Seismic Unwired – Cutting the Cable Can Help in Difficult Spots

David Monk said Apache Corp. uses wireless seismic receivers when it cites advantages over traditional systems connected together by cables.

jpt-2013-03-seismicfig7hero.jpg
Rows of seismic receivers are stored in racks where they are recharged and information gathered from onshore seismic work is downloaded to be combined for a large-area seismic survey of the Cook Inlet by Apache Corp. Pointing out are stakes used to plant the devices made by Fairfield Nodal.

David Monk said Apache Corp. uses wireless seismic receivers when it cites advantages over traditional systems connected together by cables.

Examples offered by Apache’s worldwide director of geophysics include a prospect that straddled the border of Argentina and Chile where equipment and radio signals were not allowed to cross; a survey in the shallow waters of the Gulf of Mexico where production platforms would get in the way of streamers used to pick up seismic signals; onshore areas where the cost of crews is high, which is the case in much of the US and Canada; and the Cook Inlet where the extreme environment offshore and onshore made wireless the choice for a number of reasons.

jpt-2013-03-seismicfig1.jpg
A crew in Portugal installs wireless seismic receivers made by Sercel along a median for a seismic shoot in a city there.

“These things are more efficient systems over more difficult terrain,” Monk said. At the time of the interview nearly all the seismic exploration projects for Apache were wireless.

Eliminating the bright orange or yellow cables reduces the weight of the system, can make it easier to work around obstacles or in difficult waters, and can significantly reduce the profile of seismic exploration in areas where that activity may not be welcome.

A case study by Geospace, a maker of wireless receivers, of a seismic shoot done on the plains of Colombia by Pacific Rubiales Energy Corp., concluded a crew of 21 could lay down the wireless seismic receivers that would have required workers if it had been wired.

And a side-by-side comparison by Apache in the Cook Inlet showed the quality of the seismic data gathered on land was similar for wired and unwired receivers, but wireless performed far better offshore in an area known for its difficult tides and currents.

Apache is a wireless pioneer and is far from the norm in an industry where the largest seismic receiver provider, Sercel, estimates 90% of the seismic channels in use are wired. But the rapid takeoff of wireless receiver sales suggests that is changing.

In most cases one channel, with a geophone that detects the echoes used for underground mapping, is equal to one wireless unit. Wireless multichannel units, though, are a growing part of the market.

In the past, wireless receiver sales represented about 11% of new seismic receiver equipment, said Roy Kligfield, chief executive officer of Wireless Seismic, which began selling units in the past year. In the past year, he said, the share has gone to 25% of sales.

The biggest wireless player, with nearly half of that market segment, is Geospace. The company, which recently dropped Oyo from its name, delivered 45,000 channels from 2008 to 2010, rising to 72,000 in 2011, and 158,000 in 2012. Over the last 3 months of 2012, the units sold or put into its rental inventory exceeded the total for the previous year.

Total wireless units rose about 50% in 2012, said Jack Caldwell, chief geophysicist at Geospace. Those totals are small compared to the wired market where the dominant player, Sercel, has sold about 4 million channels of its two current models of cabled equipment, as well as 150,000 cableless ones, said Malcolm Lansley, vice president of geophysics for Sercel.

Those interviewed for this article see wireless nodes growing over time, though far more in some parts of the globe than others. Lansley said in 5 years unwired units could dominate the North American market due to the costs and the obstacles in many places, both the physical barriers and permitting issues.

Wired is likely to remain the choice in open areas, where cable can be quickly laid using a vehicle, such as deserts, where drifting sand can swallow up wireless units and the valuable data they hold.

As for sound quality, that appears to be a non-factor, because the geophones or hydrophones at the heart of the units are generally the same whether the unit is wired or unwired.

jpt-2013-03-seismicfig2.jpg
A wireless node planted in the ground near Cook Inlet for a seismic shoot by Apache.

The evolution of seismic data-gathering techniques is a factor generally favoring wireless. Seismic tests designed to generate large, detailed images are using extremely large numbers of closely spaced receivers, with some exceeding 100,000 units spaced less than 10 m apart. That adds to labor costs. Generally labor expenses are now a bigger factor than the cost of the electronics, Lansley said.

