Coming to light

April 26, 2021 by James Steinbauer

This article appeared in Berkeley Engineer magazine, Spring 2021

The East Bay Municipal Utilities District, or EBMUD, provides water for 1.4 million people in a roughly 332-square-mile area on the eastern side of the San Francisco Bay, operating more than 4,000 miles of water pipelines in one of the most seismically active places in North America. The Hayward Fault, which produced one of the most destructive earthquakes in California history and may be ready for another, runs right down the middle of the entire length of the utility’s service area.

Until now, there’s been no way for EBMUD to monitor its underground network of pipelines. By the time the utility knows a pipe is stressed, it’s often too late to do anything about it. But new technology developed by Kenichi Soga, professor of civil and environmental engineering, has the potential to change that, allowing engineers to monitor for tension and compression in real time so they can fix stressed pipes before they break.

Civil and environmental engineering professor Kenichi Soga, at UC Berkeley in Berkeley, Calif. on Monday, March 8, 2021. (Photo by Adam Lau)

Distributed fiber optic sensing, or DFOS, is a system of fiber optic cables and sensors that has enormous potential for changing the way we monitor infrastructure, agriculture and a wide range of complex systems. DFOS uses standard fiber optic cables — the type used for high-speed internet connection — to detect potential vulnerabilities in large-scale structures and surfaces. When light is beamed through a fiber optic cable, a portion of that light bounces back toward the source, a phenomenon known as backscattering. Because the speed of light is constant, the time it takes for the light to return to its source can tell engineers where the backscattering was generated. Physical changes along the light’s path — such as temperature and strain — alter the magnitude and frequency of the backscattered light.

Unlike other monitoring systems, DFOS can provide virtually seamless measurements over great distances. Most conventional strain measurement systems focus on a single point. They might offer some gauge length, but it tends to be on the order of just a few inches. This is fine if an engineer is looking to measure strain in a specific area, but when it comes to monitoring system-wide behavior, DFOS is hard to match.

“With DFOS, we can get information from every point along a fiber optic cable, and we can get it for tens of miles,” Soga says.

Tracking a tectonic terrain

When David Katzev, a senior civil engineer for EBMUD, was first introduced to Soga’s technology, he immediately recognized its potential to monitor the performance of EBMUD’s underground pipes. The East Bay’s tectonic terrain causes between 800 and 1,000 water main breaks every year, Katzev says. EBMUD currently replaces 20 miles of the pipeline a year, and they need to double that number to achieve the 1% replacement rate targeted by most
utility companies.

“We can’t double our staff to double pipe replacement,” says Katzev, who helps lead the utility’s Pipeline Rebuild program. “So we’re constantly looking for innovative ways to improve our efficiency and create a sustainable long-term replacement strategy.”

Graduate student Peter Hubbard sits atop a water pipe, sourced from EBMUD for testing. DFOS allows researchers to monitor any changes in shape or potential cracks in the pipe itself. (Photo by Adam Lau)

DFOS would also give EBMUD the ability to fix its pipes faster — and with fewer resources. Katzev laid out one potential scenario: Say a 1,000-foot-long water main breaks repeatedly in its center. Without a distributed sensing system, there’s no way of knowing if the breaks occurred due to strain in just one area or if the entire pipe is compromised, so EBMUD might replace the whole thing. With DFOS, engineers would know how every inch of the pipe is responding to its environment and could target specific portions of the pipe that need to be replaced.

“We know that when that 7.0 magnitude earthquake on the Hayward Fault hits, there are going to be a lot of main breaks. This technology will help us get the system up and running faster,” Katzev says. “It has tremendous potential. It could really work for us.”

Katzev says EBMUD has pinpointed a few pipes that run adjacent to and across the Hayward Fault where he would like to incorporate DFOS. And as the utility begins to adopt the sensing technology more, it will gather data on not only its pipes but the landscape as well. It will get a clearer picture of the geographic distribution of earthquake damage and learn where the ground is subsiding or where soil corrosion poses the largest threat. The more EBMUD knows about ground movement in its service area, the better it can design its pipeline system for the future.

“Future infrastructure will need to be able to adapt to changes in its environment,” Soga says. “With distributed fiber optics, we no longer have to suffer from aging infrastructure. We can transform it.”

Seeing the inner workings of infrastructure

For the California Department of Transportation, or Caltrans, DFOS isn’t just a better system for monitoring existing infrastructure, but a superior tool for building it.

Graduate students Ruonan Ou and Peter Hubbard, from left, check on fiber optic cables installed beneath a paved road at Richmond Field Station. The cables are so sensitive that when a dog walks over the pavement, they can pick up tiny deformations in the road itself. “It’s like turning an entire street into a touchscreen,” says Hubbard. (Photo by Adam Lau)

In 2016, during construction of the new 9,000-foot-long Gerald Desmond Bridge in Long Beach, California, Caltrans ran into a monitoring problem. Plans for the bridge’s foundation consisted of 352 underground concrete shafts bolstered by base grouting, a process of injecting grout beneath a foundation pile to reduce settlement and improve stiffness.

But during the base grouting process, engineers found they couldn’t inject very much grout before the pressure started to skyrocket. “Everyone had an expert opinion on what that meant,” says Tom Shantz, a senior research engineer at Caltrans. “But it wasn’t clear if those shafts should be accepted.”

