How shall we describe the carbon copy concept an intrepid U.S. shipbuilding industry turned to during World War II to resolve a desperate dilemma? Outmoded? The author does not think so.
One of the most fundamental changes wrought by Secretary Robert S. McNamara is that the defense budget is no longer planned in terms of Army, Navy, or Air Force. Funds are allocated to military missions, not services. For instance, the size of the Polaris submarine force is related, in this process, to the Minuteman missile and other strategic nuclear forces, not to the number of nuclear attack submarines, attack carrier task forces, and other naval systems. Similarly, attack carrier forces compete for funds with Air Force tactical aircraft and Army tactical missiles.
As strategic, technological, and political changes take place, new missions for the Armed Forces will develop, some old ones will take on new importance, some will decline. The size and importance of the Navy of the future will depend upon how imaginative and convincing the Navy is in demonstrating the superiority of sea-based systems for both new and old missions. We must continue to evolve and develop new, ingenious, and increasingly effective ways of using ships. However, the budgetary competition is so much more immediate and direct than ever before that the Navy ships carrying the new weapons or assigned the new tasks must be designed, constructed, and operated with maximum efficiency if they are to compare favorably.
Naval shipbuilding today is not contributing its part to the ability of the Navy to compete and serve. During World War II, the United States developed the greatest shipbuilding industry the world has ever seen. The industry produced thousands of ships using modern mass production techniques, with some yards capable of producing a ship within a few weeks of laying the keel. The huge surplus of ships available after the war sent shipbuilding in the United States into a decline from which it has never recovered.
Foreign nations now produce the vast majority of the world’s ships. U.S. shipbuilders are almost limited to producing naval ships or commercial ships for which the government subsidizes half the cost.
In the U. S. aircraft industry, on the other hand, wartime planning and production lessons were retained and improved upon. As a result, this country is the leading supplier of the world’s commercial aircraft, and it does not need direct subsidies to compete successfully with the aircraft industries of other nations that pay much lower wages.
It is interesting to compare presently used techniques in the U. S. aircraft industry with both U. S. and foreign shipbuilding practices. Such examination suggests that certain changes in the methods by which U. S. naval ships are designed, contracted for, constructed, and operated would produce substantial benefits in terms of the effectiveness, modernity, and economy of our naval forces. These benefits are of two categories—initial construction cost savings and operating efficiencies.
As to the first category of benefits, those associated with design and construction, it appears that there are at least four ways whereby we could obtain much more for the money spent on shipbuilding. The first of these would be by:
Standardizing designs, building or modernizing classes of ships in large blocks, e.g., 25 to 75 ships, and awarding the contract for all ships of one design to one firm. This would take full advantage of the ability of shipyards to produce each successive one-design ship more efficiently. For some naval ship classes, this could cut the shipbuilding cost in half. This was amply demonstrated during World War II in the United States.
The standardization of design and concentration of production would lead to economies in the procurement of components for the ships and also permit inventory savings. In view of the relatively small number of ships we produce in peacetime, the production of all one-design ships would have to be concentrated in one shipyard to gain these advantages of series production. Of course, that yard might procure components or subcontract to many areas of the country as is now done in the aircraft industry.
Shipyards in various areas could be the prime contractors on different ship types. At present, there is a continuing and serious problem of wide fluctuations in employment at shipyards. These could be greatly reduced if a winning contractor knew that he could reasonably plan his labor force in the expectation of building five to ten identical ships per year for perhaps five to six years.
Another method would be the ensuring, in the design process, that each specification is justified on the basis of its cost as well as the extra effectiveness provided. A review of each element of Navy shipbuilding standards should be made along with trade-off studies of the over-all design and installed systems for each ship type.
Thirdly, we should carry out studies of the relationship between initial ship construction cost and lifetime operating cost very early in the design process. In business, and in the Pentagon, emphasis is properly placed on the total costs of both buying and operating a factory or a weapons system. If spending more on a machine initially will reduce the cost of its operation it may well be desirable to do so.
Although it is convenient for expository purposes to make a distinction between ship construction and operating costs, it would be wrong to ignore their real and important relationship. Some examples are the effects on manning and maintenance requirements of automation of machinery, the use of various types of propulsion, and the employment of protective coatings. The emphasis here should be on early and complete analysis carried out in a competitive environment.
