Engineering.com| BY Zeeshan Hussain

Major players in oil and gas invest in big data analytics and predictive maintenance solutions in the wake of increasing instability in the industry.

Eni SpA, an Italian oil and gas (O&G) multinational, recently announced a partnership with Google Cloud. BP and Amazon also agreed to further collaborations. Why are supermajors in the O&G industry collaborating with tech companies, and what benefits come out of such enterprises?

There has recently been unprecedented upheaval in the global O&G industry due to various complex challenges, including upstream volatility, midstream constraints, industry consolidation and shifting customer demands based on an increasing swing toward greener alternatives. With an unprecedented drop in demand due to COVID-19 becoming the proverbial straw on the camel’s back, it has become imperative for businesses to go on a cost-cutting drive.

Contrary to popular belief, the oil business is very technologically advanced. Cutting-edge tools are used to map out and drill into complex rock formations thousands of feet below the ground.  For example, Total S.A, the French oil major, and Eni are running two of the most powerful supercomputers in the world. Total’s Pangea has clocked speeds of more than 5 quadrillion calculations a second (yes, that is extremely fast), according to The Top 500 List.

Nevertheless, the industry has fallen behind in terms of digital transformation and leveraging real-time data. In late 2014, MIT Sloane Management Review and Deloitte scored the O&G sector’s ” digital maturity ” at 4.68 out of 10. By the fall of 2019, things got even worse, with O&G scoring a mere 1.3, easily the lowest amongst all the sectors. However, this is changing due to the perfect storm of advancements in technologies, falling costs of digitalization and lower oil prices, which may never reach the high levels previously attained. PwC has reported that O&G executives see the most potential in cloud computing, energy analytics and machine learning.

To accomplish this, big oil has finally joined other industries in the digitalization drive and turned to big tech to provide.

Data Is the New Oil

An enormous amount of data is constantly being generated by the O&G industry:

• Data volume for a single well exceeds 10 TB per day, thanks to optical fibers combined with various sensors being used in wells to record different parameters, such as fluid pressure, temperature and composition.
• A single offshore drilling rig can create over one terabyte (TB) of data per day, especially due to recent innovations in drilling tools, such as logging while drilling (LWD) and measurement while drilling (MWD).
• A large refinery generates one terabyte of data daily.
• Pipeline inspections generate approximately 1.5 TB for every 600 km inspected and ultrasounds around 1.2 TB for eight hours of scanning.
• Seismic surveys collect around 10 TB each.

But what can be done with such oceans of data? This is where big tech is providing expertise, innovation and new technologies for the digitalization of big oil.

A Match Made in Cyberspace

Within the past couple of years, Microsoft has announced collaborations with Chevron, ExxonMobil, Suncor and Petrobras. These partnerships have established Microsoft Azure as their primary cloud provider in order to harness the power of cloud computing, big data and machine learning.

Halliburton is using Microsoft’s technologies, such as machine learning and augmented reality (AR), an example being DecisionSpace365 created on Azure. It enables real-time data streaming from devices in oil fields and the ability to apply deep-learning models for optimized drilling and production, consequently lowering costs for customers. Predictive deep-learning algorithms help optimize field assets and enable exploration and deep-earth models by using software to fill gaps in sensor data while reducing the number of steps and time required for rendering models.

Suncor, a Canadian energy company specializing primarily in oil sands, is applying Microsoft’s technologies in various ways. One example is its use of fully autonomous trucks at its oil sands mines. Employees at some of their facilities wear wireless badges, allowing the company to track and analyze frontline maintenance work with the goal to improve safety and productivity.

Amazon Web Services (AWS) counts BP and Shell amongst its customers. BP in particular has significantly expanded its relationship with AWS by agreeing to supply renewable energy to power Amazon’s operations. In exchange, Amazon will help BP digitize its infrastructure and operations. This includes applications migration from BP’s own European mega data centers to the AWS cloud, in addition to collaborating on different AI and machine-learning initiatives. In this way, BP can reduce energy use and emissions from its own digital infrastructure and data centers. The resulting lower operating costs are the cherry on top of the cake.

Total and Google Cloud signed an agreement to jointly develop AI solutions. These can be applied to subsurface data analysis for O&G exploration and production, most notably from seismic studies (using Computer Vision technology) and to automate the analysis of technical documents (using Natural Language Processing technology).

Eni has also linked up with Google Cloud to construct a new digital platform, Open-es, to support sustainability in the industrial supply chain. It will be open to all players in the energy sector to pool data, best practices and sustainability models. Knowledge-sharing will set in motion an upsurge in safety and efficiency across the industry.

Schlumberger, partnering with Google, is deploying its O&G software suite, including the WesternGeco Omega geophysical data processing platform and Software Integrated Solutions DELFI cognitive E&P environment on Google Cloud Platform (GCP), to perform seismic processing, interpretation and subsurface modeling.

