Energy Symposium San Francisco May 1981 of the Exxon ENERGY RESEARCH DEVELOPMENT SYMPOSIUM: Sclence, Technology, and the Energy Transl?on San Callfornla, May 19, 1981 MARGERY HELD Symposium Coordinator EXXON RESEARCH AND ENGINEERING COMPANY Florham Park, New Jersey Canyrigm 2 mm Exxon lnseurh m1 Inginewins Lompiny CONTENTS Preiacc wymam Shtk, I: . Overview. Edward David, .. New Fronuev; in Energy Snowa .V .. .7 New Sourtes and The" TcdmoVoglcaV ProspeclerankR Sprow . . W37 Advances Exp?olznun and Vrodurnon Vechnohngy. erlvarn Energy Efficient Refineries Makmg More 1mm Le" CIarcheM mun . . as Energy RM) Sympasmm Auendees . . . . . ..1o7 ABOUT THE AUTHOR William T. Slick, Jr., is a Senior Vice President of Exxon Company, U.S.A. He graduated from Brown University with a 3.5. degree in Mechanical Engineering. He joined Exxon Company, Production Department in 1948, serving in a number of engineering field assignments and in the Houston Headquarters. In 1966, Mr. Slick became Assistant General Manager of the Natural Gas Department. Following an assignment in New York City in 1967 as Assistant Manager of Coordination and Planning for Exxon Corporation, he returned to Houston in 1968 as Manager of Planning for the Company?s Supply Department. in 1969, he was named Assistant Manager of the Corporate Planning Department, and in June 1972 he became Man- ager of the Public Affairs Department. He assumed his present position in May 1973 and is currently responsible for Exxon, U.S.A.'s Marine, Pipeline, and Public Affairs activities. A Registered Professional Engineer, he is a member of the American Institute of Mining, Metallurgical, and Petroleum Engineers; the American Petroleum Insti- tute; and the Council on Trends and Perspective of the U.S. Chamber of Commerce. He is Vice Chairman of the Texas Research League, ViCe President of the Association for Community Television, and a trustee of Brown University. W. T. Slick, Jr. PREFACE it is my pleasure to welcome readers to the Proceedings of this sym- posium on Science, Technology, and the Energy Transition. At Exxon, our commitment to scientific research goes back nearly a century. in 1882. our organization established the world?s first laboratory for test- ing fuels and lubricants for quality. Today, our research activities span the globe. This year, Exxon-funded expenditures will reach $635 million, more than triple what they were just five years ago. This growrng financial commitment is evidence of our continuing belief that we cannot do business in this industry. and in this age, without a strong technological base. THE ENERGY TRANSITION In these proceedings, we have described some of our thoughts on the future technological challenges in energy and on what we are doing in research and development in our efforts to deal with them. Here, however, I will discuss from the vantage point of Exxon Com- pany, management some of the parameters of the "Energy Transition? referred to in the title that we have chosen for this confer? ence. and the effect of this transition on our needs for technology. Over the long term, this transition will carry the United States?and the world?from today's petroleum-dependent economy to one that will rely principally on nondepleting and renewable energy sources. Conservation, the more efficient use of energy, will be essential in managing this transition successfully. Much has already been ac- complished in conservation, but our projections show that, despite our best efforts, US. demand for energy will increase by 2092 by the end of this century. Consumption must rise if there is to be sufficient economic growth to create new jobs, to improve standards of living, and to achieve other social goals to which the nation is committed. The substitution of other fuels in end uses now supplied by pe- troleum will be another important factor in this transition. During the rest of this century, coal, nuclear fission, and, to a lesser extent, hydro- electric and geothermal energy will supply an increasing share of total demand. More and more end?use consumption will be in the form of electricity. The demand for liquid petroleum has peaked and will continue to decline throughout the balance of this century. But there will remain large segments of demand, such as in transportation, where electricity or other forms of energy are not practical for economic or technologi- cal reasons. There will also be a continuing need for liquid and gase- ous materials as petrochemical feedstocks and as specialized fuels. A look at the supply side quickly tells us that oil and gas will be our most critical fuels. While demand for them is declining, so is our domestic supply. For almost a decade, one could say we have been living off past glories. Additions to reserves have not kept pace with consumption, either at home or abroad. The recent increases in domestic prices for crude and natural gas have had a substantial impact on exploration and development efforts, including tertiary recovery. However, even this increased activity will only serve to slow the decline in production capacity. Accordingly, while energy consumption will continue to rise over the next two decades, we can expect a widening gap between the demand for energy in liquid and gaseous forms and our ability to meet this demand with conventional, domestically produced petroleum. We could attempt to fill this gap with imported oil, and to accept the risks that come with that, or we can choose to fill it with fuels manufactured principally from domestic supplies of coal and oil shale. in our view, will be the key to making the long-term transi- tion in an orderly way. In effect, they offer a bridge from our depen- dence on petroleum to that time when we can rely on nondepleting energy sources. I will have more to say about later. First, however, I?d like to comment on several aspects of our basic business that place a premium on the development of new technology. INVESTMENTS AND TECHNOLOGY Conventional oil and gas will, as I mentioned, represent a significant part of supply for many years to come. However, thay are becoming increasingly difficult to find, to produce, and to process. PREFACE VII In the upstream part of our industry?petroleum exploration and production?hadding to our conventional petroleum reserves will re? quire working at greater drilling depths, in deeper waters, and in harsher environments, often looking for smaller deposits. These opera- tions bring with them a continuing need for new technology, to say nothing of the higher associated costs. Let us look at a few examples. A recent oil discovery in the Gulf of Mexico is in 1,000 feet of water, which approaches the limits of design for conventional production platforms. We have chosen to develop this field by using, for the first time, the new guyed-tower technology developed by Exxon Production Research Company. This effort will cost some $850 million. The current incentives for higher recovery efficiencies are obvious. Exxon now has over 400 active secondary?recovery projects, and we are moving to accelerate a variety of tertiary processes. Current tertiary-recovery activities include steam stimulation in California, a polymer flood in Mississippi, an inert-gas flood in Texas, a pilot surfac- tant flood in Illinois, and a miscible inert-gas flood in Florida. We and the industry have a long way to go, however, to develop widely appli- cable tertiary-recovery technology, especially the use of surfactants. As for offshore exploration, the search for oil in the Arctic holds great potential, but it presents unique problems, both logistical and technological. Such problems are compounded when we consider offshore exploration in the Arctic. By way of example, in the Beaufort Sea we are contemplating the use of man-made gravel islands as dril- ling sites where the use of conventional steel platforms is impractical. The end of the petroleum business?refining and marketing?presents some different challenges, brought on by two di- verging trends. On one hand, more and more of the new crude supplies?both domestic and foreign?are in the lower-gravity, heavier ranges and contain increasing amounts of sulfur and metallic impurities. At the same time, demand is shifting more and more toward the lighter, higher-gravity liquid products of higher quality. The resul- tant need for refining operations of higher severity places a premium on the development of more effective and more efficient processes, to say nothing of the obvious need for a public-policy cli- mate that encourages refiners to make the substantial investments necessary to process these poorer raw materials. Of course, the ultimate in heavy feedstocks are the solids such as coal and oil shale, the raw materials for fuels. As I mentioned EXXON ENERGY REID SYMPOSIUM earlier, offer the potential solution in the transition to the nondepleting-energy future. Our own programs are quite ambitious. You?ll read more, later, about the technical work that we have under way. For now, I?ll limit my comments to current and planned domestic operations. We are the operator and 60% owner of the Colony Shale Oil project on the Western Slope of Colorado. Field construction has begun, and startup is scheduled for late 1985. This project will involve room-and- pillar mining of 60,000 to 70,000 tons per day of oil shale and will utilize the Tosco retorting method to produce 47,000 barrels of shale oil per day. The price tag is over $3 billion. I We are also studying a project to mine and to gasify East Texas ligntte, using Lurgi technology. The resultant intermediate?Btu gas could be sold as an industrial fuel or as a petrochemical feedstock or it could be further processed to make liquid fuels. I Current design efforts will lead to a decision on appropriation in early 1983 and to potential startup in about 1988. The expected in- vestment is 33-54 billion. We currently have in operation a coal-liquefaction pilot plant at Baytown, Texas, that will test the Exxon Donor Solvent Process using several varieties of coal as feedstock. Its construction cost of $118 million and a total program cost of over $350 million are afar cry from some ofour intuitive thoughts on the costs of research pilot plants. This pr0ject is a joint Government-industry program underwritten about 50% by the US. Department of Energy; Exxon's share is about 24%. ASSESSING THE RISKS To place in context these Exxon efforts on fuels and my earlier references to the transition, I will try to capsulize for vou some of the results of our studies on the potential emergence of a US. industry. In our judgment, it would be possible for this country to have a industry capable of producing 15 million barrels per (lav of oil equivalent early in the next centurv.lThe resources are available. The technology is available. The capital could be raised. The construction and operating personnel could be trained. This would obviously be a massive job. It would require investments of some $3 trillion ($850 billion in 1980 dollars). It would require 30 years to accomplish. Most importantly, it would require a national commitment the like of which we have probably never seen in peacetime. PREFACE In many respects, however, it is at the level of the individual project that the effort to build a industry will succeed or fail. Individual investments in such projects will be virtually unprecedented in size. A single plant and associated mine to produce about 50,000 barrels per day of oil equivalent from coal or oil shale will probably cost billion. This one-plant investment ex- ceeds the 1980 assets of more than 80% of Fortune 500 companies. Investments of this magnitude obviously present some very large and unique risks. Initially, there is the risk of scaling up to commercial size with technology that has been previously applied only in large pilot plants or in one-module demonstration plants. The long lead times involved in planning, designing, and building these plants?six to eight years?subject the projects to a wide range of possible changes in Governmental policies or regulatory requirements, including changes in world economic conditions. There are also the risks of completion delays or the cost overruns common to large new ventures, and the risk inherent in implementing projects of such unprecedented magnitude. For a significant industry to come into being in any reasonable time frame, there is another unique risk. Because of the long lead time for individual plants, second and subsequent generations of plants must be in design and construction before their predecessors have provided the benefits of extended operating experience. Clearly, to build an industry of 15 million barrels per day from scratch in 30 years is an ambitious notion, to say the least. If such a scenario is to come about, one thing is certain: the managements of the participating companies must place great confidence in the ability of their technical organizations to develop individual projects?each of which requires an unprecedented capital commitment?and then to provide the continually evolving technology essential to keeping these investments economically viable. For our part, we at Exxon view all of our future activities?be they oil and gas exploration and production, refining, or fuels?as areas in which new technology and higher investments will play cru- cial roles. Just to drive this point home, we expect that 70% of Exxon Company, profits in the year 2000 will come from operations that will require the successful development and commercial applica- tion of new or extended technology. In this volume, Dr. E. E. David, Jr., and his associates will be sharing with you highlights of some of their efforts at the leading edges of technology that, at the risk of sounding poetic, give us the courage to dream so boldly. ABOUT THE AUTHOR Edward David, Jr., has been President of Exxon Research and Engineering Company since 1977. He received his BS. degree in electrical engineering from the Georgia Insti- tute of Technology in 1945 and his 5c.D. from the Mas sachusetts Institute of Technology in 1950. At Bell Tele: phone Laboratories, he rose to the position of Executive Director, Research, Communication Principles Division leavmg in 1970 to become Science Advisor to the Presi?I dent of the United States. In 1973, he became Executiv Vice Presrdent of Gould Inc. He served as Chairman of th: Board of Directors of The Aerospace Corporation and is a member of the Boards of the American Association for the Advancement of Science, Materials Research Corporation Massachusetts Institute of Technology, Carnegie Institutiorl of Washington, University of Rochester, and of the Facult of Arts Sciences of the University of Dry Davrd IS U.S. representative to the NATO Science Commit: tee and on the Advisory Councils of the Stanford Research Institute, Princeton University, and the Universit of Chicago. He holds seven honorary degrees and pla Islan active leadership role in the National Academy of/ Sci- ences, the National Academ - . - En meerln IE E. E. David, Jr. OVERVIEW In his Preface to these Proceedings, Mr. W. T. Slick, Jr., closed his remarks by referring to the importance of the economic climate in the innovation process. For those of us engaged in commercial this is an inescapable fact. The demand at the business end of the pipeline is critical in determining the rate of innovative flow in all parts of the research and development process. But there are other factors as well, factors related closely to the purpose of this symposium.* Economists have often noted that a private company captures as profits only a small fraction of the economic value of its own innova? tions. The rest escapes in various forms to other companies, workers, and consumers??in a word, to society. The other side of this coin is that no company creates new technology independent of the economic, social, educational, and legal systems that create a framework for innovation. In particular, scientists and engineers know by long experience that their common enterprise thrives best on an abundant interchange of fact and opinion within their community as a whole. Senator Patrick Moynihan once said that science and technol- ogy are not a secret, they are a system. That is central to the purpose of our energy symposia, which are intended to help to strengthen that system, to the mutual benefit of all concerned, by furthering dialogue in areas of broad scientific and technical interest. The energy situation *This symposium was held in San Francisco on May 19, 1981, as part of a continuing program of symposia that will be held in cities in at least four countries in 1981. It was the third held in the United States, the first two having been in New York City. The audience at the San Francisco symposium represented all sectors of the science and technology community on the West Coast and in the Rocky Mountain area?government, industry, and academia. FT EXXON ENERGY SYMPOSIUM that Mr. Slick has just described lends to many of these areas a political interest as well. Before introducingthe papers in these Proceedings, permit me first to provide context by telling you something about Exxon's strategy for research and development, and to follow this by a quick breakdown of our budget. Certain aspects of the energy transition remind people of the mad tea party described in Alice in Wonderland. When the tea and cakes were exhausted at one seat, the natural thing for the Mad Hatter and the March Hare to do was to move on and to occupy the next seats. When Alice inquired what would happen when they came around again to their original positions, the March Hare changed the subject. Obvi- ously, this Mad Hatter philosophy neglects the creative role of the scientific and engineering intellect. True, oil and gas will be exhausted one day. However, our premise at Exxon is that by then we will have kept the table full by replacing those resources by using new informa? tion and knowledge. For this to happen, we know that science and technology will have to play an even larger role in our industy than they have in the past. You may already have gathered that, in energy, the focus of Exxon?s is on petroleum exploration and production, fuels, lubricants, petroleum processing, and fuels, as well as alternatives such as photovoltaic cells and batteries. As the papers that follow will suggest, this involves work in all of the obvious areas of science and engineering, besides related ones ranging from metallurgy to mi- crobiology and advanced computing. However, this cannot be a mere collection of projects. It must be animated by a coherent, yet flexible, strategy. What are the key considerations? The future that Mr. Slick has por- trayed is a reasonable one, but we know that it will certainly not turn out exactly as expected. People simply don?t have the gift of prophecy, and attempts at prophesying in the energy and environmental field have often had a political rather than a factual base. We have tried to avoid this here, but we are making certain assumptions that, in effect, constitute a scenario that, of course, is not the only one possible. One assumption that we are making is that, despite all of its uncer- tainty, the future will bring evolutionary, not revolutionary change. Another is that the diversity of technologies required to fuel the world will increase rather than diminish. We believe that industry will de? velop a diverse kit of tools to deal with specific resources and markets. For example, you will read about the specialized technologies re- 3 OVERVIEW quired to extract oil in severe Arctic environments and about the vari- ety of resources that may be Important procll?uclz?g fuels like the ones we use today, or unconventiona methanol. All of this suggests the most important element otlt aegis- strategy: emphasis on a diversity of approaches in a diversm re cal areas. No single option is likely to be the key to the ene?gyf/ uwe. Many options will fall by the wayside. More fully than ever de (aft, the are recognizing that old truth: "Half of all is waste . is, which half?" the viability of all our approaches, we rely he-aVIIyF on communication and feedback throughout the Exxon organizatiop em]: commercial innovations can succeed Without that interp ayt: influences?from the scientist at the bench to the salespersonl inO field, and including financial, operating, and marketing peopleve people like to run to the technical daylight, but that does not reoi:lg are of the responsibility for ensuring that technical and - in harmony. This can mean changing one or the other, or on. b. Incidentally, the Exxon strategy makes room tor small as we aiscef; improvements in technology. After all, an improvement of one pein re- in production recovery by enhanced reserv0ir management, or ver finery yield through development 3f a be?tesr can give fits, because the energy eman . - larT?iESt22tegy also aims at broadening our understanding of thisaegl; tific base underlying present and potential technologies. Moret an percent of our total effort is devoted to developing this base, a perccens: age comparable to those in the chemical and communication inf phe tries. The first of the papers in these Proceedings highlights slomeo fundamental research that we are doing. However, you note. in a If of the papers the attempt to develop a fundamental understanfkingfe the principles operating in our technical and engineerilng . and more we expand our information greater our ext id y, the more able we will be to respond to new needs an (LDDO - tunities. The energy change will probably be evolutionary, but a:ng will come more rapidly than in past decades, and the premium WI a tabilit'. . onltagogs without saying that one of the most importanlt elec-i ments of our strategv is our belief in the continued importance of and gaseous fuels derived from hydrocarbon resources(.j thcl) oktfoer papers presented here deal with the technical status an 0: look at such fuels. The first examines fuels, and the secon 0.0 It advanced technologies in petroleum exploration and production. 