It is desired to present for consideration the feasibility of the use of water motors on board ship for the purpose of driving the dynamos and also the ventilating fans. An outline of the proposed scheme is as follows: That there should be duplicate dynamos, of a commercial high speed type, designed for running at about the commercial speed when belt driven; these dynamos to be direct coupled to the shafts of water motors of the most approved and economical type in the market, and these motors to be actuated by water from the steam fire pumps, also in duplicate; the waste water to be pumped overboard by a pump in the dynamo room, or pumped into the flushing system with an overflow, or to be returned through a return pipe to the pump in the fire room. In the following comparison the last of these three methods will be chosen to illustrate the working of the system.
In connection with the above outline the points to be considered are: 1st. Original cost; 2d. Expense of maintenance; 3d. Advantages and disadvantages.
First—Original Cost.
With reference to the expense of installation, the only considerations are those involving changes; the wiring, fixtures, etc., remaining the same, the cost of the generating set only is to be considered. The cost of the piping will be practically the same, for while it will take slightly larger pipes for the water than for the steam, yet there will be saved the cost of the separator and reducing valve, which together will easily cover the additional cost of the increased size of the pipes. The articles that will be dispensed with by using the water system will be the dynamo room electric motor and fan, the separator and the reducing valve. It will be necessary to increase the fire pumps in size over what they would naturally be for the ordinary service for which they are intended, and in the table at the end of this paper, under the head Steam Pump, is meant the increase in size of pump necessitated. By reference to the table just quoted it will be seen that the cost of the complete generating set of the steam system is $6500, while that of the water system is but $1380, showing a saving of $5120 on each of the generating sets of the ship. The prices given in the table were obtained as quotations direct from the manufacturers and are correct.
Second—Cost of Maintenance.
This head is subdivided into Efficiency, Attendance, Repairs, and Lubricants.
Efficiency.—The specifications of the Bureau of Equipment require for the different sizes of the generating sets efficiencies varying from 76% with the 8 kilo-watt set to 82% with the 32 K.W. set, at full load. It will be seen by the table already referred to that the efficiency of the steam plant at full load is greater than that of the water system ; but at half load, which is about what is carried nineteen out of the twenty-four hours, the efficiencies are about equal, or slightly in favor of the water plant. The above comparison is made allowing that the efficiency of the steam engine remains as high as when tested on shore before installation, which I have no hesitancy in saying—and experience has corroborated my statement—it does not. Further, the use of a slow speed pump admits of the installation of a compound steam end, which will result in a saving of at least 20% (pump makers claim 30%) in fuel, which, after all, is the only true basis of comparison. The water motor and dynamo will be as efficient at the end of a cruise as they were the day they were installed, and the fire pump being a slow running machine, with very little friction and very few parts, and being driven at less than half its maximum speed, should have very little wear, so that the efficiency of the entire plant should remain nearly as high during the entire cruise as it was when first tested. The details of the computation of the data in the table referred to and the data of the tests made of the Pelton Water Motor are given in an appendix to this paper.
Attendance.—At present we have at least three, and generally four, gunner's mates to attend the dynamos and engines on board ship. On a vessel carrying duplicate 24 K.W. sets, there would be two gunner's mates, first class, at a pay of $40 per month each, and two gunner's mates, second class, at a pay of $35 per month each, or a total cost of attendance of $150 per month. By the introduction of the water system, the expert attendance needed only by the high speed engine would be unnecessary, and one of these gunner's mates, first class, and both of the gunner's mates, second class, could be replaced by apprentices, first class, at a pay of $21 per month. This would leave the one gunner's mate, first class, in charge and to attend to the repairs to the wirings, search lights, fixtures, etc., and the three apprentices would stand watch in the dynamo room. This would effect a saving of $47 per month, which represents about 25% of the coal bill of the dynamos. Further, it would help in giving electrical instruction to all the seaman apprentices, first class, of the ship.
