Sorting out reflected signals was also a challenge for Sercel SA, a French manufacturer of geophysical instruments, as they worked on the next generation of their geophones-seismic sensors that map underground formations by sensing reflected sound waves.
As used in geophysical exploration for natural resources, these waves-created by “thumper trucks” or underground explosions-bounce off of the various types of subterranean materials and structures, enabling researchers to generate a three-dimensional map that can span geologic formations miles wide.
Geophones are remarkable for their coverage. A line of the devices-that is, a set of sensors each connected to the next by a single cable- can stretch for five miles, says Daniel Boucard, Sercel's product development manager. Such a line can have as many as 1,000 geophones connected together with an in-line amplifier added every 20 to 24 sensors to boost the signal.
Multiple lines of sensors, spaced between 100 and 3,000 metres apart, will be placed along side each other so that, when complete, a spread of sensors can cover an area that may stretch five miles long and two miles wide.
Once the geophones have recorded a set of readings, lines on one end are taken up and reconnected to the other end so that the sensor net moves in a rolling fashion across the area being surveyed-typically to map out potential oil and gas deposits.
The real challenge comes in setting up the sensor “spread” or network. Crews can take a week to 10 days with 100 people working on laying out the lines, says Boucard. Moreover, the geophones have to be placed precisely-so precisely that most spreads require a surveyor to establish the precise location for each geophone. The labor expense is significant, especially in high-cost parts of the world like Europe and North America.
But work doesn't end there. Each geophone must be aligned before the test, in most cases oriented toward magnetic north.
The standard geophone still be being used by most people is a single-axis device, according to Gordon Ryley, NovAtel Product Manager, measuring the seismic waves that reach the sensors from different directions. “Basically it is capturing the Z component,” he says. “You have to have a geophone oriented in the right direction to catch the different reflections of the waves.”
To set up a traditional geophone survey, crews would go through the survey area, flagging each spot that needed a geophone. Then crews laying out the sensors would place them at the flagged locations and adjust their orientation using a compass or similar tool.
Such measurements are neither especially easy nor precise, he adds. Moreover, once the manual measurement was done, surveyors could not recheck the location and orientation of the sensors without returning to the geophone locations.
To simplify the survey process NovAtel and Sercel began working together to develop a geophone that would use a built-in GNSS receiver/dual-antenna combination to determine its own location and orientation.
“Back in 2006 we had a customer who approached us,” says Ryley. “Sercel had developed the DSU3, a new revolutionary 3 component (3C) digital geophone design that could capture all components. They could actually grab all the seismic waves with one geophone. The old way you'd have different types of geophones or adjust the orientation to capture the waves from different directions. But in order to use the data from the 3C geophones, you need to know the exact orientation of the geophone or you have to accurately align the geophone to a reference direction such as magnetic North.”
Initially the firm was only interested in using GPS to derive accurate timing once the geophones were deployed. “They would either use a thumper truck or an explosive device to generate the seismic waves,” said Ryley. “They had to have accurate timing so that, when they post-processed the data, they would know exactly-based on the exact time the seismic waves were generated and the time of the received reflections-what the structure of the ground looked like below the geophones.”
The customer brought the problem of needing precise position, time, and orientation to NovAtel- and NovAtel was able to create a unique solution.
“Timing is easy-anybody can do that,” Ryley says. “We came up with a solution that says 'What if you can actually identify the position and orientation of the geophones?'”
Working with Sercel, NovAtel developed a dual GNSS receiver/antenna that fit on Sercel's existing geophone platform. The GNSS card used in the geophones is a custom design from NovAtel that incorporates a 14-channel GNSS MINOS6Lite ASIC that can track GPS, GLONASS, and satellite-based augmentation signals. The MINOS6Lite is a reduced lighter subset of the MINOS6 ASIC used on all of the NovAtel OEM6 cards. Using two antennas enabled the sensor to determine its own location and, using interferometric techniques, orientation. With that information, and the benefit of post-processing, the new geophone design-dubbed DSUGPS for digital sensor unit with GPS-eliminated the need for a separate survey of the instrument spreads.
