A Forward Look At The Refinery Of The Future

REFINING

A forward look at the refinery of the future A forecast of the way business and operating fundamentals in a long-established industry are likely to affect future technical applications. The author sees present and emerging technology providing the adaptability to deal with plant disruptions Calvin B Cobb Invensys Inc

T

he refining industry has been operating in a difficult environment for over two decades. There have been only a few years since the early 1980s when refiners produced decent returns on their assets. As a result, capital investments decreased because of a lack of performance. The good news is that the industry seems to be on the verge of a new era of performance that is attainable if companies are willing to transform the fundamentals of how they operate their business. Though certainly not peculiar to the refining industry, consolidation among major oil companies as well as independent refiners in markets such as the USA and Europe has been the name of the game of late. For example, not too long ago few industry observers would have thought that only 10 companies would own over 80% of US refinery capacity. This surefire sign of industry maturing is generally accepted as a permanent state that has been caused by the levelling-off of demand for refined products. Lower demand caused the capacity reductions that have been observed since the early 1980s. Even so, there is slightly more refinery capacity today than demand for those products. And most of the time, refining overcapacity translates into low profitability. When refinery margins, prices or profits increase, it is usually due to the alignment of factors like crude supply constraints, weather, local demand, refinery outages, pipeline operating problems, or new product grade introductions. Even though refining is a tough business, it remains a good business segment for oil companies. Refining generates cash, stabilises cash flow and earnings, provides a connection to endmarkets and customers, offers a secure outlet for crude, and is the key enabler for the upstream exploration and production industry. Refining is an essential business to all the world’s economies – there are no current economically priced alternatives to prod-

ucts derived from crude oil that can supply energy demands. Thus, the industry finds itself in a paradox. Despite having a unique marketplace, market conditions can at best be classified as neutral to unfavourable. Expectations for return on investments (ROI) are low, despite ever-increasing need for large capital investments because of increasingly stringent requirements for cleaner-burning fuels. Industry executives anticipate continuing high investment requirements in the foreseeable future for refining process technology, because it is expensive to produces fuels that have lower sulphur content, lower aromatics content, oxygenates, and other cleaner burning formulations

Enabling systems In addition to investments in new process technology, companies have been forced to make other large capital investments. As the world readied itself for the anticipated Year 2000 computer complexities, nearly all refiners took the opportunity to install new enterprise resource planning (ERP) software, primarily SAP. Given the massive effort to replace most of the legacy computer systems within a company, the costs were very high. In many cases, ERP installation costs reached or exceeded the capital costs for clean fuels processes such as hydrotreaters. The refining industry landscape in all of the world’s mature regions (USA, Europe and Japan) continues to be changeable. Without the strong pull of product demand growth, the many factors that affect refinery financial performance can line up to increase margins or decrease them. In this setting, refiners must work to make their businesses more adaptive to the disruptions that accompany the business dynamics in those mature economies. These scenarios place enormous pressure on downstream executives to improve financial performance when the fundamentals are not conducive to

21 P T Q SPRING 2004

w w w. e p t q . c o m

higher profits and return on capital. What strategies will work? First, consider the challenges accompanying the industry’s situation.

Today’s challenges The challenges facing the refining business include its capital-intensive nature, excess gasoline production capacity, and mass merchandisers in the retail marketing segment, among others. Beyond those, perhaps the largest challenge is dealing with the disruptions to business operations that are inherent in industry. Disruptions are a way of life in refining and most executives plan on them continuing. Companies usually can’t control all the factors that cause disruptions, so they need to learn to adapt to them. First among the possible sources of disruption is the use of diverse technology in one of the world’s longest and most complex supply chains. No other industry deals with a value chain beginning at a natural resource in the ground and continuing all the way through to the end-use consumer, from wellhead to pump. The visibility of this long chain adds to the complexity of running and optimising the business. Another unique disruption coming to the industry is the anticipated “greying” of the workforce and the great “crew change” of the future. Many of the experienced professionals in the industry are reaching retirement and companies are increasingly aware that their workforce will change dramatically in as few as six years, when an estimated 30% to 50% of the professionally trained workforce becomes eligible for retirement. Indeed, much of the intellectual property embodied in today’s cadre of technical workers will go out the door when they leave. That’s because there is a huge shortage of trained talent waiting in the wings to replace the experienced workers. New technological changes are in the offing as well. Two are of a “game

