EAGE Workshop on Velocities: Reducing Uncertainties in Depth

With more than 3000 hydrocarbon wells drilled, the Netherlands ranks as a highly mature petroleum province. Virtually all well planning relies on 3D seismic which is often of high quality. Velocity model building and time-depth conversion is a key step in all depth predictions. EBN, the Dutch state oil company, conducted an extensive review of the technical performance of the different operators in the country. Interesting observations can be made in the area of predicting reservoir depth. The analysis of 253 recent wells disclosed that at least one third of those wells with disappointing results (i.e. lower well rates, smaller volumes proven) suffer from poor depth prognosis. The depth errors at reservoir level range from − 219 m to +130m. The standard deviation amounts to 38 m which equals 1.2% of the (average) depth. Interestingly, there is a clear bias of 10 m in the depth errors towards being deep to prognosis. The number of wells being deep (64%) is almost double the number of wells (36%) being shallow to prognosis. A possible explanation for the bias is given by a mechanism that can be referred to as selection bias. It is important to realize that we do not have a precise knowledge of the subsurface. Our evaluations, including the depth maps, are the result of seismic interpretation and velocity assumptions which do contain noise. If we would conduct random drilling on these maps and compare actuals versus prognosis, -most likely- no bias would show up. However, in reality we do not drill randomly, moreover we put a lot of effort in selecting our targets carefully. In many cases an important selection criterion is structural height. In those cases where modest hydrocarbon columns are probable (or where the contact is already pinned down) often the planned well is aiming for a crestal position. The depth map available, with its inherent uncertainty, is a key factor in guiding the location picking. Due to the uncertainty, the crestal areas, as expressed on the depth map, are partly genuine highs, partly spurious highs. Selection bias acts equivalent to Darwin’s principle. The ranking of the drillable targets in a portfolio is analogous to the survival of the fittest. Whether the selected location was really crestal (fit) or only perceived crestal; that will show only after drilling. This effect will show up as a statistical bias in the depth prognosis errors.

We present a complete technical package to tackle the complex wave propagation and anelastic energy losses associated with mega gas clouds. We started by running FWI to resolve the velocity of the shallow gas clouds, followed by reflection tomography. We then conducted FWI-guided Q tomography to obtain high-resolution absorption model. For the deep gas clouds, since their depth and the incurred low signal-to-noise ratio in the CMP gathers are beyond the limit of these geophysical methods, we moved on with geologically-guided scenario testing in intense collaboration with geologists. Finally, we carried out visco-acoustic TTI reverse time migration (Q TTI RTM) to better deal with the issues of multi-pathing and strong attenuation. This complete package brings significant uplift to the image as compared to the vintage QPSDM result, and therefore can serve as an effective option before turning to C-wave imaging.

Horizontal drilling for development of thin reservoirs requires seismic depth conversion to be more accurate than the bed thickness, which may be of the order of 2% of depth or less. In highly heterogeneous overburden, traveltime tomography may not sufficiently resolve lateral velocity variations causing depth “busts” in the underlying image. Full Waveform Inversion (FWI) provides model resolution to half a seismic wavelength or better, which in turn provides an accurate, data-driven velocity model with reliable depth conversion accuracy away from the wells, following geostatistical velocity model calibration.

At an oil and gas development field in the NW shelf, Australia, 3D TTI FWI resolves the detailed morphology of Miocene carbonates at around 2km depth. Geostatistical scaling of the reflection tomography model did not capture this rugosity and led to development well depth errors at the Triassic reservoir level at 3 km of up to 50 m, just a few hundred metres from exploration wells. The same geostatistical scaling applied to the FWI model provides a statistically stable uncertainty of approximately 15 m, or less than 2% of depth at reservoir level. The robust depth conversion model can be confidently used for future development well planning and reservoir volume estimates.

Currently it is estimated that large percentage of PETRONAS’ reservoirs are affected by gas masking. Reserves estimation accuracy can be improved with enhancement in the structural interpretation through gas cloud. Significant velocity changes in the overburden can cause simple time-to-depth conversions to be in error. Gas cloud (PP) Time-to-Depth Conversion Methodology is carried out to develop and provide advanced standard procedure specifically for converting time structure map to depth in gas cloud affected areas. Synthetic and real data are tested to analyse the complexity and quantify the velocity variation in the area of shallow gas. It is important to establish proper velocity control and predict uncertainty in order to optimize time-to-depth conversion accuracy. Provided that there is no depth processing data, we have demonstrated that the proposed workflow provides high depth accuracy in the area under the gas cloud. We have shown that this method takes into consideration of both seismic velocity and wells provided that there is well or no well inside the gas cloud. With having well inside the gas cloud, there is only 1–4% depth error compared to without well inside gas cloud, 15–37% error.

When exploring in the emerging abrupt margin play, it is often preferable to interpret in depth rather than time to remove the regional tilt on the seismic data associated with the dipping seabed. Such margins typically contain deeply incised seabed canyons. These canyons create significant push-down effects on time seismic data. In addition to the impact on the time structure, PSTM velocities are also affected, which compounds the push-down when they are used in depth conversion.

