In a digital world, it’s software, not big iron, that’s driving hefty changes in well planning and construction, workflows, training, security
By Kelli Ainsworth Robinson, Associate Editor
In 1965, Gordon Moore, one of the co-founders of Intel, predicted that the number of transistors on an integrated circuit would double every two years. In other words, he believed that computing power would double every two years. According to a study released in 2015 by information technology group Experts Exchange in honor of the 50th anniversary of Mr Moore’s prediction – now known as Moore’s Law – computing power actually increased a trillion-fold between 1965 and 2015. That means that the smartphones we can fit in our hands today are more powerful than the supercomputers of the 1980s that filled an entire room.
The advances in computing power that have occurred over the past few decades have enabled software developers to build ever-more powerful and complex data analytics, machine learning and artificial intelligence (AI) tools that have transformed virtually every industry. Predictive analytics have helped the aviation and power industries avoid unplanned downtime. AI and machine learning power driverless cars. AI is also being used to improve medical recordkeeping and weather forecasting.
In the past several years, digital technologies have also started reshaping the drilling industry. For drilling contractors and operators alike, advances have targeted operations optimization by eliminating nonproductive time and invisible lost time, keeping people out of harm’s way, and maximizing maintenance efficiency. Technologies like advanced data analytics, AI, machine learning and virtual reality (VR) are also changing how companies manage their inventories, determine their well plans and fracture placements, and even how they train their employees.
The potential value that can be created with such technologies are being measured in the billions of dollars, said Andrew Slaughter, Executive Director of Deloitte’s Center for Energy Solutions. “That’s everywhere from uptime operations, increased production, preventative maintenance, logistics, supply chain and more rig automation.” Because the upstream industry spends around $500 billion or more annually to maintain and grow operations, even a 1% increase in productivity from the application of digital technologies could yield up to $50 billion in gains per year, he added.
Even as companies are increasingly putting digital technologies to use, the cost of digital technologies are falling. This is particularly true of sensors, which are the backbone of advanced analytics and digital technologies. “Low-cost sensors are really something that has come about in recent years, so putting sensors on equipment to gather the relevant data to improve reliability and the performance of your operations in the field has been a major part of the industry’s digital transformation,” Mr Slaughter said.
However, the digital transformation is about more than technologies – company cultures and business models also must evolve, said Damon Vaccaro, Principal at Deloitte Consulting. In particular, to succeed in the digital world, companies need to embrace greater levels of collaboration and seek out business partners that they might not have considered before, in addition to their traditional business partners.
“As companies become more digital, it’s not just a technology play. It’s a very comprehensive play that requires them to look at their operations and culture and the way that work gets done,” he said. “In the digital world, companies need to look to startups. This is immediately applicable to drilling, because there are different startups out there focused on AI and advanced analytics.”
Digitalizing the core
One major operator that is “digitalizing its core” is Shell, which the company defines as applying existing technologies to existing business models in the upstream space by using advanced data analytics to improve inventory management and seismic data interpretation, Martijn Dekker, Vice President of Strategy, said at the 2018 OTC in Houston on 30 April. “If you think about digitalization, it’s about turning data into actionable insights – into money. I do believe digitalization will transform our industry and create value,” he said. “I believe the biggest value in the near term for the offshore industry lies in digitalizing the core – making better use of the technology that’s already out there to bring value and create operational efficiencies.”
One way Shell has taken this approach is in terms of seismic data interpretation, by applying machine learning to high-resolution and 4D seismic data sets. When geologists manually pick faults from such data sets, this is a time-consuming process. By automating this process, geoscientists are freed up to focus on higher-value activities, Mr Dekker said. A plausible next step, he added, would be using AI to help identify drilling targets. “It’s not a big stretch of imagination to see how we get a computer to pick our next exploration prospect.”
Inventory management is another area where the operator is working to better use existing data to optimize performance. Shell has developed an analytics tool that looks at data, like when a particular piece of equipment has been ordered and shipped, to determine what equipment needs to be stocked and how many spares should be made available. Ensuring that the necessary components are always available can also help to optimize maintenance schedules and prevent delays and downtime that could occur due to waiting on a part. This approach has already allowed the operator to reduce inventory costs in the Gulf of Mexico. “By scaling this up and replicating it across the globe, we’re talking about hundreds of millions of dollars in savings that could be realized,” Mr Dekker said.
