New IIoT- and cloud-enabled digital tools and services give petroleum refiners new avenues to increase profitability and safety, but also require greater attention to cybersecurity
While several forces are creating conditions in which U.S. petroleum refiners can thrive in 2017 and beyond, success and profitability are not guaranteed (see U.S. refining outlook section below). Refiners must address changing supply and demand for individual refined products, fluctuations in crude oil prices and dynamic geopolitical factors, all while pursuing the industry’s ever-present imperative for efficient and safe operations. And refinery operations are taking place in an environment where the retirement of experienced workers is ongoing and the industry infrastructure is aging. The sum of these forces makes for a challenging environment for the nation’s 139 active petroleum refineries.
To strengthen their chances of success, refiners are increasingly exploring digital tools that take advantage of the emerging Industrial Internet of Things (IIoT), as well as advanced software for data analysis that can optimize process operations and reduce downtime. A host of new offerings are becoming available, and several were discussed at the annual meeting of the American Fuel and Petrochemical Manufacturers (AFPM; Washington, D.C.; www.afpm.org), which took place in San Antonio, Tex. in late March.
U.S. Petroleum refining outlook
U.S. Dept. of Energy data indicate that U.S. refining capacity grew by almost 2% in 2016, and Chet Thompson, president of the American Fuel and Petrochemical Manufacturers (AFPM; Washington, D.C.; www.afpm.org), called 2016 a good — but not great — year for the petroleum refining industry, with exports of refined products robust at 3.3 million barrels per day (bbl/d) and demand projections remaining strong.
AFPM leaders commented on the positive outlook for their industry that has been set up with the new Trump Administration in the U.S. Greg Goff, who chairs the AFPM board of directors says, “We are at a tipping point for opportunity now, and we as an industry have to rise up with a sense of duty and be leaders. We have a business-friendly White House, we have executive branch agencies that are not hostile to our industry, and we have a center-right Congress.”
Aside from the “seismic political shift” that led to Trump becoming President and the Republican Congress, Thompson pointed out that AFPM members have what he clearly sees as allies in Rick Perry (former Texas governor and new Secretary of Energy), Rex Tillerson (former ExxonMobil executive and new Secretary of State), and Scott Pruitt (former Oklahoma attorney general and new EPA administrator).
While no mention was made of addressing climate change concerns, Thompson and Goff both accepted the presence of Federal regulations for the industry. “Governmental regulations are not necessarily bad, but we need transparency in how they are made and how they are justified and how they are written,” Thompson says. “We’re not opposed to regulations as long as they are reasonable and cost-effective.
Thompson called on Congress and the Administration to bring about corporate tax reform and repeal the Renewable Fuel Standard (RFS), as well as expressed support for the White House America First Energy Plan (www.whitehouse.gov/america-first-energy). The document, which appeared shortly after Trump’s inauguration has been criticized for failing to mention renewable energy, climate change and investment in utility grid infrastructure.
The historical approach to refinery operation has largely been characterized by a “run to fail” mentality, where abnormal conditions and malfunctions were detected only when alarms arose or when a component broke or failed. The IIoT enables operators, engineers and plant managers to capture and analyze data so they can predictively identify potential issues before problems arise. A plant enabled by IIoT is equipped with a combination of sensors, automation systems and cloud-based technologies that are integrated with its current systems and data analytics capabilities. Streaming data from sensors and instruments allow plants to quickly assess current conditions and identify warning signs for abnormal operations. Beyond that, digital tools that enable plants to access the benefits of the IIoT and cloud computing are becoming instruments for boosting profitability.
The recent proliferation of sensors and software, combined with advanced analytics capabilities, has allowed plants to move to a predictive-maintenance system, says Paul Bjacek, the chemicals and natural resources research lead for business consulting firm Accenture (www.accenture.com). “But we’ve also seen what we call a ‘digital decoupling’ in the chemical process industries (CPI) and elsewhere, in which digital technology, including IIoT tools, is becoming a primary driver of value that goes beyond being a system to improve conventional processes,” Bjacek says.
