European Geologist Journal 54

Degassing kinetics of high salinity geothermal fluids


by Chris Boeije1*, Wolfgang Weinzierl2, Pacelli Zitha1 and Anne Pluymakers1

1 Delft University of Technology, Delft, The Netherlands

2 GFZ German Research Centre for Geosciences, Section Geoenergy, Potsdam, Germany



The production of geothermal fluids can be adversely affected by the formation of free gas bubbles (degassing). Decreasing pressures can cause dissolved gas to exsolve, which can reduce water production. This study aims to improve the understanding of the conditions under which free gas nucleates in geothermal reservoirs. The focus is on CO2 degassing from brines with varying salinity. We report a series of depressurisation experiments at high pressure and temperature using a cell that allows for visual monitoring of the degassing process using a high-speed camera along with pressure and temperature logging. A geochemical model was created for simulating the degassing behaviour at the same conditions as those used in the experiments, thus allowing for direct comparison.

Cite as: Boeije, Chris, Weinzierl, Wolfgang, Zitha, Pacelli, & Pluymakers, Anne. (2023). Degassing kinetics of high salinity geothermal fluids. European Geologist, 54.

1. Introduction

Degassing has several implications in subsurface flow applications such as the production of hydrocarbons or geothermal fluids. Free CO2 bubbles can cause more favourable conditions for the precipitation of calcium carbonate scale during both oil production [1, 2] and production of geothermal brines [3, 4]. Chemically active gases like CO2 and H2 can also cause corrosion in both oil and geothermal wells [5].

In exploration of geothermal sites, the presence of degassing is also used as a marker for the potential of the site to contain geothermal resources [6]. Surveys of CO2 degassing in soils have been used for this purpose in particular [7, 8]. Furthermore, regions of high degassing rates are also found to indicate the presence of high-permeability faults [9]. Finally, the presence of free gas bubbles can also lead to reduction of the water relative permeability of the reservoir as the bubbles take up space, thereby limiting the ability for the water to flow. This was assumed to have happened in the Groß Schönebeck field in Germany, which led to a reduction in the production index [10].

This study aims to improve the understanding of the formation of free CO2 bubbles from fluids during a depressurisation process, under geothermal reservoir conditions. The main research topic that is addressed is the dependency of the bubble point pressure on various parameters such as temperature and brine salinity.

2. Materials and methods

Experiments performed in this study focus on the formation of free CO2 bubbles during depressurisation, mimicking the changes in pressure encountered in a geothermal well. The experiments were done at temperatures ranging from 40 to 150 °C. This temperature range implies that these experiments are representative of low enthalpy geothermal sites [11].

2.1. Experimental setup

The experimental setup depicted in Figure 1 consists of the following components:

  • A Vindum Engineering VP1-12K-HC dual piston aqueous phase pump allows for injecting water at high pressure.
  • A Proserv Prolight 002990 titanium transfer vessel with a magnetic stirrer is used for creating gas-liquid mixtures.
  • A titanium, high-pressure visual cell with two sapphire sight glasses (30 mm diameter) on either side to allow for visualisation of the flow inside. The cell is cylindrical and has a 30 mm diameter and a depth of 10 mm. Fluid inlet and outlet are located at the side and top of the cell, respectively. A LED light source is installed to allow for uniform illumination of the cell’s inner volume.
  • A heating spiral is wrapped around the cell so that it can be heated up to the desired temperature in combination with a PID thermo-controller.
  • A Photron FASTCAM Nova S6 camera is installed in front of the cell’s window for high-speed imaging of the contents of the cell. This camera is used here at a rate of 500 frames per second. The camera is paired with a Laowa 100 mm f/2.8 2x Ultra Macro APO lens.
  • A Sensata 84HP pressure transducer is connected to the cell to monitor the pressure during the experiments at a frequency of 100 Hz.
  • Two thermocouples are connected to the cell as well: one for connecting to the thermo-controller and one to a data acquisition PC.
  • A second Vindum Engineering VP1-12K-HC pump is used downstream of the cell in fluid receiving mode to reduce the pressure in a reproducible manner.
  • Thicker tubing is installed downstream of the cell, which acts as a vessel for the hot fluids to cool down.
  • A hardware switch is connected to both the camera and the data acquisition PC and is used to simultaneously stop the pressure and image recording, thus allowing for synchronisation between the two.

