Document created: 6 February 03
Air University Review, March-April 1977

Fuel Supplies in Time of War

ACCORDING to Napoleon, "An army marches on its stomach." Traditionally, this was true, but today's army is so mechanized that the nourishment it needs is not solid but liquid. An army needs fuel. It would be inappropriate to extend the simile to talk about an air force flying on its stomach. However, the inference is obvious: an air force requires large quantities of fuel to fight any kind of war, but in general the standards of quality control are higher than those for an army.

The logistic supply of fuel, both in rear echelons and in forward areas, will be the focus of this article. The region assumed is northwest Europe, and it is further assumed that the conditions pertain to those after the outbreak of limited war.

Forward transfer of fuel to the British army has long been achieved by means of the jerry can, a copy of that used by the German army in the North African campaign of World War II. The German models were far superior to the "flimsy" used by British forces, so they were acquired wherever possible. Each can contains 20 liters, equivalent to 4.4 British Imperial or 5.5 U.S. gallons.

A rubber container is also available from a French company; the container is shaped like a bed pillow and contains 20 liters, the same volume as a jerry can.¹ The container may be dropped from a transport aircraft in a parachute-retarded or nonretarded fashion.

A great deal of money and time has been spent in exploitation of the inherent flexibility of a system based on jerry cans (as opposed to the use of tanker vehicles), and the system has been used for about 75 percent of the total logistic fuel supply for the British army on the Continent of Europe. In addition to the work put into the development of the jerry can, many people are needed to fill them. Mechanized can fillers are available, one employing a reciprocating action which, when properly manned, fills 150 cans per hour. The rotary can filler (nicknamed "rotary cow") is designed to fill 720 cans per hour, but, in the author's experience, it achieves a filling rate of only 650 cans per hour. (See Figure 1.)

During World War II, the percentage of packed stocks of the overall logistic load moving forward was 95 percent, and in the Korean War the percentage was only slightly less. Jerry cans will probably continue to be used, and in order for the cans to be handled in bulk, efforts are being made to devise suitable pallets for them. A British example is shown in Figure 2(a) and a German version in Figure 2(b). The British one folds flat for the return journey, but the German version occupies the same space full or empty.

Thus, recent efforts have been made to handle jerry cans in bulk, but they constitute an awkward load. Although the army still uses this system to refuel armored brigades--the Royal Air Force used it in all parts of the world right through World War II--jerry cans have long since become unsuitable for aircraft fuel replenishment. In order to distribute fuel to the user, a bulk supply system must be used in the rear areas.

The supply system used must be compatible with a transport and load transfer concept based on the maximum use of standard vehicles and mechanical handling aids. There will have been opportunities for peacetime reconnaissance and pre-positioning of men and material in northwest Europe, and although much of this will have been done, the force deployed will still depend on substantial air freight capability. Additional logistic problems would arise if the strike aircraft were required to operate in widely dispersed formations and if the indigenous fuel supply system was meager.

Entry by air uses drums with a capacity of 40 Imperial gallons (50 U.S. gallons),² but a large proportion is scheduled to be delivered by the 500-gallon Sealdbin--Figure 3(a). These are expensive items which cost $1000 each in Britain ($700 in U.S.) and weigh 250 pounds when collapsed and empty. They can be dropped by parachute, towed, rolled, bounced, floated, and filled from roadside tankers--Figures 3(b) and 3(c). Sealdbins are easy to handle and can be used for direct refueling, but they have a short life, an average of 15 cycles of use.

