Technologies
- 1. Some Interesting Technologies I Have Worked on or Consulted on lately: “The way to success is no longer knowledge and information only, it is experience and insight into information.” ANALYSIS OF NON-LINEAR DEPENDENCIES AND FACTORS IN NETL DATABASE OF COAL-FIRED POWER PLANTS Further to our review of the NETL report entitled: “Reducing CO2 Emissions by Improving Efficiency of the Existing Coal-fired Power Plant Fleets”, DOE/NETL 2008/1329, by Chris Nichols, et. al, July 23, 2008, REDUCT and Lobbe Technologies undertook an analysis of non-linear relationships between the plant (boiler and generator) efficiency and plant design factors listed in the database. The preliminary findings of the analysis are as follows: 1. Similar to the NETL analysis, a significant number of factors were found to affect plant performance, many of which are redundant (cross-correlated) and perhaps spurious (Table 1). However, in contrast to NETL’s findings, REDUCT and Lobbe found that the key variables defining plant efficiency depend, in many instances, on the plant design. 2. The central conclusion of the NETL report, that the observed variance in plant performance is related primarily to the plant’s operating procedures, rather than to plant design, is incorrect. Our analysis identified clusters (patterns) of plant data that characterize similar performance and design factors, and the performance of plants with these characteristics is determined by a narrower range of boiler and generator efficiencies. 3. Within each cluster of plants (patterns), operating variables have a different effect on plant performance than is the case for plants which can not be characterized by clear design patterns. This will provide information on what are the key factors in terms of efficiency improvement and reduction of CO2 emissions and fuel input for different plants. 4. Identification of the clusters of plants constrained by design and operating factors is important, therefore, in designing realistic targets for plant improvements. For example, for stokers fired with fuels which have low heating value, the standard deviation of boiler and generator efficiencies is half of those for “opposed fired” boilers, meaning that more improvement can be expected for the opposed fired boilers through optimization of their performance and operations. 5. The analysis clearly indicated that there are patterns of plant performance data (clusters of plants), which are determined by the plant design and operating constraints. However, the patterns are non-linear in their characteristics and do not
- 2. Some Interesting Technologies I Have Worked on or Consulted on lately: correspond to clusters of plants identified as having the 10% highest and 10% lowest efficiencies in the database that provided the foundation for the NETL report. Identification of different plant clusters and recognition of the extent to which their efficiency can be improved through optimization of operations provides useful and important information regarding the extent to which plant efficiency can be improved and carbon dioxide can be reduced. It will provide information to Canadian plant managers what is the best performance for plants of similar characteristics and input in North America. It will also help us identify the plant characteristics that are important in the design of the next generation coal-fired plants. The results presented in Figures 1 show selected examples of boiler efficiency patterns. By pointing the directions there improvement can be made, utilities have opportunities to reduce CO2 while saving billions of dollars. Figure 1. EXAMPLES OF PATTERNS FOR BOILER EFFICIENCY Boiler Efficiency (Primary Fuel) at 100% Load (%) Opposed Stoker BOILER EFFICIENCY, % 100 95 90 85 80 75 0 200 400 600 800 1000 1200 BOILER
- 3. Some Interesting Technologies I Have Worked on or Consulted on lately: If you open up the mind, the opportunity to address both profits and social conditions are limitless –Jerry Greenfield Heavy Oil or Oil Sands The present oil sands/bitumen recovery and upgrading technologies follow a well established route to bitumen upgrading. They are based on a simple understanding of the properties of heavy oil/bitumen, i.e., bitumen becomes less viscous at higher temperatures. A significant part (~50%) of bitumen can be removed by atmospheric and vacuum distillations; to achieve higher conversion bitumen has to be converted to lighter crude by coking and/or the addition of hydrogen. Most conventional bitumen upgrading processes are based on the simple understanding of the properties of bitumen described above. These processes differ from each other mainly in selection of the energy medium (steam, combustion gases, solvent); and in selection of processing temperature, type of catalyst and hydrogen pressure. The relative advantage of each process (higher efficiency) is defined in relation to operational characteristics such as better loop control tuning and better maintenance, or in relation to reduced operational mistakes (e.g. preventing upgrader fires). The co-processing technology is different from the approaches described above in a number of ways. The primary objective of co-processing is separation and conversion of heavy bottom components of bitumen such as asphaltenes. Firstly, co-processing recognizes that converting asphaltenes into distillable oils is the key step to increasing the effectiveness and profitability of bitumen upgrading. Secondly, co-processing is based on advanced knowledge of coal science, i.e., the role that some coal molecules can play in removal of sulfur, nitrogen, and oxygen from both bitumen asphaltenes and coal. Figure 1. 3rd GENERATION BUTUMAN PRODUCTIOIN AND UPGRADING BITUMEN UPGRADING SYNTHETIC PRODUCTION CRUDE OIL MARKETS CO-PROCESSING NORTH COAL AMERICA MINING OVERSEAS BY-PRODUCTS
- 4. Some Interesting Technologies I Have Worked on or Consulted on lately: This knowledge comes from research on the chemistry of coal tar formation, and coal hydrogenation and conversion to liquid fuels. Co-processing is based on an understanding of the hydrogen-donating capacity of coal hydroaromatic molecules, and the catalysis-like role that coal mineral matter plays in the process. This eliminates the need for costly high-pressure hydroprocessing of bitumen heavy bottom components. In summary, co-processing offers a new and much more effective alternative for processing heavy oi/bitumen deposits. The process focuses on synergisms offered by integration of bitumen and coal co-processing. It removes existing barriers to utilization of coal and bitumen, and addresses the main environmental and export constraints resulting from poor resource quality. It offers the owners new strategic opportunities to enhance the value of their resource holdings.