Repairs.—The experience of the users of the water motors of the Pelton class shows that absolutely no repairs are necessary, while the repairs on the pumps will probably be no greater than those arising from the wear and tear on the flushing and fire pumps which they replace. 'The pumps would, of course, be designed for the work to be done, and should have very little repairs; and it is well known that pumps if run much below their maximum capacity will keep up their efficiency for years of steady running, with an occasional renewing of the packing or the refitting of the valves. On the contrary, the marine dynamo engine is constantly being repaired and its parts renewed at considerable cost. In all the reports from the Bureau of Equipment on the subject of "Electricity on Shipboard," the breakages of the engines take precedence in the list of accidents to which the generating set is liable. In doing away with the high speed engine and substituting a machine that is practically indestructible in use and never needs repairs, we do much towards making the plant perfectly reliable.
Lubricants—The present method of running the dynamo engines without oil in the steam spaces reduces the consumption of oil very considerably, but it is very questionable whether it does not do so at the expense of the engines themselves. I am fully aware that engines can be and are so run, and that the interior of the cylinder gets a beautiful polish on it, and shows no sign of wear after long use; but that is the case on shore where the engines are fixed to firm foundations, get the best of care by trained men, and are not tossing about, frequently at an angle of 30°. But even without the use of oil in the steam spaces the marine dynamo engine has a dozen oil cups where the water motor has two, and without having any data at hand that will positively show what the saving in oil would be, it is a very safe estimate to say that there would be required less than one-half the amount of lubricants in the dynamo room that there is at present.
Summing up under the head of cost of maintenance, we see that the efficiency of the proposed system would be at least equal to the present, that the saving in attendance, repairs and lubricants is marked, and that therefore the cost of maintenance of a plant on the water system would be less than on the steam system at present in use.
Third—Advantages and Disadvantages.
The evident additional advantages of the water system over the steam system are as follows:
1st. Ordinary temperatures in the dynamo room.
2d. Absence of dirt and oil in the dynamo room.
3d. Absence of thumping of engines in the dynamo room,
4th, Great reduction in weight (see table).
5th. Reduction in the space required for dynamo room.
6th. No need of carrying spare parts.
7th. No lagging of pipes nor expansion joints.
8th. Possibility of replacing dynamo or motor in any market.
9th. No danger to human life due to defective or ruptured pipes.
10th. Possibility of coupling either dynamo to either motor, or both dynamos to the same motor, or both motors to either dynamo, and, if desired to run both dynamos at exactly the same speed, both motors and both dynamos could be all coupled together.
These advantages require no explanation nor data to show that they would exist; they are self-evident, and the first-mentioned, that of absence of all unusual heat in the dynamo room, and therefore on the berth deck or other living spaces or store rooms in the neighborhood, is of enough importance in itself to merit an investigation into the claims made for this system.
In connection with this subject there are two points not yet touched on: that of governing and that of reserve power. Of the former I know that the Tuerk or Syracuse Water Motor will govern as efficiently as the present steam engine governor; but the efficiency of this motor is so low that it would cost too much coal to run it, and, further, it has a more delicate and intricate construction, one that might get out of order, and would require a certain amount of attendance. The testimonials published by the Pelton Water Wheel Co. go to show that they also have a perfectly reliable governor. It is unfortunate that I have not as yet had an opportunity of testing the Pelton governor to see if it would meet the requirements of the Bureau of Equipment for governing. The small governor made by this firm will meet all demands on it from full load to no load, if the load is not varied by too large a quantity; but it would be a very simple matter to have a test of the Pelton differential governor made, so as to see just how well it would meet the demands. If their governor meets the claims of the Pelton Company as well as their wheel does, it will be all that would be necessary. At the same time I am inclined to believe that a much simpler governor could be designed, something on the general plan of the Willans steam governor, and I have sketches of two designs for this purpose: one controlling the valve motion by electricity, and one by the speed of the motor shaft. If it was found that the electrically controlled governor would work satisfactorily, it would admit of the discarding of the hand regulator in the shunt, and then the only duty of the attendant would be to fill the three or four oil cups on the generating set and see that the dynamo brushes were properly set. Either of the designs of governor just mentioned would require no power to run them except at the moment of use, with the exception that the shunt coil of the electrical governor would take probably 20 watts to keep it excited, which amount of power would be practically nothing.
