Innovations in the world of underwater vehicles are coming at a furious pace. New build, as well as individual vehicle component manufacturers, are feeling the pressure to refine their products to match the needs of the deepwater oil and gas industry.

Remotely operated vehicles (ROVs) are already established tools for tending to deepwater developments, and their abilities are becoming ever more specialized. Autonomous underwater vehicles (AUVs) are relative newcomers to the field, but the reliability and technical ability of these vehicles is on an upward curve. UnderWater asked the leaders in the industry to recount some of their more interesting vehicle development projects. The resulting case studies indicate that the future is now.

Sonsub’s Innovator
As far back as 1994, Sonsub International realized that its offshore operations experience and engineering solutions effected significant changes on new vehicles. Not yet a player in the new build business, Sonsub was thus channeling product development expenditures and efforts in the open market without ultimately enjoying the benefits of exclusivity. In an effort to address this issue, Sonsub undertook a New Build Program designed to produce a vehicle system that would set a new industry standard. A logical evolution of existing technology, the system utilizes proprietary developments and concepts already applied to present vehicles. The new system has been named the Innovator.

In 1995, Sonsub set up an internal ROV tracking system as part of their Continuous Improvement Initiative (CII) program, involving a systematic tracking of ROV faults, and a training program ensuring that all relevant personnel were fully conversant with the system. This would serve to effectively minimize human error occurrences. All failures were categorized into four types: design, wear, accident, and maintenance.

A study of analysis results identified two main issues: First, the majority of failure modes were attributed to the design category; the call for improved component selection and engineering increased ROV reliability. Also, accurate diagnostic and condition monitoring of the subsystems produced considerable reductions in failures, allowing scheduled maintenance to be carried out during non-critical activities. Logistic and component repair costs were immediately decreased. The results of this analysis were continuously incorporated in the New Build Program, making The Innovator a highly reliable, easily operated and cost effective ROV system.

The Innovator system is a high specification work class system rated in excess of 11,480 feet (3,500m) water depth. It has been built for construction support, deepwater drilling, and cable installation support. The ROV has a shaft output power of 150 HP. The electric motor features a dual shaft that facilitates the connection of the work packages to the ROV. The vehicle is also provided with a triple armored umbilical, which, with a 400 Electric HP capacity, easily allows work packages to be powered from the surface. Sonsub International’s New Build Program currently totals eleven units. Two units are now headed for the Gulf of Mexico, and nine more will enter the offshore market in the future.

ROPOS
The limited funding available for scientific vessel and submersible operations increases pressure to maximize the results from every dive. Versatility and adaptablity are vital. A typical multidisciplinary expedition includes geologists, biologists, and chemists from several institutions in several countries. Every scientist needs specific data or samples, often requiring the use of unique equipment. Each needs basic navigation information, imagery of sampling sites, and other data to provide a context for the samples or data.

The design of the new Remotely Operated Platform for Ocean Science (ROPOS), operated by the Canadian Scientific Submersible Facility (CSSF), incorporates 20 years of experience supporting multidisciplinary scientific research, first with the Pisces IV manned submersible and then the original ROPOS. The new ROPOS was built after the original was lost when its support vessel lost propulsion during a violent storm.

Users of the system have been closely involved in the ongoing development process, including scientists from more than a dozen universities in Canada, the U.S., Germany, the U.K., NOAA, the Woods Hole Oceanographic Institution, the American Museum of Natural History, the Field Museum, the Monterey Bay Aquarium Research Institute (MBARI), the British Antarctic Survey, GEOMAR, and the Canadian government.

ROPOS exists to help its users gather specific information and samples to improve knowledge and understanding of the seafloor, its inhabitants, and the behavior of seafloor installations.

Sampling can be simple, like collecting small rocks or animals, or demanding, such as collecting hot (up to 300 degrees C) fluid from hydrothermal vents for analysis of chemistry and dissolved gasses. A delicate touch is required to avoid damaging soft rocks or creatures. Samples may be microscopic, such as bacteria, or as large as an entire 800kg hydrothermal vent chimney. Devices may be general purpose, such as the manipulators and suction sampler, or specialized, such as a drill for taking rock cores. The samples must be brought to the surface in good condition.

The New ROPOS
The new ROPOS successfully addresses these needs. The basic vehicle is a 30/40-horsepower electro-hydraulic, tethered ROV designed and built to operate at depths up to 16,400 feet (5,000m) and based on a proven platform, the International Submarine Engineering (ISE) HySub. A Kongsberg-Simrad-Mesotech color imaging sonar, color (3-CCD) and Insite Systems low-light silicon intensified target (SIT) video cameras, and seven- and five-function manipulators are mounted on the front of the vehicle.

The whole system is portable, and can be shipped to meet the support vessel in a convenient port. ROPOS has operated from seven different support vessels from four countries, and has mobilized as far afield as Capetown, South Africa.

The high-quality visual information needed for scientific operations is provided by a Sony DXC-950 broadcast quality video camera with a 16x zoom lens, supplied by Kongsberg-Simrad, and Deep Sea Power and Light high-power lights on the front of the vehicle.

