Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
3 TECHNICAL ASPECTS OF SYSTEM DESIGN This chapter provides an analysis of the present design, construction, and inspection standards utilized with regard to tourist submersibles. In the case of inspection practices, relatively detailed suggestions are presented for the development of a periodic inspection program specifically for these vessels. The life support systems and emergency rescue equipment aboard these submersibles are also reviewed briefly in this chapter. DESIGN AND CONSTRUCTION Compared to workboat DSVs, the technologies employed for the metal-hulled tourist submersibles are conservative and fairly simple. These vessels are shallow diving (30-50 meters, or 98-164 feet), do not require elaborate operational equipment, and have short mission times. The major differences, when compared to submersible workboats, are the need for adequate life support, emergency equipment for a relatively large number of people (30-50), efficient and safe means of loading/unloading passengers, and the need to withstand a much larger number of dive cycles. Also, the interior of the submersible must be designed from both a business and safety standpoint to provide the passenger with maximum comfort, good viewing opportunities, and a relaxing interior design. A major consideration in design and construction is that these assets will probably have an operational life of 20-25 years. Consequently, the hulls will have the highest number of diving cycles of any submersible built as they potentially pass through the hands of many owners. Initial design should therefore take into account a long service life as well as simplicity of maintenance and operation. From a safety standpoint, the emphasis in construction must also be placed on reliability. Highly reliable methods of construction lead to successful and, hence, safer operating systems. Methods of construction" should be construed to encompass both materials and vehicle fabrication processes. As described in Chapter 2, the major classification societiessuch as the American Bureau of Shipping (ABS), Lloyd's Register, and Det norske Veritas (DnV)have each established rules for the design and construction of steel-hulled submersibles. To date, all but two tourist submersibles have been built to ABS class, and ABS has the only set of classification rules the Coast Guard recognizes for submersible design. In addition, the American Society for Mechanical Engineers (ASME) has published standards for pressure vessels for human occupancy (PVHO). ABS and the Coast Guard both recognize ASME's standard for PVHO as a satisfactory standard for man-related undersea pressure hulls. *These submsersibles generally operate in waters where bottom depth does not exceed design operating depth. 19
20 The present ABS rules make no major distinction between commercial work submersibles and those used for tourist service. Therefore, there have been few difficulties in classing the limited number of steel- hulled submersibles specifically designed and built for tourist service. ABS is currently revising its rules to add provisions for tourist submersibles. For main pressure vessel (MPV) design (and hence fabrication), ABS will approve designs based either on ASME for PVHO formulas (Section VIII, 1 or 2), or on ABS' own design formulas, which are based on Windenburg's critical buckling formula.7 Design factors of safety (f.o.s.) compare the actual design stress to the yield stress (strength) of the material used. For example, for an f.o.s. of 1.5:1, Material Yield Stress = 1.5 Actual Design Stress An f.o.s. of 3:1 or 4:1 on the material's yield stress is used in PVHO design based on ASME standards, while classification societies such as ABS will accept a minimum of 1.5:1 on the yield stress. The criteria and factors used depend on the designer.8 Also, the method by which ABS inspects the hull during and after fabrication varies with the design philosophy. For example, if an MPV is designed to PlIHO standards, ABS does not require strain gauging of the hull during its hydrostatic (drop) test. The question concerning the choice of the f.o.s. for submersibles raises complex engineering design issues and variables, such as operational depth limitations related to the capabilities of emergency response and the operating environment for a design, as in the case of military submersibles. The committee views the choice of the passenger submersible f.o.s. as an engineering design issue of concern to both the Coast Guard and the classification societies, but outside the scope of this study. Materials To date, nearly all MPVs have been built using steel (either ASTM A516 or the equivalent). The implications are positive, since the specifications for steel are well documented, as are the weld procedures required for fabrication. This experience base with steel tends to enhance reliability. However, some cautions are in order, particularly in regard to stainless steel. Stainless steel (e.g., Type 304 and Type 316) is used extensively in critical systems, including the oxygen system, the compressed air systems, the ballast systems, and all hull penetrators. Some stainless steel formulations have been shown to be susceptible to both stress corrosion cracking and crevice corrosion cracking.9 In the submarine environment, where these systems are continually exposed to salt water, both of these forms of attack on the material can occur. Some austenitic stainless steels (Fe-Cr-Ni-Mo alloysy, which are highly corrosion- resistant, have been used with a high degree of success in marine applications~°~lthough not yet in submersibles. Other materials that are more suitable for this application are the mild steels, the inconel/monel/copper-nickel alloys, and several types of titanium. While titanium is more expensive than other more commonly used materials, it may be cost effective. Aluminum ___= ~ ~ ~ ~ A ~ ~ ~ A_ ~ ~ ~ ~ ·^_ _~ ~ ^ ~ ~ A ~A A · J Utilization of other types of material for the MPV presents additional problems. Aluminum is the second most popular choice for submersible construction material; but, depending on the series of aluminum *In the drop test, a hull is lowered in water to 1.25 times the design depth for 1 hour to test its capacity to withstand hydrostatic pressure. **W. W. Kirk, International Nickel Laboratory, personal correspondence with C. M. Jones, October 1989.