Finally, we can increase productivity in the shipbuilding industry by modernizing production facilities. A new flow-line shipyard employing automated machinery similar to that used in the latest yards abroad should be built in the United States.
In summary, then, the benefits of reduced ship procurement costs call for procurement practices much like those used for aircraft or, for that matter, land vehicles. To accomplish this, we should consider adopting a new set of policies for the shipbuilding program.
First, ship designs should result from a competitive contract definition process. The Navy would present the prospective bidders with a carefully defined task which the ship should be designed to accomplish. The firms competing for the contract would submit the best designs that their staffs can devise to provide the greatest effectiveness for the cost, and would include specific guarantees of initial price performance and operating costs. In other words, we should spend more time and money on analysis and research and development before going into production, and then make up for it by mass production.
Second, a review of these competitive proposals would lead to a multi-year contract with one firm for a number of identical ships, the government obtaining fixed-price options on successive lots of ships. These techniques were used in the procurement of the Navy’s new light attack aircraft, the A-7A, and have been further refined in the competition for the Air Force’s heavy logistic aircraft, the C-5A.
Third, this change in policy from the parceling out of contracts for one-design ships to a number of firms to a policy of awarding a multi-year contract to a single firm will encourage the competitors to modernize their facilities or construct new ones for mass production. Providing a substantial termination agreement in the contract for the first few years would reduce the risk to contractors and offer further incentives for modernization.
As to the second category of benefits, the implications of these proposals for the operating efficiency of the Navy may be as important as their direct effect on shipbuilding. As an example, the standardization of ships of a particular type would make it possible for any crew to man any ship of that type in the Fleet. It would be more feasible, therefore, to have a different operating cycle for men than for ships (as is done with commercial aircraft and crews) without adversely affecting performance. It would be possible to optimize the use of men and ships separately. If the crews are not tied permanently to one ship, a larger proportion of the Fleet could be kept forward, while at the same time, fewer personnel would have to face extended periods away from home, fewer personnel and ships would be necessarily involved in long non-productive transits, and skilled operating personnel would not be used to provide unskilled labor in shipyards.
A ship could be kept on foreign station from overhaul to overhaul, i.e., two or three years, during which period several crews could be rotated to the ship by air. If desirable, there could even be different periods of rotation for different parts of the crew. Some of these changes can and are being made without a high degree of standardization of ships, but standardization would make the changes easier, more widespread, and more profitable. As pointed out below, a decision to implement these possible changes requires a careful study of each ship type to determine desirability and feasibility.
Standardization would permit great inventory savings on spare parts and, at the same time, provide greater readiness. Standardization might make feasible a specialization of the repair and overhaul or modification of a class of ships. With a shipyard specializing in this work, the crew would not be used to provide expensive unskilled labor as is now done during the overhaul or repair period, but would be free to operate other ships of identical design, receive training, or take leave.
The attention paid to total lifetime systems cost early in the design stage could well lead to additional economies in the use of our naval manpower. Machinery automation, for example, could reduce watch requirements, freeing men for maintenance, training, or other duties, and ultimately reducing operating costs.
Last, the use of large-scale series production would permit a substantial reduction in the lead time associated with introducing large numbers of a new ship type into the Fleet. This would provide a valuable hedge against uncertainty about force requirements, because forces could then be augmented more quickly. Further, a reduction in lead times means that we can respond more quickly to changes in the threat as are, for example, uncovered by intelligence. At the same time, we can reduce the need to build large numbers of competing, but somewhat different, systems to provide hedges against such changes. In those cases when production of a program of ships is completed using series production methods it is possible that it might take more time for an initial restart of production and for the construction of a small number of ships than if a number of smaller yards with some experience in building the type were available. Hence, in those cases where we may wish to restart production rapidly, we should consider making arrangements which facilitate the rapid reopening of production. For large numbers of ships, the series production techniques would deliver the total program of ships earlier even if it took longer to produce the first few.
The following questions and answers permit us to discuss each of the points that have been mentioned above in more detail.
What can be learned from foreign shipbuilders? A comparison of United States and foreign shipbuilding costs gives an indication of ways in which the cost of building ships in the United States can be reduced or the capability provided by given shipbuilding expenditures increased. The Maritime Administration estimates that the cost of ships built abroad is 40 to 55 per cent of U. S. costs.
The high U. S. cost is frequently attributed to high U. S. wage rates. While wages are, of course, a partial cause, several countries with substantially different wage structures are in effective competition on the world market. Of particular interest is the Swedish case. Labor rates are 61 per cent of the U. S. rates, yet Sweden has been successful in obtaining foreign shipbuilding contracts and has prices close to Japanese yards which enjoy wage rates that are 44 per cent of Sweden’s. For comparison, the 1964 hourly earnings, including fringe benefits, of shipbuilding employees in selected countries were as follows: United States, $3.15; Sweden, $1.92; Germany, $1.18; Netherlands, $1.17; United Kingdom, $1.07; and Japan $0.85.
The successful competition of the Swedish yards despite high wages, appears to be the result at least in part, of high productivity. The Swedish yards are modern, well planned and well equipped.
Increased labor productivity could reduce U. S. shipbuilding costs substantially. It is clear, however, that labor costs alone do not constitute a large enough part of the total cost of ship construction to explain the large difference in U. S. and foreign ship costs. The cost breakdown of a $12,000,000 cargo ship built in the United States is as follows:
| Base Vessel Cost | Per cent |
Steel | $ 900,000 | 7.5 |
Other Material | 5,700,000 | 47.5 |
Labor | 4,500,000 | 37.5 |
Overhead | 900,000 | 7.5 |
| $12,000,000 | 100.0 |
Assuming equal productivity, this cost breakdown indicates that labor cost differences as large as three to one could account for total ship cost differences of only 25 per cent.
Unfortunately, U. S.-foreign cost differences cannot be readily explained on the basis of the other elements of ship cost. Overhead
and steel plate costs are a small part of total cost, and steel prices do not vary widely from country to country. About half the cost included in the category “Other Material” is accounted for in the following way:
Approx. Cost
Main Engine and Gearing $ 570,000
Boilers 350.000
Turbine Generator Sets (2) 240.000
Hatch Covers 800.000
Accommodations (crew and 12 passengers) 600,000
52,560,000
One estimate, made by the Maritime Administration in its June 1965 report to the Congress, indicates that the total cost of all materials (including steel and externally manufactured components) in foreign countries is 67.2 per cent of the U. S. cost. This would account for an additional 18.0 per cent of total cost which, if added to the labor cost differential of three to one, would imply that foreign ship cost should be 57.0 per cent of U. S. cost as opposed to the MARAD estimate of 40 to 55 per cent of U. S. costs. Construction differential subsidies awarded since June 1963, have all exceeded 50 per cent and have implied that foreign ship costs are 46.2 per cent of U. S. costs.
Thus, labor and materials cost differences alone are inadequate to explain fully the U. S.-foreign cost differences. More important, they do not indicate ways, other than modernizing our shipyards, of achieving improvement in U. S. industry. There are, however, significant differences which do provide clues. Labor costs per ship are not only a function of the wage level and the quality of shipyard equipment and layout, but also of labor learning in the production of a single ship type. Many foreign producers tend to build greater numbers of more standardized ships than do U. S. producers. Yet, the higher U. S. wages would make series production relatively more attractive here. As well, the high U. S. costs noted above of ship components manufactured outside of the shipyard are Probably less a reflection on the efficiency of IL S. suppliers of turbines, etc., than they are an indication of the "customizing" of nearly every detail of our present-day ships.
Finally, the ships are not truly comparable as there is some evidence of greater “over-design” in U. S. ships. In other words, the difference in ship costs may be attributed in part to the foreign tendency to provide quantity production of a specific ship design with greater standardization of externally manufactured components and to differences in U. S. and foreign design methods and standards. These differences indicate possible steps to be taken in the United States.
Is there a learning curve in the shipbuilding industry? There is a large body of evidence from many different manufacturing .industries which indicates the existence of a learning or progress curve. The best known form of the progress curve states that as the total quantity of units produced doubles, the cost per unit declines by some constant percentage. The “cost per unit” can either be defined as the average cost of “n” units or the cost of the “nth” unit. The latter definition is used throughout this discussion. The pronounced learning curve in shipbuilding can be attributed to organizational and individual learning at the yard and, to cost savings, as well, in the procurement of manufactured components in volume quantities obtained at lower prices because the producers of components also learn from series production.
