Adaptive Intrinsic Transducer Noise Removal in Downhole Logging
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current downhole logging methods struggle with accurately determining wellbore properties due to intrinsic transducer noise, which is difficult to remove and varies with well attributes like depth, amplitude, frequency, pressure, and temperature, leading to unreliable measurements.
Innovation Solution
The method involves dividing raw reflected waveforms into subsections based on well attributes to estimate and adaptively subtract intrinsic transducer noise using a noise model and adaptive filtering, allowing for real-time evaluation and removal of noise from the waveforms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If transducer measurements are performed in downhole logging operations, then wellbore properties can be determined, but intrinsic transducer noise contaminates the measurements and reduces reliability
Solution Approach 1:
The patent extracts and removes intrinsic transducer noise from the measured waveforms by identifying and subtracting the noise component that is generated by the transducer itself during measurement operations
Solution Approach 2:
The patent introduces an intermediary noise model that characterizes the intrinsic transducer noise behavior, which is then used to subtract the noise from the actual measurements, acting as a mediator between the raw measurements and the cleaned results
2Measurement precision
If intrinsic transducer noise is removed using traditional methods, then some noise reduction is achieved, but the noise varies with well attributes (depth, amplitude, frequency, pressure, temperature) making removal difficult and unreliable
Solution Approach 1:
The patent makes the noise removal process dynamic by allowing the noise model to adapt to changing well conditions such as depth, temperature, pressure, and frequency, rather than using a static noise characterization
Solution Approach 2:
The patent changes the parameters of the noise model based on well attributes including depth, amplitude, operating frequency, downhole pressure, and temperature to accurately track and remove intrinsic noise under varying conditions
3Loss of information
If lab replication and modeling are performed to simulate transducer noises, then noise characteristics can be understood, but the process is extensive and may be impossible to achieve complete accuracy
Solution Approach 1:
The patent allows the system to characterize its own noise by measuring the transducer's intrinsic noise directly during field operations and building the noise model from actual operational data, rather than relying on extensive laboratory replication
Data Source
AI summary
A method and system for removing intrinsic transducer noises. The method may comprise disposing a measurement assembly into a wellbore, performing a measurement operation at a depth in the wellbore with the measurement assembly to record two or more raw reflected waveforms, identifying one or more intrinsic transducer noises in the two or more raw reflected waveforms, dividing the two or more raw reflected waveforms into one or more subsections, and identifying one or more incoherent measurements in the one or more subsections. The method may further comprise deriving a noise model for each of the one or more incoherent measurements, performing an inversion for each noise model, and applying an adaptive subtraction to remove the one or more intrinsic transducer noises from the two or more raw reflected waveforms.