ANSUL_Fire Protection Solutions for LNG

FIRE PR FIRE PRO OTE TECT CTIO ION N SO SOL LUT UTIO IONS NS FOR LIQUEFIED NATURAL GAS

TABL ABLE E OF CON CONTEN TENTS TS

S __e_c_ti_o_n

P _a _g _e _

Introduction

1

The Natural Gas Fire Problem

2

The Candidate Fire Extinguishing Agents

5

The AN ANSUL Natural Ga Gas Fi Fire Ex Extinguishment Co Concept

6

The Experimental Experience

8

The General Behavior of Extinguishing Agents

9

The Specific Agent Flow Rate Requirements For Natural Gas Fires

11

The Th e AN ANSU SUL L Re Recom comme mende nded d Ag Agent ent Qu Quant antit ity y Requ Requir irem ement ents s

12

Bibliography

26

INTRODUCTION Page 1

INTRODUCTION

The liquefaction of natural gas, which reduces its volume by a factor of over 600, has made the storage and transportation of this fuel economically attractive. However, this liquefaction technique has also served to increase the amount of energy in storage, process and transportation equipment by the same amount. This tremendous concentration of energy has not been overlooked by the gas utilities, utilities, nor gone unnoticed by the authorities authorities and the general public. The safety of natural gas, especially from the fire protection standpoint, has been the subject of considerable research in recent years, and many techniques have been refined in the overall fire protection approach to the hazard. As with any other potential hazard, the fire protection for a natural gas facility consists of three elements: fire prevention, fire control, and fire extinguishment. Figure 1 illustrates these elements as they relate to LNG (Liquefied Natural Gas) processes.

The considerations for fire prevention are well documented in the National Fire Protection Association’s Standard on “Storage and Handling of Liquefied Natural Gas (LNG),” NFPA NFPA 59A1. In additi addition, on, the techniques for fire control, especially for exposure protection, are not that different with natural gas than with many other flammable materials. There is a great amount of historical experience in this area. area. The primary primary element element to which which this publica publication tion addresses itself is the extinguishment of fires involving natural gas, in the liquefied, vapor and gaseous states. A brief description description of vapor dispersion, which can minimize downwind drift of vapor clouds, and radiation intensity is also made 10. NFPA 59A recomm NFPA recommends ends that “normally “normally gas gas fires (including (including LNG) should not be extinguished until the fuel source can be shut off.” However, a gas fire which places personnel in severe danger, a gas shutoff valve which is involved in the fire, or a fire which indirectly endangers personnel through thermal failure of equipment in the fire area, may necessitate immediate extinguishment. Therefore, this publication assumes that there are a number of situations where the extinguishment of natural gas fires is not only appropriate, but desirable.

Fire Protection

Fire Prevention

Fire Control

Process and Site Design

Exposure Protection

Fire Extinguishment

Construction Material Operation Criteria Vapor Dispersion

Provisions of NFPA Standard 59A Industry Standards

Water

Dry Chemicals

High Expansion Foam

High Expansion Foam Dry Chemicals

FIGURE 1 OVERALL FIRE PROTECTION APPROACH 003380

THE NA NATU TURAL RAL GAS FIR FIRE E PRO PROBLE BLEM M Page 2

THE NATURAL NATURAL GAS FIRE PROBLEM

In the past, the natural gas fire problem was rather simple when compared to today’s situation. At that time, nearly all our natural gas was processed, transported, stored and distributed in the vapor state. With the widespread application of cryogenic techniques in recent years, the processing, transportation, storage and vaporization of liquefied natural gas has added a new dimension to the problem. problem. Instead of being concerned concerned about about the fire extinextinguishing requirements for only the vapor state, design criteria became necessary for both the vapor and liquid states. Figure 2 illustrates some of the physical and chemical properties of natural gas. The properties are approximated since the composition of natural gas covers a rather broad range. Composition ___________ Methane Ethane Propane Butane

83–99% 1–13% 0.1 –3% 0.2–1.0%

Physical Properties _________________ Normal Boiling Point

Density liquid at NBP (Normal Boiling Point) Dens De nsit ity y vap vapor or at NB NBP P (c (com ompa pare red d with air at 70 °F (21.2 °C)) Liquid to vapor expansion Heat of vaporization Theoretical vaporizing capability of 1 cu. ft. (0.3 m2) of: Dry earth Wet earth Water

Air

Combustion Properties ____________________ Flammable range

Heat of combustion Burn rate, steady state pool Pool fire flame height

–255 to –263 °F (–160 to –164 °C) 3 1/2 to 4 lb/gal (0.42-0.48 kg/L) 1.47 1. 47 600 to 1 220-248 Btu/lb (512-577 kj/kg)

6 gal (22.7 L) LNG (Liquefied Natural Gas) 20 gal (75.71 L) LNG 24 gal (75.708 L) LNG (1 gal water = 3.2 gal LNG) 0.0005 gal (0.6019 L) LNG

5-14% (methane at normal temperatures) 6-13% (methane near minus 260 °F) 22,000 Btu/lb (51,172,000 J/kg) 0.2-0.6 in./minute 3 times base dimensions of pool (slight wind)

FIGURE 2 Approximate Properties of Natural Gas2 003381

THE NA NATUR TURAL AL GAS FIR FIRE E PRO PROBLE BLEM M Page 3

THE NA NATURA TURAL L GAS FIR FIRE E PRO PROBLEM BLEM (Co (Conti ntinued nued))

After analysis of the characteristics of a natural gas fire, ANSUL has concluded that the problem may be simplified to the extent shown in Figure 3. This figure essentially illustrates the following:

Preburn: The length of time that a fire has burned in an impingPreburn: The ing jet situation will proportionately increase the extinguishing agent application rate that is required.