REFINING changing” nature. Most industry leaders are vigorously pursuing gas-to-liquids (GTL) technology advances very hard. If successful, the conversion of natural gas to clean energy products will forever change the characteristics of today’s refineries in terms of feedstock and product slate. New automobile technology will likewise forever change transportation fuel demand and supply dynamics. New automobile power plants such as hydrogen, fuel cells, and natural gas will cause refineries to adapt to the change in product slates. Given these diverse challenges, there will not be one magic solution that can lead to increased short-term or long-term per-

formance. Each refiner is challenged to find the proper strategies using today’s technologies to drive performance higher. There is good news, however. The economic incentives are more than adequate to be attractive investment opportunities for refiners. Plus, today’s technologies are capable of much more.

Technology solutions In the refinery landscape previously described, along with its many challenges, it is imperative for refiners to raise their operations to new levels of performance. Merely extending today's performance incrementally will fail to meet most company’s performance

22 P T Q SPRING 2004

goals. The central question is how? Two strategies are gaining momentum. First, the concept of reshaping the refining business to be more adaptive to changing business and economic conditions is gaining wider acceptance. At the same time that refiners are learning to adapt, they must also be transforming the way they run their business. Combing both strategies provides a path to higher performance. Implementing these strategies won’t be easy and will not occur overnight. Indeed, that is why the Centre for Hydrocarbon Business Innovation was formed to look to the future of the oil, gas, and chemical industries and explore how the technology and methodology of refinery operations can be translated into increased profitability. Through research and discussions with industry executives, the Centre has identified two main strategies for successfully engineering increased performance. They are: Business transformation. Melding the people, business processes, and information technology into a focused whole that raises performance in a step change fashion rather than incrementally Adaptive processes. Processes that can adapt to changing markets, learn from experience, and can be built into the refinery of the future. The most compelling business transformation strategy must involve the huge information technology (IT) investments refiners made in the late 90s. Most companies invested heavily in IT – an estimated $300 billion – and yet most have not yet achieved the promised results. The shortfall appears to have occurred because companies did not transform the basic ways they do business to take advantage of those powerful computer systems. The “back-office” processes were transformed to install new systems but the task of a complete overhaul of the front office was too monumental to undertake simultaneously. Today, most executives recognise the need to integrate people, business processes and technology in order to drive performance. This is a change from previous years. Now, nearly everyone has come to the conclusion that installing software, no matter how sophisticated, is not the complete answer to higher performance. But there’s yet another component to business transformation that research from the Centre for Hydrocarbon Business Innovation believes of vital importance to fundamental industry transformation. Centre research indicates there is more valuable business data and information existing at the refinery and operations level than

REFINING ever before in history. This data can and should be effectively used to drive performance to higher levels. Indeed, intelligently linking technology and business offers the greatest avenue to improved performance, and at the same time addressing the issue of adaptability, for dealing with disruptions. Real time integrated operations are possible today through the applications of both existing and emerging technology.

Strategies Operational and logistical challenges provide the reason for designing new refinery business processes that combine the benefits of flexible hardware with technology-enabled decision-making. The three key operational challenges facing refiners are managing commodity risk, validating feedstock margins once the feedstocks have been processed, and balancing operational stability with commercial flexibility. Logistical challenges include cargo schedule optimisation, pipeline scheduling integration with refinery scheduling, and managing and monitoring feedstock quality throughout the supply chain. Yet another challenge is recognising that there are more supply chains that intersect the refinery in addition to oil. Among these are the power, gas and hydrogen supply chains. Some companies are beginning to address planning for four supply chains instead of just one. Process challenges include sulphur species management, naphthenic acid mitigation, and matching hardware to “cocktail” feedstocks and intermediates. In the Centre’s vocabulary, hardware flexibility includes increased fractionation, feedstock blending and employment of various sulphur removal technologies. Technology should be used to match feedstock-to-hardware availability, including real time feedstock blending, feedstock characterisation, broader use of real time feedstock analysers (such as NIR and NMR) and optimisation of the crude unit. Technology-enabled business processes will improve the modelling fidelity and molecular characterisation of refinery streams, particularly for sulphur species and multi-ringed hydrocarbons. With the integration of information technology and knowledge capture, crude slates will be balanced more and more in a real time planning environment that will maximise value. Technology-enabled business processes and event management have the potential of dramatically changing the way refiners operate and manage their plants in the future. Many convention-

al paradigms of refinery operation will be challenged. Most of the technology elements are available today but require enlightened thought and change-management leadership to implement and capture.