4 EXXON ENERGY SYMPOSIUM stretches the truth only a little to say that both can qualify today as new resources. The technology for finding and producing conventional oil and gas is anything but conventional these days. What is more, such technology will remain vital for a long time to come. Half of the world's petroleum has been found, the easier half. The industry must still meet the challenges of locating the rest, whether it is in the Arctic reaches or in the deep waters at the edges of the continental shelves, and then of getting out as much oil as ingenuity and economics will allow. Another strategic concern is to develop systems for the rapid, effi- cient deployment of technologies as they are required. A common characteristic of most significant energy projects today is their im- mense, multibillion-dollar scope. As Mr. Slick suggests, this makes the risks very high, especially because the payout is so long. Industry?s experience has already indicated that the prospects for economies in such megaprojects are sometimes dwarfed by the diseconomies of scale. Add to that the increasing unfamiliarity 0f the technologies that we are employing and you can see why the problems loom so large. For example, no one has yet built and operated a shale-oil or coal- liquefaction plant of the sizes now contemplated in the U.S. Critical here is the integration of research with engineering and with the man- agement of projects, including such mundane functions as contracting, field?labor management, and cost estimating. The task is as essential as it is challenging. Finally, certain strategic imperatives cut across all of our activities. Our last paper treats one of the most prominent: energy conservation in refining, our most energy-intensive operation. Equally as important are safeguards for protecting the environment and human health. None of our technologies move toward deployment until we are confident that these imperatives have been attained. Testimony to Exxon?s inten- sified efforts in this area was the recent dedication of a $22 million environmental?health laboratory that employs some 55 toxicologists, epidemiologists, industrial hygienists, and other health professionals. While we can argue over what is a reasonable range of scenarios for the future, we think that much of the strategy outlined here is insensi- tive to even fairly major changes in the future. We have looked at a wide range of possibilities, and we believe that much of our work on fundamentals is sound over that range. We have looked also at some Armageddon-type scenarios; these are difficult, but fundamental knowledge will still be useful, even though the major responses to sudden, events are likely to be primarily political and economic, at least in the early going. OVERVIEW 5 It remains for me to talk, briefly, about the financial dimensions of our work. in 1981, Exxon expects to carry out some $730 million. of Of this, $635 million will be Exxon money, with the rest coming from the U.S. Government and others. However, these figures do not include the even larger amounts allocated to exploring for new. re- sources and to erecting the facilities that bring new technology into actual use. Exxon?s worldwide exploration and capital budgets amount to some $11 billion in 1981. Since 1973, Exxon?s expenditures have been growing at an average rate of about 25% per year, well above the general price inflation. In 1981, the expenditures will be allocated about 35% to oil and gas; 15% to coal, shale, and tar sands; 10% to solar, nuclear, and other advanced forms of energy; 30% to nonenergy areas, primar? ily chemicals and information systems; and 10% to ?basic? corporate research. These numbers should tell you that we don?t believe that the future will be a repetition of the past. Indeed, the future seems never to have presented a more enigmatic face, nor to have demanded more thorough preparation. However, I, for one, am that the world will escape the fate of the Mad Hatter. Certainly, we at Exxon intend to do our part to help to create the information that the world will need for the energy transition ahead. Again, we will strive to do this through a strategy aimed at developing flexibility in our technolog- ical capabilities; through a deepened understanding of fundamentals; and through integration of our research, engineering, and busrness judgments. ABSTRACT The coming decades will see a gradual shift from our cu;-f rent dependence on oil and natural gas to greater coal, shale, heavy crude oils, and, probably, soar g3; Scientific discoveries will lead to new n: I materials, or new processes that can make a s?u stafn iar difference to this energy future. SpeCIfic examp es 0. o: energy-related corporate research are Cited. SomehijOJecar: such as characterization of the structure of large rtoc al bon molecules, are suggesting better ways to convir coed to high-value liquid fuels. A novel concept forr1 en Iancd absorption of CO2 and H25 from gas streams hashahrehat been commercialized. Concluding examples ig - novel solar photovoltaic material and an interesting nomenon: enhanced Raman emi55ion. ABOUT THE AUTHOR Peter I. Lucchesi joined Exxon Research and Elngiineecizg Company as a chemist in 1955 and is current yt President responsible for the_Corporate Researc Laboratories. During his career With Exxon, he has een involved with development, exploratory and lonl'g-rarigei research. He has had staff assignments in ExxonhC eBrrgician Company and Exxon Enterprises. He received is .d.his chemistry in 1949 from New York University, an h. there in 1954. Dr. Lucche5i spent a .yearteac (ling undergraduate chemistry at New York Univer5ity an 5 year and a half teaching undergraduate chemistry courste at the Illinois Institute of Technology. While at the Institt?u j, he carried out and published research on infrared met 5 and on the dissolution of insoluble salts in aqueous ia. Dr. Lucchesi is a member of the American Chemica fgci: ety, the American Association for the Advancefr?nsnt ot ence, Sigma Xi, and the Association of Researc irec . P. J. Lucchesi NEW FRONTIERS IN ENERGY SCIENCE In this paper I shall discuss some of the research that will form the scientific base for tomorrow?s new energy technologies. First, how- ever, permit me to give you a brief energy scenario for this future. In it, there will be a gradual shift from our current dependence on petroleum and natural gas to more reliance on coal, fuels, and renewa- ble energy resources, although petroleum will still be dominant well into the Zist century. Much conventional oil remains to be recovered, but we shall have to learn how to deal with more refractory, lower quality supplies of crude oil. The chemical industry will switch to feedstocks ranging from simple building blocks to com- plex coal liquids. Meanwhile, steadily rising costs will cause pressures for ever-more-efficient operations in terms of selectivity, conversion, and energy utilization. For many of the new technical requirements, however, we have little or no science available. Much of the scientific base that Exxon will need will have to be created. That is the job of our Corporate Research Laboratories, where most of Exxon?s fundamental research is undertaken. The centralized scientific program of these laboratories forms the scientific part of an innovation system. We work at those frontiers of science that we judge most relevant for the longer-term scenario that I have mentioned. From a small beginning in 1968, we have grown to our present size of about 250 doctoral-level scientists. Figure 1 shows what I mean by an innovation system for which our Corporate Research provides the scientific foundation. Moving from the bottom, a flow of scientific discoveries lead, for example, to new new materials, and new chemistry. After we have carried out the applied research, development, and engineering, we hope that these new findings will enter the marketplace. The upward flow repre- sents the generation of new options for the Corporation. It involves 7 8 EXXON ENERGY SYMPOSIUM Operations Needs and Applied Research and Engineering New Options Limitations Materials. Chemistry st 1? Science Figure 1. Exxon?s Corporate Research the science part of an innovation system. various company units, among which people as well as science and technology move freely. An opposite flow moves from the marketplace all the way back to the science base. This flow transmits the long-term needs and technical requirements that guide the scientists. One of the most difficult questions that we face is the selection of the scientific areas in which to invest our efforts. The number of exciting areas always exceeds our resources. First of all, we choose those areas that we feel are the most likely to have an effect on a number of energy technologies. Secondly, we prefer those areas that tend to reinforce each other. This is a point that I want to emphasize: the complexity of our technical problems demands a truly multidisciplinary approach. Finally, we need to make certain that each area can be staffed effec- tively to achieve a critical mass. A few areas of science are obviously relevant to the energy-futures scenario that I mentioned earlier. Analytical science is central to the characterization of the complex organic molecules in residua, coal, and shale oil, and to an understanding of their reactivity. New chemis- try is essential?everything from polymer chemistry to the chemistry of acid gases such as SOX, NOX, and H25. Separations have always been central in our business and will increase in importance. Finally, materials are the key to direct, economical utilization of solar energy, and to energy conversion in general. While we are currently working in many other fields, such as polymers, bioscience, catalysis, and laser science, for this paper I shall draw chiefly from the four areas that I have just mentioned. NEW SCIENTIFIC FHONTIEHS 9 With this as a general introduction, 1 shall turn, now, to specific examples of some of our energy-related research. Some of these exam- ples are discoveries that have already become inventions. Others will deal with research still in the discovery stage in which new tools, better understanding, and enhanced capabilities are being generated. COMPLEX ORGANIC MOLECULES As an example of frontiers of science that will have to be pushed back, let us look at some of our work on complex organic structures such as petroleum asphaltenes and the large molecules in coal. A detailed knowledge of the physics and chemistry of these complex molecules is central to the future technology of fuels, both conven- tional and We have had a sophisticated program to elucidate these structures. My example will deal with heavy crudes, but, as you will see later, much of the information learned spills over into work with coal, oil shale, and other hydrocarbon sources. Canadian Cold Lake crude is typical of some of the resources for heavy hydrocarbons that are now being utilized commercially. it really is different, as can be seen in Figure 2 by comparing its viscosity with that of a typical high?grade crude such as Arabian Light. We can quantify what this mean to refiners by comparing the crude?s distilla- tion curves, a common way of characterizing crude oils. Basically, we perform a batch distillation in which we measure the cumulative amount of material vaporized as the temperature is increased. As shown by the right-hand curve, about 30% of the Arabian Light crude boils off below which is typically considered the upper end of 600n Cold Lake Arab Light 500i 400? Boiling Point, ?c 300' 20c 100- 0 1 2'0 4'0 6'0 Figure 2. Canadian Cold Lake crude oil (left) is much more viscous and high boiling' than a typical high-grade crude such as Arabian Light (right). 10 EXXON ENERGY SYMPOSIUM the boiling range for the gasoline fraction in crude oil. However, Cold Lake crude consists mostly of cc?mdensed, aromatic-ring structures called asphaltenes, and only about 2% is volatile below To state the problem in another way, in Figure 3 we note a con- tinuum of molecular weights and hydrogen-to-carbon ratios as we go Molecular 1G to? 10? 10-l 105 Wt Reduction Gas Gasoline Distillates Vacuum Hesid Asphaltenes Hydrogen Carbon Flatio I I I CH. Shale Asphaltenes Cakes CHJ- Oil Heavy Crude Coal Gasoline Jet Fuel Figure 3. Continuum of molecular weights and hydrogen-to-carbon ratios. from methane, a gas containing four hydrogen atoms and one carbon atom in each molecule, to liquid fuels that have a ratio of hydrogen to carbon of about two, and finally to cokes that are essentially all carbon with almost no associated hydrogen. To produce distillate fuels of commercial interest we must efficiently add hydrogen to (or delete carbon from) the heavy resource materials shown on the right side of the bar charts. We can gasify heavy fractions all the way to hydrogen, carbon monoxide, methane, and other light hydrocarbon gases and then put these gases back together to form liquid fuels of commercial interest. Alternatively, we can crack the heavy molecules in the pres- ence of hydrogen to increase the yields of gasoline, jet juel, diesel oil, and the like. However, in refining these heavy feedstocks by current technology, the consumption of expensive hydrogen is several-fold higher than it needs to be. We do not have the chemistry to do better. To develop that chemistry, we must delve deeply into the structure of these complex structures. Past representations of asphaltenes, such as that shown in Figure 4, were based more on intuition than on fact and resembled very large ?chicken-wire? assemblages. Little was known of the structural fea? tures of minor constituents, chiefly sulfur and nitrogen, which greatly SCIENTIFIC FRONTIERS CH, cu,cn,cn,cn,?C3 Figure 4. Past asphaltene model. influence the cost of refining. The hexagonal shapes that you see here are basrc molecular building blocks based -on benzene, Figure 5, which contains six carbon atoms and six hydrogen atoms, with the carbon atoms forming a hexagonal ring in which each is attached to a hydrogen atom. As we fuse the benzene rings to make more-complex, condensed aromatics?for example anthracene, with three rings?the hydrogen?to-carbon ratio decreases because some of the carbon bonds that were attached to hydrogens are now attached to carbons in adja- cent rings. There is, now, no single analytical tool that will elucidate the struc? ture of asphaltenes, but a multifaceted team effort is yielding an in- formed consensus that we feel is close to the mark. We are particularly Benzene Anthracene 071 Figure 5. Condensed rings such as those of anthracene are defrcrent in hydrogen. 12 EXXON ENERGY RGD SYMPOSIUM proud of our characterization work using nuclear-magnetic-resonance (NMR) spectroscopy. In a strong magnetic field, the individual carbon and hydrogen atoms in hydrocarbon molecules can be made to absorb specific signals to produce spectra that are characteristic of how they are bound in the molecule. The intensity of the spectra tells us how much of each kind of bonding there is. Consider the liquid toluene (Figure 6): its molecule has a benzene ring of six carbon atoms, but one of the six hydrogens in benzene has been replaced by a methyl group, I i H- lcl: I H- 6-H Figure 6. Toluene (C7H8). There are three primary kinds of carbon atoms in toluene: one carbon atom bonded to three hydrogen atoms, five carbon atoms each bonded to one hydrogen atom, and one carbon atom that is not bonded to any hydrogen atom. The different carbons will emit distinc- tive NMR signals whose relative intensities will be in the ratio of 1 :521. Similarly, the NMR spectrum of the three hydrogens on the methyl group differs from the spectrum of the five hydrogens on the aromatic ring. It is easy to see why the NMR spectrum of a molecule of unknown composition can help us deduce its structure. However, while nuclear?magnetic-resonance spectroscopy, espe- cially carbon NMR, is potentially the most powerful tool available for the characterization of heavy petroleum fractions, a major problem has been the overall lack of detail in the NMR spectra for both hydrogen and carbon, and the resultant problem of extracting meaningful results from them. Figure 7 shows, at left, a composite NMR signal from a complex molecule containing four different kinds of hydrogen atoms. How can we resolve this curve into its component parts? The problem is analogous to having to deduce what coins I have in my pocket if I tell you that I have 7 coins totalling 62 units without your knowing which country minted the coins. If I tell you that there are four different kinds of coins involved, the problem is still indeterminate. Only when you know the exact coinage system pennies, nickels, dimes, and quarters) can you deduce the correct answer. Our work involved a novel deconvolution of this complex NMR spectrum, using nonlinear regression techniques to resolve the spec- NEW SCIENTIFIC FHONTIERS Intensity from TMS Spectral Response Spectral Response Model 0 Data from Spectrum Figure 7. Complex NMR spectrum of a large hydrocarbon molecule (deconvoluted spectrum). trum numerically into four component peaks that represent the four kinds of hydrogen atoms present. The mathematical model used in- volves a Iorenzian lineshape for the peaks, an assumption that is founded in NMR theory. The mathematical model shows excellent agreement when applied to complex mixtures of known compositiOn that were similar to the asphaltenes under study. In the central positon of Figure 7, we see the deconvoluted hydrogen spectrum for the four kinds of hydrogen in this system. If we add up the four curves from our mathematical model, the calculated composite spectrum shown at right in Figure 7 agrees very well with the measured values in the left part of Figure 7. Similarly, we have resolved the carbon NMR spectrum into its component parts, as shown in Figure 8 Now, let us go back to the historical idea of the asphaltene structure shown In Figure 4. Using the NMR techniques just described, and ectral Ingensity MOdel Data lrorn "c Spectrum 170 160 150 140 130 120 110 100 90 Spectral Response, Figure 8. Numerical deconvolution of spectrum. 14 EXXON ENERGY SYMPOSIUM several other methods including model-compound studies, a some- what different structure has emerged, as may be seen in Figure 9. This is characterized by smaller condensed aromatic clusters, longer side chains containing up to 35 carbon atoms, and a realistic representation of the nitrogen and sulfur moieties. Figure 9. Current View of asphaltene structure. Cooperative efforts in our Corporate Research and Analytical Divi- sions have specifically determined the size of the aromatic clusters, the of the aliphatic chains, the types of heteroatoms, and the nature of the peripheral groups. The locations of the heteroatoms are still being pursued. The structure of coal is even more complex, although in Figure 10 you can see quite a resemblance to the asphaltene structure shown in Figure 9. In addition to the condensed aromatics, sulfur, and nitrogen that we saw in asphaltenes, coal also contains considerable bound oxygen. This is usually present in the form of ether, ester, hydroxyl, or carboxylic acid groups. In order to unravel the structure of coal, we must deal with this organic oxygen chemistry and its implications. We have developed techniques that allow us to count different oxygen groups?for example, for a given coal to count the number of ether or carboxylic acid groups per 100 carbon atoms. I shall not discuss these further here; I want, instead, to show how we exploit this knowledge to search for more-selective thermal or pyrolytic chemistry for producing liquids from coal. We are using model compounds to elucidate the pyrolytic reactions that coal undergoes when heated in various atmospheres. We are studying the thermal chemistry of the reactive oxygen groups in order on, m, NEW SCIENTIFIC FHONTIERS 15 Ether on, o? Alcohol Hydrogen Bond 0 Mt Ester 0 Figure 10. Structure of coal. to improve our understanding of coal conversion. If the bound oxygen Isconverted to carbon monoxide or carbon dioxide during pyrolysis, this represents a loss of both heating value and yield of liquid product. In our mechanistic studies, we used a number of simple molecules that contain the same functional groups that are found in complex coal structures. I shall discuss only one of these, benzoic acid, which can serve as a model for the acidic groups in coal (Figure 11). Benzoic acid us, of course, benzene with one of its hydrogens replaced by a car- boxylic acid group, as shown at the top of this figure. We found that pyrolysis under an inert nitrogen atmosphere strips off the carboxylic acid group, forming CO2 and two reactive free radicals. The benzene (or phenyl) radicals polymerize to form low-value polyphenyl tars. The hydrogen radicals mostly combine with them- selves to form hydrogen gas. However, pyrolysis with excess hydrogen present yields benzaldehyde at lower right), benzyl alcohol toluene and no C02. These compounds are all useful liquids, so this thermal chemistry is in the right direction. 13 EXXON ENERGY SYMPOSIUM <9 . CO, 0 - Reduction Pathway (W 0H 1 1H2 CHO CH20H CHJ Polyphenyls 1 2 3 Figure 11. Control of oxygen chemistry is critical in pyrolysis. Similar results are obtained in pilot-plant pyrolysis of a Wyoming coal under hydrogen-starved (nitrogen) and hydrogen-rich conditions. Figure 12 plots percentage of original carboxylic acid content going to CO2 versus total liquid yield. As you can see, total liquid yields vary 80"} I Nitrogen 0 Hydrogen so-l CO, Gas Yield Wt?/o ot Acid in Dry 40- Feed Coal 20Total Liquid ?eld (incl. H20) Wt% on Dry Coal Figure 12. Carbon dioxide pathway mechanism confirmed in coal pyrolysis (dry Rawhide coal, inversely with yields of the later being a waste of valuable carbon resource. However, pyrolysis in hydrogen produces high liquid yields and little C02. Control of oxygen-involved thermal chemistry may therefore help us to find coal-conversion processes that are more selec- tive in producing liquids. NEW SCIENTIFIC FRONTIERS 17 GAS TREATING The transition to fuels will raise new technological prob- lems in the treating of fuel gas and gas made by gasification of residual oils, coal, and oil shale. The relatively large volumes of CO2 and H25 that contaminate these gases must be removed, so we have been investigating the mechanism of CO2 complexation in basic solu- tions. CO2 is typically removed from gas mixtures by contacting them with aqueous solutions containing organic amines and potassium car- bonate. We wanted to see whether the capacity and rate limitations of this process could be removed. As shown in Figure 13, we found that CO2 reacts with amines by two competing reactions: carbamate formation and bicarbonate formation. Gas Liquid RNH: Carbamate CO: Amine (FINHZ) Hco; FINHJ Bicarbonate Figure 13. Carbon?13 NMR confirms that CO2 reacts with amines by two competing reactions. The first reaction is fast; in fact, that is why the amine plus carbonate is used. The amine promotes the fast absorption of C02, whereas the reaction of CO, with carbonate to form bicarbonate is normally slow. But the carbamate is quite stable, hydrolvzing only slowly back to the amine and bicarbonate. Thus, the capacity ot' these absorber systems is severely limited because carbamate formation ties up the amine promoter. We found it necessary to develop new techniques for studying the carbamate-forming tendency of various modified amines in potassium carbonate gas?absorption systems. Polarography, pH measurements, and other conventional analytical tools are inoperative in these hot solutions. We now have special sealed systems and unique sampling probes that allow us to apply to hot, complex solutions and to unravel their thermodynamic properties. 18 EXXON ENERGY SYMPOSIUM By keeping the system above its vapor pressure, which is several atmospheres, a single, homogeneous, hot liquid phase can be scan- ned. This gave us reliable quantitative information on the concentra- tion of all chemical species in the system at temperatures of commer- cial interest. From these data it was possible to calculate equilibrium constants for both bicarbonate and carbamate formation. As shown in Figure 14, we discovered that amines with large, bulky substituents on the amine group have greatly improved capacity for Linear Amine -- 0 I N. OH Cc OH Stable Carbarnate Sterically 0? .. u- H10 Hindered cl; I 3 Amine Unstable Carbamate Figure 14. Undesirable carbamate formation is suppressed by steric hinderance at amino nitrogen site. CO2 absorption. This is because the bulky, hindered amine de- stabilizes the carbamate that forms and thus facilitates rapid bicarbo- nate formation. The higher CO2 capacity of the hindered-amine systems was con- firmed by vapor?liquid equilibria measurements and actual plant tests. The plot in Figure 15 shows the partial pressure of CO2 in the gas phase in equilibrium with solutions containing various loadings of C02. As this pressure approaches the pressure of CO2 in the gas being cleaned, the solution cannot pick 'up any more CO2 and must be regenerated, usually via heating with steam. It is clear that the novel hindered- amine systems offer almost twice the capacity versus the state-of-the- art, linear amines. Not only is the capacity higher, but the rate of CO2 absorption is also increased, thus permitting easy expansion of existing plants, lower operating costs reduced stripping?steam require? ments), and lower investments for new facilities. PFIOGRAMMED COMBUSTION While we hope that much of our pioneering research will affect the emerging fuels business, we must also remember the need for NEW SCIENTIFIC FRONTIERS 19 Linear Hindered Pco2 I 0.0 0.5 1.0 Solution Loading. cozlAmine Ratio Figure 15. The higher CO2 capacity of the hindered amine system is confirmed by vapor-liquid equilibria measurements. improved technology for the direct utilization of coal?burning?in an environmentally acceptable way. Our work has concentrated on a process we call staged combustion for controlling both nitrogen oxide and sulfur oxide emissions (NOX and SOX, for short). What we have done to improve the SOX situation is most easily explained in terms of another advanced concept for coal combustion: fluidized-bed combustion in the presence of limestone. The curve in Figure 16 shows, under adiabatic, or maximum, conditions, the tem- perature of a coal flame as a function of the coal?to-air ratio. The region to the left of the maximum, the fuel-lean region, is where conventional coal-fi red devices run. One can easily add calcium to the coal to try to absorb sulfur as calcium sulfate, but calcium sulfate is 1800 1700 1600 1500 1400 T, 1300 1200 03304 C85 1100 1000 900 Normal Operation Fuel Lean 1.0 Fuel Rich? Equivalence Ratio Figure 16. Adiabatic combustion temperature and stability of versus air/fuel ratio. 