No abstract

The rate of penetration (ROP) is an important indicator affecting the drilling cost and drilling performance. Accurate prediction of the ROP has important guiding significance for increasing the drilling speed and reducing costs. Recently, numerous studies have shown that machine learning techniques are an effective means to accurately predict the ROP. However, in petroleum engineering applications, its robustness and generalization cannot be guaranteed. The traditional empirical model has good robustness and generalization ability. Based on the quantification of data similarity, this paper establishes a hybrid model combining a machine learning method and an empirical method, which combines the high prediction accuracy of the machine learning method with the good robustness and generalization of the empirical method, overcoming the shortcomings of any single model. The AE-ED (the Euclidean Distance between the input data and reconstructed data from the autoencoder model) is defined to measure the data similarity, and according to the data similarity of each new piece of input data, the hybrid model chooses the corresponding single model to calculate. The results show that the hybrid model is better than any single model, and all the evaluation indicators perform better, making it more suitable for the ROP prediction in this field.

We systematically discussed the principle of the dual-comb ranging system and built a theoretical model. To correct the phase distortion caused by femtosecond frequency comb noise, a correction approach is proposed and numerical simulations are conducted subsequently. In the simulations, the performance of the method under noise of repetition rate and offset frequency is analyzed respectively. The results indicate that ranging accuracy can be effectively improved by the method. It is verified by the experiments.

In marine floating drilling, emergency disconnection of a drilling riser is required in harsh environments or loss of dynamic positioning control. After disconnection, drilling mud in the riser discharges directly from the riser bottom, and seawater refills the vacancy emptied out after mud falling through refill valves. This paper presents two new simulation procedures for this unsteady flow. The first one is a whole fluid column(WFC) model and it is solved by a cubic equation. The second one is a computational fluid dynamics(CFD) procedure, in which two high-resolution CFD schemes are applied for the first time to discrete a special governing equation, and Level Set method is adopted to track the interface between mud column and refilled seawater at each time step. Two methods can respectively provide variations of flow velocity, fluid pressure, whole column weight and flow friction force in a whole mud release duration. In particular, WFC method can easily predict overall trends of several parameters during a whole discharge period, which are prerequisite parameters for dynamic analysis of a riser in hanging state; CFD method is very sensitive to every detailed fluctuation of velocity and pressure in the initial moment of mud discharge, and can be integrated into a structural model for riser recoil response analysis. For a drilling riser with 2150 m, it takes 195.83 s to replace all mud column by seawater, and the maximal discharge velocity is 14.61 m/s. During mud falling, the top of mud column keeps static for 2.29 s before the first pressure wave reaches, and fluid pressure of part column section drops to zero and lasts 1-3 s. In addition, the maximal values of fluid weight-loss and friction force are both close to half of the whole column weight. These results are beneficial for riser system design and risk control of riser recoil.