Scientific sampling operations and instrument deployment demand a great deal of manipulative dexterity and the ability to grasp small to medium sized objects. The jaws on the ISE manipulators have been replaced with smaller double-sided jaws constructed of stainless steel. These allow better manipulation, but also have a much larger opening so that larger objects can be grasped. A clamshell-shaped sampler can replace the jaw end effector of the manipulator on either arm; it excels at sampling soft or fragile items.

The manipulators are rated for 300-kg. lift at full extension. Manipulator feedback has been upgraded to use solid state, Hall effect sensors, giving better reliability with fewer seals and no moving parts. A two-axis sensor for the ROPOS manipulators provides better linearity and more consistent feedback.

Software Engineering Associates developed a version of its SeaScape Precision Navigation System to meet the needs of ROPOS users. This PC-based, long baseline acoustic navigation system delivers 3-D real-time positioning and vectoring information for ROPOS, its cage, and the support vessel to ROV pilots, science teams, and ship control personnel.

Several other devices have been developed to meet common needs. One is the variable speed, reversible suction sampler, an original design that can also be used for jetting. Samples of water, bacterial mat, tube worms, clams, small fish and crabs, sediments and small rocks are collected in eight two liter jars with a filter mesh on the outflow.

A rotary sample tray for large geological and biological samples takes four to eight compartments. A hydraulically actuated "biobox" with larger compartments has thick Lexan walls to provide thermal insulation for temperature-sensitive organisms sampled in deep water. An "elevator" is also used for deployment and recovery.

The scientific telemetry system multiplexes up to seven bi-directional RS-232 channels, permitting real-time communication with many instruments. An external junction box is provided for the user interface, providing analog and digital input and output, DC power, and access to the RS-232 ports. A very wide variety of equipment has been successfully interfaced with the scientific telemetry system, including an Imagenex downward looking scanning sonar used for detailed surveys of bottom topography.

The hydraulic power packs on the new ROPOS provide eight separate hydraulic functions to support the steadily increasing number of hydraulically actuated scientific tools and frequent need for several tools on a single dive. A rock coring drill and a Stanley Underwater Hydraulic Tools rock-cutting chainsaw have been successfully used.

The SIPPER
The task of capturing images of microscopic marine particles such as zooplankton or suspended sediment has relied for years on using oceanographic imaging systems with analog video or film recording methods. While generally adequate, video resolution is limited by the two-dimensional detector array used. High-resolution images are attainable, but are typically produced with a reduced imaging area.

The University of South Florida’s Center for Ocean Technology (COT) has developed an imaging system based on high-speed digital line scan cameras to address the shortcomings of the current technology. This particle measurement instrument allows rapid data acquisition and analysis for large volumes of water. Known as the Shadowed Image Particle Profiling and Evaluation Recorder (SIPPER), it uses two high-speed line scan cameras and two collimated laser generated light sheets to image suspended particles in sampled water. Transient particle shadow images are captured in two dimensions to permit maximum characterization. The system uses two EG&G available high-speed line scan cameras with arrays containing over 4,000 pixels, a pair of commercial line generating semiconductor lasers with additional optics, and a custom designed 10-megabyte per second data storage system for imaging and recording. System resolution is 23m per pixel in the across flow axis by 69m along the flow axis, at two knots water velocity.

This novel imaging system makes these cameras ideal for use in mobile oceanographic imaging applications that require high-resolution outputs. The digital data imaging system provides a continuous record of all particles passing though the sampling tube, and stores views from two orthogonal axes. Lower resolution images are displayed in real-time to allow the user to optimize imaging parameters.

High-resolution images are stored digitally, thereby reducing the processing time required before computer analysis. A different digital line-scan camera is used for each axis to allow optimization of the resolution at different water velocities. The array in one axis has 4,096 pixels with a maximum scan rate of 9,300 lines per second. The other axis contains a 2,048 pixel camera that is capable of scanning up to 15,000 lines per second. At two knots over the entire sampling area, the effective pixel size for one axis is 46 x 69 microns, and 23 x 137 microns for the other. Resolution in the flow direction improves with slower water velocities. This system can sample 34 cubic meters per hour. The current data storage system permits collection of up to 80 minutes of data before offloading via a high-speed ethernet network.

The SIPPER system has been successfully deployed on towed and autonomous underwater vehicle platforms. The instrument is controllable as an Echelon LON network node via a single twisted pair or RF link, and may be configured in real time or before deployment.

H Scientific’s Remote Guidance
Recent surveys of remote-controlled underwater vehicles list over 90 manufacturers and about 300 vehicles worldwide for tasks ranging from survey and inspection to cable burying. Whatever the task, all of these vehicles have the need to position themselves with some degree of accuracy. This involves two functions: measuring the actual vehicle position, attitude, and velocity; and controlling the thrusters and actuators to drive the vehicle at the desired velocity to the desired location.

This suggests a modular approach to the design of the vehicle and its essential subsystems. Already, many of the key components for this approach are available off the shelf: thrusters, actuators and powerplants are commercially available, as are navigation devices and systems including a standard interface (NMEA 0183) which has been inherited from the surface ship navigation industry and makes it possible for the autopilot and remote control system to use a standard electrical specification and communications protocol. The protocol standardizes data for navigation sensors, Electronic Chart Display and Information Systems (ECDIS), autopilots, and the like. Nevertheless, the standard makes it possible for most instruments to communicate with one another without substantial modification.