A two-method is presented for correcting for seabed canyon effects both on time structure and on seismic velocities, to arrive at an interpretable depth seismic dataset from time-migrated seismic. Firstly the time push-down below the canyons is removed from the seismic, and secondly the velocity imprint of the canyons is removed from the processing velocities before they are used in depth-conversion.

A case study is presented of the method being applied on the Equatorial Margin of Brazil, and compared to the PSDM which arrived a year later. Although the imaging from PSDM is significantly improved, the depth structure and velocity field from this method are shown to be comparable to the final PSDM results.

The waveform analysis is a powerful tool to investigate the physical properties in the areas of interest. Since the wave propagation is influenced by all elastic parameters, it is necessary to include these parameters in the inversion. On the other hand, multi-parameter FWI is a challenging problem because plural elastic parameters increases the dimension of the solution space, in other words desensitization of each parameter occurs due to the difference of each radiation pattern.

Some previous works used preconditioned gradient method using approximate Hessian that takes radiation pattern and geometrical spreading into account in order to compensate the desensitization. However, such methods solve many forward calculations so the computational cost becomes expensive.

In this abstract we suggest new preconditioned gradient method that seeks preconditioning operator by scattering theory obtained by Wu and Aki (1985) instead of many forward calculations.

Our study presents the new preconditioned gradient method based on scattering theory and numerical results show the effectiveness of our method.

In a deepwater foldbelt offshore Malaysia, old sediments are strongly folded along with thrust faults and younger sediments are draped on the structures, which indicates change of stress regime with depth. Well logs show different compaction trend between the two sediments as compaction is poroelastically predicted to be driven by mean effective stress.

It is often difficult to analyze pore-pressure in such area because recognizing normal compaction trend in burial depth is difficult, and the common methods of pore-pressure analysis like Eatons’s method and equivalent depth method usually utilize vertical stress only. We propose to extend Equivalent Effective Stress Method to incorporate the change of horizontal stresses which are estimated by stress regime. A case study in the foldbelt offshore Malaysia shows that our approach better agrees with measured pore pressure than conventional method.

It is important to take anisotropy properties into account for estimating accurate geological structure in seismic velocity survey. Most previous studies for wave propagation in anisotropic media (e.g., Thomsen, 1986 ; Bansal and Sen, 2008) assume that the degree of anisotropy is weak. However, it reveals that subsurface materials are much anisotropic than expected and near surface materials could also be strongly anisotropic due to unconsolidated near surface sediments. Therefore, it may not appropriate to apply theories of these previous studies directly to strongly anisotropic subsurface media in seismic exploration and velocity analysis. For improving velocity analysis method, it is necessary to understand the effects of the anisotropy in the behavior of seismic wave propagation in strongly anisotropic media. We found that the influence of anisotropy appears strongly in arrival time of direct P-wave and converted-waves (P-to-S). Our study indicates that we would obtain information about anisotropic property from residual waveforms that show difference between anisotropic and isotropic wavefields.

Well depth was measured until about two decades ago using a relatively standard procedure based on wireline logs, and it was then essentially fit-for-purpose. The measurement was subject to errors but once identified, these were generally understandable by all involved. The measurement and definition of well depth have since become less set due to a combination of complications, including: the frequent use of Logging While Drilling (LWD), of highly deviated well trajectories, of floating rigs, the evolution of tool string configurations, the curtailment of logging programs, the systematic use of computers to store and exploit the depth-based data, etc. Concurrently, the training of the personnel involved in the acquisition, exploitation and management of well data including the well depth has also been reduced, while new or improved data acquisition technologies, integration methods and exploitation techniques require or assume well depths of greater accuracy and precision. The industry is now acquiring higher value / higher cost well data indexed against a depth that has a greater uncertainty than before, and generally without the information required to mitigate and quantify this uncertainty. Remarkably, operators have not expressed widely a concern about this deterioration of well depth.

It is submitted that blunders in well depth, i.e. errors due to human mistakes rather than to uncertainties intrinsic to the measurement methods are now much more common than before. This will be illustrated by simple real examples and by compilations of observations made on large sets of wells, to demonstrate that the examples are symptomatic rather than anecdotal. When these errors become apparent, they are frequently airbrushed and the problem itself is repressed on account of it being “too hard” or perhaps “too late”. Errors tolerated at the start of a workflow or a project, for instance when converting time to depth or when propagating well data into an inversion cube, are then carried throughout all the subsequent work, amplifying uncertainties at every step.

This is not “the best we can do” with respect to well depth: there are in fact simple ways to mitigate this consequential uncertainty, the prize being a lesser uncertainty in most activities that rely on or use depth-indexed well data. The solutions are essentially procedural and they often have no incremental cost. In practice, the results of work using a better well depth tend to fall into place more readily and more consistently, yielding tangible reductions in uncertainty.