes à brouillard d’eau et les générateurs de gaz à froid employés dans les systèmes d’extinction d’incendie, sont également examinées.

14. KEYWORDS, DESCRIPTORS or IDENTIFIERS (Technically meaningful terms or short phrases that characterize a document and could be helpful in cataloguing the document. They should be selected so that no security classification is required. Identifiers, such as equipment model designation, trade name, military project code name, geographic location may also be included. If possible keywords should be selected from a published thesaurus, e.g., Thesaurus of Engineering and Scientific Terms (TEST) and that thesaurus identified. If it is not possible to select indexing terms which are Unclassified, the classification of each should be indicated as with the title.)

fire suppression; water mist; water mist additives; aerosols; alternative gaseous agents; compressed air foams; prescriptive rules; performance based