SEQUIM — The morning of Jan. 6 dawned cold and clear near the mouth of Sequim Bay. The silence was broken only by a gunshot coming from across the water: someone out duck hunting.
Here at Marine Sciences Laboratory, part of Pacific Northwest National Laboratories, a team of researchers was putting two inventions to the test.
One is a prototype device called an AMP, for Adaptable Monitoring Package. It’s an elongated fiberglass enclosure outfitted with numerous instruments designed to take measurements from the sea floor.
The other device is the remote-operated vehicle, or ROV, that they plan to use to take the AMP down to the bottom of the bay, attach it to a base station on the sea floor, and return to the surface without it. It’s an off-the-shelf vehicle called a Seaeye “Falcon,” to which the researchers attached an external frame to hold double the number of thrusters so it can carry the 600-pound AMP.
The team of about a half-dozen researchers, led by James Joslin, senior mechanical engineer at the University of Washington’s Applied Physics Laboratory, and Brian Polagye, an assistant professor of mechanical engineering at the UW, began developing these technologies at the behest of the Snohomish County Public Utility District.
For several years, the PUD pursued research into using undersea turbines to generate electricity from tides, and needed a way to install monitoring technology down near their turbines.
In 2014, however, the U.S. Department of Energy backed out of funding the project after cost projections just about doubled from $20 million to $38 million. The PUD abandoned the project after spending a little under $8 million.
The Applied Physics Lab team kept at developing its technology, however, and repurposed it so it can be used for basic research on the health and ecosystem of Puget Sound or other projects requiring undersea monitoring for long periods of time.
The end result is a flexible and simplified monitoring and deployment system that is much cheaper to build and use than high-end custom-built equipment.
The AMP’s instrumentation includes two kinds of sonar, a current profiler, three visual cameras, four hydrophones as well as strobe lights. It’s connected to computers on shore that gather and record data.
Ideal research location
Admiralty Inlet, the mouth of Puget Sound between Whidbey Island and the Olympic Peninsula, is an ideal test laboratory for these sorts of instruments.
It’s where Snohomish PUD planned its tidal energy project, because of the strong currents in the reach. It’s also a key location for studying how oceanic conditions affect Puget Sound, especially when it comes to factors like nutrients, oxygen levels, acidity, and the health of marine wildlife populations.
One ongoing study at the UW has sensors mounted on the bottoms of two Washington State Ferries plying the Port Townsend-Coupeville route, to measure water currents and conditions at various depths.
That study began in May 2014 and is expected to run for five years, the average service period for a state ferry before scheduled dry dock maintenance.
“We’d like to see how much ocean water, which would be deep, cold, dense water, is coming in at the bottom of Admiralty Reach, and how much fresh surface water is going out,” said Jim Thomson, principal oceanographer with the Applied Physics Laboratory, who is conducting the research.
The ocean is a source of some nutrients in Puget Sound that might deplete the oxygen level in the water or promote toxic algae blooms.
“When we have these fish kills and other events around Puget Sound, some of it can be locally sourced, and some of it is natural, coming in from the ocean,” Thomson said.
“What’s coming in the gateway; what’s the ocean’s influence on water quality; and that in turn will help us understand what our local influence is on water quality,” he said.
NVS Explorer collects a vast amount of information on currents, winds, salinity, water temperature and other data from buoys, ship-based and shore-based instruments, overflights, and research stations on land or on the sea floor, and presents them all in an interactive map.
In that sense, it’s exactly the sort of data-driven service that devices like the AMP were designed to augment.
Into the water
On the morning of Jan. 6, the APL team gathered on the deck of the R/V Jack Robertson, the APL’s custom-built work boat, for last-minute checks, before lowering the ROV with the AMP into the water.
The team had tested the full system since June, both here, in Lake Washington, and in the UW’s Marine Science Building, which has a 13-foot-deep saltwater tank for submerged testing.
They had encountered several problems along the way. In the indoor tank, a rope line got sucked into one of the thrusters; a wire got pinched and needed to be replaced; and some dirty connectors needed to be cleaned. One field test went awry when the AMP developed an electrical ground fault and stopped functioning, forcing the test to be cut short.
The team was confident most of those issues had been worked out for the Jan. 6 field test.
“The only major thing that can go wrong here is loss of power with deployment,” Polagye said shortly before starting.
Lowering the device into the water and cutting it loose from the ship went smoothly. Joslin, wearing a shoulder-mounted control unit, watched the view from the cameras on three screens on the ship while Emma Cotter, a doctoral student, monitored the feed from sonar sensors on a laptop.
The water was cloudy, so Joslin at first had to lean over Cotter’s shoulder while searching for the telltale cylinder of the base station they’d lowered onto the seabed the day before.
They quickly picked up the echo of an object, but when the ROV got close enough to see it on the cameras, they found they were looking at a barnacle-encrusted oblong object lying on the sea floor.
“What’s that?” Joslin said.
“That’s not it,” Polagye said.
They discovered later that day they’d found an intake pipe for the marine lab. They brought the ROV back to the boat to try again. This time they located their target.
Joslin piloted the ROV close as Cotter called out distances: “20 meters, 15 meters, 10 meters.” The base station came into view.
Joslin pushed the vehicle forward into it. Its “U” shape is designed to guide the vehicle and the AMP precisely into place so the AMP can connect its instruments to the base station.
“It’s like an underwater game of horseshoes,” said Paul Gibbs, an APL engineer.
When things go wrong
Step one was complete. Step two was to connect the device to the base station and bring the ROV back to the boat. Shortly after Cotter powered up the AMP from a computer room on shore, all the monitors went dark.
They diagnosed a ground fault in the ROV. No power, no way to detach the AMP from the vehicle, or to get the vehicle back to the boat. The engineers huddled to find a solution.
“We have had a lot of problems with faults, especially in seawater,” explained Chris Siani, a senior engineer with APL who designed some of the electrical systems. “It’s one of the things that will really take you down.”
The team decided to turn off the AMP to clear out the electrical system, then restart it. The trick worked, in part, and Joslin used a screwdriver to make the fault detection system on the control box less sensitive and less likely to trip that relay again. Power was restored to half the ROV’s thrusters.
It was enough power to pilot the ROV away from the AMP and base station, but with half the thrusters down, it was hard to control. “It’s a little squirrelly,” Joslin called out to the team waiting on the deck to retrieve the vehicle.
They managed to get it back on board and secured for its return trip to UW. On the shore, the computers were already starting to collect the AMP’s data.
The software, still in development, was largely written by Cotter and Paul Murphy, a research scientist with the Northwest National Marine Renewable Energy Center at UW. The current profiler is used to measure ocean currents, while the visual cameras and sonar are good for identifying marine life, Cotter said.
The “adaptable” in its name means that any variety of instrument can be installed on it, from cameras to specialized gear like a cetacean click detector, which listens for sounds whales make.
“We weren’t sure exactly what instruments they wanted it to work with. So we designed it to house anything we could imagine,” Joslin said.
The trick is to make sure the instruments collect data only when there’s something interesting to collect, and not fill up the equivalent of thousands of hard drives with imagery of empty seas, for example.
There is also the problem of “cross-talk,” which occurs when the signal pulse sent out by one instrument gets picked up by another one.
The result can range from interference to false readings — one camera thinking it’s spotted a marine mammal that isn’t there, Cotter said.
“It just demonstrates how difficult things can be when you have to do this underwater,” Joslin said. “We learn something new each time we put it in the water.”