Prediction of reservoir quality in carbonates via porosity spectrum from image logs

Abstract The Majiagou Formation is an important gas-bearing stratigraphic unit in the Ordos basin of west China. The highly complex lithology and heterogeneous pore systems make it difficult to predict the reservoir quality via petrophysical and image logs. This study investigates the lithology, pore systems and well log responses of the Majiagou Formation Member 5 (Ma 5) carbonate reservoirs using a comprehensive analysis of cores, thin sections, routine core analysis, conventional and porosity spectrum derived from image logs. The results show that the lithologies of the Ma 5 are dominantly of mud-sized to silt-sized crystalline dolomite, and there are some dolomites containing gypsums mudstones and limestone. The pore systems include vugs, intercrystalline pores, dissolution pores and microfractures. The vuggy dolomites and the fractured dolomites are the best quality reservoir rocks, and the mud-sized to silt-sized crystalline dolomites containing intercrystalline pores are good quality reservoirs with high porosity and permeability. Open fractured zones are recognized as dark sinusoidal waves, and the vuggy dolomites can be recognized by the dark spots on the image logs. However, the fractured and vuggy dolomites are only occasionally detected, and most of the dolomite reservoirs in Ma 5 contain only intercrystalline pores, which are difficult to be recognized by image logs. In addition, good quality reservoirs have very similar log responses (low natural gamma-ray (GR), low-moderate bulk density (DEN), and moderate-high resistivity) with the poor reservoir quality intervals. By applying the Archie's formula to the flushed zone, a total of 150 porosity curves can be obtained from the XRMI image logs. Then the porosity spectrum is derived from the histogram of porosity distribution for a certain interpretation interval. The best quality reservoir rocks (fractured or vuggy dolomites) have very broad porosity spectrums with long tails due to the presence of vugs and fractures. The abundance of intercrystalline pores containing in the mud-sized to silt-sized crystalline dolomites also result in the broad porosity spectrums without any tails. In contrast, the very narrow porosity spectrums correspond to the poor reservoir quality intervals. The interpretation results are verified by core observations and thin section analysis. By using the porosity spectrum analysis derived from the image logs, the reservoir quality can be predicted. The results can help improve the understanding and prediction of the reservoir quality in carbonates using well logs.