Caltrans determined it could not approve a base  grouted shaft for future projects unless it could verify that the grouting process transferred a sufficient load up into the shaft. The question, then, was: How?

Typically, engineers will measure the flow rate and pressure of grout as it is injected into the ground below a shaft — a method that, at best, produces an educated guess. Short of conducting a load test on each shaft, which would be prohibitively expensive, there’s no way of knowing if the pressure from the grout is pushing against the base of the shaft.

So Shantz turned to Soga, who told him that DFOS could provide a relatively inexpensive and scalable way to measure strain during base grouting.

“As engineers, it’s important for us to understand how the infrastructure we are designing really performs,” Soga says. “This technology can give us the information we need to make better decisions and design safe and cost-effective buildings.”

Postdoctoral student James Chien-Chih Wang tests DFOS technology at a San Francisco Public Utilities Commission construction site. DFOS cables buried underground monitored changes in the surrounding sediment as a drill rig (seen at right) ran over a prescribed track. (Photo by Adam Lau)

During a base grouting study in 2019, Soga and one of his Ph.D. students, Andrew Yeskoo, wove fiber optic cables through the steel rebar cages of 12 test shafts and connected each end to a fiber optic analyzer, about the size of a desktop computer tower, called an interrogator. Once the shafts were poured and base grouting began, the interrogator pulsed a beam of light through the fiber optic cables and gathered the data that returned, giving researchers the ability to look down into the inner workings of the shaft.

“We found that there was definitely load being put into the shaft during grouting, but that load was not necessarily uniform,” Yeskoo says. “In some shafts, we would see it more concentrated on one side. We would also see the distribution vary depending on the method of grout delivery.”

There are a few different ways to base grout a shaft. Before the study, Caltrans used an open system, where grout is injected directly into the soil. But with Soga’s measurements, Schantz was able to see that a closed system, where the grout fills a rubber bladder attached to the base of the shaft, was better at transferring pressure up into the shaft.

“We were really flying blind without these measurements,” Shantz says. “This technology will give us a clearer picture — and, in some cases, a surprising picture — of system behavior in walls, bridges and foundations. I can see the day when we start putting fiber optic cables in all of our infrastructure projects.”

Minimizing greenhouse gas emissions

Far from seismically active California, in the southern English countryside, the Rothamsted Research Center lays nestled among a patchwork quilt of centuries-old farm fields. Here, Soga is partnering with researchers in the hope of leveraging DFOS to study how land management affects the release of major greenhouse gases — such as carbon dioxide, nitrous oxide and methane — into the atmosphere.

Globally, agriculture contributes between 10–12% of all greenhouse gas emissions, and 40% of those emissions are estimated to come directly from soils. Plants absorb CO2 from the atmosphere during photosynthesis and store carbon in their roots. When plants die and decay, some of that carbon is released back into the atmosphere, but part of it lies buried in damp, oxygen-poor soil and stays there as organic matter.

A study at Rothamsted that began in the 1940s found that unplowed fields maintain a higher level of organic matter over time. Intensive plowing decreases organic matter in the soil by exposing it to oxygen and creating CO2. Researchers estimate that plowing since the dawn of agriculture has unearthed about 133 billion tons of carbon — an amount equal to more than a decade of global emissions at current levels.

Drought also exacerbates the release of carbon. When soil dries out, it shrinks, forming cracks that allow oxygen to seep in. Linqing Luo, a former postdoctoral student in Soga’s research group, said DFOS can detect the strain caused by cracking soil in the same way it monitors compression and tension caused by base grouting or an earthquake. In a preliminary test at the Colorado School of Mines, Soga’s research team buried fiber optic cables at different levels of soil in a wind tunnel set up to simulate a drying field. DFOS successfully monitored strain as cracks formed at the surface and spread deeper into the soil.

Luo is now testing custom designs for fiber optic cables that can measure greenhouse gasses as they escape into the atmosphere. One allows gas to infiltrate the fiber optic cable through a special porous cladding, or skin. Different gasses absorb different wavelengths of light. By monitoring how much of a specific wavelength is absorbed as light pulses through the cable, Luo hopes he can detect concentrations of greenhouse gasses in the surrounding soil. A second design calls for coating the porous cladding with a chemical indicator that will combine with gasses in the soil to generate different colored light. The intensity of the colored light corresponds to the concentration of gas along the cable.

“Right now, the intensity of light that’s produced when the fibers absorb different amounts of gas is very small,” said Luo, now a research scientist at Berkeley Lab. “The signal is very sensitive. Somehow, we need to figure that out.”

Civil and environmental engineering professor Kenichi Soga holds a length of fiber optic cable. (Photo by Adam Lau)

Eventually, Soga and Linqing will bury the cables in different fields at Rothamsted to study how certain farming practices alter soil structure. Linqing and others said they hope the data DFOS provides will not only help farmers keep more carbon in their fields, but conserve water and prevent the damage caused by over-fertilization, such as harmful algal blooms.

“By telling us how soil structure is changing in time and space, this sensing technology could help minimize the effects of agriculture on the environment,” says Richard Whalley, a soil scientist at Rothamsted Research and the study’s principal investigator. “It could substantially improve our understanding of how to manage land — and reduce emissions of major greenhouse gases.”