Development and Application of Fiber-Optic Sensing Technology for Monitoring Soil Moisture Field

Accurate acquisition of the moisture field distribution in in situ soil is of great significance to prevent geological disasters and protect the soil ecological environment. In recent years, rapidly developed fiber-optic sensing technology has shown outstanding advantages, such as distributed measurement, long-distance monitoring, and good durability, which provides a new technical means for soil moisture field monitoring. After several years of technical research, the authors’ group has made a number of new achievements in the development of fiber-optic sensing technology for the soil moisture field, that is, two new fiber-optic sensing technologies for soil moisture content, including the actively heated fiber Bragg grating (AH-FBG) technology and the actively heated distributed temperature sensing (AH-DTS) technology, and a new fiber-optic sensing technology for soil pore gas humidity are developed. This paper systematically summarizes the three fiber-optic sensing technologies for soil moisture field, including sensing principle, sensor development and calibration test. Moreover, the practical application cases of three fiber-optic sensing technologies are introduced. Finally, the development trend of fiber-optic sensing technology for soil moisture field in the future is summarized and prospected.

Water Injection Profiling Using Fiber Optic Sensing by Applying the Novel Pressure Rate Temperature Transient PTRA Analysis

Fiber Optic Systems, such as Distributed Temperature Sensing (DTS), have been used for wellbore surveillance for more than two decades. One of the traditional applications of DTS is injectivity profiling, both for hydraulically fractured and non-fractured wells. There is a long history of determining injectivity profiles using temperature profiles, usually by analyzing warm-back data with largely pure heat conduction models or by employing a so-called "hot-slug" approach that requires tracking of a temperature transient that arises at the onset of injection. In many of these attempts there is no analysis performed for the key influencing physical factors that could create significant ambiguity in the interpretation results. Among such factors we will consider in detail is the possible impact of cross-flow during the early warm-back stage, but also the temperature transient signal that is related to the location of the fiber-optic sensing cable behind the casing when the fast transient data are used for interpretation such as the "hot slug" during re-injection. In this paper it will be shown that despite all such potential complications, the high frequency and quality of the transient data that can be obtained from a continuous DTS measurement allow for a highly reliable and robust evaluation of the injectivity profile. The well-known challenge of the ambiguity of the interpretation, produced by the interpretation methods that are conventionally used, is overcome using the innovative "Pressure Rate Temperature Transient Analysis" method that takes maximum use of the complete DTS transient data set and all other available data at the level of the model-based interpretation. This method is based on conversion of field measurements into injectivity profiles taking into account the uncertainty in different parts of the data set, which includes the specifics of the DTS deployment, the uncertainty in surface flow rates, and possible data gaps in the history of the well. Several case studies will be discussed where this approach was applied to water injection wells. For the analysis, the re-injection and warmback DTS transient temperature measurements were taken from across the sandface. Furthermore, for comparison, injection profiles were also recorded by conventional PLTs in parallel. This case study will focus mostly on the advanced interpretation opportunities and the challenges related to crossflow through the wellbore during the warm-back phase, related to reservoir pressure dynamics, and finally related to the impact of the method of DTS deployment. In addition to describing the interpretation methodology, this paper will also show the final comparison of the fiber-optic evaluation with the interpretation obtained from the reference PLTs.

Classification and Localization of Low-Frequency DAS Strain Rate Patterns with Convolutional Neural Networks

Distributed acoustic sensing (DAS) has been used in the oil and gas industry as an advanced technology for surveillance and diagnostics. Operators use DAS to monitor hydraulic fracturing activities, to examine well stimulation efficacy, and to estimate complex fracture system geometries. Particularly, low-frequency DAS can detect geomechanical events such as fracture-hits as hydraulic fractures propagate and create strain rate variations. Analysis of DAS data today is mostly done post-job and subject to interpretation methods. However, the continuous and dense data stream generated live by DAS offers the opportunity for more efficient and accurate real-time data-driven analysis. The objective of this study is to develop a machine learning-based workflow that can identify and locate fracture-hit events in simulated strain rate response that is correlated with low-frequency DAS data. In this paper, "fracture-hit" refers to a hydraulic fracture originated from a stimulated well intersecting an offset well. We start with building a single fracture propagation model to produce strain rate patterns observed at a hypothetical monitoring well. This model is then used to generate two sets of strain rate responses with one set containing fracture-hit events. The labeled synthetic data are then used to train a custom convolutional neural network (CNN) model for identifying the presence of fracture-hit events. The same model is trained again for locating the event with the output layer of the model replaced with linear units. We achieved near-perfect predictions for both event classification and localization. These promising results prove the feasibility of using CNN for real-time event detection from fiber optic sensing data. Additionally, we used image analysis techniques, including edge detection, for recognizing fracture-hit event patterns in strain rate images. The accuracy is also plausible, but edge detection is more dependent on image quality, hence less robust compared to CNN models. This comparison further supports the need for CNN applications in image-based real-time fiber optic sensing event detection.