Until now there has not been a commercial, off-the-shelf system which can provide the remote-control function in a generic manner. The tendency has been for vehicle manufacturers to develop systems for each vehicle design, often incorporating it within a central "mission computer" responsible for on-board as well as positioning and navigation functions.

The Solution
Faced with the challenge of creating a generic remote control autopilot that is applicable to almost any vehicle configuration, H Scientific Ltd. developed a solution based on a new range of self-tuning autopilot algorithms originally developed for surface vessels. These include simple heading control, speed control, track keeping (waypoint following), point and area hovering with and without operator-imposed heading setpoint, controlled drift, and rotation on a hoverpoint or within a hover area. The autopilot system uses NMEA 0183 inputs, and is also equipped with an efficient serial data protocol which allows data and commands to be sent both ways between the autopilot unit and a remote control workstation. The remote control system allows the underwater vehicle to be controlled manually or by sending a set of parameters for the autopilot, while monitoring the position and vehicle status on a graphical display on the remote control workstation.

The AUV Simulator
The system was developed and tested using an AUV simulator, allowing the user to configure the vehicle to match the layout and physical characteristics of a particular vehicle. The navigation instruments can also be configured to reflect the instruments on board, including the data rates, latencies, and error characteristics. Having configured the craft, the user may drive it manually using sliders to control all the individual actuators and seeing the vehicle responses on the simulator screen. The vehicle motions may be viewed from different angles, so as to observe motions in all six degrees of freedom.

Before an autopilot can be applied to the vehicle, it is necessary to make connections between the autopilot outputs and the vehicle actuators. On a surface ship, this is a relatively simple task. There is usually one actuator which moves the rudder, which in turn controls the yaw rate and heading. On an underwater vehicle, control may not be so straightforward. For example, a vehicle might be equipped with two thrusters mounted aft and two horizontal hydroplanes near the front. In this case, yaw and heading are controlled by differential thrust and forward hydroplanes control the vehicle attitude. In the simulator, this configuration information is provided by the user and is used to ensure that the control signals are routed to the appropriate combinations of actuators.

A key feature of the autopilot algorithms is their ability to learn the vehicle response characteristics and self-adjust to achieve rapid, precise control without over-shooting. This is achieved using a short sequence of maneuvers. The heading controller, for instance, is tuned by performing a short maneuver involving a zigzag and straight-line sections. Thereafter, the autopilot is able to calculate appropriate settings for its gain coefficients and limits, whereupon the heading controller is set up for future use. Variants of this self-tuning process are used for tuning the speed control, forward and lateral position control, roll and pitch, and hover controllers.

Having configured and tested the vehicle simulator, application to a real vehicle is achieved by downloading the configuration file across the serial data/command link. The simulator software is used once again, this time in its role as a front panel for remote control. This is done by selecting a "remote control" mode instead of the "simulation" mode. This means that the operator is already familiar with the same menu structure and controls for operating the real vehicle.

Given the wide range of mission profiles, it is necessary for the remote control system to be as flexible as possible. Using a combination of low, intermediate, and high level commands, it is possible to achieve almost any maneuver within the constraints of available power and the configuration of the vehicle and its actuators. A "standby" command, for example, may be used to leave the vehicle parked or hovering within a predefined distance of the current position to await further instructions. A joystick may then be used to control the vehicle so as to bring instruments to bear on a target, and a "return home" command may bring the vehicle back for recovery at the end of a mission.

The remote control function uses an efficient protocol based on a query-response structure. This structure provides for error checking and positive confirmation that each message has been received without errors, a necessity in the underwater environment, particularly over extended distances. A watchdog facility allows the vehicle to execute a predefined maneuver or sequence if no intelligible messages are received within a predefined period, so that the vehicle will return to a recovery location following failure of the datalink.

Using a modular design approach and abstracting the remote control and positioning task to a separate, generic and fully configurable unit, it is possible to build a fully functioning vehicle, including remote control facilities, from off-the-shelf components. The wide variations of vehicle configurations, mission profiles and their positioning or autopilot requirements, have made it difficult to design a generic remote control and positioning unit. H Scientific has shown, however, that it is possible to design a generic unit to suit a very wide range of vehicles. By providing a wide range of commands, spanning from the low level, direct machinery signals up to the high level commands using waypoints and hover areas, and an equally wide range of automatic control modes, it is possible to cater to a large portion of vehicle mission profiles. The result is a system which is user-configurable for almost any vehicle, and self-tuning; it gives the operator a very powerful, intuitive console workstation from which the vehicle may be controlled and monitored.

