In several industries using pipelines to transport different products from one point to another is a common and indispensable process, especially at oil/hydrocarbon industries. Thus, optimizing the way this process is carried out must be an issue that cannot be stopped. Therefore, the performance of the control strategy implemented is one way of reaching such optimal operating zones. This study proposes using Model Predictive Control strategies for solving some issues related to the proper operation of pipelines. It is proposed a model based on physics and thermodynamic laws, using MATLAB® as the development environment. This model involves four pumping stations separated by three pipeline sections. Three MPC strategies are developed and implemented. Accordingly, the results indicate that a centralized controller with an antiwindup back-calculation method has the best results among the three configurations used.

This paper presents a novel non-invasive monitoring method, based on a Liénard-type model (LTM) to diagnose single and sequential leaks in liquid pipelines. The LTM describes the fluid behavior in a pipeline and is given only in terms of the flow rate. Our method was conceived to be applied in pipelines mono-instrumented with flowmeters or in conjunction with pressure sensors that are temporarily unavailable. The approach conception starts with the discretization of the LTM spatial domain into a prescribed number of sections. Such discretization is performed to obtain a lumped model capable of providing a solution (an internal flow rate) for every section. From this lumped model, a set of algebraic equations (known as residuals) are deduced as the difference between the internal discrete flows and the nominal flow (the mean of the flow rate calculated before the leak). Once the residuals are calculated a principal component analysis (PCA) is carried out to detect a leak occurrence. In the presence of a leak, the residual closest to zero will indicate the section where a leak is occurring. Some simulation-based tests in PipelineStudio® and experimental tests in a lab-pipeline illustrating the suitability of our method are shown at the end of this article.

En este artículo se desarrolla una estrategia de control para un proceso de producción, conformado por un reactor con calentamiento, adquirido a través de un serpentín con vapor. Además de la variable controlada que es la concentración del producto, CB, se controlan adicionalmente el nivel y la temperatura. Se sintonizaron los controladores PID, empleando el procedimiento del modelo de primer orden más tiempo muerto (FOPDT). Finalmente, se realizaron varias simulaciones bajo condiciones de operaciones diferentes. La primera prueba realizada consistió en operar el sistema en condiciones nominales, haciendo CB(set)= 0.4. Se observa un sobre-pico de 0.55% para la concentración de B, que se considera bastante bajo, en cuanto al tiempo de establecimiento se observó una reducción considerable, se pasó de 30 min en un lazo abierto a 15 min en un lazo cerrado; adicionalmente, se logra un error en estado estacionario nulo con lo que se verifica la efectividad del lazo de control.