Entry by sea can be accomplished by means of an 8-inch-diameter ship-to-shore submarine hose. The working pressure is 200 pounds per square inch (psi), and a throughput rate of 230 tons per hour is possible. An alternative method is to use the towed flexible barge or dracone. This is a large tube made from nylon-reinforced neoprene rubber 225 feet long by 9 feet wide and 6 feet deep. It is filled and emptied through a 4-inch armored hose fitted to the bow of the vessel, and pumping can take place when the ship is stationary or under way. Each dracone carries 105,000 gallons when full. (See Figure 4.) Except for biannual pumping exercises in the United Kingdom, little information is available, and wartime experience by the British is nonexistent. Experience gained by the U.S. with limited use of dracones in Vietnam shows that they are not very satisfactory from the point of view of hazard to other shipping and users of shore facilities. ³

When the fuel reaches shore, it can be stored in bulk bolted steel tanks of various capacities from 15,000 to 125,000 gallons, as shown in Figures 5(a), 5(b). However, these tanks are falling out of favor since they take a large amount of sapper support, and fabric tanks seem now to be the norm. In the British system, these tanks, fabricated from nylon-cord-reinforced neoprene rubber, come in two sizes, 40,000 and 12,500 gallons, the latter shown in Figure 6(a), and can be interconnected to form tank farms. In the U.S. system, in addition to tanks having similar capacities, they are available with capacities of 10,000 barrels and 25,000 barrels--Figure 7(a). It may be noted that these tanks and their positioning require some ground preparation, if they are not to cause trouble when filled. An indication of the degree of preparation required for such so-called berm-mounted tanks is shown in Figures 7(b) and 7(c).

The complex of tanks on shore may be connected to the airheads by means of a 6-inch victualic pipeline* as shown in the overview drawing, page 54. This pipeline comprises aluminum alloy pipes, each 20 feet in length and weighing 80 pounds as compared with a poorly designed 200-pound steel equivalent. Helicopter movement of pallet loads is possible, as shown in Figure 8(a), but carriage of half a mile's worth on a 4½ ft ton truck is more usual. Figure 8(b) shows the pipeline being laid, transported by this means; a load of such pipes is shown in Figure 8(c). The line works at 400 psi and at a flow rate of 540 gallons per minute. However, it is economical to establish such pipelines only when the throughput exceeds 1250 barrels per day. The pumping is accomplished by means of two of the pumps of a three-pump station used in series. These stations, which contain their own water separators, filters, and quality control point, occur at intervals of 12 miles in the pipeline.

*The term victualic pipeline derives from victualic coupling, in which a groove (rather than threading) is cut at the end of a pipe and the pipe ends are joined by a rubber ring around the joint; a sleeve is placed about the ring and bolted, compressing the ring and making a strong, liquid-tight joint.

The U.S. system will shortly incorporate a main line pump driven by a gas turbine having speeds between 5800 and 10,000 revolutions per minute. The turbine drives the pump through a reduction gearbox with 1080 horsepower output under standard conditions, with an operating pressure of 1000 psi. Three of these units are capable of pumping more than 37,000 barrels per day of finished product 90 miles through an 8-inch pipeline on level terrain. This system should handle most theater requirements. The pump is shown in Figure 8(d).

Other means for conveying fuel from the rear echelons to the forward areas are those associated with the highly developed transportation facilities in northwest Europe: rail, truck, and barge. The main problem with all these conveyances is that the individual containers have limited capacity, require prepared media on which to travel, and manpower requirements are generally high. The supply is also discontinuous and could be sporadic or have large gaps in times of war.

When the fuel has been moved forward and stored in large, collapsible, neoprene rubber, 50-gallon or 500-gallon tanks, it is then pumped to the aircraft by means of portable pumps, having filters and separators integral with them. The equipment used for this is readily air-portable and may be carried and operated by two men.

The British army is scheduled to use high mobility load carriers, Figures 9(a) and 9(b). These four and a half ton vehicles should replace the medium mobility standard load carriers and may have installed on them two 600-gallon aluminum tanks from which fuel is pumped through attached separators and filters by means of a centrifugal pump. These devices would be eminently suitable for replenishing fuel for helicopters and all V/STOL aircraft.

