Figure 1. The rule and the exception of Q1 in Oman. (Figures courtesy of PDO)

One of the key challenges facing geophysicists is to ensure that the most appropriate geophysical technologies are applied in a timely manner to attain maximum business impact.

That was the thrust of Shell’s 2007 Geophysical Conference in Muscat, Oman, which was organized by Petroleum Development Oman (PDO) and chaired by PDO’s Chief Geophysicist Bob Sambell. Shell is PDO’s primary oil company owner with a 34% stake; the Government of Oman holds a 60% stake.

Worldwide, particularly in compact tertiary deltas, seismic quantitative interpretation (QI) successfully predicts the subsurface or reservoir properties such as fluid fill, reservoir development and lithology variations from seismic data, calibrated by well and geological data. However, many common techniques are not feasible in Oman because of unfavorable rock properties and poor seismic data quality. Because PDO and Shell have developed extended capabilities as well as a comprehensive QI study framework, they were able to overcome the challenges Oman presents. This framework ensures that only feasible and most appropriate QI techniques are applied and thus provide business value.

Feasibility of QI in Oman

The key question is: Can seismic be used to predict the lateral variations (if any) of the relevant subsurface parameters at a scale of interest? In many cases in Oman, it is not possible, as evidenced by the poor well-to-seismic matches.

The poor quality of the land seismic data is one element contributing to the overall QI challenge in Oman; the other is the complex geology in terms of structure, burial history, deposition, diagenisis, compaction processes and stratigraphy (ranging from Cretaceous to Cambrian). To date, amplitude versus offset (AVO) could not be used (modeling predicts only small AVO effects, but the real gathers look very different than the models), direct hydrocarbon indicators (DHI) such as structurally conformable amplitude anomalies are hardly observed (but there are a few), and inversion is rarely used to directly predict a reservoir property such as porosity (Figure 1).

In general, the feasibility of a QI technique is a function of the reservoir impedance profile or acoustic log “character,” the reservoir lateral variability, and the seismic data quality.

Mapping the QI feasibility

Based on this previous analysis, QI feasibility can be assessed through a matrix that combines acoustic reservoir character and seismic data quality (Table 1). The well-to-seismic match also relates to both the seismic data quality and the impedance log “character” and quantifies the feasibility.

Poor “erratic” acoustic reservoir character often results in poor seismic data quality at reservoir level and thus also a poor well-to-seismic match. Good “blocky” acoustic reservoir character has the potential to have good seismic data quality (subject to surface/overburden impact). If the acoustic reservoir character is good but the seismic data quality is poor, seismic corrections through reprocessing or re-acquisition can be considered.

Given the different characteristics, we can map the various stratigraphic units in the QI feasibility matrix. Note that this is a gross generalization, and each case should be evaluated separately. But it should at least provide a rough reference.

Best practice

It is essential to commence a QI study with a feasibility review to ensure that the right QI technique will be applied. The feasibility study is split in two parts. The first part of a QI feasibility study investigates whether the expected change of the variables under study would produce a clear effect on the “ideal” seismic. No seismic, only well data, is required at this stage. Some useful analysis tools are

  1. A correlation panel with impedance profiles and synthetics zoomed in around reservoir interval;
  2. A table of average acoustic impedances and properties of reservoir and bounding lithologies; and
  3. Impedance-Property multiwell cross plots at various scales.

The second part of a QI feasibility study looks at whether the real seismic can predict the reservoir property variations of interest. There is often a large mismatch between the measured and the ideal seismic data, which is quantified by the well-to-seismic match. It is important to ensure that the acoustic well log data is properly quality-controlled and edited before attributing the cause of the mismatch to the seismic. A zero-offset vertical seismic profile, when it matches the synthetic and not the seismic, supports the validity of the synthetic.

At this point, a milestone has been reached. Can the seismic be used to predict the lateral variations of the relevant subsurface parameters at a scale of interest? If so, proceed with the QI project. Otherwise, stop.

Seismic responses

Understanding the seismic response, that is, separating the rock reflections from the artefacts, is of crucial importance. It is the coherent “noise” such as mode conversions, ground-roll, guided waves, scattering and multiples that are often wrongly interpreted as stratigraphic events.

As an example, locally dipping features have often been interpreted as clinoforms. A detailed review of the seismic and acoustic models showed that these interpreted clinoforms are coincidental with seismic multiples and should not be used. This had a huge business impact. It changed the dynamic model and impacted preferred waterflood or steamflood patterns because these clinoforms could have disconnected the flowpath from injectors to producers.

To push the envelope on QI reliability and opportunities, a step-change in seismic data quality is required. There are clearly some fundamental limitations to what can be achieved, but substantial processing improvement was achieved over the last years, and further improvement should be possible. The drive in Shell geophysics for a “next-generation seismic” is very timely. The quest here is for high-density, multicomponent and wide-azimuth seismic. This raises many new opportunities. But it will require considerable input from all disciplines. From the QI side, a key contribution will be the understanding of (near) surface imprint on seismic data, which will suggest improved ways of acquiring and processing the data. Full-wave form modeling is essential in this endeavor.

Right place, right techniques

The challenge is to identify the right opportunity in the right location. The identification of the most suitable technique for a certain area avoids wasting considerable amounts of time and money on trials. These are often the robust semi-quantitative techniques and workflows that can extract information from seismic and that delivers business value in a wide range of assets. To maximize the business impact of QI in Oman, we developed a matrix that describes which, and at what stratigraphic level, QI techniques are feasible and can be meaningfully applied (Table 2). This includes some novel techniques.

Structurally conformable amplitude anomalies are hardly observed in Oman (but there are a few). DHI scanning, as successfully applied in many hydrocarbon basins, will thus not work in Oman.

Volume interpretation, particularly the quick scanning through large multiple seismic volumes to QC seismic data, has proven itself very powerful to highlight stratigraphic trends and structural settings or to identify artefacts. Body-checking is particularly valuable in the intra-salt stringers.

Sparse Spike Inversion is a routine application in Oman, but it is mainly used to aid the structural and stratigraphic interpretation, as its validity for quantitative interpretation is very limited. Only in a few selected cases within the Natih or Shuaiba Formation is a full seismic conditioning of the static model, making use of model-based inversion, with possible and potentially valuable benefits.

Geophysical reservoir monitoring techniques, developed together with Shell Technology, are a key focus activity for the reservoir geophysics team and add considerable value to the PDO business, particularly in enhanced oil recovery. The main techniques are microseismic, surface deformation, cross-well electromagnetic imaging, surface time-lapse seismic and Virtual Source downhole permanent seismic monitoring which, when all properly integrated, provide an areal picture of the reservoir and/or fluid changes.

Conclusion

QI feasibility is mainly related to the reservoir impedance profile, its lateral variability and the seismic data quality. It is shown that there are considerable limitations to QI in Oman. Therefore, before considering QI studies, it is essential to evaluate the feasibility. The right QI technique must be applied in the right location. In combination with unraveling the complex seismic response, significant business impact can be established in difficult seismic areas, as PDO and Shell discovered. Thus we only apply QI techniques that are feasible and polarize the odds of finding hydrocarbons, assist in determining the reservoir quality or add other relevant information.