Digital Twins

Graphics processing units (GPUs), due to their enormous processing power, offer benefits through the concept of digital twins. A digital twin is a continuously learning virtual digital copy of all assets, systems and processes. It can predict asset behavior and capacity to deliver on specific outcomes within given parameters and cost constraints.

NVIDIA is assisting Baker Hughes in employing GPU-accelerated computing to work on platforms and build AI-enabled services for major O&G operators across the globe.

Similarly, Shell has chosen Bentley’s iTwin platform, which is an Azure cloud-based platform providing interoperability across supply chain systems. This will allow Shell to manage and analyze data, integrate with existing systems, and increase collaboration across business operations.

Data Can Predict the Future

According to research by Kimberlite, just 3.65 days of unplanned downtime a year can cost O&G companies approximately $5 million. An average offshore O&G company experiences about 27 days of unplanned downtime a year, which can amount to $38 million in losses. In some cases, this number can rise to as much as $88 million.

Driven by the Internet of Things (IoT), operations and maintenance are turning more proactive and predictive as opposed to reactive and planned. Data can be leveraged from sensors (e.g. temperature, vibration, flow rate sensors, etc.) to identify any anomalies in equipment behavior and forecast failure modes within a certain timeframe.

Predictive-maintenance solutions help O&G companies forecast equipment breakdowns before they can have an adverse impact on their safety levels and bottom line. Schneider Electric reports that applying IoT-enabled predictive-maintenance solutions can help save $4 million due to early identification of rotating machinery damage, $500,000 due to early identification of coupling failures and $370,000 due to early identification of heat exchanger valve problems.

It can also improve asset utilization and increase productivity by making operations more flexible and agile. By comparing operational data across multiple pieces of equipment, IoT solutions can help estimate machine utilization, identify the periods of best performance and establish best practices to improve performance across the entire O&G supply chain—from exploration to refining and distribution.

While the O&G sector produces 29 percent of methane emissions, the greenhouse effect of methane is 86 times higher than that of carbon dioxide. In the U.S. alone, the O&G industry releases 1 million tons of methane into the environment every year due to leakages. IoT helps identify and reduce pipeline leaks, thus decreasing environmental damage.

Augmented/Virtual Reality (AR/VR)

O&G companies are also adopting AR/VR for subsurface studies, training and simulation, maintenance and planning. Hard hats equipped with AR can project instructions for the technician directly onto the equipment to conduct an inspection or maintain a system. Instead of being dependent on manuals, AR enables this information to be delivered graphically, thus increasing efficiency and reducing errors.

VR can be used for practical training instead of in a classroom or on location. Trainees use a VR headset to enter an environment or interact with a piece of equipment virtually. During the process, it provides invaluable hands-on training at a fraction of the cost. Likewise, VR apps connected to sensors enable engineers to monitor equipment in real-time without needing to be onsite. Geoscientists are also able to visualize seismic data through VR, and even drill virtually, so that they can better determine where to explore.

Socially Distanced Drilling and Operations

The three largest service providers in this field are Schlumberger, Halliburton and Baker Hughes. Before the pandemic, O&G companies always counted on these specialists to be onsite and control drill bits and interpret real-time data. Now, to adhere to travel restrictions and social-distancing recommendations, drillers had to work from home.

For ongoing operations, during the second quarter of the year, Baker Hughes and Schlumberger both had two-thirds of their drilling activity supported by remote work. For Schlumberger, this was up 25 percent from the first quarter. For Baker Hughes, it was up 20 percent. Halliburton, the largest U.S. fracking company, has reduced the number of onsite engineers by shifting work to real-time operation centers. All of them have cited the adoption of remote work, which allowed them to shut down many operational sites, as the leading cause of significant operating cost reductions.

For drilling new wells, due to travel restrictions, many companies might not have been able to achieve their targets without the adoption of remote technology. Chevron was able to continue directional drilling in the Permian Basin of West Texas and New Mexico by setting up a remote team based mostly out of their homes in Houston because they were able to get real-time data through Azure.

ExxonMobil was able to make its Permian Basin operations totally cloud-based, leading to secure and reliable collection of live data from the oil field assets. This has allowed for quicker and more accurate decision-making on drilling optimization, well completions and personnel deployment.

There is no doubt that the confluence of technologies is beneficial not only for the O&G industry but also the whole world. It will reduce health and safety incidents and lead to better-trained personnel by allowing remote work and simulations unless absolutely necessary. Predictive maintenance will save tons of money by streamlining maintenance plans, as well as an overall reduction in the number of catastrophic failures over time. Goldman Sachs estimates that a 1 percent reduction in the O&G industry’s capex (capital expenses), opex (operating expenses) and inventory management can result in savings of about $140 billion over a 10-year period. Last but not least, all these factors will lead to considerably minimizing adverse effects on the environment, thus protecting our precious lands, oceans and wildlife.