21 selected, it may prove problematic with respect to its weldability. The soon series aluminums, such as 5083 or 5456, are excellent materials in the pre- and post-weld condition. This series has not indicated susceptibility to early onset of stress and fatigue or cracking in the heat-affected zones in the material on either side of the welds. Series 6000 and 7000 aluminums are quite the opposite. Due to its availability and corrosion resistance, 6061-T6 has been utilized extensively in the submersible industry, both commercially and by the U.S. Navy (with disappointing results). The post-weld condition of the material in the heat-affected area has exhibited tendencies to crack under stress levels far below those predicted. This problem has been brought to the forefront by the Naval Sea Systems Command (NAVSEA), which has issued instructions that 6061-T6 will not be used for structural purposes when it must be welded without first being reviewed on a case-by-case basis. It is noted that the ABS allows a maximum stress of 9,000 psi for 6061 aluminum, versus the 8,000 psi the U.S. Navy allows for repair work on existing structures. NAVSEA's requirement for specific case-by-case technical review does seem prudent and presumably would be mirrored by the Coast Guard and ABS in their review processes. Acrylics Other candidate materials include thermoplastic materials such as acrylic. Acrylic spheres have been utilized on five submersibles to date. The thickness required for the Johnson Sea-Link submersibles' new spheres has been increased from 4 in. to 5 in. by ASME's PVHO. The original spheres were designed with a 20-year life expectancy but showed signs of failure within 15 years of service. This problem is a particular concern with plastics or composite materials utilized as pressure vessels subject to external (hydrostatic) pressure. To date, three manufacturers have presented designs utilizing acrylic for the MPV. However, the classification societies have not established rules governing design, construction, and testing of this type of hull. To further compound the problem, HYCO Technology's ARIES, SEA VIEW, and COMEX's proposed SEABUS are proposed as cylindrical (not spherical) pressure vessels. No relevant test or operational histories exist here, since the history is based entirely on spherical shapes such as the U.S. Navy's NEMO submersible or the Johnson Sea-Link research submersibles. If large area or extensive use of acrylics as a principal structural material is to be approved and rules for it written, the analytical and testing work will have to be contracted and paid for by the organization requesting the rules. It will be expensive and might require test-to-failure of a full-sized hull section, as well as development of extensive fatigue data. Even if acrylic hull standards are developed, the potential market for this type of submersible may be greatly diminished by the time this can be accomplished. Me best operating locations may be already taken by steel-hulled submersibles before the first acrylic hull submersible enters tourist service. There is another related problem with immediate significance. Present rules, based on PVHO standards, call for replacing all acrylic viewports after 10,000 diving cycles or 10 years in service (unless a potential designer/fabricator can demonstrate, via testing, that either or both criteria should be extended for their application). At an average cycle rate of nearly 2,000 dives per year, the viewports would have to be changed after 5 years. (Private research is being conducted by Sub Aquatics to extend the time period specified by those rules.) If the 10-year or 10,000-cycle requirement were extrapolated to massive acrylic hulls, the entire hull would have to be replaced. This cost factor alone might be reason enough to stop any further consideration of this hull material. The classification societies and the certificating agen~the Coast Guardmust be specifically careful to ensure the adequacy of design and particularly cautious in their interpretation of test data, since the ARIES, SEA VIEW, and SEABUS (Comex) applications break new ground in the utilization of acrylic material. As noted earlier, materials such as steel, with an extensive history of use in these areas, exhibit Letters from Commander NAVSEA to Lockheed Advanced Marine Systems: paragraph 2, letter dated May 17, 1985; and paragraph 4, letter dated April 1, 1988.