A. Sta State: The natural gas at the source of the fire problem will be in either the vapor vapor or the liquid liquid state. state.

Obstructions: The presence of obstructions in the fire area will Obstructions: The influence the number of extinguishing agent application points required to insure adequate agent coverage.

B. Conf Configu igurat ration ion:: A natural gas release release may be rapid, producing producing a pressur pressurized ized flow. flow. If the release occurs outdoors, outdoors, the problem is simplified. If, however, it occurs in a contained volume, flammable concentrations may produce potentially explosive conditions. Liquefied natural gas leaks may take the form of a pressurized flow and, if the leakage rate is adequate, the problem may be further complicated by the formation of a liquid pool. C. Var aria iabl bles es:: In the case of pressure fires in both the vapor and liquid states, there are three very important variables that will directly influence the ease or difficulty of extinguishment:

Impingement: If the natural gas jet is impinging on a vertical surface (process equipment) or a horizontal surface (ground), a fire will be significantly significantly more difficult difficult to extinguish than if the jet is not impinging impinging on a surface. surface.

Within a contained volume, an important variable to be considered is that other flammables (refrigerants, etc.) may be present. These other flammables could behave quite differently than natural gas with regard to flammable and explosive limits. The behavior of LNG (Liquefied Natural Gas) in a spill situation is an important consideration in determining extinguishing agent application requirements. The characteristics of the surface on which a spill occurs will influence the initial rate of vaporization. vaporiz ation. However, However, an approximation approximation of the initial rate of vaporization on both solid surfaces and water can be said to be in the range of 50 ft3 per minute of vapor per ft2 (1 (15. 5.24 24 m3 2 per minute per m ) of LNG sur surface face area area..

Natural Gas

State

Configuration

Variables

Liquid

Vapor

Pressure

Impingement Preburn Obstructions

Contained

Pressure/Pool

Spill

Other Flammables

Impingement Preburn

Vaporization Rate

Obstructions

Obstructions

FIGURE 3 Definition of the Natural Gas Fire 003382

THE NA NATU TURAL RAL GAS FIR FIRE E PRO PROBLE BLEM M Page 4

THE NA NATUR TURAL AL GAS FIRE PRO PROBLEM BLEM (Co (Contin ntinued ued))

The steady-state vaporization rate, in contrast, is approximately 1 ft3 per minute of vapor per ft2 of LNG surface area (0.3048 m3 per minute per m2). This rate is equivalent to a 1 ft (0.3 m) deep pool evaporating in 10 hours, assuming that steady-state had already been reached. While a fire situation will produce a higher rate of vaporization at steady-s steady-state, tate, a fire of greater intensity will occur in an initial spill situation. These factors are taken into account in the design criteria (See Figure 12). With this definition of the characteristics of a natural gas fire, it was then possible to review candidate agents to determine their compatibility with the problem.

THE TH E CAN CANDID DIDA ATE FIR FIRE E EXT EXTING INGUIS UISHIN HING G AGE AGENT NTS S Page 5

THE CANDIDATE FIRE EXTINGUISHING AGENTS

Historically, the only extinguishing agents accepted as effective on natural gas vapor fires were dry chemicals and carbon dioxide. Furthermore, Furtherm ore, due to the dry chemicals’ tremendo tremendous us effectiveness effectiveness advantages over carbon dioxide, the latter is usually employed only in areas where the dry chemicals may damage sensitive equipment or where a total flooding technique can be employed. Such agents as water, protein foam, aqueous film forming foams (AFFF) and other water base agents have been found to have little or no effectiveness in the extinguishment of vapor fires, or for that matter, pressure fires in general. Hence, most fire extinguishment experimentation and actual fire extinguishing experience in the natural gas vapor fire field have been restricted to the dry chemical agents. With the advent of LNG (Liquefied Natural Gas), most of the water base agents were immediately ruled out since they were not only ineffective, but their application on an LNG spill could worsen the situatio situ ation. n. NFPA NFPA 113 (“Standard for Low-, Medium-, and HighExpansion Foams”) cautions against the use of foam or AFFF on refrigerated or cryogenic fluids due to severe boiling and increased vapor release that would follow. One noteworthy exception to the use of water base agents on LNG is high expansion foam, which has an extremely low water content. High expansion foam experimentation on LNG fires has demonstrated that this agent does have vapor dispersion and fire control capabilities. Use of high expansion foam is discussed later in this document. At the moment, the only known agents that have demonstrated an ability to completely extinguish LNG fires are the dry chemicals. In this agent category, three types presently account for 95% of the applications in the United States: A. Sodiu Sodium m Bicarb Bicarbon onate ate Base Base (ANSUL (ANSUL PLU PLUS-F S-FIF IFTY) TY):: This agent, which is the dry chemical first developed, has been largely replaced by the more effective potassium bicarbonate base material in the oil and gas industry. B. Monoamm Monoammoni onium um Phosph Phosphate ate Base Base (ANSUL (ANSUL FOR FORA AY): This agent is approximately as effective as the sodium bicarbonate base material on flammable liquids and vapors. It has the added advantage of being an effective extinguishing agent in Class A (ordina (ordinary ry combustibles) combustibles) fires. C. Potassiu Potassium m Bica Bicarbon rbonate ate Base (ANS (ANSUL UL ‘Pur ‘Purpleple-K’): K’): This agent, which was introduced commercially in the United States in the 1960s, has been shown to be more effective than the sodium bicarbonate base material. Hence, it is becoming the standard dry chemical in high intensity fire applications.