20 EXXON ENERGY BSD SYMPOSIUM thermally stable only at the temperature shown in the shaded region at the lower left of this figure. The fluidized-bed combustion system solves this problem by putting heat-transfer surfaces into the combustion zone. This forces the com- bustion to occur at much lower temperatures, so that calcium sulfate is stable and the emissions of sulfur oxides can be kept very low. This approach has two problems, however. First of all, having heat- transfer surfaces in a combustion zone is undesirable from an engineer- ing viewpoint; secondly, doing the job this way gives more emissions of NOX than one might wish. That is a little surprising, because ther- modynamically the NOX concentration should be minimal at these low temperatures. However, this situation is kinetically controlled, and once NOX is formed at the high coal-particle combustion temperature, there is insufficient residence time to shift the equilibrium as the flue gas cools. Our approach was to work with the fuel-rich portion of the process. If one increases the coal-to?air ratio until there is more coal than air, the flame temperature drops. ln the shaded region at the lower right of Figure 16, calcium sulfate ceases to be stable, but calcium sulfide is. In this zone, we can do combustion without heat removal and, further- more, retain the sulfur in the coal ash as calcium sulfide. That is the thermodynamics. As a matter of kinetics, it turns out that the calcium has to be in the right form: atomically dispersed through- out the coal as calcium humate. Fortunately, nature provides many coals with calcium in exactly that form. If the coal is deficient in calcium, we can add it in the appropriate form during the coal- washing pretreatment. Our concept is a real improvement over diffusion-limited, state-of- the?art systems, for example, low-temperature fluidized beds contain- ing limestone (Figure 17). When the coal and the calcium carbonate Figure 17. Conventional approach coal mixed with limestone. NEW SCIENTIFIC FHONTIERS 21 are in discrete particles, there are very serious diffusional limitations on the system. The sulfur released from the coal, usually as H25, must find its way to the limestone, where it can react to form solid calcium sulfide, before it is swept out of the bed with the fuel gas. By keeping the sulfur in the coal ash (Figure 18), we avoid this problem. 1; Coal} Air: ~(Fuel Rich) T. Figure 18. Exxon approach calcium atomically diSpersed within the coal. Another requirement, of course, is not to lose too much unburnt carbon in the ash. As you can see from the arrow in Figure 19, 70% sulfur capture has been achieved at a high carbon conversion of 90%. 4 1400heated so 1.3 Sulfur Retention so - in Ash 4? 30 20--l 10-4 30 20 1o 0 Original Carbon Remaining in Ash Figure 19. Sulfur retention versus carbon retention in ash. For NOX control, this same fuel-rich combustion scheme can con- vert the nitrogen compounds in the fuel to molecular nitrogen (Figure 20). In the laboratory, under carefully controlled combustion condi- tions, we have shown that 90% of the fuel?s nitrogen can be converted to N2. This result is consistent with thermodynamic calculations. Figure 20 shows our results compared to EPA requirements and thermodynamic 22 EXXON ENERGY HID SYMPOSIUM I 4>Fuel Rich Fuel Lean 10.00014 Exxon woo-l Experimental Data . current Level - Regulation -- --Proiected Future Limit N01. WG-1 1 mi Thermodynamic Limit I 1.0 2.0 3.0 Fuel/Air Ratio Figure 20. Strategy for NO, control: optimize temperature and fuel/air ratio. perfection. Note the narrow window of conditions where this result is possible. Although state-of?the-art furnaces burn fuel lean, and cur- rently meet EPA requirements, future specifications may be much more difficult to meet. There are drawbacks to our scheme. One cannot simply operate a combustor in a fuel-rich mode, because that would waste a considera- ble portion of the fuel?s heating value. instead, we do a staged combus- tion, with the first stage comprising operation in this region. Figure 21 shows the process concept. First, we gasify under fuel?rich conditions in a cyclone reactor, where sulfur, ash, and unreacted carbon are removed from the bottom. The hot, clean fuel gas is sent to a low-Btu First Stage: Gasltioation Second Stage: Combustion Steam cor H20. "2 ?5 Water Air cmuns??yopa. M, . a? Coal N, Hot.Low BTU Gas Gas Boiler Slag with Figure 21. Application of the Exxon concept for the direct utilization of coal. NEW SCIENTIFIC FFIONTIEFIS 23 gas boiler where the residual heating values are recovered. While this scheme increases system complexity and creates engineering prob- lems, it also creates opportunites. The design studies we have been doing indicate that a cyclone reactor based on this chemistry may be a very attractive way to retrofit existing gas- and oil-fired boilers and furnaces for coal firing. SOLAR ENERGY Let us turn, now, to some of the less conventional energy sources, which i shall typify by reviewing some of our work in the area of solar energy. State?of?the-art technology for photovoltaic devices has strongly emphasized ultrapure silicon, a material that is theoretically able to convert sunlight into electricity with an efficiency of about 16%. However, this material is inherently expensive, and foreseeable cost reductions do not seem likely to bring solar-electric costs into a range where large-scale use would be attractive. Our work has centered on amorphous silicon hydride, a substance that can, in our opinion, improve the cost-effectiveness of photovoltaic devices. While this semiconductor is inherently less efficient than crys- talline silicon in converting light to electricity, it is a much better light absorber. This means that thinner sections (less material per watt) can be used in solar cells, and also that the purity requirements are much less stringent. The interplay of these factors offers the potential of signif- icant cost reduction. Before going into detail on this, a brief overview of how semicon- ductors convert light into electricity may be helpful. With apologies to the solid-state physicists among the readers of this paper, shall use some analogies to give the sense of what is occurring. When a photon of light interacts with an electron in a molecule, the electron can absorb some of the light energy and be elevated to a higher energy state. A semiconductor is special in that these higher energy states can only have certain discrete values, and also in that it is somewhat difficult for the electron to return to its original energy level, releasing heat in the transition. Imagine the electron to be a ball being energized and propelled up a flight of irregular stairs from to as shown in Figure 22. Some of its original kinetic energy will be captured as potential energy, relative to its initial ground?state position. It can fall back to C, but it cannot be trapped at any level between and S. If we have some way of collect- ing the balls at and returning them to via some extraneous path, 24 EXXON ENERGY FMD SYMPOSIUM I Potential Energy Voltage Figure 22. Quantum energy levels in a pure, semiconductor. - then we can recover this stored potential energy as useful work. Thus, the semiconductor solar cell generates a discrete voltage differential that can be used to move electrons through an external circuit. An amorphous material is analogous to an irregular steplike slope (Figure 23). The ball may be propelled up in the same manner, but it can land at almost any elevation. Further, it is likely to roll down immediately to some lower level or back to its initial ground state. Therefore, because the energy is hard to store and collect, such mate- rials either cannot function as solar cells or do so with very poor efficiency. In Figure 24, we see a model of the lattice of pure silicon. The balls represent silicon atoms and the rods represent bonds to four neighboring Si atoms. The perfect regularity of this lattice yields a sharp, discrete voltage between photon-excited electrons and their ground state. In practice, a layer of this silicon is sandwiched between two very thin, transparent conductors. As men? tioned before, this layer has to be thick enough to absorb?rather than Figure 23. Energy levels in an amorphous solid. NEW SCIENTIFIC FRONTIERS 25 t; Figure 24. silicon. A reflect?most of the incident light, and the time required for the elec- trons to migrate to the collector electrodes must be small relative to their tendency to return to the ground state within the semiconductor. Amorphous silicon somewhat resembles silicon, but each silicon is not attached to four silicon neighbors. The many structural dislocations result in dangling bonds, so that this material is a very poor photovolatic converter. However, if we heal these dangling bonds via the introduction of hydrogen, we find (Figure 25) that the resulting silicon hydride (Sin) has many of the properties of its expen- sive relative, silicon. Returning to our analogy, this would correspond to a slope that is heavily, but not uniformly, terraced, like the vineyards on the Rhine. The nonsteplike portion corresponds to disordered states that degrade semiconductor performance, and we are learning to control these by careful attention to the route used to our silicon hydride. Our scientific contribution rests on the theoretical model that we developed, which seems to explain electron mobility in amorphous silicon hydride. We are able to infer the statistical distribution of the irregularities in the slope from measurements of the mobility of elec- trons at different temperatures. While some disorder is inevitable in an 26 EXXON ENERGY BSD SYMPOSIUM Figure 25. Amorphous silicon hydride. amorphous semiconductor, the measured disorder leads us to believe that efficiencies as high as 12?13% may be realized, versus 16% for silicon. Of course, this assumes that other defects, caused, for example, by trace impurities, can be eliminated. One route to the desired degree of property control is by manage- ment of the hydrogen species present during (Figure 26). Conventional Sin is made by decomposing silane vapor, in a glow discharge. The charged products, comprising fragments of this molecule, are deposited on a conducting surface to produce the semiconductor film. The ratio of Si to is essentially determined by the composition of the silane. We are using a sputtering technique in which an argon-ion plasma bombards a silicon target, dislodging silicon atoms that are drawn onto a substrate. Hydrogen is introduced extraneously, thus affording more degrees of freedom. The conventional wisdom had held that the high- energy electrons also associated with the argon plasma would likewise bombard the substrate, causing defects in the Sin film. However, we found process conditions where high-performance, high-density films can be produced. The plot in Figure 27 shows the concentration of defect states versus hydrogen partial pressure. You can see in the NEW SCIENTIFIC FRONTIERS 27 Glow Discharge Decomposition of Silane Charged Fragments Si Power Supply 1 s: I Substrate H- 1412-: Sputter Deposition 4 Target . Ar SI H2 Power Supply Substrate Figure 26. film fabrication. lower right that some of our sputtered films have fewer defects than the best glow-discharge?produced films reported in the scientific literature. In regard to the electron bombardment, we have actually found that it can improve our films via an operation analogous to annealing. We have been able to produce the world?s first Sin solar cell made entirely by sputtering. In the performance curve of current versus vol- Defect Density 1011 .1: Flange ol' Glow Discharge Results (Literature) Best to Date 0.2 0.4 0.6 0.8 1.0 1.2 1.4 Hydrogen Concentration Figure 27. Defect density. in sputtered films. - 28 EXXON ENERGY FMD SYMPOSIUM tage in Figure 28, we see that the cell yielded an open-circuit voltage of 0.82 and current densities up to 8 milliamps/cmz. The cell?s effi- ciency of 3% is still well below that available with the more expensive silicon cell, but, now that we have an idea of the scientific principles Normalized Current MA I 5? cw inns-madam Voc=0.821v I 1 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 Voltage Figure 28. Sputtered solar cell (efficiency involved, we forecast steady improvement. This will come from delving more deeply into these new thin-film semiconductors, using modern tools of materials characterization, defect identification, and control of growth chemistry. ENHANCED RAMAN SCATTERING My final example deals with a new aspect of light scattering that has potential for studying the dynamics of interfaces, and that may also lead to novel applications in the collection and use of light energy. However, allow me to digress for a moment to set the stage. Consider a single molecule, represented in Figure 29 by a ball-and- Spring model. When a photon of light strikes the molecule, there are in general only two possible outcomes that result from scattering of the photon. First, the photon could be scattered elastically by a process known as Rayleigh scattering. No transfer of energy from the photon to the molecule results in this case. Secondly, the photon could be scattered inelastically. One process involving inelastic scattering is Raman scattering. In this scattering, there is transfer of energy, the energies of the molecule and photon being altered in an energy-conserving manner. The energy of the scat- tered photons provides a direct measure of the allowed quantum- energy states of the molecule. Thus, Raman scattering from molecules, NEW SCIENTIFIC FRONTIEHS 29 Incident Photon Scattered Photon II . l' .l.l.ilV.I.l.l t_i_ Rayleigh Scattering Elastic Scattering- No Energy Transferred to the Molecule . MoIecule Incident Photon . Scattered Photon Raman Scattering Inelastic Scattering- Energy is Transterred . Molecule from the Photon to the Molecule Figure 29. Light scattering. as on a solid surface or in a liquid, provides an excellent way to identify and to study the molecular species. These features of light scattering are demonstrated in Figure 30, which shows a Raman spectrum of pyridine. Note the characteris- tically sharp lines in the spectrum of this liquid that allow us to finger- print the pyridine and to study its vibrational-energy states. The I I 2000 1500 1000 500 0 Energy Transferred to the Molecule Figure 30. Raman scattering. Liquid pyridine spectrum (10? molecules). 30 EXXON ENERGY nan SYMPOSIUM number 1018 in the upper-right corner means that we need 1018 molecules in the sample in order to get a signal strength high enough to yield a spectrum with easily recognizable peaks. In other words, Raman absorption by molecules is normally not very efficient; At the far right, i also show the elastic, or Rayleigh peak that is always many orders of magnitude more intense than the very weak, inelastically scattered light. Now it happens that there is a remarkable enhancement of the Raman scattering from molecules in the vicinity of certain surfaces. This effect was first observed a few years ago by British scientists. I shall develop this concept by looking at a Raman spectrum ofa smooth versus a rough silver surface. The smooth surface (Figure 31) exhibits no Raman spectrum. All that we see is background noise and the Wk? I I I I I 2000 1 500 1000 500 0 Energy Transferred to the Molecule Figure 31. Normal Raman scattering (from smooth surface). Rayleigh peak of the elastically scattered light. When a specially roughened silver surface is employed (Figure 32), however, we see two strange peaks to either side of the Rayleigh peak. More about this later. Now, if we absorb a monolayer of pyridine on this rough silver surface, we see (Figure 33) the Raman spectrum that we saw previ? ously. However, it is remarkable that we are able to detect any signal at all, as the number of moleCUles irradiated on the surface is only about 10?, compared with 1018 in the liquid sample shown earlier. On an ordinary surface this spectrum would not be visible over background noise. We see, therefore, that there is an unexpected, million-fold enhancement in the Raman scattering on the silver surface. NEW SCIENTIFIC FHONTIERS 31 0-1 I 2000 1500 1000 500 Energy Transferred to the Molecule Figure 32. Surface-enhanced Raman scattering (from a rough silver surface). Scientists in our laboratory were led to an understanding of the enhancement that I have described by an unusual experimental obser? vation. By looking very closely at the elastically scattered light, they discovered the new, very intense Raman line that I mentioned earlier. The behavior of this line was unlike that ever before seen in Raman scattering. For example, as the exciting frequency of the laser light Ml I I I 2000 1500 1000 500 0 Energy Transferred to the Molecule I Figure 33. Surface-enhanced Raman scattering (pyridine on a rough silver surface, 1012 molecules). 32 EXXON ENERGY SYMPOSIUM source was increased, this peak in the Raman scattering moved further away from the elastic peak. We determined that this Raman scattering must be due to vibrations of the rough metal surface itself, and we developed a model that describes our observations. We have found that a key to this effect is the roughness of the silver surface. To understand its role, let us first consider the unusual interac- tion that light can have with an isolated sphere of silver. At certain frequencies, this sphere will be resonantly excited by the light. This is analogous to your TV antenna, which can detect certain frequencies very sensitively if it has the right size and shape. Although the signal reaching the TV antenna is extremely weak, it will excite an antenna circuit of appropriate size into producing a signal strong enough for your TV set to amplify. Similarly, a specific light energy is strongly absorbed by a unique size of silver sphere characteristic of that light (Figure 34). For visible light, we need a surface roughness that yields bumps or spheroids in Light Spheroid Light Silver Figure 34. Submicrosc0pic roughness plays a critical role. the 100A size range. This strong absorption occurs because the elec- trons in the silver spheroid respond collectively to the incident light field by setting up a large oscillating charge on the spheroid. This explains why the Raman enhancement is seen on a silver surface, which is an excellent conductor of electricity one with high elec- tron mobility), and not on semiconductors or insulators. When the correct frequency of light impinges on the silver spheroid and the charge oscillation occurs, a very strong electric field is gener- ated on the spheroid. This increases the intensity of the light at the surface, and this, in turn, causes an enormous increase in the light scattering from any molecules at or near the surface. In a sense, this NEW SCIENTIFIC FHONTIERS 33 can be thought of as storing the light energy in the silver and using this stored energy to scatter from the surface moleCules. The behavior of the oscillating charges on the silver surface follows the physical laws of acoustics. In fact, a useful analogy involves the oscillatory behavior of a tuning fork. If the fork is stimulated by con- tinuous mechanical vibrations of a given frequency and amplitude, it can be made to emit higher-amplitude resonant beats of a frequency that is characteristic of the forks dimensions. Of course, the average energy of the emitted sonic beats must be less than the average energy of the stimulating energy, because some of the latter is invariably dissipated as heat. We envision the rough silver surface as being composed of a ran- dom array of small bumps of different shapes and sizes (Figure 35). For any incoming light, there will be some subset of bumps that will act as antennas to amplify the light scattering, thereby giving the observed enhancement. Laser Field Figure 35. A rough metal surface has bumps of many shapes has antennas at all visible frequencres. The unusual optical properties of these surfaces makes them very interesting to study, and perhaps quite useful. The ability to detect a monolayer of materials would be very attractive in any studies of pro- cesses that occur at surfaces, such as, for example, catalysis and corro- sion. In addition, the enhancement of optical processes suggests that there may be ways to harness this unusual phenomena for collecting light energy. EPILOGUE This brings me to the end of my story. I have attempted, here, to describe some of Exxon?s research in areas of energy-related science. All of the examples cited involve new science, generated as a result of 34 EXXON ENERGY SYMPOSIUM a multidisciplinary effort and carried out with an awareness of the broad range of future energy needs of the entire industry. I emphasize that we are part of a total system. Why is this so essen- tial? For one thing, the conventional wisdom says that it takes 8 to 20 years to convert a scientific discovery into an invention and then into an innovation that people can use. We cannot wait that long. The only way that I know to cut the time is to carry out the scientific work in a freely communicating, dynamic, total system. I also emphasize the multidisciplinary character of our work. This is much more than a catch word. it is essential to our future. Our technological systems are chemically and physically complex. Often, they are highly interacting. Their complexity is often beyond our scien- tific capabilities. Accordingly, we must create the new science?the new chemistry, new characterization tools, new materials, and new models?needed for the complex systems involved. This will require bringing together the awesome capabilities of modern science to un- ravel some truly difficult puzzles. The lone specialist looking through his own microscope will have limited effect. The challenge for science has never been greater. ABSTRACT Yet another in a series of energy transitions is underway, this time from fossil to renewable sources. This paper dis- cusses Exxon?s on the fuels needed to bridge the transition from petroleum to the renewables. Overviews are presented of Exxon?s major development projects on coal liquefaction and coal gasification, as well as longer-range research on fuels. Recent guid- ance studies that examined the outlook for various tic fuel technologies are summarized. Finally, the pros- pects for renewable energy sources are considered briefly. ABOUT THE AUTHOR Frank B. Sprow is Vice President for Fuels Re- search at Exxon Research and Engineering Company. He received his BS. and M.S. degrees in chemical engineering from MIT and his in chemical engineering from the University of California at Berkeley. He began his career with Exxon in November 1965 at Baytown, Texas, where he worked in various engineering and supervisory posi- tions. In May 1971, Dr. Sprow joined Exxon Company, Supply Department in Houston, and later became responsible for negotiations for purchase and sale of Ex- xon?s U.S. crude oil supplies. He then became Technical Manager of Exxon, Bayway, New Jersey, Refinery in August 1975 and was named Operations Manager in August 1977. Dr. Sprow joined Exxon Research and En- gineering in March 1979 as General Manager of Petroleum Programs. He assumed his present position in January 1980. Dr. Sprow belongs to the American Institute of Chem? ical Engineers and the Society of Automotive Engineers. F. B. Sprow NEW SOURCES AND THEIR TECHNOLOGICAL PROSPECTS One of the important factors that has shaped the industrialization of the modern world has been the availability of inexpensive, convenient, liquid and gaseous fuels from petroleum and natural gas. As a result, these fuels play a crucial and strategic role. Now, however, we are being forced to confront the fact that the resource base for oil and gas is finite. There must ultimately be a transition to renewable or essentially inexhaustible sources of energy such as solar and fusion. An important part of the bridge for that transition will, we believe, be fuels. By fuels I mean liquid and gaseous fuels made from non- petroleum sources such as oil shale, coal, tar sands, and biomass, fuels that can function as replacements for those derived from petroleum. in this paper i shall be describing Exxon?s research and development work on fuels from these new sources and something about their technological prospects. i shall start by discussing what we mean by an integrated approach to process development. My purpose is to provide you with a framework for viewing some snapshots of our fuels that are at different stages in the development process. 