Marion A. Keyes

Multi-Level Control and Optimization Methods for Desalination Plants

Abstract This paper describes multi-level control and optimization techniques for desalination plants for improved product quality, productivity, and minimum cost of operation. Methods of implementation by distributed process controls are also discussed. Specifically, the conservation of energy and materials, minimization of equipment outages, and increase production rate are primary goals to be achieved. A dual-purpose combined-cycle power generation and distilled water production plant is considered. The power generation portion consists of multiple co-generation units, consisting of gas turbine generators, waste heat boilers, and steam turbines. The desalination portion operates on the distillation principle having multiple trains of multi-flash evaporators with recycling capability. In addition, a vapor compression cycle in evaporative distillation is included to describe the opportunities in this type of desalination. A multi-level control system is considered in three levels: (i) dedicated sequential and regulatory controls; (ii) supervisory and optimizing controls; (iii) management information and control.

Methods of energy efficient control and optimization for combined-cycle cogeneration

Abstract This paper discusses the fundamentals and implementations of an energy efficient operation of various cogeneration arrangements. These arrangements include industrial and commercial energy users with in-plant generation facilities, both trading with electric utilities. A multi-level control and optimization approach is followed. Instrumentation and delicated feedback controls are presented at the first level. Supervisory and optimizing controls providing the load values (set points) of generating units are discussed as a second level hierarchy. Coordination with the utility is discussed at the third level within the purchase/sale formula. Implementation by distributed control systems is discussed. The functions performed at each level are described, including the control hardware. The features of the equipment and advantages of multilevel control system are presented as related to particular applications.

Kappa Number Estimation, Control and Management in the Pulping Process

The primary goal of this work is to develop a uniform method and implementation for the control of pulping reactions. Specifically, the desired pulp Kappa number and black liquor residual alkali are provided for, under the various operating conditions of different mills. A general model is developed which accommodates parameter estimation for different charging and cooking procedures. Having physical meaning, these parameters are measurable and readily available from the mills process knowledge and information base. For this general and multivariable model, a supervisory control strategy is developed to maintain the desired Kappa number and residual alkali concentration. This is accomplished by coordinating and simultaneously setting of the two primary reaction inputs, namely, the initial alkali charge and cook time/temperature targets. Prior to implementation of any supervisory control scheme, measures are first taken to stabilize the underlying process. Actual implementation of the controller is then done on an incremental basis. This design is highly manageable and provides good performance despite significant modelling error. Further, the system also incorporates on-line statistical process control analysis for real-time closed loop actuation. Implementation of the control stratgegy is discussed including the distributive microprocessor algorithms. The supervisory pulp quality controller functions as the highest level of an integrated pulp mill control system coordinating with other elements to optimize process production, yield and efficiency. In total, the system provides a cost-effective means of realizing significant process economic benefits and, consequently, a high rate of return on investment.

Self-tuning and Self-adaptive Control of a Heating Process by Distributed Microprocessors

Abstract This paper investigates and evaluates two different control approaches, namely: (i) Self-adaptive discrete control and identification techniques based on Z-transformation, and (ii) Self-tuning continuous control and parameter estimation techniques, both implemented by microprocessor based distributed control system, The process is a heating equipment designed to deliver the heated medium (water) at a specified temperature from a tank under load (flow) variations. The heating of the water is done in an enclosure before entering the tank whose temperature is controlled. Flow fluctuations through the tank and level fluctuations within the tank are imposed and the algorithms are evaluated for comparison. Recursive least square methodology with a variable “forgetting factor” is used for identification. Model-based parameter estimation techniques and measurements are utilized to design the self-tuning controls based on continuous system approaches. A self-tuned Smith predictor along with a self-tuning PID control were used. The effect of sampling time in both algorithms is discussed. The chosen implementation is with a distributed microprocessor system. The system used has both discrete, and continuous algorithm forms available as configurable “function blocks” (functional programming).

Evolution of Adaptive Control Algorithms and Products: A Critical Review and Evaluation

Abstract Adaptive control literature exceeds far beyond the respective product development activities except the recent years. The justified credit for product goes to distributed controls resulted from microprocessor developments In this work, the historical perspectives on adaptive algorithms and product developments including the current status of implementations by distributed controls are reviewed Recent product developments, the suitability and the need of distributed controls are des-cribed. The need of combining the theory, application knowledge, and physical in-sight for a successful product is presented. Adaptive control products marketed as a part of total system and as a single product are discussed and evaluated. As current challenges, the need of using artificial intelligence, pattern recognition, learning systems, and neural network structure for providing selective and intelligent algorithms is emphasized.

Hierarchial Language Processing

Abstract Since the earliest applications of general-purpose digital computers to process control in the late 1950s and early 1960s, the need to establish a more efficient means of designing, programming, and debugging application and systems software has been apparent. The use of higher level languages to satisfy this need has accelerated. A proliferation of languages and dialects has resulted with little general agreement on any language form which is universally suitable for all applications. The consideration of application suitability is further compounded by the advent of the microprocessor with its own unique set of restrictions and capabilities. The continued trend toward more powerful control processors at greatly reduced cost has been completely obscured by the honerous cost of software development. The challenge today is to reverse this trend. We believe that the divergent language requirements arise from a failure to separate the communication level of the language processor from the implementation level. A hierarchial language structure is proposed which accomplishes this separation. Utilizing this structure, a single language is delin-eated allowing the user to optimize the implementation specifically for this application.

A novel method and procedure for on-line measurement of fluid properties

Abstract This work describes an on-line method and procedure for calculating fluid properties in real time while the system is in operation. The method utilizes a distributed control system which gives the same results as computer programs, but without compilation, and in machine language. Thermodynamic relations of the fluid from tables, along with a fluid property formula, are imbedded into the proposed function blocks. The pressure and temperature measurements are provided by conventional transmitters. The proposed system provides other fluid properties, as well, in every sampling period, which can be as short as 20 ms.

Engineering Work Station as a Tool for a Process Control Design and Implementation

Abstract This paper describes the total role of a PC-based Engineering Work Station (EWS) in designing and implementing the control functions for a plant. The plant is controlled, monitored, and managed by a microprocessor-based distributed control system which interfaces with EWS. The steps and procedures of the work is described from control design to the plant start-up. During the design and implementation phase EWS was also used as a personal computer and many off-line calculations were performed. The plant for the work was a chiller system which is part of the power house supplying the steam and chilled water into the chemical and polymer processing plant as well as buildings for heating and cooling. The role of EWS as a significant control design and implementation tool, which received wide acceptance and user’s appeal, has been demonstrated. The main advantages of EWS have been: (i) the ability to configure and edit all control system configurations before installation is complete, (ii) to have on-site permanent drawings of control logics at any time, (iii) to integrate various control symbols and load them onto control modules by operator-friendly softwares, (iv) to monitor the process operation for diagnosis and troubleshooting as a support to the user.

Energy Management for Process Industry: How Much Control is Needed?

Abstract This paper provides an overall and unified knowledge on industrial energy management applicable for a broad base of industrial plants. Energy conservation and management opportunities are presented for various processes and equipment. Energy savings are discussed on particular power equipment such as: boiler, turbine, utility tie-line, and chiller which are common for process industry. Furthermore, few energy intensive processes are included to demonstrate the significant energy conservation opportunities. The methods for energy savings are described, which generally are: improved feedback controls; supervisory and optimizing controls; and coordinating controls. Considerations in implementing the savings by distributed controls including the hardware and software requirements are discussed. In conclusion, this paper is a survey of industrial energy management on fundamentals, methods, and implementation to reflect the current status of technology.