Eliminating wires can also eliminate limits. “It gives you the freedom to design a survey not based on a set interval (determined by cable spacing) but on the object of the survey itself,” said Jorgen Skjott, vice president sales for Geospace.

Dumb but Dependable

Early sales of wireless seismic devices have been dominated by nodes that are described as dumb, because they do not communicate. They are small, relatively simple units designed to “sit there and record data,” Monk said.

Replacing the wires in these nodes is a GPS unit with an extremely precise clock synchronized using a satellite connection, enough memory to store a large amount of data, and a rechargeable lithium ion battery.

Nodes are set out, usually for no more than 2 weeks, then collected. The data is time-stamped using the clock in the GPS unit and stored inside the unit. That allows it to be downloaded and combined with data from other nodes to create a wide-area underground seismic image.

Based on its experience using nodes for large-scale projects, Apache is comfortable without real-time updates. The reliability rate for wireless nodes can exceed 99%, according to Mike Yates, a senior staff geophysicist for Apache, whose projects include the Cook Inlet survey. That is well within the company’s specifications for seismic contractors.

When shooting a survey using dumb nodes, “You do not know if you are recording data or not,” Monk said. While Apache regularly uses them, some companies will not use dumb nodes because “they want to know what the data looks like while it is being recorded.”

The reliability of wireless nodes has improved rapidly since their early days, Monk said. Apache and BP backed one of the first, the Firefly developed by Ion. Early on there were times when many of those early units were not working, he said. Now when a node does fail, it is not often the result of an internal failure of the unit.

“Typically when we get one (node) that is bad, something is wrong that is explainable. Kicked over by a cow. It is not typically an instrument failure,” Monk said. Many cable-less units come in two pieces connected by a wire that is sometimes severed by chewing rodents.

Lansley said there are two groups of users: Some clients are comfortable with the risk of failures—he puts the risk of missing or bad data due to electronic failures at less than 0.1%—while others insist on seeing the data as it is gathered to ensure quality.

With large spreads of nodes, a few units failing will not affect image quality. External threats can be a bigger issue. “The problem is if you have theft or damage to equipment it usually is caused by groups of locals in area,” Lansley said. “It becomes a problem for the shoot if there is a large hole in the acquisition spread that can severely degrade what you get.” That risk is there whether the equipment is wired or not.

To respond to these threats, some makers of nodes have added communication capabilities in the event of trouble. These features range from status updates indicating if the unit has been moved and if it is functioning correctly to broadband connections able to allow remote downloads. In places where seismic surveys draw a hostile reaction, wireless receivers can also be buried.

Cabled seismic systems offer what can be a reassuring stream of information of the status of the system and the data as it is coming in. But if a wire fails it generally means work has to stop to fix it because a string of receivers may be offline.

“We recently completed work in the Middle East in an agricultural area where we were having the cable cut by farmers nine or ten times a day,” Monk said. “It was slow conducting the operation and getting the work done.”

What Is Really Smart?

Wireless covers a wide range of systems, including smart ones designed to allow real-time monitoring and data collection. There is a large part of the market that requires it.

“With some of those who work with the older geophone systems, there is a comfort factor in seeing data. I have to admit I was one of those people who worked when we were recording with 24 channels,” which made monitoring for quality problems an important job, Lansley said.

Wireless Seismic sells units for recording and transmitting data that transfer data along the line of receivers during a shoot “like a bucket brigade,” Kligfield said.

“The connectivity is exactly like a cabled system without the cable,” he said. “The technology will succeed if we make it easy to do that.”

Sales to a few early users began last year for the company, backed by Chesapeake Energy and Energy Ventures. Its customers include MicroSeismic Inc. which is using it to monitor fracturing jobs because the system is easy to set up and the data is available for processing and interpretation as it is collected.

Data flows along the line, bypassing units if they are down, is consolidated at gathering stations and sent along to a trailer. Rather than collecting the nodes after the shoot and then combining the data, like putting together a puzzle, the system provides immediate feedback, Kligfield said.

Sercel has a Wi-Fi system for “harvesting data” by driving by with the data collection unit in a vehicle, or even using a helicopter flying comfortably above the tree line, Lansley said.

There has been some blurring between wired and wireless. Sercel has done jobs using cabled receivers supplemented by unwired units in difficult-to-reach spots.

As for what systems are smart to use, Monk said the choice depends on the job. A smart system has the advantage of allowing an operator to track what is going on at any time, but compared to dumb units, “these systems get complicated, are more expensive, and are more likely to have a problem.”

jpt-2013-03-seismicfig3.jpg
Wireless seismic receivers, which are roped together for easier retrieval, are lined up to be dumped off the back of one of the boats in the small fleet used by Apache to survey more than 200 square miles of Cook Inlet.

Battery Problems Included

When using wireless seismic receivers, battery life can become an obsession. Makers of wireless units are offering longer-lived units, but for Apache the rule of thumb is the batteries in the units can go about 14 days on a single charge.

Making sure all the batteries are working “is not hard to do if things are going smoothly. If you are having problems in the field such as bad weather or local people disrupting the operation” it can become a logistical nightmare, Monk said.

jpt-2013-03-seismicfig4.jpg
Workers from SA Exploration ready wireless seismic nodes for deployment. The water-tight devices were used by Apache for the largest ever 3D survey of the Cook Inlet in the summer of 2012.

For competitors in this crowded market, battery life is a critical point of comparison. Wireless Seismic’s Kligfield emphasizes his unit has a battery life of up to 21 days for normal seismic duty. One maker has said it is working on a unit that could go 60 days without a charge. Monk said he has yet to test it. Geospace’s Skjott said the company is selling nodes designed to last 45 days, but rarely remain out for more than a couple weeks because they are constantly being moved for the next shot.

Battery weight is also a factor when comparing wired and unwired. Eliminating the weight of the cables requires adding the weight of the battery.

In a 2008 study Lansley estimated the weight of a wired system is about equal to an unwired one when the receivers are spaced 50 m or more apart. Since then nodes have gotten smaller, so the breakeven length is shorter, but if the job calls for nodes 12 m apart, he said wired is lighter. Since equipment purchases are long-term investments, though, buyers typically buy wired recievers with a standard 50-meter spacing allowing them to do a wide range of jobs and get around obstacles, so the weight of the wires is usually present even when not needed, said Caldwell of Geospace. Plus the cables add volume, which is an issue when deliveries are made to remote areas, and Lansley said crews can put out unwired units faster because wires are bulky and workers have to “wrangle with the cable.”

The common wisdom is that a shakeout in the crowded field of wireless providers is likely in the years ahead, leaving only a few companies as is the case in the wired business. In the short run the field may grow. One of the biggest names in the computer business, Hewlett-Packard, is working with Shell to create a small node built on a sensor chip mass-produced for use in computer printers, called the MEMS chip.

The pair has said they want to create a seismic receiver with the cost low enough to allow seismic shoots with massive numbers of receivers—the number 1 million is often mentioned in the industry—for exceptionally detailed images.

In the meantime, other makers are pushing to create wireless devices that will leapfrog those two corporate giants, in a fast-changing business where there is a common belief that things are finally going their way. “Our market research suggests cabled systems will be a minority in 3 to 5 years,” Kligfield said.

Seismic Unwired – Going Wireless in Difficult Terrain

The Cook Inlet has a long history of oil production, with more than 1.3 billion bbl produced and the potential for more. But ever since production peaked at 230,000 b/d in the 1970s, there has been little exploration as it dwindled to around 12,000 b/d.

jpt-2013-03-seismicfig5.jpg
Crews unload land bags filled with seismic receivers for recharging and downloading the data gathered for a survey around Cook Inlet.

Stories about the decline point out that the Cook Inlet was eclipsed by far larger finds on Alaska’s North Slope, such as Prudhoe Bay.

