Essentials in Handling Large Oil Fires

Essentials in Handling Large Oil Fires

Conclusion of Paper on This Subject—Improvements in Fire-Banks— The Foam System of Extinguishment—Protecting Fuel Oil Plants

THIS paper, which is concluded from page 37, FIRE ENGINEERING, January 10 issue, is the result of careful study and research into the origin, peculiarities and methods of fighting fires in oil tanks. It is full of practical suggestions for the fire chief who has this problem to wrestle with:

Improvements in Fire-banks

The most serious menace from a crude oil fire comes when it boils over. If the boil-over can be held within prearranged firebanks, there is no further difficulty; but fire-banks have not always proved effective, and we therefore turned our attention to the possibility of improving fire-bank design. Here again we turned to small model apparatus, as tests on full size tanks with fire walls would have been impracticable. We therefore set up the apparatus shown diagrammatically in Fig. 8. A segment of the “tank” from which the “boil-overs” were discharged is shown, and model walls of various sizes and shapes were set up—the whole outfit being made to a scale of about one inch to the foot. Altogether in these tests we made some 66 of these walls, and allowed 1,000 miniature boil-overs to flow against them. These tests showed that all the present types of wall could be vastly improved by the application of the old sea wall principle of folding the advancing waves back on themselves.

Fig. 10. Some Practicable Methods of Applying the Principle of the Fire-bank.

F’ig. 9 shows the comparative height of different walls which will fully retain the same spill. A one foot vertical wall with a coping is as effective for turning back the wave of oil from boil-over as a four foot plain vertical wall or a five foot earth dyke with a slope of 1 to 1, or a seven foot earth dyke with a slope of 1 1/2 to 1.

Fig. 10 shows a number of practicable methods of applying this principle. That this principle is not just an experimental theory was demonstrated soon after the tests by a fire in one of the California fields, as shown in Fig. 11. A tank of crude burned and boiled over, the tank being distant 160 feet on practically level ground from the company’s wooden frame pumping plant. By building temporary earth fire walls topped with a flare-back of corrugated iron, the station was protected from ignition.

So much for the tests that we have conducted with the view to increasing our knowledge of behavior of burning oil tanks and protecting against spread of fire by boil-overs.

The American Petroleum Institute made in the early part of this year a very comprehensive survey of the records of all oil fires that have occurred in the United States during the past ten years. I believe that this is one of the most thorough investigations of fire records that has ever been made in this country, and it has yielded some very interesting results. The entire report has been published, and any of you can obtain a copy by addressing the American Petroleum Institute in New York City.

Contributions to Knowledge of Oil Fire-Fighting

There have been several other contributions in recent years to the oil industry’s knowledge of fire protection. Summarized, the more important developments have been—

Fire resistive oil storage tanks.

  1. Tank appurtenances which minimize fire hazards.
  2. Recognition and elimination of fire hazards.
  3. Knowledge regarding the behavior of oil fires.
  4. Fire walls of a design which will more effectively control a flow of oil.
  5. Effective oil fire-extinguishing systems.

Going into a little more detail regarding these developments, we can say that a modern oil tank, properly equipped, is very fire resistive. Many cases are on record where flames from exposing fires have enveloped tanks containing gasoline without damage to tank or contents.

The modern all-steel tank has an excellent fire record as regards lightning, and is believed to be immune from the lightning hazard. This is important because statistics show that 55% of the oil-tank fires in the last ten years were caused by lightning.

Fig. 8. Model for Testing Wave-Stopping Ability of Firebanks.

Static electricity has been blamed for many oil fires, and has quite likely caused several. During the last few years knowledge has been gained regarding static phenomena, and practices have been introduced which tend to protect oil risks from this hazard.

We have learned what to expect when an oil tank is on fire. We know now that if a tank contains a refined oil (gasoline, kerosene, distillate, etc.,) and is ignited, it will burn quietly until it is extinguished or the contents entirely consumed. There is no danger of an explosion or a boil-over which will force the oil out of the tank.

Protecting Fuel Oil Plans

In the case of fuel oils, which arc usually the only other oils found in cities, we know that there is a possibility that the oil will ultimately boil out of the tank if the fire is not extinguished within the first few hours after ignition. We have learned, however, to fairly accurately predict the time when we may expect a possible boil-over, so that we can utilize the time before this boil-over to prepare fire walls, etc., to control any flow of oil, and to take the necessary precautions to avoid anything from being endangered by the flow of oil from the boilover. As a matter of fact there is little likelihood of. the occurrence of a fuel-oil fire. Most of the fuel oils have a flash point above 150° F., and it is extremely difficult to ignite a tank of fuel oil with paper and matches.

