CINEMar/Open Ocean Aquaculture Annual Progress Report for the period 1/01/02 through 12/31/02

Principal Investigators: James D. Irish, Walter Paul, Woods Hole Oceanographic Institution

I. Accomplishments
Introduction ­ The Woods Hole Oceanographic Institution (WHOI) has been collaborating with the UNH Open Ocean Aquaculture Program for almost four years. A close cooperation has developed between the WHOI group and the UNH OOA team over the years. We believe that this teamwork is an essential requirement for the development of the OOA finfish culturing installation, servicing, and operation.

A. Scheduled Tasks
Each listed task depends on UNH directions and has changed from what was originally proposed two years ago. The work has evolved and continues to evolve as a combination of responses to events happening at the site, through the development of new instrumentation, components, and technology, and due to better understanding of mooring behavior through numerical computer modeling. The main tasks of the WHOI contribution, in close collaboration with the OOA team, during the past year are:

• Advise UNH OOA researchers on materials, established technology and methodology for mooring systems.
• Design a feed buoy mooring system with compliant mooring feed hose with electrical conductors and compliant elastic tethers.
• Continue to develop instrumentation, which enhances the monitoring capability necessary to understand the requirements for mooring offshore fish cage systems.
• Compare modeled and actual net cage mooring behavior under waves and current forcing through evaluation of at-sea observations.
• Provide input to OOA researchers and operators to assist in the development of a system of administer feed to the fish, remotely monitor the fish cage environment and image fish behavior

B. Progress on Tasks
B-1. Continued Analysis of results: ADCP Currents and Acoustic Backscattering
The ADCP on the Bub/Morey UNH environmental mooring deployed in 1999, 2000 and 2001 recorded velocities at 2-meter intervals from 6 to 48 meters depth (the top and bottom data are sometimes contaminated by sidelobe returns, but sections are good). The ADCP was mounted on the buoy in a downward looking configuration, so the data may be contaminated by buoy motion. The data (Figure 1) were averaged to hourly values to reduce instrumental noise to below 1 cm/sec and the results analyzed for tides and low frequency variations.

The tides were not the major component of the currents, but the wind (weather forced) lower frequency signals dominated the variance. The M2 component of the tides (the largest constituent) was about 4 cm/sec Eastgoing and 3.4 cm/sec Northgoing. The tidal component of the signal is about 5-6 cm/sec (Figure 2). The weather-forced currents exceed 25 cm/sec, were strongest in the upper parts of the water column, and decreased with depth. The tides were nearly barotropic with major axis (4.9 cm/sec) of the tidal ellipse twice the minor (2.2 cm/sec), and oriented in the on-off shore direction (126º True).

In late 2001, the ADCP was moved to the wave rider mooring and deployed in an upward looking configuration to eliminate possible motion contamination of the data (see section "Open Ocean Aquaculture: Moored Instrument Buoy"). The currents are in reasonable agreement with the earlier observations although the currents appear a bit larger. The amplitude of the backscattered acoustic signal was also analyzed and shows a distinct diurnal signal, strongest near the bottom, but visible at all depths (Figure 3). This is related to the diurnal migration of organisms at night. The longer-term variations in backscattering are related to the change of total number of scatters (and their size) in the water column. The signal is biased toward larger organisms (cm size) because of the 300 kHz acoustic frequency of the ADCP, but still gives an indication of the variation in biological activity at the site.

B-2. Recovery of load cells from Northern cage
During the repair of Northeast cage and mooring after its collision with a barge in Summer 2002, the load cells were removed from the mooring lines. They had been deployed since August 1999, so were in the water for nearly two years. Most of the load cells and mounting hardware showed no corrosion and appeared in relatively good shape (Figure 4). However, three showed significant corrosion that may have affected the operation, but not mooring strength. The load cell with the most severe corrosion appeared to be working well. Three of the load cells appeared to be working fine, three appeared to be working, but had significant offsets in the "zero," and the last three were not working. The load cells are now at Sensing Systems of New Bedford for evaluation as to what may have caused the failures so that we can improve on future designs.