This blogpost is originally from Aquantico | Challenge, innovate & deliver value

Plantservices.com | By Sheila Kennedy, CMRP, contributing editor

Maintenance and reliability best practices are continually improving and so are the technologies that support them. Firmly embedded in the realm of the “best” is predictive maintenance (PdM), which combines real-time monitoring of asset condition, environmental, and/or operational data with smart analytics to detect, assess, and forewarn of impending problems.

Prescriptive maintenance (RxM) takes PdM a step further by prescribing corrective actions for deteriorating conditions and including them in the alerts. RxM is a newer concept just gaining ground that is made possible by machine learning (ML), artificial intelligence (AI), and the internet of things (IoT).

“Preventing dreaded unplanned downtime provides a clear business benefit and driver for adoption of PdM,” said Ralph Rio, vice president at ARC Advisory Group. “Unplanned downtime often cascades into major losses including revenue, WIP materials, and larger equipment repair costs.”

Additionally, the costs for implementation and support of a PdM application have dramatically lowered with the IoT, cloud platforms, microservices for analytics, and wireless sensors, Rio observes. “Clear benefits and lower costs have improved financial justification and driven adoption in a wide range of industries,” he added.

The six case studies summarized below illustrate some of the numerous ways PdM and RxM are transformative.

BASF / Schneider Electric

BASF, the largest chemical company in the world, has digitalization as a corporate strategy. This includes using data to better forecast maintenance requirements and reduce unexpected shutdowns.

One such initiative involved an electrical infrastructure expansion at BASF’s Beaumont, TX, plant. Since production efficiency depends on having predictable electrical power, the plant chose to enable remote monitoring and management of its new power distribution substation’s operations and asset health with EcoStruxure Asset Advisor from Schneider Electric. The data-driven, IIoT-enabled service employs asset sensors for continuous condition monitoring along with predictive analytics to identify threats that could lead to asset failures.

BASF also has access to customized, proactive advice about how to prevent failures and improve its maintenance strategies through its partnership with Schneider Electric’s Connected Services Hub (formerly Service Bureau).

Read “How 7 companies are accelerating PdM and RxM at their plants”
The use case involved having more than 100 condition variables continually collected, measured, and computed for 63 substation assets. The asset data is monitored and analyzed from a digital dashboard that provides 24/7 insight into the substation’s global health index and specific asset statuses, from any location.

“We were surprised by some of the things that we found out. Asset Advisor is helping us to prevent catastrophic failures,” said Lee Perry, electrical design engineer at BASF. The cloud-based service “connects assets so we can look at the health of not just our electrical distribution equipment, but also the motor control centers and the motors that actually drive the process,” he explained.

“If there’s an issue, then we get an email from the Service Bureau saying, ‘We’re noticing this.’ I can log in to it remotely, look at the same information, and help troubleshoot that issue,” added Perry.

With around-the-clock access to data and expert guidance, the plant has the information it needs to make the right decisions at the right time, perform PdM to optimize asset health, and take steps to improve the efficiency of critical electrical distribution assets. These actions consequently benefit plant uptime, performance, productivity, and safety.

ALCOA / Senseye

PdM is a key part of Alcoa’s strategy to become a more stable and profitable organization. To help automate PdM and reduce downtime and maintenance costs, a proof-of-concept (POC) project involving 50 assets across two casthouse systems was conducted at its Fjar∂aál aluminum smelter in Iceland. The solution is now being expanded to at least 1,000 assets in Fjar∂aál and is scalable enterprise wide.

Senseye was chosen to unify and synchronize the equipment sensor data in the plant’s OSIsoft PI ecosystem with maintenance data in its Oracle eAM solution. From the current and historical ingested data, it automatically builds models and starts learning, without having to set up parameters or alarm levels. Isolated peaks in raw data create signatures that reveal trends and hidden failures. Predictive analytics identify what is happening and why, and can provide prognostic insights on the asset’s remaining useful life.

Operators are automatically notified of cases needing attention and can drill down for further details. It is “like having a thousand eyes in the plant” letting you know when data streams start to move away from normal, explained Árni Einarsson, reliability implementation manager at Alcoa in his presentation at OSIsoft PI World San Francisco 2020.

For instance, an idling current increase provided an indication of a fault in the HDC saw motor system. It was determined a belt guard had come loose and was in contact with the sawing drive, damaging the belt. Replacing the belts and fastening the belt cover during a maintenance shutdown resolved the issue, avoiding 12 hours of unplanned downtime.

A sensor failure detected in a coiler rod cropping shear motor prompted the discovery that a lower pinch roller sensor had come loose, leading to a sharp increase in shear motor torque. Re-fastening the sensor returned the torque to normal levels, avoiding three hours of unplanned downtime.

