Fernando Javier D'Amato

Industrial application of a model predictive control solution for power plant startups

This paper reports on the development and a successful MPC implementation for startups of combined-cycle plants. Minimizing startup times is important for energy utility companies to reduce operating costs due to lower fuel consumption and lower emissions. The new controller regulates the gas turbine loads while keeping the main operating constraints (steam turbine stresses) within their allowable ranges.

Efficient algorithms for mistuning analysis

Abstract This paper describes two algorithms to calculate bounds on the largest frequency response function that is attained when a large number of parameters is perturbed simultaneously within a box. The algorithms have been created to calculate worst-case responses of mistuned (perturbed) bladed disks, which is an important step in the analysis and optimization of advanced turbomachinery components. Numerical comparisons with other algorithms demonstrate the relative efficiency and accuracy of the proposed algorithms.

New Tools for Analysis and Optimization of Mistuned Bladed Disks

This paper describes methods to calculate the maximal magnitude attained by a frequency response function under bounded perturbation of a set of parameters. These methods enable accurate and ecient determination of the worst-case (maximal) response amplitudes that a bladed disk may experience in the presence of unpredictable but bounded perturbations. The accuracy and eciency of the proposed algorithms is demonstrated using a bladed disk model subject to 56 simultaneous perturbations in the blade-alone natural frequencies. Accuracy and eciency are achieved even in the presence of a large numbers of perturbations. Therefore, the proposed algorithms are well suited for conducting realistic parametric studies required to design robust bladed disks that consistently perform their function despite manufacturing variations, wear, and other real-world imperfections.

Reduced order modeling for clearance control in turbomachinery

Finite element models (FEMs) are extensively used in the design optimization of utility scale steam turbines. As an example, by simulating multiple startup scenarios of steam power plants, engineers can obtain turbine designs that minimize material utilization and at the same time avoid the damaging effects of large thermal stresses or rubs between rotating and stationary parts. Unfortunately, FEMs are computationally expensive and only a limited amount of simulations can be afforded to get the final design. For this reason, numerous model reduction techniques have been developed to reduce the size of the original model without a significant loss of accuracy. When the models are nonlinear, as is the case for steam turbine FEMs, model reduction techniques are relatively scarce and their effectiveness becomes application dependent. Although there is an abundant literature on model reduction for nonlinear systems, many of these techniques become impractical when applied to a realistic industrial problem. This paper focuses in a class of nonlinear FEM characteristic of thermo-elastic problems with large temperature excursions. A brief overview of popular model reduction techniques is presented along with a detailed description of the computational challenges faced when applying them to a realistic problem. The main contribution of this work is a set of modifications to existing methods to increase their computational efficiency. The methodology is demonstrated on a steam turbine model, achieving a model size reduction by four orders of magnitude with only 5% loss of accuracy with respect to the full order FEMs. These practical implementations enable the calculation of multiple additional design scenarios.

Adding code generation to develop a simulation platform

Efficient and accurate control technologies require extensive simulation capabilities to validate the control software and demonstrate the impact on the business and equipment. To create a platform for rapid development and simulation of complex dynamic models, the authors and their colleagues have designed an object-oriented architecture. A portion of the architecture framework is constructed using code generation based on XML component definitions. This paper describes the key aspects of the architecture of the control simulator platform and its code generation capabilities. The simulation platform consists of defining a collection of components represented by differential equations, the capability to select, configure and interconnect components, and the ability to solve the coupled set of equations. The code generation is custom-built, however, the generated code results in more consistency and improved reliability by eliminating error prone steps and allowing the simulation engineer to focus on component mathematical description. As the number of components that are generated increases, the investment in a custom-built code generation is quickly realized by significantly reducing the amount of time required to create or update the code template for each component. The paper also discusses the importance of handling updates as well as the initial creation of code files. The techniques leveraged within this project have been learned through the use of other code generation tools, including GUI development tools. In particular, care has been taken to minimize the accidental loss of manually introduced code and handle version updates of the code generator. Our team is using this framework to develop simulation tests for power plant optimizations and have plans to add custom code generation to other areas of the platform.