Reserve Power.—A water motor has a wider range of power, at a high efficiency, than any other class of engine. It is simply a question of the amount of water that can be provided for it. With the same Pelton wheel I have obtained i6-horse power at an efficiency of 86%, and ¾-horse power at 82% efficiency. Where is there a steam engine that can equal that performance? This same motor is capable of developing at least 60-horse power at the same high efficiency. It therefore becomes a matter of how much water the pumps can provide. In the pumps selected in this comparison, either pump, without running at its maximum speed, is large enough to run both motors with power enough to drive both dynamos to their full capacity. If, therefore, either pump was to break down, the other would be equal to the emergency in the very rare case of the necessity of running both dynamos.
Another advantage not mentioned is the possibility of starting up the generating set in a few seconds. With the steam set, if anything happens to the dynamo or engine it becomes necessary to warm up the other engine before it can be started, which requires at least five minutes, during which time the entire ship would be in darkness, which might delay the supply of ammunition or the training of the guns at a most critical time, or cause great confusion in the engine room.
The Ventilating System.
It is generally conceded that the present system of ventilation is wrong. We have the tremendous air ducts running through the ship, occupying berthing or coal space and, in all probability, ruining the water-tight bulkhead system, and the engines driving the ventilating fans of this system use steam, thereby heating the living spaces and store rooms by their admission and exhaust pipes. The fans must be run on the entire system in case an improvement in the ventilation is desired at any one point, and as the air inlets in each store room will be left open, the unnecessary work done is great. It has therefore been advocated by many officers of the service that between every two water-tight transverse bulkheads a separate blower be established, and that these blowers should be run by electricity. This would be a great improvement, but it would be still better to run the blowers by water motors. By the use of the latter there would be a number of advantages: first, a saving in coal, as the small water motor is more efficient than the small electric motor; second, a saving in cost, as an electric motor having the same power as a water motor would cost six or eight times as much; third, a saving in weight, as the same relation would hold as in their cost; fourth, the water motor would need absolutely no attendance, merely filling the oil cups once every four hours would be all the attendance necessary. With the electric motor, theoretically, no attendance is needed, but it is very liable to be damaged unless entirely encased, in which event it is very likely to heat excessively; and while the brushes, if once properly set, should require no further attention, still if a little dirt should get on the commutator, and it was not removed, it would be a matter of a very short time before the entire surface would be ruined, requiring turning down in the lathe. In the event of a hot bearing of the armature it will either slow down and bum out, or the fuze will blow, which is not desirable, as it would be at least a 20 ampere fuze, and that in itself would be liable to set fire to some of the woodwork in the neighborhood. In case of a hot bearing with a water motor, an event much more rare than with an electric motor, the machine would slow down and no damage would be done. The only advantage the electric motor has over the water motor is the ease of conveying the power to it,—wires in the one case and pipes in the other.
I have given an outline of the system of water motors that could be used on board ship, and the more the matter is looked into and discussed, the more apparent the advantages become. The water motor could be applied directly to the anchor hoist, saving a large amount in cost and weight; and, as in this event it would be located above the water-line, the waste water would run overboard, so there would be the saving of one pipe in case the exhaust is connected to the condenser, which it usually is. Here again comes the saving of undue heat in the living spaces, as every one knows what a nuisance an anchor engine is, on account of the heat given off by it for hours after it has been used. I think, however, that the advantages of the use of a water motor are nowhere so apparent and real as in running the dynamos and ventilating fans, even if the present system of large fans is continued.
APPENDIX.