22 their individual weaknesses in ways that are well understood; newer materials such as acrylic or composite materials will likely exhibit new problems. For examples some experts ascribe little validity to scale model testing of acrylic hulls, believing that acrylic does not scale representationally. A U.S. Navy technical expert has suggested to PVHO that designers and builders of acrylic-hulled tourist submersibles build a minimum of two scale models (approximately 1/4 or larger) for testing. Under this proposal, one model would be tested to destruction to validate the design calculations. The second model would be subjected to the maximum number of pressure cycles that can be withstood through its destruction. In this context, a pressure cycle would replicate a dive cycle. The maximum number of cycles derived from the test would then be used to determine the replacement interval. In contrast to the opinion mentioned above, this expert believes that acrylic scales very well, with scale effects being reasonable. These contrasting views simply reflect the lack of wide experience with these materials and the uncertainty facing the regulatory agencies. Flammability Considerations The selection of materials associated with electrical equipment and wiring and the interior of the submersible should include consideration of flammability properties, including ease of ignition, flame spread, and composition of combustion products. The design should proscribe the use of materials that exhibit low flash or fire points. In addition, any potential fire should be restricted to minimal, well-defined, isolated areas within the submersible without propagation paths. Material selection should minimize the toxic hazard associated with combustion products. A National Aeronautics and Space Administration Handbooks provides criteria governing materials selection, evaluation, and control. The Coast Guard will find this handbook useful as a guide. Structure The predominant method of hull construction used in U.S. operational submersibles is to reinforce cylindrical shell sections with external ring stiffeners. The rules for designing main pressure vessels using ring stiffeners are well known. The external ring stiffener approach maximizes the internal volume available to passengers without sacrificing the additional strength required for the stiffeners. Some European designs offer a variation by allowing the stiffeners to penetrate the continuous shelf plate. In the tourist submersibles observed by the committee, the stiffeners did not penetrate through the cylindrical shell. It is well that they did not. Such penetration could present numerous problems with regard to maintaining a true cylindrical cross-section, and allowable local deviations from circularity could be exceeded. Additional problems presented by allowing stiffeners to penetrate the MPV shell occur in welding. The number of welds in the MPV would increase by a factor of two (at the minimum) with penetration. This doubles the possibility of failure due to poor welding, shell pressure cycle. fatigue. early onset `,f hil~klina Or ae.ner~1 instability. Materials selection is also crucial for maintaining structural integrity. For example, 6061 aluminum, which is susceptible to cracking after heating, is widely utilized as exostructure material for tourist submersibles. The exostructure is the framing around the MPV that supports the superstructure (deck and conning tower) and encloses the ballast tanks. When the tourist submersible surfaces to unload and load passengers, there is always contact (impact) with its tender vessel (the ferry boat). Depending on wind and wave action and the relative position (to the ferry boat) of the submersible when it surfaces, the impact forces can replicate minor collisions. It is possible that a collision during operations could exert the force required to buckle the submersible's exostructure and, through it, deform the pressure hull. The damage might appear minor during a visual inspection, but could be more critical in reality. ~ ~ 7 ~ ~ ,7 ~7^ a_ ._. a. *Dr. Jerry Stachiw, Head, Materials Technical Staff, Naval Ocean Systems Center, San Diego, CA (personal correspondence).
23 If the exostructure, as it collapses, damages the shell of the MPV by causing it to deform permanently, catastrophic failure is only one step away. With a permanent set (deflection) in the MPV shell, the circularity of the shell is compromised. If the dimensional irregularity caused by this impact is greater than the percentages allowed by classifying agencies, a classical buckling failure could occur if, for instance, the submersible were to dive to a depth close to its design (or operating) depth. The amount of allowable deviation from the design diameter (or dimensions) is small; ABS allows 1.0 percent of the design diameter. This possibility, coupled with various construction methods (e.g., some European designs) places special emphasis on the need for close inspection during annual surveys. At present, ABS rules state that during special surveys (done at approximately 3-year intervals) the surveyor Could call" for critical dimensional checks. Since tourist submersibles are drydocked on land during mandated surveys every 18 months and are basically disassembled for inspection, it would be a relatively simple task to check the hull circularity and verify that it is within the allowable limits. Pressure Cycling Tourist submersibles operating in resort areas such as Grand Cayman and the Virgin Islands can average 6 dives per day, 6 days a week. This yields a total of 1,872 cycles per year, which is an extremely high number compared to experience with work submersibles and Navy deep submergence rescue vehicles (DSRVs). Military and industrial submersibles would not see that number of cycles in their lifetime. For example, the two U.S. Navy DSRVs average less than 100 dives (cycles) per year. Increased pressure cycling causes structural fatigue to appear at earlier stages than the expected design life. This was evidenced by an Aloha Airlines 737, which lost its forward cabin roof inflight in a 1988 accident. Because these aircraft island-hop, they undergo a higher than average number of pressure cycles during their operational life. The same case can be made for submersibles. In this connection, it is noted that ABS requires only 1 or 2 cycles of hydrostatic testing, while the Naval Sea Systems Command requires a minimum of 10 cycles. While such testing is adequate to verify basic design parameters, it does noteven at 10 cyclesaddress the fatigue problem. Increased cycling in operation will eventually lead to fatigue failures, which emphasizes the need for intensified inspection and testing of vehicles that have been in service for several years. There is a sufficient data base on steels to predict performance with a given number of pressure cycles. I~here is not an adequate data base for predicting performance of new materialsparticularly acrylics. Redundant and Backup Systems The safety of a system is partly a function of its reliability; that is, that the subsystem or component will be available and operate on demand. System reliability, in turn, can be partly a function of system redundancy or backup availability. While a less than perfect subsystem or component operational reliability is usually acceptable in the commercial working industry where human safety is less at risk, the tourist submersible industry requires a significantly higher reliability for each subsystem. This can be achieved only by having a substantial reserve or a dedicated emergency backup system. Strict redundancy is accomplished by duplicating critical components, such as with an extra set of gauges, communications gear, and secondary sources of compressed air that are directly accessible. Other systems may be provided with a backup capability different in design from the primary system. For example, there is a requirement to carry sufficient solid ballast (usually lead or concrete) to be utilized as drop weights. This is needed in the event that the submersible is grounded by flooding of its largest pressure vessel, aside from the main personnel capsule. The drop weight system on one submersible visited by the committee (ATLANTIS class) is activated by a hand-driven hydraulic pump, and this activation arrangement may be common to other tourist submersibles now in operation. In current designs that were observed, there is no provision for a backup in case the hydraulic pump fails. Further, the system can be activated
24 only from inside. Other methods of drop weight release, utilized in work submersibles, are pyro-ignited links, cable cutters, etc. The committee observed that the drop weight system on ~ . . . . . .. . . , v , the ATLANTIS submersible was not actually tested in situ; that is, the weights were not dropped. Since the first response to an emergency during diviners pointed out to the committee by several operatorsis to surface immediately, the ability to release ballast rapidly is critical, either as water blown out of tanks or as release of weights. Full testing of the drop weight system at sea in at least one vessel of each class would provide significant improvement in safety assurance. In addition, further assurance would be provided by having a manual backup for dropping the weights from inside the vessel. In submersibles whose operating depth is 150 feet or less, the drop weight system should also incorporate a feature that allows divers to jettison the weights from outside the hull. Stability If a tourist submersible grounds due to a system failure and the drop weights must be jettisoned, the Coast Guard does not specify a minimum required stability. The minimum stability required to recover a submersible in its normally upright (trim) attitude is governed by the Metacentric Height (see Appendix A). For a submersible to remain upright while submerged and thus exhibit positive stability, the center of buoyancy must be vertically located above the center of gravity while submerged (see Appendix A, Enclosure 2, Figure 1~. The center of buoyancy can be below the center of gravity while the submersible is on the surface if the Metacentric Height is positive. Stability is an important parameter; its adequacy could be verified in a simulated (test) emergency ascent. The test envisioned would both test the ability of the drop weight system to function properly in water and verily that the stability required is adequate to guarantee that the attitude of the submersible during recovery is upright and not severely inclined. Emergency trim control could also compensate for the possibility of unusual loading in case of an event that would cause passengers to retreat to one location in the hull, and a simulated test could include this loading variable. Quality Control While the submersible construction observed by the committee was conducted with quality control (QC) standards, it is essential that requirements to assure quality standards be established and maintained for the industry as a whole. QC standards and detailed documentation are important for ensuring safety. Based on site visits and observations, the committee is concerned about the present level of recordkeeping by the manufacturers. The Coast Guard should formalize the QC system, establishing what records must be kept and who should to keep them. Conclusions and Recommendations Regarding Design and Construction The method of tourist submersible design used within both the domestic and international industry is still maturing; hence, there are inconsistencies in industry design standards between those of the certification and classification agenciese.g., factors of safety utilized in pressure vessel design, design parameter terminology, extent of data required during post-construction hydrostatic tests, and selection of materials used. The Coast Guard should reevaluate the areas of inconsistency with ABS (indicating a lack of adequate definition) that exist in tourist submersible design procedure, material selection, and testing. Discussions between ABS, Lloyd 's Register, and DnV could be of value in the course of the Coast Guard 's assessment.