B-3. Reconfiguration of Wave Rider into the Environmental Mooring
As part of the OOA program restructuring for 2002, the WHOI wave measuring buoy/mooring was combined with the environmental monitoring mooring. WHOI added a larger flotation collar and increased the number of compliant elastic tethers to make a stiffer mooring to better measure the waves. An acoustic release with line recover package was added so that UNH divers did not have to recover the mooring. In addition to UNH sensors, WHOI added an ADCP (current profiler) with SeaCat (T&S) at 53 meters, a mount for the UNH SeaCat (T&S) with OBS and Fluorometer at 22 m depth, and mounted the UNH Microcat (T&S) at 1 m depth on the buoy. A serial port capability was added to the data system and interfaced with the 1 m Microcat to relay information to shore. This system is continually evolving and will be improved and expanded in the coming year. For a more detailed discussion of this activity, see section "Open Ocean Aquaculture: Moored Instrument Buoy."

B-3. Developing Monitoring Capability for Aquaculture Operations
As part of the proposed effort for 2003, we were going to evaluate the telemetry and control options and start development of the monitoring capability to control/monitor an offshore aquaculture operation. This work was accelerated and started in earnest in spring 2002 and consists of several components: (1) imaging the fish and feeding operations with video and acoustic techniques, (2) spread spectrum and Ethernet modem data telemetry and command communications, (3) integrated feed buoy controller system.

Imaging: As part of the MIT/WHOI Joint Program Instrumentation class, three students teamed up with Jim Irish (WHOI), Hanu Singh (WHOI Deep Submergence Laboratory), and Michael Chambers (UNH) to investigate the potential of using acoustic imaging (which might provide the possibility of getting biomass estimates in the fish cage), video (digital pictures to send back over telemetry link to aid in management and operations decisions) and to test the telemetry link. Preliminary work on a mid-April 2002 Gulf Challenger cruise showed the Imagenex 881A sonar head could not image haddock (who do not have swim bladders) Figure 5, but could easily image juvenile cod in a shallow cage under the coastal lab dock (Figure 6).

An underwater camera and video equipment, borrowed from the WHOI Deep Submergence Laboratory, allowed some tests to be made in the fish cage and under the dock at the Coastal Laboratory to check out the video imaging capability. In the fish cage, the biofouling on the net was so great that the light level was reduced below that at which the camera worked (Figure 7). The biofouling cleaned the water in the cage so that any added lights in the cage would work well (little scattering from "marine snow"). The camera worked well at that depth outside the fish cage and imaged the divers and biofouling. In an upward looking configuration under the coastal lab dock, the camera system imaged the fish and feed well (Figure 8). The camera was controlled by an AXIS Video server that digitized the pictures and sent them over an Ethernet link to a computer to display and store the images, and to control the number of frames taken, image quality, etc. The results of the optic and acoustic imaging were written up and presented as a student poster at the OCEANS'02 conference this fall, winning second place.

Telemetry: As part of the waverider buoy and the feeder controller efforts, a telemetry link is being developed and tested. This has involved a more thorough understanding of the radio behavior over the 5-mile link to shore over water, and how to optimize the capabilities and minimize the weaknesses. A scheme has been developed and tested during the second half of 2002 on the wave rider buoy, and has demonstrated the capability to transmit water property and wave data to shore on an hourly basis while recording more detailed information on flash card in the instrument for later analysis. The 900 MHz radios had adequate bandwidth for routine data telemetry to shore, and commands to control operations offshore. This requires a two-way link, which was implemented on the environmental wave buoy, but not fully tested.

For further tests, a system with a Persistor data system, 900 MHz spread spectrum radio was assembled in a pressure case with battery and deployed on the feed buoy with software for testing the telemetry link. This failed due to a broken connector in the radio, which was subsequently repaired. This system was tested from the coastal lab to the SeaCoast science center and was found to work reliably. A "Control-C" interrupt of the program allowed the user access to the low-level PicoDOS operating system. This would allow new software or control files to be sent to the remote instrument and the system restarted.