THE ANS ANSUL UL NA NATUR TURAL AL GA GAS S FIR FIRE E EXT EXTING INGUIS UISHME HMENT NT CON CONCEP CEPT T Page 6

THE ANSUL NA NATURAL TURAL GAS FIRE EXTING EXTINGUISHME UISHMENT NT CONCEPT

ANSUL has given very careful consideration consideration to the charact characteristic eristics s of the natural gas fire, the compatibility of, and experimental information on, the available fire extinguishing agents. Combining this with the practical aspects aspects of the fire situation, ANSUL ANSUL has developed a conceptual approach to the extinguishment of natural gas fires. This concept, which outlines the selection and application of most appropriate extinguishing agent for the various potential fire situations, is illustrated in Figure 4.

The ANSUL ANSUL concept is based on the following: following: A. Vapor Vapor – Pressu Pressure re Fire Fires: s: The only extinguishing agents commercially available in a wide range of equipment and capable of extinguishing flammable gas fires are the dry chemicals and carbon dioxide. Of these two types, the dry chemicals are more effective and have the added advantage of concise experimental data to support the design criteria in this application. Of the two more common dry chemicals, the potassium bicarbonate base agent is more effective, but is also more expensive than the sodium bicarbonate base agent. Therefore, some users prefer the sodium bicarbonate base agent from an economical standpoint.

Natural Gas

State

Configuration

Best Solution

Liquid

Vapor

Pressure

Dry Chemical

Contained

Carbon Dioxide

Pressure/Pool

Spill

Dry Chemical or Dry Chemical and High Expansion Foam

Dry Chemical or Dry Chemical and High Expansion Foam

FIGURE 4 The ANSUL Concep Conceptt 003382

THE ANS ANSUL UL NA NATUR TURAL AL GAS FIR FIRE E EXT EXTING INGUIS UISHME HMENT NT CON CONCEP CEPT T Page 7

THE ANSUL ANSUL NA NATURAL TURAL GAS FIRE EXTINGUISHM EXTINGUISHMENT ENT CONCEPT (Continued) B. Vapor Vapor – Contai Contained ned Fire Fires: s: The most appropriate means for extinguishing a fire or inerting the atmosphere prior to a fire in an enclosed volume is by using a gaseous extinguishing agent and a total flooding approach. In enclosed volumes, these systems are normally operated automatically when gas detectors sense a concentration of 1/4 to 1/2 the lower explosive limit of the fuel involved.

Since there may be flammables other than natural gas in the protected volume, the system should be designed to produce an agent concentration adequate to inert the most difficult fuel present. C. Liquid Liquid – Pres Pressure sure/Poo /Pooll Fires: Fires: LNG (Liquefied Natural Gas) pressure fires of any significance will usually produce pools of the fuel in the vicinity of the failure. For the same reasons outlined for pressure fires with the vapor, the dry chemicals are the most effective agents for LNG pressure fires. However, the presence of obstructions (process equipment, piping, etc.) is extremely significant since the dry chemical may not extinguish flames that are substantially shielded from the agent stream. In this case, one has two alternatives:

Provide enough dry chemical application points to preclude the possibility of any flames being shielded by obstructions. Utilize high expansion foam to bring the spill fire under control by vapor dispersion and radiation reduction, after which it may be desirable to extinguish the remaining flames with dry chemical. D. Liquid – Spill Fires: In this type of fire, there are two significant considerations that must be taken into account during the design of the fire extinguishment equipment. One is the rate of natural gas vaporization anticipated as a result of the spill of LNG on the surrounding surface. The design criteria developed for both dry chemical and high expansion foam were based on experiments where the burning LNG was vaporizing at an approximate rate of 0.5 in./minute (1.27 cm/minute). A “fresh” LNG spill on the ground, especially if the ground has a high moisture content, will result in an increased vaporization rate up to 3.0 times steady state conditions17. Th This is hig highe herr vaporization rate will increase the fire intensity. This problem is very important in automatic systems where the agent is intended to be applied very quickly (within seconds) after ignition. This problem is not so significant with manually operated fire extinguishing equipment as the LNG (Liquefied Natural Gas) spill will usually freeze the ground to such an extent that the vaporization rate will have reached equilibrium before the extinguishers are manned. This does not, however, imply that it is sound practice to delay the application of the agent until a stabilized condition is attained. The minimum dry chemical application rates which will just extinguish a steady state LNG spill fire (negligible ground heating effect and maximum radiation-induced burning rates) are increased by a factor of up to 2.5 for the burning rates that exist for fires immediately following the LNG spill on land. (See Figure 12.) A second important important consideration consideration is the presence of obstructions in the spill area. Like pressure/pool fires, two alternatives are available: Use of dry chemical from sufficient application points to preclude the possibility of shielded flames; or use of high expansion foams to control the fire followed by dry chemical to extinguish the remaining flames.

It should be recognized that in both pool and spill fires vapor concentration reduction may be desirable under certain conditions. The application of high expansion foam can accomplish this as previously stated. Specific reference to its use is found on Page 14.

THE EXPER EXPERIMENT IMENTAL AL EXPER EXPERIENCE IENCE Page 8

THE EXPERIM EXPERIMENT ENTAL AL EXPERIEN EXPERIENCE CE

The basis for ANSUL’s concept and design recommendations is a direct result of five major testing programs involving the control and extinguishment of natural gas and LNG fires. The programs are illustrated in Figure 5.

S_it_e _

D __a_te _ T _e _s _t_s_

Longview, Texas7

1951 91

Types of T _e _s _t_s_

Agents T _e _s _t_e_d_

Vapor-Nonlmpinging Jet

Sodium Bicarbonate

Vapor-Horizontal Impinging Jet Vapor-Downward Impinging Jet Vapor-Split Pipe Impinging Jet Six Lakes, Michigan8

1965 48


Sodium Bicarbonate

1969 107


LNG Pool Fires

Potassium Bicarbonate

The 1973 tests, conducted at Norman, Oklahoma, determined that “fresh” LNG LNG spills with with accelerated accelerated boil-off rates increased increased dry chemical flow rates for extinguishment.