In total, Alcoa reduced unplanned downtime of the machinery by up to 20 percent and achieved full ROI in 4-6 months. “With the POC completed, we were able to scale the solution to other parts of the business,” said Einarsson. “The focus is on finding the best business cases to continue.”

Duke Energy Renewables / Seeq

Duke Energy Renewables, an owner/operator of wind and solar farms across the U.S., used advanced analytics and ML to automate profiling and detection of failing contactors on one of its wind turbines. “This was a trial run with one use case that was interesting, the data was available, and it seemed like a good candidate. We wanted to get up and running fast and to see if this even works for us,” said Abhi Hullatti, manager of performance analytics at Duke Energy Renewables, in his presentation at the ARC Industry Forum Orlando 2020.

Hullatti described how each turbine has six contactors that help to ensure that when the generator kicks in, it ramps up smoothly and synchronizes with the electrical grid. When any of the six contactors fail, the turbine goes offline for 2-10 days for diagnosis and repair. Turbine contactors at one site tended to fail more frequently than at others. Automating the prediction of impending failure would enable PdM, improve uptime, reduce maintenance costs, and allow better management of spare parts inventory.

Using an automated profiling tool from Seeq, a model was trained to look for the contactor fault error code and plausible leading-indicator signals (reactive power, active power, current, wind speed, rotor speed, and generator speed), to recognize what normal behavior looks like, and to provide one-hour advance notification of a fault.

The model ran against 2.5 years of cleansed signal data for the turbine and it found 12 occurrences of the error code, mostly within the few months preceding a failure. There were no false positives leading up to that outage, and no further error codes or predictions followed once all the contactors were replaced. The model can run continuously once it is validated and it will give automatic notification when something is going wrong, said Hullatti.

“The results are extremely promising, so we want to next expand this model to the rest of our turbines at that site, and also, more interestingly, look at other failure modes. That will be a lot more complex but also far more financially rewarding,” observed Hullatti. The return on the ability to catch a generator that is about to fail, and act proactively instead of reactively, is in the hundreds of thousands of dollars per event, he explained.

Global Mining Company / Uptake

A leading global mining company needed to predict and prevent failures that can cause unplanned downtime in its material transport operations. The company owns and operates a network of private railway tracks, rolling stock, and signals that move iron ore from its mining sites to the seaport. Any faults, delays, or breakdowns in the process of moving the high-value material is not only costly from an equipment maintenance and operational performance perspective, but it can also be potentially catastrophic.

The trains and rail infrastructure already produce immense amounts of data from wayside telematics devices. If harnessed, the Wheel Impact Load Detection (WILD), Wheel Condition Monitoring (WCM), Hot-Box Temperature Detection (HBD), and Bearing Acoustic Measurement (BAM) data could provide operational and condition insights into anomalies with sufficient time to take corrective actions.

The mining company chose Uptake, an industrial AI software company, to provide the necessary visibility into degrading asset conditions across the network and advanced analytics to prescribe recommended solutions. The software is now analyzing the four telematics data sources for indications of faulty wheels, wheel bearings, and axles on 10,000 rail cars.

Events are ranked according to criticality. Each high-severity wheel or bearing downtime event for a rail car costs the company at least $6,000 and requires immediate attention to return it to operation. Previously unforeseen, the mining company now has roughly a weeks’ advance notice on high-severity wheel alerts, 11 days’ notice of high-severity bearing alerts, two weeks on medium-severity alerts, and a month on opportunistic events.

With the advanced notice of issues and prescriptive alerts, the company can proactively bundle its maintenance activity, significantly reducing downtime and avoiding rail car failure. As a result, it is able to reduce its single-car unscheduled maintenance events by an estimated 50 percent, from 1,850+ per year to 925+ per year, which represents approximately $34 million in savings over five years.

Cement Plant / Aspentech

A cement plant that had struggled with cyclone blockage in its cement kilns was one of several PdM case studies shared by Chris Williams, global director of asset performance management (APM) services for AspenTech, at his ARC Industry Forum Orlando 2020 workshop. Frequent blockages in filters and cyclones limit production and increase maintenance costs, he explained. Subtle changes in the preheater process can eventually lead to blockage.

To reduce downtime and optimize maintenance, the plant sought a solution to predict and avoid cyclone blockages through early and accurate detection. Having the added benefit of RxM capabilities to resolve emerging problems quicker would minimize production losses.

The cement plant chose a PdM and RxM solution with AI and ML capabilities from AspenTech. Maestro for Aspen Mtell was used to build enhanced agents capable of predicting cyclone blockages. The failure agents trained on past blockages to identify the failure signatures that precede degradation, breakdowns, and process disruptions, and learned to recognize the pattern.