Workflow As already described, the refinery of the future must implement a number of key strategies, but one common element of the transformation will revolve around the central theme of workflows. In simple terms, workflow-centred transformation begins by identifying business improvement goals, like margin, costs, safety, etc. Company leaders always know these goals and good management make these straightforward and easy to comprehend. Once established, the company creates diagnostic and design efforts to achieve the goals. Ultimately, the result can be simply stated as defining what management wants people to do in their jobs and providing them with the proper data, information, and knowledge to make better decisions. Towards this end, key processes and workflows must be identified in order to gain the greatest potential benefits. The three most important among them are: — Refinery reliability and predictive operations/maintenance — Advanced planning and modelling — Agile operations. Can process and workflow optimisation really be achieved? Invensys has completed three projects that connected plant-level systems-to-business systems with the objective of increasing performance. All three – a crude oil production/pipeline/heavy oil upgrading facility, an LNG liquefaction plant, and a petrochemical plant – demonstrated what had been missing in the ERP implementations as well as how the gaps could be filled in order to operate the businesses for greater profit. In all three cases, information from plant smart devices, including program logic controllers, DCS systems etc, were connected to a middle technical layer, then to the business systems. Today, technological advances like Invensys’ ArchestrA open framework provides the interoperability to connect all the information sources in a refinery and map the data and information to the knowledge workers who can make better decisions. This therefore provides the confidence in asserting that the refinery of the future will include the three previously noted key elements.

Predictive operations In terms of refinery reliability and predictive operations/maintenance, refin-

24 P T Q SPRING 2004

ers abhor unplanned shutdowns. Reducing their number and frequency is always a primary objective. One refiner recently responded to a reporter’s question about how many unplanned shutdowns were acceptable with a short quote: “Zero”. While the refinery of the future may not consistently reach the zero objective every year, it will utilise real time plant information to predict equipment and process health to reduce unplanned downtime by one to two percent of today’s average of around 4%. This will be achieved by integrating equipment monitoring technology and predictive maintenance, predictive process degeneration (heat exchanger fouling, catalyst activity, furnace efficiency), and dynamic performance management. Technology to accomplish the individual steps exists today but it is not integrated or mapped to the work processes of refinery engineers. The ArchestrA system is again an example of technology that can link the technologies together and provide operators and engineers the predictive capabilities to improve refinery reliability.

Advanced planning In terms of advanced planning and modelling, crude oils are becoming highly variable, with higher concentrations of sulphur. At the same time, more stringent anti-pollution regulations are forcing greater restrictions on fuel specifications. But chemists know there are fundamental limitations on how far empirical systems can go in achieving proper control over processes that require new knowledge for their solution. Today’s regulatory constraints are written in terms of aromatics and sulphur, not boiling range “lumps”. The only way to change this is to change the basic information by which all process calculations are based. In other words, infuse the system with new knowledge that, in turn, will provide the opportunity to develop new knowledge. While it is not often that there are radical leaps in technology in the hydrocarbons industry, it may now be possible to think about refining streams as a chemist would; that is, by considering the molecular components of crude oil rather than as pseudo components characterised by the traditional boiling fraction methods. Historical attempts to build molecular refinery linear programs (LP) were limited to the lighter boiling components (those that boil at less than 400°F) because molecular level physical property data were not readily available for higher boiling components. Invensys’ first experience in building such

REFINING LPs was in 1985, when a refiner was assisted in evaluating new refinery processing schemes to deal with specific components like benzene and C7 to C9 paraffins. Useful results were produced but the usefulness of such models was clearly limited by not being able to represent all the crude oil components. Building molecular-based models is not new, however. In 1897, C W Marbury discovered that crude oil was comprised of a homologous series of hydrocarbons. From 1926 to 1972, the American Petroleum Institute funded research on the subject, but it was ultimately abandoned because of the complexities. In 1982, J W Bunger conceptualised a computational chem-

