The MARS Ocean Observatory Testbed
History of the MARS observatory

The cable was successfully installed in March 2007. Prior to the installation, the MARS cable was stored on Alcatel’s leased cable-repair vessel, the C/S Maersk Defender in Portland, Oregon. As this vessel approached the end of its five-year lease, Alcatel scheduled the vessel to install a cable in Alaska in September, then to steam south to install the MARS cable in Monterey Bay in October, before it returned to its owners in Europe. Unfortunately, the Maersk Defender ran into problems with high currents during the Alaska cable installation, and Alcatel’s five-year lease expired before they could perform our installation in Monterey Bay. The cable ship

The C/S Global Sentinel installs the MARS node.
In mid-March, 2007, the cable-laying ship Global Sentinel began laying the MARS observatory cable just offshore of MBARI's facilities in Moss Landing, California.
Image: Todd Walsh (c) 2007 MBARI
offloaded the MARS cable in Portland and steamed off to Europe. Alcatel then subcontracted Tyco to install the MARS cable with their cable-repair ship, the C/S Global Sentinel. What followed was a frantic effort by the Global Sentinel crew to mobilize their new ROV so that MBARI’s cable could be installed in the tight window between the end of the winter storms and the start of the salmon season on April 5th, 2007. (Our permits forbade cable installation during salmon season.) The cable installation was completed on April 2nd, 2007, three days before the deadline, despite many late-season storms.

Alcatel finally delivered the medium-voltage converter prototype in September 2007. The following months at MBARI involved integrating the medium-voltage converters into the housing, with several modifications incorporated to provide system stability, which proved to be a problem with negative impedance loads. This was followed by system-level integration, testing the Voltronics power supply, the cable simulator, and  the medium-voltage, low-voltage, and data communications subsystems end-to-end, first in the lab, then in the test tank.

After many years of hard work and complications, on February 26th the R/V Point Lobos installed the MARS node into the trawl-resistant frame on Smooth Ridge 35 kilometers (20 miles) offshore, and the power switch on shore was finally turned on. The system came up perfectly, and then, after 20 minutes, the 10kV power connector failed, the current surged, and the voltage plummeted. Repair would require the TRF to be recovered, which would require a vessel with a large lift capacity and a cable-splicing capability.

The cause of the failed connectors was never really established. A major contributing factor was thought to be the cable-termination-assembly (CTA) half of the connector being out on Smooth Ridge for 12 months before the node was installed, despite ODI stating that the connector was designed to withstand this condition.

At the time of the MARS connector failure, ODI was in the middle of a nine-month qualification program of a second generation of medium-voltage converters that were being developed for the NEPTUNE Canada observatory. These connectors had a higher breakdown voltage design margin. ODI agreed to install the first of these new generation high-voltage connectors to the MARS node medium-voltage housing and CTA as soon as they were available. The MARS connectors and their harnesses themselves underwent thorough factory acceptance testing at ODI, including high-voltage testing (at 10kV) under pressure (10,000 psi) for 18 hours at a time, with intermediate connector mate/unmate cycles under pressure. In July 2008, technicians and engineers from ODI installed the new connectors on the MARS system in the lab.

As soon as they were installed, the entire MARS system from the Voltronics power supply through the cable simulator, CTA, and new medium-voltage connector harnesses to the medium-voltage power housing, with the seawater return, were tested in the MBARI test tank for three days, with several mate/unmate cycles of the new connectors, and with periods left unmated underwater for hours at a time. The system was now ready for installation at sea. All we needed was a repair vessel.

The MARS node aboard the C/S Intrepid.
The MARS node is prepared for redeployment aboard the Cable Ship Intrepid

The at-sea repair would need a cable ship to bring the trawl-resistant frame to the surface, remove the failed cable-termination assembly, and splice on the spare cable-termination assembly. Fortunately, back in 2003, MBARI had contracted with Alcatel to deliver two CTAs, as the lead time was over 12 months. Therefore, MBARI had a spare CTA to refurbish so that it could be switched out at-sea with minimal delay. Without this spare, the repair would not have been affordable. However, finding a cable vessel to undertake the repair at an affordable price proved to be extremely difficult due to an upturn in the telecommunications cable market. After months of searching to find a suitable repair vessel, MBARI finally managed to sign up IT International Telecom Inc. for its cable vessel, C/S Intrepid, to undertake the repair in November 2008 during its transit between Hawaii and the Panama Canal. The repair was successful and the MARS system is now fully operational.

The MARS node was installed with its first science instrument, a “CTD” sensor that records conductivity, temperature, and depth with MBARI’s plug-and-work software and protocol. This enabled the team to view the data stream in real time and to easily monitor the operation of the system. As soon as the CTD was plugged in to the science port on the node, it started streaming data together with its metadata to the Shore Side Data System (SSDS). Being equipped with a plug-and-work software and protocol enabled interoperability between observatories. The instrument can be unplugged from MARS and plugged in to any of the MOOS buoy observatories to operate without any modifications. Both the MARS shore station and the MOOS mooring understand the protocol, and both can execute the same Java driver code that is stored in the instrument. This capability to run the same instrument driver software on multiple observatory architectures was developed as part of the MOOS initiative to provide interoperability of instruments between observatory systems.

MARS node on the seafloor.
The completed hub of the MARS observatory rests on the seafloor of Monterey Bay. The MARS Observatory will allow scientists to perform a variety of real-time experiments 900 meters below the ocean surface.
One month later, a prototype of the Free Ocean Carbon Enrichment (FOCE) experiment—an early deliverable from a new MBARI initiative—was installed on MARS. This was the first experiment designed specifically to work on a cabled observatory. Within seconds of connecting FOCE system to the MARS science node, the project team on board the R/V Point Lobos was receiving data from the FOCE sensors. This data had been delivered to the SSDS on shore through the MARS cable, published from SSDS, and made available via a wireless connector from MBARI back to the ship.

In early 2009, the Ocean Research and Conservation Association’s Eye-in-the-Sea camera was installed on MARS and began to send low-light video images back from the deep. MBARI is looking forward to the installation of several more instruments later in the year including a broadband seismometer, an Environmental Sample Processor, the Benthic Rover, and a bottom pressure recorder with precision tiltmeter to monitor submarine volcanic areas.

Previous 1 | 2 | 3

Last updated: Oct. 07, 2009