skip to main content
OSTI.GOV title logo U.S. Department of Energy
Office of Scientific and Technical Information

Title: Process feed-rate estimation and event detection using least-squares techniques

Technical Report ·
OSTI ID:6219755

During operation of the ICPP first-cycle process, the process feed rate equals the tank depletion rate for feed tank G-106 and can be determined from G-106 volumetric data collected by the existing Process Monitoring Computer System (PMCS). The process feed rate evaluated here was obtained using plant precision Scanivalve data collected during August 1981. The calculations are basically appropriate for on-line use but have been demonstrated only in an off-line manner on the PMCS vax 11/780 computer. The selected technique assumes a second-order polynomial model and performs least-squares smoothing over a sliding N-point data window. The leading edge of the data window is at the new (real-time) data point, and the results obtained are filter estimates evaluated at that moving time point. The data points can be unevenly spaced. The estimated variables in general are liquid volume, volume rate of change (feed rate), and volume acceleration. Included are their precision error statistics, and fundamental estimation theory is used to project the standard error values from the window center to the new data point. The process feed rate information is useful for process estimation and control applications; the volume acceleration information is useful for determining batch-input transfer times via suitable threshold detection criteria. The latter provides event time determination using analog data alone and is useful in lieu of (or associated with) event determination by logic analysis methods that require use of plant digital signals.The data used contain considerable noise which limits the feed rate time response, and suggestions are made for improved performance. An adaptive algorithm is considered but does not perform significantly better than a large fixed window size. The event-detection algorithm performs quite well for the data used.

Research Organization:
Exxon Nuclear Idaho Co., Inc., Idaho Falls (USA)
DOE Contract Number:
AC07-79ID01675
OSTI ID:
6219755
Report Number(s):
ENICO-1129; ON: DE83009742
Resource Relation:
Other Information: Portions are illegible in microfiche products. Original copy available until stock is exhausted
Country of Publication:
United States
Language:
English