الفهرس 
Regional Geologic SettingOnly 14 pages are availabe for public view
 |  | from | 356 |  |  |
| | | from | 356 | | |
Abstract
The Nile Delta basin is the most prolific, prospective gas and condensate province in Egypt. Nile Delta is the main gas producing province in the northern part of Egypt with approximately 62 TCF proven reserves. The study area is Hapy Field which lies in Ras El Barr concession, in the eastern part of the offshore Nile Delta, about 40 km away from shoreland, in 80 m of water.
The integration analysis between petrophysical evaluation of well logs for Kafr El-Sheikh reservoir units and different seismic interpretation techniques in Hapy Field prove to be useful for accurate delineating the lateral and vertical extension of the sand reservoir units (A22, A24, A30 and the main sand unit of A20). Also, it is evidence for discriminating between the gas-bearing sand, brine sand and shale since the Base Gas and Base of the main sand unit of A20 are clearly overlapped in several seismic sections.
Finally, the gas reserve of the A20 reservoir is estimated based on geobody extraction technique to show the efficient profits of the reservoir.
The available data for this study consists of well and seismic data. The well data includes a complete set of well logs for six wells, composite logs and deviation surveys for the wells, MDT data for two wells, core reports for two wells and check-shot data for two wells. The seismic data is a 3D poststack seismic cube (MAZ volume). Techlog (2015.3), Petrel (2017.4), OpendTect (6.6) and Humpson-Russel (HRS 10.1) softwares have been used in this study.
Well logging analysis of the Kafr El-Sheikh reservoir units for the available six wells in the study area resulted in the following conclusions:
1. The main pay zone in Hapy Field is A20 sand which exists in all the available wells with a large thickness and optimum effective porosity and hydrocarbon saturation values. The A22 pay zone exists only in two wells; H-1 well as a thin sand layer and H-4 well as sand mutually intercalated with shale. The A30 pay zone exists only in two wells; H-2-ST-2 well as three sand units separated by shales and H-8 well as four sand units separated by shales. Also, a thin pay zone (A24) exists only in H-2-ST-2 well and pinching out into shale in the other wells.
2. Analysis of the pressure test data shows that, the gas-bearing sand members of H-1 well are separated from each other and have different fluid densities. However, the gas-bearing sand members of H-8 well are vertically connected. The analysis of the multi-well pressure depth plot illustrates that, there is no horizontal connectivity between the sand reservoir units of H-1 and H-8 wells, due to the production from different reservoirs, as well as producing from two sand bars isolated by a permeability barrier of shale.
3. Iso-parametric maps of A20 sand illustrate that, shale volume decreases westwards and southwards, while effective porosity increases westwards and southwards to a maximum value in the southeast where hydrocarbon saturation and net pay thickness have maximum values, which indicates that, the southeastern side of the field is the most promising area unlike the northeastern side which exhibits minimum values of effective porosity, hydrocarbon saturation and net pay thickness.
Seismic data interpretation involved seismic features analysis to delineate A20 reservoir and discrimination between Base gas and Base of A20 reservoir. Synthetic seismograms were generated and used with time-depth charts to tie well to seismic. Time maps were constructed for Top, Base gas and Base of A20 reservoir clarifying the four faults affecting it, then they were converted to depth maps using an average velocity equation derived from check-shot data.
Seismic interpretation concludes that, A20 reservoir is trapped in a tilted fault block between a major northeast-southwest (NE-SW) listric growth fault and a major northwest-southeast (NW-SE) listric growth fault. The down-thrown side of both faults is toward the north. A20 sand is also dissected by two minor faults trending west northwest-east southeast (WNW-ESE) which splay from the major northeast-southwest (NE-SW) fault.
The four seismic attributes that have been studied in this study are:
• 3-D amplitude auto-tracking attribute.
• Spectral decomposition attribute.
• RGB color blending attribute.
• Geobody extraction attribute.
The 3D amplitude auto-tracking for the interpreted horizons (Top A20, Base Gas A20 and Base A20) illustrates the gas and water distribution within A20 reservoir effectively. Spectral decomposition (SD) is a technique which breaks down the seismic signal into narrow frequency sub-bands. When these sub-bands are examined in a spatial context (i.e., plan view), they reveal interference which is happening through the available signal bandwidth. Hence, it makes use of much lower seismic frequencies to image the reflective nature of the subsurface rock. This decomposition gives higher resolution and detection of the layer stacking boundaries, heterogeneity and thickness variation than are possible using conventional broadband seismic attributes. A tuning cube was constructed by using SD-tuning cube module of OpendTect software (6.6). The used frequency range for creating the tuning cube is 5-65 Hz. OpendTect software (6.6) extracts and flats this slab of data, then apply discrete Fourier transform to this slab of data to generate the tuning cube. According to the amplitude spectrum of the seismic data, all frequencies have been analyzed with increment of 5 Hz starting from 5 Hz up to 65 Hz. It is noticed that, the best frequencies representing sand bodies are the lower frequencies (less than 30 Hz) because of their large subsurface depths and the best low frequencies which delineate the sand units are 5 Hz, 20 Hz and 30 Hz. RGB color blended maps were constructed by mixing these three frequencies into the same plan view of a 3D survey. A set of RGB color blended maps for A20 reservoir were constructed. These maps delineate the geometry and lateral extension of the sand reservoir units. They illustrate that, the sand bars are trending toward the west. Geobody has been extracted from the RGB color blended volume which delineates the lateral and vertical extension of the A20 reservoir unit.
The gas reserve of A20 reservoir has been estimated using the average values of the petrophysical evaluation results for the available wells, the area of the geobody extracted from the RGB color-blended cube and reservoir engineering parameters from PHPC internal report. The calculated gas reserve was 2.527 trillion cubic feet (TCF) of gas which is slightly higher than the A20 gas reserve calculated by PHPC company of 2.107 TCF of gas using the results of additional 9 wells which are not available for us in this study.
Rock physics analysis was performed to create a bridge between elastic properties (Vp/Vs ratio, density, acoustic impedance, Poisson’s impedance, bulk modulus, shear modulus, young’s modulus and Lamè parameter) and reservoir properties (shale volume, effective porosity and hydrocarbon saturation) to guide for lithology and fluid discrimination. In this study, Poisson’s impedance has proved to give the best discrimination between sandstone and shale lithologies in addition to the best discrimination between the gas-bearing and water-bearing sand as compared to the other attributes.
The model-based inversion technique is used to extract complete volumes of the reservoir properties (shale volume, effective porosity and hydrocarbon saturation) from the inverted P-impedance by creating a relationship between these reservoir properties and P-impedance logs calculated from the available well logs to derive a linear equation relating between them, then applying these equations on the inverted P-impedance volume. These reservoir properties volumes show perfect matching with the inverted P-impedance which proves its efficiency. Then these reservoir properties were estimated using three methods of neural network; single attribute, multi-attribute and Probabilistic Neural Network (PNN). It is noticed that, PNN gives the lowest error and highest correlation between the original and predicted logs. Therefore, the PNN method was used to predict complete volumes of these properties all over the seismic data. Horizon slices are created through the inverted P-impedance volume and the other reservoir properties volumes derived from the inverted P-impedance and predicted using PNN. These horizon slices show perfect discrimination between the gas-bearing sandstone, brine sandstone and shales. These horizon slices indicate that, the reservoir properties volumes predicted using PNN are relatively more precise than those derived from the inverted P-impedance.