According to proponents of IIoT-enabled digital systems and advanced analytics, the new tools can allow improved decision-making by aggregating data from multiple sources — cost-effectively generating data not available previously. It can then allow pattern recognition and analytics to guide actions based on that wealth of data. Benefits of such IIoT-enabled tools are said to include the following:
- Increasing the rate of asset utilization by reducing unplanned downtime
- Minimizing small efficiency losses from sources that may not have been detectable previously
- Raising operating efficiency through improved monitoring of energy usage
- Improving operations by continuous monitoring and by providing instant access to information that supports decision-making
- Maintaining the effectiveness of control loops, controllers and models over time, so the benefits of advanced process control are sustained
- Lowering overall process risk, thus improving safety
- Reducing maintenance costs
Focus on economics
In order to realize these benefits, though, refineries need ways to transform all of the captured data into information within a real-world, operational context. A host of companies have been developing systems for providing tangible value for IIoT-related data collection and analysis.
Martin Turk, a global solution architect for industrial clients at Schneider Electric SE (Rueil-Malmaison, France; www.schneider-electric.com) says, “There is a need to begin with the problems that need to be fixed and to ask how these new [IIoT-related] technologies can help solve them, instead of starting with the tools and trying to find what problems they could address. At Schneider, we’re taking a value-focused approach to IIoT, where the objective is to leverage the IIoT to make petroleum refiners more profitable,” he says.
In February 2017, Schneider introduced patented software known as Profit Advisor (Figure 1), which uses data analytics to measure financial performance of industrial operations in realtime. Profit Advisor works with process data historians to mine both past and realtime operating data, and then crunches those data through proprietary segment-specific accounting algorithms, the company says, to determine realtime operational profitability and potential savings.
Developed in collaboration with Seeq Corp. (Seattle, Wash.; www.seeq.com), Schneider’s Profit Advisor helps make economic-based decisions, in part by using continuous comparisons between designed performance and actual performance, Turk says. “It allows us to predict the impact of operator decisions on plant economics, making each operator more like a proprietor,” he says.
Profit Advisor measures the realtime profit performance of each major plant asset and unit operation, and the whole plant, so it is a departure from current cost-accounting systems that only measure financial performance of the overall plant, Schneider Electric says. The product is designed to allow individual plant personnel to “see and understand the return-on-investment and business value of their actions… in realtime,” the company adds, empowering them to make better decisions about operational profitability.
The system can also make it easier for workers to focus efforts on activities likely to provide the greatest financial returns and allows them to predict the profitability of possible changes before they are made, which can minimize risk and eliminate waste, Turk explains. For example, assessing the cost of a given period of downtime to fix a component could be compared to the costs of continuing to run a piece of equipment in a slightly degraded or suboptimal state for a certain period, Turk says.
Schneider Electric’s Profit Advisor exists within a larger system of digital tools that includes Avantis PRiSM (process information signal monitoring), a predictive asset-analytics solution that can provide early notification of equipment health issues days, weeks or months before failure, and ARPM (automated rigorous performance monitoring). ARPM is a model-based online application designed to provide operators and engineers with realtime information about the performance of plant assets (for example, compressor efficiency) so that they can make better and faster decisions regarding what to do to correct for deviations from expected behavior.
PRiSM was originally designed for rotating equipment in other sectors, allowing operators to detect deviations and examine likely causes of problems, Turk explains, but his company is now moving this tool into refineries and adapting it to handle other equipment classes, such as heat exchangers and reactors.
With all of the IIoT-related technology available, it has become relatively easy to collect data, but using those data thoughtfully to really make smart decisions about what to do with those data is what we are focused on, says Don Empie, communications director at Honeywell Process Solutions (HPS; Houston; www.honeywellprocess.com).
Honeywell is in the early stages of implementing its HPS Connected Plant initiative, which uses IIOT-enabled data collection and predictive analytics to enhance profitability across multiple facility sites, Empie says (Figure 2). To support the effort, HPS has created what it called “an ecosystem of OEMs [original equipment manufacturers],” each of which brings deep and specific expertise in different equipment classes. HPS Connected Plant is designed to harness the IIoT to tap into the deep knowledge of Honeywell and its network of suppliers and partners, Empie says, and by doing so, end-users are better able to make use of data enabled by IIoT systems.