Still it does offer the potential for significant conventional oil finds onshore and in shallow waters offshore.  A 2011 assessment by the US Geological Survey (USGS) covering the entire inlet estimated there could be up to 1.3 billion bbl of oil to discover—its mean estimate was nearly 600 mm bbl of oil plus 13.7 Tcf of natural gas from conventional formations. Plus significantly more gas from unconventional ones.

Unlike the North Slope, there is a ready market for natural gas. The nearby city of Anchorage depends on the inlet for natural gas and is concerned about dwindling local supplies, and the location offers transport links not available in the far north.

Past exploration suggested the underground structures likely to hold hydrocarbons could readily be identified using 3D seismic.

But there have been few surveys using advanced seismic technology and they generally covered small areas. That is not surprising considering the inlet’s treacherous tides, punishing winter weather, endangered Beluga whales, and the brown bears thriving in the wildlife preserves on its shores. Not to mention the permitting process.

That all changed when Apache Corp. began an aggressive leasing program a few years ago, gaining rights to explore more than 1 million acres. The big independent exploration company saw the potential for discoveries, particularly in unexplored deeper horizons, and based on its experience elsewhere it believed it could deal with the problems associated with doing a large-scale seismic survey.

Obtaining permits to shoot seismic in the inlet required creating a monitoring program, with specially trained and equipped observers making sure the operation followed regulations that can shut down seismic work if certain endangered mammals are within a 6-mile radius, and where sounds could affect animals.

Putting out the sensors needed to record those echoes presented another set of natural challenges in waters where tides can change the water level by as much as 35 ft. Currents up to 7 mph (6 knots) rule out towing streamers that  move with the current, which also has presented problems for strings of seafloor receivers wired together using cables.

“There were horror stories in the past of cables drifting in the currents. Currents are a big problem,” said Mike Yates, senior staff geophysicist for Apache, which also experienced electrical problems with a cabled system, caused by the salt water.

On land, laying out long lines of receivers connected by cables was a problem because the permits banned cutting through the brush to create the relatively straight paths needed for cables, and forced detours around protected areas, such as bear dens.

The work on land was mostly done in the winter, despite extreme cold and blizzards that dumped more than 11 ft of snow in 2011–2012.  Crews had to dig out the snow for the drills used to place explosives and to land helicopters that delivered bags full of the wireless receivers, known as nodes. Each of the wireless receivers weighed less than 5 lb each, about one-quarter the weight of the single wired receiver and the cable that goes with it, allowing more wireless stations to be delivered per load.

Doing the work in the summer was out of the question because during that time of year “southeast Alaska is like a jungle though it is not quite as warm,” Yates said. “There are parts you cannot get across because of swamps and bogs and there is the ever-present danger of bears.”

Wireless Options

Given the obstacles, using seismic receivers without cables was an obvious possibility for Apache, whose experience with them dated back to helping fund development of one of the first wireless seismic nodes a few years before.

But first it did a test of its seismic options with two lines running side by side over 18 miles from the land, through the mudflats bordering the inlet, and then into water as much as 200 ft deep.

One line used traditional strings of seismic receivers connected by cables carrying power and data, and the other used wireless nodes stuck in the ground with a spike. Since the test was done in late winter and early spring, an armed wilderness expert accompanied crews to protect them from the bears living there, or assist if there was an emergency or they were stranded overnight.

A number of other variables were considered from the most effective charge size for use onshore—more was found to be better—to the best depth to place it—35 ft based on problems with drilling deeper holes.

When it came to the receivers, wireless nodes won based on how they performed in that difficult environment. “The nodal recording system matched the data quality of the cabled data system for the onshore program, but clearly outshone it offshore and in the transition zone,” said Yates in paper OTC 23771, delivered at OTC’s 2012 Arctic Technology Conference.

Onshore wireless receivers allowed the crews to walk around the obstacles, planting the nodes in spots marked by surveyors. Wires limit detours, which is a problem on land where straight-line paths were not available.

