Background

EBSILON^®Professional is a calculation program for energy and mass flow balancing of power plant processes.

EBSILON^®Professional permits the balancing of

• individual components
• component groups
• sub-systems
• complete systems

without taking into account, whether these components or systems build a closed or an open model.

EBSILON^®Professional has an uninterrupted model structure with

• a standard component, which is used for modeling the usual power plant model
• an additional component for modeling complex power plant processes.

In this way, EBSILON^®Professional is also capable of balancing,

• water/steam processes
• gas turbine processes, and
• combination processes

with heat extraction.

EBSILON^®Professionaluses the

• IAPWS-IF97 of the IFC-67 steam table, and
• cp-polynomial for air/flue gas

as the data basis.

EBSILON^®Professional is a variable program system, with which you can balance all the power plant models with the help of a closed solution system (based on a sequential solution method). Controllers are necessary for a versatile use of this system. They are used as iteration controllers of how the components and their characteristics must be specified by the user. The iteration run is optimized by the "self-learning" properties of the controller.

Based on this philosophy, EBSILON^®Professional proves to be a very flexible system. Thanks to the experiences, all the possible power plant systems can be balanced with the help of this system.

The cycle is modelled as a non-linear equations system, solved by a series of linear equation system. The variable coefficients and the right-hand sides are formed by using the values of the preceding iteration step. The iteration ends, when the basis variable no longer change.

To accelerate the convergence, the variable right-hand sides and the coefficients are developed through the Taylor series during the entire preceding iteration step. In this way, the non-linear equations system is solved with the Newton method.

To solve a series of linear equation systems, an iteration method is used, since the matrix is filled only slightly. A Gauss-Seidel method is selected after extensive comparisons of computing time and convergence.

The user must do this essential work, before he can start with the calculation: The user must model the cycle as exactly as possible by using the available components. The rules for modeling are as follows:

Rule 1: Check, whether the selected component really has the physical properties, in order to be simulated.

Rule 2: Connect the components reasonably, under absolute consideration of the meaning of each connection.

Rule 3: In case none of the available components possess the properties of the process components, it must be checked, whether another corresponding menu component can be used in place of this component.