This research is focused on the simulation, optimisation and synthesis of utility systems, in particular combined heat and power (CHP) systems. These systems involve gas and steam turbines, steam generation at different pressure levels, condensing equipment and auxiliaries. The CHP systems are of substantial industrial interest for the efficient supply of heat and power. As they are highly integrated processes their size and implicit complexity requires the use of an equation oriented (EO) framework including the models developed in this research. An EO mathematical model for the simulation, optimisation and synthesis of CHP systems has been developed. It includes models for the simultaneous solution of all process streams, every major piece of equipment and investment and operating costs. Several EO simulation examples, from simple unit operations, a whole real cogeneration plant involving a commercial gas turbine with 1275 variables and equations up to a synthesis model with 3042 variables for a fixed structure, are used to demonstrate the applicability of the CHP model and the EO framework. A number of energy and economic optimisation problems were solved using a SQP (Sequential Quadraic Programming) method. Both the EO model and the use of the SQP code was fully explored by experiments in a model with two steam turbines. In addition the following utility systems were optimised: a model of a combined heat and power plant; industrial size problems including a model for a cogeneration plant currently in operation and a synthesis model for a fixed structure. An important contribution made to solve EO simulation problems for CHP systems was to obtain converged solutions to plant sections which then were used as starting guess for larger plant sections until whole systems are simulated. Also this strategy was used to provide a warm starting guess for the efficient solution of large continuous optimisation problems of CHP systems.

Direct drive renewable energy converters provide a low speed, high torque input to the electrical generator. As a result these generators must be larger than their high speed counterparts. Because of this size and the large airgap-closing force, the structural design must be stiff and robust. This typically results in heavy generators, with structural (‘inactive’) material dominating the electromagnetically ‘active’ material. Design tools are set out, validated and used to model the inactive material in high torque axial-flux and radial-flux type machines. Simple optimisations on generator aspect ratios are carried out to find lighter designs. Axial-flux and radial-flux permanent magnet synchronous machines for wind turbines are compared in terms of mass and a cost criterion, with and without the inactive mass. Some machines are designed in such a way that the normal force is nonexistent or significantly smaller than in conventional electrical generators. The design and modelling of a speed air-cored permanent magnetic machine is described. Discussions and conclusions highlight impacts on direct drive design philosophy.

This thesis describes research carried out by the author between 1986 and 1990 at the Department of Chemical Engineering, University of Edinburgh, under the supervision of Dr Colin Pritchard. The aim of the research was to devise and evaluate a novel compression heat pump cycle as a potential technology for the utilisation of work and heat from environmental sources in industrial applications. The principal requirement of such a heat pump is that it can accept time-varying inputs of work and heat whilst supplying a controlled heat load at constant delivery temperature to an external load. This requires a cycle with the ability to 'self-regulate' its capacity to match external variations in <i>either</i> energy input <i>or</i> energy takeoff. The use of a nonazeotropic mixture as working fluid offers the potential to vary the composition of circulating fluid in a heat pump cycle, thereby varying the capacity, and this was chosen as the basis for research. The prediction of thermodynamic properties of halogenated hydrocarbon refrigerants is reviewed with special emphasis on: acquisition of sufficient data for preliminary plant design from minimum information, and on the prediction of binary vapour-liquidequilibrium data (VLE) in the absence of experimental measurement. The Cubic Chain-Of-Rotators (CCOR) equation of state is assessed as a route to these properties; it is shown that this equation offers improved liquid-phase property prediction compared to other cubic equations of state. The CCOR equation is also shown (by comparison with experimental measurements from the literature) to predict binary VLE to the same degree of accuracy as the Redlick-Kwong-Soave, Lee-Kesler and Carnahan-Starling-DeSantis equations in the absence of any parameter optimisation. Procedures are described for the optimisation of the CCOR equations's performance as a predictor of both pure-fluid and mixture properties by the fitting of empirical parameters to experimental data. It is demonstrated that this optimisation procedure allows quantitative description of the liquid phase of pure fluids and of mixtures.

The digital displacement hydraulic piston pump is a hybrid device which combines a microcompressor with an established form of high-pressure pump to create a highly integrated machine which can produce a variable high-power output directly from an electronic command. The actively controlled inlet poppet-valve in each cylinder can be held open against the discharging flow in order to disable it during a single cycle. Cylinders can be disabled in this manner, following a maximally smooth sequence, allowing a controlled output flow to be achieved. A compliant device located near the pump, such as an accumulator, provides time-averaging of the flow pulsations in order to minimise the effects of the quantisation error caused by cylinder disabling. The advantages of this approach over the conventional variable-swash axial piston pump lie with both the response speed and the inherent energy efficiency of real-time cylinder selection. Disabling cylinders in this way restricts parasitic losses to very low levels since unused cylinders are not pressurised nor do they incur loads on their associated bearings. The response time of the pump is related to shaft speed, with the pump able to attain either full or zero output from any starting condition, in less than a single shaft revolution. At induction motor speeds this allows large-signal response times of the same order as those achieved by commercial proportional valves. The thesis chronicles the development of the Digital Displacement pump. It begins with the formulation of a simulation model which is able to predict the behaviour of the machine in both flow and pressure control modes. The valve control possibilities are then explored and the design of active valve latches using finite-element analysis described. The sinusoidal flow forces on the disabled poppet are evaluated through a large range of experiments and the results condensed into parametric equations useful for predicting the valve latching requirements of most machines. The mechanical and electronic hardware design, leading to the construction of the prototype, is then discussed.

This thesis addressesth e analysisa nd design of commercial power plants by using the Combined Pinch and Exergy Approach. Current practice in design for commercial power plants heavily relies on experience and computer simulation and lacks systematic design methodologies. On the contrary, Pinch Technology allows systematic and generic approaches to chemical process design in which targets are set prior to design. These approaches can address issues related to process integration and optimisation. This thesis exploits the analogy between power plant design and chemical process design and applies the philosophy of Pinch Technology to the field of power plant design. In this thesis, the "onion" model used to represent the hierarchy of chemical process design is applied to power plant design. This model decomposes the whole design problem into three relatively simple tasks, including turbine system selection, heat exchanger network (HEN) design and fuel supply determination. Complex interactions exist between these individual components. To describe the complex interactions between the different components, a qualitative tool called the Combined Pinch and Exergy Representation (CPER) has been developed. The CPER allows engineers to visualise the overall performance of a power plant and the interactions between components. This diagram can also help engineers to screen design options. A quantitative tool, called the shaftwork targeting approach, has been developed in this thesis to evaluate each possible design option and identify the most promising one ahead of detailed simulation and designA tool called the Exergy Remaining Problem Analysis (ERPA) has been developed to guide HEN design. This allows the design to achieve the shaftwork targets. By evaluating the impact of individual matches on the remaining problem, the ERPA can determine the influence of individual matches on shaftwork generation. By detecting inappropriate matches, the ERPA can ensure that the HEN design meets the shaftwork targets. Based on the "onion" model of power plant, a systematic and generic approach to power plant design has been developed. In this approach, power plant design starts with the turbine system, then moves to the heat exchanger network and the fuel supply. This approach is entirely general which can be applied for design of different power plants. The significance of the new approach is that it enables engineers to screen possible design options with physical understanding and identify the most promising design option ahead of detailed simulation and design. This speeds up the overall design process and ensures that an optimal solution is obtained. vi

No description available.

No description available.

No description available.

No description available.

No description available.