Figure 1: Schematic of experimental setup.

 2.2. Experimental procedure

Brine and CO2 are premixed in the desired proportions and pressure in the transfer vessel and then homogenised using a magnetic stirrer. The visual cell is initially filled and pressurised using the same brine. The initial pressure is based on the predicted bubble point pressure from the model and is chosen to be significantly higher than this value. The brine-CO2 mixture can then be pumped from the vessel into the cell, displacing the initial brine in the cell. Five times the cell’s volume of the CO2-brine mixture is injected into the cell to ensure that the cell contains the correct CO2 concentration. Next the cell is heated to the desired temperature. The CO2-fluid mixture inside the cell is continuously monitored during both the filling and heating stage. Subsequently, the depressurisation is performed by letting the downstream pump receive fluid from the cell. The receiving rate is scheduled such that the pressure reduction rate is gradual. The full depressurisation process takes approximately 40 seconds to complete, which coincides with the time it takes to fill up the camera’s memory buffer. Once the pressure has been reduced sufficiently, free gas bubbles will start to form within the cell. The pressure at which the first gas bubbles are observed is the bubble point pressure and this pressure is the main result that is analysed in this study.

2.3. Data analysis methods

The main results from each experiment are a set of images that show the emergence of bubbles during the degassing process and the accompanying pressure log. An image analysis routine was developed using the MATLAB Image Processing Toolbox to identify individual bubbles and their properties such as size and growth rate. Since the images captured during the experiment and the pressure log are synchronised, the bubble data from the analysed images are combined with the pressure data to determine at which pressure bubbles are formed during the degassing process.

The image analysis consists of the following steps: first the background image (i.e., the image at the start of the experiment before bubble formation) is subtracted, resulting in net images with bubbles showing up as regions of high intensity. These net images are then converted to a binary format. Individual bubbles are identified in these images using MATLAB’s regionsprops function. Finally, the number of bubbles (i.e. individual regions) are counted on each image and can thus be plotted as a function of pressure.

2.4. Overview of experiments

Table 1 shows an overview of the experiments that were carried out within this study and lists their experimental conditions. Different brine compositions, temperatures and CO2 concentrations were studied. For some of the experiments this means that CO2 can already start forming bubbles while it is in its supercritical state, since the critical point is at: pcrit,CO2 = 73.8 bar, Tcrit,CO2 = 31.0 °C.

Table 1: Overview of experimental conditions for elevated temperature experiments.

Brine composition


Temperature (°C) Initial CO2 concentration [mol/kgw]
1 M NaCl 40, 60, 100, 150 0.20
1 M NaCl 40, 60, 100 0.50
1.5 M CaCl2 + 2 M NaCl 40, 60, 100, 120 0.15

2.5. Model description

This study mainly focuses on determining the bubble point pressure as a function of the studied parameters. The experimental results are compared to a geochemical model that predicts the solubility of CO2 in brine. This model uses the R coupled version of PHREEQC [12]. PHREEQC (pH-REdox-EQuilibrium) is a software package for simulating chemical reactions and transport processes [13].

At low pressures and following Henry’s law the molar fraction of CO2 in the aqueous phase depends on the partial pressure of the gas phase (PCO2(g)) divided by Henry’s constant KH:

For high pressures the partial pressure needs to be replaced by the fugacity:

where ϕ is the fugacity coefficient of CO2. This can be computed by an Equation of State (EoS: [14, 15] or any other variant of a cubic or virial EoS); the exponential term is often called “Poynting correction”; vCO2 is the partial molar volume of CO2 (aq) averaged over the P range [1,PCO2]. In the following, vCO2 is equal to 0.032 L/mol [16]. At specified pressure and temperature conditions the PHREEQC model requires CO2 fugacities as an input, which were computed with individual EoSs.

