Haghighi, M (The University of Adelaide) | O'Reilly, DI (Chevron Australia Pty Ltd, The University of Adelaide) | Hunt, AJ (Chevron Australia Pty Ltd) | Sze, ES (Chevron Australia Pty Ltd) | Hopcroft, BS (Chevron Australia Pty Ltd) | Goff, BH (Chevron Australia Pty Ltd)
This paper demonstrates how good technical evaluations and focused operational application can enhance the value of a mature asset. The Windalia reservoir underlies Barrow Island (BWI), situated 56 km from the coast of Western Australia, and has produced oil since 1965. Waterflooding commenced shortly after initial production, in 1967, and remains the main drive mechanism in the field today. Throughout the life of this onshore field, water injection and oil production have varied according to asset strategy and economic conditions. In this case study, we share how recent improvements made in the areas of Reservoir Surveillance and Operations activities have increased water injection efficiency and total oil recovery.
Through the use of new methods and workflows, the BWI Sub-Surface team was able to target specific areas of the field to distribute water to in order to increase injection and maximise oil production. For example, new workflows were built with the real-time PI monitoring system to analyse Pressure Fall Off (PFO) tests from each of the 147 waterflood patterns in detail. Capacitance-Resistance-Modeling was also leveraged to guide individual well target injection-rates. Operationally, several projects were also initiated to increase water injection into the right areas of the field.
The new Reservoir Management approach has significantly increased the volume of water being injected into the areas of need, supporting improved levels of oil production. For the first time in almost 10 years, the stream-day water injection rate has exceeded 90,000 bwipd. The results from PFO transient interpretation and pattern balancing proved effective in directing water to low-pressure, high-GOR areas of the field. They also provided valuable information about formation perm-thickness and skin. The phenomenon of water-cycling was also largely avoided, owing to close monitoring of production well tests and water injector transient surveys.
The present work addresses reservoir and operational aspects of Australia's largest active waterflood. The lessons shared are highly applicable to a low oil price environment, as they show how fit-for-purpose and low-cost acquisition of reservoir data can lead to improved field performance.
In this paper we demonstrate the use of the capillary pressure equilibrium theory (CPET) model to address the effects of partial saturation in order to estimate hydrocarbon saturation in a reservoir volume using acoustic impedances derived by seismic inversion. The data set used here has been donated by BHP Billiton, and is from an offshore oilfield called the Stybarrow field. The set comprises of a well with a 20-foot sandstone oil saturated pay section and 3D pre- and post-stack seismic volumes. Using the provided angle stacks and well log data, a statistical wavelet, and low impedance model, the final impedance model is computed. There are two final impedance models, derived from post-stack, and pre-stack data. The final impedance models are in agreement with one another at each of the well locations, with low impedance at the oil saturated well, and high impedance at the water saturated well. The corresponding CPET model is built based on the empirical porosity from the well log. The rock and fluid properties are available from the logs and petro physical reports provided by BHP Billiton. The CPET model has difficulty distinguishing between 0 and 30% water saturation. The impedances predicted by the CPET model are in good agreement at the two well locations (blind wells), predicting 98% oil saturation in the 97% oil saturated section, and 8% water saturation in the 5% water saturated section of the reservoir. Finally using the CPET workflow, a 3D distribution of saturation was computed from inversion derived acoustic impedance and the CPET model estimated from well log. Unlike conventional approaches of estimating saturation, our method is able to discriminate between patchy and uniform saturation. Our results on Stybarrow field data reveal that the Stybarrow field behaves in a manner very close to the uniform curve at low water saturation. However, starting at 30% water saturation or higher the distribution becomes slightly patchy.
Presentation Date: Wednesday, October 19, 2016
Start Time: 3:35:00 PM
Presentation Type: ORAL
We present a new method for seismic reservoir characterization and reservoir-property modeling on the basis of an integrated analysis of 3D-seismic data and hydraulic flow units, and apply it to an example of a producing reservoir offshore Western Australia. Our method combines hydraulic-unit analysis with a set of techniques for seismic reservoir characterization including rock physics analysis, Bayesian inference, prestack seismic inversion, and geostatistical simulation of reservoir properties. Hydraulic units are geologic layers and zones characterized by similar properties of fluid flow in porous permeable media, and are defined at well locations on the basis of logs, core measurements, and production data. However, the number of wells available for hydraulic- unit analysis is often extremely limited. In comparison, the lateral coverage and resolution of 3D-seismic data are excellent, and can thus be used to extend hydraulic-unit analysis away from well locations into the full 3D reservoir volume. We develop a probabilistic relationship between optimal 3D-seismic-data attributes and the hydraulic units that we determine at well locations. Because porosity and permeability distributions are estimated for each hydraulic flow unit as part of the process, we can use the 3D seismic probabilistic relationships to constrain geostatistical realizations of porosity and permeability in the reservoir, to be consistent with the flow-unit analysis. Reservoir models jointly constrained by both 3D-seismic data and hydraulic flow-unit analysis have the potential to improve the processes of reservoir characterization, fluid-flow performance forecasting, and production data or 4D-seismic history matching.