Alstom Schilling’s Electric Quest
The Quest is Alstom Automation Schilling Robotics’ new electric work-class ROV system, scheduled for availability in March 2000. Because it is based on the Remote Systems Engine (RSE), the Quest uses a set of simplified and integrated sub-systems for propulsion, control, and communication. The Quest delivers equivalent performance to current work-class systems, but is smaller and lighter and can be easily configured for a wide range of functions. All subsea components are rated for operation at depths to at least 11,480 feet (3,500m). One of the Quest’s most significant innovations is its SeaNet communication and telemetry system, which uses compact, lightweight devices (called hubs) to route power and signals between the operator control station and the accessories on the tether management system (TMS), ROV, and accessory tool packages. The hub’s sphere is only 6.75 inches in diameter, and the port array measures approximately 18 x 8 x 2 inches. Weighing just 25 pounds in air and seven pounds in sea water, the hub allows a substantial reduction in overall ROV weight and size. These hubs replace the large control pods used in current systems.

A single type of cable/connector assembly attaches to the hubs, connecting all accessory items to the system, including sensors, cameras, thrusters, and tools. The SeaNet design makes installing equipment extremely easy. An item is simply fixed in place and a standard SeaNet cable/connector assembly is attached between the item and one of the hub’s numbered ports.

The hub comprises two sections, an atmospheric sphere (composed of two hemispheres) which contains most of the electronics, and an oil-filled, pressure-compensated cable port array, to which the ROV or TMS accessories are connected.

A look inside the SeaNet hub reveals two other important differences from current control pods: the small number of printed circuit boards and the absence of wiring harnesses. While control pods typically require one or more printed circuit boards for each accessory used on the ROV system, the Quest hub contains only four unique boards. Signals and power are routed directly from a printed circuit board to external connectors on the hub’s cable port array, eliminating the need for wiring harnesses. Since this design feature eliminates approximately 75 percent of the electrical interconnections typically required for ROV accessories, there is significantly less risk of malfunction caused by intermittent or incorrect connections.

When maintenance is needed inside the hub, access to the electronic components is particularly easy. Without undoing any electrical connections to the tools and accessories attached to the hub, the technician can simply remove the two sections that compose the sphere and replace printed circuit boards as necessary. The boards have easy-to-understand visual diagnostic indicators for component health. LED indicators built into the connector head of each accessory cable provide information about voltage and signal status. Each circuit has independent ground fault monitoring and remotely resettable fusing.

A standard Quest system contains three identical SeaNet hubs (two on the ROV and one on the TMS), each with up to four composite video channels and 26 serial channels. Because the hubs are modular, additional hubs can easily be added to the system if more channels are required. This modular design has the additional benefit of parts commonality, which simplifies maintenance procedures and reduces required spares inventories.

As an alternative to adding hubs to the Quest ROV, a hub can be permanently installed on an auxiliary equipment skid. It can then simply be connected to the ROV system when the tool skid is used. Because there will typically be two hubs on each ROV, and because hub location is flexible, the devices can be located near the accessories to which they are attached. This convenient access, which minimizes the length of accessory cables, is enhanced by the presence of connectors on both sides of the hub.

With the Remote Systems Engine at its core, the Quest ROV is designed to significantly reduce the cost of remote operations through a strategy of component commonality and simplicity.

Air-Independent AUV Power System from the University of Calgary
The selection of a power system for large underwater vehicles such as naval submarines is usually based on previous operational experiences, familiarity with the technology, and defense policy. For example, for almost a century, the heavy, low energy density, but cheap and reliable lead-acid battery has been the most favored in conventional submarine designs, despite the advances in electrochemical technology. The breadth of operational experience regarding autonomous underwater vehicles (AUVs) is even more limited. There is also, because of the lower power demands, a wider choice of energy conversion systems.

The majority of AUV designs still rely on secondary batteries such as silver/zinc or lead acid. However, with the increasing design emphasis on vehicle underwater endurance and hence cost effectiveness, more exotic, compact, and lightweight high energy density power systems are under investigation for AUV power requirements.

Numerous attempts have been made at theoretically, qualitatively, and quantitatively evaluating vehicle designs and associated power systems. The power system’s energy density (gravimetric and/or volumetric) or power density is often the primary consideration and specification factor. The limitation here is that, for example, if a suitable system energy density is used in isolation from other selection criteria, the power output, which directly dictates the speed, is not considered. Hence, the vehicle may not attain the desired operating speed. Figure of Merits (FoM) approaches have also been investigated and in some cases applied to the selection of the energy source and conversion system. The FoM is the product of a number of factors, each relating to a specific characteristic of the particular power system, such that each evaluation factor (fi) is given a numerical weight. While evaluations using the FoM have attempted to provide an objective analysis of differing systems, they are biased to the extent that they reflect their originator’s priorities in selecting the values to be used for each weighting factor. In addition, the application of the FoM technique is hindered because too much of the available data is ambiguous or ill defined. The weight (fi) to be given to each criterion still relies on a certain amount of subjective judgment, and the mathematics of how to generate an overall FoM number are arbitrary. Hence, despite the growing sophistication of the FoM type of analysis, it cannot as yet provide a definitive solution to the AUV power system selection problem, although it is a useful tool for conceptual or preliminary design assessments in determining the best group of systems available.