In the downstream sector, Shell has moved beyond digitalizing the core and applied new technologies to create new business models. For instance, Mr Dekker said, the company has developed a digital platform through which consumers can order gasoline to be delivered to their home or workplace, rather than having to fuel up at a gas station. “It’s not unforeseeable that we could do something similar (in the upstream sector), creating an offshore Uber where essentially ships show up when you need them,” he said.
In order to develop solutions at a pace that keeps up with the advances in computing, Shell is taking what it calls a minimum viable product approach to digital projects, Mr Dekker said. Often, the development of a new solution or technology can take three years. However, he noted, because computing power advances so quickly, by the time a solution is ready, it may already be obsolete. At the same time, the end user typically doesn’t see or handle the project until it’s complete, so no feedback is provided during the development process. Under the minimum viable product approach, a usable product is rolled out after three or four months. Based on end-user feedback, more features or capabilities can be added to the product over time, he said.
Applying AI to the subsurface
The industry has made considerable progress adopting digital technologies in areas like inventory management and predictive maintenance, where companies can apply lessons learned from other industries. In fact, those are the areas where there have been the biggest gains in oil and gas. “Whether you’re in retail or automotive, you use similar AI tools to stock your inventory,” said Barry Zhang, CEO of Quantico Energy Solutions, a geoscience-based AI company.
On the other hand, certain challenges are unique to drilling, like characterizing the rock properties of a well that’s being drilled. “But how do you drill that well in deepwater as best as you can or how do you drill that well in West Texas as best you can? That, you can’t look to other industries for.” Quantico is looking to advance the development of AI to characterize the subsurface and rock properties to optimize drilling efficiency and fracture placement.
The company applies proprietary AI algorithms to data gathered during drilling and stored in the electronic drilling recorder (EDR), along with historical field data supplied by the operator and from Quantico’s previous field runs, to produce well logs that describe the rock qualities, including shear, compression and density. Logs created using its AI algorithms have achieved the same accuracy as logs produced with traditional logging tools, according to Mr Zhang. In 2017, a supermajor qualified the AI-based logs to be the same accuracy as their deepwater LWD tools. However, since the logs are produced by analyzing data that’s already been collected, there’s no need for an operator to run an LWD tool while drilling. The technology has been deployed in more than 150 wells to date, primarily on US land but also in the deepwater Gulf of Mexico and the North Sea.
Over the past nine months, the company has been rolling out a real-time drilling service, QDrill. It analyzes the gamma ray, survey and data from the EDR, such as weight-on-bit (WOB) and rate of penetration (ROP), to determine the unconfined compressive strength (UCS) of the formation and the mechanical specific energy (MSE) going into the rig. “The UCS is the energy needed to break the rock. If you’re drilling efficiently, the energy going to the rig, or MSE, should trend similarly to your UCS,” Mr Zhang said. “In other words, you should be applying the amount of energy to break the rock plus the amount of energy needed to rotate the drill string.” If the MSE and UCS are not in sync, it means the rig is not drilling efficiently, which may mean a drilling dysfunction is occurring and/or the drill bit and cutters could be experiencing premature wear.
This information could help the driller to make real-time decisions, such as to slow down in order to save the bit, or to speed up in well intervals when the formation isn’t as hard and drilling at high speeds won’t cause inefficiency. “If a driller has no visibility into what’s actually happening in the rock in real time, that’s the equivalent of driving your car based on the feel of your steering wheel and suspension without seeing the road,” Mr Zhang said. “Showing how the rock is changing with the AI, you actually get to see the road.”
The baseline QDrill service uses AI to provide real-time data on UCS and MSE, and an add-on service can advise the driller on how to drill more efficiently by applying AI to determine how parameters, like WOB or RPM, should be adjusted.
Because the service leverages data that’s already being collected, no new hardware or software is installed directly on the rig. If the drilling data is collected and streamed through an EDR system, the oil company can give Quantico permission to access the streaming data. That data is then processed through software and interpreted by AI algorithms, and finally shared with the oil company in real time through a drilling dashboard delivered via the web. If an oil company has an in-house data management and streaming software, Quantico can access the data directly from those internal systems.
To date, QDrill has mostly been used as a lookback service – to review data that was collected while drilling to produce logs showing where dysfunction occurred. Currently, Quantico is pursuing more field deployments. As a rule, the company targets up to a 20% reduction in total drilling costs through the use of the service in real time. In one lookback application in Permian, it was applied to data collected from three wells from three different pads, each with approximately 5,000-ft laterals.