- 5. Some Interesting Technologies I Have Worked on or Consulted on lately: Automated Fault Detection and Diagnosis Systems for Industrial Process Energy and Performance Improvement Executive Summary Automated fault detection and diagnosis systems (AFDDS) are computer software and hardware used to identify and manage abnormal system operations including dysfunction and/or malfunction of the system parts. In the last 10 - 15 years, AFDDS have been widely used in military, aerospace, automobile, nuclear, and chemical industry sectors to increase safety and reliability in these sectors and to enhance maintainability and availability of their operations. However, as emphasized in this report, relatively few AFDDS have been applied in the pulp and paper sector and the petroleum sector – here jointly defined as the Target Sectors. This report recommends, therefore, to support a number of actions and technology enabling strategies, which will allow Canadian industry to benefit from AFDDS technologies. The actions and strategies recommended include: 1. Specific barriers to AFDDS are addressed by proposing that CETC establish an AFDDS Network Group in order to provide industry with the needed knowledge and training in support of AFDDS. The goal is to strengthen and consolidate Canadian capabilities in automated fault detection and diagnosis systems. 2. The lack of AFDDS developments in the Target Sectors is addressed by proposing a phase-in approach to support proof-of-concept demonstrations of simple AFDDS infrastructure and implementation methods (learning by doing). The goal is to reduce the cost and increase the capacity of AFDDS. 3. The research goals and objectives recommend for universities call for adding value to AFDDS applications such as prognosis capabilities, automated updating, etc. (learning by research). The goal is to improve the cost/benefit ratio of the AFDDS that are implemented. We list below reasons for the above recommendations. Less than 50 systems addressing fault detection and diagnosis in the Target Sectors were identified in the course of this project, with more than two-thirds of the systems being prototypes, demonstrations or university studies, worldwide. Because there are so few applications of AFDDS in the Target Sectors, key barriers to the implementation of AFDDS were examined for technical and scientific constraints, and for constraints related to industrial practices and operations. It was concluded that, although there is always some room for improvement of AFDDS technology, the key barriers to adoption of AFDDS in industry are not lack of scientific and engineering knowledge, but rather industry- related constraints such as financial, human, or organizational factors, and the complexity of the operations in the Target Sectors.
- 6. Some Interesting Technologies I Have Worked on or Consulted on lately: Although many AFDDS developers and universities believe that large improvements could be made in implementation of AFDDS if higher levels of support for the technology were provided by industry management and operating personnel, industry points out rightly that their adoption of any new technology is constrained by limited resources, pressures to increase profitability, the need to respond to environmental and regulatory requirements, and reduced numbers of expert personnel due to plant closures, to name a few examples. Plant resources are first allocated to important strategic investments irrespectively of the advantages offered by AFDDS, or any other technology applications. As stressed in the proposed AFDDS adoption criteria, identifying AFDDS applications as a strategic investment is the surest way to promote the implementation of new AFDDS. As matters stand today, there is a large gap between work in universities and industrial practice – a gap created by the fact that universities and industry have different goals and different priorities. This gap was carefully considered when the AFDDS adoption criteria in this report were developed and when recommendations were made. One of the key findings of this research project is the high value of AFDDS in the case studies examined. There is a misconception in industry and the research community about the high value of automated control and optimization technologies as compared to the value of an investment in automated fault detection, diagnosis and prognosis technologies. This misconception results in more investment in the area of process control and optimization and less investment in AFDDS. While automated control and optimization can provide an improvement of operations’ efficiency and availability on the order to 2-3 percent; in contrast, AFDDS can provide improvements on the order of 3-5 percent by going beyond fault avoidance and better maintenance, and providing the ability to: 1) handle large process disturbances; 2) explain and correct behavior of controllers in order to meet planned targets; 3) provide immediate advice on alarm management including early detection of problems, etc. The case studies reviewed show that the benefits of AFDDS in the Target Sector are on the order of $ 0.5 to $ 5 million dollars per system/plant. For the Canadian pulp and paper sector, the potential benefits are on the order of $50 to 80 millions from the reduction of maintenance and outages costs, and the addition benefits of $70 to 110 millions from energy savings. The total benefits range from $120 to $190 million dollars or an “average” $2 to $3 million dollars per plant, depending how an “average” plant is defined. This is in line with the actual benefits from implementing AFDDS as reported by companies like Norske Skag. The payback for the majority of AFDDS implemented in the pulp and paper sector is less than two years, with many systems paybacks of less than eight months. For the Canadian petroleum sector, the potential benefits from the reduction of maintenance and outages cost are about $76 millions; additional benefits from energy efficiency savings are $70 to $100 million, for total benefits ranging from $146 to $176 millions or an “average” $2 to $4 million dollars per plant and a payback period of less
- 7. Some Interesting Technologies I Have Worked on or Consulted on lately: than a year. This is in line with potential benefits of $4 to $6 million dollars per plant that have been reported by US studies of AFDDS applications in refineries. Why is it then that despite large benefits and excellent paybacks so little AFDDS capacity has been implemented in the Target Sectors? Some of the answers to this question lie in the barriers and constraints to the implementation of AFDDS cited above, but there are also other considerations. The significant barriers to the development of AFDDS products (required knowledge of the Target Sectors operations, plant control and automation, information technologies and software, advanced data analytics, plant-wide information systems, etc.) and the small size of the AFDDS markets in the Target Sectors (< $ 40 millions) discourage AFDDS vendors from commercializing products for the Target Sectors. The situation is critical because AFDDS technologies are needed in the Target Sectors more than ever before. They are needed in the pulp and paper sector because of the sector’s poor economic state, and in the petroleum sector because of a critical need to reduce GHGs emissions by refineries. It is important, therefore, that the right steps be taken by Canadian industry to increase utilization of AFDDS. Different strategies are required to effectively support different types of technologies at different stages of growth, e.g., emerging, evolving, reviving and mature technologies. There is a general lack of understanding, however, of the management of technology growth including the best ways to assist commercialization, and the role that government can/should play in reducing risks and improving the odds of success. While research and development funding organizations like PRECARN are effective in supporting emerging technologies developed at universities, and organizations like FPInnovation and Petroleum Technology Alliance Canada (PTAC) are effective in supporting reviving technologies developed in industry/university partnerships, technologies like AFDDS in the Target Sectors need a different approach and require the establishment of a mechanism that connects government laboratories, technology vendors and industry. This report seeks, therefore, to identify policies to support growth of AFDDS that use “learning curves” as a measure of potential and available best development approaches to increase AFDDS capacity in the Target Sectors. Based on the characteristics, applications, barriers and documented progress of AFDDS, this report classifies AFDDS in the Target Sectors as evolving technologies (have been utilized for a shorter period of time and have experienced improvement during that time), which indicates a strong potential to increase AFDDS technology capacity through ‘learning by doing” and “learning by research.” This study also shows that a simple cost/benefit analysis of AFDDS does not provide sufficient insight into how to select the best AFDDS applications or design AFDDS support programs, i.e., there are many AFDDS applications with a high payback that have not been implemented. Broader adoption criteria which consider the presence of niche applications and the need for strategic investment by industry provide much better measures and guidance for development of AFDDS support programs. For example, 70 percent of AFDDS installed in the petroleum sector are used for Fluid Catalytic Cracking
- 8. Some Interesting Technologies I Have Worked on or Consulted on lately: units and almost half of the systems installed in the pulp and paper sector detect and diagnose paper quality. Looking for and exploring such niches offers direction regarding the most attractive opportunities for AFDDS in the Target Sectors. Finally, the small AFDDS market in the Target Sectors is addressed by recommending that the proposed new CETC’s program includes Industries-of-the-Future technologies, i.e., integrating automation, informatics, robotics, and AFDDS technologies. The goal is to decrease both industries energy use by over 133 PJ per year and GHGs emissions by over 8 MT per year, at less than $10 per ton of CO2 reduced. The potential is exciting. Based on industry experience of 4 to 1 payback from automation technologies, the proposed CETC program alone will likely result in over $ 40 million worth of benefits to companies in the Target Sectors.