The next step was carried out jointly with UNH (Stanley Boduch) and WHOI to implement a video system of monitoring the fish. Two cameras were ordered by UNH, an AXIS Video server procured by WHOI, and two Ethernet modems were purchased by UNH. These subsystems were integrated with the Persistor controller and 900 MHz DataLinc spread spectrum radio into one package and tested by Stan Buduch. He established the base station at the SeaCoast science center whilethe remaining components were in a pressure case that could be easily moved around. Tests were run from the beach, from the coastal laboratory and from a boat en route to the OOA site. The Ethernet system worked well out to the site, but failed on station. As the power in these radios cannot be changed (e.g. no amplifiers added), any gain must be from a change of antennas. A 24-db high-gain antenna was purchased for the SeaCoast science center and a 14 db gain antenna for the feed buoy. These fixed antennas need precise alignment or the system will not work. As part of the feed buoy operation in 2003, this link will be further tested, refined and improved (with new antennas) while the system is in operation in the feed buoy.

B-4. Feed Buoy Controller
A central part of any aquaculture operation is feeding. It is a critical technology that cannot be directly carried from inshore operations in protected waters to offshore sites. To implement an offshore, remote feeding operations several components are required, including the feed buoy, mooring system feed hose to the fish cage, feeding mechanism, feeding controller, and diagnostic and operational monitoring Again this effort was split between two groups at UNH, and WHOI who is working on the controller. The basic concept is for a controller to conduct the offshore operations on a preplanned schedule, relaying operations and diagnostic information (battery voltage, currents drawn by the various pumps, the rate of charging, etc) to shore on the feeding operation as well as environmental parameters and providing video images of fish and feeding operations.

To control the feed buoy operations, a Persistor CF1 microcomputer with 4 channel serial port, 8 channel 12-bit A/D and 256 MB compact flash card was utilized. The system runs PicoDOS, a variant of DOS, and code was written in C, assembled and downloaded to the Persistor’s flash RAM. To control the feed buoy operations, two data files on the compact flash card are utilized. One tells the computer what hour to do what task (all tasks start on the hour), and the second controls the details of the feeding operation. To change the feeding schedule and timing, these files are edited. The most reliable manner is to edit the new files on shore, test them, then down load them to the compact flash card via XMODEM protocol over the spread spectrum telemetry link.

To control the power, the 3.3-volt logic of the Persistor was used to control a FET to throw a relay that actually switches the power to the various motors controlling the feeding operation. This power control system was divided into a 12-volt section inside the controller pressure case, and a 24-volt system in a separate control box. To supply the necessary current to the run motors at reasonable current drains with minimum losses, a 24-volt power supply was used. This reduced the current drain and was still compatible with the solar panel and wind generator hardware. The 24-volt system power then changed the 12-volt controller battery through a regulator. This allowed the controller (and radio relaying information to shore) to be separated from the basic feed buoy 24-volt power, so that if a failure occurred in the feeding mechanism that discharged the battery, the controller would be able to relay diagnostic information to shore to aid in optimum response. The controller, battery, and video system was housed in an underwater housing (from the motion pack design, Irish et, al, 2001). This case is connected with the inputs, antennas, control lines, etc. though Impulse underwater connectors, so that the system is not subject to damage if significant water were to get into the buoy (or even if the buoy were to sink) (Figure 9).

B-6. Design of Feed Buoy Mooring and Compliant Feed Hose
Performance Requirements: The OOA feed buoy mooring system must permit the feed buoy to move freely in 12 meter (~40 ft) high observed storm waves and all tide levels at the OOA site without feed buoy collision with the net cage while in the deepest wave trough. The mooring must also tolerate the wave following feed buoy to be raised to the crest of the highest storm waves at the site without overstretching its mooring or pulling the feed buoy under water. A mooring scheme was developed using a combination of rubber tethers, ropes, and highly stretching feed hose. This mooring system allows the feed buoy to follow the contours of the highest storm wave peaks, raising it 47 ft above the top of the net cage (20 ft wave crest at MHW) without being pulled under. The system also prevents a collision between net cage and feed buoy while the bottom of the feed buoy is 5 ft below the top of the net cage (20 ft wave trough at MLW). That the feed buoy can move to the highest wave crest position is possible due to the very high stretch capacity of the feed hose. The buoy’s lowest wave trough position is achievable without collision through the ability of the compliant rubber tether mooring component to pull the buoy away from the net cage. Figure 10 shows a side view of the feed buoy mooring. The feed hose is designed to stretch over 130 percent at its maximum working tension. The elastic buoy mooring tethers provide the needed stretch and retraction capability through a combination of prestretched BuoyTech compliant rubber tethers and Spectra fiber rope sections.