When the drilling data was analyzed using QDrill, the service was able to identify several barriers for drilling performance. For instance, there were frequent transitions between hard and soft rock. Transitions from soft to hard rock can accelerate bit cutter wear, while transitions from hard to soft rock can cause toolface deflection and unintended micro-doglegs, Mr Zhang said. The service, had it been run in real time, would have identified these transitions through fluctuations in UCS, and the oil company would have been able to adjust RPM and WOB and minimized premature bit wear.
Additionally, when the service was applied to the drilling data that was collected from the well, it was able to identify a rise in MSE but flat UCS, which meant the rock was staying the same, yet the energy going into the drill string was increasing. This change indicated stick-slip was occurring and the driller needed to slow down ROP. Further, the service was able to identify areas where the rock was soft and the driller could have picked up speed. Overall, “we were able to identify intervals where the operator could have drilled faster or remediated dysfunction earlier in about 30% of the lateral footage had they used the QDrill service (while drilling the well),” Mr Zhang said.
For fracturing applications, the company’s QFrac service leverages related AI algorithms to determine the Poisson’s Ratio, Young’s Modulus, brittleness and minimum horizontal stress of the formation. This data is useful for targeting fracture placement. By knowing which portions of the formation have similar stress measurements, for instance, the operator can better place their frac stages and perforation clusters to ensure consistent fracture propagation, which improves stimulation, Mr Zhang said. Additionally, the data supplied by the service can help operators see where stages could be lengthened or where two stages could be combined into one. This knowledge can translate into lower completion costs. “Operators can pump that job with less time if they’re combining stages – without compromising production,” he said. “With fewer stages by definition, you have less intervention. You don’t have to drill out as many plugs, which means you don’t have as many opportunities to stick tools. The well could require fewer fishing runs, which can cost you hundreds of thousands of dollars a job.”
Two years ago, Quantico ran this service for an operator on two wells in the Bakken. QFrac was used to calculate the minimum horizontal stress along the length of the lateral. The operator used this information to group intervals of rock with similar stress profiles by adjusting the depths of the stages and perforation clusters, Mr Zhang said. “This enables the perf clusters to be initiated more uniformly and the stages to be pumped in a fashion that reduces the chance of some fracs growing deep into the formation, due to low stress rock, and other parts of that stage having fracs that are short and, thus, under-stimulate that interval.” Because the fracture placement was optimized, stimulation was improved, and one well’s production was 40% higher than offset wells in the area, and the other well’s production was 70% higher.
Although LWD tools often aren’t run in the US land market due to high costs, many operators prefer direct measurements gathered from tools touching the rock rather than statistical methods using AI, Mr Zhang said. “The shift to AI, when done properly, brings about a hybrid approach of statistics constrained by the geophysics – however, it does take some time to get one’s arms around the statistical nature of these new tools and workflows,” he added. “We just need time to get used to using AI to identify – and react with lower cycle times – the variability that arises from the rock, which, over time, is a primary contributor to inconsistent drilling and frac operations that hurts the bottom line.”
Training a new generation
Training new and rehired employees as contractors reactivate rigs is a major challenge for the drilling industry – but it’s another challenge that can be met through the adoption of digital technologies. Drilling Systems’ Virtual Reality (VR) and On-the-Rig simulators (OTR), mobile versions of the company’s high-fidelity simulators released in 2017, are being used to create immersive training environments that realistically simulate the working conditions on the rig.
“Due to the downturn, there’s been a significant amount of the workforce that has left the industry voluntarily or involuntarily,” Jim Krupa, Regional Director for Drilling Systems, said. “We’ll have a lot of people coming into the industry that are completely green, and then you’ll have people that have been in the industry for a long time, but maybe they haven’t worked in 12, 18 or 24 months and they might be in need of a competency reverification. So that’s one of the things that we’re really focusing on.”
Many of the employees entering the industry are from the Millennial and Generation Z age groups and are digital natives who have been using computers, smartphones and tablets since they were children. Technology advancements in VR and computer modeling that create highly realistic simulation environments provide the sort of hands-on, high-tech means of learning that younger employees are used to, Mr Krupa said. In fact, because of cultural and technological changes, training in general is shifting from 70% classroom and 30% practical to 70% practical and 30% classroom-based, with a large emphasis on simulation exercises.
At the same time, as technologies on the rig become more complex, it makes little sense to train employees in the classroom and on low-fidelity simulators if they’re going to be working with high-tech equipment and controls. “Working on a rig now requires understanding computers and technology, so that’s why using simulation’s a great way to enhance that,” Mr Krupa said.