The multivariate (MV) models helped the plant to better understand the root cause of the condition and provided prescriptive guidance on feed composition and kiln operating conditions to the operators, enabling them to quickly bring the process back to normal operation. Within weeks, the solution was able to provide several days of early warning. “Eliminating just 25 percent of blockages generated over $1 million in savings,” said Williams.

Wärtsilä / Pega

Wärtsilä, a Finnish technology and service provider for energy and marine markets around the world, needed to improve its ability to process and analyze asset condition data received daily from thousands of equipment installations. An IoT engine upgrade was needed to help preempt performance issues with its customers’ power plants and vessels, and to automate feedback to each customer. Its legacy system could not sufficiently scale to handle the desired automation.

Many of Wärtsilä’s installations have multiple engines per site, and each engine has attached equipment and hundreds of sensors generating condition-based data. To better harness the global data, the company needed to map it to a normalized data structure for quick processing and apply a rules framework to assess the sensor data for anomalies. This would enable PdM practices to preempt equipment failures and also help to optimize service support.

Read “RxM: What is prescriptive maintenance, and how soon will you need it?”
The company worked previously with Pega on other projects and chose the Pega Platform to extend its existing IoT capabilities. The digital transformation platform for Digital Prescriptive Maintenance is designed to help optimize the cost and duty-cycle of devices and systems.

With the new IoT-powered solution, Wärtsilä now compiles all incoming sensor data in a data lake in the cloud, where it is normalized and fed into the Pega Platform. Using a complex rules framework, the data is processed to identify exceptions that may indicate a developing risk of engine failure or a reduction in equipment performance. Daily feedback reports document the findings and prescribe recommended actions, enabling the company to save time and money by avoiding unplanned downtime and production problems.

Benefits to the customers include improved environmental and economic performance and long-term predictability of their vessels or power plants. Wärtsilä is also benefitting from optimized equipment maintenance, improved reliability and performance, and increased sales of proactive maintenance contracts.

Each of these industry leaders is enjoying the rewards of using maintenance and reliability best practices. With the benefits of PdM and RxM being realized, the old ways of doing business seem antiquated and companies like these are not looking back.

This blogpost is originally from Aquantico | Challenge, innovate & deliver value

Oil & Gas Engineering | BY DR. HIROAKI KANOKOGI

When other methods fail, Artificial intelligence is a critical finding and solving tool

Oil refineries and petrochemical plants are meant to run continuously, usually for years at a time, to maintain critical production obligations and achieve financial goals. Consequently, problems disruptive enough to interrupt production or impair product quality are serious and to be avoided. Disruptions can be related to equipment failure, such as a motor burning out, or stem from a process problem beyond operators’ ability to correct.

Plant managers look for ways to determine when a problem is on the horizon, before it escalates into an unscheduled outage. For equipment, trouble brewing can be spotted with basic diagnostic sensors, such as bearing noise monitors, which can warn the maintenance department when a critical parameter is changing.

Process problems are often more subtle but can be just as disruptive, as we’ll discuss in two examples later in the article. In these situations, some process variable begins to move into dangerous territory and can’t be detected by the automation system, leaving operators to determine corrective action.

The question emerges, was there anything in the production data prior to the incident that pointed to the situation developing? Operators and process engineers may pore over data for hours looking for clues, only to throw up their hands. This challenge spurred a search for new technologies to provide a solution.

Help from AI

Artificial intelligence (AI) is the topic of many discussions these days with many suggestions as to how it will improve manufacturing in the future. A more practical approach is exploring what AI is doing right now to solve problems, as described below.

Human operators rely on the automation system to show them the overall state of the process. This should help them identify when there are signs of abnormalities developing and the cause. In reality, the automation system shows operators a wide range of variables, leaving them to make their best interpretation. If a problem occurs, operators must respond quickly and correctly to forestall an incident. But what they really want is a system capable of indicating the state of the process and identifying which sensor to focus on when an error occurs.

AI can identify the main factors contributing to abnormal situations and point out the specific sensors indicating the causes. This allows operators to concentrate on a small number of manageable elements to solve the problem, rather than trying to deal with a much large number of sensors, most of which do not relate to the immediate problem.

This process begins with operators identifying a specific problem. Data scientists work with domain experts — the process engineers — to build a system incorporating a learning model. This combination of artificial and human intelligence (AI+HI) works with an observe-orient-decide-act (OODA) loop as an effective decision-making procedure:

1. Observe — intelligent sensing
2. Orient — advanced analytics
3. Decide — real-time and astute decision making
4. Act — agile actions for value creation.

Observe includes identifying the problem and setting the goal, defined as the process state in which the problem will be solved. This calls for narrowing down the process data and maintenance information necessary for analysis, and then translating the problem into specific tasks to set issues.

Orient determines which direction the analysis proceeds to solve the defined tasks. It combines AI technology, domain knowledge, and data scientists’ expertise to dig into the data for analysis.