1 shall then discuss our development programs in both coal liquefaction and coal gasification. This, I hope, will give you a current, slice?in?time look at where we are and what we are doing with these process de- velopments. Then, i shall briefly discuss a recent research-guidance study. We have used this guidance study to shape our longer-range research program, which I shall illustrate with an example of some current work. Finally, 1 shall close by considering renewable sources and how they relate to the outlook for PROCESS DEVELOPMENT WORK Let us first look at process development work as we practice it in Exxon. The process of process development is conventionally broken 37 38 EXXON ENERGY SYMPOSIUM down into activities that are distinguished from one another by their goals. In Figure l, i have listed the usual labels applied to these ac- tivities and to their associated goals. Guidance studies suggest areas in which to work and directions for the research to take. They also pro- vide evaluations of progress. Longer-range research produces and de- fines new ideas that are then moved into applied research to determine their technical feasibility. If the idea then appears to be technically Activities Goals Research Guidance New Ideas Technical Feasibility Commercial Feasibility Data for Design Guidance Studies Longer Range Research Applied Research Development Scale-Up Planning/Design ConstructionlStart-Up Operation Profitable Operating Plant Figure 1. Process development activities and goals. feasible, it is moved into development. Development is really a matter of creating commercial feasibility by identifying and solving problems. if and when a process can be developed to the point where it appears to be commercially feasible, it can be advanced to scaleup. Here, the goal is data for the design of a commercial plant. The final goal, of course, is a safe and profitable operating plant. The time scale for process development can typically span decades. The time scale shown in Figure 2 is meant only to be illustrative, Activities Timing of Activities Guidance Studies Longer Range Research Applied Research Development Scale-Up Planning/Design Construction/Start-Up Operation there trom Start at Deveiopment Figure 2. Time scale for process development activities. '1 NEW 39 because each development is unique. By and large, however, the more complicated the technology and the larger the development facilities required, the longer it takes from initial idea to a commercial operating plant. Unfortunately, fuels technologies are among the more complicated. Within Exxon, research guidance and longer-range re- search are continuing activities. Applied research begins with the birth of an idea, but it often continues in one form or another well into the operating life of a new technology. An important aspect of each activity is the scale at which it is prac- ticed (Figure 3a). In the early phases, only information or a few grams of material are involved. Scale then escalates rapidly in the develop- ment and scale-up phases to thousands of tons per day in the commer- cial phase. This rapidly escalating physical scale results in a similar rapid escalation in the costs associated with each succeeding phase, from thousands through millions to billions of dollars (Figure 3b). The number of people also escalates rapidly (Figure 3c) and peaks during construction at several thousand people per plant. The different activities call for different kinds of skills and backgrounds. The key to successful development is the ability to focus and to integrate the skills and expertise of these diverse groups over an extended period of time. Obviously, information must flow from group to group as the de- velopment proceeds, but it is equally important that feedback flow in the other direction (Figure 4a). A critical aspect of good integration is the active involvement of the engineers and the research laboratories in the research-guidance studies (Figure 4b). Equally important is the active involvement of both the engineers and the operating company in the scale-up phase (Figure 4c), because it brings practical commer- cial experience to bear on the development. Finally, of course, the results of the guidance studies are fed back to focus and guide the other activities (Figure 4d}. The point of all this communication is to ensure that the resultant technology is safe, operable, and economic. EXXON DONOR SOLVENT PROCESS With this as framework, then, let me begin by giving you a view of the current status of our work on the Exxon Donor Solvent coal liquefaction process. This project is in the final, scale-up phase of development, which is currently planned to extend to 1984. The goal of this phase is to bring the technology to a state of commercial readi- ness, such that a commercial, pioneer plant can be built at an accepta- ble level of risk. 40 EXXON ENERGY SYMPOSIUM Activities Scale of Operations Guidance Studies Longer Range Research Applied Research Development Scale-Up Planning/Design ConstructionlStart-Up Operation I 10: Activities 10? Tons/Day [bi Scale of Costs Guidance Studies Longer Range Research Applied Research Development Scale-Up Planning/Design Construction/Start?Up Operation 15: ?is ?is Cumulative Costs 5 [Cl Activities Scale of People involved Guidance Studies Longer Range Research Applied Research Development Scale-Up Planning/Design Construction/Start-Up Operation 10? 102 16: 164 People Figure 3. Scales for process development activities: operations, costs, and people involved. NEW 41 [81 Activities Organizations Guidance Studies Staff Longer Range Research lie Basic Research Lab Applied Research Applied Research Lab Development Scale-Up Planning?Design Engineean Construction/Start-Up Engineering 5. Operating 00. Operation Operating Co. Activities Organizations IA Guidance Studies 1 Stett Longr Range Research Basic Research Applied Research pp esearch Lab Development Scale-Up Ptannin .?Desi Engineerin Construction/Start-Up Engineering 3. Operating Co. Operation Operating Co. . . . Activmes Organizations Guidance Studies Staff Longer Range Research Basic Research Lab Applied Research Applied Research Lab Development (Scale-Up} PlanningJDesign Engineering ConstructioniStart-Up En ineering 3. Operating Co. Operation Operating 0. Id) Organizations Guidance Studi@\ Staff Longer Range Researcth I Basic Research Lab Applied Research I Applied Research Lab Activities Development Scale-Up Planning/Design Engineering Construction/Start-Up Engineering 5 Operating Co. Operation Operating Co. Figure 4. Organizational information flow in process devel- opment: feedback at all stages, invoivements in research guidance studies, involvements in scaleup, and Use of guidance studies. 42 EXXON ENERGY SYMPOSIUM Research on this process started in the mid-1960s and proceeded under Exxon sponsorship until 1976. At that time, the Exxon Donor Solvent project was formed, a $367 million joint undertaking of the U.S. Government and an international consortium of private com- panies. Participants include the U.S. Department of Energy; Exxon Company, the Electric Power Research Institute; the Japan Coal Liquefaction Development Phillips; Atlantic Richfield; the German firm Ruhrkohle; and the Italian company AGIP. All partici- pants, including the U.S. Government, share in the guidance of the project and in any eventual rewards in proportion to their financial support. Exxon Research and Engineering Company is the program manager. Until now, costs have been split 50/50 between the Gov? ernment and industry, but the Government?s share is being reduced somewhat as part of the Reagan administration?s drive to cut Gov- ernmental spending. The main focus of our current effort is the large pilot plant. It is designed to convert 250 tons of coal per day into 400 to 500 barrels of liquids per day. However, its primary product is not oil but data that can make possible the design of a viable commercial-size plant. The pilot plant cost $1 1 8 million to build, and it costs nearly $3 million per month to operate. design was based on a careful engineering analysis of the major technical uncertainties that will face the designers of a commercial plant. We believe that it is the minimum size that will still permit us to duplicate and to test actual commercial practices. Coal was introduced into the unit on June 24, 1980. From then until the middle of October, when the plant was shut down for inspection and turnaround, we were dealing with mechanical problems rather than process concerns. The key to successful operations during this period was avoiding plugging in the process lines, caused either by the solids in the slurries or by solidification of heavy tars. The plant is unforgiving in this respect, so that its service factor?the percentage of time that the plant is feeding coal?is a strong function of the time required to unplug the equipment after a coal outage. in spite of this, we achieved our target of a 50% service factor in the early, shake- down phase. Coal feeding was resumed December 30, 1980. We have now ac- cumulated a total of over 3,000 hours on coal. The longest continuous run has been 31 days, and the service factor for the period since restart has risen to over 65%. This reflects the increasing level of experience and skill of the operating staff, which are among the important ac- complishments thus far obtained. NEW 43 The heart of the scale?up activity is a comprehensive test program that was planned by our technical specialists and design engineers. I shall use the flow plan of the pilot plant shown in Figure 5 to describe the major elements of the test program and their status. Coal if Solvent Hydrogenation Gas Swept Ml" th. I J, I :58; Naptha Distillate V, Li ue? ln?iion a Slurry "new" rJ 7 Drler A VALVE SIurryI Pre- I . Heater. 50L.le Bottoms ?2 I H: WITHDRAWAL Processing Fuel Figure 5. EDS pilot-plant flow plan and test program. The gas-swept mill both crushes and dries the coal. We have dem- onstrated the operability of the process on coal ground to particles of 3 millimeters and smaller and dried to 4% moisture in the gas-swept mill. We also have another drying system to test, consisting of a mill to reduce the wet coal to the same size range, followed by drying in the slurry mix tank. That test has been deferred to later in the program. Satisfactory operation of the pumps for the slurry feed has been demonstrated. Erosive wear was a prime concern, because the pumps operate on a solid-liquid slurry at high pressures. While the slurry-feed pumps performed well, we were surprised to experience very short operating life of the packing in identical pumps that pump the bottoms from the atmospheric distillation tower into the vacuum distillation tower. These pumps operate at lower pressures but higher temperatures. This situation has now been much improved by a combination of new packing materials and better maintenance. While we can live with present performance in the pilot plant, it is not good enough for a commercial plant. Therefore, we shall be investigating changes in pump design, seeking even longer packing life. Low?rank coals can form solids in the reactor because of their high calcium content. Later in the program, when we run Wyoming coal, we plan to test methods for withdrawing. these solids. 44 EXXON ENERGY SYMPOSIUM The rate of coke formation in the preheater furnace for the slurry was of concern to the design engineers. A major goal is to learn how to achieve acceptable coking rates with the kind of furnace coil designs that could be used on a commercial scale. At this point, our experience has been mixed. We are presently working to understand the causes for some sporadic episodes of coking in the slurry preheater furnace. I We are also concerned about coke formation in the preheater fur- nace for the distillation towers and in the flash zones of the towers themselves. Tests for coking at the design temperatures have been completed successfully. We plan to test the limits of operability later in the program. Early in the development, before construction of the large pilot plant, researchers identified the pressure?letdown valve leading from Valve Inlet Plug Figure 6. Pressure letdown valve. the liquefaction reactor to the distillation section as a potentially seri- ous weak point. After coal liquefaction at pressures of over 400 atmo- spheres, the next processing step, distillation, operates at near atmo? spheric pressure. The valve is called upon to release this pressure and at the same time control the rate of flow of material from the liquefac- tion reactor. Figure 6 shows a drawing of one of two designs under test in the pilot plant. The hot, abrasive mixture passes through the valve at the speed of sound. No previously known letdown valve could handle those conditions, so one of our engineers designed this valve. The cladding of tungsten carbide steel that is used in this valve is standard in the valves that broke down in the same service in other pilot plants for coal liquefaction. The key, here, is the special interior design that streamlines the flow of materials through it. The result has been a valve NEW 45 that has shown very little wear in the more than 3,000 hours that the plant has now operated. An interesting aspect of this story is that the computer played little or no role in it. The inventor, an engineer named Bob Platt, was given computer-generated data about the composition and behavior of the materials coming out of the liquefaction reactor. However, he recog- nized its unreliability, and he drew instead on his experience and creativity. On the other hand, he now thinks that a program can be created to help to design scaled-up valves such as the one he invented. However, we were not quite that smart in all areas subject to ero? sion. Figure 7 shows a drawing ofan elbow in the transfer line between the preheater furnace for the vacuum distillation tower and the tower Elbow Eroded Elbow 74? Figure 7. Erosion in the elbow of a transfer line. itself. A hole eroded through the 11/4-inch thickness of the metal in this elbow in just 21 days of operation. The original elbow has been re- placed with one having a smoothly contoured bend, lined with a special refractory material that is metal-fiber reinforced. This lining material has performed very well in test sections in other lines in the plant, and it is now performing well in the transfer line. Another major aspect of the test program is comprehensive materials testing. This program goes well beyond the usual corrosion racks and test coupons. It includes studying specific equipment components, using online monitoring and stream analyses to correlate corrosion rates with operating conditions. Before we leave reference to Figure 5, let me point out that this pilot plant contains no bottoms-proCessing facilities?the bottoms are so- lidified, collected, and stored. However, 30?50% of the carbon fed to 46 EXXON ENERGY SYMPOSIUM the process ends up in the bottoms. To be economic, the process must use the energy value of this carbon. Present project plans do not call for development of a particular bottoms process. But studies are underway to define the work needed to demonstrate both gasification and burning of bottoms and to define incentives for such an effort. Applied research and development have been continuing in parallel with the scale?up activities, and one result has been a substantial im- provement in the process. This improvement involves recycling some of the bottoms back to the liquefaction reactor. While this sounds simple, a considerable amount of research was necessary to define the optimum conditions of pressure, temperature, residence times, the ratio of feed to solvent, and the ratio of recycled bottoms to solvent. Figure 8 shows the increase in liquid yields for the improved process 50/ Base Process 50/ Improved Process 40 Liquid . Yield, wt 30 willlli'? ll .j l ll? 'll Subbituminous Bituminous Figure 8. Process improvements from continuing yields from different types of coal. for a range of coals of different types. However, even more important is the effect on the product slate. Figure 9 shows the distribution of the total product over the boiling range from methane to fuel oil for the base process, while Figure 10 shows the distribution of products for the improved process. The effect of the improved process is to shift product from the boiling range for fuel oil into that for gas and naphtha On the basis of our present plans, we expect by late 1982 to have the data and information from the 250?ton-per-day pilot plant that are necessary to permit the design of the liquefaction section of a commer- cial plant. Planning studies are now underway to screen the Exxon Donor Solvent process along with other fuels technologies for possible Exxon commercial projects. NEW 47 50? 40Boiling Point Figure 9. Product slate from base process. wt c1 c1 c5 Boiling Point Figure 10. Product slate from improved process. COAL GASIFICATION Let us turn, next, to our work on coal gasification. Here, we have a novel catalytic gasification process in the development phase in a one-ton-per-day pilot plant at our laboratories in Baytown, Texas. Be- fore I discuss some of our work in this unit, however, let me first describe the process and how it differs from conventional technology. Current technology for similar processes involves three steps in three separate reactors (Figure 1 1). In the first, coal is reacted with steam to form carbon monoxide and hydrogen. That mixture is reacted with more steam to convert some of the carbon monoxide to more hydro- gen (the water?gas shift reaction), and the resulting mixture gas) is then reacted in a third reactor to form natural gas. The first reaction soaks up heat, which is usually supplied by injecting oxygen into the reactor to_burn some of the coal. How? ever, heat is released in the next two steps?about the same total 48 EXXON ENERGY nan SYMPOSIUM Coal Steam Steam co+3H2 Methane l. Gasification 7\ A. ?20 If my 54m. i cram . awestr' Figure 11. Conventional coal gasification processes for methane. amount as the heat absorbed in the first step?but it is recovered at a lower temperature. This represents a thermodynamic inefficiency. In our catalytic process for coal gasification (Figure 12), all three reactions are carried out in the same reactor. The catalyst lowers the gasification temperature and also promotes the reaction of carbon monoxide and hydrogen to form methane. Thus, a mixture of gases is produced that is rich in methane. The carbon dioxide is removed, and the methane is separated and can be sent to a pipeline. The remaining carbon monoxide and hydrogen are recycled back to the reactor. As a result, the net products from the reactor are methane and carbon dioxide. This forces the gasification and methanation reactions into balance, and it means that we can capture the heat of the methanation reaction and use it to drive the gasification reaction. In effect, when we recycle the hydrogen and carbon monoxide, we recycle the heat of the methanation reaction back to the gasification reaction. Catalyst Coal Steam CO, Gasitication - Methana?on CO H2 CO2 Separation ?)Methane Hze?? Figure 12. Catalytic coal gasification process. NEW 49 Thus, the catalytic process is a more efficient route to natural gas than current technology, both as a result of recycling the heat of reaction and the lower temperature of operation. Lower operat? ing temperatures also ease material problems. There is no need for an oxygen plant, and the slagging problems that can attend oxygen injec- tion are eliminated. The presence of the catalyst reduces the process?s sensitivity to coal properties, and it also eliminates the troublesome tars that are formed in some processes. However, everything has its price, and in this case the price is the need to recover catalyst and to add make-up catalyst to replace that which is not recovered. More-stringent requirements for gas cleanup are imposed by the need to separate the methane from the recycle carbon monoxide and hydrogen. Finally, there are some concerns that the solids residue from the process might pose additional disposal problems over those normally associated with coal-ash disposal be- cause of the presence of residual catalyst and carbon. We believe these disadvantages are more than outweighed by the advantages. In the development phase, there are usually two main goals. The first is simply to demonstrate that the process will operate. We experienced some problems of a mechanical nature in the startup of our one-ton- per-day unit, but they were quickly solved. At this point, we can say that we have been able to operate both the gasification and catalyst recovery sections of the plant successfully. However, it is not enough simply to operate the process. It is critical that the process be made to operate economically. This involves identifying potential problems and potential improvements. Problems are always encountered when a process is moved out of small, laboratory equipment and into a pilot plant. The earlier they can be found and solved, the better. We have identified three problems: low density ofthe fluid bed, low rates for the methanation reaction, and production of fines. We have also identified potential improvements: a lower-cost scheme for coal drying and steam generation, and staged gasification for better utilization of coal. I now propose to use one of these problems, that of the low density of the fluid bed, as an example of the kind of activities that are typical at this stage of a process development. When we started up our one-ton-per-day, process-development unit, we found that the density of the bed was only about one-third of that expected. Samples from the bed showed that the char occupies a larger volume than does the coal that was fed into the bed. It appears that when a coal particle enters the bed and begins to heat up, it partially melts and becomes plastic?. As it heats up further, gases 50 EXXON ENERGY RGID SYMPOSIUM are evolved that enlarge the particle into a porous, solid foam. This is akin to blowing plastics to make foams or to popping popcorn. Foamed-up particles are more fragile than compact particles and there- fore tend to produce more fines. Our first step was to develop a laboratory