| Resources are the key to quality and sustainable development under national policy.Geological exploration is gradually shifting towards unconventional,low-permeability,high-density,and ultradeep wells in response to the growing difficulty of finding large oil and gas deposits and mineral reservoirs at shallow and medium depths.Promoting exploration to the depths is the key to innovation in drilling technology.Unlike shallow geological structures,deep geological structures are subject to complex factors,and drilling is associated with serious safety and efficiency challenges.It is common to encounter complex operating environments in the drilling process,such as extreme pressure,high temperatures,and variable BHP.This can result in technical difficulties such as frequent changes in work conditions,difficulty monitoring the system’s operation status,and low density of data values.Intelligently monitoring and evaluating geological drilling processes is one of the most effective means of ensuring safe and efficient drilling.Considering the characteristics of the drilling process,an intelligent monitoring method based on data feature resolution is proposed to achieve high-performance monitoring of the drilling process.The main works in this thesis are as follows:(1)Intelligent monitoring schemes for geological drilling processesWith the goal of efficient and safe drilling,an intelligent monitoring solution is proposed for the complex geological drilling process.By integrating mechanism analysis and data-driven methods,a distributed monitoring model is established to identify abnormal drilling factors when the operating status fluctuates and switches.Likewise,a multivariate statistical analysis model was developed to solve the problem of assessing the drilling operation process’ s operational quality.From the perspective of engineering quality management,a performance assessment strategy with multiple modes is proposed to determine the current performance grade.(2)Distributed drilling operation condition monitoring based on integrated probabilistic principal component analysisGeological drilling process is affected by the complex geological environment and the disturbance of mechanical equipment.As a result,the operating status often deviates from the set interval,resulting in drilling accidents and property damage.To improve the utilization of instrument measurement data,a distributed monitoring method using integrated probability principal component analysis(IPPCA)with minimal redundancy and maximum relevance is developed to address the problem of monitoring the operational status of the drilling process.The related process variables are divided into sub-blocks by using the minimal redundancy maximal relevance(m RMR)algorithm.In the following phase,local detection was combined with IPPCA to construct global monitoring statistics to realize the entire monitoring scheme.Lastly,the feasibility and superiority of the proposed method are verified through real-world production processes.According to the experimental validation of the field operation data,the average false alarm rate and non-detection rate are respectively 5.45 % and 4.42 %,illustrating the feasibility and engineering adaptability of the proposed method.(3)Decentralized operating performance assessment method based on multi-block total projection to latent structuresAs drilling is a complex process involving several process variables and working modes.When disturbed by the external environment,the system’s operating state can drift and cannot always operate under optimum conditions.Therefore,timely,accurate,and comprehensive operational status evaluation is critical to process improvement.Geological drilling is a complex industrial process that involves multiple systems and stratigraphic uncertainties,making it difficult to assess its performance.Using multiblock total projection to latent structures(T-PLS),a decentralized operating performance assessment is proposed for the geological drilling process.The most related variables can be grouped in the same block according to the variational trends of the detection variables.Following this,a T-PLS algorithm-based operating performance assessment model and Bayesian inference is realized.The online assessment is conducted by calculating the scoring vector based on the offline model and incoming data.It is then compared with each state’s feature vector to achieve an online evaluation of the operating state.Whenever a drilling system is not operating optimally,the variable contribution analysis method is used to identify the dominant factors and to correct the operation as soon as possible.Experiments demonstrated an accuracy rate of 88.4 % using operational data from the drilling site.This is better than the current standard operating evaluation method and demonstrates the proposed method’s effectiveness in accurately describing the current status.(4)Operating performance assessment method with multiple drilling modesConsidering process quality management in the drilling process,the performance of drilling operations is divided into several grades throughout the production period.In essence,it involves the evaluation and classification of time series.Since drilling is a batch process with multiple working modes,establishing an accurate model for estimating performance levels is difficult.However,due to subjective and external factors such as the geological environment,it is impossible to accurately and reasonably classify and describe the operational performance grade of the system.Motivated by the above discussions,a performance assessment for multiple modes of the drilling process is presented.First,variables were selected according to Pearson index and information entropy theory to minimize the interference of irrelevant variables in the modeling process.The process capability index is used as a performance indicator to add performance labels to drilling data.Then,drilling operating modes are classified according to different process characteristics.Finally,an operating performance assessment model is performed,and real-time drilling performance grades can be obtained.As a result of the proposed method,geological constraints are overcome and better assessment results are obtained.With an average accuracy rate of 87%,the proposed method can be used to analyze and evaluate the drilling operation in a comprehensive manner.(5)On-site implementation and engineering applicationSeveral geological resource exploration wells in China have been industrially installed based on the intelligent monitoring and assessment scheme presented in this paper.This is to verify the effectiveness and engineering applicability of the proposed drilling intelligent control system.To meet site process and construction requirements,the algorithms are adjusted and optimized according to the conditions at each well site.During the practical application period,the drilling intelligent control system integrated with the proposed method demonstrated the ability to achieve high-performance monitoring and accurate system operation status evaluation at the drilling site,as well as an increase of 20.79 % in drilling efficiency during the application period,ensuring safe and stable drilling operations during the application period.This can help improve drilling efficiency while ensuring the safety and stability of drilling operations.It is evident from the field application model of the unified function that the intelligent monitoring method has the potential to improve drilling efficiency and has the potential to prove to be both theoretically and practically valuable. |
PDF Full Text Request