In general, the evaluation techniques have been very subjective and open to wide applications and interpretations of design philosophies. What seems evident is that the complete design of the AUV to be used and the missions to be accomplished are essential elements in the evaluation and design process. Although the evaluations have usually addressed design, the power system evaluation and integration analysis is still severely limited in application. Where the more advanced analyses have considered a vehicle with a variety of power systems for a particular mission, the power system has been assumed to operate with a fixed energy and power density. Hence, the effects of the mission profile, and therefore the performance of the power system operation, have not been examined to evaluate the influence on total vehicle design and vessel operational performance.

Over the last decade, research at the University of Calgary (Alberta, Canada) has focused on a detailed systematic investigatory study of the design of underwater vehicles in general, but predominantly AUVs, to elucidate the effects of subsystem selection options and integration. A recent example of this specialized work was funded by the Office of Naval Research (ONR) for underwater vehicle applications in areas where underwater ordnance was known or suspected. Mainly due to the low magnetic signature, the selection of any electrically based or operated propulsion system was not possible. After detailed vehicle application analysis by the ONR, a fossil-fueled heat engine system was initially targeted for further development, specifically a Rankine cycle turbine or Stirling engine.

The Stirling Engine
Subsequently, based on the Stirling engine development, underwater vehicle design, and total system integration knowledge, researchers at the University of Calgary designed and manufactured a hydrocarbon-fueled Stirling engine for the vehicle. Vehicle design and integration studies proved useful and enlightening, especially with the additional constraints imposed on the vehicle. Some concerns included limited vehicle diameter, high performance operation, low power requirements, safety, and a non-magnetic signature. The engine had to be compact and lightweight. When coupled with a suitable compressed gaseous oxygen source, the total vehicle endurance was driven by a multitude of factors, such as combustion system operation, exhaust gas management, oxidant storage, thermal efficiency, lightweight non-magnetic materials, construction, and system packaging, that had to be traded off against each other.

More research is currently underway to further quantify and discern the characteristics of the oxidant and exhaust gas management systems. The application of Stirling engines for the ONR’s AUV application required a leap of faith into a technology not usually thought to be viable in small vehicles, even though the Swedish Navy has been operating larger systems for several years in its conventional submarines. The University of Calgary was pleased to be the recipient of the ONR contract and hopes to continue the work and expand the application of the developed system to other vehicles and operational scenarios.

Bluefin Robotics’ MIT Roots
Bluefin Robotics Corporation, a spin-off from the Autonomous Underwater Vehicle (AUV) Laboratory at the Massachusetts Institute of Technology (MIT), began building vehicles in 1989. Since then, its robotic submarines have been all over the world, from the hostile winter waters of the Labrador Sea to the Gulf of Mexico. The vehicles have returned safely from each of their independent missions and consistently produced high quality data sets.

In total, Bluefin Robotics has logged 60 weeks of field expeditions, totaling thousands of AUV dives without losing a vehicle. The AUVs are rated for 19,680 feet (6,000m) depth and come equipped with a wide range of sensors such as echosounders, acoustic arrays, sidescan sonar, and CTD. Bluefin also builds and operates small AUVs in deepwater 9,840 feet (3,000m). The company has built 12 AUVs, all completed on time and within budget. Bluefin continues to work closely with the MIT AUV lab to develop technology. Current funding levels at the Lab are around $3.5 million per year, virtually all of which is provided by the U.S. Navy.

The company is convinced that the AUV will be a major player in the future of the industry. An AUV can dramatically improve offshore site surveys and pipeline route surveys. By eliminating the tether, data quality improves and costs drop from 25 to 75 percent, depending on the operation. Navigation is more accurate, sonars can be positioned at a constant height off the bottom, and surface vessels do not have to make long turns due to tethers and tow lines. Bluefin’s vision is to automate the entire survey and operate directly from land, thus eliminating the need for expensive surface vessels.

Hydro-Lek’s ROV Cutting Tool
With the high mobilization costs of large work class vehicles, operators are now recognizing the advantages of using small, easily deployed ROVs whenever possible. A new cutting tool skid has been developed by Hydro-Lek which enables small observation class ROVs to perform some of the tasks of bigger work class vehicles. The new tool package can be fitted to any small host ROV enabling it to cut cable, pipe, or hose up to three-inches in diameter.

The first of the new cutting tools has been delivered to Dutch Diving b.v. of Den Helder in the Netherlands. The company is using it on a Sea Eye Tiger electric observation class ROV for cutting grout hoses and steel cables around platforms in the Dutch sector of the North Sea. The cutter now also makes it feasible to use small, easily deployed ROVs for a range of salvage, harbour engineering, and pipelaying tasks.

The Hydro-Lek cable cutter tool skid can be built to fit any type of small ROV. It uses a 380V 3 phase AC 2.2 kW motor which powers a hydraulic pump and cable gripper, control valves, telemetry and a diamond tipped or carborundum rotary saw. The jaw, blade and feed control system can all be rotated through 360 degrees enabling vertical or horizontal cuts to be made. When in use, the jaw locks firmly onto the item to be cut and with such rigidity that the blade can cut the workpiece from any angle without twisting. Rigidity is such that cutting can be carried-out in zero visibility by monitoring the current drawn on the three phase power pack to control the feed of the blade. The tool skid can also be used to provide the ROV with manipulators or cleaning tools.