A rule of thumb based on extrapolation of recent Maritime Administration contracts is that the average unit cost of a group of 15 ships built in one yard is 75 per cent of the single ship order price; the fifteenth ship costs about 65 to 70 per cent of the single ship cost indicating an individual unit learning curve with a slope of 0.9. It should be noted that this is not due to high fixed initial engineering services cost. Unlike naval ships, the merchant ship “lead-ship” costs are small, being usually less than one million dollars, or less than ten per cent of the cost of one “followship.”
Wartime shipbuilding experience also indicates that the construction of a number of ships of identical design by one shipyard reduces the average cost per ship substantially. The experience during World War II of building Victory ships and destroyer escorts indicates a well-developed shipbuilding individual unit learning curve with a slope of .86. But other examinations of World War II shipbuilding experience, notably that by A. D. Searle, in the December 1945 issue of Monthly Labor Review, indicate greater learning with curves having slopes of .84 to .78. The following table shows the average number of man-hours required to construct Victory ships:
What are the present U. S. practices? In the United States, the common practice of the shipbuilding industry is to bid on “custom” designs, while many foreign yards have their own designs and offer ships of that design, thereby taking advantage of the economies of quantity production of a single-type ship. As pointed out above, this practice may account for a substantial portion of U. S.-foreign cost differences.
Series production—and the contractual arrangement it implies—requires that a contract or contracts for a number of ships be let over a relatively short time. It might be thought that this would make the application of this concept to naval ship construction inappropriate because it would build in bloc- obsolescence. A review of our postwar naval shipbuilding program, however, indicates that in fact this is the way we do buy ships (i.e., concentrated production of a few categories of ships, although often of more than one design, followed by shifts to similar concentration on other categories). Unfortunately, we do not take advantage of this fact to obtain the economies of scale possible with series production.
Table I shows the construction programs of all but a few ship types for the Fiscal Years 1948 to 1966. As well, the increased efficiency in building, manning, deploying, repairing, and altering ships which standardization might make possible, coupled with a system which could permit us to build rapidly and modernize a particular ship type would more than compensate for the potential “bloc- obsolescence” problem.
What are the potential economies of series production? The present destroyer escort construction program provides an opportunity to consider the possible economies of series production. The 26 ships in the Fiscal Year 1964 and 1965 segments of this program were awarded at the same time; seven ships each went to three yards; and one yard was awarded a contract for five. The shipyard contracts plus funding for labor/material price rises, change orders, etc., accounted for slightly over half the cost, or an average of 13.2 million dollars per ship exclusive of “lead ship” costs. If a learning curve slope ranging between .86 and .9 is assumed to exist for the short production runs of each of the four shipyards this portion of the initial ship cost amounts to about 15.2 million dollars.
Government-furnished equipment accounted for the remaining cost or an average of 11.2 million dollars per ship. Thus, a total initial ship cost of 26.4 million dollars results under present contracting arrangements.
In Fiscal Year 1966, an additional ten new DEs have been included in the program. There are a large number of aging World War II destroyers which will require replacement. If we can assume, for the sake of argument, that the annual buy will continue for the next five years at the present level of ten, the total program of DEs of this type will total 86.
Let us examine how the procurement program costs might vary under alternative purchase arrangements. Four procurement methods for the DE construction program are compared in Table II on page 30.
Procurement Method A assumes no learning; it would be equivalent to awarding contracts to different shipyards for individual ships, or to a situation in which uncertainty about future contracts caused the shipyards to forego the possible savings which could accrue if longer production runs could be planned for in advance.
Procurement Method B assumes that learning does occur, but that the runs are limited to the three contracts for seven ships each, with all the rest limited to five ships each (i.e., each year's program to of ten ships is awarded equally two yards).
Procurement Method C represents the most optimistic situation likely under present methods. It assumes that no more than six yards win contracts (four already have contracts), and that they have enough confidence that this will happen to take full advantage of the economies of scale.
Method D assumes that one shipyard wins a multi-year contract and builds all 86 ships.
Two learning curve slopes are examined: 0.86 (which equates to labor learning in World War II ship construction) and 0.90 (which is approximately the MARAD estimate). The procurement cost of government-furnished equipment is treated on the basis of alternative assumptions that (1) there are no cost changes, and (2) that a learning curve slope of 0.85 occurs. We should expect something between these two costs of Government furnished equipment to occur. The cost of much equipment already reflects cost reductions as it is bought as part of larger orders for a number of ship types, and therefore we ought not to expect further cost reductions on it. Other equipment is bought primarily for these ships, and cost reduction may occur as the quantity bought increases.
Table II indicates the considerable advantage which could be realized through contracting with a single producer for series production of all ships (Method D). For the same total cost, even under the more conservative assumptions, this method would permit the procurement of about 15 per cent more ships. (Method D in comparison with Method C, using six shipyards, and assuming a shipyard learning curve slope of .9 and constant costs on government-furnished equipment: 114 ships versus 99 ships.) Under the more optimistic assumptions, the increase in the number of ships would amount to 87 per cent, (Method D in comparison with Method A, in effect, using 86 shipyards, and assuming a shipyard learning curve slope of .86 and a learning curve slope of .85 on government- furnished equipment: 161 ships as opposed to 86 ships.)
In this hypothetical DE program, Method A would not occur because, as pointed out above, contracts for 26 ships have already been let together to only four shipyards and their bids clearly implied learning. Additionally there are not 86 shipyards in the U. S. capable of building DEs (although a practice of spreading contracts which resulted in too low a rate of production could have the effect of preventing progress).
We might reasonably expect costs in an 86 DE ship program between those resulting from Methods B and C. On one hand, there are normally about 16 private shipyards capable of building DEs. On the other hand, the shipyards winning initial contracts would be in a strong competitive position in future competition as their unit costs on the subsequent ships would be lower. The uncertainties to the contractor of yearly or bi-yearly contracting would reduce the efficiency of shipyard planning and procurement even if only a few yards were consistent winners. Furthermore, the yards which are already in production might recognize their cost advantage over yards which are not and adjust their bids upward. Therefore, restricting consideration to comparison of Method D with costs halfway between Method B and C, we might expect between 18 per cent and 45 per cent more ships using Method D.
Before proceeding, it may be well to consider the implications of series productions of ships in the context of an important and expanding military mission, one in which the Secretary of Defense’s new budget planning procedures have put the Air Force and the Navy in direct competition. Considerable publicity has been given lately to the need for further increases in our ability to deploy rapidly our Army forces and to the resulting development of a new, huge Air Force transport aircraft, the C-5A. But the Navy has not been sitting still. Elaborate studies have been conducted which show that ships have an important part to play, not only in the resupply of deployed forces, but in getting their equipment to a trouble spot quickly. A system of ships, preloaded with Army equipment protected from deterioration in dehumidified holds, and operating with or near other naval forces on world-wide stations has been proposed as an economical, effective way to meet the future need for rapid delivery of forces.
The ships, called Fast Deployment Logistic Ships (FDL), would deliver the Army equipment; the Army personnel would be flown in to marry up with it. The ships have advantages not shared by aircraft. For example, with strategic warning, they can proceed to an area and standby at sea ready to off-load on short notice without causing an international crisis. Aircraft also have advantages. There will likely be some of both in any long-range plans, but the question of how many ships or aircraft we will have is very much open.
The mix of ships and aircraft to carry out the rapid deployment mission will be determined in large measure by their relative marginal effectiveness per dollar. In the present context, this simply means that the choice between adding one more ship or aircraft to the force will depend on how much additional capability the extra ship or aircraft gives to a once per dollar. Therefore, an important cost to consider is that of the extra or next ship in e series. If the price is reduced on each successive ship, it becomes more likely that a ship will be added to the force in preference to an aircraft. If the price of all ships, after the lead ship, is the same, the likelihood will cease to exist.
The potential effect of series production of Fast Deployment Logistic Ships (FDL) production is indicated in Table III.
Table III shows the large absolute savings that series production of FDL ships might permit. But focus attention on the marginal costs. If the learning curve has a slope of .86 (or .9), the cost of the 21st ship would be only 16.6 million dollars (or 20.1 million) as compared to 32 million dollars if the ships were built individually. In the calculus of a cost and effectiveness study, a difference of that magnitude could be of central importance. Series production, combined with the construction economies possible in a new shipyard and the operating economies described below, could have a significant effect on the composition of the best mix of airlift and sealift for rapid deployment.
Can we modernize our shipbuilding facilities? It should be noted that series production implies a fairly high degree of concentration in the industry with large factories accounting for nearly all production. Table IV compares, by size of establishment, the degree of concentration of employment and output in the aircraft industry, which is characterized by series production, with that of the U. S. shipbuilding and repair industry, which is not. The aircraft industry is clearly more concentrated. In order to implement series production of ships, the present structure of the industry may have to be changed.
The initial cost of a highly automated, flow line shipyard capable of building eight to 16 destroyer escort ships a year is estimated to be between 70 million dollars and 80 million dollars including real estate. (Gotaverken’s yard at Arendal, Sweden, of this capacity and design is reported to have cost 40 million dollars to build. Of course, we would not necessarily want a carbon copy of the Swedish yard, but we should take the conditions peculiar to the United States into consideration in the design of our hypothetical yard, e.g., the high quality of U. S. steel plate, climatic conditions, types of ships to be built, the capabilities of ship component manufacturers, etc.)
Learning, and some of the benefits of series production, can occur in any shipyard, new or old. Hence, series production can be achieved in many existing yards, but the slope of the learning curve may be considerably different than the one experienced in a new yard and initial ship cost might be different. Referring to Table II again, we see indications of the large effect of changes in learning curve slope and hence the effect on shipbuilding savings. On an 86-destroyer-escort purchase the savings increase from 522 million dollars to 673 million dollars as the slope varies from 0.90 to 0.86. ($277 million to $341 million when Method D costs are compared with costs halfway between Methods B and C.) Assuming no change in the slope of the learning curve, a simple reduction of five to six per cent in the shipyard costs (exclusive of government- furnished equipment) of the 86-ship DE program would pay for the yard. On an FDL program of 40 ships, 416 million dollars to 547 million dollars might be saved due to series production or five to eight times the cost of a new yard. We have been considering whether we can afford more modern shipyards. In the long run the question might properly be whether we can afford not to modernize our nation’s shipyards.
How do design practices in the shipbuilding industry compare with those of the aircraft industry? The absence of a process like the design competitions prior to aircraft contract awards may be another important element in the relatively high cost of ships. The design competition in the aircraft industry forces the competitors to design aircraft which will give effective performance yet can be built at the lowest cost. Every aspect of the proposed aircraft is studied and restudied by the large engineering design teams of each firm to ensure that the effectiveness justifies the cost.
The design of ships, on the other hand, is largely accomplished by agents who are not connected with shipyards. The present policy of most U.S. private shipowners who build here is to employ design agents to develop designs and prepare plans and specifications. All of the basic design decisions are, therefore, made before the first bid is submitted. Naval ship design follows a similar pattern; the over-all design is normally determined without benefit of detailed and competitive study of costs and effectiveness.
The long production runs of one aircraft type and the design competitions which precede contract award make it profitable and necessary in the aircraft industry to maintain large engineering staffs. This is also true of other major defense industries but not of the shipbuilding industry.
Professional, technical, and kindred workers account for 25 per cent of all defense-related employment in the aircraft and ordnance industries, for example, as compared to seven per cent in the ship and boat building and repair industry.
There has also been a definite increasing trend in the percentage of non-production employment to total employment in the aircraft industry since World War II. The number of non-production workers in the aircraft industry rose from 36,000 in 1947 (25 per cent of total employment) to 126,000 in 1958 (33 per cent of total employment). In shipbuilding, the number has been low, 13,000 to 18,000 (an average of 13.5 per cent of total employment) and shows no clear trend.
It is, of course, true that non-production workers are part of overhead costs, but skilled engineering, cost analysis, and management teams are an essential part of modern production. The great success of the U. S. aircraft industry in the world market in spite of high labor costs, provides evidence that this overhead cost is more than compensated for in increased over-all efficiency.
One of the ways in which design and management teams operating in a competitive situation might increase the effectiveness that is provided by our shipbuilding expenditures would be to examine shipbuilding standards. These may be unnecessarily high; i.e., the cost may be too high in relation to the extra capability that is provided over those ships with different standards.
Little can be said definitely in this area without a careful examination of the costs and benefits of each specific item. However, in gross terms, Navy standards about double the cost per ton of ships over commercial standards. American commercial standards are higher than European or Japanese standards and are estimated to account for a 10 to 15 per cent cost differential. It is by no means clear that the cost differential is justified. There is some evidence that substantial "over design" of commercial ships occurs as a result of tradition.
There may be an element of increased cost in designs executed without intimate knowledge of and communication with the actual production facilities of a specific shipyard. If ships of one design are to be built in several shipyards of different layout and capability, designs tailored to the production facilities are, of course, impossible. The value of ship-yard-oriented design teams and series production in this respect is obvious.
The over-all problems discussed above are being recognized in the merchant fleet.
As W. J. Dorman and J. J. Henry observed the in Vol. 69, 1961 issue of The Society of Naval Architects and Marine Engineers.
then there is an area, for which we do not have a good name at the moment, which involves the shipbuilder and designer. The usual U. S. practice is to have a naval architect prepare contract plans and specifications. These are used for obtaining bids and guiding the construction after award of a contract. These plans and specifications are not tailored to the construction methods of any one yard; sometimes they are too restrictive and in the over-all picture tend to eliminate initiative on the part of the shipbuilder . . .
In turn, the shipbuilder has not thoroughly entered into the job of reducing construction costs. But if we are to get as many ships as possible for the available money, all parties involved—owner, government, labor, naval architect and shipbuilder—must co-operate to reduce the cost of ships by eliminating overdesigning and overbuilding.
In summary, then, the high cost .of U. S. shipbuilding may be the result of a vicious circle. Fewer ships are built because the cost is high, yet the cost remains high because short production runs preclude a redesign of the industry which would lower costs. The traditional practices of custom designing most ships, of parceling out construction contracts for ships of similar design, and of non-competitive design arrangements all tend to perpetuate the high cost.
The proposed way to break the circle and to improve the return from our naval shipbuilding expenditures would be to conduct a competitive contract definition phase with firms, which would be bidding on production, submitting designs. The contractors would offer the government a multi-year fixed-price incentive contract proposal offering options on successive blocs of ships. The requirement to meet arbitrary construction standards might be removed in favor of detailed specification and guarantees in the proposed design of performance, reliability, and cost. If the final contract were to include long-run production and encourage construction of a new modernized facility, the number of possible bidders could be increased to include representation from industries familiar with parametric design, large-scale production, and the close relationship between design work and production.
As to what changes in methods of manning and operating ships might be instituted, several other questions seem appropriate.
Can we separate planning for personnel and ship utilization? The large number of different types of ships in the U. S. Navy and the differences between ships of one type built in different yards compound the problems of manning Navy ships. It is difficult to rotate crews under these circumstances. When individuals are transferred from one ship to another, they often undergo a “breaking-in” period that either means less efficient and less safe performance, or a temporary crew augmentation while the new man is being instructed.
If ships of one type can be built identically, then crews could be moved readily from one ship to another and the use of manpower could be divorced from the ship’s employment cycle. If the number of ship types can be reduced, greater benefits might result. As pointed out above, ships could be kept overseas from overhaul to overhaul with crews rotating by air; crews would not be assigned to ships in overhaul, the full repair work being done by specialists. The resulting reductions in under-use of manpower and ships could be translated into increased numbers of forward- deployed units, decreased manning for any given deployment level, or, possibly, increased likelihood of retention of personnel.
A rough estimate of the potential for more efficient use of personnel is provided by the average percentage of time that a naval ship is in overhaul or transiting. During the 1964 calendar year, naval ships spent, on the average, 9 per cent of their time in overhaul and 9 per cent transiting. Under the suggested concept, the crew need not be assigned during overhaul and the number of transits could be reduced by keeping a ship deployed and rotating crews by air. On the remaining transits during peacetime there would be little reason to maintain a full crew.
The converse of these considerations is to be found in possible ineffective employment of the ship because of the needs of the personnel attached to it. Including overhaul time, ships spend over 60 per cent of the time in port. Some part of this time is required for repair, upkeep, and provisioning, but a part can only be attributed to the needs for rest and recreation of permanently assigned personnel.