25 System redundancy is prevalent in most designs. In one notable exception however, the drop weight systems, which are critical in the event of a pressure vessel (non-main pressure vessel [MPV]) flood, did not have a backup externally operable release on the vessels, the committee observed. The drop weight system for each submersible design should be tested in sim by vessel designer and manufacturer. The simulated emergency ascent test should account for the possibility of passenger overloading in one area of the submersible. The present main pressure vessel (MPV) design used for tourist submersibles in service in the United States uses a ring-stiffened cylinder where the stiffeners do not extend into the pressure vessel. Submersibles that use other MPV designs may become available for use in U.S. waters. In some cases these submersibles may utilize stiffeners that penetrate the MPV shells imposing additional strength and maintenance problems. This potential concern places greater emphasis on inspecting and maintaining hull dimensional accuracy. To ensure that structural failure due to shell buckling or general instability does not occur from either repeated impacts (collisions) with the submersible's tender .(feny boat) or because of problems inherent with designs that allow stiffeners that penetrate the main pressure vessel, greater emphasis should be placed on post- constn~ction inspection and critical dimensional checks during 18-month surveys. Special attention should be given to symptoms of fatigue. Failure in aging submersibles also should be a part of the survey. There is a serious lack of information on the performance of acrylics in repetitive loading such as is experienced on tourist submersibles. Expanded testing is needed to validate design cntena. This testing should be done by the designer or manufacturer proposing to use acrylic materials to the extent that they may pose signif cant questions about the integrity of the main pressure vessel. Both the Coast Guard and ABS should provide guidance in the planning and conduct of such tests. LIFE SUPPORT SYSTEMS Air Supply/Regeneration ABS rules require that the onboard air system be sized for 72 hours plus a normal dive.) The normal dive is usually counted as one full day. Adequate overcapacity in air required to blow the ballast tanks and in onboard battery capacity (for emergency power) is required and was built into the systems observed by the committee. Air tank capacity for blowing tanks must be adequate for at least the number of dives projected between recharging opportunities. In addition, some excess should be provided for emergencies. This air provides the primary means of returning to the surface. The air system design (valves, strainers, etc.), as well as total capacity, therefore merits special attention. The onboard carbon dioxide (CO2) scrubber system (Soda Sorb and blower motors) is essential in maintaining the quality of the air. According to ABS rules, the allowable CO2 limit is 0.5 percent of gross. Redundancy is provided by arranging alternative circulation patterns in the event of scrubber system failure and by providing extra capacity for scrubber reactants. Adequate sizing in the basic scrubber designs is essential; this capacity must be tested and any deviation from classification requirements observed and rectified. A problem of this nature was evidenced during a joint Coast Guard and ABS inspection of a tourist submersible in Guam on April 6, 1988. In that case, the CO2 level during the second scrubber test
26 was greater than 0.5 percent (due to improperly installed fans); but the result was approved based on first test results that were less than 0.5 percent. Fire Suppression Fire suppression in confined spaces, where humans are present, is difficult. Heat build-up is very rapid, especially with closed hatches, in which case the temperature can pass 100°C in less than 40 seconds. Under these conditions, smoke rapidly obscures v~sibilin,r, hindering both fire-fighting and escape. Carbon monoxide (CO) is also an immediate problem, as it affects decision-making capability long before it affects motor functions. The present Coast Guard requirement is simply for an approved, portable fire suppression system. The current systems in use on submersibles are an interim response in the absence of a specified fire-suppression system. The requirement for a fire suppression system illustrates a problem in that all the available options have various safer and health concerns associated with them. For example, use of the gas Halon is currently the prevalent method; the ATLANTIS class of submersibles uses Halon 1301 as a fire suppressant. This gas is effective for the purpose but has several drawbacks, including decomposition resulting in acidic byproducts, limitations on breathability for humans (10 min. at 6 percent maximum), and the fact that the Halon cannot be removed by carbon filters. Despite these deficiencies, Halon 1301 is the best available short-term fire-suppression alternative when the time to surfacing is only a few minutes. Halon is a fluorocarbon and for that reason will probably be phased out within a few years. - The committee noted that nitrogen is an alternative fire suppressant currently under study by the Office of Naval Research and the Naval Research Laboratory. Nitrogen is nontoxic and merits consideration as the primary fre-suppressant gas. Like other