Christophe Barnes

Monitoring underground reservoirs is essential to ensure their efficient and sustainable operation. This study investigates the detectability and imaging, using seismic methods, of modifications in elastic properties caused by cyclic injection and production of geothermal fluid in high-temperature aquifer thermal energy storage (HT-ATES) systems. In such systems, the induced variations in elastic properties are expected to be very small, requiring high-sensitivity methods for reliable detection. The work is conceived as a feasibility study in a setting where field data are not yet available. Synthetic modelling is used to evaluate the effectiveness and sensitivity of Full Waveform Inversion (FWI) to reservoir variations. FWI is selected for its ability to exploit the full recorded wavefield, which is critical for resolving subtle elastic changes in thin, low-contrast reservoirs. The DeepStor HT-ATES research project in Karlsruhe, Germany, where layers thinner than 10 m at 1.3 km depth are targeted, serves as a case study. Two elastic models were developed: one unperturbed and one incorporating anomalies expected after five years of cyclic operations. Due to the thickness and depth of the targets, surface seismic methods were excluded in favour of cross-well configurations using Distributed Acoustic Sensing (DAS). Active sources are adopted to ensure controlled and repeatable illumination required for time-lapse monitoring. Results indicate that geometries with receivers intersecting the reservoirs significantly improve imaging and parameter estimation. This study evaluates the applicability of established FWI approaches in thin, stratified geothermal systems and highlights the importance of survey design for successful HT-ATES monitoring.

Despite the efforts reported in the literature to explain contrasting experimental observations, the evolution of seismic attributes (velocity and attenuation) of rocks across the saturation range remains ill understood. In a comparative study, we monitored the evolution of ultrasonic P wave attributes in a porous sandstone subjected to two experiments of moisture adsorption and water spontaneous imbibition. Both experiments highlighted a significant (i.e., by 1 order of magnitude) and similar drop in P waves amplitudes, although the maximum saturations reached in each experiment is very different (i.e., about 2% vs. 70%). However, only moisture adsorption leads to a dramatic elastic softening (velocity reduction). This difference might be explained by the coupling between two competing physical mechanisms taking place during water imbibition, namely, elastic softening driven by water adsorption at the grain contact and elastic stiffening driven by full saturation of the grain contacts, at the ultrasonic frequency of the measurement.

Full-waveform inversion for borehole seismic data is an ill-posed problem and constraining the problem is crucial. Constraints can be imposed on the data and model space through covariance matrices. Usually, they are set to a diagonal matrix. For the data space, signal polarization information can be used to evaluate the data uncertainties. The inversion forces the synthetic data to fit the polarization of observed data. A synthetic inversion for a 2D-2C data estimating a 1D elastic model shows a clear improvement, especially at the level of the receivers. For the model space, horizontal and vertical spatial correlations using a Laplace distribution can be used to fill the model space covariance matrix. This approach reduces the degree of freedom of the inverse problem, which can be quantitatively evaluated. Strong horizontal spatial correlation distances favor a tabular geological model whenever it does not contradict the data. The relaxation of the spatial correlation distances from large to small during the iterative inversion process allows the recovery of geological objects of the same size, which regularizes the inverse problem. Synthetic constrained and unconstrained inversions for 2D-2C crosswell data show the clear improvement of the inversion results when constraints are used.

We tested on a data set including 12 different rocks the assumption that, during water imbibition experiments, the interaction between direct P waves and reflected P waves on the imbibition front could explain the early P wave amplitude variation occurring before any velocity variation. Our calculations show that (i) the observed distance between the water front and the ultrasonic sensors is always larger than the estimated critical distance below which the first peak amplitude is impacted by the reflected P wave, (ii) the magnitude of the coefficients of reflection is generally very small, and (iii) using simulations of composite waveforms, an impact on the first peak amplitude is expected only in 25% of the tested samples. P wave interaction is definitively a mechanism which can lead to amplitude variations, but it is very likely that such a mechanism is not responsible for the early amplitude variation detected in most of the tested rocks. Moisture diffusion is still a very relevant mechanism to account for our observations on P wave amplitude variations during capillary rise.

Capillary rise experiments (spontaneous imbibition tests) were conducted in the laboratory with ultrasonic and X‐ray monitoring on the Sherwood sandstone and the Majella grainstone. The aim was to provide a direct comparison between the variation in seismic attributes (amplitude, velocity, spectral content, and energy) and the actual fluid distribution in the rock. Two pairs of ultrasonic P wave sensors located at different heights on a cylindrical rock specimen recorded every 5 s the waveforms when capillary forces make water rise up into the rock from the bottom in contact with a water tank. Simultaneously, computerized tomography scan images of a vertical cross section were also recorded. Two important results were found. (i) The amplitude of the first P wave arrival is impacted by the upward moving fluid front before the P wave velocity is, while the fluid front has not yet reached the sensors level. In contrast, the P wave velocity decreases when the fluid front reaches the Fresnel clearance zone. The spectral analysis of the waveforms shows that the peak frequency amplitude is continuously decreasing without noticeable frequency shift. (ii) A methodology based on the calculation of the analytical signal and instantaneous phase was designed to decompose each waveform into discrete wavelets associated with direct or reflected waves. The energy carried by the wavelets is very sensitive to the fluid substitution process: the coda wavelets related to reflections on the bottom end face of the specimen are impacted as soon as imbibition starts and can be used as a precursor for the arriving fluid.

Fluid substitution processes have been investigated in the laboratory on 14 carbonate and siliciclastic reservoir rock analogues through spontaneous imbibition experiments on vertical cylindrical specimens with simultaneous ultrasonic monitoring and imaging. The motivation of our study was to identify the seismic attributes of fluid substitution in reservoir rocks and to link them to physical processes. It is shown that (i) the P wave velocity either decreases or increases when the capillary front reaches the Fresnel clearance zone, (ii) the P wave amplitude is systematically impacted earlier than the velocity is, (iii) this precursory amplitude decrease occurs when the imbibition front is located outside of the Fresnel zone, and (iv) the relative variation of the P wave amplitude is always much larger than that of the P wave velocity. These results suggest that moisture diffuses into the pore space ahead of the water front. This postulate is further supported by a quantitative analysis of the time evolution of the observed P wave amplitudes. In a sense, P wave amplitude acts as a precursor of the arrival of the capillary front. This phenomenon is used to estimate the effective diffusivity of moisture in the tested rocks. The effective moisture diffusivity estimated from the ultrasonic data is strongly correlated with permeability: a power law with exponent 0.96 predicts permeability from ultrasonic monitoring within a factor 3 without noticeable bias. When the effective diffusivity is high, moisture diffusion affects ultrasonic P wave attributes even before the imbibition starts and impacts the P wave reflectivity as evidenced by the variations recorded in the waveform coda.