test with which we could study the problem (Figure 13). This test consists of sealing a small amount of coal in a quartz tube about 4 millimeters in diameter and about 8-centimeters long. The tube is heated very quickly to about held there for a few seconds, and then cooled. The swelling index is defined as the ratio of the final height to the initial height. The test correlates fairly well with actual performance of coal in the one- ton-per?day, process-development unit. Using this test, we have found that one of the factors contributing to the problem is pressure. Our early work was done in a smaller unit that was capable of operating at only 7 atmospheres. As you can see in Figure 14, our preferred pressure of 35 atmo- spheres seems to produce the worst swelling. We have developed a couple of approaches for controlling this problem, the results of one of which is shown at the bottom of Figure 14. This involves pretreating the coal; as you can see, it yields an acceptable swelling index over the pressure range of interest. We are also investigating some other possi- ble solutions, looking for the most economic one. The last step in the development of the catalytic process for coal gasification will take place adjacent to the E550 refinery in Rotterdam. We are now carrying out the planning studies for a 100-ton-per-day pilot plant to be constructed there by a group of Exxon affiliates. De? sign should begin in 1982, with completion of the pilot plant sched- aoow Final Height Inltlal Height Fluld Bed 200- I I I I 0.90 0.95 1.00 1.05 1.10 1.15 Swelling Index Initial Height Figure 13. Laboratory swelling test predicts fluid bed density. at? NEW 51 Swelling Index Preheated I 50 100 Pressure, Atm Figure 1'4. Swelling in untreated coal is strongly dependent 0n pressure. uled for late 1985. We expect to operate this pilot plant for three to three?and-one?half years. Thus, the earliest that the design of a com- mercial plant could be started is the late 19805. Such a plant would come onstream in the early 19905. RESEARCH GUIDANCE There are, of course, other approaches to converting coal to tic fuels, and other resources such as oil shale and biomass. In order properly to focus and to provide direction to our research program for fuels, we look at the prospects for a variety of fuels technologies in our research-guidance studies. We formed a team of experts from our various laboratory and en- gineering units. In consultation with outside experts, this team first assessed the relevant scientific trends, and from those trends it pro- jected specific technical advances in the area of fuels technology. These projected advances were then used to project economic potentials. Finally, economic comparisons of the various alternatives were made, based on a systems analysis that considered costs from the resource to the end use. We concluded that the potential exists to improve the efficiency, to reduce the investment, and to lower the costs of all the fuels?some, of course, more than others. We also concluded that a range of technologies will likely be required to allow optimum utiliza- tion of a broad variety of worldwide resources for fuels. Based on these guidance studies, we have reoriented our longer- range research program on fuels. We are continuing our 52 EXXON ENERGY SYMPOSIUM longer-range research in the areas of direct liquefaction of coal and gasification of coal, because there appears to be substantial room for improvement of current processes. Several technologies now exist for recovering oil from oil shale, and these technologies are being applied to the first wave of commercial plants. We have broadened our longer-range research program to explore for new and more efficient approaches to shale oil conversion and processing. Technologies also exist for the of liquid fuels from the carbon monoxide and hydrogen available from gasifying coal. Here too, opportunities for improvement justify further exploration. Finally, we are exploring the nature of coal and oil shale for clues to better approaches. LONGER-RANGE RESEARCH To illustrate our longer-range research, I shall now discuss some recent work in the area of the direct liquefaction of coal. Earlier in these Proceedings, Dr. Lucchesi has pointed out the complex nature of coal. Here is another look at the kind of molecular structures present in coal, focusing on the carbon-carbon bonds. These structures (Figure 15) are hooked together into large molecules that have high melting points. Direct liquefaction involves breaking bonds in these structures to create smaller molecular fragments that are liquids instead of solids. Figure 15. Carbon-carbon bonds in the molecular structure for coal. We have been attempting to understand the mechanism of the reac- tions in which these bonds break. The strength of the different bonds shown here varies greatly. For simplicity, I shall focus on two kinds of if? NEW 53 bonds, a relatively weak one and a relatively strong one. I shall also simplify this diagram further (Figure 16) to show just these two bonds, representing the groups attached to them by circles. The weak bonds can be broken simply by raising the temperature. When the bond breaks, a pair of free radicals is formed, and these react either with Bond Breaking with Heat Gaga?[Q .. Bond Breaking with Hydrogen Atoms Figure 16. Carbon-carbon bond breaking. hydrogen or with a hydrogen-donor molecule to pick up hydrogen. in the process, some hydrogen atoms, represented by are formed. One of our scientists, Dr. Lonnie Vernon, has been able to show that these hydrogen atoms, which are free radicals themselves, can attack those bonds that are too strong to be broken by temperature alone. This hydrogen-atom attack yields a smaller molecule with a hydrogen at? tached and a free radical that can react with hydrogen to form another small molecule and a hydrogen atom. This understanding has been helpful in guiding our efforts to optimize the Exxon Donor Solvent coal liquefaction process. It has also served to suggest new directions in our continued exploratory research on coal liquefaction. How do the prospects for renewable sources mesh with those for the fuels that have been discussing? Solar energy, of course, is the classic?renewed again each morning. As Dr. Lucchesi has shown you, both our research and that of others on solar photovoltaics is progressing. Since photovoltaics are built on semiconductor technol? ogy, their future seems more promising, and more open ended, than the rather conventional technology of solar thermal collectors. We expect that research will meet many, if not all, of the challenges re- maining to solar-electric technology?:high cost, low efficiencies, 54 EXXON ENERGY RED SYMPOSIUM low-cost energy storage, and integration with low-cost, back-up sys- tems when the sun does not shine. For baseload electrical generation, fusion energy may ultimately fit our energy needs best. Recently, fusion has begun to look more and more promising, but we must not fail to appreciate the magnitude of the job ahead for it. Once all the science was in hand, it took some 30 years to engineer commercially viable fission reactors and to make them a significant source of power. Serious engineering for fu- sion power is just now beginning, and it is certain to encounter major engineering and commercialization problems, including many that we cannot now imagine. Fusion is not a way out for'the near or inter- mediate term. lt is not a substitute for the expanded use of available technologies based on coal and fission reactors, or for the need to pursue on other energy sources with more immediate promise, such as the breeder reactor. But solar, coal, fission, and fusion produce heat and electricity, not the liquid fuels that are our most critical problem. Some people think we can side-step the problem by shifting to electric vehicles in a major way, as early as this decade. In our opinion, the superior ability of chemical fuels to store energy means there will be a place for liquid fuels in transportation for the foreseeable future. And this brings me full circle to where fuels. fuels offer the greatest promise as a supplementary source of transportation fuels, and clearly their time has come. U.S. coal and oil-shale resources are huge. The estimated total-resource base in the .S. for is over 13 times the size of the .8. resource base for oil and gas. Development of these supplies would not only relieve U.S. pressure on the scarce world supplies of petroleum, it would also promise, eventually, to clamp a lid on energy prices. We believe that existing technology makes shale oil and gas from coal competitive with OPEC oil. Furthermore, it appears likely that, before too many years, advances in technology will make liquid fuels from coal another practical, economic alternative to imponed oil. As I have discussed, we have active development programs undenivay on coal liquefaction and gasification, and we are pursuing a broad range of etic fuel technologies in our longer-range programs. ABSTRACT Advances in four major areas of technology are improving our capability to find and to produce petroleum, coal, and minerals. Multispectral satellite imagery enhanced by computer processing is being used to identify surface ex- pressions of important subsurface structures, to help to identify drainage patterns, formation contacts, rock types, mineral deposits, coal or oil?shale outcrops, and hydrocar? bon seeps. Systematic analysis of global. changes in sea level allows us to determine the ages and compositions of rock strata before the first test hole is drilled. A guyed? tower system permits the use of conventional platform dril? ling and production technology in water depths up to 600 meters, and subsea production systems are being en- gineered for even deeper waters. Enhanced oil recovery methods, including chemical flooding, miscible gas, and thermal processes, are also reviewed. ABOUT THE AUTHOR L. William Welch, lr., is President of Exxon Production Research Company. He returned to EPR in Houston in July 1975 after serving in London for five years as Vice Presi? dent of Esso Europe, Inc., and President of Esso Exploration and Production (U.K.), Inc. Mr. Welch received a BS in petroleum engineering in 1943 from the University of Tulsa. He joined the Exxon organization in Seminole, Ok- lahoma, at that time. In 1944 he took a military leave from Exxon and served in the US Navy as Lt. until 1946. He has worked for Exxon in Oklahoma; Kansas; Louisiana; Alberta, Canada; London, England; and Texas, holding positions in production, planning, and research. Mr. Welch is a member of the Society of Petroleum Engineers, the American Petroleum Institute, and the Industrial Re- search Institute, and is a Vice Chairman of the US Na- tional Committee for the World Petroleum Congress. L. W. Welch, Jr. ADVANCES IN EXPLORATION AND PRODUCTION TECHNOLOGY It is always challenging to discuss the latest developments in the technology of petroleum exploration and production. Here, I shall begin by describing how we obtain information about the surface of the earth from satellites, and how we use this information to look for oil, gas, and minerals. Then, I shall tell you how geologists are turning the geologic record of global changes in sea level into valuable predril- ling information. I shall follow this with a discussion of our ongoing program to develop offshore-production capabilities in deeper water and more hostile areas, such as the northern and western parts of the North Sea. Finally, I shall discuss enhanced oil recovery and, I hope, shall give you an idea of the contribution various techniques might be expected to make to increase the world?s supply of crude oil. SATELLITE IMAGERY Satellite imagery is a topic that we can all appreciate, in light of the current success of the space shuttle and heightened interest in using space tools to solve down?to-earth problems. Study of the surface out- crops of formations (Figure I) has long been a standard exploration tool of gelogists. From the surface?s geology they make extrapolations into the subsurface, where oil and gas are found. For example, geologists look for areas where an impermeable cap rock overlays a permeable reservoir that, in turn, is in communication with an organic-rich source rock. In such situations, oil and gas fields may exist in subsurface structures or other traps. In the 19205, geologists found many major structural features on the surface by using aerial photographs, such as the one of the Sheep Mountain anticline in northern Wyoming shown in Figure 2. With later 57 58 EXXON ENERGY HID SYMPOSIUM SqurceRockm 4 Figure 1. Relationship of surface geology to subsurface structure. improvements, they could map in one day in the office what formerly had taken several months in the field. Geologists are now being aided by an even more powerful tool: special satellites for photographing the earth. In the early 19705, a 10Km Figure 2. Aerial photograph showing surface configuration of Sheep Mountain anticline. ADVANCES IN 59 satellite program now known as EROS (Earth Resources Observation Systems) was started (Figure 3). Each EROS satellite circles the earth in a near-polar orbit at an altitude of approximately 900 kilometers, carry? ing a multispectral scanner that covers the green to near infrared part of the spectrum. Data obtained are digitally recorded on board the satel- Multispectral Scanner Altitude 900krn Flight Path Figure 3. Earth Resources Observation Systems (EROS). lite on magnetic tape with a four-channel system. As the satellite orbits the earth, it images an area 185 kilometers square at a nominal resolu- tion of about 80 meters. Data are transmitted immediately, or are periodically relayed, from the satellite to a half?dozen ground stations. Data tape from the satellite can be purchased from the United States Geological Survey's Earth Resources Observation System Data Center, or from similar facilities in Canada, Japan, Brazil, Italy, or Australia. The data (Figure 4) are read into a large computer and corrected for spatial position, detector malfunction, and latitude. The corrected tape is then used as input to Exxon's own system for enhancing desired images. First, the data are displayed on an interactive system through which a geologist can work with them to enhance specific natural features or to emphasize small differences between features of interest. At this point the skill and experience of the geologist become critical to success in oil or minerals discovery. When the specific enhancement techniques to be used on the image have been determined, the scene can be interpreted with the interactive system or returned to the large computer for enhancement processing: 60 EXXON ENERGY RED SYMPOSIUM Digital Processing System Correct for: Satellite a Spatial Position Corrected Data Tape Large compmer Detector Mallunclion Tape Latitude Digital Enhancement System Enhance: Large Computer Emphasize Differences Corrected Interactive Visual Tape '9 Display System a Tri?Color 1:1 Composite Photograph NV Digital Plotter Tape Figure 4. Processing and enhancement of satellite imagery. if a hard copy of the image is desired, the enhanced tape is sent to a digital plotter that produces a composite color photograph by exposing photographic paper successively to blue, green, and red light. The satellite photographs pick up a number of apparent faults and linear features that could not readily be mapped by surface geogogical techniques. Different rock types and mineral deposits often have unique spectral signatures. By means of various mathematical proce- dures, the computer can systematically modify the spectral data and then assign arbitrary, contrasting colors that further enhance these data and help to identify these deposits. For example, a particular color combination can be used to emphasize the beds of oxidized sandstones found in the wash zones along large creeks, while another color combination can help to identify areas within the beds where alteration of the formations has taken place, such as extensive oxida- tion or alteration by the physical forces of nature. A specific color combination can permit the geologist to differentiate between alkaline-rich beds and carbonate beds. Another can be par- ticularly useful to identify contacts between different types of rocks. Satellite photographs can show the presence of coal or oil shale. One way to do this is to locate a known mineral deposit on the ground and then determine its reflectance characteristics from the satellite data. Next, we can try several color combinations until we find the one that best delineates the deposit. Other areas that have the same spec- tral characteristics will then be assigned the same color as the known mineral deposit. The same techniques can be applied to other surface features, including mineral deposits, hydrocarbon seeps, and changes ADVANCES IN 61 in mineral type and vegetation. New sensor systems to be installed in future satellites will have much-improved resolution and will detect a wider range of frequencies. Thus, this valuable tool for understanding surface geology promises to make even more significant contributions to the future discovery of oil, gas, and minerals. CHANGING GLOBAL SEA LEVELS Now, I would like to turn to our research on prehistoric global changes in sea levels. This seemingly academic subject has become a powerful geologic tool. We frequently use it in our exploration efforts, especially in frontier areas where we have little or no information on subsurface wells. Basically, this tool helps us predict more accurately from seismic data the age, depositional environment, and rock types of potential targets for exploration. The concept of global changes in sea level is not new. A century ago, earth scientists recognized evidence in sedimentary rocks of syn- chronized, worldwide rises and falls of sea level. In the past 20 years, Exxon has advanced this concept by painstakingly matching sedimen- tary sections around the globe in order to define quantitatively the cycles of sea-level rise and fall. As a result, numerous cycles having durations of 1?50 million years were documented for the last 500? million-year period. More importantly, this documentation led to the discovery that most of the changes in sedimentary sequences depo? sited in subsiding basins are controlled, not by local uplifts of the surrounding land surface, but by global changes in relative sea level. Over millions of years, changes in sea level occurred when either the volume of the ocean basin or the volume of sea water in it changes. Over time, both things happened simultaneously, but at varying rates. Let us look at these two processes separately. First, let us consider two situations where the volume of the ocean basin changes in re- sponse to internal earth forces that push the continents apart. These forces come to the surface at the midocean ridge. In the first situation (Figure 5), the continents are moving apart rapidly in terms of geologic time, and hot, newly formed ocean crust is defining a broad, midocean ridge. The center of the ocean basin is filled in with hot new crust, and the sea level rises to compensate. In the second situation, the opposite occurs (Figure 6). The conti- nents are moving apart slowly, and the ridge of hot, newly formed ocean crust is narrower. Therefore, the volume of the container is greater, and, consequently, the sea level is lower. 62 EXXON ENERGY R810 SYMPOSIUM Sea Level Figure 5. Sea level is raised by thermal expansion of a fast- spreading midocean ridge. Calm Moat ?Flow Figure 6. Sea level is lowered by thermal contraction of a slow- spreading midocean ridge. Geotectonic events like these alter the volume of the ocean basins slowly and are responsible for long-term changes in sea level having a duration of 10?50 million years. Short-term changes having a duration of 1?2 million years may be caused by the addition or subtraction of water from the ocean?basin container. The melting of major continental glaciers in polar regions adds water and raises the global sea level. Conversely, widespread glaciation can store substantial volumes of water on the continents, thus lowering the sea level. Let us take a closer look at how changing global sea levels control the pattern of deposits in a model ocean basin. Block diagrams, includ- ing that in Figure 7, will be used to represent an ocean margin where land-derived sediments onlap the continent. The cross-sectional dimensions are variable, ranging from about 20?200 kilometers hori- zontally and about 300?1500 meters vertically. The relative positions of sea level during a complete cycle, from high to low to high, are shown at the right in these diagrams. ADVANCES IN 63 When sea level falls rapidly (Figure 7), say over a period of half of a million years, the continental shelf is exposed, and a rapid rush of coarse material that normally is deposited on the shallow, flat shelf is flushed farther seaward. A lowstand submarine fan may accumulate in the adjacent deepwater environment (Figure it generally contains sands of reservoir quality. When the sea level reaches its lowest point, the geologically rapid flush of coarse sediment ceases, and the streams return to equilibrium. A smaller volume of coarse sediment is trapped back on the coastal plain, while only fine sediment coming from rivers reaches the deep basin areas. This finer-grained sediment, which is transported to the basin, backfills behind the fan to form the lowstand delta. It is generally shaly and of little or no reservoir potential. saw Falling Sea Level High I 1., Shell 300-150er sno- Oldov ?nch Time 20-200km Figure 7. Depositional sequences and sea-level changes start of cycle. Sea Level High Sea Level High . Low Low Stand . Ig?t'i'l?wnanne - 004.: Ice? u, 11 Low Stand Delta High Stand Sequence me mneservoir Quality Sands Shale Figure 8. Depositional sequences-and sea-level changes completion of cycle. 