3. Results and discussion

3.1. Comparison of experimental and modelled bubble points

Figure 2 shows the modelled CO2 solubility in 1 M NaCl brine as a function of pressure and temperature with input CO2 fugacities computed using an EoS described by [14]. The contours in the plot are iso-solubility (mol/kgw). Experimentally determined bubble point pressures for this same brine are given in the same plot for comparison for both the 0.20 and 0.50 mol/kgw CO2 concentrations, which is why these contours are plotted using thicker lines. For the higher concentration experiments, the measured bubble points coincide fairly accurately with the predicted solubility, that is, the measurements are close to the 0.50 mol/kgw contour. This is not the case for the 0.20 mol/kgw experiment, where the measured bubble points are significantly lower than the model predicts. Measured bubble points here are between the 0.10 and 0.15 mol/kgw contours.

An even better fit to the high concentration experiments can be obtained by using another EoS. Here we use the volume translated version of the Original Soave-Redlich-Kwong (SRK: [17]) EoS as described by [18]. In this case, the EoS’s parameter values are fitted to the experimental data to obtain an improved match between experiment and model, as shown in Figure 3. Visual observation of the model prediction shows a better fit for the 100 °C experiment compared to the Duan and Sun EoS. Still, a good fit was not found for the lower concentration experiment, where data points remain significantly lower than the model predicts. The experimental approach can introduce a certain degree of error in the results, which may help explain the observed deviations between experiments and model. In order to assess the impact of the experimental approach, the experiments were first repeated using the same conditions. A comparison between the original and repeat experiments allows for determining whether the experimental procedure was carried out correctly as both experiments should ideally give the same results. Both sets of experiments were found to yield very similar results with variations in bubble point pressure limited to just 1 bar. Therefore, the experiments are considered to have been performed correctly. Secondly, the image analysis procedure was changed by lowering the threshold in the binary conversion step. This implies that smaller bubbles can be picked up earlier, but also that noise can erroneously be identified as bubbles. This still only caused very small deviations in the obtained bubble points, again with differences of around 1 bar compared to the values found earlier. So neither the repeat experiments nor the change in the image analysis are sufficient to explain the discrepancy between measurement and model prediction, requiring further investigation.

Figure 2: P-T dependence of the CO2 solubilities in a 1 M NaCl brine solution. Results are obtained with PHREEQC and CO2 fugacities computed using the Duan and Sun (2003) EoS. Experimentally obtained bubble points are also given for both the 0.20 mol/kgw (●) and 0.50 mol/kgw (■) CO2 concentrations.

Figure 3: P-T dependence of the CO2 solubilities in a 1 M NaCl brine solution. Results are obtained with PHREEQC and CO2 fugacities computed using the volume translated SRK Eos. Experimentally obtained bubble points are also given for both the 0.20 mol/kgw (●) and 0.50 mol/kgw (■) CO2 concentrations.

Figure 4 shows the model prediction of CO2 solubility for the higher salinity brine (1.5 M CaCl2 + 2 M NaCl) using the EoS of Duan and Sun along with the experimentally obtained bubble points using a CO2 concentration of 0.15 mol/kgw. The lower concentration compared to the experiments using the 1 M NaCl brine was chosen here due to the model predicting significantly lower CO2 solubility here. Results are thus compared to the 0.15 mol/kgw solubility contour, hence the increased line thickness for this contour.

Experimentally determined bubble point pressures found here agree reasonably well with the model predictions, although the deviations increase with increasing temperatures. The model also predicts increased CO2 solubility for this concentration at temperatures beyond 100 °C. This behaviour was not found in the experiments, as the bubble point pressure at 120 °C is significantly higher than that at 100 °C. Additional experiments at higher temperatures are recommended to validate the model predictions.

