"The changing current is a really big issue because it's so fast through there," said Doc Church, TNC's safety engineer. "If you fall in the water, it's probably going to be a long time before they catch up with you."
This month, TNC is embarking what engineers say will be its biggest current-related challenge during the five-year, $849 million construction project. After seven months of work in Seattle and the Port of Tacoma, the beginnings of the first bridge foundation, or caisson, is ready to haul into position.
"The big problem is, we're putting the new bridge just 60 feet from the old one," said Brenda Lichtenwalter, a TNC engineer coordinating the anchoring system. "We're sitting right there in these vortices. It's a really confused flow pattern.
To figure out exactly what the forces on the new caisson will be, TNC ran two types of modeling - for a margin of safety Lichtenwalter calls a "belt and suspenders" approach.
The first model was a virtual one, created by computers, using computational fluid dynamics to predict the direction and intensity of forces on the caissons.
"It's cutting-edge computer stuff," Lichtenwalter said.
In that model, software engineers subdivided the moving water into millions of tiny cubes, and tracked each of them individually during virtual current tests. Millions of individual calculations were combined to create an aggregate picture.
The other model was a real one, a 1/100th-scale replica constructed by H.R. Wallingford Ltd, a British hydraulics research and consultancy firm. Modelers at the firm's testing laboratory in Wallingford, England, spent a month duplicating the bathymetry of the Narrows in a flow tank 100 feet long and 30 feet wide.
The point of the modeling, Lichtenwalter said, was "to optimize the anchor array" - that is, to configure anchors so they would share the load equally at all tides throughout the construction process. (When the caissons are securely set on the bottom, the anchors will be disconnected and the caissons will stand on their own.)
Simply arranging the anchors and chains at equal distances around the caissons, like spokes on a wheel, did not work, Lichtenwalter said. And even if it had worked in theory, she said, it would not have worked in practice, because the old caissons are in the way.
Instead, Lichtenwalter and other designers came up with an arrangement of 32 anchors for each caisson, positioned at various distances along the arcs of two concentric circles. One circle is 300 feet from the center of each caisson. The other is 600 feet away.
When the caissons are in place, the problems associated with currents still will not be over. The certainty of erosion has presented TNC with another expensive and complicated challenge.
Computer modeling indicates that the powerful vortexes created by water whipping around the old and new caissons will be capable of digging trenches 70 feet into the seabed, threatening the stability of both bridges.
To counteract this scouring action, TNC is placing 70 tons of rock around the bases of the caissons.
For engineers looking for design parameters for the new bridge, the simple tide charts used by fishermen and clam diggers are next to useless. The currents in the Narrows are far too complex for that.
Instead, TNC put devices on the Narrows floor that sense the speed and direction of currents for every vertical foot of the channel.
The devices, called acoustic Doppler current profilers, are similar to instruments used at airports to measure wind shear. From the bottom of the Narrows, about 1,000 feet north of the bridge, the little devices emit pinging sounds, like the submarines in World War II movies.
The sounds move up through the water column, sending back an echo whenever they strike particles drifting in the water. The Doppler hears the echoes and uses them to calculate the direction and speed of currents. It transmits the date electronically to TNC, where it's displayed, in real time, on computer monitors.
"It sends the data with little pulses, like a thousand busy beavers tapping out a code with their tails," said Ebbesmeyer, whose former company, Evans/Hamilton, installed the devices under contract with TNC.
The devices are working well, Ebbesmeyer said, with a couple exceptions: The currents are strong enough to roll rocks as big as croquet balls around on the floor of the Narrows and the resulting clatter is so loud during heavy tidal exchanges that it drowns out the pinging. The acoustics don't work.
At peak flows, Ebbesmeyer said, using sonar in the Narrows can be "like trying to shout across the freeway."
Also, the tumbling rocks threatened to crush the instruments. To protect them, technicians had to design little steel bunkers for them.
"People take the Narrows for granted," Ebbesmeyer said, "but it's really something to be respected.