Numerical Modeling of the Feed Buoy System: UNH’s Igor Tsukrov and Oleg Eroshkin have numerically modeled the mooring arrangement through Finite Element Analysis. The analysis computes the response of the feed buoy and its mooring components to ocean currents and waves. Results of this analysis include the heave and horizontal position of the wave following feed buoy relative to the netcage and its mooring under selected sea state and ocean current conditions. In addition the mooring forces of the mooring elements are determined, this information allows the selection of mooring elements with sufficient safety factors and compliance for mechanical endurance. It was found that the feed hose has to be as short as possible to prevent clogging with feed food at low tide, but must have the ability to stretch over 130 percent at maximum survival tension. Also the rubber tethers have to allow significant stretch and retraction forces to prevent the feed buoy from collision with the netcage. WHOI determined and provided essential component mechanical property data to OOA to allow realistic modeling predictions and modification of components to improve mooring performance. This work has progressed significantly and is continuing next year.

Feed Hose Development: A 40 ft long nylon tire cord reinforced rubber hose with 100 percent maximum stretch, developed under an ONR program, was made available by WHOI and deployed to connect between the Feed Buoy and the top of the net cage. It was found that the 40 ft long hose developed too much sag and slack at low tide, trapped the feed, clogging the hose.

A special 26 ft long hose with at least 130 percent stretch at maximum mooring load was designed and specified by WHOI, and was built at HBD Thermoid in North Carolina in November 2002 for the feed buoy. The load-elongation behavior of this hose was determined with the WHOI Snubber Software as combined response of its rubber cross section and the its nylon tire cord reinforcement. Through optimization of the tire cord reinforcement geometry, the high required working stretch is possible. The calculated maximum working tension range is between 4,750 and 5,600 lbs, depending on the amount of water fill pressure at over 140 percent of stretch.

The new feed hose also was built with a total of 13 conductors, which are arranged in an approximately stretch neutral helical configuration to minimize the conductor stretch while the hose stretch up to 140 percent. Two types of WHOI designed (#18 AWG each) conductors allow about 2 percent stretch and axial compression, needed to follow the unavoidable deformation of the rubber hose wall during load application. These copper conductors were specially manufactured by Cortland Cable Company for this application. The stretchy conductor link is considered critical to survive in the high stretch environment of the feed hose. Figures 11, 12, 13, and 14 show the hose in various stages of manufacturing.

Sensing Systems Inc. of New Bedford built a specially designed load cell with a center hole for feed to WHOI specifications. This system will measure the tensions in the feed hose at the same time that environmental forcing is being measured to compare with modeling results. The load cell will be recorded by the system developed for the fish cage load cells (Irish, et al., 2001), which will record 20-minute bursts on a schedule with the wave rider buoy. Then the feed buoy controller will continue to monitor the tension statistics hourly to keep track of the tension on the feed hose. The load cell is designed to work linearly though 6,000 lbs tension and fail well beyond the feed hose failure.

The completed hose was furnished with electrical connector pigtails at WHOI, see Figure 15. Special care was taken to assure their survival at sea.

The hose was pull tested and load cycled to 2,500 lbs maximum (for safety reasons), see Figure 16. Conductance was monitored on one wire during the test, and all wires were checked before and after the test, there was no change in nine conductors monitored.

C. Important Results and Findings
Discussion of work given above has more details, but a summary of results is:

• Environmental forcing of the fish cage and mooring were lower than initial estimates, so the drag and forces acting on the fish cage and mooring system are lower that initial design. These observed environmental results will allow better modeling of the OOA site, and more economical components to be used in future work.
• The waverider buoy/mooring was augmented with sensors to provide a local "observatory" for environmental monitoring at the OOA site. Telemetry of data to shore was improved so that hourly reports of environmental conditions can be done and served to others on the WWW.
• Tests with video cameras showed that fish could be imaged, and relayed over short distances to shore for monitoring offshore feeding operations.
• First order feed buoy controller was constructed and is being tested for control of feeding operations and relay of diagnostic data to shore.
• An improved mooring scheme for the feed buoy was developed and constructed that consisted of a compliant rubber feed hose and compliant elastic mooring components. This keeps the feed hose taut so that gravitation settling of feed and subsequent clogging of the hose is less likely.
• The new feed hose also has specially developed "stretchy" electrical conductors that permit observations to be made in the fish cage and the data relayed to the feed buoy for telemetry to shore.

