Most of oil and gas equipment is welded shell-type structures. Their operation may cause accidents of different complexity. Despite the fact that such structures go through technical diagnostics and life cycle analysis, the current accident statistics does not suggest any improvements. Nowadays there is a rapidly-developing line of research that focuses on creating so-called digital twins. These are numerical models that are based on real mechanical characteristics, dimensions and changes that may occur during the operation of the facility. These models allow using diagnostic data for assessing the remaining service life. While mechanical engineering deals with mass-produced products and has a large database of equipment failures, shell-type oil and gas facilities are mostly specific and their reliability database can barely be useful for predicting their performance. A number of studies on digital twins suggest heterogeneity in physical, mechanical and geometric properties pattern of shell-type structures. Some attempts to cover these properties in numerical models proved that it is actually possible to predict the areas where damage would occur most frequently and cause breakdown. However, in order to keep improving the quality of the modelling system, it is necessary to consider the properties of welded joints.This article deals with analyzing welded joints properties which are essential for developing an algorithm for a digital twin of a welded cylindrical shell. The study revises welded joints of pipe billets made of 09G2S steel. The equipment used meets all the standards concerning welded joints. Properties are distributed unevenly across the welded joint, the weld-affected area and the base metal. These areas are exposed to intensive sound waves within the human hearing range. As a result, they generate secondary acoustic emission signals. When intercepted and analyzed, the signals indicate local heterogeneity. Therefore, welded joints produced by joining weldless billet pipes were inspected by secondary acoustic emission analysis and the Expert-19 device. The test allows analyzing the acoustic properties of a sound wave that went through the welded joint. Before the experiment, it was determined how the measured properties correlate with the distance between the Leeb (D-type) sensor and the microphone that detected the sound wave and specified its characteristics for the numerical analysis. The analysis involved a special computer program. The sound-wave propagation was analyzed along the axes of the shell, its circumference and across the width. Since the properties of the welded shells were measured both before and after the welding, the data on distinctive changes all along the structures’ surface has been obtained for the first time. Also the test has for the first time determined the peculiar features of integral parameter distribution within the entire shell. Response level of acoustic-wave propagation to the heterogeneity has been illustrated in terms of both quality and quantity.