“Instead of having people to measure exactly the location of each point,” says Boucard, “with DSUGPS the measurement is made by the sensor itself.”
The dual antenna also makes it possible to determine the geophone's orientation using interferometric techniques, says Ryley. “We can then tell from the angle of each of the geophones which direction they are pointing.”
Two antennas enable the DSUGPS units to obtain an azimuth accurate to within three degrees, which was good enough for the customer.
With the exact position and the exact orientation of the geophones, the firm's post-processing program could create a 3D map of the area beneath the test area to a very high accuracy.
The new technology not only cuts labor costs and setup time, it increases the chances that a survey campaign will be successful.
Poor orientation “has been the cause of failure on other systems,” says Boucard. “Usually people have been using a compass or something like that [to determine the orientation of the geophones]. One, the accuracy is not very good. Secondly, sometimes between the time you come and measure [a geophone location] and the time the wave is recorded, the sensor may move a little bit-and you don't know what was going on in between. Having the measurement inside the unit, you are sure about the measurement-which is very important.”
The self-surveying capability of the DSUGPS geophone and the power of post-processing enable customers to use the instruments across terrain where seismic campaigns might not previously have been possible or cost-effective.
The dual-antenna capability solved another problem as well. To correctly place sensors without GPS capability, testing crews may have had to inflict unwanted damage on local ecosystems.
“When they were deploying seismic surveys, they would actually have to cut down trees and remove shrubs-basically destroy the environment in order to create lines, cut lines, to plant the geophones in,” says Ryley. “So, one of the things that we came up with was 'OK we can save you a bit of that on the survey side if we can tell you the position on each geophone within a metre of accuracy,' which is easy enough to do. But the second part of that was: 'We can do that even in a moderate to dense forest.'”
“The customer can put the geophones in the ground anywhere they want, and we'll be able to identify the accuracy of the geophones within 12 to 24 hours,” adds Ryley.
Because the geophones are not normally moved for several days, the time of data acquisition is not really a problem. There is a day or two to set the instruments up; next the surveys are conducted; then the crews move the geophones the following day-a far cry from the week to 10 days required by traditional geophysical survey techniques and instruments.
“Within that period we'd be able to tell them the position of each geophone in the field down to a meter of accuracy,” Ryley say. “So that prevented them from cutting down a lot of trees. It was environmentally much more acceptable for people who were doing oil and gas surveys, and it saved them all that time for surveying.”
Thinking Inside the Box
Getting the dual antenna to work given the small platform was a challenge, said Ryley.
The baseline was also just eight inches between the two patch antennas, Ryley says. “To try to get three degrees of accuracy on an eightinch baseline is quite tough.”
That means a lot of processing takes place on the receiver card that goes into each DSUGPS and the NovAtel-provided postprocessing software that runs on the Sercel system software.
“We provided the receiver and the antenna as two separate components. We worked together with the vendor on with their industrial design and the interface into their geophone including the communication network,” said Ryley. “We basically worked with them to get it built into an enclosure that fit on top of the geophone”
In order to save time on signal acquisition, NovAtel provides a base station to provide timing and control for the GPS portion of the geophone system. The base station receiver tracks the GNSS satellites and passes along the ephemerides or orbital locations of all the spacecraft in view of the ground so that the geophone/GNSS units don't have to search constantly for the signals. However, although the base stations improve signal acquisition, they could potentially add delays into the system.
“That is one of the things that we have to manage,” says Ryley, “to keep the timing so tight. They have a very high-speed digital communications system between all of the geophones. We had limits on how much delay could go through the system, and it was critical that we minimized delay in the design.”
A lot of effort also went into keeping the costs down. “The cards had to be at a decent enough price to go into the customer's product,” says Ryley, “because when you are selling 10,000 geophones, you can't really have $500 or $600 worth of receivers and antennas on each one of those. We had a very, very tight price point that we had to hit.”
But all's well that ends well.
The system has now been tested successfully in France, in British Columbia in an area of dense forest, and across a giant, swampy area in central Mexico-a test that lasted more than six months.