Nanbao shallow reservoir is a typical complex fault block reservoir. According to oil products, 60.9% of the reserves belong to conventional thin oil reservoirs. The reservoir has the geological characteristics of active marginal and bottom water, small fault block and small oil area and has come into extra high water cut stage. Because of low water flooding production rate, it is urgent to carry out technical research to further improve reservoir recovery. The paper regarded the reservoir of NmⅡ4② layer in Gao 63⁃10 block as a typical unit. According to the geological characteristics of small oil sands area and sufficient natural energy, the paper puts forward a new oil drive method named deep huff and puff. Based on the oil well，this new oil drive method can improve the swept volume and flushing efficiency and has the advantages of small scale, low investment and flexible adjustment. It provides technical reference for development and adjustment of similar reservoirs at extra⁃high water cut stage.

When the traditional CFD method is used to solve the Navier⁃Stokes equation of incompressible flow, the pressure equation has an elliptical shape and cannot be solved, and the convergence is poor. Therefore, the traditional CFD method has certain limitations. In order to study the computational efficiency and accuracy of the lattice Boltzmann method compared with the traditional CFD method, taking the thermal diffusion problem of plates with constant temperature boundary as an example, the temperature field between the plates is solved by the finite element volume method, the finite difference method and the lattice Boltzmann method. By comparing the solution results and the number of normalized iteration steps, the feasibility, accuracy and efficiency of the lattice Boltzmann method are clarified. The results show that under the same physical model, the lattice Boltzmann method and the finite difference method are slightly better than the finite element volume method. The lattice Boltzmann method has the fastest solution speed and the least iteration steps, followed by the finite difference method, and the finite element volume method has slowest solution speed.

The prediction and analysis of natural gas demand can provide help for reasonable production scheduling and resource allocation, and then bring economic benefits to the network companies. This paper discusses the current research progress of natural gas demand prediction and other energy demand prediction, analyzes the research status and development direction in this field at home and abroad, and introduces the characteristics and development trend of the models used specially. The problems in current study and future research direction are pointed out, and the research results and the research results that can be used for reference are summarized, which can provide reference for future research on the further development of natural gas demand.

To study the plugging properties of novel profile control agent carboxymethyl starch, it uses laboratory-made solvent carboxymethyl starch block experiment as the plugging time under the same filter volume to determine the performance indicators. It is found that the blocking performance is affected by carboxymethyl starch concentration, solution salinity and temperature and time of storage. With increasing the concentration of carboxymethyl starch and salinity, the sealing performance of the solution is improved; the solution of the filtration time is increasing trend at 70 ℃ and is decreasing trend at 90 ℃ with increasing storage time.