Search

El presente trabajo de investigación inició con la fundamentación teórica concerniente a la inyección continua de vapor y a las variables que afectan el proceso como tal, teniendo en cuenta lo realizado en la fase 1 de esta investigación, por las investigadoras Lina Nova y Gina Becerra, quienes determinaron los KPIs influyentes en el proceso. Para poder determinar el abordaje más adecuado y así utilizar los datos de los 56 campos estudiados, consignados en la base de datos elaborada por los autores, se estudiaron siete metodologías propuestas por diferentes autores en el pasado, escogiendo finalmente las dos más próximas a la resolución de los objetivos planteados. Estas fueron las propuestas por Bursell y Pittman en 1975, y por Chieh Chu en 1985, la primera haciendo uso de una aproximación gráfica, y la segunda utilizando la regresión lineal como método de estimación de la relación vapor/aceite. Basándose en lo anterior, se tomaron, modificaron y mejoraron las metodologías previas, para así lograr estimar la eficiencia de un campo potencialmente sometible a inyección continua de vapor, donde el campo Palagua fue el escogido para este trabajo.

LONGLIFE (“Treatment of long term irradiation embrittlement effects in RPV safety assessment”) was a collaborative project of the 7th Framework Programme of EURATOM under the umbrella of NULIFE/NUGENIA, aiming at an improved understanding of irradiation effects in reactor pressure vessel steels under conditions representative of long term operation. The LONGLIFE project was completed by end of January 2014. The paper gives an overview of the main project results and their implications for future research, as discussed at the final project workshop. The microstructural database for neutron-irradiated RPV steels was extended considerably and existing gaps on mechanical property data were closed. Indications of late blooming effects (LBE) were found in some cases, but clear criteria for the occurrence/exclusion in terms of irradiation conditions and chemical composition have still to be developed. The commonly accepted trend, that low flux and low irradiation temperature promotes LBE, is supported. A significant flux effect on the size of defect clusters was observed in two high Cu weld materials, while the changes of mechanical properties are not affected by the neutron flux. The database requires completion in particular for low-Cu RPV steels. The shift of reference temperature T0 over the thickness location of a VVER-440 welding seam does not follow the prediction Russian code, because of the strong variation of the intrinsic weld bead structure. Therefore, the effect of the initial microstructure and of the heterogeneity on the radiation behaviour has to be addressed in future works. Existing embrittlement trend curves models were applied to the LONGLIFE data base. None of the trend curves could predict the behaviour of the entirety of the LONGLIFE materials sufficiently. A guideline for monitoring radiation embrittlement during life extension periods was developed.

For efficient firing in sloping grate bark ovens, it is important to control the ash line position on the grate. Here, the problem of detecting the ash line with an image processing system is treated. The presentation of trend curves to the operator is of special interest. The trend curves shall indicate how the ash line position changes with time, and in that way improve the manual control of the process. Future development can result in a closed loop feedback system. The goal has been to obtain algorithms suitable for a small computer e.g. IBM-PC. Three algorithms for detection of the ash line are presented. One is an optimization method where the position of the ash line is computed by maximizing a criterion. The criterion is designed especially for the application, and maximized by dynamic programming. The other two algorithms are based on thresholding techniques. One is based on preprocessing before a global thresholding, and the other is based on thresholding in subimages. A threshold selection algorithm has been developed and used in the implementation of the algorithms. The algorithms have been tested on images taken from a video tape, showing the interior of an industrial bark oven. Experimental results show that the algorithms can be used for the generation of trend curves.