| dc.description.abstract | The work presented in this thesis is aimed at studying the potential for applying ultrasonic Doppler measurements during a borehole drilling process using ultrasonic logging-while-drilling (LWD) tools. The primary intention for the application of Doppler measurements is the identification and quantification of influx/efflux of formation fluids/drilling mud through fractures and pore networks in boreholes. This thesis is composed of six chapters and all chapters may be read individually. The results presented in this thesis are generated using laboratory scale experiments and some simulations. This work was funded by the Centre for Innovative Ultrasound Solutions (CIUS), which is a Norwegian Research Council appointed centre for research-based innovation (SFI). CIUS was hosted by the Department of Medical Imaging and Circulation (ISB), at the Faculty ofMedicine at Norwegian University of Science and Technology (NTNU) in Trondheim, Norway.
Chapter 1 provides a brief background about the project and introduces the ultrasonic LWD tools and technologies currently used in the field. A detailed summary of literature on ultrasonic Doppler measurements for borehole applications is provided so as to set the basis for this project. Finally the objectives set for this project during its inception are outlined along with links to relevant chapters within the thesis where they are addressed.
The following chapters are grouped into two independent parts. Part I comprises of chapters 2-4 and are based on the topic of Doppler ultrasound for influx/efflux velocity estimation from borehole fractures. Part II comprises of chapters 5 and 6, and are based on the topic of influx gas detection and quantification in boreholes using ultrasound backscatter.
Chapter 2 discusses the optimal parameters for ultrasonic Doppler acquisition in typical LWD conditions along with the expected upper and lower limits on velocity estimation. The use of short pulse-lengths of 2-4𝜆 for flow velocity estimation is explored along with the corresponding measurement error. The possibility of achieving a high signal to noise ratio of about 30 dB, using instrumentation with technical specifications similar to ultrasonic LWD tools, is demonstrated while using water as the working fluid. The improved imaging capabilities of Doppler ultrasound compared to conventional pulseecho imaging is demonstrated for fractures with fluid influx. Estimation of the fracture shape using the power Doppler image with a resolution approaching the point-spread-function of the transducer is demonstrated.
Chapter 3 extends the work done in Chapter 2 by evaluating the performance of Doppler ultrasound for velocity estimation and fracture shape imaging using real drilling fluids used in the field. Experimental studies using five different drilling fluids under two categories viz. oil based and water based, and density range of 1200 kg/m3 - 1800 kg/m3 are discussed. Experiments studying the influence of drilling fluids, flowing in orthogonal directions mimicking the annular flow and influx/efflux flow, on Doppler measurements is discussed. The effect of attenuation in the drilling fluids and subsequent influence on SNR of the power Doppler images are discussed.
The estimation of fracture shapes using the power Doppler images was done using conventional thresholding approaches in chapters 2 and 3. Chapter 4, which is a collaboration with Sigurd Vangen Wifstad, a fellow CIUS PhD, demonstrates the use of convolutional neural networks (CNNs) to improve the estimation of fracture shapes beyond the point spread function limit. The CNN was trained using simulated power Doppler images from several procedurally generated fracture shapes. The trained model was tested on experimental power Doppler images generated during the work done in chapter 3. The CNN was able to estimate the fracture areas with a significantly lower mean absolute error of 4.9 ± 4.1 mm2, compared to 22.9 ± 1.7 mm2 using the conventional thresholding method. The CNN also enabled the estimation of fracture shape and area using a single frame of the power Doppler image, compared to about 30 frames required for the thresholding method, facilitating a possible increase in scanning speed of the logging tool.
Chapter 5 discusses a method for detecting free gas in drilling fluids from the ultrasonic signal using conventional LWD tools. The use of higher order statistics to make the method noise-robust is also described and demonstrated using simulations and experiments. Gas influx in deep boreholes or under high pressure/high temperature conditions is usually in the supercritical state of matter and as such would be dissolved into the drilling fluid. A concept for applying ultrasonic methods for detecting dissolved gas in deep boreholes is also discussed.
Several prior studies suggest a good correlation between ultrasound backscatter/ attenuation and the gas content in drilling muds, and thereby propose methods for its quantification in boreholes. However, the aforementioned studies neglect the influence of gas bubble size, which can vary significantly over time and has a significant influence on the ultrasound backscatter/attenuation. A model combining existing theories on ultrasound backscatter from bubbles depending on their size is presented. The proposed model is demonstrated using simulations and experiments, where the ultrasound backscatter is evaluated from bubble clouds of varying bubbles sizes. It is shown that the size and number of bubbles strongly influence ultrasound backscatter intensity, and it is correlated to gas content only when the bubble size distribution is independently known. Such information is difficult to obtain under downhole conditions during drilling. Consequently, it is difficult to reliably apply methods based on ultrasound backscatter, and by extension its attenuation, for the quantification of gas content during influx events in a borehole. | en_US |