alternatives, nitrogen has inherent control problems and will require highly engineered control systems. Emergency Breathing Apparatus Another illustration of the differences and complexity of submersibles compared to surface vessels is the need for emergency air purification or air supply. ABS rules for submersibles require emergency breathing life support for the duration of the dive or two hours, whichever is longer. (In the case of tourist submersibles, the two-hour requirement pertains.) Currently there is no Coast Guard standard for devices to meet this need. There are two aspects to the problem of emergency breathing apparatus: viz., providing emergency breathing for the passengers and crew (pilot and attendant). Emergency breathing requirements for passengers, in the event of a hire or loss of power with attendant failure of the air regeneration system, are usually met via individual units. One builder examined by the committee elected to solve this problem by providing an MSA (Mine Safer Appliances Company) rebreather device designed to remove carbon monoxide by oxidation and with the ability to remove certain other particulate materials via a filter. The device does not, however, remove other contaminants besides CO and will not support life if oxygen levels are too low or if for any other reason the atmosphere itself will not support life. The crew is provided with a full face mask, air fed, so that they breathe off the high-pressure air system. Thus, they have a life support system that will last as long as there is high-pressure air. The MSA self-rescuer may well be the best available solution to the problem of emergency breathing for passengers in the event of a hire. However, since the device does not provide air, it does not *Rapidraft Letter from Commander USCG Section Marianas to Commander USCG Washington, DC, April 19, 1988, regarding the completed quarterly control verification exam of the MARIEA I submersible. **This is a problem of much broader scope than tourist submersibles, as Halon is used as a fire suppressant on many vessels.
27 appear to satisfy the rules as they exist at this time. The alternative of using air-fed masks for passengers presents a problem of pressurization of the compartment; sufficient exposure to the pressurized atmosphere would require passengers to be treated for decompression upon rescue, although this would not be a problem if the submersible is able to execute the prescribed emergency ascent to the surface within a minute or two. Also, air-fed masks are not portable, perhaps making evacuation of passengers on the surface more difficult. This issue needs further attentionparticularly as to whether the MSA rebreather meets both the letter and intent of the ABS rules. Perhaps the most troubling possibility is the risk of fire. The atmospheric contamination in such an event is extreme and creates an immediately life-threatening situation. Therefore, every possible precaution must be taken to avoid fire, to minimize the toxic byproducts of any fire should one start, and to provide a breathing apparatus that will protect every passenger from inhaling toxic contaminants and smoke. Personal Flotation Finally, the committee observed that the inflatable life jackets used on ATLANTIS are not Coast Guard approved. Indeed, the Coast Guard has not approved any inflatable life jackets for such service, in part because no user has been willing to pay the cost of testing and approval. (Noninflatable jackets are not compatible with the storage and access constraints of submersibles.) The use of inflatable jackets in this context is in line with airline practice, and while not specifically approved for general use they have been accepted in current submersible operations. Moreover, aircraft-type inflatable jackets lend a familiarity derived from air travel and therefore may provide some psychological benefits in understanding and accepting their presence and use. This approach to personal flotation is compatible with an operating profile in which an attending surface craft is present at all times. Conclusions and Recommendations Regarding Life Support problems. The gas Halon, used in current fire-suppression systems, presents possible safety and health Because nitrogen gas is nontoxic, consideration should be given to the development of a nitrogen-based fre-suppression system, as a possible replacement for the Halon-based systems. There is an indication that CO2 absorption systems (scrubbers) used on some tourist submersibles have occasionally been unable to maintain CO2 below required levels. These systems either lacked sufficient capacity or were incorrectly installed; these deficiencies were overcome by reversing the airflow pattern. The Coast Guard should require that CO2 absorption systems have adequate capacity and redundancy, in accordance with a hazards analysis, and should enforce this requirement in design and periodic inspection. Failure of this system could result in injury or fatality to passengers. The current rebreather devices for providing emergency air supply for passengers do not satisfy ABS rules and are not adequate protection against some types of atmospheric contamination. The Coast Guard, in cooperation with ABS, should promulgate standards on emergency breathing requirements and systems for tourist submersibles. All participantskoast Guard, ABS, and the users-should agree on the application of those rules.