Decide examines what actions are suggested by the analysis results. If directions seem to be going off on a tangent, plant personnel go back to the first step and see if the problem is defined correctly. The participants must agree on whether or not to implement a given plan.

Act puts the consensus plan into effect. This may involve adding capabilities such as edge computers, cloud computing, and data storage. If it proves necessary to redefine the tasks, plant personnel must return to the first step and reconsider the overall approach.

Applying AI

An ethylene producer contacted Yokogawa and asked for assistance in solving a list of reoccurring process-related problems. Working with the company’s internal problem-solving team, all the participants engaged in a workshop to become familiar with the methodology and understand how the project would proceed.

Together, the team identified possible causes for each problem from hundreds of sensor parameters. The parameters were used to monitor the operational state of the equipment and create an AI model to detect anomaly in the equipment and understand the plant status. This methodology was applied to eight projects, we will look at two.

Case 1: Benzene Production Reactor

An ethylene plant produces cracked gasoline, which is converted to benzene by adding hydrogen in a reactor with a catalyst. This also removes impurities. To maintain a stable reaction, it is necessary to modulate hydrogen flow and reaction temperature to match raw materials.

The catalyst deteriorates gradually, resulting in lower catalytic performance. However, since there is no method to quantify activity, catalysts are periodically activated or replaced following a schedule based on run time or a calendar. This means that some catalysts are replaced even though they are still serviceable, creating extra maintenance.

Conversely, unexpectedly rapid deterioration of the catalyst ahead of replacement time causes an increase in impurity levels, resulting in defective and unsellable product. This creates several major problems:

• Reduced production with missed targets
• Loss of raw materials
• Increased cost of disposal
• Shutdown for emergency maintenance.

The observe phase concluded the operators needed KPIs for catalyst activity because without it they couldn’t tell when conditions called for a catalyst change prior to costly production problems. This knowledge would optimize maintenance and avoid product losses.

The orient phase examined production data for the previous two years, during which time there were three catalyst changes: two following the normal schedule, and one emergency replacement due to production problems.

Process data during stable operation and the time just before the emergency maintenance became training data and was analyzed by AI to create a training model, which was then applied back to the data. This resulted in a catalyst health index, formed from a synthesis of multiple measured process variables identified by the AI analysis, which became the very KPI the operators needed.

When the two years of data was examined using this index, it became clear that catalyst health could be determined, and it was possible to decide when it needed to be replaced prior to product degradation.

The decide and act phases became clear. Operators now follow the catalyst health index in real time and schedule catalyst changes based on condition. This results in maximum catalyst life while avoiding product degradation and emergency shutdowns.

Case 2: Cracking Furnace Cooling Tower

In ethylene plants, ethane, naphtha, and other raw materials are heated in a cracking furnace. To prevent excessive cracking, the hot gas moves to a cooling tower, where cold water is sprayed into the stream to reduce the temperature to <35 °C. The heated water is sent to various heat exchangers in the plant and eventually recirculated.

During the summer, the cooling tower seemed to lose capacity, making it difficult for operators to control the process. The cracking reaction continued and impeded separation of the desired components, causing poorer product quality and yield. For several years, this situation was reluctantly tolerated as an unavoidable seasonal effect, at least until the problem unexpectedly vanished during the summer of 2019. There were no obvious climatic reasons, so the engineering team wanted to find out what had changed, and how to keep the problem from returning.

The observe phase set the two-part goal: from plant data, establish an indicator of an operating condition causing the loss of cooling effectiveness and identify a parameter closely related to this temperature rise. By changing the parameter, operators could improve operation.

The orient phase examined data from the previous two years, 2018 and 2019 (Figure 3a), when the temperature increased and stayed flat, respectively. Whatever happened in 2019 solved the problem, but no one could positively identify the specific change.

The AI analysis used to build the training model suggested several candidate parameters, including cooling tower temperature and cooling water flow rate. The analysis found these parameters affected the temperature and flow rate of the heat exchanger adjacent to the cooling tower, and both were closely related to the temperature in the cooling tower.

The training model created an index capable of predicting the effectiveness of the cooling tower. This was a synthesis of process variables, both upstream and downstream from the cooling tower itself. The higher the index, the less likely there would be conditions capable of causing loss of cooling capacity and product problems. The pleasant surprise of improved temperature control experienced in the summer of 2019 became reproducible at will.

Understanding what AI is and what it can do is difficult to pinpoint because it takes so many forms. Within process manufacturing, it can help solve many types of problems because it is a methodology as much as a technology. It requires engagement between HI and AI as illustrated in these examples. AI becomes the tool to extend the capabilities of HI once the human domain experts define the root problems to be solved.

