About a year ago the author was unexpectedly placed in charge of the cargo oil of a Navy fuel ship. Never having had any engineering experience, the author approached the subject of the measurement of fuel oil with an unbiased if not a blank mind. It soon became apparent that although there were ample instructions covering the physical changes of oil due to temperature there was no literature on the subject of the practical measurement of the oil—or if there were, its existence was unknown to most of the service and to the large commercial companies.
One method of measuring the amount of oil delivered to a vessel is by means of an oil meter placed in the discharge line. However oil meters are not at present supplied to navy fuel ships and are not used by commercial companies to the author’s knowledge. The general opinion among destroyer engineer officers seems to be that meters are unreliable. This is based on the fact that it is difficult to tell when the meter is out of order and (possibly) because the meter is influenced by the temperature of the oil and consequently its reading should be corrected for the mean temperature of the oil during the fueling. This mean temperature would be very difficult to determine as it would not be the average of the temperatures before and after the pumps were used.
The usual method, which is used by both the Navy and commercial companies, is to compute the amount of oil in the tanks before and after fueling. The difference is, of course, the amount of oil discharged or received. The quantity of oil each time is obtained by taking soundings or ullages of the tanks and entering tables made out for each tank which show the amount of oil in the tank for any given sounding or ullage. A temperature correction must then be applied to the figures to reduce the oil to the volume that it would occupy at sixty degrees Fahrenheit. The Navy uses the formula given in chapter 38 of the Engineering Instructions while commercial companies generally use the tables in the Bureau of Standards Circular No. 154. These two formulae give fairly similar results for small quantities of oil but different results for large quantities.
The method may be put in mathematical form using the formula for temperature correction as given in the Engineering Instructions:
Let: A = Gals, of oil at temperature t as shown by soundings or by ullages.
A'= Gals, of oil sixty degrees.
A"= Gals, of oil received or discharged.
t = Temperature of oil at time that soundings were taken.
Then: A'= A - .0004A(t-60) (1)
= 1.024A - .0004At (2)
A"=A'1-A'2 (3)
= 1.024(A1 – A2) - .0004(A1t1-A2t2) (4)
Formulae 1 and 3 are the most convenient to use in actual practice as the subtractions involved are made more rapidly than the multiplications involved in the other formula.
It seems needless to state that the temperature of the oil must be taken at the same time that the soundings or ullages are taken. However the author had one prominent commercial company propose using the temperature of the oil at the time that the pumps were stopped instead of the temperatures at the time that the ullages were taken. The oil had been heated by steam- to facilitate pumping, and as the ullages were taken about four hours after the steam was turned off the temperature had fallen about twenty degrees since the time the valves were closed and the pump stopped. The temperature correction was to be applied to a little over 2,000,000 gallons of oil and the twenty-degree error in temperature would have resulted in a large loss, a loss which incidentally would not have been in the company’s favor.
Furthermore, it seems needless to state that the temperature applies to the volume of oil in the tanks from which it is obtained, and not to the volume of oil given to or received from the vessel alongside. For example, if there are about 2,000,000 gallons of oil on the tanker and the destroyer alongside receives only 60,000 gallons, the temperatures of the oil as taken from the tanker’s tanks are applied to the 2,000,000 gallons and not to the 60,000 gallons. This is obvious from formula 1 and 2. The author has, however, had innumerable destroyer engineers tell him that they did not see the necessity of taking the temperatures on the tanker both before and after oiling as the temperatures could not change over a couple of degrees and this would make no difference when applied to the small amount of oil that they were taking.
A short cut in applying the temperature correction, which is often suggested, is to apply a correction, based on the mean of the temperatures before and after fueling, to the volume of oil pumped as shown by the change in sounding or ullages. This method is not exact as may be seen when it is put as a formula:
A"= (A-A)-.0004 ( t-t/2 – 60) (A-A)
This formula cannot be made to correspond to formula 3 or 4 by any method of simplification or transposition; and the amount of error which will result in its use is dependent upon the magnitude of the temperatures and volumes of oil involved. A large commercial company once suggested using this incorrect method on a 2,000,000-gallon cargo, apparently, because they did not know how to handle the temperature correction when the temperatures before and after fueling were not the same.
If formulae 1 and 3 are correct and if the tables for converting soundings or ullages are correct, the accurate measurement of oil depends entirely upon the accuracy with which the soundings or ullages and the temperatures are taken. No mathematics can be more accurate than the data on which they are based. The Bureau of Standards no doubt had this in mind when they stated that the assumed weight of a gallon of water is 8.32828 pounds.
In measuring oil, a sounding is the distance that the surface of the oil is above the bottom of the tank, and an ullage is the distance that the surface is below the top of the tank. Destroyers generally take soundings of their tanks, and fuel ships and commercial companies generally take ullages. The author has had no experience in taking soundings and will consequently confine himself to the methods of taking accurate ullages. Some of the precautions necessary in taking ullages apply equally as well to soundings.
Before taking ullages or soundings, it is important to make sure that there is the same amount of oil in the pipe lines before fueling as will be there after fueling. Shore stations generally keep all their lines full of oil as the only way for them to clear the lines is to pump it back up into the tanks. Commercial tankers and navy fuel ships generally find it more convenient to keep the suction lines to their pumps full and the discharge lines empty. This can be assured by opening the pump gate valves, the suction valves of one tank and the discharge valves to one or more tanks. As the discharge lines are above the level of the oil in the tanks, this lets the oil run out of the discharge lines through the pump (which is the rotary type) and back to the tank via the suction line. If the suction lines are empty the oil will flow by gravity from the tank into the lines when the valves enumerated above are opened as the suction lines are all below the oil level in the tanks. To overlook these precautions may mean an error of several thousand gallons of oil.
It is essential that the suction lines be opened to only one tank. If two or more tanks are opened to the suction lines and if there is a difference in oil level in these tanks, the oil will be flowing by gravity from one to another while the soundings are being taken. Of course any soundings or ullages taken under these conditions will be incorrect.
Vessels receiving or discharging cargo are subject to changes in list and trim. Consequently the question comes up as to what effect these changes will have on the accuracy of the soundings or ullages. If the opposite sides of the tank are parallel, if the place where the soundings are taken is midway between the forward and after bulkheads, and if the top of the tank remains above and the bottom below the surface of the oil, the soundings will not be affected by a change of trim. Under the same restrictions, changes of list will not affect the soundings if the place from which they are taken is midway between the inboard and outboard bulkheads. Also ullages will not be affected under these conditions provided that the changes of list and trim do not exceed four or five degrees. Commercial tankers generally have the openings in the center of the tanks. Navy fuel ships have them correctly placed fore and aft, but very close to the inboard bulkhead. This gives correct ullages regardless of changes of trim, but leaves a possibility of error in the case of a change of list. However, the navy fuel ships have tanks arranged in pairs, port and starboard, the capacities of a pair of tanks are very nearly equal, and their sides are approximately parallel to each other. Consequently the total amount of oil in the two tanks, as shown by the soundings or ullages, will not be affected by a change of list provided that both tanks have such a quantity of oil in them that the oil does not expose their bottoms or touch their tops. For example, if the ship takes a half a degree list to starboard, the ullage in the starboard tank may be a quarter of an inch too large; but the ullage in the port tank will be a quarter of an inch too small, and the two errors will cancel.
In taking ullages, a steel tape is used with a bob on its end. It is most convenient to use a brass bob one foot long, and graduated in quarters of an inch from the top down (i. e. the zero mark on the bob is next to the zero mark on the tape). The tape is held against the ullage mark at the top of the tank, and the bob lowered until it reaches the oil, and then it is further lowered until the next foot mark on the tape is against the ullage mark. The bob is then withdrawn and the distance of its oil mark from its top is added to the number of feet that the tape was lowered.
The bob should be lowered with no swinging. If the bob is swinging with a circular motion, too large a reading may result.
All movements of the tape before the reading is taken must be down; and all movements of the tape after the reading is taken must be up. If this is not done, the ullage taken will be too small. The reason for this is that the oil mark on the bob shows only the minimum distance that the bob’s zero mark was above the surface of the oil; and therefore it is essential that the tape be read when the bob is in its lowest position.
If the oil mark on the bob is found to be biased, the bob was probably resting on the bottom of the tank or on a submerged frame and the ullage will be too large. In this case another ullage should be taken using a few inches less of tape.
The ideal thermometer for taking the temperature of the tanks is one with graduations to every degree on a large copper scale, and with a bulb enclosed in an open cup. The standard navy issue is not satisfactory because the graduations are generally for every two degrees, the narrow painted scale is very hard to clear of the oil, and the thermometer must be lashed in a paint pot or improvised bucket to prevent the temperature of the air from affecting its reading while it is being drawn out of the oil and while the scale is being cleaned.
Most commercial companies take the temperatures of all their tanks and apply a temperature correction to the volume of oil in each tank, although some are content to take the temperature of only one tank and to assume that it is the same for all the others. After the laborious process of taking the temperatures of ten tanks (all of which happened to be full) on a commercial tanker, the author was convinced that there was practically no variation in temperatures of the different tanks under normal conditions. However, on the suggestion of Lieutenant G. W. Clark, U. S. Navy, temperatures were taken in a full tank on the sunny side of the ship and a half empty tank on the shady side of the ship. The results were such that it seemed probable that the temperature of no tank was uniform throughout its depth. This was investigated; and it was found that the temperatures at the bottoms of the tanks were that of the sea water while the temperatures at the surfaces of the oil were apparently determined by the weather conditions acting upon the above water portions of the tanks and by the distance that the surface of the oil was below the top of the tank.
Theoretically this is what might have been expected in a climate where the air is warmer than the water. Oil is a poor conductor of heat; and as the heat is applied at the top of the tank, no convecting currents are formed. Whether or not the temperature is uniform in a climate where the water is warmer than the air or in a tank which is heated by steam, the author has had no opportunity to determine. Also the author has been unable to determine if the temperature varies with the horizontal distance from the side of the ship although this seems to be probable. Any one desiring to investigate the temperatures in a tank can easily do so by lowering a maximum and minimum thermometer, such as is used in a magazine, to various depths in the oil.
From the above it seems probable that the best method of applying temperature corrections to the oil is to apply a separate correction to each tank based on the mean of the surface temperature of the oil and the temperature of the sea water.
It is interesting to note that there is a pneumatic instrument installed on most large ships, whereby the hydrostatic pressure at the bottom of the tank is shown on a column of mercury. There is a scale at the side of the mercury column which shows the height of the oil in the tank. This instrument has the advantages that the soundings can be taken very rapidly, that they can be taken fairly accurately when the ship is rolling (which cannot be done with a tape), and that if the scale is made out correctly no temperature correction would have to be applied. However, it has two fatal defects which are that it is a delicate instrument and that the manufacturing company does not furnish sufficiently detailed drawings to enable it to be overhauled by a ship’s force.
In conclusion the following odds and ends are offered. The measurement of oil is not an exact science, but the average error in fueling a small ship should not be over 600 gallons. Most commercial companies expect more or less large losses. One company was protected by Lloyd’s; but as their tanker had no tables to convert their soundings into gallons, they and the insurance agent could only take the Navy’s figures on the amount of oil transferred.