Feasibility of Using Ambient Fluctuations of Pore Pressure to Characterize Gas Saturation and Other Formation Properties

Water levels in wells usually fluctuate in response to periodic loads caused by barometric pressure or tides, and this response depends on the characteristics of the well and the elastic and fluid flow properties of the formation. The fluctuations in wells produced by variations in barometric pressure or tides are typically small, on the order of 1 cm of water or less, but they can be measured using readily available pressure transducers. Theoretical analyses are available that link the phase lag and amplitudes of the periodic pressure fluctuations in wells to formation characteristics. This has led to a method for interpreting water level fluctuations in wells to estimate formation properties as an alternative to pumping or slug tests. This method is appealing because it requires minimal labor and can be used to characterize temporal changes in properties, such as permeability changes following earthquakes, for example. Pressure fluctuations in wells should be sensitive to changes in gas saturation, which would make this technique attractive for monitoring storage of CO2 or natural gas, production of natural gas, air sparging for remediation, or other subsurface process where the gas content may change. Small changes in pressure in deep wells are often only detectable when the well is shut-in, or isolated from the atmosphere, a configuration that is not included in the available analyses. Moreover, these analyses assume that the well is perfectly coupled to the formation, whereas many wells are enveloped by a low permeability skin that will likely affect the response to periodic loads. The objective of this thesis is to evaluate the feasibility of analyzing ambient fluctuations of pressure in wells to estimate gas content and other formation characteristics, for wells that may be shut in and affected by well skin. A method is proposed that considers the effects of fluid compressibility when the wellhead is sealed from the atmosphere or tidal influences. The skin factor, which is commonly used in well testing analysis, is also included in the analysis of ambient pressure fluctuations. Two different cases were studied in this work. One case used data from three producing wells in shut-in conditions located in Oselvar site, an offshore oil/gas reservoir, where a periodic load was applied by variations in pressure on the seafloor caused by the ocean tides. Another case study used data measured in a monitoring well in a confined aquifer near Clemson, South Carolina. Barometric pressure caused periodic variations in applied load at the ground surface. Data were analyzed when the well was open and when it was sealed to the atmosphere. The approach for analyzing the observed data involves characterizing the phase lag and amplitude ratio between the observed pressure fluctuations and the periodic applied loads. The theoretical analysis was used to create a plot relating phase shift, storage coefficient, amplitude ratio, shut-in correction term and transmissivity for different values of skin factor. The plot appears to be a convenient and practical tool to estimate formation properties and well skin, although numerical inversion of the data is also possible. The ocean tides show strong signals in deep formations, with different responses according to different locations. The estimated formation transmissivity in Oselvar site is between 0.4E-6 and 1.9E-6 m2/s, and is similar to the known value for the site, in the order of 0.8E-6 m2/s. The values of gas saturation found were between 0 and 0.04, always with the maximum gas content near wells A-1 and A-2. Well A-3 has the lower gas content values. The skin factor ranges between 0 and 2 for Oselvar site, and the theoretical shut-in correction factor is in the order of 0.43. In the case of Clemson site, the pressure data is dominated by the confining unit, with an estimated transmissivity of 0.5E-6 to 3.2E-6 m2/s, while values of transmissivity obtained by slug tests are between 0.8E-6 and 3.1E-6 m2/s. Gas is not detected in the near well area, and the skin factor is in the order of 10 or higher. The theoretical shut-in correction factor is 0.11, while the one estimated graphically is 0.03. The consistency on the calculated formation property values at the Oselvar site over time and space supports the conclusion that the proposed methodology is feasible to be used in deep wells to determine gas saturation and other formation properties. In addition, the graphical method provides a practical tool to evaluate the skin and shut-in effects. The method looks encouraging, but from the results at the Clemson site the need arises to evaluate the effect of heterogeneities on the formation response to surface load