Sodium Bicarbonate Potassium Bicarbonate High Expansion Foam Monoammonium Phosphate

Norman, 1973 100 17 Oklahoma

LNG Pool Fires (Accelerated Boi oill-Of Offf Ra Rate tes) s)

The 1969 Six Lakes program established the potassium bicarbonate base agent requirements for low flow rate (200-1600 ft3 /sec (5.7-45.3 m3 /sec)) gas fires and also also served to compare the relative fire extinguishing effectiveness of potassium bicarbonate and potassium chloride base dry chemicals.

Monoammonium Phosphate

Potassium Chloride Marinette, 1972 43 10 Wisconsin

The 1965 Six Lakes program was conducted to compare the effectiveness of potassium bicarbonate, monoammonium phosphate and sodium bicarbonate base dry chemicals on two of the four gas transmission hazards tested in the Longview program. From this experimentation, definite design criteria for the potassium bicarbonate base agent were developed for the two hazards tested, and correlations between the relative extinguishing effectiveness of sodium and potassium bicarbonate base agents produced the potassium bicarbonate base agent design criteria for the other two hazards.

The 1972 program, conducted at ANSUL’s Fire Technology Center,, was performed Center performed to determine the minimum agent requirerequirements for sodium bicarbonate, potassium bicarbonate, monoammonium phosphate and high expansion foam on LNG pool fires of 400 (37.2 m2) and 1200 (11 (111.5 m2) ft2 in area.

Vapor apor-Ho -Horiz rizonta ontall Potass Potassium ium Impinging Jet Bicarbonate

Six Lakes, Michigan9

The 1951 Longview program established the technical information for the use of sodium bicarbonate base dry chemical on four variations of gas pressure fires that are typically found in the natural gas transmission industry.

Sodium Bicarbonate Pot otas assi sium um Bicarbonate High Expansion Foam

FIGURE 5 ANSUL Large Scale Natural Gas Fire Fire Testing Testing Programs

THE TH E GEN GENERA ERAL L BEH BEHA AVIO VIOR R OF EXT EXTING INGUIS UISHIN HING G AGE AGENT NTS S Page 9

THE GENERAL BEHAVIOR OF EXTINGUISHING AGENTS

In situations other than total flooding, it is generally accepted that if an extinguishing agent is not applied to a fire at a sufficient rate, the fire will not be extinguished12. It is also also known known that, up to a certain point, increasing the agent’s application rate will result in a shorter extinguishment time. This extinguishing time and agent application rate relationship has been found to be hyperbolic as shown in Figure 6.

) c e s – t ( E M I T G N I H S I U G N I T X E

tminute

Rminute

AGENT AGE NT APPLI APPLICA CATIO TION N RATE RATE

(R – lb/sec lb/sec (kg/s (kg/sec)) ec)) 003385

FIGURE 6 General Relationship of Agent Rate and Extinguishing Time

THE GEN GENERA ERAL L BEH BEHA AVIO VIOR R OF EXT EXTING INGUIS UISHIN HING G AG AGENT ENTS S Page 10

THE GENERAL BEHAVIOR OF EXTINGUISHING AGENTS (Continued)

Another illustration of this behavior is shown in Figure 7, where the agent quantity and agent application rate are plotted. In a number of experimental programs, it has been determined that there is an optimum agent application rate (Ropt) at which which rate the leas leastt amount of agent (Qminute) will be required for extinguishm extinguishment. ent. Application rates less than Ropt result in longer extinguishment times and the expenditure of more agent than at Ropt . Furthermore, if the application rate is less than Rmin, an infi infini nite te quantity of agent would theoretically be unable to extinguish the subject fire. Rmin has been found to be in the range of 0.4 to 0.5 Ropt, wh whic ich h accounts for the 2.0 factor of safety usually put on Rminute to arrive at a design rate. If the agent is applied at a rate greater than Ropt, the time of extinguishm extinguishment ent is usually not not reduced to any significance (as shown in Figure 6) resulting essentially in the wasting of agent (Q >> Qminute).

) b l – Q ( Y T I T N A U Q T N E G A

Qminute

Rminute

Ropt

003386

AGENT AGE NT APPLI APPLICA CATIO TION N RATE RATE

(R – lb/sec lb/sec (kg/s (kg/sec) ec)))

FIGURE 7 General Relationship of Agent Rate and Quantity

THE SPE SPECIF CIFIC IC AG AGENT ENT FL FLOW OW RA RATE TE REQ REQUIR UIREME EMENTS NTS FO FOR R NA NATUR TURAL AL GA GAS S FIR FIRES ES Page 11

THE SPECIFIC AGENT FLOW RATE REQUIREMENTS FOR NATURAL NA TURAL GAS FIRES

After all the experimental information was analyzed, recommended design criteria were developed for the application of the extinguishing agents to the various natural gas fire configurations. These recommendations are graphically shown in Figures 8 through 15. Figure 8: Recommended Dry Chemical Design Application Rates for the Extinguishment of Non-lmpinging Natural Gas and LNG Pressure Fires.

E. The design design rate rate selected selected for high expansion expansion foam must produce fire control with at least 90% reduction of the radiant heat flux under the conditions described in Figure 15. It is generally accepted that a minimum application rate of 6 ft3 per minute per ft2 (1.8288 m3 per minute per m2) is desir desirable able as determined by testing. Under some circumstances faster control times may be essential, or longer control times acceptable. The entire foam application rate/fire control time relationship has been included in Figure 15.

Figure 9: Recommended Dry Chemical Design Application Rates for the Extinguishment of Horizontal Impinging Natural Gas and LNG Pressure Fires.

F. In the combin combined ed use of high expansio expansion n foam and dry chemicals, the high expansion foam application must be continued until the dry chemical has completely extinguished all flames.

