Simulation of yearly maximal flood discharges of reservoirs of Dnieper cascade

Probabilistic estimation of reliability of hydrotechnical structures of cascades of hydro power plants requires investigation of probabilistic nature of loadings and influences that hydrotechnical structures experience during operation. Cascades of hydro power plants are situated at water courses th...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Mozgovyi, Andrii, Spirande, Karina, Butnik, Svitlana
Format: Tagungsbericht
Sprache:eng
Schlagworte:
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:Probabilistic estimation of reliability of hydrotechnical structures of cascades of hydro power plants requires investigation of probabilistic nature of loadings and influences that hydrotechnical structures experience during operation. Cascades of hydro power plants are situated at water courses that have considerable geographic extent. Therefore, there are correlation relationships between natural climatic factors of such complex objects as hydro systems, that occur during passing from one hydroscheme of the cascade to another. Under condition of linear correlation, pair and partial correlation coefficients for pairs of series of statistical data of annual maximal discharges of Dnieper, that were recorded in Vyshgorod, Kyiv, Kremenchuk, Lotsmano-Kamyanka, were calculated. The developed algorithm was tested on hydrological regime of hydroschemes of the Dnipro cascade of hydro power plants. Numerical results obtained with help of the proposed procedure were compared with known results. The largest deviations of points of the plot of linear function of regression of annual maximal discharges of r. Dnipro from the registered annual maximal discharges are recorded at p = 99.95 %, σ = 933 m3/s; p = 0.5 %, σ = 3687 m3/s, and standard deviation is σ = 780 m3/s. The obtained data correlate with data of other authors in whose investigations the difference between factual and predicted data reaches 10 % – 12 %. From the viewpoint of approximation of predicted data to factual ones the results can be improved by the use of the more complex nonlinear polynomial regression models of high orders. The use of approaches of multi-factor correlation and regression turned out to be acceptable during simulation of hydrological regime of Dnipro under cascade arrangement of hydroschemes. Confidence intervals of regression equations and accuracy of the obtained results were assessed.
ISSN:0094-243X
1551-7616
DOI:10.1063/5.0120185