Figure 4: P-T dependence of the CO2 solubilities in a 1.5 M CaCl2 + 2 M NaCl brine solution. Results are obtained with PHREEQC and CO2 fugacities computed using the Duan and Sun (2003) EoS. Experimentally obtained bubble points are also given for a CO2 concentration of 0.15 mol/kgw (■).

3.2. Gaseous vs. supercritical bubbles

As mentioned above, some experiments performed here already have bubbles performing above the critical point of CO2. This is especially true at higher temperatures and CO2 concentrations. Figures 5a and b show the bubbles for the experiments using the 1 M NaCl brine with 0.5 mol/kgw CO2 concentration at 40 and 100 °C respectively. At 40 °C, bubbles only start forming at a pressure of ~40 bar, which means that here bubbles are gaseous, as the pressure is below the critical point. This results in fairly large bubbles that quickly rise up through the aqueous phase. For the 100 °C experiment, the bubbles already start forming at a pressure of ~125 bar, which is above the critical point of CO2, and thus bubbles are supercritical here. These bubbles are much smaller compared to the gaseous bubbles at lower temperatures, which causes them to be much less buoyant and not to rise as fast inside the cell. The rate of bubble formation is also much larger for supercritical bubbles, as shown in Figure 6, which shows the observed number of bubbles as a function of pressure. Almost all of the supercritical bubbles form instantly as the bubble point is reached, obscuring the entire visual window and thus making it hard to distinguish individual bubbles. The gaseous bubbles at lower temperatures form far more gradually, starting with only a few bubbles near the bubble point and then gradually increasing as the pressure is reduced further.

Figure 5: Experiments using 1 M NaCl brine and 0.5 mol/kgw CO2 concentration showing (a) gaseous bubbles at 40 °C and (b) supercritical bubbles at 100 °C.

Figure 6: Number of bubbles vs. pressure for the experiments using 1 M NaCl brine and 0.5 mol/kgw CO2 concentration at 40 and 100 °C, respectively.

4. Conclusions and recommendations

  • For the 1 M NaCl brine, the bubble point of a mixture with a high, 0.50 mol/kgw, CO2 concentration, a good match is found between experiments and model for the investigated pressure and temperature conditions using either the Duan and Sun or volume translated SRK EoS model.
  • At lower CO2 concentrations, significant differences are found between the model and experiment. These deviations may be caused either by errors in the experimental approach or the model. Further evaluation of the degree to which these errors affect the results is recommended. The model description may also be improved by using pressure or temperature dependent formulations for certain physical parameters.
  • The higher salinity brine (1.5 M CaCl2 + 2 M NaCl) has approximately three times lower CO2 solubility compared to the 1 M NaCl brine. A good match is found for experiment and model for this brine using a CO2 concentration of 0.15 mol/kgw, although deviations between experiments and the model increase at higher temperatures.
  • Supercritical bubbles are much smaller and also form in much greater numbers at or near the bubble point pressure compared to gaseous bubbles.
  • Further experiments at higher temperatures are required to validate the solubility behaviour that is predicted by the geochemical model. The experimental approach (i.e., high-speed imaging inside a visual cell) used here is well suited for performing experiments at higher temperatures, although some parameters can be improved upon to capture the process in greater detail. These include running the camera at a higher frame rate and using a more zoomed-in view of the visual cell. This is especially relevant when observing the formation of the supercritical bubbles, which are very small and form almost instantaneously in large quantities.


As part of the REFLECT project, this project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No. 850626. The authors gratefully acknowledge this support. Technical support from Michiel Slob in setting up the lab experiments at Delft University of Technology is also acknowledged.


  1. Cosmo, R. d. P., Pereira, F. d. A. R., Ribeiro, D. d. C., Barros, W. Q., Martins, A. L. , Estimating CO2 degassing effect on CaCO3 precipitation under oil well conditions. Journal of Petroleum Science and Engineering 181, art. 106207 (2019).
  2. Fu, Y., van Berk, W., Schulz, H.-M. Temporal and spatial development of scale formation: One-dimensional hydrogeochemical transport modeling. Journal of Petroleum Science and Engineering 112, 273–283 (2013).
  3. Arnórsson, S., Deposition of calcium carbonate minerals from geothermal waters — theoretical considerations. Geothermics 18(1), 33–39.
  4. Stefánsson, A., et al., Quantifying mixing, boiling, degassing, oxidation and reactivity of thermal waters at Vonarskard, Iceland. Journal of Volcanology and Geothermal Research 309, 53–62 (2016).
  5. Pátzay, G., Stáhl, G., Kármán, F.H., Kálmán, E. Modeling of scale formation and corrosion from geothermal water. Electrochimica Acta 43(1), 137–147 (1998).
  6. Rodríguez, F., et al. Exploration of deep-seated geothermal reservoirs in the Canary Islands by means of soil CO2 degassing surveys. Renewable Energy 164, 1017–1028 (2021).
  7. Chiodini, G., et al. Carbon dioxide degassing at Latera caldera (Italy): Evidence of geothermal reservoir and evaluation of its potential energy. Journal of Geophysical Research: Solid Earth 112(B12) (2017).
  8. Frondini, F., Caliro, S., Cardellini, C., Chiodini, G., Morgantini. N. Carbon dioxide degassing and thermal energy release in the Monte Amiata volcanic-geothermal area (Italy). Applied Geochemistry 24(5), 860–875 (2009).
  9. Peiffer, L., et al., Soil degassing at the Los Humeros geothermal field (Mexico). Journal of Volcanology and Geothermal Research 356, 163–174 (2018).
  10. Blöcher, G., et al., Hydraulic history and current state of the deep geothermal reservoir Groß Schönebeck. Geothermics 63, 27–43 (2016).
  11. Chandrasekharam, D., Bundschuh, J. Low-Enthalpy Geothermal Resources for Power Generation. London: CRC Press. 2008.
  12. De Lucia, M., Kühn, M. Geochemical and reactive transport modelling in R with the RedModRphree package. Advances in Geosciences 56, 33–43 (2021).
  13. Parkhurst, D.L., Appelo, C.A.J. Description of input and examples for PHREEQC version 3– A computer program for speciation, batch-reaction, one-dimensional transport, and inverse geochemical calculations, in U.S. Geological Survey Techniques and Methods, book 6, chap. A43. Reston, VA. 2013. p. 519.
  14. Duan, Z., Sun, R. An improved model calculating CO2 solubility in pure water and aqueous NaCl solutions from 273 to 533 K and from 0 to 2000 bar. Chemical Geology, 193(3), 257-271 (2003).
  15. Peng, D.-Y., Robinson, D.B. A New Two-Constant Equation of State. Industrial & Engineering Chemistry Fundamentals 15(1), 59–64 (1976).
  16. Spycher, N., Pruess, K., Ennis-King, J. CO2-H2O mixtures in the geological sequestration of CO2. I. Assessment and calculation of mutual solubilities from 12 to 100°C and up to 600 bar. Geochimica et Cosmochimica Acta 67, 3015–3031 (2003).
  17. Soave, G., Equilibrium constants from a modified Redlich-Kwong equation of state. Chemical Engineering Science 27(6), 1197–1203 (1972).
  18. Le Guennec, Y., Privat, R.,  Jaubert, J.-N. Development of the translated-consistent tc-PR and tc-RK cubic equations of state for a safe and accurate prediction of volumetric, energetic and saturation properties of pure compounds in the sub- and super-critical domains. Fluid Phase Equilibria 429, 301-312. (2016).

This article has been published in European Geologist Journal 54 – Geothermal energy – A geological contribution to the energy transition

Read here the full issue: