The system will be unavailable for updates from 12:30 on Tuesday 23 May to prepare for the upgrade of the software platform.

Revised regional maximum flood method and regionalisation

Masule, Simenda (2021-12)

Thesis (MEng)--Stellenbosch University, 2021.


ENGLISH ABSTRACT: Estimating flood peaks is necessary for the design and safety evaluation of hydraulic structures. The flood peaks are estimated using methods developed during the late 1960s to the late 1980s with only limited data and methodology available at that time. The Regional Maximum Flood (RMF) is one such method, which is empirically established and uses regional envelope curves to determine the maximum flood that can be expected in a region. The method further uses regional ratios derived from the recurrence interval flood peaks as a fraction of the RMF for estimating the recurrence interval of the latter. The RMF method was last updated in 1988, availing more than 30 years of additional data for further analysis. This research revisits the RMF method as applicable to South Africa, where an evaluation and update of the method are presented. The results from the evaluation indicated that the regional envelope curve established in 1988 were exceeded at two sites (Lorie dam and Pongolapoort dam). As a result, currently these exceeded envelope curves do not adequately represent the RMF. The excess above the RMF at each site was calculated as 25% and 8% respectively. This resulted in the need to update the method and its regionalisation. The regionalisation resulted in the determination of 15 RMF K regions and their associated RMF K regional envelope curves. The determined RMF K regions, in contrast to the 1988 RMF regions, were observed to have regions of smaller boundaries with a highest RMF K regional envelope value of 5.8 and the lowest value 2.8. The RMF K regional envelope curves were drawn 15% above the maximum observed flood peaks for each region. It was concluded that the newly determined RMF regional envelope curve adequately represents the RMF and is therefore applicable for determining the expected maximum flood at site of interest. Regional ratios for estimating the flood peaks with different recurrence as presented in the 1988 RMF method were evaluated and it was found that the ratios significantly overestimated the flood peaks. New regional ratios corresponding to a recurrence interval of two to 200 years were then derived and evaluated. The evaluation revealed that the ratios were associated with an extremely low correlation coefficient. In most cases, the new regional ratios underestimated flood peaks. The percentage mean error of the new ratios for six out of 11 RMF K regions evaluated for 50 to 200-year recurrence interval were between -15% to +15%, which was deemed as acceptable. However, because of the low correlation coefficient, it was concluded that these new regional ratios should not be used to estimate flood peaks with different recurrence intervals. Further investigation into these ratios is recommended. Topics from this research on which further research would be beneficial for the improvement of the presented RMF method were highlighted.

AFRIKAANSE OPSOMMING: Die beraming van vloedpieke is noodsaaklik vir die ontwerp en veiligheidsevaluasie van hidroliese strukture. Die vloedpieke word beraam deur metodes wat tussen die laat 1960s en die laat 1980s ontwikkel is met slegs beperkte data en metodologie beskikbaar. Die Streeks Maksimum Vloed (SMV) (Eng: RMF) is een so ‘n metode, wat empiries ontwikkel is en dit gebruik streekgrenskurwes om die maksimum vloed wat verwag kan word in ‘n streek te bepaal. Hierdie metode gebruik verder streekverhoudings wat afgelei is van die herhaalperiode van vloedpieke as ‘n breuk van die SMV, om laasgenoemde te beraam. Die SMV metode is laas opgedateer in 1988, met meer as 30 jaar se addisionele data vir verdere analise beskikbaar. Hierdie navorsing het die SMV metode weer ondersoek soos van toepassing op Suid-Afrika en ‘n evaluasie en opdatering van die metode is aangebied. Die resultate van die navorsing het aangetoon dat die streekgrenskurwes wat in 1988 vasgestel was, by twee terreine (Loriedam en Pongolapoortdam) reeds oorskry was. Die resultaat is dat hierdie huidige oorskryde streekgrenskurwes nie die SMV voldoende verteenwoordig nie. Die oorskryding van die SMV by elke terrein is as 25% en 8% onderskeidelik bereken. Dit het gelei tot die noodsaaklikheid vir die opdatering van die metode en sy afbakening. Die afbakening het verder gelei tot die vasstel van 15 SMV-K streke en hulle geassosieerde SMV-K streekgrenskurwes. Daar is waargeneem dat die SMV-K streke, in kontras tot die 1988 SMV streke, kleiner gebiede beslaan, met die hoogste SMV-K streekgrenswaarde 5.8 en die laagste waarde 2.8. Die SMV-K streekgrenskurwes is as 15% bo die maksimum waargeneemde vloedpieke vir elke streek getrek. Die gevolgtrekking was dat die nuut bepaalde SMV streekgrenskurwes die SMV voldoende verteenwoordig en dus toepaslik is vir die bepaling van die verwagte maksimumvloed by die betrokke terreine. Streekverhoudinge om vloede te beraam met verskillende herhaalperiodes soos deur die 1988 SMV metode bepaal, is beoordeel en daar is gevind dat die verhoudingedie vloedpieke merkbaar oorskat het. Nuwe streeksverhoudinge vir vloede met n herhaalperiodes van 2- tot 200 jaar is verder afgelei en beoordeel. Die evaluasie het getoon dat die nuwe streekverhoudinge die vloedpieke onderskat. Die gevolgtrekking was dat hierdie nuwe streekverhoudinge nie gebruik moet word om vloedpieke met verskillende herhaalperiodes te bepaal nie. Daar word dus aanbeveel dat verdere navorsing uitgevoer moet word rakende die verhoudings.

Please refer to this item in SUNScholar by using the following persistent URL:
This item appears in the following collections: