An artificial intelligence software program designed to identify and react to drilling issues like a human worker would can increase drilling efficiency, reduce costs and retain knowledge of best practices and lessons learned in drilling. This form of artificial intelligence, case-based reasoning, looks at data streaming from a well site in drilling operations and makes human-like decisions based on data that can help less experienced users make correct choices and disseminate best practices in an automatic form, Phil Wade, COO of Verdande Technologies, told Rigzone in an interview.
Verdande’s software, DrillEdge, allows oil and gas companies to leverage Big Data to bolster efficiency in their operations, said Wade. By focusing on early identification of non-productive time (NPT), such as differential stuck pipe, NPT can be reduced to zero. Verdande’s software can identify potential complications and optimize performance in terms of bit wear, mud motor damage, hole cleaning, lost circulation and hole ballooning and drillstring washout.
“All these things have symptoms that people may not be trained to see,” said Wade, who has served in several roles at Verdande since joining the company from Halliburton in 2010. These symptoms may be minor, but frequent, and should be noted. “In an ideal world, if industry reacts to the case studies like it’s supposed to, then NPT doesn’t have to happen.”
The company’s radar system encompasses all of Verdande’s solutions and puts them on a radar screen in pie slices for reach potential problem. The screen is blank when no potential problems are on the horizon. When a potential problem arises, a blip, or case, will appear on the system’s interface screen. A case, which can contain best practices or other information that a customer wants in the case, is then presented to the end user on steps they should take to correct a problem. The user must click on a case due to the possibility of more than one case being on the radar at the same time.
When the case is clicked on, the case will present the steps a worker needs to take, such as talking with the well site workers and alerting the drilling engineer. It may then ask the worker to reference a company’s best practice. If nothing works, the case alerts workers to call the well site and stop drilling operations. The information is presented in a way so that end users don’t have to hunt for what they need at that moment, said Wade, adding that no data needs to be manipulated.
“Anyone working at a site can use it.”
The system automatically alerts users not logged in by email of potential problems so they can log on and address the issue. It also includes collaborative tools that allow end users to comment on what they see, and for other end users logged onto the system to see those comments. Data also can be accessed online from a computer or iPad in the field.
“As long as you have access to the cloud, you have access to these cases,” said Verdande.
Developed through research at one of the world’s foremost experts in case-based reasoning, the Norwegian University of Science and Technology (NTNU) in Trondheim, Norway, this technology allows an operator to take information they’ve learned in previous operations and apply that information to situations that are similar, but not exactly the same. In the past, case-based reasoning had been used to analyze historical data, but Verdande’s now Chief Technology Officer, Frode Sormo, wrote a paper on applying case-based reasoning technology to recognize patterns in real-time data.
Norway-based Statoil ASA saw the benefit of the technology, and allowed its historic data to be used in developing the idea into a prototype, said Wade. The application of case-based reasoning to a flowing datastream had never been attempted before; theoretically, if successful, then nearly any real-time data could be used from any industry or application, Wade said.
In 2004, Verdande raised enough money to begin designing solutions for certain data sets; the technology underwent several revisions and was used in smaller projects from 2004 to 2010. Through its analysis of well data sets to develop cases – for which Wade gave a conservative estimate of 400 sets of data from 400 wells – Verdande was able to see patterns and symptoms emerge, which can enable better decision-making in drilling globally.
“There isn’t any place in the world you could drill today where the symptoms leading to a mechanical stuck pipe would be different,” said Wade.
This analysis can be conducted for thousands of wells worldwide or in a specific region.In 2010, the company received an infusion of investment due to Verdande’s success with stuck pipe solutions. That same year, the company opened its Houston office, which houses the sales and operations arm for the group. Since 2010, the company has developed case-based reasoning for other challenges operators may face in drilling, such as twisting off. Over the past three and a half years, the company has created six new case-based reasoning solutions.
Verdande has used its analysis to build a case library. At this time, the start case library consists of an average of two cases per problem area, which is 17, but this number changes as Verdande builds more robust generalized cases or adds new problem areas. Wade believes these cases can be deployed on any given well, and the behavior will be accurate enough to alert someone to a potential problem. A case has multiple events, or symptoms, said Wade. Each event is generated by an algorithm captured by a case.
Case-based reasoning applications are being used in other industries, Verdande is the sole owner of the intellectual property for case-based reasoning for oil and gas, Wade noted.
Though it’s targeting the onshore and offshore drilling markets, Wade believes Verdande’s case-based reasoning software is a better fit with land operators at this time due to the lower margins in operations that operators can risk, lower level of services and greater need for economies of scale.
“Land-based operations have not typically required the same amount of advanced to ultra-advanced solutions that deepwater uses,” Wade noted. “However, this is changing with greater use of rotary steerable assemblies in unconventional plays, as well as greater overall automation process and uptake of technology like ours.
“I think we are one of the companies influencing this sea change,” Wade commented.
The software can be deployed across 50 wells in a factory-drilling approach to development, and by identifying inefficiencies in the drilling process for a number of wells, an end user can take immediate action and improve their operational efficiency. The software can allow workers to monitor a greater number of rigs.
“If you start to expand the number of rigs that four workers might be monitoring in a real-time operations center to 80, it can be overwhelming for the people to monitor,” said Wade.
Traditionally, more workers would be added, by Verdande’s software, which can add additional human-based decision capability without increasing the headcount level.
“Software like ours is helping to drive the quality of data in the oil and gas industry today,” Wade commented. Historically, time-based data was around 30 second range, even five years ago, but applications such as Verdande’s is driving the increase in data resolution down to the one to 10 second range, with 10 seconds becoming less common.