The decision to implement a program to divorce crew and ship utilization would, of course, require a careful and detailed analysis of the optimum cycle for each ship type. This would include consideration of the relative advantage of having various proportions of the force deployed forward. The peacetime advantages of the program in terms of reduced operating cost and better working conditions for naval personnel must be examined in the context of wartime use of the ships.
What trade-offs between ship construction and operating costs are possible? The required manning levels of naval ships might be reduced with no loss in effectiveness by greater use of automated control equipment. These devices are being installed in some foreign combatant ships. In May 1963, the U. S. Maritime Subsidy Board approved a change under contract for certain subsidized ships to provide centralized control of machinery, automatic data recording, and other mechanization features which could reduce manning scales by 20 to 25 per cent. The cost per ship to effect this manpower saving was in the range of only $250,000 to $500,000.
The operating costs and manning levels of ships should be related to initial investment costs in other ways. For example, future costs including manning can be reduced by spending more money initially on protective coatings, aluminized rigging, and constant tension winches. Inorganic zinc silicate coatings are relatively expensive at $0.50 to $1.50 per square foot (about $83,000 extra for an FDL size ship) but they require little upkeep, thus eliminating a large amount of costly, tedious, time-consuming chipping and painting.
Finally, there are other possible advantages. It may be possible to reduce the lead times associated with ship construction and thereby gain the advantages noted above. With more modern facilities, construction time can be reduced substantially. Gotaverken’s Arendal yard can complete a large commercial ship in 19 weeks. (The same company’s yard in Gothenburg requires about twice as long to do the same work.)
Further, as one company gains expertise in construction, learning curve phenomena reduce construction time dramatically. Again looking at World War II experience in building Victory ships and destroyer escorts, we find the following:
While the implication of these figures for production expansion is clear, there is no evidence s on what the restart time of construction of ships would be if the run of a particular type had been completed and the yard shifted to some other type. Shipbuilding, itself, may not be the long-lead-time item. Administrative lead time, the lead time for weapons or other component manufacture may be dominant. However, the concepts discussed above would tend to reduce these times also In the case of administrative lead time, the government would be dealing with one producer, familiar with the product and with experience in building it. As for components, concepts above imply greater standardization of parts which should mean that there is a greater likelihood that essential components would still be in production.
By having identical ships, the problems associated with modernizing and converting them could be simplified and contracted for in the same way as for initial construction. It is also possible that sufficient changes in the relative cost of new production and conversion might be effected to make it more attractive to build new ships rather than convert old ones. For naval ships, this is a distinct possibility, as the electronics and weapons systems are a major cost item, and they, in general, are the equipment changed in conversion. If basic ship costs can be reduced it may be preferable to junk old ships, and build new platforms for new weapons rather than convert old platforms.
By adopting these new concepts in shipbuilding and in ship operation, it may be possible to effect large and important changes in the composition and quality of the Navy. More and newer ships could result or better use be made of present numbers of ships. The Navy might be able to attract and retain higher quality enlisted men. Sailors would spend fewer hours on tedious, boring jobs, have greater assurance of rotation, and face fewer extended periods away from home. The percentage of ships we could keep deployed might increase.
This discussion has concentrated on possible long-run basic changes. There are, of course, important transitional problems which would have to be faced if it were decided to proceed along these lines. Nevertheless, it seems quite possible to introduce many changes which might improve performance relatively quickly. We might, for example, rotate at least part of the crews of selected deployed ships by air or transfer crews off ships in overhaul.
The shipbuilding concept could be very effectively started on a program of new fast deployment logistic (FDL) ships. This would permit us to implement all of the proposals which appear desirable after further examination and do it on a new class of ships with a new mission. The next logical step might be the amphibious assault ship program.
Devising new and imaginative ways of using ships to project fighting power may not be enough. Our ships must be designed, constructed, and operated in the most efficient way possible to ensure that the Navy of the future continues to occupy a central place in the nation’s military establishment.
A graduate of the U. S. Naval Academy with the Class of 1956, Lieutenant Commander DiBona served in the USS Monssen (DD-798) for a year before attending Balliol College, Oxford University, as a Rhodes Scholar from 1957 until 1960. He attended Submarine School at New London, Conn., and was then assigned to the USS Irex (SS-482) from April 1961 until 1963, when he assumed his present duties in the Office of Assistant Secretary of Defense (Systems Analysis).