Direct Modeling of Reservoirs Through Forward Process-Based Models: Can We Get There?

Miller, James; Sun, Tao; Li, Hongmei; Stewart, J.N.O.; Genty, Coralie; Li, Dachang; Lyttle, Colin

doi:10.2523/iptc-12729-ms

Cited by 9 publications

(6 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…To date such coastal morphodynamic models are generally able to adequately simulate morphological change due to concurrent tides, waves and currents for short to medium-term time scales; events of 3-4 years [10][11][12][13]. Prediction of the long-term dynamics of sandy interventions have however remained a major challenge for process-based models [14].…”

Section: Introductionmentioning

confidence: 99%

Morphodynamic Acceleration Techniques for Multi-Timescale Predictions of Complex Sandy Interventions

Luijendijk

Schipper

Ranasinghe

2019

JMSE

View full text Add to dashboard Cite

Thirty one percent (31%) of the world’s coastline consists of sandy beaches and dunes that form a natural defense protecting the hinterland from flooding. A common measure to mitigate erosion along sandy beaches is the implementation of sand nourishments. The design and acceptance of such a mitigating measure require information on the expected evolution at time scales from storms to decades. Process-based morphodynamic models are increasingly applied, together with morphodynamic acceleration techniques, to obtain detailed information on this wide scale of ranges. This study shows that techniques for the acceleration of the morphological evolution can have a significant impact on the simulated evolution and dispersion of sandy interventions. A calibrated Delft3D model of the Sand Engine mega-nourishment is applied to compare different acceleration techniques, focusing on accuracy and computational times. Results show that acceleration techniques using representative (schematized) wave conditions are not capable of accurately reproducing the morphological response in the first two years. The best reproduction of the morphological behavior of the first five years is obtained by the brute force simulations. Applying input filtering and a compression factor provides similar accuracy yet with a factor five gain in computational cost. An attractive method for the medium to long time scales, which further reduces computational costs, is a method that uses representative wave conditions based on gross longshore transports, while showing similar results as the benchmark simulation. Erosional behavior is captured well in all considered techniques with variations in volumes of about 1 million m 3 after three decades. The spatio-temporal variability of the predicted alongshore and cross-shore distribution of the morphological evolution however have a strong dependency on the selected acceleration technique. A new technique, called ’brute force merged’, which incorporates the full variability of the wave climate, provides the optimal combination of phenomenological accuracy and computational efficiency (a factor of 20 faster than the benchmark brute force technique) at both the short and medium to long time scales. This approach, which combines realistic time series and the mormerge technique, provides an attractive and flexible method to efficiently predict the evolution of complex sandy interventions at time scales from hours to decades.

show abstract

Section: Introductionmentioning

confidence: 99%

Morphodynamic Acceleration Techniques for Multi-Timescale Predictions of Complex Sandy Interventions

Luijendijk

Schipper

Ranasinghe

2019

JMSE

View full text Add to dashboard Cite

show abstract

“…Process-based methods (Sylvester et al 2011;Pyrcz et al 2005;Zhang et al 2009;Karssenberg et al 2010;Miller et al 2008) are dynamic algorithms that attempt to mimic the processes of sediment transport and deposition that originally created the now buried stratigraphy (Koltermann and Gorelick 1996). These are most commonly process-imitating rule-based approaches, since full fluid-dynamical computations are currently too slow to be used iteratively to fit data.…”

Section: Introductionmentioning

confidence: 98%

A New Approach for Conditioning Process-Based Geologic Models to Well Data

et al. 2015

View full text Add to dashboard Cite

Generating a realistic earth model that simultaneously fits data observed at multiple well locations has been a long-standing problem in petroleum geology. Two insights are offered for solving this problem in a Bayesian framework. The first is conceptual-it connects geologic inversion to the new field of probabilistic programming and shows that the usual description of a Bayesian problem in terms of a graphical model is inadequate for describing a process-based geologic model due to the dynamics of the generative algorithm. This is a paradigm shift in probabilistic modeling where stochastic generative models are represented using a syntax resembling modern programming languages. Probabilistic programming allows one to generalize this structure to include complex programming concepts, while also simplifying the process of developing new inference algorithms. The second insight is algorithmic and involves using variational inference to derive a simpler, more computationally tractable approximation to the posterior probability density function. If this surrogate distribution is close to the true posterior, it allows for very fast simulation of an arbitrary number of models that all fit the data equally well. This study focuses on the particular geologic formation known as submarine lobes: elongated pancake-like formations which are sequentially laid down, one on top of the other over geologic time, forming potential petroleum reservoirs. The location and orientation of the lobes conclusion is that, even though the variational approximation is crude, it produces better predictions than any point-based method, including maximum likelihood. The fact that probabilistic programming outperforms conventional Bayesian approaches in the case of lobe models offers the potential for attacking more complicated forward models where multiple geologic processes are simultaneously active.

show abstract

“…Forward stratigraphic modeling can be applied to predict the location and heterogeneity of depositional systems and petroleum reservoirs (Miller et al, 2008), as well as the depositional response to controlling factors (Piper and Normark, 2001). For example, in an exploration scenario with low-resolution seismic-reflection data (generally two-dimensional, 2-D, profiles with frequencies of 5-20 Hz; Normark et al, 1993;Prather et al, 2012) and no lithologic control from well penetrations, seismic-stratigraphic interpretation and structural restoration can be applied to create a paleotopographic surface for modeling the location, size, shape, and sub-seismic heterogeneity and stacking of deposits (Deville et al, 2015;Hawie et al, 2015;Barabasch et al, 2018).…”

Section: Introductionmentioning

confidence: 99%

Grain-Size and Discharge Controls on Submarine-Fan Depositional Patterns From Forward Stratigraphic Models

Hawie¹,

Covault²,

Sylvester³

2019

Preprint

View full text Add to dashboard Cite

Submarine fans are important components of continental margins; they contain a stratigraphic record of environmental changes and host large accumulations of oil and gas. The grain size and volume of sediment supply to fans is thought to control the heterogeneity of deep-water deposits; predicting spatial variability of sandy and muddy deposits is an important applied challenge in the characterization of fans. Here, we use DionisosFlow stratigraphic-forward models to evaluate the sensitivity of submarine-fan deposition to a range of grain sizes, with corresponding diffusion coefficients ranging from 10 to 100 km2/kyr for coarse sand to silt/clay, and discharges. In general, finer grains are transported farther in our models because they have larger diffusion coefficients. Coarser grains typical of a sand-rich fan tend to pile up and compensationally stack at the mouth of a proximal feeder channel. Finer grains tend to be distributed across the model domain; however, finer load resulted in fewer channel avulsions because finer sediment did not build topography as high as coarser sediment. Increasing sediment-gravity-flow discharge resulted in a thicker depositional system; however, relatively coarse sediment piled up at the mouth of the feeder channel, which created a slope that promoted basinward sediment transport. Our modeling results can be applied to predict the overall geometry, stacking, and grain-size distribution of submarine fans. Improved understanding of grain-size and discharge controls also informs interpretation of the stratigraphic record of submarine fans. For example, outcrop observations of heterogeneity and compensational stacking of depocenters can be quantitatively related to changing boundary conditions, namely changes in the caliber and overall supply of sediment delivery to deep-water basin margins.

show abstract

Direct Modeling of Reservoirs Through Forward Process-Based Models: Can We Get There?

Cited by 9 publications

References 0 publications

Morphodynamic Acceleration Techniques for Multi-Timescale Predictions of Complex Sandy Interventions

Morphodynamic Acceleration Techniques for Multi-Timescale Predictions of Complex Sandy Interventions

A New Approach for Conditioning Process-Based Geologic Models to Well Data

Grain-Size and Discharge Controls on Submarine-Fan Depositional Patterns From Forward Stratigraphic Models

Contact Info

Product

Resources

About