(371f) Conceptual Framework for Integrated Control and Design of Coupled Processes

Authors: 
Himmel, A., Otto-von-Guericke-University Magdeburg
Sundmacher, K., Max Planck Institute for Dynamics of Complex Technical Systems
In view of increasing variabilities in the price and quality of raw materials, process engineering systems have to be able to operate dynamically. Especially when using renewable energies, medium- and short-term periods are crucial for the operational management.

At the same time, it is often essential to compensate for these variabilities along the process path. For this purpose, material and energy buffers as process elements play an important role, since temporary over- or underproduction is not transferred to subsequent process elements. A further possibility to compensate variabilities on the raw material side is offered by the special topology of integrated processes that we consider. Thereby, only processes are considered generating a similar product from different raw materials, which are then mixed to form the main product. Furthermore, these processes exchange secondary produced material and energy flows via buffer systems to complement each other. For example, the production of methane via a biogas plant and catalytic methanization using hydrogen from an electrolyzer and carbon dioxide from the biogas plant is such an integrated process.

The special process topology allows us to formulate a general operating strategy for the processes on the one side and a design concept for the intermediate buffer system on the other side. This conceptual framework is applicable to arbitrary processes.

The operating strategy we use is divided into two parts and follows a classic paradigm [1]. First, the production levels are determined such that both processes operate in an economically attractive area. Second, a control signal is determined that takes the processes to the new production level as fast as possible. The two problems — determination and transition of the production levels — are linked, since we use information about the dynamic behavior, such as the transition time of the individual processes, to evaluate the magnitude of the load change. In general, this leads to a bilevel optimization problem that is hard to solve online. In order to reduce the numerical complexity, we separate both problems by an offline generated surrogate model of the transition times.

Another important aspect in our approach is the consideration of buffer sizes when determining new production levels. This is related to the design of the buffers, which are dimensioned in accordance to feasible scenarios for variabilities in raw material prices and qualities. In order to reduce investment costs, the buffer capacities are favored to be minimal. However, it is crucial to consider the dynamic properties of the buffer system in such a way that no further dynamics are imposed on the overall process.

The control strategy that we use to realize the transition follows the strategy proposed in [2].

In summary, the objective of this contribution is to present a unified approach for integrated control and design of two coupled processes. This ensures a fast response to price and quality variations of the raw materials, so that the operation of the entire plant within an economically optimal range is guaranteed.

[1] Lam, D.L. and Swartz, C.L.E. (2006), Optimal steady-state transitions under constrained predictive control, 16th European Symposium on Computer Aided Process Engineering and 9th International Symposium on Process Systems Engineering

[2] Himmel, A., Sager, S., Sundmacher, K. (2018), Time-optimal set point transition for nonlinear systems, Automatica, under review.