AI going forward

Today, AI in process manufacturing is limited to analysis, rather than a primary method of real-time process control. However, this is changing. Yokogawa and others have been involved in experiments to replace traditional PID-loop-based control strategies with a single comprehensive AI system able to learn how to optimize control of a single process unit or an entire refinery. After all, if AI techniques can control a self-driving car, why not an ethylene plant?

The answer is, it can and will. Yokogawa has introduced machine learning as a practical technology for process manufacturing, and leveraging the results, is accumulating application cases in laboratory settings and actual facilities. This requires developing technology to ensure safety and versatility, with key abilities, including:

• Safe learning methods
• Robust ability to handle disturbances
• Fast response to changes in set points
• Continuous online learning
• Applicability and transferability of models to multiple facilities.

The goal is an AI-based control system able to deal with challenges quickly and effectively to improve process performance and plant operations.

This blogpost is originally from Aquantico | Challenge, innovate & deliver value

Plant Services| By Sheila Kennedy

Amid a lot of industry noise, there’s no better way to get a sense of the value of modern AI driven maintenance approaches – and the imperatives associated with implementing these – than to hear stories directly from practitioners. In every industry and across the world, companies driven to improve reliability at their facilities are testing and investing in new ways to optimize their assets. Their efforts are paying off in greater uptime, efficiency, safety, and cost savings.

Run-to-failure is increasingly reserved for rare and unique circumstances. This trend started when increasingly capable condition inspection and monitoring tools shifted the asset management focus from “fix what’s broken” to “keep it from breaking down.” Today, unprecedented opportunities afforded by the industrial internet of things (IIoT) have further changed the playing field, and there are potential benefits yet to be realized.

For example, predictive maintenance (PdM), originally based on selected asset condition data, has grown to accommodate online, real-time streams of multiple types of condition data received via sensors and even drones. Some companies are applying machine learning (ML) to further refine their predictive analytics and prognostics.

The newest opportunity, prescriptive maintenance (RxM), is a multivariate approach that merges asset condition data with any combination of operating, environmental, process safety, engineering, supplier, or other related data to better diagnose conditions and prescribe specific options for corrective action. The advanced analytics, pattern recognition, modeling, ML, and artificial intelligence (AI) that empower RxM may help companies finally greatly curtail, if not eliminate, the need for reactive maintenance on critical equipment.

We asked experts at seven manufacturing companies to describe their PdM and RxM progress, including successes that are helping keep their programs alive. Here are their stories, followed by some examples of enabling solutions.

• Machine learning for PdM: Covestro
• Convincing POC: Borealis
• Zero-failure objective: Mercer Celgar
• Greenfield opportunity: Saudi Aramco
• Growth management: AGL Energy
• Engineering advantage: Church & Dwight Co.
• New business model: Kaeser Compressors

Varied AI & PdM approaches to a common goal

Numerous software and technology solutions work together to enable PdM and RxM, though some companies won’t name specific choices. A sampling of products involved in some of these stories includes, in alphabetical order: Aspen Mtell from AspenTech, IMx and WMx from SKF, Leonardo IoT from SAP, ORM Digital Twin from Sphera, PI System from OSIsoft, Predict-It from Engineering Consultants Group, and PRiSMfrom AVEVA.

Human nature is another consideration. Despite the advancements enabled by IIoT technologies in the ability to practice PdM and RxM, indications are that workplace culture continues to be a tug-of-war between operations and maintenance (O&M), cautions Paula Hollywood, senior analyst at ARC Advisory Group.

While the functions and objectives of O&M are different, they share common goals, and an asset performance management (APM) approach helps to synchronize the efforts. Hollywood explains: “APM is not just about maintenance; it is evolving to become more about the two groups collaborating along with engineering to systematically utilize assets to maximize profits and reduce risk factors.”

Nevertheless, as the manufacturers above have demonstrated, it is easy to become committed to the practice after taking that first step.

This blogpost is originally from Aquantico | Challenge, innovate & deliver value

Hart Energy| By Philippe Herve, SparkCognition

In an industry like oil and gas, unexpected failures can be catastrophic for many reasons. The cost of maintenance or new parts is often exorbitant, and with how much capital is committed to wells by a company per day, the lost opportunity cost from unplanned downtime can be shocking.

Despite this, the industry has never had particularly great methods of predicting and preventing unexpected failures before they can cause problems. Too many operators still take either a reactive or scheduled approach to maintenance, both of which have tremendous downsides for reliable production. Reactive maintenance runs the risk of creating safety hazards or large operational damage, while scheduled maintenance often halts operations unnecessarily. Enter predictive maintenance, which shows great promise as a new paradigm for monitoring operations but has been a challenge to implement efficiently and effectively. New technology is needed to enable oil and gas operators to truly enjoy the benefits of predictive maintenance and keep their rigs running safely and continuously.

That new technology is none other than automated model building (AMB), which aims to deliver artificial intelligence (AI) to the fingertips of industrial companies, hence ensuring the success of predictive maintenance programs without the need for large data science organizations. As an AI company working in the oil and gas space, SparkCognition already has helped oil and gas operators apply AMB to their operations with great success.

Current state of predictive maintenance

Predictive maintenance, which involves analyzing patterns in sensor data to forecast impending asset failures before they occur, is an invaluable tool for oil and gas operators. It allows companies to optimize their maintenance strategies, performing the needed maintenance on the asset while it is not in use or bringing in a replacement beforehand if the failure cannot be averted. This greatly reduces unscheduled downtime while eliminating many of the uncertainties from the drilling process.

According to research by the Electric Power Research Institute, the annual cost of scheduled maintenance is $24/ hp. Reactive maintenance costs $17/hp per year—before considering the additional costs of the safety hazards or operational damage that may be incurred by asset failure. Predictive maintenance was found to be by far the most cost-effective method, costing only $9/hp annually and including no hidden costs or dangers.

The quantitative benefits of predictive analytics also have been well documented within the oil and gas sector. McKinsey & Co. found that advanced analytics can yield as much as 30 to 50 times the initial investment within the first few months of implementation.

In spite of this, the oil and gas industry has been slow to adopt predictive maintenance, as it comes with some barriers to entry. Predictive maintenance requires data analytics, but data science talent can be scarce. Even for operations that do have an in-house data science team, the exorbitant amount of time predictive maintenance requires from it in collecting and analyzing new sensor data is neither feasible nor scalable.

Predictive maintenance also relies on the use of models based on prior data to predict asset health and operating states. Traditionally, these models were built on predefined rules worked out by human analysts. Such models can take enormous amounts of time and human talent to build, perform poorly in extreme or unusual conditions and are subject to human bias. Additionally, they are inflexible; if a single variable of the asset changes, the model is rendered useless.

AMB is the answer

AMB opens up the full potential of predictive maintenance to the oil and gas industry. AMB platforms ingest operational, sensor and historical data, including production data (oil, gas, water, etc.), sensors from compressors, pumps and vapor recovery units. Then they automatically build machine learning models that predict the operating state and remaining useful life of a given asset. This reduces the time required for the process of model creation by orders of magnitude, bringing it down to weeks or even days.

AMB also amplifies a data scientist’s work by enabling more effective model building. It gives data scientists the ability to iterate through thousands upon thousands of different models and architectures rapidly. In doing so, AMB allows data scientists and nondata scientist model builders alike to focus on their data problems rather than parameter tuning within models.

As AMB platforms leverage machine learning, models produced by AMB are continuously learning and refining themselves, allowing them to adapt to changing conditions in assets and catch the never-before- seen failures or edge cases that traditional models struggle with.

Case study

SparkCognition proved out the utility of AMB in predictive maintenance for oil and gas for a major E&P operator looking to leverage the power of AI to predict workover needs and production in oil fields.

As oil and gas maintenance operations go, workovers do not come cheap. They require large crews, heavy equipment and the complete shutdown of production. For a fairly normal well that produces an average of 50 bbl/d priced at $60/bbl, ceasing production translates to an average of $3,000 lost income per well for each day it’s not running. Add to that the costs of a workover rig with support equipment and crew at about $10,000 per day and it’s a hefty price tag. All told, workover operations can cost roughly $70,000 to $100,000 for a single well.

Unfortunately, workovers are a necessary evil of oil and gas production. On average, 10% of wells in a field require a full workover each year. Savvy operators are always on the lookout for ways to reduce this number while maximizing returns, and this operator decided to turn to AMB as a possible solution. Specifically, the operator wanted an automated, reliable solution to monitor the state of its wells, assist process engineers in identifying which wells needed maintenance (and what kind of maintenance), and predictive maintenance needs enough in advance to allow the operator to minimize downtime. It also wanted a solution that would help process engineers prioritize maintenance operations by assessing the potential return of a given well after a workover event.

Working together with SparkCognition’s AMB Darwin platform, the operator rapidly created cognitive models to predict well maintenance needs and future well production accurately. These models were created using a limited dataset with only one input per month. This set included data on well static properties, including location, reservoir characteristics, and monthly oil, water and gas production. Darwin ingested these data, extracted the 40 most relevant features and built optimized models to identify seven types of maintenance events for wells in the field.

The models Darwin built predicted workover, rod change and cleaning operation needs in 12 out of 17 wells in the field with accuracy between 70% and 80%. These predictions were made three to six months in advance, allowing the operator to plan maintenance operations. In doing so, the operator was able to improve safety and reduce its operating costs by replacing expensive emergency repair with informed, as-needed scheduled maintenance events.

The models also provided valuable insights for process engineers on well production potential. The engineers were able to maximize their returns by focusing efforts on the most promising wells.

This blogpost is originally from Aquantico | Challenge, innovate & deliver value