Figure 10: Recommended Dry Chemical Design Application Rates for the Extinguishment of Downward Impinging Split Pipe Natural Gas and LNG Fires.

For the graphs in Figures 8 through 15, the criteria shown in solid lines are based on actual experimentation and those shown in dashed lines are correlations (based on relative extinguishing effectiveness of the agents) or extrapolations. The design information on LNG pressure fires are theoretical and it assumes that the LNG completely and immediately flashes to a vapor at 70 °F (21 °C). upon exiting the failure point. point. The dry chemical rates are then based on the free volume of natural gas using an expansion factor of 600. This approach is justified on the basis of reported correlations attained in experimentation with gaseous and liquid propane.14

Figure 11: Effects of Dry Chemical Application Rate on Fire Extinguishment Time for LNG Spill Fires with a Total Evaporation Rate of 0.5 Inches per Minute. Figure 12: Recommended Dry Chemical Design Application Density for the Extinguishment of LNG Pool Fires for Various Vaporization Rates. Figure 13: Recommended Dry Chemical Design Application Density for the Extinguishment of LNG Fires for the Steady State Vaporization Rate. Figure 14: Recommended Dry Chemical Design Application Density for the Extinguishment of LNG Fires for Initial Accelerated Vaporization Rates. Figure 15: Effects of Foam Application Rate of Control Time for LNG Spill Fires Using 500:1 High Expansion Foam. Figures 16 Through 20: Recommended Dry Chemical Design Quantities Based on the Recommended Application Rates Shown above, using 30 Second Effective Discharge Time. These figures can be used to estimate total agent design quantities desired.

In general, the following additional criteria apply: A. Dry Chemical Chemical Fire Fire Extinguisher Extinguishers s utilizing utilizing high veloci velocity ty dry chemical streams are superior to soft or “fan” streams for the extinguishment of natural gas or LNG fires. Care should be exercised on LNG spill fires to avoid disrupting the liquid surface of the fuel with the agent which would cause an increase in the burning intensity. B. All the design design criteria criteria for for dry chemical chemical on natural gas pressure pressure fires employ a safety factor of two (2.0) on the minimum rate found necessary to effect extinguishment in the experimental programs. When designing automatic fixed nozzle dry chemical systems, the applied safety factors would be increased substantially to achieve much higher application rate densities (Ib/sec/ft 2). The minimum design rate for LNG spills in Figure 11 also has a safety factor of 2.0 times the rate found necessary to effect extinguishment in the testing. C. Dry chemical chemical extinguishe extinguishers rs and extinguishi extinguishing ng systems systems should be selected to produce optimized discharge times according to application conditions. D. From NFP NFPA A 11 “Standa “Standard rd for for Low-, Low-, Medium Medium-, -, and Hig HighhExpansion Foam”3: “In (testing), (testing), control was establishe established d with expansion ratios greater than 250:1, although an expansion ratio of about 500:1 proved most effective.”

THE ANS ANSUL UL REC RECOMM OMMEND ENDED ED AG AGENT ENT QU QUANT ANTITY ITY REQ REQUIR UIREME EMENTS NTS Page 12

RECOMMENDED DRY CHEMIC RECOMMENDED CHEMICAL AL DESIGN APPLICA APPLICATION TION RATES RAT ES FOR NON-IMPINGING NATURAL NATURAL GAS AND LNG PRESSURE FIRES (2.0 SAFETY FACTO FACTOR R APPLIED)

(Based on data from References 7, 8 and 9.) LNG agent requirements are theoretical and assume that the LNG completely vaporizes upon contact with the air and immediately expands to its 70 °F (21.1 (21 .1 °C) con condit dition ion (60 (600 0 times expansion).

70 (31.8)

60 (27.2)

50 (22.7)

) c e s / g k ( c e s / 40 b l (18.1) – e t a R n o i t a c i 30 l p p (13.6) A n g i s e D l a 20 c i m (9.1) e h C y r D

T Y F I - F U S L P

- K ’ e e l r p ‘ P u

10 (4.5)

0 0

500 (14.2)

1000 (28.3)

1500 (42.5)

2000 (56.6)

2500 (70.8)

Natural Gas Flow Rate – ft3 /sec (m3 /sec)

0

500 (1893) 1000 (3785) 1500 (5678) LNG Flow Rate – gal/minute (liters/minute) FIGURE 8 003387

THE TH E ANS ANSUL UL REC RECOM OMMEN MENDED DED AGE AGENT NT QUA QUANTI NTITY TY REQ REQUIR UIREME EMENTS NTS Page 13

RECOMMENDED DRY CHEMIC RECOMMENDED CHEMICAL AL DESIGN APPLICA APPLICATION TION RATES FOR HORIZONTAL HORIZONTAL IMPINGING NATURAL NATURAL GAS AND LNG PRESSURE FIRES (2.0 SAFETY FACTO FACTOR R APPLIED)

(Based on data from References 7, 8 and 9.) Dashed lines indicate extrapolations or correlations: LNG agent requirements are theoretical and assume that the LNG completely vaporizes upon contact with the air and immediately expands to its 70 °F °F (21. (21.1 1 °C) °C) condit condition ion (600 times expansion).

70 (31.8)

60 (27.2)

) c e s / g k ( c e s / b l – e t a R n o i t a c i l p p A n g i s e D l a c i m e h C y r D

50 (22.7)

40 (18.1)

Y T F I F S U L P

30 (13.6)

20 (9.1)

’ - K e e l r p ‘ P u

10 (4.5)

0 0

200 (5.7)

400 (11.3)

600 (17)

800 (22..7)

1000 (28.3)


0

200 (757) 400 (1514) 600 (2271) LNG Flow Rate – gal/mi gal/minute nute (liters/minute (liters/minute)) FIGURE 9 003388


RECOMMENDED DRY CHEMIC RECOMMENDED CHEMICAL AL DESIGN APPLICA APPLICATION TION RATES RAT ES FOR DOWNWARD IMPINGING SPLIT PIPE NATURAL GAS AND LNG LNG PRESSURE FIRES (2.0 SAFETY FACTO FACTOR R APPLIED)

(Based on data from References 7, 8 and 9.) Dashed lines indicate extrapolations or correlations: LNG agent requirements are theoretical and assume that the LNG completely vaporizes upon contact with the air and immediately expands to its 70 °F °F (21. (21.1 1 °C) °C) condi condition tion (600 times expansion).