28 Inflatable life jackets used aboard tourist submersibles are not Coast Guard-approved, although the Coast Guard is currently accepting them. The Coast Guard should consider testing and approving one or more inflatable life vests. INSPECTION The U.S. Coast Guard routinely conducts periodic inspections of surface vessels to ensure compliance with safety and pollution abatement requirements as defined by Coast Guard regulations. These inspections are conducted by personnel at the local level. They are effective because of the knowledge and training of the inspectors and because application of the rules results in similar systems and equipment on most vessels. Years of experience with surface ships has allowed the Coast Guard to develop an inspection program under which the local inspectors are sufficiently knowledgeable to work directly between the rules and the hardware without interpretation difficulties. Inspection of tourist submersibles is not so straightforward and routine because of the relative newness of these vessels and the relative lack of experience with them on the part of Coast Guard inspectors. A number of vehicle inspections take place at different stages in the life of a tourist submersible. They include: construction and post-construction inspections; periodic certification inspections and reinspections; and quarterly control exams (for foreign-built vessels). Construction Inspection During MPV construction, which could last 2-3 months, the classification society is on-site weekly. In addition, spot (surprise) inspections occur at random intervals. Almost every inspection taking place during construction includes participants from both the Coast Guard and the classifying society (e.g., ABS). This phase of construction is critical, since the main pressure vessel is the single most important structural component of the submersible. The entire construction (or assembly) period could last as long as 12 months. The final assembly (after all subsystem components are assembled) takes 5-6 weeks. During the vehicle construction period there are 25-40 inspection visits. When Coast Guard personnel are not available they will utilize the information gathered by the classification agency. Among the MPV tests witnessed or performed are critical dimensional checks, material tests, hydrostatic tests, and functional tests. Before the hydrostatic (drop) test is done, radiographic and ultrasonic testing is performed. After the hydrostatic test, ABS performs another dimensional check and non- destructive inspection (dye penetrant and magnetic particle) examination of all welds to ensure compliance with the specified procedure. Periodic Inspections Organized periodic inspections are an important aspect of maintaining safety in tourist submersible operations. In general, all vessels carrying passengers must have on board a valid Coast Guard certificate of inspection (COI). A COI is issued by the cognizant Coast Guard Officer in Charge of Marine Inspection after an inspector has examined the vessel and determined that it is in satisfactory condition and fit for the service for which it is intended and that it complies with the applicable regulations. (This is the "certification" process described in Chapter 2.) A COI for small passenger vessels less than 20 meters (65 feet) in length (all . . ~ . . . ~ , =, `~ existing tourist submerses are In this category) is valid for three years. At least two reinspections must be made within the triennial period. When possible, these reinspections will be made at approximately equal intervals between triennial inspections for certification.