A key part of HPS Connected Plant — and an example of how HPS is taking advantage of existing expertise — is Honeywell subsidiary UOP LLC’s (Des Plaines, Ill.; www.uop.com) Connected Performance Service (CPS) offerings, which were launched in autumn of 2016. UOP’s cloud-based software services continuously monitor streaming plant data and apply advanced analytics and machine learning to identify latent or emerging underperformance, alert plant personnel and make specific operational recommendations. The objectives include reducing unplanned downtime, increasing safety, raising efficiency and improving supply chain management.
Leveraging UOP process models and best practices, the CPS services create a “digital twin” of a plant that operates virtually in the cloud. “This ‘digital twin’ is kind of a utopia plant operating in the cloud that allows realtime comparisons between actual and simulated plant performance,” explains Zak Alzein, Honeywell UOP vice president and general manager for CPS.
“We are offering a holistic approach to optimizing asset capabilities and maximizing uptime,” Alzein adds, “by bringing together rigorous knowledge of process technology with new software tools.” These IIoT-enabled tools take into account equipment inputs and feed properties and link them via cloud computing to maintain performance over time and provide a platform for continuous innovation and improvement, he says.
Two critical strengths for CPS services are their machine-learning algorithms and the open partnership between HPS and UOP. Since each petroleum refinery is unique, broad process technology experience is important. “Our fundamental knowledge of the chemistry is married to the data analytics and the machine learning,” Alzein says. The UOP vice president thinks the technology world has reached an inflection point in machine learning, where these types of algorithms are found in many places, including in ordinary web browsing and smartphone applications. “Machine learning can eventually create almost a ‘self-healing’ plant that can use the IIoT to quickly introduce software updates and security patches, and proactively manage its own maintenance, for example,” Alzein says.
Thus far, Honeywell UOP has announced three plants in which the services will be used, with more announcements forthcoming. The facilities announced to date are the Binh Son Refining and Petrochemical Co. Ltd. complex in Quang Ngai, Vietnam, the Delek Refining Inc. refinery in Tyler, Tex., and the Al Waha Petrochemicals Co. facility in Jubail, Saudi Arabia.
“Adoption is slow in this industry, but the plants are recognizing the potential benefits of these tools and these approaches,” remarks Alzein.
Remote process support
The March AFPM meeting also saw the launch of the KBC Co-Pilot Program, which is a service using simulation technology with IIoT and cloud computing tools to access the expertise of strategic and technical consultants at KBC Advanced Technologies (Walton-on-Thames, U.K.; www.kbcat.com). In Q3 2016, KBC became a wholly owned subsidiary of automation company Yokogawa Electric Corp. (Tokyo, Japan; www.yokogawa.com).
The Co-Pilot program is the initial manifestation of the KBC Production Core, which envisions automation of all aspects of production operations, with integrated technology and consulting best practices that leverage cloud computing and the IIoT.
The first release under the program is a Refinery Unit Performance Co-Pilot, says Jason Durst, Co-Pilot Program Manager at KBC, and is focused on driving value for clients by providing them with the tools and expertise to collaboratively maximize the potential from oil-refinery process units. Future releases will add Co-Pilot solutions for other asset types.
The Refinery Unit Performance Co-Pilot service monitors process operations at a facility in realtime to remotely support the plant with expertise and insight that supplements the plant’s own capabilities and resources. It tracks data from multiple sources, including actual operating units and simulation programs, says Durst, and through Web-based dashboards, allows both clients and KBC subject matter experts to analyze the raw data and standardized unit performance indicators to make decisions to increase unit performance.
Co-Pilot is focused on bringing value to the client, Durst adds, and it is suited to process operations where the following may be true: managers are not confident that their operating plan is always realistic and achievable; an inexperienced workforce means the unit operation often misses plan; engineers lack the tools and knowledge to maximize profit or reduce risk; or operators do not always automatically know when they are deviating from plan.
Co-Pilot assures asset operators that their simulation and planning tools are up-to-date through the cloud, and that any adjustments made or recommended by their engineers result in optimal process performance and safe operation of equipment within recognized limits, KBC says.
