| Abstract | The Makikit_1 well is a vertical exploration well drilled located in the Northern Bonaparte Basin, offshore of Timor-Leste. It’s lay on in the Joint Petroleum Development Area (JPDA) between Timor-Leste and Australia. Where there are many wells, some of which are still producing oil and gas. However, some wells were depleted and abandoned. The study area is located in block 06-102 (JPDA), 13.1 km NE of Bayu_1 well, 17.2 km NW of Mistral_1, 22.3 km E of Elang_1, 13.6 km E of Elang_3 and 44.5 km NW of Fohn_1. Structurally, it is situated about 300 km SW from Dili, the city of Timor-Leste, and 500 km NW from Darwin supply base, Australia. The well intercepted sandstone and shale from Plover and Elang formations, the former was considered as reservoir rock and the later as source rock in the Upper to Lower Jurassic age. The well log curves used in this study are GR, CAL, RHOB, NPHI, LLD, LLS, DTC and DTS. By well log analysis we have identified 6 reservoir zones, zone 4, 6, 8, 10, 12 & 14. The result shows that there is high resistivity in zone 6, which is considered as a hydrocarbon zone while others are water-bearing zone. The porosity by different logs ranges as follows: {u1D719}{u1D441} 0.05 – 0.13, {u1D719}{u1D437} 0.04 – 0.08, {u1D719}{u1D452} 0.06 – 0.10, {u1D719}{u1D460} 0.04 to 0.09. For permeability, we have used two models to calculate i.e., Wyllie_Morris and Wyllie_Timur because the Makikit_1 well has no core data. The permeability by Wyllie Morris’s method ranges from 7.22 to 25.20 mD, and that by Wyllie Timur’s method varies from 0.14 to 1.22 mD. The water saturation was calculated by four models such as Archie, Simandoux, Indonesia, and Kbul, and the result ranges for each model as follows: Archie from 0.46 to 0.82, Simandoux from 0.37 to 0.70, Kbul from 0.03 to 0.16 and Indonesia from 0.43 to 0.88. The water saturation by Kbul’s model saturation result is very low. Consequently, water saturation by Archie, Simandoux and Indonesia models are considered more reasonable. Notably, in this study the machine learning technic (ANN) was applied to predict permeability. The result of machine learning prediction is positive matching quite well with the empirically calculated permeability by Wyllie_Morris model. |