FOCUS ON DRILLING OPTIMIZATION
The company is now solely focused on incorporating more optimization practices into its technology. In the factory drilling approach being used by operators on unconventional plays, operators can learn quite a bit from the few 10 wells out of 300 wells planned. By taking lessons learned from the first 10, the drilling of other wells can be planned more efficiently. This process is continuous, whether it’s the eleventh or thirtieth or hundredth well.
“Most unconventional plays involve a very cookie cutter drilling process,” said Wade. “As you drill more and more wells, you learn what works and what doesn’t and you start to effectively eliminate NPT.”
The increase in drilling efficiency is now putting stress on other things, such as equipment. Tools failures are becoming a top problem. One company that Verdande worked with in the Bakken play in North Dakota, Hess Corp., helped the company reduce NPT. Then, the operator asked Verdande if it could use surface measurements to determine how to make a mud motor live longer downhole. Verdande accomplished this by aiding the operator in drilling smart. By extending the life of the tool, the operator can drill further, effectively drilling the ideal well.
“By helping the crews at the wellsite identify when the rig’s autodrill equipment began inducing damaging oscillations as it attempted to maintain a predetermined rate of penetration, the crew could reset the equipment and continue to the end of the run without needing to come out of the hole due to the potential resulting damaged mudmotor,” Wade noted. “Our solution can work to protect measurement while drilling/logging while drilling as well, but in the Hess example, that equipment is not as important as the mudmotor problems for a variety of reasons.”
Verdande also is looking at how to expand the life of bits, or how to make downhole repairs more economic. By using surface measurements to determine whether vibrations at the surface are being transferred to downhole tools and shaking them apart, drilling can be done more smoothly, Wade explained.
The company also is looking at calculating surface stick-slip to enhance drilling optimization. When a hole is not being cleaned effectively in drilling operations, or a bit is wearing, the hole is narrowed down, increasing the friction that downhole tools experience. If the drillstring is getting a lot of friction, it will stop downhole, even if it’s turning at the surface, winding the drillstring up a like propeller on a bathtub toy, Wade explained.
This results in either the tool twisting off, or a fast unwinding of the drillstring, which is very abusive on the equipment. Sensitive electronics in the drillstring, although quite robust, will be torn apart from the 100s or 10s of hundreds of G-forces at which the drillstring will spin.
GREAT CREW CHANGE, EFFICIENCY FOCUS CREATE NEED FOR ARTIFICIAL INTELLIGENCE SOLUTION
The Great Crew Change – which is resulting in drilling knowledge being lost as Baby Boom-aged oil and gas workers prepare to retire – and a need to reduce NPT in oil and gas operations has created a need for Verdande’s software by land operators. “With a couple of our customers, we’re seeing them take six months to train a drilling engineer, and when they’re finished and the engineer’s ready to go, another company recruits them due to the attrition of workers,” said Wade.
The U.S. oil and gas industry is grappling with increased demand for workers due to the surge in U.S. shale activity. In 2013, more than 1 million jobs were added in the industry due to shale activity, according to a report from the Deloitte Center for Energy Solutions. Despite significant recruitment and retention efforts, oil and gas companies are still having a hard time meeting their hiring needs.
This difficulty is compounded by the fact that the U.S. oil and gas industry workforce is shrinking as older workers retire. The 2008 economic downturn delayed some workers from retiring, but the industry will likely soon feel the full impact of the Great Crew Change as workers’ 401K plans and savings recover and they move ahead with retirement plans.
Oil and gas companies are seeking to reduce costs and boost operational efficiency as a means of coping with the higher capital investment required to develop shale, according to a panel of industry experts at the Bloomberg Oil & Gas Conference last fall.
Oil and gas companies are focusing on decreasing costs and improving efficiency in their operations. To achieve these goals, companies are implementing oilfield service technology, said Herb Listen, Ernst & Young Oil and Gas Assurance leader, in a recent press briefing with reporters in Houston.
The manufacturing approach to shale drilling has prompted operators such as BHP Billiton to pursue “relentless” improvements in productivity, in which it aims for each shale well to be drilled in a more cost-effective and productive manner than the previous well. This focus means that companies have to be able to learn quickly from their mistakes to improve their performance, said Rod Skaufel, head of BHP’s U.S. onshore business, in a media presentation in Houston in early June.
Statoil, another company active in the Eagle Ford shale play in South Texas, also is seeking to increase efficiency in its shale operations through its holistic application of technology, Rigzone reported earlier this year.
Statoil is an investor in Verdande, along with ProVenture Management, Investinor, and Verdane Capital. Verdande’s partners include Baker Hughes Inc. and Suvira Energy. In late 2011, Baker Hughes acquired a 7.5 percent interest in Verdande, and utilizes Verdande’s software solution to enhance its own product, WellLink radar. Verdande also still sells its software directly to operators. Statoil currently uses the technology through Verdande’s relationship with Baker Hughes.
Earlier this year, Verdande partnered with the AMiDST (Analysis of Massive Data Streams) Consortium, a research program funded by the European Union, to examine advanced artificial intelligence methods to allow Big Data analysis of data streams.
The objective of the AMiDST research project is to develop a toolbox that facilitates analysis in order to improve predictability across a variety of cases. AMiDST will work with research and industry partners to test predictive analysis of data streams across a number of industries, including oil and gas, banking and the automobile industry, according to Verdande’s website.
The company also spent about a year and a half opening the healthcare and financial services verticals, and experienced success with the prototypes for these verticals.
“We worked with some very blue chip companies in that process, but have since decided that those vertical would be best as standalone efforts and we are now focused solely on expanding our energy offerings while we consider the best way forward with the other verticals,” Wade explained.