Abstract: In the fracturing process of tight gas reservoirs, due to the high fracturing displacement, water hammer is prone to occur when stopping pump during fracturing , which causes obvious low-frequency oscillations in the surface pressure data, leading to problems such as poor quality of double logarithmic curve for injection pressure drop, difficulties in applying analysis methods in the field, and etc. Due to the pressure dropping data shows a typical characteristic of low-pass filtering signal when stopping pump, the pump-stopping data can be denoised and processed with three kinds of methods, namely Savitzky-Golay filtering, fast Fourier transform and wavelet transform. Apply this method to some well in Daniudi. The on-site analysis results showed that the data quality improved significantly after denoising and processing with wavelet transform, and the fracture parameters could be obtained accurately with processed data. The research method has a very important guiding significance for improving the speed and accuracy of the effect evaluation after fracturing.

Key words: fracturing effect evaluation, pump-stopping pressure dropping data, water hammer denoising, injection pressure drop, unconventional gas field