Fig. II. How a Wooden Pumping Plant of California Oil Company Was Protected by Temporary Earth Walls, Surmounted by a Flareback of Corrugated Iron, When a Tank of Crude Oil Boiled Over.Fig. 9. Comparative Height of Different Walls Which Will Fully Retain the Same Spill.

Progress in Extinguishing System

Considerable progress has also been made in developing effective extinguishing systems for oil fires. Fire records show cases where tanks containing 80,000 barrels of gasoline on fire have been quickly extinguished with foam systems. These tanks are 120 feet in diameter and 40 feet high, and when we consider that at the surface there are over 12,000 square feet of gasoline burning at one time, it shows that the method of extinguishment is quite effective.

Fig. 12. Triple Combination Cars of a Fire Department Functioning as Pumpers for Foam Solutions, in an Oil Field Fire.

The details of many of these problems are more for the fire protection engineers of the oil companies than for the fire chiefs, who have no time for anything except high lights on the subject. Probably what the fire chief is principally interested in is how to prevent these fires, and what he can do to more effectively combat a fire of this character. I will therefore briefly discuss these two phases of the subject.

Fire prevention in the oil industry is largely a matter of proper construction detail; although, of course, good housekeeping is Essentials in Handling Large Oil Fires of paramount importance here, as in any other industry. Details of construction involve a special engineering study for each case. This study combines safe structural practice with knowledge gained through experience by the oil industry regarding the hazards incident to the commodities handled. Tank construction has been greatly improved in the last few years, but to show that the type of tank which has been in use during the last ten years has been a relatively good fire risk. I will give you some statistics from the survey of the American Petroleum Institute. In oil marketing stations, which arc usually the only oil plants found in cities, it was found that out of a total of over 11,000 there have been less than two fires per year. This is equivalent to one fire every 5,860 years for each plant, or in other words there is one chance in about 6,000 of any one particular plant catching on fire during a year’s time.

(Continued on page 87)

(Continued from page 78)

As for fire protection—it is difficult to outline definite suggestions for municipal fire departments’ aid in fighting fires in oil plants. It is always a good idea for the fire department officials to occasionally visit each large plant, and have a talk with the superintendent who can assist in outlining a method of procedure and cooperation in case of various contingencies. Where there is a foam system, the fire department can sometimes help with the operation of the system. This is particularly true in cases of fires at night where there may be but few employees at the oil plant, and where the operator of the foam system can be considerably helped by efficient aides.

The Foam System of Protection

For the benefit of those unfamiliar with the foam system, it will perhaps be well to bring in here a brief description of the foam system of protection.

Experience has shown that when a fire resistant froth is floated on top of a burning oil surface the heat of the flames is cut off from the oil surface. As the oil surface cools, its vaporization is decreased so that the flames soon die down from lack of fuel to feed them.

To produce a proper froth for this purpose, two chemical solutions are mixed together in equal proportions. These solutions are stored separately, one in each tank. In order to prevent the foam from being broken down by mechanical agitation before it reaches the fire, it is desirable to bring the solutions together as near as possible to the oil surface where they are to be applied. In practice separate pipe lines are usually run to the oil tank protected by the foam system and the lines are siamized into a mixing chamber at the edge of the oil surface. Here the foam is produced and allowed to flow over the oil surface.

Fig. 13. Various Operations in the Use of a Portable Fire Foam Mixer, for Use in Fighting Fire in Tank Not Equipped with Mixing Box or When the Box Is Out of Commission.

The solutions are supplied in proportions as nearly equal as possible by means of two pumps of the same size and operating at the same rate of speed.

Where a hose stream is desired two hose lines are used—one for each solution—and the solutions brought together in a short length of a larger sized hose siamized to the twin solution lines. A nozzle is used on the end of the short length of single hose, and the foam stream controlled as any other fire stream.

Fire Apparatus as Temporary Foam Pumpers

The fire department can sometimes provide pumping facilities in case the foam pumps are, for some unforeseen reason, disabled at the time of the fire, or in case foam solutions are available where there is no foam system. Chiefs Short and Scott have both had occasion to call upon their departments to function in this way. (See Fig. 12.)