The proliferation of internet-connected devices and sensors associated with IIoT technologies, coupled with increased use of cloud computing and data-as-a-service models, has further raised cybersecurity concerns for industrial control systems. Attention on the topic continues to grow and AFPM meeting organizers included a session about cybersecurity and automation systems. Among the themes explored by speakers was the need to merge the operational technology (OT) sector of the business with the information technology (IT) area, the traditional “home” to cybersecurity countermeasures (for more information, see Chem. Eng., June 2014, pp. 30–35).
Eddie Habibi, founder and CEO of PAS Inc. (Houston; www.pas.com) spoke about the need for petroleum refineries and other CPI companies to undertake a comprehensive inventory of what he calls “cyber-assets,” which includes all control-system sensors, input-output devices, computer workstations, mobile devices, and others. “You can’t secure it if you don’t know it exists,” Habibi says. “If you have a complete inventory of cyber assets, you can identify vulnerabilities and determine if unauthorized changes have occurred,” he notes (for more information, see Chem. Eng., October 2016, pp. 60–64).
The traditional “IT-centric” view of cyber endpoints for industrial control systems neglects many parts of the distributed control systems (DCS) and programmable logic controllers (PLC) that exist below the level of information networks. That portion, which Habibi termed the “production-centric” cyber endpoints, consists of 80% of the assets that require inventorying, he says.
Aside from the comprehensive cyber-asset inventory, Habibi also recommended that companies conduct a prioritization exercise for the costs and consequences of various types of cyberattacks, or other incidents in which cybersecurity may be at risk unintentionally or non-maliciously. This can better ensure that resources are devoted to cybersecurity in a thoughtful way. In addition, attention should be paid to how cyber assets are backed up, and how recovery from a cyberattack would be accomplished.
According to Habibi and others at the meeting, the cyberattack “threat landscape” is growing, but so is the recognition of cybersecurity’s importance at a grassroots level. It is important to realize that industrial control systems have characteristics of a living organism that continually changes, Habibi remarked, and also that eliminating the problem of cybersecurity will never be accomplished with a single solution.
Another speaker at the cybersecurity session was Gavin Mead, principal, cyber services for KPMG LLP (New York; www.kpmg.com/us). Mead also addressed the need for cybersecurity to extend into the OT world. Cybersecurity has been a hot topic in IT for several years, but the wave of interest in securing OT components from cyber threats has been more recent.
Mead pointed out several reasons for why the cybersecurity threat is growing. These include the fact that industrial automation systems are more sophisticated now, and that realtime business decisions are increasingly made with information from the control system. In addition, commoditized IT systems are common, and they support the OT system. Meanwhile, cyber attackers are increasingly sophisticated and well-funded.
The greatest risk to companies comes from failing to spend their financial resources for cybersecurity in the smartest way, Mead says. “We spend a disproportionate amount of money on assessing the problem and not enough on what the remediation will look like,” he says. “That should come sooner in the process.”
“The topic of cybersecurity is well discussed now, but there is still not enough sharing of information about cyberattack incidents,” Mead says.
Jeff Melrose, principal technology strategist for cybersecurity at Yokogawa Corp. of America (Sugar Land, Tex.; www.yokogawa.com/us) added a new dimension to the AFPM session by discussing potential cybersecurity threats associated with drones. He says drone technology, even that available to hobbyists, has evolved to the point where their range is up to three miles, and they have the ability to maintain a stable hover or follow a target autonomously for 30 minutes more.
A drone equipped with electronic transmitters could theoretically follow a target and be directed remotely to disrupt wireless communications or surveil. Melrose suggested that refineries should begin instructing physical security personnel to look for drone activity near plants and should update procedures to include what to do if a drone approaches a facility.
Innovations place more robust instruments in demanding applications and enable processors to take advantage of the industrial internet of things…
NTT Communications Corp. (NTT Com; www.ntt.com) and Mitsui Chemicals Inc. (both Tokyo, Japan; www.mitsuichem.com) have demonstrated the ability to predict,…
In late June, Total (Paris; www.total.com) announced the launch of its second Plant 4.0 start-up "incubator," with the goal of…
An overview of how the chemical process industries are looking to the IIoT to help form new business models in…
Predictive process models enable more efficient operations and higher-value products When they hear the words “process modeling,” some may envision…