A petrological and geochemical study regarding the suitability of the Vaalputs Radioactive Waste Disposal Facility

Abstract

The National Radioactive Waste Disposal Facility at Vaalputs in Bushmanland was selected for the disposal of intermediate- to low-level radioactive waste. The selection followed intensive investigations involving a wide spectrum of the earth sciences. Amongst others, mineralogical, petrological and geochemical studies were applied to study the rocks as a natural analogue to waste disposal. The aim of the investigation was mainly to identify the geochemical processes in the geological environment which gave rise to the present element distribution and furthermore to assess the extent to which the processes could retard element mobility. Using this knowledge, supported by experimental data, it was attempted to predict how these processes could contribute to retard the migration of radionuclides, comprising mainly Cs, Sr and Co, from the buried waste to the biosphere in the event of accidental release. The suitability of the rocks for the disposal of high-level radioactive waste was only briefly investigated. The geological investigations led to the identification of an argillaceous surficial succession overlying a crystalline basement, comprising mainly granitic gneiss of the Namaqualand Metamorphic Complex, into which noritic rocks of the Koperberg Suite, amongst others, are intrusive. The surficial succession consists of a rock unit of argillaceous character, termed the white clay, overlying the weathered basement and is suggested to have been derived by in situ weathering of the basement. The white clay is overlain by a red clay unit, informally named the Vaalputs Formation, in which the Si02/A1203 ratio is significantly higher than that of the white clay indicating its more arenaceous character. Calcrete development is evident at various horizons through the entire surficial succession, but is especially prominent near the present ground surface. The red clay is significantly more depleted with respect to most elements, except Ca and Mg in the calcrete, when compared to the white clay. The entire succession is covered by a thin veneer of red sand belonging to the Gordonia Formation. The main clay minerals of the red clay are illite with subordinate kaolinite and smectite, while smectite predominates in the white clay and in the weathered and fresh basement, with subordinate kaolinite and illite. Smectite and kaolinite generally increase in depth while illite decreases, but exceptions to these trends occur. The former two formed by hydrolysis of the feldspar minerals and their stability is confirmed by the ground water chemistry. Illite presumably formed by physical and chemical reconstitution of micaceous minerals of which especially bioti te is abundant in the Vaalputs rocks. These clay minerals, especially smectite and illite, have good adsorption and ion-exchange properties and can account for high concentrations of many elements, including the natural isotopes of the radionuclides. High concentrations of notably Nb, V, Cr, Th, Ni, Co, Zn and Pb can probably be attributed to adsorption onto, or co-precipitation with Ti-, Fe- and Mn-oxide minerals, which are known to be "scavengers" of these elements. Some trace elements such as Sr, Ba, Pb and Y are concentrated in a similar manner as above by calcretes, while the alkaline environment in the vicinity of calcrete layers is considered to be responsible for the immobilisation of Cs, Rb, Ba and V. The sediments are generally impoverished with respect to uranium, except for a few instances where locally favourable physico-chemical conditions were possibly conducive to its adsorption onto clay minerals. Iron-oxide minerals have also been shown to adsorb uranium, but hardly any correlations of Fe with U were found. Zirconium and hafnium are thought to be present mainly in zircon. Their correlation with Ti-, Fe- and Mn-oxides towards the base of the red clay suggests their presence in heavy mineral concentrations.