When the fire department pumps foam solution with its regular pumpers, best results arc obtained if positive displacement pumpers (rotary or plunger) are used. These should be preferably of the same type and rating, and should be run at the same rate of speed. It is best to take suction from the top manholes of the solution storage tanks, if sufficient suction hose is available, otherwise the oil companies can probably make some kind of pipe connections for getting solutions to the pumpers.

The fire department can furnish hose lines to replace pipe lines in case the foam solution lines should become broken or otherwise be out of commission, or in case an extension is desirable.

How to Use Foam Hose Streams

Foam hose streams are often used to supplement regular foam mixing boxes on tanks, or for extinguishing pools of oil on the ground. A foam hose stream is used effectively in a different manner than a water stream. The foam should not be directed into the oil, nor should it be sprayed around over the oil surface. Best results are obtained if the stream is impinged against some vertical object, and the foam allowed to float quietly over an oil surface. In extinguishing a tank fire with a foam hose stream, if possible, direct the stream across the flames to the other side of the tank shell, attempting to hit the inside of the tank shell above the oil surface, thus allowing the foam to flow on to the oil surface where it will spread itself across the fire. I have seen instances where the oil surface was almost covered with foam and the fire practically out, when careless handling of a foam stream swept the foam off the surface, and allowed the fire again to flash over a large part of the oil area.

(Continued on page 94)

Essentials in Handling Large Oil Fires

(Continued from page 87)

In directing a foam hose stream on a pool of oil on the ground, the same method should be followed as in covering the oil surface in a tank with foam. Where there is no wall or other object alongside the pool against which to direct the stream, it is best to hit the ground at the edge of the fire. The foam which is thus formed on the ground is pushed by the stream onto the fire.

When it is desired to apply foam to a tank not equipped with a mixing box, or on which the box has perhaps been put out of commission by the collapse of the roof or some other cause, we have found a portable mixer, as shown in Fig. 13, to be very useful. The tube is made of a light strong alloy so that it can be easily carried to the tank and up-ended on the rockers provided at the bottom by a few men.

Fig. 14 shows a test fire in a 30′ tank in which various types of mixers, including one of these portable types (indicating), are being tried out. On this matter of foam protection, our Company has done a large amount of experimental work to determine the proper rates and methods of application to tanks of various sizes; the most effective formula for the two chemical solutions, etc.

So much for the use of foam in fighting oil fires. Water should never be used on the same fire with foam because the foam is broken down by the water. The effectiveness of foam is greatly decreased when a little water is sprinkled on it. The shell of a tank to which foam is being applied, can, however, be cooled with water, and a tank thus cooled can be extinguished with less foam and in shorter time than where the shell is not cooled. This is particularly true with high vapor pressure oils, such as gasoline. The drainage of water must be watched very carefully around an oil fire to prevent spread of fire by oil floating on top of water.

These are some of the more important special features connected with the work of fighting oil fires. I will not take time to discuss any more, for it is likely that you will have some questions to ask and I will welcome a general discussion of this subject for the remainder of the time allotted me. To sum up briefly the message that I have tried to bring: During the past few years we have emerged from uncertainty to quite definite knowledge of what is required for safe oil storage construction. We have learned that a properly constructed gas-tight steel roof tank, containing refined oils, has an excellent fire record, and is relatively a good risk.

Fig. 14. Test of Fire in 30-foot Tank. Various Types of Mixers Are Being Tried Out, Including a Portable One, the Yoke Poles of Which Can Be Seen at Right of Picture.

As for fire fighting, we have, through extensive experiments and actual experience at fires, learned about what is required to make most effective use of both foam and water. I have tried to point out how the local fire departments may be of great assistance to the fire fighting organizations of the different oil companies when fires occur.

In closing I wish to say (and I am sure I speak for the fire protection engineers of the other oil companies as well as for myself), that we invite the cooperation of the fire chiefs in an effort to keep down the fire loss record. We will be glad to work with them at any time on such of their problems where our specialized knowledge can help.

Essentials in Handling Large Oil Fires


Essentials in Handling Large Oil Fires

What Causes “Boil-Over”—Determining Accurately When It Will Occur—“Heat-Wave” Progress Shown by Sensitive Paint Stripe

THIS paper is the result of careful study and research into the origin, peculiarities and methods of fighting fires in oil tanks. It is full of practical suggestions for the fire chief who has this problem to wrestle with:

Large oil fires have always been the cause of great anxiety to fire chiefs. This has probably been largely due to the fact that, until recently, we have had quite limited knowledge regarding the behavior of these fires. At first it appears surprising that more knowledge has not been available on oil fires, but, strange as it may seem, oil tank fires are of such infrequent occurrence that very few men have witnessed many of them. Probably a majority of fire chiefs have not seen a really large oil fire, and even those chiefs who are located in the oil industry centers, have seen comparatively few.

Fig. 1. Model Apparatus for Determining What Goes on Inside Burning Tank Full of Oil.

On account of this infrequent, scattered occurrence of oil fires it was difficult for any one man to get enough experience to teach him much about them. Several years ago the company with which I am connected started a systematic effort to collect information regarding oil fires. We studied all available fire records, and for the past few years I have personally witnessed, or investigated on the ground, every large oil fire which has occurred on the Pacific Coast. But even this effort left us still in the dark regarding the reasons for many things that we observed in connection with these fires. And so about a year and a half ago we undertook an extensive series of tests with a view to clearing up some of these questions. The results of these tests have contributed so much to our knowledge of the subject that I am going to write briefly about several of them.

Uncertainty as to Boil-over and When It Will Occur

Probably the element of greatest uncertainty in connection with oil tank fires that have occurred in the past has been the question of whether the burning tank would boil over, and, if so, when this would occur. Available records of tank fires seemed to be entirely conflicting on this point and so we set up the model apparatus, shown in Fig. 1. In this we provide means for determining just what goes on inside a burning tank full of oil. The tubes (sample draws) tapped into the tank at different elevations and terminating on this side of the wall enabled us to draw off samples of the oil at various stages of the fire for laboratory analysis. An electrical instrument is also shown for reading temperatures indicated by thermocouples—which are really electrical thermometers—that we have placed at various points inside the tank. Fig. 2 illustrates the equipment diagrammatically (thermocouples, sample drawoffs, gauge well, etc.).

All told we burned more than 100 test fires in tanks equipped with test apparatus as shown; and in sizes ranging from 110 gallon drums shown in the first picture to a 15’ x 5’ tank containing about 6,600 gallons.

The Cause of the Boil-over

Fig. 3 shows that we obtain typical boil-over conditions with out test apparatus. This is a fire in crude oil, the picture being taken at the moment of the boil-over.

Fig. 2. Diagrammatic Illustration of Equipment Shown in Fig. 1.

Conditions Necessary to Cause Boil-over

The hundred or more test fires that we burned showed very definitely that a boil-over is caused by water in the oil becoming heated above its boiling point, and three conditions are necessary if a boil-over is to occur from a burning tank (see Fig. 4). The first is of course that there must be some water present. A dry oil with no water in the bottom of the tank will not boil over. The second is that the oil must be of such a nature that what we have called a “heat wave” will carry heat down well ahead of the burning surface and turn the water in the bottom into steam. The third necessary condition is that the oil must be sufficiently viscuous so that the steam rising through it will form a lot of foam..

Fig. 3. Fire in 15-foot Tank Containing Crude Oil, Taken At the Moment of the Boil-over, Showing How Typical Boil-over Conditions Are Obtained with Test Apparatus.

All three of the conditions that I have mentioned—water, a heat wave, and viscosity—are practically always found in crude oils; but never, so far as we know, in refined oils.

Heat Wave Most Interesting of Conditions

The matter of the heat wave is perhaps the most interesting of these conditions (see Fig. 5). The heat is not conducted or radiated down through the oil, but must be actually conveyed down by the hot oil when its specific gravity becomes great enough to cause it to gradually sink. A wave will form only if the oil contains a wide range of boiling points so that, as the light volatile fractions burn out of the top layer, there will be heavy portions left which will sink into the cool oil below carrying a lot of heat down with them.

Fig. 5. Diagram Showing Action of the Heat Wave in Qil Tank.

The curves in Fig. 6 represent a graphical record of what went on inside of the oil tank during a typical test. The temperatures are shown vertically, and the time, in hours, shown horizontally. Each curve is the temperature record of the thermocouple with the corresponding number, the numbering being from the top of the tank downward. It will be noted that the top thermocouple was quickly raised to a high temperature, while the lower thermocouples showed no temperature change. After a little over a half an hour the heat wave reached the second thermocouple, which jumped to the temperature of the heat wave. Finally the heat wave reached the water in the bottom of the tank and, there being a viscuos oil present, a boil-over occurred.

Apparent Inconsistencies in Records of Oil Fires Cleared

The information that we obtained from these tests cleared up the apparent inconsistencies in the records of past fires. It explained why boil-overs had never occurred from refined oil tanks, and why they should not be expected to occur. As our tests had shown widely different rates at which the heat wave traveled down through different kinds and gravities of crude oil, we also had a clue as to why some boil-overs had occurred within perhaps four or five hours after the fire had started, whereas others did not take place until the tank had burned for as much as twenty-four hours.

Predict Boil-over With Accuracy

This information enabled us to predict with considerable accuracy when a boil-over would occur in a recent crude oil reservoir fire in Bakersfield, Cal. This reservoir contained when fired about six feet of heavy oil. Our records indicated that, in such oil, a heat wave might be expected to travel about one foot an hour. The fire started at 5:30 in the evening; and so we advised our superintendent in charge of operations at the scene that a boil-over might be expected around 11:30. He used all available men for throwing up emergency fire wall protection until 11:15, at which time he ordered all men back to positions of safety. The field around the reservoir was completely cleared before 11:30; and at 11:40 the boil-over occurred.

Fig. 4. What Will Prevent and What Will Produce Boilover, Shown Diagrammatically.

While we were thus able to make a quite accurate and useful prediction in the case of this heavy oil fire out test reports, and also the records of actual fires in lighter crude oils are not quite so consistent—the rate of heat wave travel in these oils ranging all the way from eighteen inches to over three feet per hour. Hence, in an effort to get some more accurate means of predicting when a boil-over might be expected from a light crude fire, we developed in the course of our tests a means of observing the advance of a heat wave on a tank by painting on the outside of the shell of this tank a stripe of paint made from a special pigment that changes color when its temperature is raised above 150 degrees. Fig. 7 shows the sharp front made by the heat wave. In this case there was some oil spilled on the outside of the tank which had been volatized where heated above the line of the “heat wave” front, but which had remained below this line.

Follow Progess of Heat Wave by Stripe of Special Paint

In the early stages of a tank fire it is generally possible to approach the windward side of the tank without discomfort. Hence when a fire starts in a tank containing crude or fuel oil we plan to apply a stripe of our temperature indicator paint by sending up to the tank a couple of men equipped with a brush on a long extension handle with which they can paint a stripe from somewhere near the top down to the ground. Then by watching the progress of the change in color of this stripe we can tell how fast the heat wave is going down; and if we know the level of the top of whatever layer of water or wet settlings there may be in the bottom of the tank, we can estimate quite closely the time at which a boil-over will occur. This information will enable us to make maximum use of men and equipment that we have available for throwing up emergency fire walls, etc.

(Continued on page 37)

Essentials in Handling Big Oil Fire

(Continued from page 10)

Thus the information gained from our boil-over tests is of distinct value to those in charge of oil fire fighting, and it has proved to be of greater benefit to us in planning safe construction.

Fig. 7. This Picture Shows Sharp Front Made by Heat Wave. Oil Spilled on Outside of Tank Volatilized Where Heated Above Line of Heat Wave, but Remained Below This Line.

Cannot Entirely Prevent Boil-overs

Having learned from our tests the reason why boil-overs occur, we next tried to figure out some way of preventing them. We could not change the physical characteristics of the oil, so the only condition which it seemed possible might prevent boilovers in crude, would be to get the water out. This we tried in every way we could think of, even building some of the test tanks with cone-shaped bottoms for easy drainage. This was successful in a few cases, but we found we could never depend on it. Crude oil frequently contains emulsions of water and oil, and the water will not easily settle out. As water expands some 1,700 times when turned to steam, it takes only a very small quantity of water in the crude to cause a boil-over. We were therefore compelled to give up the idea of completely preventing boil-overs and turn our attention to trying to make those which we can not prevent as nearly harmless as possible.

(To be continued)

Stanley, N. Y., Elects Chief—Frank Thompson has been elected chief of the fire department at Stanley, N. Y.

New Chief for Sharpsville, Pa.—At the election of the fire department at Sharpsville, Pa., John Joyce was elected chief.

Wausaukee, Wis., Purchases Chemical—The village board of Wausaukee, Wis., has purchased a Boyer chemical car.

Orangeville, Ill., Buys Fire Apparatus—A Ford truck on which was mounted chemical tanks was purchased by Orangeville, Ill.

Assistant Chief Ogdensburg, N. Y., Dead—James St. Andrews, assistant chief of the fire department at Ogdensburg, N. Y., died on December 17 at the age of sixty-five years. He was the oldest man in the department in the point of service, having served for thirty-five years.