70 (31.8)

60 (27.2)

) c e s / g k ( c e s / b l – e t a R n o i t a c i l p p A n g i s e D l a c i m e h C y r D

50 (22.7)

40 (18.1)

Y T F I F S U L P

30 (13.6)

20 (9.1)

’ - K e e l r p ‘ P u 10 (4.5)

0 0

100 (2.8)

200 (5.7)

300 (8.5)

400 (11.3)

500 (14.2)

Natural Gas Flow Rate – ft3 /sec (m3 /sec) 0

100 (378.5) 200 (757.1) 300 (1135.7) LNG Flow Rate – gal/minute (liters/minute) FIGURE 10 003389


DRY CHEMIC CHEMICAL AL APPLIC APPLICA ATION RATE RATE VS. EXTINGUISHMENT EXTINGUISHMENT TIME FOR LNG SPILL FIRES WITH BURNING RATE RATE OF 0.5 IN./MIN IN./MINUTE UTE (1.27 cm/minute) cm/minute)

Based on data from Reference 10. Design Application Rate is Based on 2.0 Safety Factor Applied to Minimum Rate

30

Minimum ‘Purple-K’

25 ‘Purple-K’ ) s d n 20 o c e s ( – e m i T 15 t n e m h s i u g n 10 i t x E

Minimum PLUS-FIFTY

PLUS-FIFTY PLUS-FIFTY Design Application Rate

5

‘Purple-K’ Design Application Rate 0

0

0.01 (0.05)

0.02 (0.10)

0.03 (0.15)

0.04 (0.2)

0.05 (0.24)

0.06 (0.29)

0.07 (0.34)

Dry Chemical Application Rate – (lb/sec/ft2) FIGURE 11 003390


RECOMMENDED DRY CHEMIC RECOMMENDED CHEMICAL AL DESIGN APPLICA APPLICATION TION DENSITIES DENSIT IES FOR A RANGE OF LNG POOL POOL BURNIN BURNING G RATES RATES (2.0 SAFETY FA FACTOR CTOR APPLIED) APPLIED)

0.07 (0.34)

0.06 (0.29)

) 2 t 0.05 f / c (0.24) e s / b l ( – y t i s n e 0.04 D (0.2) n o i t a c i l p p A n g 0.03 i s (0.15) e D l a c i m e h C 0.02 y r (0.10) D

Y F T I F U S L P

K ’ e - K l e p p r u P ‘ P

0.01 (0.05)

0 0.5 (1.27)

1.0 (2.5)

1. 5 (3.81)

LNG Linear Burning Rate – in./minute (cm/minute) FIGURE 12 003391


RECOMMENDED DRY CHEMIC RECOMMENDED CHEMICAL AL DESIGN APPLICA APPLICATION TION RATES FOR LNG POOLS BURNING AT 0.5 IN./MINUTE (1.27 cm/min cm/minute) ute) (2.0 SAFETY FA FACTOR CTOR APPLIED APPLIED))

1000 (453.6)

500 (226.8)

300 (136.1) 200 (90.7) ) c e s / g 100 k ( c (45.4) e s / b l – e t 50 a R (22.7) n o i t a c 30 i l p (13.6) p A n 20 g i s (9.1) e D l a c i m e 10 h C (4.54) y r D

5 (2.27) 3 (1.36) 2 (0.91)

1 (0 (0.45 .45)) 10 (0.9)

50 (4.6)

100 (9.3)

500 (46.5)

1000 (92.9)

5000 (464.5)

10000 (929)

LNG Are Area a – ft2 (m2) FIGURE 13 003392


RECOMMENDED DRY CHEMICAL DESIGN APPLICATION RATES FOR LNG POOLS BURNING AT AT 1.5 IN./MINUTE (3.81 cm/minute) (2.0 SAFETY FA FACTOR CTOR APPLIED) APPLIED)

1000 (453.6)

500 (226.8)

300 (136.1) 200 (90.7) ) c e s / g 100 k ( c (45.4) e s / b l – e t 50 a R (22.7) n o i t a c 30 i l p (13.6) p A n 20 g i s (9.1) e D l a c i m e 10 h C (4.54) y r D

Y T F I F ’ S K U e L l P r p

P u ‘

5 (2.27) 3 (1.36) 2 (0.91)

1 (0 (0.4 .45) 5) 10 (0.9)

50 (4.6)

100 (9.3)

500 (46.5)

1000 (92.9)

5000 (464.5)

10000 (929)



EFFECT OF FOAM APPLICATION APPLICATION RATE RATE ON CONTROL TIME FOR LNG SPILL FIRE USING 500:1 HIGH EXPANSION EXPANSION FOAM

300

Fire Control is defined as when the radiant heat flux has been reduced by 90 percent or more.

