Risk reduction and accelerating drill or drop decisions using DHI and AVO anomalies: a case study from the Nyk High, NCS Pierre-Yves Raya1, Peter Brown2, Jeet Singh2, Stuart Greenwood2, James Raffle2, Ross O’Driscoll2 1: Wintershall Norge, 2: ION Summary In 2016, Wintershall Norge AS and PL847 partners were awarded a production license covering blocks 6706/5 and 6706/6 of the Norwegian Continental Shelf. The area is located in the Vøring Basin on the Nagflar Dome, north of the Hel Graben. The Aasta Hansteen gas development is located approximately 60 kilometres to the South. Four Upper Cretaceous prospects have been identified in addition to the Hvitveis discovery. Hvitveis is a faulted 4-way dip closure. Marisko, and the three other prospects are tilted fault blocks. The proven reservoir in Hvitveis and the targeted reservoir in the prospects are sands of the Upper Cretaceous Nise Fm level (Figure 1). All recent discoveries at the Nyk High and Vema Dome show flat spots or amplitude anomalies related to hydrocarbon fill. In the Greater Aasta Hansteen area, the reservoir properties are generally good, with porosity values exceeding 30% and permeabilities in the Darcy range. However, in the Hvitveis discovery well 6706/6-1, the reservoir properties are much poorer with an average porosity of 15% and estimated permeability values ranging from very tight to 5mD. These poor reservoir properties could partly explain why no obvious flat-spot or hydrocarbon related amplitudes anomalies were observed in Hvitveis or in the four other prospects on the vintage seismic at the time of the application. The initial part of the work program was a re-processing of the available 3D seismic to identify potential direct hydrocarbon indicators and allow a more detailed interpretation and understanding of reservoir distribution. To address these goals, a Pre-Stack Depth Migration was run using a dataset re-processed with the latest de-multiple and de-ghosting technology. Input data Quality and Data processing The input seismic data covered an area of approximately 1250sq.km, acquired in 2000 using a conventional towed source and streamer configuration at 6m and 7m tow depth respectively. During the application phase, the vintage seismic was considered to have optimised acquisition parameters for improvement through pre-SDM reprocessing with today’s algorithms.

Source & receiver de-ghosting proved to be extremely beneficial to this dataset and was applied to suppress side-lobes as well as to boost low frequencies (Figure 2). (Zhou et al, 2012, O’Driscoll et al, 2013).

Figure 1 Arbitrary re-processed seismic line showing Hvitveis and four other prospects targeting sands from the Upper Cretaceous Nise formation.

Risk reduction and accelerating drill or drop decisions using DHI and AVO anomalies

Figure 2 Autocorrelations before (Upper left) and after (Upper right) source & receiver de-ghosting. Lower figure: Amplitude spectrum before (blue) and after (red) source & receiver de-ghosting and associated statistical wavelets showing a clear reduction of side lobes after de-ghosting application. Five tomographic model building updates using Kirchhoff and beam TTI migrations were used (Jones 2010, 2015) to generate a detailed velocity model. Non-parametric generalised move-out picking (GMO) was used to resolve fine velocity detail in the subsurface. (Fruehn et al., 2014, Luo et al., 2014). The main challenge was to adequately pre-condition the data to prevent the GMO picker poorly picking across AVO anomalies with polarity reversals, which would otherwise cause serious problems with tomographic velocity inversions. GMO picking has been tailored to avoid the transition zone between positive and negative polarity (AVO class 2 and 2P) with two sets of picks from near and far angle ranges (Figure 3). Utilising careful consideration five tomographic iterations allowed an extremely accurate velocity model to be generated, producing an excellent geological image which defined fine structural details and provided improved definition of Direct Hydrocarbon Indicators in the form of flat spot events.

Figure 3 Use of angle muted gathers to effectively account, pick and invert for AVO anomalies. Interpretation Figure 4 shows migrated stacks comparisons between the vintage and newly re-processed images. The reduction of side lobes has resulted in a sharper wavelet and enhanced the low frequency data content. This resulted in improved continuity, fault definition as well as more detailed geological structure, including identifications of DHIs.

Risk reduction and accelerating drill or drop decisions using DHI and AVO anomalies

Figure 4 Migrated inline stack vintage (left) and re-processed (right). Note the reduction in side lobes resulting in enhanced frequency bandwidth and leading to an improved continuity, fault definition and more detailed geological structure. In the Marisko prospect, two reservoir intervals, Upper and Lower Nise Fm, of around 250 and 200 metres thick respectively are identified on the newly re-processed seismic. Bright amplitudes conformable with structure and the presence of a flat-spot in some areas probably reflect good reservoir properties. Indeed, seismic modelling shows that an increase of 5% in porosity compared to the in-situ conditions observed in the Hvitveis discovery would favour the development of a clear amplitude response to the fluid. In addition, a flat-spot can only be observed in case of significant increase in Net to Gross. Explanation of the prospect anomalies therefore requires an increase in porosity and sand quality (Figure 5).

Figure 5 Seismic modelling demonstrating that an increase in reservoir properties quality allows the development of an amplitude response to fluid. In the two reservoir intervals, AVO class III anomaly is shown as amplitude brightening on far stack compared to near stack data and as strong positive AVO product (intercept x gradient). The amplitude map extracted at Top Nise horizon shows a very good conformance with depth contours in the area of strong interpretation confidence (eastern part) and where polarity reversal can be followed (Figure 6). The very well-defined flat spot area at 3080m defines the hydrocarbon contact. Reservoir effectiveness is the highest risk for the Marisko prospect. However, the strong DHI response indicates that the porosities are likely to be higher with a corresponding higher permeability, reflecting a location within a more channel setting of the turbidite depositional system.

Risk reduction and accelerating drill or drop decisions using DHI and AVO anomalies

The interpretation of the new seismic data indicates that the lower of the two reservoirs pinches out to the west shown by a thinning of the section plus a truncation of amplitudes indicating the reservoir does not extend beyond the interpreted pinch-out position. Conclusions A combination of improved, modern re-processing, along with detailed Kirchhoff pre-SDM has provided the asset team a much improved, higher resolution seismic image compared with their vintage product. Benefits of broad-band re-processing were shown to produce detailed fault imaging on both small and much larger scale.

This has helped de-risk and improve overall interpretation. DHIs are well observed in the two reservoir intervals and are in accordance with the seismic modelling performed using the Hvitveis discovery well. Amplitudes conform to structure and the near/far angle stacks comparison show proper, consistent AVO in gas and water legs. Acknowledgements The authors would like to thank PL847 partners as well as Wintershall Norge AS and ION for permission to publish this work.

Figure 6 Upper figures: Near and Far stacks showing brightening with offset. The far stack also shows the flat spot as well as the amplitude termination expected from the seismic modelling. Lower figures: positive AVO intercept × gradient in the reservoir interval (left) and RMS amplitude map at top Nise showing the conformability of the amplitude termination with depth contour (right).