Hydro-Lek’s managing director Chris Lokuciewski says, "Anyone with a 10 or 25 hp ROV will know that one of their most attractive features is the ease with which they can be deployed. You do not need massive A-frames to launch them and their correspondingly thinner umbilicals are a real bonus for work in deep water. Because we are able to design and build hydraulic power packs for any type of vehicle, no matter how small, we can add a wide range of tools to observation class ROVs which enable them do much more than just go and look."

Annual ROV Design Competition
The intent of the Web site www.rov.net, originating in 1995, is to promote the understanding and knowledge of ROVs and associated technology, as well as preserve and record all the different aspects of the world of ROVs for the future.

The site hosts an annual ROV Design Competition open to private individuals. Out of 58 entrants for this year’s event, only three were successful in building an ROV.

The 1999 winners are divided into two categories: Design and Build an ROV or Design of an ROV. The winners in the build category are Steve Thone and David Thompson, each building an ROV that actually works in water and can move along three axes. Steve powered his ROV with the traditional thrusters and propeller blades, while David used water jet propulsion. The ROV Design only winner is Gavin Chait of South Africa. Along with competition sponsors Hydrovision Limited and Tritech International, ROV.net awards prize money to the winners. The next ROV Design Competition kicked off during June 1999 and will run through June 2000.

SMD’s Purpose Built CMROV
Soil Machine Dynamics (SMD) began developing a high powered jetting ROV in 1993 and built its first prototype machine in 1994. Sufficiently impressed, KCS placed an order for the Marcas II Cable Maintenance ROV (CMROV) system in 1995. This was an 8,200 feet (2,500m) ROV with 180kW of power which could be used with a jetting skid or a trackbase for improved burial capability. Nereus, built for Temasa in 1996, was a development of Marcas II with a jet tool integrated into the ROV chassis and a bolt-on track module.

In late 1997, Charles Tompkins Consultants (CTC) Ltd. of Middlesborough approached SMD with a request for a high powered 8,200 feet (2,500m) bespoke trenching ROV. CTC, which has particular experience and expertise in marine trenching gained in the offshore oil and gas sector, wanted a state-of-the-art machine based on proven technology. The order was placed in January 1998 and the latest generation of CMROV (C-Trencher 1 or CT-1 for short) was delivered in June 1998.

There are a significant number of advances which differentiate CT-1 from SMD’s previous CMROVs, all of which are directed at producing a swimming machine with superior jet burial performance. SMD has put considerable effort into developing its own jet burial technology, including conducting jetting trials in both sands and clays.

The conclusion of these trials and subsequent experience with machines has resulted in a significant amount of knowledge for SMD. For instance, it was found that high water flow, low pressure fixed jets are effective at trenching in sands and clays up to 50kPa shear strength. Configurations of nozzle size, angle, and spacing for forward and rear facing jets can be optimized for the soil conditions. Also, the greater the amount of power that can be applied to the jetting, the better.

The tracked jetters produced by SMD have performed extremely well in contrast to ROV-mounted systems. The basic concept behind CT-1 is to make its jetter as close as possible to the tractor versions, and to this end it is fitted with the same high flow jetting system and cable depressor, and with a power level similar to SMD’s latest tractor - 280kW (375hp).

The jet tool uses 3.28 feet (1m) deep penetrating swords with forward facing jets to fluidize the seabed. Rear facing jets keep the trench fluidized and also counterbalance some of the couple created by the reaction force of the forward jets. The legs pivot at the top to fold up into the chassis to vary depth or to stow them. As with all SMD flexible product burial tools, the jetter uses a cable depressor which is designed to be friendly to the cable and to guarantee the depth of its burial. The depressor is fitted with a cable sensor in the shoe at the bottom.

The results to date from CT-1’s work on the SEA-ME-WE 3 project are very promising. Over the first 8km trenched, a cover of one meter or more has been achieved at typical trenching rates of 500m per hour in a single pass through soft to medium sands.

Considerable effort has been put into the design of the ROV layout which has resulted in improvements in its performance, efficiency, and maintenance. The jet tool is integrated with the main chassis and is not a bolt-on skid as with adapted work class CMROVs. This has a number of benefits. A more compact design reduces the front and side areas and, hence, drag. It produces easier access to parts for maintenance. The water pump is mounted high at the rear of the machine so that it runs in clearer water, reducing wear, especially in the seals.

The profile of the buoyancy has been made as low as possible, producing a 20 percent reduction in front and side area compared with SMD’s first commercial CMROV. This reduces the aspect ratio and thus the drag of the machine. A major problem with increasing the power on free swimming un-garaged vehicles can be the increased size of the umbilical. By working with the umbilical manufacturer to optimize the cable design for the increase in power from 180kW to 280kW there was only a 10 percent increase in diameter to 44mm.

The hydraulic transmission system is designed to give total power sharing so that all of the available power can be put into any of the individual systems. The real benefit of the power sharing system can be seen when the optional track modules are fitted to the machine. These give the operator improved control compared to running on skids when jetting and are especially useful over long distances. But their real advantage is that they are a far more efficient traction unit than a swimming ROV. A tracked vehicle can produce 0.5 tons of bollard pull at 500m per hour for as little as 4kw, compared to 65kW for an efficient thruster system. This means that 250kW can be devoted to jetting, enabling the burial depth to be increased to 4.92 feet (1.5m) by fitting longer swords and an additional water pump.