250 Six (6) ft3 /minute/ft2 (1.83 m3/ minute/m2) is a generally accepted minimum design rate.

s d n 200 o c e S – e m i T 150 l o r t n o C e r 100 i F

50

0 0

5 6 (1.5) (1.83)

10 (3.05)

15 (4.6)

High Expansion Foam Application Rate – ft 3 /minute/ft2 (m3 /minute/m2) FIGURE 15 003394

If LNG pools are burning, the common practice is to provide foam discharge for 3 times the average response time for fire fighting personnel to arrive on site and extinguish the fire with dry chemical. In the absence of this information, it has been generally accepted for the purpose of design that a minimum 60 minute continuous foam discharge is adequate for foam concentrate storage tank sizing. ANSUL recomm recommends ends continuous continuous foam discharge for burning LNG situations. If LNG pools are not burning and foam is being used for vapor mitigation, it is desirable to keep a minimum of 3 ft. (0.91 m) depth of foam over the spill area. Manually ON/OFF cycling the discharge as required is recommended to maximize available foam concentrate supplies. After initial foam coverage based on 3 minutes of discharge, it is possible that reapplications may only be required every 30 minutes. This can be affected by individual site conditions.

Steady state LNG pool evaporation is approximately 0.025 in. (0.0635 cm) per minute. When maintaining a 3 ft (0.91 m) foam depth over the spill area of non-burning LNG, the evaporation rate may increase in the range of 0.050 in. (0.127 cm) to 0.075 in. (0.191 cm) per minute from the heat input provided by the foam drainage. Evaporation rates of continuously foamed LNG that is burning may be in a range above 0.075 in. (0.191 cm) per minute. The evaporation data listed above is based on JET-X Agent and Hardware testing conducted at ANSUL’s R&D facility in a cement containment pit using LNG that was above 99% Methane.


RECOMMENDED DRY RECOMMENDED DRY CHEMIC CHEMICAL AL DESIGN QUANTITIE QUANTITIES S FOR NON-IMPINGING NATURAL NATURAL GAS AND LNG LNG PRESSURE FIRES

(Based on Recommended Application Rates and 30 Seconds Effective Ef Discharge Time)

1400 (635)

1200 (544.3)

Y T F I F S U L P

1000 (453.6) )

g k ( b l – s e i t 800 i t n (362.9) a u Q n g i s e D l 600 a c i (272.2) m e h C y r D

’ K e e l p u r P ‘

400 (181.4)

200 (90.7)

0 0

500 (14.2)

1000 (28.3)

1500 (42.5)

2000 (56.6)

2500 (70.8)


0



RECOMMENDED DRY RECOMMENDED DRY CHEMIC CHEMICAL AL DESIGN QUANTITI QUANTITIES ES FOR HORIZONT HORIZ ONTAL AL IMPING IMPINGING ING NATURAL NATURAL GAS AND LNG PRESSURE FIRES

(Based on Recommended Application Rates and 30 S ec o n d s Effective Ef Discharge Time)

1400 (635)

1200 (544.3)

) g k (

1000 (453.6)

b l – s e i 800 t i t n (362.9) a u Q n g i s e D 600 l a (272.2) c i m e h C y r D 400

Y T F I F S U L P

’ K e l r p u P ‘

(181.4)

200 (90.7)

0 0

200 (5.7)

4000 (11.3)

600 (17)

800 (22.7)

1000 (28.3)


0

200 (757.1) 400 (1514.2) 600 (2271.2) LNG Flow Rate – gal/mi gal/minute nute (liters/minute (liters/minute)) FIGURE 17 003396


RECOMMENDED DRY RECOMMENDED DRY CHEMIC CHEMICAL AL DESIGN QUANTITIE QUANTITIES S FOR DOWNWARD IMPINGING SPLIT PIPE NATURAL NATURAL GAS AND LNG PRESSURE FIRES

(Based on Recommended Application Rates and 30 Seconds Effective Ef Discharge Time)

1400 (635)

1200 (544.3)

1000 (453.6) ) g k (

Y T F I F S U L P

b l – s 800 e i t i (362.9) t n a u Q n g i s e 600 D l a (272.2) c i m e h C y r D 400

’ K e e l r p u P ‘

(181.4)

200 (90.7)

0 0

100 (2.8)

200 (5.7)

300 (8.5)

400 (11.3)

500 (14.2)


0



RECOMMENDED DRY RECOMMENDED DRY CHEMIC CHEMICAL AL DESIGN QUANTITI QUANTITIES ES FOR LNG POOLS BURNING AT 0.5 IN./MINUTE (1.3 cm/minute) (30 SECOND DISCHARGE TIME)

10000 (4536)

g k / b l – y t i t n a u Q n g i s e D l a c i m e h C y r D

1000 (453.6)

Y T F I F ’ S K U e L l p P r P u ‘

100 (45.4)

10 (4.5) 10 (0.9)

100 (9.3)

1000 (93)

10000 (929)



RECOMMENDED DRY RECOMMENDED DRY CHEMIC CHEMICAL AL DESIGN QUANTITIE QUANTITIES S FOR LNG POOLS BURNING AT 1.5 IN./MINUTE (3.8 cm/minute) (30 SECOND DISCHAR DISCHARGE GE TIME)

10000 (4536)

1000

) (453.6) g k ( b l – y t i t n a u Q n g i s e D l a c i m e h C y r D

Y T F I F S U ’ L K P l e p r P u ‘

100 (45.4)

10 (4.5) 10 (0.9)

100 (9.3)

1000 (93)

10000 (929)


COMMERCIA COMM ERCIALL LLY Y AVAILA AILABLE BLE FIRE SUPPR SUPPRESSIO ESSION N EQUI EQUIPMENT PMENT Page 25

COMMERCIALLY AV COMMERCIALLY AVAILABLE AILABLE FIRE SUPPRESSION EQUIPMENT A. High High Expan Expansio sion n Foam: Foam: Foam expansion rates of 500:1 are favored for fire control and are well-suited for vapor dispersion. ANSUL recommends the following high expansion foam generators for LNG with performance characteristics as shown. Calculating Foam Quantity For Local Application (LNG) High Expansion Generators Typical Discharge Characteristics

