Large Scale Experiments on the Tightness of Cemented Boreholes under Cyclic Loading

The exit from nuclear and fossil-fuel energy and the increase in renewable energy conversion lead to a higher fluctuation in energy supply. To meet the demand in times of energy shortages from sun and wind, this effect can be compensated by extracting and using gas from underground gas storages. As long as enough renewable energy is available, storages can be filled again. However, this results in increasing injection and extraction frequencies, leading to faster occurring pressure and stress changes and therefore posing an additional challenge for reservoir rock, cap rock and technical components.

To evaluate the effects of this additional cyclic loading on the rock-cement-steel-compound, we used an autoclave system on a realistic scale. It simulates abandoned drillings and consists of a 2 m long cemented steel casing with an autoclave system. To simulate injection and extraction, gas pressure (N₂) is applied and released on both ends. Additionally, temperature can be raised up to 100 °C. Between loading cycles, permeability can be measured to determine the effect of pressure and temperature variation on the tightness of the cemented well system.

We present results from the analysis of four cemented casings. Since the hardened cement isn’t connected to the steel casing after experiments, we assume an annular gap as main gas path in most cases. This gap is modelled and fitted to the experimental data. After pressure variations between 0 bar and 60 bar, tightness of the system decreased in every experiment, which leads to an increased modelled annular gap width. Temperature variations between 30 °C and 70 °C didn’t have an effect on the first two casings, but increased tightness and therefore decreased the modelled gap width in the third casing.

Additionally, we observed an anomaly in the second casing, which was extraordinarily tight before the first pressure drop. However, when small amounts of pressure (around 3 bar) were released from an autoclave chamber, around 30 % of the released pressure built up again within a few minutes, while the rest took several hours. We assume a releasing induced gas cooling in the autoclave, while the surrounding warmer steel heats the gas up to the original temperature. This can only be observed for experiments with a rather tight cement-steel plug. In other cases this is not observable because it is superimposed by the pressure build up through the annular gap. The results of finite difference model taken this temperature induced pressure build up into account and are compared to the results of independent permeability and porosity measurements.