The basis of the British Army of the Rhine supply system is the NATO pipeline network. This is a European grid with main line pipes of 8-, 10-, and 12-inch diameters; laterals of 6 and 8 inches, connectors of 4 and 6 inches, and with port, canal, and road junction take off points. The line is supplied by tanker ships discharging into storage at major port facilities and operates at high pressures, about 1200 psi. The flow rate through the pipeline is greater than 500 gallons per minute. Because three types have to be processed, namely CIVGAS, DIESO, and AVTUR (equivalent to MOGAS, Diesel Fuel, and JP 4), multi-product working is used. In this, one product follows another down the line with the region of mixed fuels being pumped away into a slops tank for later separation. Supply of large quantities of any fuel to any part of the network is thus possible.

Protection of pipelines. Most of the NATO pipeline is in a buried, protected environment and, thus, requires little manning for patrolling. Any temporary pipeline that is joined with the NATO pipeline would, however, need patrolling to protect the line against accidental damage and that due to surface action. It should be possible, though, to combine the running of the pumps with such duties, and only in extreme emergency would the manpower required exceed that used for conveying fuel by tanker vehicles. Also, the use of redundant or alternative lines could obviate some of these difficulties. 

With the advent of lightweight materials for pipelines, such as new metal alloys, polymers, and metal/plastic composites, and of pumps with high power-to-weight ratios, it is possible to foresee greater use being made of pipelines in spite of the problem associated with damage. It should be possible to position these pipes and pumps rapidly so that they would link together the larger stocks concentrated in rearward areas with smaller tank farms at airheads. This method has the advantage that the pipelines can go across terrain not always negotiable by road. It may also become feasible to use similar lightweight equipment on high mobility tankers operating in forward areas and to fuel aircraft directly, even though they may be located a hundred feet or so from the tanker itself. This modified supply system would be more flexible than that envisaged for the near future and would be able to keep pace with the fluidity of modern warfare.

For the first role, fuel pumping rates should be of the order of 540 gallons per minute in each line, and for the second a flow rate of about 30 gallons per minute in each line is required. It is contended that most of the needs of the air force in the NATO area could be met with these two sets of conditions, and the design parameters involved in this procedure are dealt with in the succeeding part of this article.

Pressure loss and power loss in pipelines. It is possible to derive the relationship between the pressure loss for a standard length of pipe (taken here to be one statute mile) and the flow rate in that pipe. To be more specific, for JP 4 the pressure loss, H, per mile is given by

H = 95.3f Q² psi
            D⁵

where Q is the flow rate (gallon/minute), D is the pipe diameter, and f is the value of friction coefficient obtained by the Colebrook-White formula ⁴ using an absolute pipe wall roughness of 0.0018 inches and allowing a contingency factor of 5 percent for joints and lack of knowledge of roughness. Values of pressure loss per mile are shown in Figures 10 and 12(a), 12(b). Figures l1(a) and 11(b) show the maximum possible pipe run for different values of pressures and flow rates, and it can be seen that adoption of small bore pipes necessitates pumping at high pressures in order to maintain the flow rate at a distance.

The power required for pumping is calculated by multiplying the pressure loss along the pipe by the flow rate. Thus

horsepower = 0.583 pQ

where p is the pressure loss along the pipe given in thousands of pounds per square inch, and Q is given in gallons per minute. The variation of power loss with flow rate is given in Figures 12(a) and 12(b).

Pipe wall thickness. The wall thickness, t, is derived from considerations of pipe hoop stress. Thus

hoop stress = _P D
                         2t

It is customary in a particular application to permit the walls to be stressed to a fraction of the ultimate tensile stress (UTS) of the material. The wall thickness and density of material assumed, coupled with the pipe diameter, will then lead to the weight of pipe per unit length.

In the case of the British Army's Emergency Fuel Handling Equipment (Surface) system, where the pressure loss per mile is 34 psi, the external diameter is 6.65 inches, and the UTS of the material used is 63,000 psi. In this case a maximum pressure of 1200 psi is assumed, and this leads to the law following for the empirical thickness of pipe wall using any material.

t (inches) =17.7 D (inches) pressure loss (psi/mile)
                                     UTS of tube material (psi)

For pipes of 4- and 6-inches diameters pumping 540 gallons per minute at pressures of 400 psi, the wall thickness and pipe weight per unit length are shown in Table 1.

Weight of pumping equipment. Table 2 gives some of the relevant characteristics of power installations currently used by the British army. The present generation of pumping equipment shows an average installed weight of about 125 pounds per horse-power. A new generation of gas turbine pumps now under development could lower this figure to about 35 to 40 pounds per horsepower. The power needs envisaged have thus been calculated and the values shown in Table 3.

Pressure loss. The calculated results shown in Figure 10 are quite startling and enable us to reach two very significant conclusions. First, when pipe diameter is very large, the head losses are low, and a substantial flow rate can be maintained over considerable distances by comparatively small power inputs.

At 540 gallons per minute a 10-inch pipe loses about 6 psi per mile, so if pumped at 180 psi, it would need boosting every 30 miles. A 6-inch pipe loses 40 psi per mile, and if pumped at 240 psi, it would need booster stations at 6-mile intervals, or at a pressure of 400 psi boosting every 10 miles. On the other hand, if a smaller pipe diameter is chosen, say 4 inches, the head losses become very large, about 300 psi per mile, and high flow rates can be achieved only by working at much higher pressures. If we should choose 1200 psi, this would require the installation of booster pumps in the line at about every 4 miles. A flow rate of 540 gallons per minute is equivalent to about 900 tons a working day, sufficient for a number of squadrons of aircraft and a large helicopter fleet. If we wish to replenish an airhead tank farm from a large-capacity store or pipehead at daily flow rates of this order, then it is useless to consider pipe diameters of less than 4 inches.

The region of the graph concerned with the small pipe diameters is enlarged in the inset of Figure 10. It is of interest because such pipes could be used in forward areas to replenish Harriers or similar V/STOL airplanes while it hides some distance from the supplying tanker. For each aircraft a flow rate of at least 30 gallons a minute would be necessary, but the graph shows that head loss increases by a factor of more than 10 for each half-inch reduction in pipe bore, and despite the handling advantages of small-diameter pipes, it would seem that we could not afford to opt for a pipe size below one inch.

With regard to the distance pumped along given pipes at given pressures, it can be seen from Figure 11 (a) that pumping through a half-inch line at a pressure of 1200 psi would produce 24 gallons a minute at 240 feet but only 12 gallons a minute at 900 feet. The situation is better at a l-inch bore, shown in Figure l1(b): 300 psi would maintain 24 gallons a minute out to 1500 feet if necessary.

Fluid horsepower. It can be seen that power is a function of flow rate and pressure loss, and the power has been plotted in Figures 12(a) and 12(b). For completeness, the pressure loss per unit length has been plotted on the same figures for some pipe diameters in dashed lines. For a 10-inch pipe the power needed to convey 540 gallons a minute is some 1.07 horsepower per mile (not shown on the curves), and for 6-inch diameter pipes 12.6 hp is needed. If the diameter is reduced to 4 inches, however, the power needed rises to 92 hp per mile, and to 362 hp per mile for a 3-inch diameter pipe.

At the smaller diameters, the variation of power required as diameter is changed is even more staggering, despite the lower flow rates. At 24 gallons per minute, 1.54 hp per mile are required to pump through a 1½ inch pipe, 11.9 hp for 1 inch, and 357 hp for the half-inch size. It is more realistic to relate these figures to the probable battlefield situation of refueling Harriers and helicopters by quoting them as power requirements per 300 feet-run of pipe: 0.08, 0.70, and 20.3 hp respectively. The first two could be achieved from a vehicle power take-off unit, or a small portable pump, but the power needs of the half-inch diameter pipe are excessive, and its use can be quickly dismissed.

Pipeline materials. It can be seen from Table 1 that selection of the less-strong materials for the pipeline, such as plastics and fiber-reinforced rubber, would result in a very large ratio of wall thickness to bore and, hence, a stiff, unmanageable line. To make the line thinner and more flexible, it would be necessary to employ much lower pressures and, hence, reduced flow rates. The alternative is to use metal pipes; they would be thinner and, therefore, lighter, but they would be more cumbersome to handle.

THERE IS a firm impression that the logistic chain of the British army and the Royal Air Force has worked well in spite of equipment and systems that are outmoded and ripe for revision. The fact that armored brigade, helicopter detachment, and airfield commanders have always received a rapid response to their call for fuel has led to the thought that revision of existing techniques and of plant design is not necessary. This is a fallacious assumption, and, in my view, some thinking will have to take place in the near future if helicopter and Harrier refueling is to take place expeditiously.

It is also a fact that most systems in use by Western forces postulate that future conflicts will be long in duration. The implication of shorter duration of hostilities, together with the use of minimum or no sapper support, should be investigated.

An argument has been put forward in this document which demonstrates that a reasonable solution to the problem of conveying fuel forward with the above considerations is by the extended use of pipelines. In the quantitative survey, the main factors involved in making use of such pipelines have been examined. Two appropriate applications are those of moving some 900 tons a day at 540 gallons per minute over distances around 30 miles and of replenishing helicopter or V/STOL aircraft when they are located some 300 feet from a high mobility tanker vehicle.

In the first application, the relative merits of pipe diameter as small as 4 inches and as large as 10 inches have been considered. For the smaller gauge 4-inch pipe, 540 gallons per minute can be achieved only by pumping at high pressure, say 1200 psi through a high-strength, rigid metal pipe, but the large head losses which result can only be made good by installing every 4 miles a 368 horsepower pump weighing 14,750 pounds. As diameter increases, so the advantages conferred by lower pressures and lower power requirements can be realized, the pipes can be thin-walled, lightweight, and collapsible. At a diameter of 6 inches, 540 gallons per minute can be achieved at a pressure of 400 psi, and collapsible fabric lines are feasible. Pumps rated at 126 horsepower, weighing 5050 pounds, and placed every 10 miles should suffice to maintain the flow. For a 10-inch fabric line, the figures are even better: 180 psi pressure, maintained by 32.0 horsepower pump, weighing 1280 pounds, placed every 30 miles.

To refuel helicopter or V/STOL aircraft directly from a high mobility tanker over a distance of 300 feet, say 4 at a time each at 30 gallons per minute, lines of at least 1-inch diameter could be used, operating at pressures of the order of 200 psi. The pumping power needed would be approximately 14 horsepower, well within the capacity of a power take-off from a stationary fuel carrier. A number of related facts have been excluded from the analysis, but it is hoped that those included here are sufficient to give some food for thought on the problem in the appropriate quarter.

Ames, Iowa

Notes

2. In this paper when gallons are cited, U.S. gallons are intended unless otherwise indicated.

4. C. F. Colebrook and C. M. White, "The Reduction of Carrying Capacity of Pipes with Age," Journal of the Institute of Civil Engineers (London) vol. 7, 1937, p. 99.

Acknowledgment

Preparation of this article was conducted partially under the auspices of the Engineering Research Institute of Iowa State University at Ames, Iowa. This support is gratefully acknowledged.

Contributor

Dr. Arthur Akers (Ph.D., University of London) is an Assistant Professor in the Department of Engineering Science and Mechanics and Engineering Research Institute, Iowa State University, Ames. For ten years, he worked for Her Majesty's Government in England, first at the Royal Naval College, Greenwich, and then at the Royal Military College of Science, Shrivenham. He held a Queen's commission and was for some years a serving officer in the British army reserve, specializing in problems of fuel supply to the battlefield.

Air & Space Power Home Page | Feedback? Email the Editor