Development of an effective operational energy envelope for process monitoring

Abstract

An Effective Operational Energy Envelope (EOEE) is proposed and developed for the purpose of process monitoring, wherein desirable operating conditions form a part of the EOEE. This concept builds on previous work conducted in energy visualisation of systems. Thereafter, the EOEE was evaluated on a FLOWNEX model of the Pebble Bed Micro Model (PBMM). This system was chosen because it is an energy conversion system, which ideally lends itself to energy analysis studies.

The system was simulated in FLOWNEX with a constant Power Turbine (PT) set point of 100 kW. The degradation of compressors and turbines was simulated through the use of the respective characteristic charts. The PBMM system in FLOWNEX compensated for this by increasing the pressure, thereby increasing the mass flow within the system. The results from these simulations were captured using MATLAB and characterised by energy attributed graphs. Following this, the Heterogeneous Euclidean-Overlap Metric (HEOM) function was used. This function quantifies the difference between two matrices of two different operating conditions. The normal (non-degraded simulations/operating conditions) and degraded operating conditions were illustrated through Singular Value Decomposition (SVD).

The Machine Efficiency (ME), or energy conversion efficiency of the system, was used as a performance measure to determine which results would be included in the EOEE and which would not. All operating conditions with an ME below 14% were deemed to fall outside of the EOEE. A second method of creating the EOEE was also proposed. This included a warning front that would act as a buffer between the EOEE or desired operating conditions, as well as the results that were deemed too undesirable. This warning front operated between the machine efficiencies of 13% to 15%. These values formed the basis for the EOEE conceptualisation of the PBMM.

SVD was used to create 3D graphs to visualise the EOEE. The first three SVD values represented the axis and multiple 2D versions. In the 2D versions, the axes were represented by a combination of the first three SVD values. In another version, a single SVD value was plotted against its respective operating point number. This resulted in overlap between points in the EOEE and points not in the EOEE. Consequently, this version is less usable than other methods. These graphs clearly illustrate the difference between the operating points that are included in the EOEE and the points not included in the EOEE or the warning front. Following this, a second set point of 150 kW was simulated to validate the EOEE as being operating point independent. The same patterns and trend lines were observed for both the150 kW and the 100 kW points. This could indicate that the EOEE can be used to determine if the system is operating as intended or if an error or degradation has occurred to the level where maintenance is required for multiple operating points.