There has been a comprehensive attempt to improve the reliability of the machine through all stages of the design and assembly. For example, SMD chose to replace the two pole motor with a four pole unit thoroughly tested prior to assembly. Integrated construction of the vehicle and jet tool helped reduce the interfaces. Selection of critical components was based on proven performance or developments of existing technology.

CT-1 has been so successful that SMD is now in the final stages of completing CTrencher-2 for CTC. This vehicle is based on CT-1, but with several improvements. It has a highly effective 4.92 feet (1.5m) jet tool and a track package which can be fitted for trenching in harder ground. It boasts a powerful deburial tool for excavating cables to 4.92 feet (1.5m) deep. CTrencher-2 is capable of operation in rough weather (sea state 7) and has a lighter 300kW motor design, more efficient (reduced density) buoyancy package, and is depth rated to 9,840 feet (3,000m).

SMD has developed the world’s most powerful CMROV to meet the ever increasing demands of the telecommunications submarine cable market. It is a highly reliable machine backed up by the company’s commitment to after-sales support.

Cybernetix S.A.’s SWIMMER
With the increasing depth of offshore oil and gas exploration, the cost of ROV spreads for inspection, maintenance, and repair (IMR) tasks represent a growing part of the total field’s operational expenditures. The main objective of the Subsea Work Inspection and Maintenance with Minimum Environment ROV (SWIMMER) Project is to demonstrate that it is possible to operate a standard work class ROV for IMR tasks on deepwater fields without long, heavy umbilicals or special support vessels and their associated costs.

The SWIMMER shuttles the ROV with a special AUV from the surface to a subsea docking station. When the AUV is docked, the ROV is reconnected to a power/control umbilical which is integrated into the production umbilical and is ready to be operated from the surface like a traditional system.

The ROV is attached onto the AUV shuttle (called a BAT skid) with the following features: electrical thrusters and power supply from on-board high energy density battery pack; hybrid LBL/dead-reckoning long range positioning system for autonomous transit trip to target location; low speed/high reliability long range acoustic bi-directional telemetry system for vehicle control and monitoring from the surface unit throughout the mission, and HF telemetry for real-time surface control using joysticks; short range sub-metric positioning system using RF metrology and acoustic camera for automatic docking on subsea station; and an underwater mateable connector set allowing connection of the umbilical to an integrated TMS for ROV deployment.

Mission parameters are downloaded into the onboard computer and the system is ready for launching. The AUV can be launched with classical cranes or A-frames either from the surface production facilities or an ordinary supply boat. Unlike conventional deepsea ROV systems, there is no need for sophisticated DP support vessels.

The transit phase is a trip from the launching point to an area located 98 feet (30m) right above the docking station. The BAT skid’s navigation is controlled onboard the vehicle using data from the long range positioning system, which is designed to give a metric accuracy at a maximum slant distance of 10,000 meters. The transit phase ends when the skid stabilises itself above the docking station with a pre-defined bearing.

The approach phase is the final vertical descent followed by the landing of the BAT skid on the docking station. The short range positioning system provides the onboard computer with sub-metric precision data. The docking station has a V-shaped structure which acts like mechanical guides for the final landing.

When the BAT skid has landed at the docking station, it is mechanically locked and the wet connectors are mated, thus allowing power/control of the ROV, power/control of the TMS, and refueling of the skid’s batteries. The ROV is now ready for operations.

Since the ROV is now powered and controlled from the surface control room via the pre-installed umbilical, there are no limits to the complexity or duration of the tasks it can carry out. The reachability of the ROV depends only on the length of the tether. It will be operated within a 250m radius area around the docking station. During recovery, the different phases described above are reverted. The SWIMMER operation ends when the BAT skid is recovered from the supply boat or the platform.

Advantages of the Concept
Since a surface support vessel is not needed for ROV operations, economical simulations made for a typical deepsea field (IMR tasks only) show that an operator can be paid back from his initial investment in as few as three or four years. There are many other valuable advantages in using SWIMMER: fast mobilization, operational flexibility, use of standard ROVs, new deepwater field design, and retrofitting of existing fields.

Sea trials will be conducted to reflect actual operational conditions, but it will be necessary to perform trials on a real production site prior to putting SWIMMER on the market. To achieve this, Cybernetix intends to work with oil and gas companies to obtain access to a production field and build an industrial prototype capable of interventions at 9,840 feet (3,000m) depth.

Carbon Fiber Reinforced Plastic in Vehicle Design
Urenco (Capenhurst) Ltd., in England has been manufacturing high-tech glass, aramid and carbon fiber reinforced plastic (CFRP) products for almost 30 years and over the last six years has developed thick wall composite products. This thick-wall composite is ideally suited for use on ROVs and AUVs to contain electronic equipment or instrumentation, or for buoyancy, as it is capable of withstanding extremes of compressive pressure. Most ROVs and AUVs are currently rated up to 9,840 feet (3,000m) water depth, and it is expected demand will arise for greater depth ratings over the next few years.