Comparison of Efficiencies of the Steam and Water Systems for Driving Dynamos.
The size of the generating set used as an illustration will be the 24 K. W. set of the present type, compared with a similar-sized dynamo driven by a 24-in. Pelton water wheel, using water at 150 pounds pressure and making about 700 revolutions.
The Dynamos.
The commercial efficiency of the present naval multi-polar dynamo of 24 K. W. capacity is about 90% at full load, the losses being proportioned about as follows:
Watts. Per cent.
Shunt loss CR 778 3.3
Armature loss CR 833 3.7
Hysteresis, friction, etc. 700 3.0
2311 10
If now the load be reduced to one-half, the losses in the shunt and those due to friction, etc., will remain the same, or, summed together, will be 6.3% of the full load, or 12.6% of half load. The armature loss will be (½)2 X 3.7%, or .9% of full load, or 1.8% of half load, so that the entire loss will be 14.4% and the dynamo efficiency will fall to 85.6%.
For the high speed commercial dynamo for the water set, I have selected a 24 K. W. compound wound Lundell dynamo, running at 700 revolutions and having the same voltage as the present naval type. The compact and ironclad form of this make of dynamo, together with its very high efficiency, makes it particularly suitable for the work required of it. The losses are as follows at full load:
Watts. Per cent.
Shunt loss 684 2.5
Armature loss 800 2.95
Friction, etc., 690 2.55
2174 8
This gives a final efficiency of 92% at full load. At half load we have losses as follows : Shunt loss 5%, armature loss 1.45%, friction, etc., 5.1%, combined loss 11.55%, or the efficiency at half load falls to 88.45%.
The data for the Lundell dynamo were obtained directly from the Interior Conduit and Insulation Co. of New York and were from actual tests. As the dynamo was belt driven when tested, its efficiency would probably have been higher had it been connected direct to the shaft of a water motor, thereby saving the binding friction on the bearing at the pulley end of the machine.
There are many instances quoted by reliable authorities giving 94% and in some cases 95% efficiency for high speed dynamos. The General Electric Co. quote a 4 pole 25 K. W. dynamo, running at 1050 revolutions, on which it will guarantee 90% efficiency, but state that it will run higher than that figure. The C. and C. Electric Co. quote on a bi-polar 25 K. W. dynamo 91% as a guaranteed commercial efficiency. Therefore the 92% quoted by the Interior Conduit Co. is probably correct, and can easily be attained.
The Pump.
The efficiency of the modem direct-acting pump runs very high. Correspondence with the Worthington, Blake and Dow pump-makers establishes at least 90% as their efficiencies when everything is in good working order. At first thought this seems too high, but a little consideration and investigation will show that this is not unreasonable. An old-fashioned pump having 80% efficiency was considered very good; but that class of pump had a fly-wheel, crank, connecting rod, etc., and not as much thought had been expended on the details of the design; the openings were frequently insufficiently large and the valves were leaky. Further, the presence of the fly-wheel, so advantageous generally where unequal stresses were brought on a machine, was detrimental for two reasons: first, because its own weight and the action of the thrust of the connecting rod on the crank pin absorbed a large percentage of the work, due to friction; and, secondly, because the weight of the fly-wheel actuating the plunger of the pump, which was moving an incompressible and inelastic body, caused its power to be delivered like a blow, and there was great loss by impact; so that now, with a slow-moving, direct-acting pump, the steam cushions its own blow, giving it the effect of a push instead, and with almost no friction, we naturally can expect and do get high efficiencies.
Professor Robert H. Thurston, of Cornell College, in an article on the Contemporaneous Economy of the Steam Engine, in the Transactions of the American Society of Mechanical Engineers for 1894, gives all the data of an actual test of a pumping engine having a commercial efficiency of 90.78%, and states: "The friction of the engine is remarkably low for this type of pumping engine, but it is, of course, still above the figure obtained from the best direct-acting machines, which, in the Newtown, Mass., trial reported recently, for example, was found to be 4.2%, giving a mechanical efficiency of .958 as compared with that here obtained of .9078." By friction of the engine Prof. Thurston means the difference in horse power between the I.H.P. of the steam engine and the hydraulic horse power of the pump, as the data accompanying his report show. He further states that the Newtown, Mass., percentage of efficiency is somewhat higher than usual. It sounds paradoxical, but a little reflection will show that the greater the number of expansion cylinders, the lower the efficiency of the engine; therefore for a duplex simple pump the efficiency will be higher than for a triple or quadruple expansion engine. The reason of this is that there being more parts to the compound engine there is more friction, and therefore lower mechanical efficiency, but when we compare the coal consumed, then the advantage of the compound engine asserts itself. Quoting from a letter from the Blake pump manufacturing firm, we have: "Under favorable conditions, when running at its maximum, the efficiency will be in the neighborhood of 90%, and when only doing one-half the work this will probably be increased to 95% or 94%. The enclosed tracing shows cards taken by the writer from a Blake triple compound pump, with steam jacketed high pressure cylinder. In this case the efficiency was 93.2%, which is quite remarkable when all the places where friction is produced are considered."
Mr. Dow, of the Dow Pump Works, writes: "We find the loss in I.H.P. of our non-compound direct-acting steam hydraulic pumps, as we have furnished the U. S. Gov. boats Oregon, Olympia and Monterey, to in no case exceed 10% at full capacity, and 5% at half speed…The high efficiency you refer to, 95-96%, in large pumping engines is correct."
It is therefore safe to assume that the pump can have an efficiency at least as high as claimed for it, 90%, at full load for the dynamo; this will be but half load for the pump. The reasons the efficiency rises as the load is decreased in a pump are as follows: The friction depends directly on the pressure and speed, and as the speed varies directly as the amount of water required, at half load there will be but half speed, and as the pressure remains the same there will be just half the mechanical friction or the same loss per cent; but the water friction varies as the square of the velocity of flow, and as this at half load is half what it is at full load, the loss by friction will be but one-quarter. Therefore the efficiency should be and is higher. Still, to be on the safe side, I have taken the efficiency at half load, which will be only one-fourth the capacity of the pump, as 90%.
The Engine.
The efficiency of the high speed engine is about 89%, the loss being due entirely to clearance and friction. This loss, as the speed remains constant, is always the same; therefore at half load the engine efficiency will be 78%.
The Water Motor.
The water motor used in the before-described comparison is manufactured by the Pelton Water Wheel Co. of San Francisco and New York. I have tested other makes of water motors, and read of still other tests, and it has been my experience, as well as that of others making comparative tests with these motors, to find that they are at least 15% and in some cases 35% more efficient. The firm claims an efficiency of 85% when the wheels are set in accordance with their instructions,, and I have found their claims not only true, but below what is really obtainable. During the past two months Ensign W.H.G. Bullard, U.S. Navy, and the writer tested one of the Pelton wheels bought out of stock two years ago, and without any expectation at that time of its being tested for efficiency. Tabulated below are the results of our tests, showing an efficiency at the higher pressures in excess of that claimed by the firm. Tests Nos. 1, 2, 3, 10 and 11 were made with the motor running at incorrect number of revolutions, in order to determine the efficiencies when governing, in case the pressure on the whole jet was reduced and the revolutions kept Up. Test No. 13 was the last one made, and it was noticed that the brake was binding closely on one side, which may have occasioned the lower efficiency.