“To use cables would be very difficult. There are exclusion zones for bear dens or no-permit areas,” Yates said. “We were not allowed to cut lines (through brush). There had to be no impact on the ground. We had no cutting crews whatsoever.”

In the water, the problems with cables were not limited to the currents.

“Multiple efforts were made to lay out and power up the cabled line, but none were 100% successful; problems with line power and leakage prevented it from ever being fully operational,” Yates said in the paper, adding that with the cabled system “there are lots of electrical connections in the water. Electrical connections in salt water are prone to failure.”

A second problem was the wired receivers need to remain connected to a boat during a shoot. Keeping a boat in place would be extremely difficult given the rapid flow of the tides.

In contrast the self-contained, water-tight nodes from Fairfield Nodal performed reliably in the saltwater, and the 65-lb units, which are shaped like curling stones, tended to stay put. If a node was out of place, Yates said the data could be adjusted based on location information from the instruments inside the receivers.

Another positive for the one-piece units, which record sounds for later retrieval, is they look less imposing than wired ones which often come with a large battery to power a section of the line.

jpt-2013-03-seismicfig6.jpg
This outlines the conventional and unconventional resources evaluated by the US Geological Survey. The yellow defines the area with two conventional units holding oil and gas. The red holds two gas-rich unconventional deposits. Apache’s survey covered a swath across the central section of the Inlet about 50 miles from Anchorage.Source: USGS

“It simplified a lot of permitting issues,” Yates said. “With cabled system you see cables and batteries and hydrophones or geophones and it looks very intrusive. The thing with the node system is the battery is inside. To the public it looks like a lot less intrusive system.”

Going Beyond Expectations

The ultimate test was Apache’s year-long survey, running survey that covered 320 sq miles. They began in the fall of 2011 and worked through a winter on land, covering about 100 sq miles, during a period when the snowfall set a record in Anchorage, the nearest weather reporting station.

During the spring, work proceeded in the shallow waters in the transition zone on the edge of the inlet, and during the summer Apache covered the deeper waters with several boats, seismic source boats created the carefully timed sounds. Other vessels placed the nodes, dropping them off the end of the boat in an assembly line fashion, and later retrieved them using ropes attached for the purpose.

“We got through the entire summer season on Cook Inlet offshore without any spread-related downtime,” Yates said. “We acquired 220 sq miles of marine data. That is unheard of in that environment.”

One positive was there were no recordable Beluga whale sightings, which could have delayed work.

And Apache exceeded its planned work in Cook Inlet for the summer of 2012 by changing its routine to reduce the number of shots needed. The original plan was to lay out a 5-mile-long array with 6 rows of 160 receivers each, then shoot the patch and move all the receivers to the next block and repeat the steps.

Instead it changed to a method used in land surveys, known as a segmented roll. Yates said that once the shoot had progressed 3 miles down the line, they would pick up those nodes at the rear and move them to the front. “The original plan would result in a lot of inefficiency. You would have to take more shots twice,” he said.

At the end there were only two minor fuel spills, and the seismic contractor on the job, SA Exploration, reported no recordable incidents.

While Apache is not offering predictions of what it might find when it drills, the seismic survey suggests the potential is around the high end of the USGS estimate. John Bedingfield, vice president for exploration and new ventures for Apache, said in an investor briefing that, “The company’s analysis ‘strongly’ suggests another 1.3 billion to 1.4 billion bbl of oil yet to be discovered.”

Drilling began late last year based on what was learned from the seismic program that is expected to require 2 more years to complete.


For Further Reading

  • OTC 23771 Seismic Acquisition in Alaska’s Cook Inlet by Mike Yates, et al. Apache Corp.
  • SEG 2011–0112 Wine Country 3-D: Cable-less Seismic Acquisition in Mendoza Province, Argentina by Mike Yates, et al. Apache Corp.
  • SEG 2009–0112 Seismic Sans Frontières—Cross-Border 3D Acquisition in Tierra Del Fuego by Mike Yates, et al. Apache Corp.
  • US Geological Survey Assessment of the Cook Inlet http://pubs.usgs.gov/fs/2011/3068/fs2011-3068.pdf