Designs can vary from pure cylinders to cylinders containing internal stiffening rings also manufactured in CFRP and bonded in position. Finite element analysis and computer modeling techniques are used to determine the optimum fiber lay-up angles and wall thickness requirements. The quality of the composite is of utmost importance and all winding and curing parameters are tightly controlled. Sample off-cuts are taken and analyzed for volumetric fraction of fiber to resin content and voidage levels. Checks for cracking and delamination are also carried out. The benefits of using CFRP in this application are its corrosion resistance and its lightness compared to metal hence providing opportunities for increased pay-loads.

Both simple cylindrical constructions and reinforced structures have been used or developed for these applications by Urenco and various other organizations. Glass fiber pressure structures are relatively common but suffer due to the weight of the glass and limited structural properties. Besides carbon and glass composites, various other reinforced plastics that have been developed include aramid and hybrids such as glass/carbon mixes. Each has its own merits or demerits. Carbon fiber remains the preferred material to achieve the greatest savings in weight and the greatest depths.

Since one of the problems that besets the producer of such components is the well known variability in composite performance, it is essential to maintain tight control on all aspects of the production process. The overhead cost can be minimized by use of wet filament winding which can be developed to give very consistent products without the need for more expensive materials such as prepreg.

Most vehicles for deepwater deployment are relatively thick walled structures. This causes particular problems for the manufacturing design because of the difficulties of controlling the cure process adequately. Thick composites are prone to either prompt or fatigue-driven delamination and local cracking caused by the interaction of the resin and fiber systems. Urenco employs a variety of techniques including material, design, and process optimization and modelling to greatly reduce these so that the vessels operate within stress and stress cycling parameters.

Experience has shown that values of compressive strength obtained by sample testing do not well represent the performance under duty. Safety factors should be determined by testing of vessels. As well as the triaxial stress state that exists in service in the CFRP vessel walls, the effect of some manufacturing irregularities such as fiber buckling are well known contributors to the reduction of quality. Certain other contributary factors are understood but by no means are all of the processes involved quantified yet, which makes field experience and a conservative design philosophy essential for all but experimental vessels.

Similarly, prediction of buckling pressure by simple linear analysis of homogeneous and symmetric vessels can only give a guide to the upper limit of operating pressure. The interaction of the nonlinear material properties of the plastic, manufacturing irregularites and dimensional tolerances in the vessel all play a part in reducing the effective buckling pressure. Even with well controlled manufacture, this can lead to anywhere between a 30 and 50 percent loss of margin. Some progress has made toward understanding and predicting these effects, though, at the moment, correlation of prediction with test measurement is poor for certain classes of design, such as those where the vessel is optimized for materials strength and buckling. As is usual with composite structures, great attention is required to end closures and joints between the composite vessels and other load bearing components. Failure often initiates at the interfaces where the internal stresses in the composite can manifest themselves as delamination forces at the ends. This can be exacerbated by poor design of bearing loading and bending moments.

In Urenco’s experience, some designs which are not optimized according to classical laminate design methodology have been found to give consistent and satisfactory performance in the field. There is no doubt great potential in the further use of these materials for submersible shells, but the gap between theory and experience is somewhat larger than some advocates of this technology would maintain. It will probably take many years before they have proven themselves to customers from the oil and maritime industries. However, Urencon has made a start in commercializing composites for submersible pressure vessels.

Simpler is Better
Sub-Atlantic Ltd. of Aberdeen, Scotland, hold to a fundamental company design ethos: adhere to the KISS principle (keep it simple, stupid) at all times. By focusing on the task required to be performed by the vehicle, a system can be produced using the minimum number of components in the simplest manner possible. To this end Sub-Atlantic decided to produce a range of all electric vehicles. But rather than follow the herd and use brushless D.C. thrusters with the attendant complexity and high parts count, they created an A.C. thruster which has proven to be ideally suited to the ROV environment.

The Sub-Atlantic thruster is a 440V three-phase squirrel cage design which enables the rotor to be mechanically very straightforward. The integral driveshaft is supported at either end by long life bearings and exits the thruster body through a simple yet extremely effective and durable ceramic ring and static P.T.F.E lipseal arrangement. Apart from the propeller, there are no other moving parts in the thruster. Less parts means less to go wrong ,which means less down time; so much so that over the seven years the original Cherokee ROV has been in existence there has not been one thruster failure.

There is no need for bulky and heavy transformers or power supplies on the vehicle; 440V three-phase is sent down the tether to the vehicle and fed to Sub-Atlantic’s own design control card, which measures only 120mm by 130mm by 32mm thick, small enough to fit in the palm of your hand. Taking a +/- 5V proportional speed control signal from the joystick via the telemetry link, this card provides torque control of the thruster with smooth and seamless transition from one direction to the other. Again the low parts count and the well thought out design give enhanced robustness and reliability. The thruster is oil filled and pressure balanced and units are in the field operating successfully at 20,000 feet (6,000m) depth.

To simplify and add lightness is a tenet that appeals to engineers of all disciplines, and it should also impact the ROV industry. It’s time to move away from rigidly calling for 100hp vehicles when it is possible to get more for less. UW