64 EXXON ENERGY SYMPOSIUM As the sea level again rises, the deposits do not make their way as far basinward. They are trapped in the shallow areas of the shelf. When sea level remains in a high position over a long period oftime, and conditions stabilize, river systems transport new deposits and dump them in the shallow water environment. Eventually, such shal- low water areas fill with sediment. Additional debris that reaches the shoreline is progressively dumped farther seaward into previously deep?water environments. These sea-level cycles may remind you of the ebb and flow of tides that you have seen during a day at the beach. Global changes in sea level roughly repeat the same patterns on a much larger scale and over a much longer time. in fact, one full cycle from lowstand to highstand and back to lowstand, which we have summarized here, generally takes a few million years. Of course, we do not have direct observations covering a few mil- lion years of changes in sea level. Geologists have had to extrapolate these changes from cyclic depositional patterns observed repeatedly on seismic sections from all over the world (Figure 9). Indeed, Exxon has charted ancient changes in sea level over the last 500 million years in more than 60 locations around the world. the Figure 9. Locations of global sea-level studies. Figure 10 summarizes the results of our analyses of changes in sea level during the Cenozoic Era, from 65 million years ago to the present. Note both the many short-term cycles of varying magnitude as well as the major fall in sea level of approximately 300 meters that took place ADVANCES IN 65 29 million years ago. Note, also, the overall long-term decline in sea level of about 250 meters. Millions of Years Before Present High Long Term Change - 300 -l 2 -- Est-mated Level x-l 100 in Meters A 0 Present Day Sea Level Rise 100 200 Low Figure 10. Global changes in sea level during the Cenozoic Era. Using the North Sea as an example, let us now turn to the applica- tion of the global charts of sea level to the interpretation of seismic sections such as the one shown in Figure 11. This is a cross section about 80?kilometers long by 2,500-meters deep that shows various layers of sedimentary rock that reflect sound or seismic signals. Fre- quently, decisions on the petroleum potential of an area must be based primarily on seismic images of the sedimentary rocks. This is especially true in offshore frontier areas with limited or no well control. Knowing the ages, depositional environments, and rock types is critical. To determine these factors, the major sequence boundaries are first identified. These sequence boundaries, emphasized with fine lines in Figure 11, are horizons along which discontinuities occur in the paral- lel reflector patterns. They are believed to represent shifts in deposi- tional sites controlled by changes in sea level. Next, the sequence boundaries are matched to falls in sea level recorded on the chart (Figure 10). The obvious place to start in this trial-and-error procedure is by matching the most pronounced fall in sea level, recorded at the 29-million?year mark on the chart, with the sequence boundary displaying the largest discontinuity on the seismic section (Figure 12). Best fits with the sea-level curve are then made for other sequence boundaries. Using the concepts summarized previously in the block diagrams (Figures 7 and 8) on the relationship between change in sea level and 66 EXXON ENERGY RIID SYMPOSIUM 0 Seconds - 16km of Years Before Present Seconds ml 50* 60 VE-2011 16km 1mm" Low Sea Level igh Sea Level Sea Level Shale Reservoir Sand message. We (top) Figure 11. North Sea seismic-section interpretation of major sequences. (bottom) Figure 12. Geologic interpretation of North Sea seismic section, and matching sequence boundaries to major falls in sea level. depositional sequences, a geological interpretation of the seismic sec- tion is then undertaken. After a series of intersecting seismic sections has been analyzed in a similar manner, potential exploration targets are identified for further study. In this case, the depositional sequences associated with the four major cycles of sea level that occurred be- tween 60 and 45 million years ago look promising. Using information gleaned from several sections, a composite schematic section is drawn to clarify the interrelationships of deposi- tional events taking place in the general area of interest (Figure 13). The bottommost or first sequence, deposited 60 million years ago, is the starting point. Deposition in the first sequence began with the development of the lowstand submarine-fan sands, flushed seaward with the rapid fall of sea level charted on the left. ADVANCES IN 87 Low of bars Before Present Basin Low Sea Level High Sea Level Delta Shale Reservoir Sand Submarine Fan Shale Reservoir Sands Figure 13. North Sea depositional sequences versus global changes in sea level. As sea level reached its lowest point, the finer sediments settled out closer to shore, forming a lowstand delta. With the ensuing rise and highstand of sea level, sedimentation backed up even farther toward the continent. Most of the sedimentation occurred in shallow water environments. You will recall that in this highstand phase, as the area of the shelf infills with sediment, deposition occurs progressively farther seaward to complete the first cycle. This cycle of events was repeated three more times between 60 and 45 million years ago. Once the depositional events taking place in an area are interpreted, stratigraphic traps (sand bodies encased in sealing shale) can be recog- nized. These traps frequently form reservoirs for oil and gas. The submarine fan deposited during a cycle of sea level that started 50 million years ago actually represents the Frigg Gas Field of the North Sea (Figure 14). It is approximately 8-kilometers wide, 100? meters thick, and contains 10 trillion cubic feet of gas. This series of illustrations has demonstrated, I hope, why we feel that our knowledge of the global history of sea level has turned out to be a significant step forward. OFFSHORE PRODUCTION Now, let us change our focus from exploration to production technology. After an offshore discovery of oil or gas has been made, the next steps are to design, fabricate, and install a system to produce BB EXXON ENERGY nan SYMPOSIUM 0.0 1830 Meters 1.0 - Time - Seconds 2.0- 3.0- Figure 14. Seismic section for the Frigg gas field of the North Sea. the field (Figure 15). Early offshore developments in shallow water relied on conventional steel platforms to support drilling rigs and pro- ducing equipment. To date, more than 1,000 of these platforms have been installed in the Gulf of Mexico, including over 100 by Exxon. As the industry moved to deep water and to harsher operating envi- ronments, new concepts were developed. Concrete gravity platforms, for example, are now being used extensively in the North Sea in water depths to over 150 meters. New equipment and techniques have now extended the use of steel platforms to water depths of about 300 me- ters. We have also developed a deepwater structure called the guyed 60 Meters 150 Meters 300 Meters 600 Meters 1500 Meters Figure 15. Progressive development of technology for production of oil and gas in deeper-water systems. are. ADVANCES IN 69 tower that will enable us to produce oil from fields using conventional platform drilling and production technology in water depths to about 600 meters. Finally, subsea production systems are being engineered for use in water depths beyond 600 meters. Similar subsea production methods are now being used as a means of producing fields in water depths less than 600 meters that would otherwise he marginally economic. I shall discuss these deeper-water systems individually. A gravity platform is a large, concrete structure held in place on the sea bottom by its own weight. Equipment to accommodate numerous wells and oil production of many thousands of barrels per day is in? cluded on its massive deck. Such platforms are very expensive. For example, the total cost of a new gravity platform now under construc- tion is expected to be about $2 billion dollars, including platform facilities but excluding the cost of the drilling program. Our research on gravity platforms is leading to better technology and equipment, as well as providing us with the information required for an independent analysis of structural designs being offered by contractors. For example, we have developed specialized procedures for analyzing the stability of foundations composed of nonuniform soils. Technology developed for iacket platforms has been useful in determining the deckiatigue characteristics of gravity structures. Exxon Production Research Company has been a leader in develop- ing dynamic analytical techniques and complementary fatigue- analysis procedures to determine the response of deepwater platforms to dynamic wave and earthquake loads. Steel-jacket platforms such as the Hondo platform standing in 250 meters of water in the Santa Barbara Channel are designed to resist the loads imposed by severe earthquakes. An earthquake in the Santa Barbara area in 1978 was sufficiently strong to cause the drilling derrick to vibrate noticeably, yet later inspection revealed no damage to the structure. The Exxon-developed guyed tower (Figure 16) is a structure that is permitted to move with the waves instead of rigidly resisting the applied loads as does a conventional platform. The tower is held upright by guylines that run from it to clump weights on the ocean floor and then to anchor piles. Under normal operating conditions, the clump weights remain on the ocean floor, and deck motion is virtually imperceptible. During a severe storm, some of the weights lift off the bottom, permitting the tower and guylines to sway and to absorb the larger loads. The motion of the tower is small enough to permit standard platform drilling and producing operations, but large 70 EXXON ENERGY nan SYMPOSIUM Pile Anchor Guyed Tower Clump Weight Figure 16. Features of the Exxon guyed-tower system. enough to absorb and to reduce significantly the effect of wave forces. Even in a design storm, the tower sways back and forth only about two degrees, not enough to cause seasickness. We have conducted a one-fifth-scale-model test of a guyed tower designed for 500 meters of water and North Sea environmental condi- tions, with 1 2 other companies contributing to the test?s cost. The test structure was installed in 90 meters of water in the Gulf of Mexico. No major problems were encountered during 3.5 years of operation. Du r- ing 1978 the tower experienced eight storms that represent, to scale, design storms in the North Sea with waves up to 40 meters high. Analysis of these data have verified the analytical procedures that we use to design guyed towers. Currently, we are designing a guyed tower for use in developing a field discovered by Exxon in 300 meters of water in the Gulf of Mexico. In water depths beyond about 600 meters, it will become more economical to use some type of subsea production system. Therefore, Exxon has developed an integrated subsea installation referred to as a Submerged Production System, or SP5, to produce oil from potential deepwater discoveries (Figure 17). In this system, the wells and as- sociated equipment are located on the seafloor. Flowlines connect the submerged production system to a production conduit, or riser. The produced fluids are processed and stored on a floating production vessel, in this case a converted tanker. Special high-reliability compo- nents and techniques minimize equipment malfunctions. The seafloor equipment is designed in modules that can be replaced, usrng an unmanned maintenance manipulator. Production operations Will be ADVANCES IN 71 Storage ti Treatment 9.. ?If: kg -3 Vessel ?72" I It? W.- Mainlenance - - Vessel .. Dynarmcatly Posnioned Drill Vessel Maintenance l. Production Manipulator Riser Flowlines 5 . 0 . SPSTernplale \g Figure 17. Exxon?s integrated Submerged Production System (SP5). remotely controlled from the storage and treatment vessel or another surface facility. Exxon has been developing its SPS since 1968. The most significant element ofthis development effort has been a prototype pilot test in the Gulf of Mexico in 50 meters of water. This effort included design, fabrication, testing, installation, and operation of all critical system components. The structure is 30-meters wide, 38-meters long, and l3-meters high. Once the installation site was reached, the SPS was launched from the barge and then lowered to the sea floor. The wells were drilled through the central area. In 1977, installation of the production riser completed the three-well prototype system. The installation program used techniques appropri? ate for deepwater applications; that is, operations were controlled re- motely from the surface without the use of divers. The wells were successfully produced for an extended period, and the production tests are now complete. All major objectives of the field test have been achieved. Total cost amounted to about $82 million when the project was completed last year. Equipment for the first commercial application of the SPS con- cept is now undergoing land tests in Holland. This system is intended for use in the North Sea. We are also developing a caisson vessel (Figure 18) to serve as a permanent floating platform for drilling and production for use with an SPS template in deep water. We have completed a two-year prelimi- nary design study of the concrete vessel shown. This work clearly indicates that drilling, processing, storage, and offloading can be safely 72 EXXON ENERGY SYMPOSIUM 4 . Caisson Vessel Luna] if! nilFigure 18. Caisson-vessel drilling and production system. and efficiently performed on a single vessel. We have also designed a steel version of the vessel hull. This effort will further support technical and economic tradeoffs for deepwater prospects. Exxon is drilling a number of exploration prospects in water depths in the 900?1500 meter range. If a discovery is made, we can begin an orderly final design and construction of either the SPS?caisson?vessel system or the SPS-tanker system previously described. We recognize that development of an oil or gas field in over 900 meters of water will be extremely difficult and costly and will require highly skilled people. Research and prototype tests of the type that have described are essential to produce resources in deepwater areas. ENHANCED OIL RECOVERY It is now timely for me to discuss enhanced oil recovery, or the recovery of oil that has already been discovered but cannot be pro- duced economically from known reservoirs by natural depletion or by further injection of water or gas. Methods for enhanced oil recovery are promising and are being vigorously developed in many parts of the world, but their applicability has technical and economic limits. The reason why enhanced recovery methods are needed can best be explained if it is understood why conventional recovery methods in- variably leave some oil trapped in the reservoir rock. 50 let us begin by studying a reservoir rock. ADVANCES IN 73 Figure 19 shows the passages inside a rock that initially contained both water and oil. This photograph of a sandstone core from a produc- ing formation shows, through successive enlargements (Figure 20), the complicated geometry formed by the grains of sand and the intercon- necting passages or pores. Normally, salt water, or brine, coats the rock surfaces, and oil forms a continuous phase winding through the tiny passages. Primary production and secondary water injection remove part of the oil, but some portion remains, frequently as much as 40?60% of Figure 19. Passages inside a reservoir rock. 74 EXXON ENERGY SYMPOSIUM ?r Figure 20. Further enlargements of the porous reservoir rock shown in Figure 19. ADVANCES IN 75 the oil originally in place (Figure 21). The remaining oil is no longer continuous, but exists in the form of discrete droplets like that shown, trapped in pore constrictions throughout the rock. Trapped Oil water Figure 21. Oil trapped in pore constrictions in a reservoir rock. Figure 22 is an enlarged photograph ofa wax cast ofan oil droplet. It was made by solidifying molten wax in the pores of a reservoir rock and then removing the rock by acid leaching. The droplets are micros- copic in size; approximately 4,000 of them would fit on the head of a pin. The challenge of enhanced recovery is to contact as many of these droplets as possible and somehow to liberate them from their traps in the reservoir rock. The principal methods for enhanced oil recovery that are either being applied in the field or are under development fall into one of three categories: 1 . Chemical flooding?use of special soaps or surfactants to wash the oil from the reservoir rock, or of polymers to increase the efficiency of a conventional waterflood; 2. Miscible gas flooding?use of solvents such as hydrocar? bon gases and CO2 to dissolve the oil; and 3. Thermal recovery?use of steam or in situ combustion to heat and to mobilize heavy crudes. There are three sub- categories under this heading: steam stimulation, steam flooding, and in situ combustion. Miscible gas and thermal processes are commercial in many loca- tions in the world, although we clearly need to learn more about the 76 EXXON ENERGY SYMPOSIUM ~avFigure 22. Enlarged photograph of a wax cast of a microscopic oil droplet in a reservoir rock. technology involved in order to improve their efficiency. Because these two methods are reasonably well developed, I shall concentrate on describing methods for chemical flooding, which are very much in the research stage of development. I shall, however, compare the oil recovery anticipated from all three enhanced-recovery methods. Figure 23, is a schematic representation of a typical process for surfactant flooding. At the end of a waterflood, only brine can flow in the swept region of a reservoir, and the residual oil remains trapped as microscopic droplets distributed throughout the reservoir. A small bank of surfactant is injected to begin the tertiary flood. Most droplets of residual oil will be mobilized in regions contacted by surfactant, with these mobilized droplets coalescing to form a moving, growing oil bank ahead of the surfactant bank. This is shown here in highly simplified form. A somewhat larger bank of brine, thickened by adding a low con- centration of polymer, is injected following the surfactant. The thick- ened brine bank is then followed by ordinary brine. The thickened brine bank improves the efficiency with which the surfactant sweeps the reservoir by controlling the tendency of the ordinary brine to chan- are ADVANCES IN 77 nel, or finger, through the surfactant and oil banks. Once the oil bank reaches the production well, both residual oil and brine are produced. Figure 24 shows the current surfactant-flooding field activity in the United States as listed in an Oil and Gas journal survey article last April. At that time, 22 tests were under way and 8 more were planned and formally announced. From available data, it appears that only a fraction of these tests will rec0ver as much as one-third of the residual oil in the test areas. Except for one planned test, none of the tests shown on the map is a full-scale commercial project. However, two tests are expansions of earlier pilot studies designed to evaluate com? mercial viability. Currently, use of chemical flooding, which includes both polymer flooding and surfactant flooding, is more active in the United States Brine Surfactant z?x . on . rr ne Ordinary Brine Figure 23. Surfactant flooding. 78 EXXON ENERGY R810 SYMPOSIUM El ?lIl 4 Active Planned Figure 24. U.5. field activity surfactant flooding. than in other parts of the world. However, projects for chemical flood? ing are underway in a number of countries outside of the United States, as indicated on the map in Figure 25. Several technical problems still either limit the applicability of sur- factant flooding or contribute to the marginal economics and high risk associated with use of the process. First of all, chemical systems have not yet been fully developed and tested for the high temperatures and 4? . . .1 Figure 25. Field activity chemical flooding. ADVANCES IN 79 salinities encountered in the majority of light oil reservoirs. Secondly, Chemical losses through adsorption and trapping dictate actual chemi- cal requirements and therefore strongly influence economics. About one metric ton of surfactant can be adsorbed by a cube of rock ten meters on a side. Finally, performance of the process in complex, actual reservoirs must be better defined. In the near future, at least, pilot testing will be required for each new potential application. Let me return to the first item on this list and to discuss it in more detail. High salinity is a problem because common surfactants that are effective at low salinities are not effective at high salinities. Further- more, at reservoir temperatures of about many common types of surfactants, particularly those that might otherwise be used at higher salinities, become degraded over the multiyear life of a field flood. 20 0 15 - 0 Salinity100 Temperature, c?C Figure 26. Temperature-salinity grid for surfactant field tests. Let me put the temperature-salinity problem in better perspective. Figure 26, shows conditions for past and ongoing field tests on a temperature-salinity grid. Many tests were conducted in low-salinity, low-temperature reservoirs simply because these'are the easiest targets. The low-temperature, low-salinity region on this chart repre? sents less than approximately 10% of the ultimate potential reserves for surfactant flooding in the United States, excluding Alaska, while the high-temperature, high-salinity region represents about 40%. The per- centage distribution may be different for reservoirs worldwide. 1 80 EXXON ENERGY HGD SYMPOSIUM However, the overall conclusion remains the same: there IS a need to develop new chemicals for high-temperature, high-salinity conditions. In recent years, a significant portion of Exxon?s efforts has been directed toward the development of surfactant systems applicable in the high-salinity region shown in this figure. We are conducting'a surfactant-flooding field test in the United States In the Loudon field in Illinois (Figure 27). The Loudon reservoir contains water With a very high salinity. This pilot test should aid in the development of high- Loudon Pilot Area <11 . 3 Illinois Loudon Field aam?