Summary

One of the most common silicate minerals on earth is olivine, which has been a natural regulator for atmospheric carbon dioxide concentrations for millions of years. In the presence of water, CO2 is chemically converted to calcareous products. In past years, the possibility of olivine applications to acquire negative emissions of CO2 has received the attention of scientists and policy. The CO-Action project targeted the obstructions for the use of olivine to capture CO2 such as scientific knowledge gaps that inhibited the estimation of mineral weathering rates, and addresses the bottlenecks for large scale applications in natural environments. Main obstructions are: 1) uncertainty of weathering rates of olivine under natural conditions; 2) toxicity risks of nickel that is incorporated in the mineral and is released upon weathering.

A two-year field experiment was conducted in the Netherlands to study mineral weathering rates under various conditions. Variables were: mineral source (Norwegian and Spanish olivine, Canadian wollastonite), type of application (on-top versus mixed-in), availability of moisture (rain-fed versus wet), and effect of vegetation (planted versus non-planted). All field plots contained mesocosms with similar treatments for mass-balancing. Pore water was sampled periodically over various depths, and different chemical extractions were applied to the soil. The numerical model OWCS, based on the shrinking core principle, was calibrated with data obtained from the field experiments. A separate risk assessment module for nickel was developed and incorporated in the model.

Results show that 6 to 8.5 percent of the initial mineral dose was dissolved after two years. Applying the Shrinking Core Model to each grain size, the dissolution over time could be calculated accurately. Of the smallest grain fraction (<2μm), 100 % of the mass is dissolved in two years. In the same period, 42.9 % of the fraction < 8 μm is dissolved. The larger fractions show decreasing dissolved portions respectively. Modelled and measured values show a good agreement.

Only marginal differences for extractable Mg and Ni were observed between dry and wet plots. Moisture content varied around 20 – 40 % in dry plots, indicating that the moisture content under natural precipitation conditions is not significantly limiting the chemical reaction compared to saturated conditions. The effect of vegetation (alternating winter barley and summer rye), and type of olivine application was of minor consequence, but the grain size distribution of the grounded minerals plays a decisive role in weathering rates. Due to weathering, and the release of alkaline products, the pH of the soil increased 0.2 to 0.5 unit over the course of the experiments.

Nickel concentrations in pore waters are elevated compared to the reference plot. However, 96 % of all Ni measurements are below the analytical upper reporting limit. Only two measurements (Norwegian olivine source) exceed the generic quality standards but only short after application. In the first period of weathering, most nickel is released from the ultrafine fraction. In all cases, the calculated No-effect concentration (PNEC) for nickel is larger than its measured value in pore water and are (far) below the risk characterization ratio (RCR < 1). This is a strong indication that no chronic toxic risks of nickel release are to be expected for the concentrations observed during the experiment.