A Prony brake was used, with a stream of cold water running over it all the time, and two pieces of soap bearing against the pulley, like the brushes of a dynamo, kept the brake equally lubricated, and the pull on the scale was very steady. The pull was measured by a spring balance, the brake arm being kept horizontal all the time, and the balance was compared with the standard after each test, and in several cases the standard itself was used. The standard balance was verified before and after the series of tests and was found to be accurate. The amount of water used was absolutely measured by running it into the pool of the natatorium of the Naval Academy, where the level of its surface could be measured by means of a float and rod that could be read with exactness to the one sixty-fourth of an inch, which corresponded to just one cubic foot. Thus the need of any formulae or miners’ inch measurement was obviated. In this manner a number of tests were made, and the coefficient of ajutage found to be 95% with the 5/8-inch jet, and 92% with the ¾-inch jet, which agrees very closely with what they should be when worked out theoretically. After the coefficients were determined and found to be practically constant, the 5/8-inch jet not varying 1% in a range of pressures from 20 to 100 pounds, and the ¾-inch jet not varying at all in from 15 to 50 pounds, the water motor was moved to the power house, close to the pump, so as to get rid of pipe friction and have the pressures more under control, and at the same time- higher pressures were obtainable. The amount of water then used was determined by using the coefficients of ajutage found as already described. The efficiencies thus determined can therefore be taken as being reliable, and agree with those in testimonials published by the Pelton firm. In this connection it is a matter of interest to quote from a letter from the Pelton Water Wheel Co. on the subject of efficiency: "Regarding the efficiency of our wheel when carefully made, especially for test purposes, would say that several years ago we sent a special test motor to Professor Reuleaux of Berlin, who obtained an efficiency between 90% and 91% from same…Referring again to the installation at Fitchburg, the efficiency shown by the wheels there is fully 86%, the power delivered by generator is easily ascertained from the station instruments, and taking the theoretical discharge of the nozzles—no allowance whatever being made for nozzle friction—the efficiency of the wheels was 86%."
The efficiency therefore of the water motor at full load is 86%. This loss of 14% is made up of several things: friction of machine, friction of water in buckets, loss by impact on edge of bucket, etc. Of these the friction of the machine alone remains constant, as the speed remains the same; as the water is throttled down the pressure is reduced, and to keep up the efficiency the revolutions should be reduced correspondingly; but this cannot be, as the speed of the dynamo must remain constant; hence the efficiency will fall to about 75%. (See tests Nos. 6 and 11.) Here we throttle 125 pounds and 880 revolutions down to 86 pounds and 880 revolutions, and the efficiency falls from 86% to 75%. This can be obviated in two ways: first, by using a number of jets, say four, two of them being fitted with throttle governors, and by automatically or by hand cutting out a jet when the throttling reduces the power by an amount equal to the power developed by one jet By this means we retain our 86% efficiency on two of the jets as long as the load is above one-half, so that we would have, at the lowest possible efficiency, two jets of 86% and one of 75% or a combined efficiency of 82 1/3%. Another and apparently simpler method of keeping up the efficiency, and one that I think is mechanically practicable, is to have one or more jets that can be varied in size of opening according to the power required. In this case the efficiency would not fall below 84%. But we will allow that we cannot get over 82 1/3% at half load; hence we have 86% at full load and 82 1/3% at half load as the efficiencies of the water motor.
Combining these efficiencies we have for the steam system at full load 90 X 89 = 80%; for the water system at full load, 90 X 92 X 86 = 71.2%. At half load we have for the steam system 86 X 78 = 67%, and for the water system 90 X 88.45 X 82% = 65 ½%. But in every dynamo room where the dynamos are run by steam there is an electric motor and blower to help keep the room cool; the size of this blower varies from ½-horse power to 2-horse power, and as these small motors have not efficiencies over 75%, it will be safe to say that the average blower installed with a 24 K.W. set uses 1 ¼-horse power from the output of the main dynamo.
We therefore have finally for the efficiency of the steam plant, meaning thereby the ratio between the power produced at the dynamo terminals, less that required for running the blower, and the I.H.P. of the engine, at full load 89% X 86 ½% = 77%, and at half load 78% X 79% = 61.6%. In these combined efficiencies it will be seen that the power needed to run the dynamo room ventilating blower is taken into account in computing the efficiency of the dynamo where it is considered as a loss.