-o 8 Figure 27. location of Loudon Field surfactant-flooding test in Illinois. salinity technology. Although the temperature of the Loudon field IS only about the chemical system used in Loudon has the poten- tial of being modified for use at higher reservorr temperatures, to Exxon has been active in surfactant-flooding research for well over a quarter of a century. Previous Exxon field tests of various types of processes involving surfactants have been conducted In several areas of the United States, including a previous test in the Loudon field. Significant advancements have been made over the past five to seven years. Even so, the technology remains in a research stage because of the complexity of the process and the marginal-to-poor economics. Even with further improvements in technology, additional. economic incentives may be needed to stimulate commercial production. Also, a lead time of seven to ten years is required to developand to test a process for a given field. Therefore, significant commercral production from surfactant-flooding projects cannot be obtained before 1990. ?r ADVANCES IN B1 In summary, there is no question that methods for enhanced oil recovery will become increasingly important in increasing our pe? troleum reserves. Even so, the near- and intermediate-term outlook for major increases in oil production from enhanced-recovery processes is still uncertain. Nevertheless, I want to focus your attention on the upper end of the ranges shown in Table I, because we believe that most of the uncertainties will be favorably resolved. Table 1. Status and outlook for enhanced oil recovery Recovery Timing for Level, Ultimate Significant ?lo Original Potential Process Status Production Oil in Place (10" Bbl) Surfactant Currently After 1990 10-15 2?7 uneconomic Miscible gas Applicable After 1985 5?15 2-5 in special cases Thermal Attractive Now 5-35 4-8 for better prospects Surfactant processes are currently uneconomic and thus not yet ready for general commercial application. No significant production is expected before 1990. Surfactant processes should be able to recover 10 to 15% of the original oil in place in the better-candidate reservoirs and ultimately to achieve a recovery potential of from two to seven billion barrels in the continental United States, excluding Alaska. Miscible gas processes are applicable in situations in which reser- voir geology is favorable and in which there are ample supplies of low-cost injection gas. This technology offers the greatest potential for near-term production from light oil reservoirs, with some possibility of significant production as early as 1985. Miscible gas processes should recover 5 to 15% of the original oil in place in better prospects and will ultimately recover between two and five billion barrels in the continen- tal United States, excluding Alaska. Miscible gas processes should also contribute substantially to worldwide production in the future. Of the thermal processes, steam injection is the preferred process in most heavy oil fields. It is currently being applied commercially on a :1 82 EXXON ENERGY nan SYMPOSIUM significant scale. In situ combustion is already marginally attractive in special situations. Depending on whether or not steam flooding fol- lows steam stimulation, the recovery level can vary from 5 to 35% of the oil originally in place. The ultimate United States potential (lower 48 states) from thermal processes should fall in the range of four to eight billion barrels. Worldwide, the prospect for thermal-process ap- plication is very good, particularly in Canada and Venezuela where vast deposits of heavy oil exist. Our estimates of the ultimate potential for present enhanced-oil- recovery methods in known fields in the lower 48 states is up to 20 billion barrels. I am hopeful that the 20-billion-barrel number is a conservative one, as new fields are discovered and as improvements in technology occur. We have now reviewed several examples of what Exxon and other members of the oil industry are doing to improve the world?s pe? troleum supplies. Exxon is committed to making a maximum effort to mitigate the energy problem that is certain to remain with us for many years. Successful research on exploration and production is a key in- gredient in reducing our dependence on others for petroleum. ABSTRACT Refinery and petrochemical plants have many types of in- teractive energy systems, including fuel, steam, power, and process?unit operations. These systems are being signifi- cantly affected by the continuing increase in energy costs and the accompanying opportunities to engineer advanced systems of higher efficiency and to utilize nontraditional energy sources. Within Exxon, significant reductions in energy consumption in refineries havelalready been realized via an integrated management and technology ef- fort. The major features of the Exxon program are highlighted, including results achieved to date and the out- look for continued progress. ABOUT THE AUTHOR Clarence M. Eidt, jr., is General Manager of Exxon En? gineering?s Petroleum Department in Exxon Research and Engineering Company. He received an MS. degree in chemical engineering from Louisiana State University and joined the Exxon Research Laboratories in Baton Rouge in 1956. In 1967?68, he spent a year on loan to Humble Oil?s Corporate Planning Department in Houston. He became Assistant Director of the Exxon Research Laboratories in 1968. In 1972, Mr. Eidt joined as Assistant General Manager of the EE Petroleum Department. The following year he returned to Baton Rouge as Manager of the Exxon Research Laboratories. In September 1976, he transferred to Exxon Corporation in New York as Manager of Regional Planning in the Logistics Department. Two years later he became Manager of Corporate Planning Coordination in the Corporate Planning Department, the position that he held prior to his return to in September 1980. Mr. Eidt is a member of the American Institute of Chemical Engineers and the American Petroleum Institute and is a licensed Professional Engineer in the State of Louisiana. C. M. Eidt, Jr. ENERGY EFFICIENT REFINERIES: MAKING MORE FROM LESS The efficient use of energy is of growing importance to consumers and suppliers of energy alike. As an energy supplier, efficiency considera- tions permeate every aspect of Exxon?s operations, from the production of energy raw materials?crude oil, natural gas, and coal?through the manufacture and distribution of finished industrial and consumer pro- ducts, such as motor gasoline, jet fuels, and heating oils. In this paper, I shall focus on one aspect of the oil chain?refining?and shall discuss recent and current Exxon activities in two primary areas: (1) energy utilization, as related to the basic thermodynamic and thermal effi- ciency of refinery operations, per se, and (2) improved utilization of crude feedstocks by modifying refinery selectivity. Advances in both areas are steadily reducing the amount of crude oil required to man- ufacture a barrel of desired refinery product. BACKGROUND Let us look, first, at energy efficiency. Despite its current popularity, energy efficiency is certainly not a new area of interest for refiners. There have always been driving forces to reduce internal oil consump- tion and losses and to increase yields of finished products. Exxon has historically designed facilities that recognized these forces. Since the early 19705, however, rapid escalation in crude oil prices and the resultant emphasis on resource conservation have given substantial added impetus to these traditional drives. Before we look at current activities and what we see down the road, however, let me place these incentives in context and outline what has been accomplished to date. In the petroleum industry?s sequence for energy utilization shown in Figure 1, refining is the largest single consumer of energy. Producing 85 86 EXXON ENERGY ROAD SYMPOSIUM ll 9 Willa 2 20 on 5 a Consumed 4 in" 3 [3:7 2 1 (55?7019 Eli?? I I Producing Crude Refineries Marketing Transportation Figure 1. Energy consumption in the oil-supply chain. requires around crude transportation plus marketing, roughly and refineries, about All together, some 15% of the energy avail- able in crude oil is consumed in producing, refining, the remaining 85% to end users, with refineries accounting for more than half of the total. Taking a closer look at refining, Figure 2 shows that for every 100 barrels of crude oil delivered to the refinery, about 8 barrels are re- quired to meet internal energy needs for processing. At first glance this may not seem like very much; indeed, from the standeInt of utiliza- tion of raw materials it compares quite favorably With many other industries. Nevertheless, this 8% for energy consumption represents some 50% of the operating costs of a typical refinery. I . More importantly, every extra barrel that can be saved during refin- ing is one less barrel that has to be produced to meet customer de- mands. From this perspective, improved efficiency and conservation 3 100 . Ene Delwemd crUde Conleguymed Refinery Opera?ng 005?s Re?ning Products Figure 2. Energy requirements in refining. ENERGY EFFICIENT REFINERIES 87 can also be viewed as equivalent to augmenting supply. Within Exxon, for example, we have achieved a 22% improvement in the efficiency of energy utilization in our worldwide refining circuit relative to pre- embargo practices. In 1980 alone, the resultant savings amounted to 28 million barrels, equivalent to the output from a multibillion-dollar fuels plant that would require many years to plan, design, build, and start up. Efficient use of refinery energy obviously makes good business sense from the standpoint of reduction in both costs and supplies, as well as helping to conserve a valuable natural resource. These savings did not just happen. In addition to requiring signifi- cant investment, they are the direct result of an integrated management and technology effort to improve energy efficiency in refineries (Figure 3). This effort combines: (1) a management system for measuring energy efficiency, setting goals, and monitoring progress, with (2) Energy Management Systernl Set Goals I 1 Monitor Progress Energy Analyses Individuai Integrated Process Energy Units Comp ex Current Future Refinery VI Refinery Jr Advanced Technology Jr Process Non-Process Figure 3. Energy-management system. energy analysis of individual process units and total plant sites, and (3) to develop advanced technology. We believe that each of these elements is essential for continued progress toward tomorrow?s refin- ery, whose efficiency will be still higher, and I shall be using this concept of an integrated management and technology effort as a framework for outlining where we have been and where we see our- selves headed in both energy efficiency in refineries and the utilization of crude oil. First, let us consider a generalized, but nontheless representative, refinery sequence for processing crude oil into various products (Figure 4). Starting at the left, crudeoil is separated in distillation towers into several primary components?light gases, naphtha liquids, middle dis? 88 EXXON ENERGY SYMPOSIUM Li ht Gases Naphtha Octane Improvement Gasoline Middle Dist. Desulturize Jet Fuel Crude Crude ?l Diesel Oil Separation Gas Oils . . Heating Oil Catalytic Cracking Heavy Desulfurize Fuel oil Figure 4. Simplified refinery process schematic. tillates, gas oils, and heavy bottoms. The light gases are further frac- tionated into separate components, some of which are propane and butane (LPG). The naphtha liquids are reformed to raise their octane quality to a level suitable for gasoline. Depending upon the source of the crude oil, the middle distillates may require desulfurization to produce satisfactory jet fuel, diesel oil, and heating oils. Gas oil streams are fed to a catalytic cracking unit where materials of higher molecular weight are chemically converted to yield more gasoline and light distillates. The bottoms are either sold as heavy fuel oil or con- verted via an upgrading step for bottoms, such as Exxon?s FLEXICOK- process, to yield additional feedstocks for catalytic cracking and the other processes indicated. The various processing steps shown here differ in their energy re- quirements. Depending upon the quality of the crude oil feedstocks available, the desired mix and quality of products, and other factors, individual refineries will tailor this basic sequence, and hence the processing intensity, to fit their particular situations. As a result, re- fineries are not all alike, and they can and do vary widely in internal energy requirements. Figure 5 compares a typical product-demand barrel, showing the desired proportions of major refinery products, with two representative crude oil feedstocks of differing quality. Light, sweet crudes, shown on the left, characteristically contain significant amounts of natural- product precursors that can yield the desired product slate with rela- tively mild processing. Conversely, these preferred crudes are low in heavy, tar-like bottoms that require intense processing to convert them, chemically and physically, into the desired molecular species required ENERGY EFFICIENT REFINERIES 89 for various refinery products. In addition, light crudes generally con- tain relatively small quantities of contaminants, such as nitrogen and sulfur, that must be removed to avoid corrosion of equipment and deleterious effects on product quality. Moreover, light crudes are also low in their content of metals, such as nickel and vanadium, that typically deactivate the that are used to promote certain de- sired chemical reactions during refining. Heavy, sour crude oils, on the other hand, have fewer natural- product precursors, more heavy bottoms, and more contaminants such as nitrogen, sulfur, and metals. The bottom line is that the manufacture of a given product slate from a heavy, sour crude requires intense processing and may use 30?100% more energy than is needed to manufacture the same product slate from a light, sweet crude. Simi? larly, variations in the desired product split also affect a refiner?s energy requirements. As indicated at the bottom of Figure 5, a reduction in the yield of heaVy fuel oil requires additional conversion of bottoms, more intense processing, and even higher consumption of energy. Light Sweet Crude Product Demand Heavy Sour Crude Low Energy Inpm Middle Distillate Middle Distillate Jet Fue' 8: Gas Oils Diesel Oil 8? Gas 0"5 Heating on Heavy Fuel Oil Bottoms Moderate Higher Energy Energy input Input Jet Fuel Diesel Oil Heating Oil Figure 5. Energy consumption in refineries is a function of crude oil quality, product slate, 'and processing intensity. 90 EXXON ENERGY SYMPOSIUM Poor Quality Feed Good Quality Feed Energy Required Der Feed Processing Severity Figure 6. Energy-guideline factors for a catalytic cracking unit. ENERGY MANAGEMENT Turning back, now, to Exxon?s energy management system (Figure 3), these crude oil, product, and processing variations are recognized by using a building-block approach in tailoring a unique yardstick for the energy efficiency of each refinery. First, an energy guideline factor is defined for each individual pro- cess unit in the refinery, for example, for a distillation unit, a desul- furizer, or a catalytic cracker. These guideline factors are an amalgam of practical economic considerations and technical feasibility; in ef? fect, they are designed to represent a stretch goal in terms of achieva- ble energy efficiency. These energy guideline factors are expressed in terms of the energy required per barrel of feed processed in the unit. Feedstock quality and severity of processing also affect energy con? sumption, as noted earlier, and the energy guideline factor allows for such effects. Point A on Figure 6, for example, is the guideline factor for processing a feed of poor quality in a catalytic cracking unit. Table 1. Constructing a Refinery?s Energy Yardstick Process Units Guideline Target (Energy/Day) Catalytic cracker 2,000 Crude separation 2,200 Others Sum (total refinery) 5,200 ENERGY EFFICIENT REFINERIES 91 Efficiency Yardstick 160 Actual Energy 140 (6240 Guideline Ergo, 120% of Guideline (5200 Bbl/Day) 100 80 60 Figure 7. Measuring energy efficiency in refineries. We have developed energy guideline factors for some 40 different types of process units that essentially cover the total energy consump- tion of an entire refinery. Each unit's target for energy consumption is then determined from the energy guideline factor and the feed rate to the unit. By selecting appropriate process-unit building blocks, the energy targets for individual units can then be summed to yield the energy guideline target for an entire refinery (Table As illustrated in Figure 7, actual energy usage is then divided by the guideline target and expressed as a percentage of guideline. Refineries grading 100% have an energy-consumption performance that is equivalent to target, while values over 100% indicate that potential energy savings can still be achieved with modern facilities and sound management. Since the energy guideline system recognizes changes in crude oils, product slates, or refining facilities, it can be used to reset goals and to monitor progress over a period of time. Figure 8 depicts the Exxon history from 1973 to 1980 that I mentioned earlier?a 22% improve- 25? 20 1 5 Improvement Since 1973 10 5 0 i 1973 1975 1977 1979 1931 Figure 8. Improvement of energy efficiency in Exxon refineries. rw?ruv? 92 EXXON ENERGY SYMPOSIUM 28 Cumulative Savings (1974? 1980) 26 120 Million Barrels 23 19 13 Millions of Barrels .1974 1975 1976 1977 1 978 1 979 1 980 Figure 9. Energy saved in Exxon refineries. ment in energy efficiency, equivalent to 28 million barrels saved in 1980 alone, and cumulative savings of 120 million barrels over the time frame indicated (Figure 9). EARLY CONSERVATION EFFORTS The savings obtained to date are largely attributable to improved operating practices and to heat recovery schemes, more or less straightforward, that resulted from energy audits and analyses of indi- vidual process units. These early responses to rising energy prices were logically biased toward efficiency projects that could be rapidly im- plemented by using technology that was already largely in hand. A distillation column provides a good, if somewhat oversimplified, illus? tration of this sort of intra-unit energy recovery. Indeed, this type of simple heat integration was rather commonplace in the industry even before rising crude prices made economically viable the more com? plex and costlier systems for energy recovery. Distillation columns are commonly used to separate the various components in an oil stream that have different boiling points; there are many such columns in a typical refinery. In a distillation column, energy is added to vaporize the lighter-boiling materials in the feed. These lighter materials rise and are taken overhead as vapors, while the higher?boiling, heavier components exit as liquids from the bottom of the column. Referring to the example in Figure 10, to raise the feed to its boiling point and partially to vaporize it, an amount of energy equivalent to ENERGY EFFICIENT REFINERIES 93 Distillation Column %?ondenser Energy Reflux Preheater Storage 2 Energy Reboiler Energy Cooler 3 Energy Storage Figure 10. Energy balance - distillation example. Feed 7 that contained in two barrels of heavy fuel oil is supplied to the feed preheater. In addition, another eight energy-equivalent barrels are added at the bottom of the tower to boil off more of the feed. At the top of the tower, the lighter?boiling materials are taken overhead as a vapor and condensed back to the liquid state using cooling-water heat exchangers. These extract about seven barrels worth of energy from the vapor. Some of the condensed material is returned to the tower as reflux to help to achieve the desired separation, with the remainder representing the net overhead product from the distillation operation. ln a similar fashion, the heavier, higher-boiling components removed from the bottom of the tower are cooled, extracting the remaining three energy-equivalent barrels, before they are routed to product storage. One obvious step for heat integration is to exchange energy between the hot tower bottoms and the cold feed, thereby raising the tempera- ture of the feed stream while simultaneously cooling the bottoms. As shown in Figure 11, this heat-integration step eliminates the previous requirement for two barrels of equivalent external energy to preheat the feed, resulting in an overall energy savings of 20%. As noted, this is an oversimplified example. However, even when the analysis is confined to individual process blocks, quite a few alter- natives exist. In this case, for instance, if the bottoms product were going on to a second-stage distillation for further separation, it might be preferable to leave the heat in the bottoms stream and to look for another hot fluid to preheat the incpming feed. 94 EXXON ENERGY FIG-D SYMPOSIUM Distillation Column Storage ?Condenser Energy Reflux "o?c Reboiler Energy Coolor? - we iStorage Figure 11. Energy balance distillation example with heat integration. Seal?? Preheater Energy Storage PERSPECTIVE Such analyses become increasingly complex as the simpler intra- unit opportunities for energy recovery are exhausted, and we begin to look at the entire refinery as a single energy system composed of literally hundreds of heat sources and heat users that encompass a wide range of temperatures and are geographically separated, in some instances, by distances of several miles or more. in addition, heat is not the only energy system in a refinery. Power generation and the re? quirements for mechanical shaft work represent other energy needs in a different form. When we step back to this broader perspective, the vista for energy efficiency widens, new opportunities for applying ad- vanced technology become apparent, and more powerful engineering tools are required to support audits and analyses of the energy system. This is where we are today?learning to think in terms of a total, integrated energy complex composed of three basic systems, as illus- trated in Figure 12: Processes fall into two major categories: chemical reactions and separations. The chemical reactions can be further subdivided into those that give off heat?exothermic reactions?and those requiring the input of heat in order to sustain the reaction?endothermic reac- tions. Separation processes, among which distillation is the dominant technique employed today, typically require heat. Depending upon the molecular nature of the components to be separated, the tempera- ture level of the required heat input can vary from near-ambient levels to temperatures in excess of ENERGY EFFICIENT REFINERIES 95 3 Work - Steam Boilers/Turbines gum I. Gas Turbines . (Power Generation) Compressors/ Processes Fuel 5 Heat Exchangers Reactions Furnaces Separatiory Figure 12. Basic energy systems in refineries. The Heat system employs fuel-fired furnaces to supply energy input to the reaction and separations processes. Heat recovery from hot process streams, as in the distillation example that we have just con- sidered, reduces the amount of externally fired heat that must be supphed. Work, the third system also requires external fuel to fire steam boil- ers and gas turbines. Gas turbines and/or back?pressure steam turbines, in turn, provide work energy by generating power, or by compressing gases and pumping liquids. Only a few of the equipment items and operations involved in sup? plying, using, or transferring energy in each system are illustrated here. There are many others, and additional areas of interactive overlap as well. It would not be practical to comment on all of them here, but a few examples may serve to make the point. Looking at the work system in Figure 13, for instance, fuel-fired boilers can be used to generate steam that can, in turn, drive back- -.f?Work Steam Boilers/Turbines Fuel Gas Turbines .. Compressors Heat r" a: I Fuel Heat Exchangers.) Reactions . Furnaces .. \g Separations \x x_ .I Figure 13. interactions among energy systems in refineries cogeneration. 