29 ABS also conducts annual and special surveys of tourist submersibles. During the annual surveys the hull is usually examined for significant damage. The air and hydraulic systems are inspected, the vessel is repainted, and the superstructure is repaired. There is no requirement in ABS rules for a critical dimensional check, although special ABS surveys that occur every three years "could call" (according to the guidelines) for such checks. In the committee's view, annual and special surveys should include dimensional checks to confirm the roundness or circularity of the MPV. Tourist submersibles represent a new challenge for the Coast Guard inspection system. Although these vessels have much in common with surface craft, the major differences may mean that the expertise of local inspectors could be insufficient to ensure adequate periodic safety reviews. In addition to the unique aspects of submersibles as a group, it is also reasonable to expect substantial differences between the designs of various manufacturers. In time, some level of standardization is likely to occu~particularly in safety-related areas; but at present, each vessel could meet the regulations in different ways, so that a valid inspection may require detailed design knowledge on the part of Coast Guard inspectors beyond what the present system provides. The U.S. Navy's deep diving submersibles and saturation diving systems present a similar situation. Inspection and certification of these facilities are done by a specialized group of auditors backed by a second _ r _ _ ~ at ~ _ · ~ i. ~ ~ ~ _ ~ ~ . ~ ~ . .~ ~. · , _ group ot specialized engineers. Inspections by local personnel are conducted but they are preliminary to a visit by the NAUSEA 92Q Certification Audit Team from Washington, D.C. This procedure works well for the Navy, but is contrary to the Coast Guard's proven local inspection procedure. Continuation of local Coast Guard inspections appears desirable and reasonable since local inspectors can provide information about the local operational environment. The use of local Coast Guard inspectors can be satisfactory if the inspectors can be given adequate information on the submersible to be inspected to make the transition from the regulations to design and operational details. The participation of Headquarters personnel expert in submersibles (possibly with input from Marine Safety Center personnel) would ensure that effective, thorough inspections were made. The current inspection procedures appear to be adequate to maintain safe operations on systems that are essentially new. The committee has reviewed extensive experience with Navy-controlled deep- diving systems and several research submersibles and has developed an outline of inspection requirements, which is intended to help the Coast Guard and ABS in their review and development of their respective inspection procedures, as tourist submersible operations expand and become more routine. In brief, the committee believes that the designs of tourist submarines and their method of onerntion are Ninny enc,~oh to retire rletnileA knowle~lne an the nart of an in~nector to ensure an adequate safety review. Flus requirement can be met by providing a detailed inspection plan generated as part of the initial approval process and encompassing the program's equipment, operational procedures, and training programs. The inspectors should rarely be required to make technical judgments. Instead, the inspection plan should contain the necessary pass/fail criteria for each approved item or system. The specific points and criteria contained within the inspection plan should be derived from a formal hazard analysis (see Chapter 4~. As with the initial system approval, periodic inspections represent a major effort on the part of both the Coast Guard and the operators. Great care must be taken to limit the inspection criteria to those items and procedures of real importance to the safety of the passengers and operational personnel. In addition, operational inspections should be part of the inspection protocol. Inspections conducted while the vessel is moored at the dock on the surface capture the vessel in a passive mode; an active operational mode is more indicative of the actual status of the vessel. The committee's suggested inspection plan is described in greater detail in Appendix B. For foreign-built vessels, which are outside the Coast Guard certification procedure, quarterly control verification exams are conducted jointly by the Coast Guard and the classifying agency. The Coast Guard issues a certificate, CG-4504, Control Verification for Foreign Vessels to those foreign vessels it finds to be in compliance with SOLAS regulations. This form of inspection is essential for determining whether the vessel maintains its classification or certification. Because the SOLAS regulations for passenger vessels apply to ships on international voyages, their applicability to submersibles (which do not make such
30 voyages) is highly questionable. However, they do provide the agencies with a means of spot-checking the subsystem effectiveness and reliability of foreign-built vessels. Conclusions and Recommendations Regarding Inspection Inspection, both initially during construction and then continually throughout the operational life of the vessel, is extremely important for ensuring the safety of passengers and crew. The inspection plan for a given class of submersible should be defined ear) in the design phase, with consultation taking place among the Coast Guard, the classif cation society, and the prospective operator. The development of inspection plans and criteria should be based in large part on the results of a formal hazard analysis (see Chapter 4). Continuation of local Coast Guard inspections appears desirable and reasonable. However inspectors need to have detailed technical information and knowledge regarding the specific class and In they are inspecting. The inspection plan can provide much of this information. ~ ~ . At, . Personnel from Coast Guard Headquarters and the Marine Safety Center should be available to provide additional in-depth fantiliari~ with tourist submersibles systems and operations. The inspection protocol should include functional tests of critical systems, which must first be defined in the inspection plan. Hull circularity is considered to be an essential factor in this plan. Specific criteria for passing or failing each inspection point must be clearly established and firmly adhered to. Inspectors and COTPs should not be permitted to deviate from the established criteria, except in the case of specific items that are identified as such in the inspection plan (and usual with a requirement for.~n~rif~ concurrence at higher levels). , , ,