Further, high-fidelity cyber simulations – realistic 3D simulators equipped with cyber chairs – and VR training can better prepare employees for their jobs because of how faithfully they can re-create the rig environment, said Ian Hudson, CEO of 3T Energy Group. 3T Energy is a new training company formed by Drilling Systems and AIS Training in late 2017. The company’s simulators put trainees in immersive environments where they can experience low-frequency scenarios, like a blowout or major equipment failures, in a safe manner. “It could take seven or eight years to truly be a competent driller using more formal on-the-job training methods. If you’re using technology and simulation, you can get years’ and years’ worth of experiences in a much more condensed period of time and use this to verify and ensure competency, not just time-based experience,” he said.
3T Energy will leverage Drilling Systems’ high-tech simulation tools and AIS Training’s expertise in developing and delivering training curriculums and its competency management software, Transform. The software can help companies, including drilling contractors, create an individualized training path for each employee, based on what training the employee has already received, what training they need and their career advancement goals that are mapped to required competencies, Dr Hudson explained.
“It’s really a learning journey,” he added. “If you’re an assistant driller and you want to progress to driller, you’ll have a journey mapped out ahead of you that will require certain skills to be measured. You’ll be able to access the tools that give you the knowledge you need and help you demonstrate that you have the necessary skills.” For instance, he said, the training program for a rig employee might include an e-Learning well control course, a VR exercise for familiarization with managed pressure drilling systems, and a team-based simulator exercise for crew management skills. This provides a blend of technical and non-technical competency. “We want to help the industry move from a compliance-based approach to training to a competency-based approach,” he said.
Advances in computing power have enabled improvements in Drilling Systems’ Unity 3D graphics engine, allowing the company to create VR training modules. Through VR, new employees can take virtual tours of a rig to get a feel for the equipment with which they will work. They can also practice procedures and processes, like lockout/tagout, in the virtual environment. These simulations also allow employees to experience and learn from their response to events like dropped objects, Mr Krupa said. “The people who have used this VR tool report it as being very realistic. People jump backwards and are caught by surprise, just as they would if this event happened in real life.”
Over the past few years, Drilling Systems has seen the popularity of VR training increase rapidly. “We are seeing an increased demand, almost weekly, for new VR projects,” Mr Krupa said. He believes this demand will continue to pick up, especially as advances in computing power and the processing power of graphics engines increase, making VR hardware cheaper and more compact.
While there are VR headsets, like the Oculus, that connect to a smartphone, high-fidelity equipment used to create VR simulations for scenario-based training requires a powerful graphics card and strong internet connection if used in multiplayer mode. These factors limit where VR training can be run for the time being, Mr Krupa said. “The tech industry is always finding ways to make hardware, lighter, faster and cheaper, so this is likely to improve in the near future.”
As training in the drilling industry becomes increasingly practical and hands-on, universities are beginning to build simulator-based training into their curriculums, Mr Krupa said. “They’re passionate about understanding what the engineer of the future looks like. And what do companies like BP and Chevron and Halliburton want the graduating staff to look like?”
Drilling Systems has sold drilling, coiled-tubing, wireline and crane simulators to the petroleum engineering departments at universities, including the University of Oklahoma, Texas A&M University, University of North Dakota and University of Louisiana Lafayette. Most recently, the University of Wyoming ordered Drilling Systems’ simulators for the school’s Drilling and Completions Simulation Laboratory, which will be housed in the new Engineering Education and Research Building that is scheduled for completion in 2019.
While digital technologies are offering new levels of insight, connectivity and optimization to the drilling industry, they are also creating new challenges. “Interconnected and integrated systems open you up to possible cyber threats,” Davinder Singh, Principle Business Developer for Siemens, said. To combat these threats, the drilling industry needs to build its networks with cybersecurity already in place, he added. “We need to build in cybersecurity concurrently, rather than taking a Band-Aid approach later.”
In April, Siemens and Norwegian industrial cybersecurity firm Secure-NOK formed a partnership to integrate Secure-NOK’s SNOK cybersecurity offering into Siemens’ industrial communication platform. This includes the Ruggedcom RX1500 industrial router and switch and Ruggedcom APE computing platform, which power automated processes and visualizations. Under this partnership, new Ruggedcom devices will come loaded with SNOK cybersecurity modules designed for operational technology (OT) networks, which can also be retrofitted on existing devices and networks. The modules include a network intrusion detection system that monitors the internal communication of control systems and endpoint monitoring, which detects threats related to servers, data historians and the supervisory control and data acquisition software that powers automated processes. The SNOK software modules leverage machine learning to perform advanced behavioral analysis on the network to establish a baseline behavior, allowing the software to quickly detect anomalies, which could indicate the presence of a threat.
Drilling rigs are vulnerable to both internal and external cyber threats, Nina Hesby Tvedt, Sales and Marketing Director for Secure-NOK, said. Internal threats are the same threats that most people have to guard against when using their personal devices – viruses and malware that can be transferred onto the system through email or unsecure websites. These threats can then be transferred onto the rig’s drilling controls network via a connection to infected devices.
“If that happens at the wrong place at the wrong time, it can shut down critical systems,” she said. External threats include targeted attacks from state-sponsored or terror groups. “There are currently several active alerts out, from the American government, for instance, about ongoing targeted campaigns against the energy industry,” she added. “The drilling industry needs to be aware that they’re an attractive target.” The US Computer Emergency Readiness Team, a division of the department of Homeland Security, regularly issues alerts on its website.
Protecting a drilling rig’s network from cyberattacks requires three things, Mr Singh said. Awareness training can help both a drilling contractor’s employees and third-party employees who work on the rig to recognize suspicious emails and unsecured websites that could introduce viruses or malware. Companies also need to develop a process for addressing and recovering from threats or attacks. Finally, companies need to implement the right technology to meet their needs. “It has to be in a combination of all those three aspects if you want to be successful in having a secure environment,” he said.
Although Siemens has experience building and deploying communication and control networks with integrated security, it also regularly partners with other companies with expertise in providing additional cybersecurity solutions to protect those networks, Mr Singh said. “I don’t think anyone can do it alone. And as Siemens, we realize that we don’t have the answer to all the security problems out there. That’s why a partnership approach is very important,” he added.
The SNOK portfolio of cybersecurity tools is deployed directly in the rig’s control network and monitors traffic between the control network and automated equipment, like the pipe-handling system, for signs of a cyberattack, Ms Tvedt said. SNOK can also monitor vulnerable endpoints and was designed to be non-intrusive, unlike traditional IT cybersecurity solutions. It does not automatically perform any actions that could disrupt equipment function and communication, she said.
“Since we are deploying it on control and automation systems, you do not, under any circumstances, want to disturb the industrial process, which could cause even bigger problems.” The system issues an alert when a threat has been detected, so the operator can take action. Additionally, SNOK was designed to run on limited computing resources, so it consumes minor computing and processing power and doesn’t slow down the network speed.
The non-intrusive nature of the solution and the ease of installation offer advantages in securing older, legacy rigs that weren’t built with network connectivity in mind. While these rigs have been updated to enable network connectivity, they might have limited processing or network capacity compared with newer rigs that were built to run advanced software. “(It) only requires a small amount of the overall bandwidth of the control network,” Mr Singh said.
As rig automation continues to increase and rig control networks become more complex, the drilling industry’s cybersecurity needs will continue to evolve. That means companies that provide the industrial network to enable automation, as well as companies that build cybersecurity solutions, must stay agile, Ms Tvedt said. “The perpetrators (of cyberattacks) are going to find new ways to target companies constantly, so it’s a game of always being ahead and innovative.”
At the same time, drilling contractors must be willing to share experiences and lessons learned, which will enable the industry as a whole to better defend itself against cyberattacks, Mr Singh said.
In 2014, IADC recognized the need for a forum for drilling contractors to jointly identify and tackle cybersecurity concerns and formed a cybersecurity workgroup under the Drilling Control Systems Subcommittee of the Advanced Rig Technology Committee. In 2016, the task group grew from a workgroup into a subcommittee and has since developed two sets of guidelines to help the industry prepare for and respond to cyberattacks. The IADC Guidelines for Assessing and Managing Cybersecurity Risks to Drilling Assets was published in 2016, and the IADC Guidelines for Baseline Cybersecurity for Drilling Assets was published in early 2018.
Earlier this year, IADC also elevated the cybersecurity group into a full committee, still led by Secure-NOK’s CTO, Siv Hilde Houmb. The committee is currently working on four additional sets of guidelines: Guidelines for Network Segmentation; Cybersecurity Training; Guidelines for Hardening of Control Systems Focusing on Existing Drilling Assets; and Guidelines for Security Monitoring and Audit.
“What IADC is doing is very important in terms of having a forum where drillers can share their experiences as they discover the cybersecurity landscape,” Mr Singh said. “Information sharing and knowledge sharing is going to be critical for them to find the most appropriate security solution.” DC