96 EXXON ENERGY SYMPOSIUM pressure turbines to generate power, and/or to drive compressors and pumps before the steam is utilized in the process systems as a heat source. The generic term for this technique?the integrated production of process heat and shaft work to generate electric power or to drive compressors?is cogeneration. There are many variations on this basic concept, one of which is illustrated in Figure 14. Low Pressure Steam Heat High Pressure Steam Steam Turbine Generator Boiler Hot Exhaust 4? Gas Fuel Gas Turbine Generator Electric Power Figure 14. Cogeneration. When energy costs were low, gas was typically burned in a gas turbine. The hot, high-pressure combustion gases perform mechanical work by driving the turbine; this, in turn, drives a generator to make electricity. When the hot combustion gases are vented, the overall efficiency of utilization of the energy in the fuel is around 30?35%. In one embodiment of cogeneration, however, the hot exhaust gases from the turbine are used to make steam in a boiler; in turn, the steam can drive a high-pressure steam turbine to generate additional electricity or to provide shaft work to a compressor or pump. Exhaust steam from the turbine can then be used further to provide heat to process units. Schemes such as this can raise the efficiency of utilization of the energy in the incoming fuel from 30?35% for the power-generation- only mode to above 80% for the cogeneration mode. Of course, addi? tional equipment is required, and the investment for it must in turn be balanced by the energy savings achieved. Let us turn, now, to the heat and process systems. There are, as have mentioned, many sources and users of heat in a refinery that provide or need heat at many different temperature levels. Both the amount of heat required and the temperature level are important. Fig- ENERGY EFFICIENT HEFINERIES 97 ure 15 presents a typical profile showing the amount of heat required by a refinery at various temperature levels. Separation processes, such as the distillation example considered earlier, call for about 50% of the total heat requirement and tend to be situated near the lower end of the temperature spectrum. The remainder of the heat requirement is con- sumed in chemical reactions at higher temperatures. These reactions involve a change in the molecular structure of the various crude oil Amount of Heat IRecovered Heat Net Heat Demand Cat Cracking Reforming Hydrogen Production I I I I I I 100 200 300 400 500 600 700 800 Temp Separations 1 Reactions Figure 15. Typical profile refinery heat requirements. components, and high temperatures are often required to get the de? sired reactions to proceed at a reasonable rate. For example, catalytic cracking and catalytic reforming, the basic processes for manufactur- ing motor gasoline, require heat at about Hydrogen manufac- ture requires heat at the upper end of the range, around Hydrogen is a key ingredient in certain refinery processes to remove contaminants, such as the sulfur and nitrogen mentioned earlier, and to help to convert heavier liquids to lighter ones. About half of the required heat can be obtained by suitable energy exchange and heat recovery among hot and cold streams, as in the distillation example. The net heat demand is then supplied by furnaces that burn refinery gases or liquids and/or by externally supplied gas. Additional heat is potentially recoverable, but, unfortunately, as shown in Figure 16, much of it is available only at low temperature levels and/or is contained in furnace stack gases from which recovery is relatively difficult. Heat transport loops or hot belts are one means by which utilization of low level and/or other sources of waste heat can be improved. Such systems typically employ a circulating fluid, such as an oil stream, to 98 EXXON ENERGY SYMPOSIUM Amount of Heat 100 200 300 400 500 600 700 BOO Temp Stack Gases Cooling Systems Figure 16. Waste heat. recover heat from a variety of sources of waste heat and subsequently to transfer it to a number of heat users. For example, waste heat from the process system, available from products being cooled on the way to storage or from furnace stack gases in the heat system, can be used to heat the circulating fluid. The hot fluid can then be used to preheat water for boiler feed or air for furnace combustion, thereby reducing the amount of fuel required to fire the boiler or furnace. Figure 17 depicts these interactions in terms of the three basic energy systems that were discussed earlier. These belts or loops not only facilitate recovery of waste heat and transfer of energy among geographically separate equipment, but also provide certain advantages in flexibility and reliability over one-on- Work Steam Boilers/Turbines \u Fuel Gas Turbines (Power Generation) Processes Fuel a t' Heat Exchangers ions Preheat Water . Preheat Air Separations 1 Figure 17. interactions among energy systems in refineries hot belts. ENERGY EFFICIENT REFINERIES 99 one integrated systems. Since they contain multiple sources and sinks and can be designed to include back?up trim heating and cooling, they begin to resemble a distributed utility system (such as for electrical power) wherein one or more elements can temporarily drop out of the system without adversely affecting the operability of the system as a whole. TOOLS FOR COMPLEX ANALYSES OF ENERGY SYSTEMS From these few examples, you can appreciate how broadening the horizon for energy efficiency to encompass the total site?and perhaps including neighboring industries and utilities as well?dramatically increases the opportunities for conservation of energy and improve- ments in energy efficiency. Not unexpectedly, the required energy audits and analyses undergo a correspondingly substantial increase in complexity: we have moved a long way from the example for simple distillation discussed earlier. Powerful tools for computer simulation, capable of sorting through hundreds of options for energy efficiency to arrive at a preferred combination, are being developed to cope with this increased complexity. One such computer tool used by our design engineers can esize and screen alternative networks for heat exchangers (Figure 18). With this tool, some 20 or more streams to be heated and cooled via heat exchange can be thermally matched by computer to line up the various heat sources and users in a way that minimizes the external net heat required. Investment costs and energy savings for the alternative cases are then weighed to determine the design bases that are most attractive. Stream Heat or Cool Temp. Range. A 30 ?150 300 - 50 Computer 200 - 30 40 80 Investment Net Fuel Required Minimize External Energy Figure 18. Computers facilitate the design of heat exchange systems. IN 100 EXXON ENERGY erD SYMPOSIUM Another computer system that we have developed within Exxon can screen a mix of different types of projects for energy efficiency?heat recovery, cogeneration, etc. This system is based on a mixed integer- linear programming concept and has the ability to examine partial as well as yes/no-type options. In a recent study, we used this tool to SCreen some 800 possible efficiency modules involving over 70 pro- cess units in a refinery-chemical plant complex. CONSERVATION VIA IMPROVED REFINERY SELECTIVITY Thus far, I have concentrated on benefits for energy efficiency that are attributable to heat recovery, integration, etc. It is time to turn, now, to advanced refining technology, and to examine how improve? ments in process technology that affect the basic chemical and physi- cal processes involved in refining operations can also make ?more from less.? Improvements in process technology can yield benefits in energy efficiency by two principal routes: 0 Directly, through energy savings, such as with new that promote desired reactions at lower tempera- tures, and 0 Indirectly, by improving refining selectivity to provide higher yields of the products most desired from starting materials of poorer quality. Today, many refineries do not have the processing capabilities to upgrade the heavy, high boiling bottoms in crude oils to lighter, more desirable products such as gasoline, jet fuels, and diesel fuel. This is particularly troublesome on two counts: first of all, the quality of crude oil is declining, which means more bottoms and more contaminants, and secondly, the primary market outlet for bottoms has traditionally been heavy fuel oil for use in power generation. This market is shrink- ing, and lower-cost sources of energy such as coal are available as substitutes. Since coal and nuclear energy are not readily substituted in other demand sectors, for transponation fuels, there is a strong incentive to develop improved technology for upgrading heavy crude bottoms to lighter products. To illustrate this point, typical yield data from a refinery without the capability for bottoms conversion are presented in Figure 19. As may be seen, the heavy crude yields nearly twice as much heavy fuel oil as does the light crude. Although the yield of motor gasoline in this example is only lower, the yields of middle distillate products, including jet fuel, diesel fuel, and home heating oil, are only half as great. In addition to highlighting the output problems for heavy fuel oils, these data also indicate that the refiner would have to process ENERGY EFFICIENT REFINERIES 101 Light Crude Heavy Crude Gasoline Mid. Dist. Prod. Heavy Fuel Oil Figure 19. Product yields on crude) in a refinery without bottoms conversion. more of the heavy crude in order to satisfy a given demand for gasoline and middle distillate products, including jet and diesel fuels. Technol- ogy for the selective conversion of bottoms can thus have the dual benefit of reducing crude oil requirements while providing higher yields of the more desirable light products. Given these driving forces, it is not surprising that a considerable amount of activity in the industry today?R&D and commercial projects?is aimed at the upgrading of heavy oils and bottoms. In this regard, two basic approaches are currently receiving the most attention: hydrogen addition and carbon rejection. Both routes lead to the desired increase in the hydrogen/carbon ratio. Exxon?s thermal processes, fluid coking and FLEXICOKING, are good examples of the carbon-rejection approach. Commercial projects based. on these technologies are being developed for a number of locations by Exxon and others. Fluid coking is an Exxon process that was developed in the 19505. In this process (Figure 20), heavy bottoms are fed to a reactor where they Reactor Products Flue Gas to Boiler Scrubber Burner Hot Coke Feed Coke By-Product 1 Cold Coke Coke . Figure 20. Simplified flow plan for fluid coking. 102 EXXON ENERGY RGD SYMPOSIUM are pyrolyzed to produce the desired lighter liquids. Some byproduct coke is also produced. The liquids are scrubbed and distilled prior to further processing into motor gasoline and middle distillates. The coke is transferred to a burner where some of it is burned with air to produce a hot coke stream that is then returned to the reactor to provide the heat necessary for the pyrolysis reaction. The residual coke is with- drawn as a byproduct, and the flue gas from the burner, which still contains some combustibles, typically goes to a boiler where it is used as a fuel to generate steam, thereby effecting additional recovery of energy. Fluid coking is a rugged, versatile process. It is relatively immune to the metals and other impurities that adversely affect many catalytically based conversion systems. it is capable, therefore, of processing a wide variety of difficult heavy feeds, including not only the bottoms from heavy crude oils, but also materials such as tar-sands bitu'men and nonconventional heavy crudes. The latter materials are so heavy that they will not flow under ambient conditions, thus requiring the injec- tion of heat into the producing reservoir before the oil can even be pumped to the surface. As shown in Figure 21, fluid coking completely converts the heavy fuel oil. Product yields for both gasoline and middle distillates are Without Bottoms With Fluid Coker Conversion Gasoline f: 42%? Mid. Dist. Prod. m" Heavy Fuel Oil Figure 21. Product yields when processing heavy crudes with fluid coking. increased significantly, although there is a residual yield of a non- premium coke byproduct. This coke, however, can be burned directly as a fuel, or, alternatively, it can be gasified in a separate gasification step to yield a low-Btu gas. ENERGY EFFICIENT REFINEHIES 103 FLEXICOKING is an advanced version of fluid coking that integrates a gasification step into the process to effect direct conversion of the coke into a low-Btu gas. The FLEXICOKING process (Figure 22) employs three major vessels, as contrasted with the two-vessel system r?9Fleactor Low-Btu Gas Products Scrubber Heater ??jCoke Read? Hot Coke - - Feed 6??1 Gas 35"? Purge Coke Coid Coke Air Steam Figure 22. Simplified flow plan for for fluid coking. The reactor system is essentially the same, but the single coker-burner vessel has been replaced by a two-vessel, heater- gasifier system. Coke from the pyrolysis reaction passes through the heater to the third vessel?the gasifier?where it is gasified with steam and air. Gas from the gasifier is routed through the heater, transferring heat to the hot coke circulating back to the reactor before the gas exits from the system for cleanup as a low-Btu gas. After gasification, there is now only a small purge stream of residual coke. Figure 23 shows that FLEXICOKING further upgrades the product slate, because the low-Btu gas, after cleanup, is a clean, sulfur-free fuel that can be used to fire refinery furnaces or boilers or can be sent to a nearby industry, thereby freeing more valuable fuels such as natural gas and low-sulfur fuel oils. Further, by integrating coking and gasifica- tion, FLEXICOKING avoids the energy losses that would be incurred in cooling and reheating the coke with a fluid coker plus a separate, standalone gasifier system. So much for coking and the carbon-rejection approach to upgrading bottoms. I shall not comment here on the other approach, hydrogen addition, other than to note that it is also receiving increased attention throughout the industry. We fully anticipate that technological ad- vances in both areas, coking and hydroconversion, will be forthcom- 104 EXXON ENERGY SYMPOSIUM Without Bottoms With Fluid Coker With FLEXICOKER Conversion 2970 35-h Gasoline 45%? 42% Mid. Dist. Prod_ 18% I Low-Btu Gas (Diso/ocoke Coke Heavy Fuel Oil 2 Figure 23. Product yields when processing heavy crudes with FLEXICOKING. ing and will be rapidly applied as refiners around the world pursue improved refinery selectivity as one means to make more from less. OUTLOOK FOR CONTINUED PROGRESS I hope that the limited examples that I have discussed so far have given you some appreciation of the progress made to date and the potential that we see ahead for further improvements in energy effi- ciency in refineries and in the utilization crude oils. We are, of course, researching a number of other avenues, and i shall touch briefly on a few of these in closing. As refineries become more energy efficient, refinery gases formerly used to fire furnaces will become available for upgrading into other products. We are investigating process technologies to accomplish this. For example, current processes for treating gas to remove undesir- able impurities, such as hydrogen sulfide and carbon dioxide, tend to be energy intensive, so we are also actively researching alternative process concepts and improved treating agents to reduce energy re? quirements for these applications. Regarding separations, we have already observed that distillation is the major separation process employed in today?s refineries. Other separation techniques that are less energy intensive are under study, including membrane-based separation systems and adsorption schemes for purification. Finally, we have mounted a very active effort to keep abreast of and to provide input to relevant energy developments by others, such as manufacturers of compressors, pumps, turbines, and systems for com- pUter control. ENERGY EFFICIENT REFINERIES 105 Looking back, we see a record of substantial progress in energy efficiency in refineries and in the utilization of feedstocks, and we believe that a key factor in our accomplishments to date has been an integrated management and technology approach involving: 0 Shaping up operating practices, 0 Putting a management system for energy in place to set goals and to monitor progress, 0 Making optimum use of technology in hand, 0 Broadening the focus of energy analysis to a total site basis, and 0 Carrying out an effective program to provide the technological base needed for continuing advances. For the future, we see considerable potential for continued im- provement. in response to this outlook, Exxon has significantly in- creased its engineering and research efforts in these areas, and antici- pates investing in excess of several billion dollars during the 19805 in energy efficiency and upgrading projects in our worldwide refining circuit. ENERGY SYMPOSIUM A1TENDEES Mr. Robert C. Anderson TRW Defense and Space Systems Group One Space Park Redondo Beach, CA 90278 Dr. Hugh A. Baird President C.F. Braun 8: Co. 1000 South Fremont Avenue Alhambra, CA 91803 Dr. Richard E. Balzhiser VP for Research Development Electric Power Research Institute 3412 Hillview Avenue Palo Alto, CA 94304 Mr. John M. Barbieri Special Consultant to the Director Institute for Marine 8: Coastal Studies University of Southern California Los Angeles, CA 90007 Mr. Robert Barker Lawrence Livermore Laboratory Box 808 Livermore, CA 94550 Dr. Frank S. Barnes Acting Dean University of Colorado Campus Box 422 Boulder, CO 80309 Mr. Edward J. Baum Director of Corporate Relations California Institute of Technology 1201 E. California Boulevard Pasadena, CA 91125 Mr. Stephen D. Bechtel, Jr. Chairman Bechtel Group, Inc. P.O. Box 3965 San Francisco, CA 94119 Prof. A.I. Bienenstock Department of Materials Science Stanford University Stanford, CA 94305 Dr. Bruce Billings President Thagard Research Corporation 2712 Kelvin Avenue Irvine, CA 92664 Prof. Michel Boudard Department of Chemical Engineering Stanford University Stanford, CA 94305 Dr. Roger D. Bourke Jet Propulsion Laboratory 4800 Oak Grove Drive Pasadena, CA 91103 Mr. Peter Cannon Staff Vice President Rockwell International 1049 Camino Dos Rios Thousand Oaks, CA 91360 Prof. Majorie C. Caserio Department of Chemistry University of California Irvine, CA 92717 Dr. Thomas J. Clough Atlantic Richfield Company 515 South Flower Street Los Angeles, CA 90071 107 108 EXXON ENERGY R8D SYMPOSIUM Mr. Lee H. Collegeman Boeing Aerospace 1701 W. Charleston Las Vegas, NV 89102 Dr. Thomas B. Cook, Jr. Sandia National Laboratories Albuquerque, NM 87185 Mr. Floyd L. Culler, Jr. President Electric Power Research Institute 3412 Hillview Avenue Palo Alto, CA 94303 Prof. Paul E. Damon Laboratory of Isotope Geochemistry University of Arizona Tucson, AZ 85724 Dr. Frederic deHoffmann PO. Box 1814 LaJolla, CA 92037 Dr. Brewster C. Denny Dean Graduate School of Public Affairs Washington University Seattle, WA 98195 Prof. Philip F. Dickson, Head Chemical Petroleum-Refining Engineering Department Colorado School of Mines Goldeen, CO 80401 Dr. Paul F. Donovan Review Critique 277 Sheldon Avenue Santa Cruz, CA 95060 Dr. David L. Douglas Technical Manager Electric Power Research Institute P.O. Box 10412 Palo Alto, CA 94303 Dr. Richard Drobnick Center for Futures Research Graduate School of Business Admin. University of Southern California Los Angeles, CA 90007 Mr. Richard Elkus Pacific Measurement 488 Tasman Drive Sunnyvale, CA 94086 Dr. Thelma A. Estrin School of Engrg. Applied Sciences University of California Los Angeles, CA 90024 Prof. Robe? Eustis Stanford University Stanford, CA 94305 Mr. Ersel A. Evans Vice President Westinghouse Hanford Company P.O. Box 1970 Richland, WA 99352 Mr. Manuel Fernandez Zilog 10460 Bubb Road Cupertino, CA 95014 Dr. Christopher S. Foote Department of Chemistry University of California Los Angeles, CA 90024 Dr. Robert Freidheim Associate Director Marine Policy Studies University of Southern California University Park, CA 90007 Prof. Theodore H. Geballe Department of Applied Physics Stanford University Stanford, CA 94305 SYMPOSIUM ATTENDEES 109 Dr. John 0. Golden Dean of Graduate Studies 8: Research Colorado School of Mines Golden, CO 80401 Prof. Martin Gouterman Department of Chemistry University of Washington Seattle, WA 98195 Mr. Grant L. Hansen President SDC Systems Group 2500 Colorado Avenue Santa Monica, CA 90406 Dr. Joseph Harris Department of Chemistry Arizona State University Tempe, AZ 85281 Dr. C.E. Helms Department of Chemical Engineering Stanford University Stanford, CA 94305 Prof. George Hill University of Utah Salt Lake City, UT 84112 Mr. Charles J. Hitch Lawrence Berkeley Labs University of California Berkeley, CA 94720 Mr. C. Lester Hogan Fairchild Camera and Instrument Corporation 464 Ellis Street Mountain View, CA 94042 Dean Russell M. Holdredge College of Engineering - Utah State University Logan, UT 84322 Prof. George M. Homsy Department of Chemical Engineering Stanford University Stanford, CA 94305 Mr. Cuthbert C. Hurd Solar Energy Research Associates 332 Westridge Drive Portola Valley, CA 94025 Prof. M.L. Jackson Department of Chemical Engineering University of Idaho Moscow, ID 83843 Prof. William L. Jolly Department of Chemistry University of California Berkeley, CA 94720 Mr. Jack Kahn Lawrence Livermore Laboratory Box 808 Livermore, CA 94550 Mr. Donald Keach Deputy Director Institute for Marine 8: Coastal Studies University of Southern California University Park, CA 90007 Prof. Herbert B. Keller Department of Applied Mathematics California Institute of Technology Pasadena, CA 91125 Mr. H. Kenneth Lake, Jr. The Diebold Group, Inc. 4500 Campus Drive, Suite 348 Newport Beach, CA 92660 Dr. Leonard Laster President Health Sciences Center of the University of Oregon Portland, OR 97201 110 EXXON ENERGY RliD SYMPOSIUM Mr. Frederic Leder Occidental Research Corporation P.O. Box 19601 Irvine, CA 92713 Prof. J. lvan Legg Department of Chemistry Washington State University Pullman, WA 99164 Mr. I. Leibson Vice President Bechtel Group, inc. P.O. Box 3965 San Francisco, CA 94119 Dean George W. Lucky College of Engineering New Mexico State University Box 3449 Las Cruces, NM 88003 Mr. Robert J. Mackin Jet Propulsion Laboratory 4800 Oak Grove Drive Pasadena, CA 91103 Dr. Michael May Lawrence Livermore Laboratory Box 808 Livermore, CA 94550 Prof. C.B. Moore Department of Chemistry University of California Berkeley, CA 94720 Dean William M. Mueller VP for Academic Affairs Colorado School of Mines Golden, CO 80401 Dr. William A. Nierenberg Director Scripps Institution of Oceanography University of California LaJolla, CA 92093 Dr. Arlan D. Norman Chairman, Department of Chemistry University of Colorado Boulder, CO 80309 Dr. Arvid Pardo Senior Research Scientist institute for Marine 8: Coastal Studies University of Southern California University Park, CA 90007 Prof. Donald Pederson Department of Electrical Engineering and Computer Sciences University of California Berkeley, CA 94720 Dr. John R. Pierce Division of Engineering California Institute of Technology 1201 E. California PasadenaPinson Dean, School of Engineering San Jose State University San Jose, CA 95192 Dean Karl Pister University of California 315 McLaughlin Hall Berkeley, CA 94720 Dean Richard C. Potter School of Engineering California State University Long Beach, CA 90840 SYMPOSIUM ATTENDEES 111 Prof. Calvin F. Quate Applied Physics and Electrical Engineering Stanford University Stanford, CA 94305 Mr. Richard Rasmussen, President Pacific Measurements 488 Tasman Drive Sunnyvale, CA 94086 Prof. Judy Rayment Department of Materials Science University of California Berkeley, CA 94720 Mr. Delmar R. Raymond Weyerhauser Company Tacoma, WA 98401 Mr. Cloyd P. Reeg President Union Oil Company of California 376 South Valencia Avenue Brea, CA 92621 Dr. William C. Reynolds Chairman, Mechanical Engineering and institute for Energy Studies Stanford University Stanford, CA 94305 Dr. Donald B. Rice, Jr. President Rand Corporation 1700 Main Street Santa Monica, CA 90406 Prof. Gene l. Rochlin University of California Berkeley, CA 94720 Mr. Peter U. Rodda California Academy of Sciences Golden Gate Park . San Francisco, CA 94118 Mr. William Salesky University of California Berkeley, CA 94720 Prof. W.R. Salzman Department of Chemistry University of Arizona Tucson, AZ 85721 Prof. Charles Sanders California State University Long Beach, CA 90840 Mr. John Sellars TRW Defense Space Systems Group One Space Park Redondo Beach, CA 90278 Mr. Art Spaulding President Western Oil Gas Association 727 West 7th Street Los Angeles, CA 90017 Mr. Robert Spinrad Xerox Research Center 3333 Coyote Hill Road Palo Alto, CA 94304 Mr. E.E. Spitler Vice President Chevron Research Company 576 Standard Avenue Richmond, CA 94802 Mr. R.J. Stegmeier Union Oil Company of California 376 South Valencia Avenue Brea, CA 92621 Dr. Richard Steiner Department of Chemistry University of Utah Salt Lake City, UT 84112 112 EXXON ENERGY 3&0 SYMPOSIUM Mr. D. R. Stephens Lawrence Livermore Laboratory Box 808 Livermore, CA 94550 Dean Koehler Stout Engineering Division Montana College of Mineral Science and Technology Butte, MT 59701 Prof. Robert L. Street Stanford University Stanford, CA 94305 Mr. T.L. Stringer Tosco Corporation 10100 Santa Monica Boulevard Los Angeles, CA 90067 Dr. B. Samuel Tanenbaum Harvey Mudd College Claremont, CA 91711 Dr. Klaus Timmerhaus Engineering Center AD 1-25 University of Colorado Boulder, CO 30309 Dr. Don Walsh, Director Institute for Marine and Coastal Studies University of Southern California University Park Los Angeles, CA 90007 Dr. R.F. Weiss Scripps Institution of Oceanography LaJolla, CA 92093 Mr. John G. Welles Colorado School of Mines Golden, CO 80401 Dr. Edward A. Wenk, Jr. University of Washington Seattle, WA 98195 Mr. Glen Werth Lawrence Livermore Laboratory Box 808 Livermore, CA 94550 Mr. John R. Whinnery Department of Electrical Engineering and Computer Sciences University of California Berkeley, CA 94720 Mr. Robert L. Wiegel University of California Berkeley, CA 94720 Dean J. Richard Williams University of Idaho Moscow, ID 83843 Prof. Lotfi A. Zadeh University of California Berkeley, CA 94720 Proceedings of the Exxon Energy 3&0 Symposium Edited and Published by BEN H. WEIL and DAVID K. JOHNSON Designed by BEN H. WEIL, RAY D. MORITZ and WILLIAM REUTER, JR. Illustrated by DONOVAN AND GREEN Composed by EXXON RESEARCH AND ENGINEERING COMPANY and J. SCHILLER, INC, in OPTIMA LIGHT, with Display Lines in HELVETICA BOLD Page Makeup by PRODUCTION GRAPHICS Offset Printed by HOWARD PRESS on SUNRAY Vellum Opaque Natural, 80# Bound by A. HOROWITZ SONS in Yale Blue LEXOTONE, Pin Morocco Finish, with Gold Stampings on Blind Panels