istry approach based on structural indices for digitising petroleum composition and properties. This method was proven by 1995 in ARCO-sponsored research conducted by Bunger and Associates and given the trademarked name Z-BaSIC, which stands for Zbased structural index correlation method. The method’s letter Z is the hydrogen index that describes the hydrocarbons in crude oil. The method provides the following capabilities: Represents a basis for new data and knowledge refiners can potentially use in a variety of ways to improve understanding of crude oils and refining Allows true chemical composition modelling instead of correlation-based mod-

elling techniques. It calculates the property of a component through correlation of the correlating Z number and molecular weight, without having to identify it uniquely (including heavy ends which are traditionally beyond the capabilities of conventional molecular analysis techniques) Allows a meaningful analysis and modelling of the molecular composition of crude oil. Because a distribution of crude oil constituents is not assumed and individual molecules can be traced, streams can be accurately analysed. The method is able to analyse crude oil management in greater detail, in a fraction of the time and at lower costs than traditional methods. What can be done with this new approach? The learned practitioner will quickly recognise a myriad of benefits from a molecular characterisation and the advanced planning models in which it can be placed. In addition to molecular characterisation of refinery streams, improved optimiser technologies will dramatically change the way companies conduct refinery planning and scheduling. By relying on first principles, the new modelling environment will: converge the planning, scheduling and yield accounting model environments; populate inferential characteristics of product properties such as RVP, distillation, V/L, and cetane with pure property predictions; and greatly enhance model fidelity and simplify planning, scheduling, and yield accounting. How can this analytical methodology be turned into reality? Incorporating Z-BaSIC into industry-leading refinery models like the Invensys programs PETRO (LP), PROII and ROMeo is a logical starting point for using this new technology. This effort will involve, first building the Z-BaSIC physical property characterisations and, secondly, rebuilding computational algorithms to handle molecular-level descriptions of complex hydrocarbon streams. While this process will require a great deal of effort, it will provide new tools to make major breakthroughs in how refineries are operated. Furthermore, it will allow companies the ability to improve financial performance significantly. Sulphur speciation is the first practical application that fits with the clean fuels projects being initiated by most refiners where we can identify and track the nearly 200 individual sulphur compounds through the refinery.

Agile operations A senior executive stated recently: “I want to make 10 decisions a month that could add to our profitability, but

26 P T Q SPRING 2004

REFINING we do not have that capability today. We can make one-month decisions and maybe one-week decisions but we cannot make daily or hourly decisions.” He mused that he hoped seven of the 10 decisions would be good ones recognising the inherent uncertainty in making them. The scenario he described can be summed up as “agile operations”, a Holy Grail of sorts in today’s plant operating environment. But is it attainable? From Invensys’ view, today’s powerful computers combined with advancements in optimisation solvers make it possible to enhance refinery capabilities to reflect both the actual operational capability and the constraints of refineries. By extending traditional linear optimisation to include non-linear and mixed integer optimisation, large complex problems that were previously not feasible can now be robustly solved with the additional mathematical processing and advanced solver technologies. For example, advanced refinery-wide optimisation will improve refinery planning and scheduling and result in better feedstock selection, operational planning and target setting, and blending decisions. Integrating these tools with real time operational, scheduling and pricing information in order to provide continuous advisory alerts to key decision makers can further extend the benefits. Continued advances in computing power make the more sophisticated modelling methods economical and allow technology delivery of leading-edge tools, like some of those previously described.

Opportunities Dealing with disruption is possible and should be considered a mission-critical objective of management when considering the likely future course of business events. 2004 will be yet one more step down the road. Refiners are embarking on a business transformation journey that will include real time planning, non-linear optimisation modelling, next generation product blending systems, molecular modelling, predictive refinery maintenance, and interactive supply chain systems. The refinery of the future will be able to adjust refinery operations to maximise its risk-adjusted profits. The agility will be guided by sophisticated option analysis and uncertainty analysis capability to properly weigh all the factors. The concept of building adaptive business processes that transform the things done today affords a path to increased profit.

This article is based on a paper prepared for presentation at the NPRA annual meeting, San Antonio, Texas, USA, March 2004.

Calvin B Cobb is vice president and general manager, global hydrocarbons consulting, Invensys Inc, Houston, Texas, USA. He has 30 years’ experience in the energy sector and heads the company’s hydrocarbons consulting group, which includes the Centre for Hydrocarbons Business Innovation. E-mail: [email protected]

27 P T Q SPRING 2004