Generator Inlet Pressure p _s_i____(_b_a_r_) __ 50 (3.45) 75 (5.17) 100 (6.89)

Foam Output c_f_m_____(c_m __m_)_ 2 ,2 4 0 (63) 3 ,2 0 0 (91) 3,735 (106)

Solution Flow g_p_m ______(l_p_m_)_ 35 (132.5) 42 (159) 50 (189.3)

E_x_p_a_n_s_io_n__ 4 6 5 :1 5 5 5:1 5 4 5:1

JET-X-15A (LNG)

50 75 1 00

(3.45) (5.17) (6.89)

12,625 14,495 18,240

(357) (410) (516)

180 22 0 26 0

(681.4) (832.8) (984.2)

5 2 5 :1 4 9 5:1 5 2 5 :1

JET-X-20

40 50 75 1 00

(2.76) (3.45) (5.17) (6.89)

13,443 16,034 2 1 ,1 4 5 24,301

(381) (454) (599) (688)

2 12 23 8 29 4 33 8

(802.5) (900.9) (1112.9) (1279.5)

4 7 4 :1 5 0 4:1 538:1 5 3 8 :1

G __e_n_e_ra _t_o_r JET-X-2A

B. Dry Dry Chem Chemic ical al:: A complet complete e line of dry chemical chemical extinguis extinguishhment systems have been designed specifically for natural gas and flammable liquid applications. Figure 21 summarizes the ANSUL dry chemical chemical product line, illustrating illustrating the flow rates, rates, which can be related to the data contained in this report. Category

Agents

Extinguisher Capacity

Flow Rate

Hand Portable

PLUS-FIFTY

10, 20, 30 lb (4.5, 9, 13.6 kg)

1.5-2.5 lb/sec (0.7-1.1 kg/sec)

‘Purple-K’

9, 18, 27 lb (4.1, 8.2, 12.2 kg)

PLUS-FIFTY

150, 350 lb (68, 158.8 kg)

‘Purple-K’

125, 300 lb (56.7, 136.1 kg)

Hand Hose Line Systems

PLUS-FIFTY

150, 350, 500, 1000, 1500, 2000, 3000 lb (68, 158.8, 226.8, 453.6, 680.4, 907.2, 1360.8 kg)

4.5-10.0 Ib/sec (2-4.5 kg/sec) for hand lines

Vehicle Mounted

‘Purple-K’

125, 300, 450, 900, 1350, 1800, 2700 lb (56. (5 6.7, 7, 13 136. 6.1, 1, 20 204. 4.1, 1, 40 408. 8.2, 2, 612.4, 816.5, 1224.3)

25-100 Ib/sec (11.3-45.4 kg/sec) forr tur fo turre rets ts fo forr 135 1350 0 lb lb (61 (612. 2.4 4 kg) kg) capacity and larger

Wheeled

Engineered Systems

4.5-8.5 Ib/sec (2-3.9 kg/sec)

4-100 Ib/sec (1.8-45.4 kg/sec) for piped systems depending on their capacity FIGURE FIG URE 21

C. Detectio Detection n and Con Control trol:: This report is not intended to provide detailed coverage of the detection and control aspects of fire control and extinguishment. However, it should be recognized that whether automatic or manual, the detection control system design is integral to the extinguishing system design, if an optimum total system control and extinguishing capability is to be realized.

BIBLIOGRAPHY Page 26

BIBLIOGRAPHY

1. National Fire Fire Protection Protection Associati Association, on, “Storage and and Handling of Liquefied Natural Gas (LNG),” NFPA Standard 59A. 2. Walls, Walls, W. W. L., “LNG: “LNG: A Fire Servi Service ce Appr Appraisal aisal,” ,” FIRE FIRE JOURNAL, January, 1972. 3. National Fire Fire Protection Protection Associatio Association, n, “Standard For Low-, Low-, Medium-, and High-Expansion Foams,” NFPA NFPA 11. 4. RE REMO MOVE VED D 5. RE REMO MOVE VED D 6. RE REMO MOVE VED D 7. “Natural Gas Fire Tests,” Tests,” Technica Technicall Bulletin Number 32, Ansul Ansul Incorporated, Marinette, Wisconsin. 8. “Fire Tests With With Natural Gas Gas Jets – Six Six Lakes,” Lakes,” Ansul Ansul Incorporated, Marinette, Wisconsin. 9. “Fire Tests Tests With Natural Natural Gas Jets – Six Lakes,” Ansul Incorporated, Marinette, Wisconsin (1969). 10. “LNG Fire Control, Control, Fire Extinguishment Extinguishment and Vapor Vapor Dispersion Dispersion Tests,” University Engineers, 1972. 11. RE REMO MOVE VED D 12. Guise, A. B., and Lindlof, Lindlof, J. A., “A Dry Chemical Chemical Extinguish Extinguishing ing System,” System ,” NFPA NFPA QUART QUARTERL ERLY Y, Vol Volume ume 49, Number 1, July, July, 1955. 13.. RE 13 REMO MOVE VED D 14. Guise, Guise, A. B., B., “Fire “Fire Tests Tests Made Made On LP LP Gas Gas,” ,” LP GAS GAS,, May, May, 1948. 15.. RE 15 REMO MOVE VED D 16.. RE 16 REMO MOVE VED D 17. ”An Experimental Experimental Study Study on the Mitigation Mitigation of Flammable Flammable Vapor Vapor Dispersion and Fire Hazards Immediately Following LNG Spills On Land,” For AGA by University Engineers, February, February, 1974.

ANSUL INCOR INCORPORA PORATED TED MARINETTE, WI 54143-2542 715-735-7411

d e t a r o p r o c n I l u s n A 7 0 0 2 © t h g i r y p o C

2 8 5 1 5 7 F . o N m r o F

ANSUL_Fire Protection Solutions for LNG

Recommend Documents