D. Difficulties Encountered
There were the standard problems of designing and constructing complicated systems such as the feed buoy, controller, mooring, but the basic system came together reasonably well and now must prove itself at sea. No "showstoppers" were encountered in this component of the work.

E. Anticipated Success in Meeting Project Objectives in the Scheduled Project Period
As in any new development project, the completion and assembly of complicated system s appears to take much longer than anticipated. As evolving the feed buoy controller and mooring system, the first major steps have been taken, and as testing and evaluation take place next year, they systems will continue to evolve. Provided that the overall system will work reliably, this integrated system is considered a true “first” in the development open ocean finfish culture.

F. Reports, manuscripts, and presentations resulting from the project
Michel, A.P.M., K.L. Croff, K.W. McLetchie, and J.D. Irish, "A Remote Monitoring System for Open Ocean Aquaculture." Oceans 2002, 2002 Baldwin, K. C., J. D. Irish, B. Celikkol, M. R. Swift, D. Fredriksson, and I. Tsukrov, "Open Ocean Aquaculture Engineering," Oceans 2002, 2002.

Fredriksson, D.W., J.D. Irish, M.D. Chambers, G. Rice and B. Celikkol, "Fied Measurements of the Forcing and Response of an Open Ocean Aquaculture Fish Cage and Mooring System," Aquaculture America 2002, 113, Jan27-30, San Diego, CA, 2002.

Fredriksson, D.W., M.R. Swift, J.D. Irish and B. Celikkol, “The Heave Response of a Central Spar Fish Cage”, Proceedings of the 21st International Conference on Offshore Mechanics and Arctic Engineering. Oslo, Norway, OMAE2002-28441, 2002.

Fredriksson, D.W., M.R. Swift, J.D. Irish and B. Celikkol, “Fish Cage and Mooring System Dynamics Using Physical and Numerical Models with Field Measurements.” accepted Aquacultural Eng. Jour., 2003.

Fredriksson, D.W., M.R. Swift, J.D. Irish and B. Celikkol,. “The Heave Response of a Central Spar Fish Cage,” accepted Journal of Offshore Mechanics and Arctic Engineering, 2003.

G. References Cited
Irish, J.D., M. Carroll, R. Singer, A. Newhall, W. Paul, C. Johnson, W. Witzell and G. Rice, "Instrumentation for Open Ocean Aquaculture Monitoring," WHOI Tech. Rept. 2001-15, Oct, 2001.

II. Tasks and Activities for next reporting period

A. Tasks for the next reporting period

• Deploy feed buoy mooring system according to design and monitor performance
• Evaluate performance of:
  1. electrical conductors link in feed hose with video and water property instrumentation power and signals
  2. hose performance by measuring tension in feed hose and comparing with model predictions from current and wave forcing.
  3. Ethernet modem telemetry link for video imaging of fish in cage
• Continue development of feed buoy controller and operations software and user friendly interface to enhance operations
• Complete reporting of previous environmental forcing and fish cage and mooring response.

B. Brief work plan to accomplish tasks

• Work jointly with UNH personnel on feed buoy as required providing WHOI assistance.
• Work with Dave Fredriksson on data analysis and reporting
• Work with modelers on evaluation of feed buoy performance relative to observations to evaluation and reporting of findings.
• Testing the design for a coil-cord conductor around the elastic tethers to allow deeper sensor data telemetry, which will add midwater SeaCat (with oxygen, chlorophyll-a fluorometer and optical backscattering) data to the telemetry link.

C. Anticipated concerns or difficulties

• Ethernet modem telemetry link operation/performance
• Endurance of electrical conductors in elastic hose wall
• Possibility of net cage lifting with some anchor clump relocation in very high waves
• Overall system response to severe storm events
• Integration of feed buoy systems

III. Expenditures
During the past 2 years, the task list order has been changed, and work in the past year augmented as requested by UNH. Funding was adequate for proposed work, but increased requirements by OOA team has exceeded original budget estimates and a modest supplement awarded. WHOI efforts will expend all moneys within the time of the proposal and accomplish most of the tasks originally outlined.