This is the first in-depth presentation in book form of current analytical methods for optimal design, selection and evaluation of instrumentation for process plants. The presentation is clear, concise and systematic-providing process engineers with a valuable tool for improving quality, costs, safety, loss prevention, and production accounting.
From Chapter 1 Introduction
"Instrumentation is needed in process plants to obtain data that are essential to perform several activities. Among the most important are control, the assessment of the quality of products, production accounting… and the detection of failures related to safety. In addition, certain parameters than cannot be measured directly, such as heat exchanger, fouling or column deficiencies, are of interest. Finally, new techniques, such as on-line optimization, require the construction of reliable computer models for which the estimation of process parameters is essential.
"This book concentrates on the tasks of determining the optimal set of measured variables and selecting the accuracy and reliability of the corresponding instruments. The goal is to obtain sufficiency accurate and reliable estimates of variables of interest while filtering bad data due to possible instrument malfunction. An additional goal is to observe and diagnose single and multiple process faults."
From the Preface
"There is a vast amount of literature devoted to the selection and good maintenance of instruments. This literature covers the selection of the right instrument for a particular range and system, but only after the desired accuracy and reliability of measurement have been established. Little has been written on how to systematically determine the right accuracy and reliability needed when selecting an instrument, much less how much redundancy is needed for a particular system. The key variables that needed estimation come from control requirements, as well as monitoring needs for safety, quality control and production accounting. These are the starting points of the design methodology. This book concentrates on determining the optimal accuracy and reliability of instruments and their location. To determine this, certain desired properties of the system of instruments are used as constraints while the cost is minimized. These properties, among others are variable observability, system reliability and precision of certain variables.
"This book is not a textbook. Rather it is intended to be an organized collection of the most relevant work in this area…. It has been written with the intention of making it readable by engineers with some background in linear algebra, mathematical optimization and graph theory. It is organized so that the complexity of the sensor network design is addressed step by step."
The information in this new book serves the needs of chemical and other process engineers involved in instrumentation and control, maintenance, plant operations, process design, process development, quality control, safety, and loss prevention.
Illustrations and Tables
The text is supplemented with more than 100 flow charts, diagrams and other schematics that illustrate procedures, systems and instrumentation. More than 70 tables provide useful reference data.
Dr. Miguel J. Bagajewicz brings to this new book his extensive experience in design, data management, teaching and writing in the area of process engineering. He received his M.S. and Ph.D. in Chemical Engineering from the California Institute of Technology. He is presently Associate Professor, School of Chemical Engineering and Materials Science, and Director, Center for Engineering Optimization at the University of Oklahoma. He is the author or co-author of more than 100 journal articles, conference presentations, and reports, and the author of articles on data reconciliation and sensor location in the Instrument Engineers' Handbook, fourth edition. He is a member of the American Institute of Chemical Engineers (AIChE), and on the executive committee of the Central Oklahoma Chapter.
1. Plant Data Management o Introduction o Plant Information and Operations Management o Model-Based Monitoring o Quality of Data
2. Instrumentation Design Goals o Introduction o Measured and Key Variables o Selection of Monitoring Variables o Selection of Key Variables in Control o Selection of Measured Variables for Fault Diagnosis o Instrumentation Design Goals o Upgrading of Instrumentation
3. Instrumentation o Introduction o Flow Rate Instrumentation o Level Measurement o Temperature Measurement o Pressure Measurement o Density Measurement o On-Line Process Analyzers o Transmission and Transformation of Signals
4. Errors in Measurement o Introduction o Instrument Properties o Measurement Quality o Sensitivity and Speed of Response o Hysteresis and Dead Band o Calibration Curves o Accuracy of Different Instruments
5. Variable Classification o Introduction o Model o Measurement Equation o Graphs and Flowsheets o Connectivity of Systems o Observability o Redundancy o Linear Systems o Canonical Representation of Linear Systems o System Degree of Redundancy o Quantification of Observability and Redundancy o Graphs and Canonical Matrices o Nonlinear Systems
6. Design and Upgrade of Nonredundant and Redundant Sensor Networks o Introduction o Upgrade and/or Design Goals o Design for Estimability o Design for Estimability Efficiency o Compulsory Measurements and the Upgrade Case o Sensor Networks for Bilinear Systems
7. Data Reconciliation o Data Reconciliation o Background o Linear Data Reconciliation o Steady-State Linear Data Reconciliation o Nonlinear Steady-State Data Reconciliation o Dynamic Data Reconciliation
8. Design of Accurate Sensor Networks o Introduction o Cost-Optimal Design o Multiple Instruments and Hardware Redundancy o Maximum Precision Models o Generalized Maximum Precision Model o Relation Between Sensor Network Models o Solution Procedures for Linear Systems o Parameter Estimation in Nonlinear Systems
9. Precision Upgrade of Sensor Networks o Introduction o Upgrade Options o Cost Benefit Analysis o Upgrade Models Based on Addition of Sensors o Model for Resource Reallocation o Generalized Model for Resource Reallocation and Upgrade
10. Reliability of Nonrepairable Sensor Networks o Introduction o Sensor Service Availability o Sensor Service Reliability o Failure Density and Failure Rate o Markovian Model o Mean Time to Failure o Estimation Availability and Reliability of Variables o Determination of Estimation of Reliability o Estimation Reliability in Nonredundant Systems o Availability, Reliability and Degree of Estimability o System Availability and Reliability
11. Design of Reliable Linear Nonrepairable Sensor Networks o Introduction o Nonredundant Networks Featuring Maximum Reliability o Redundant Networks Featuring Maximum Reliability and Hardware Redundancy o Redundant and Restricted Networks
12. Design of Reliable Bilinear Nonrepairable Sensor Networks o Introduction o Bilinear Multicomponent Systems o Energy Networks
13. Design of Reliable and Cost-Efficient Nonrepairable Sensor Networks o Introduction o Minimum Cost Model o Minimum Number of Sensors Model o Solution Procedure o Relation to Other Models o Limitations of Previous Models o Generalized Maximum Reliability Model
14. Design of Repairable Sensor Networks o Introduction o Failure Intensity o Repair Intensity o Expected Number of Repairs o Maintenance and Total Cost o Residual Precision o Minimum Cost Model
15. Design of Robust Sensor Networks o Introduction o Origin of Gross Errors o Gross Error Handling o Test for Gross Error Presence o Gross Error Detection in Dynamic Data o Reconciliation o Inaccuracy in Gross Error Detection o Multiple Gross Error Identification o Gross Error Size Estimation o Sensor Network Error Detectability o Sensor Network Gross Error Resilience o Robust Sensor Networks o Minimum Cost Model for Robust Networks
16. Genetic Algorithms o Introduction o Genetic Algorithms
17. Design of Sensors for Process Fault Diagnosis o Introduction o Fault Detection, Diagnosis and Alarms o Fault Observability o Fault Resolution o Sensor Network Design o Sensor Location for Fault Observability o Sensor Location for Fault Resolution Index