FIELD: geophysical survey.
SUBSTANCE: invention relates to field-geophysical research and is intended for determination of volume fractions of fluids along well bore. In compliance with proposed method of determining volume distribution of fluids in well bore for at least one well at least once, training and test data sets are formed for machine learning algorithm, containing distribution of acoustic signals along well bore, structural parameters of well, affecting shape of spectrum of radial resonance modes in acoustic signal, as well as values of volume fractions of fluids corresponding to depth and time of registration of said acoustic signals along well bore. At least one machine learning algorithm is selected with the teacher. At least once, based on the formed training data set, features are generated for the selected machine learning algorithm, containing information on distribution of spectral density of power of acoustic signals along the wellbore, and generating responses containing information on values of volume fractions of fluids along well bore. Specifying the training quality metric. At least once the selected machine learning algorithm is trained on the basis of the formed training set of data, formed features and responses. Then the trained machine learning algorithm is checked for compliance with the specified quality metric using the generated test data set. Acoustic noise is recorded in the wellbore at a given depth interval for given structural parameters of the well and features are formed, which contain information on distribution of spectral density of power of acoustic signals along the length of the well, similar to features used for machine learning algorithm, after which distribution of volume fractions of fluids along well bore is determined using tested trained algorithm of machine training.
EFFECT: technical result of invention is high accuracy, veracity and reliability of determining volume fractions of fluids along well bore.
10 cl, 6 dwg, 2 tbl
Title | Year | Author | Number |
---|---|---|---|
METHOD FOR DETECTION OF INTERVALS OF INFLOW AND ABSORPTION OF FLUIDS IN OPERATING OIL-GAS WELLS | 2022 |
|
RU2788999C1 |
METHOD FOR BOREHOLE ACOUSTIC NOISE LOGGING | 2021 |
|
RU2787265C1 |
METHOD OF DETERMINING WORKING INTERVALS OF DEPTHS OF OIL AND GAS FORMATIONS | 2019 |
|
RU2728123C1 |
METHOD OF DETERMINING CHARACTERISTICS OF FILTRATION FLOW IN A BOREHOLE ZONE OF FORMATION | 2019 |
|
RU2728121C1 |
METHOD FOR CHOOSING THE OPTIMAL HYDRAULIC FRACTURING DESIGN BASED ON THE INTELLIGENT ANALYSIS OF FIELD DATA TO INCREASE THE PRODUCTION OF HYDROCARBON RAW MATERIALS | 2021 |
|
RU2775034C1 |
METHOD FOR DETECTING ZONES OF SOLIDS DISCHARGE THROUGH AN IMPERMEABLE BARRIER IN A WELL | 2020 |
|
RU2749589C1 |
METHOD AND DEVICE FOR BOREHOLE SPECTRAL NOISE LOGGING | 2012 |
|
RU2499283C1 |
METHOD FOR DETERMINING PROPERTIES OF GAS CONDENSATES AND VOLATILE OILS | 2024 |
|
RU2825279C1 |
METHOD OF EVALUATION OF PHASE PERMEABILITY PROFILE IN OIL AND GAS PRODUCTION WELLS | 2019 |
|
RU2707311C1 |
METHOD OF DETERMINING PHASE DEBITS IN OIL WELL | 1996 |
|
RU2097554C1 |
Authors
Dates
2020-07-28—Published
2019-12-20—Filed