Forward modelling is a fundamental part of time-lapse seismic feasibility studies and closed-loop seismic reservoir monitoring schemes. The forward modelling workflow represents a complex interaction between multiple domains, each contributing its own set of assumptions, models, and simplifications, based on a priori information. To allow for the best use of the modelled data, the aim ought to be to keep the level of modelling-introduced noise as low as possible to allow for an optimized decision making process on the specifics of a particular time-lapse seismic project. These considerations will have a profound impact on cost and production. It is therefore key to keep 4D signatures subsurface-relevant, avoiding any contributions from low-fidelity forward modelling. To address this challenge, the paper at hand offers a redefined closed-loop seismic reservoir monitoring framework that draws from and caters for a field-scale dynamic integrated Earth model. The validity of the framework and its implications are demonstrated using the purpose built Chimera model which comprises a four-way closure, faulted structure, and represents a turbidite-type reservoir with clastic depositional sequence.
Literature produces many successful examples of timelapse seismic case studies, with carefully analyzed 4D signatures. Calvert (2005) and Johnston (2013) offer detailed insight in this regards, by also simultaneously sketching the state-of-the-art of time-lapse seismic processing. However, quantitative and even qualitative comparisons of differences between predicted and actually measured time-lapse seismic data are much less performed and discussed in open literature, despite the obvious benefit for model reconciliation. An explanation can be found in the assumptions and methodology used for the feasibility study which can be too simplistic to warrant for a meaningful comparison.
When examining 4D signatures, it often turns out that the observed 4D signal is considerably bigger, smaller or different in shape than what was expected from the preceding feasibility study, producing an e term. Given the level of uncertainty within a 4D feasibility study performed, such a mismatch is even expected and the ε term can be further decomposed into three chief elements:
Hill, David (Schlumberger) | Lowden, Dominic (Schlumberger) | Sonika, Sonika (Schlumberger) | Paydayesh, Mehdi (Schlumberger) | Bulat, Andrej (Schlumberger) | Truelove, Leigh (Schlumberger) | Chahine, Clark (Schlumberger) | Herrera, Adrian Rodriguez (Schlumberger)
The history of closed-loop seismic reservoir monitoring (CL-SRM) dates back to early 2000's with the History Matching Using Time Lapse Seismic (HUTS) project; the authors noted at the time that the proposed forward modelling step was too costly (Gosselin 2003). From 2003 the technologies used in the forward modelling step of the CL-SRM workflow have developed in line with the availability of computing resources. Three groups of methods are currently in common use: a) direct analytical transformation of reservoir properties to seismic properties, b) variations on well log based fluid-substitution and petro-physical modelling, and c) simulation-to-seismic modelling (Gjøystdal, 2007). Both a) and b) are typically 1D and neglect realistic noise, overburden and acquisition configuration effect, while c) models noise free seismic data attributes, incorporating overburden and acquisition configuration effects via ray-tracing, but still neglects realistic noise. In all three cases the modelling is largely confined to the reservoir interval.
Summary We present a new method and a field data example for creating reservoir models that simultaneously match seismic and geologic data. The main advantage of our approach is that we can define multiple objective functions for a variety of data types and constraints, and simultaneously minimize the data misfits. The results of our study indicate that improved reservoir models can be obtained using our method, compared to current geostatistical modeling methods. Introduction 3D seismic data, due to its excellent spatial resolution, plays a key role not only in defining the reservoir structure and geometry, but also in constraining the reservoir property variations (Doyen 2007). To produce realistic models of the reservoir facies and corresponding petrophysical properties, and avoid nonphysical results, seismic information should be actively integrated in the reservoir modeling process.
We present a new method for seismic reservoir characterization and reservoir property modeling based on an integrated analysis of 3D seismic data and hydraulic flow units, and apply it to an example of a producing reservoir offshore Western Australia. Our method combines hydraulic unit analysis with a set of techniques for seismic reservoir characterization including: rock physics analysis, Bayesian inference, pre-stack seismic inversion and geostatistical simulation of reservoir properties. Hydraulic units are geologic layers and zones characterized by similar properties of fluid flow in porous permeable media, and are defined at well locations based on logs, core measurements and production data. However, the number of wells available for hydraulic unit analysis is often extremely limited. In comparison, the lateral coverage and resolution of 3D seismic data is excellent, and can thus be used to extend hydraulic unit analysis away from well locations into the full 3D reservoir volume. We develop a probabilistic relationship between certain 3D seismic data attributes and the hydraulic units we determine at well locations. Since porosity and permeability distributions are estimated for each hydraulic flow unit as part of the process, we can use the 3D seismic probabilistic relationships to constraint geostatistical realizations of porosity and permeability in the reservoir, to be consistent with the flow unit analysis. Reservoir models jointly constrained by both 3D seismic data and hydraulic flow unit analysis can therefore help to improve the accuracy of dynamic reservoir flow simulation and production history matching.
Gerd Kleemeyer, SPE, is manager of integrated geophysical services in Shell's Global Solutions Upstream organization in Rijswijk, Netherlands. During his 18 years with Shell, he has worked on exploration and development projects in the Netherlands, Norway, UK, and Russia and as geophysical consultant for global new venture exploration. Kleemeyer holds an MS degree from the Technical University of Clausthal, Germany. The advanced seismic technologies available today require timely planning and close integration to maximize their value. This requirement has been a red thread in all my previous feature reviews, and this is likely to continue.
The Stybarrow Field is a moderately sized biodegraded 22° API oil accumulation reservoired in Early Cretaceous sandstones of the Macedon Formation in the Exmouth Sub-Basin, offshore Western Australia. The reservoir is comprised of excellent quality, poorly consolidated turbidite sandstones up to 20m thick. The field lies in approximately 800m of water and has been developed with five near-horizontal producers and three water injection wells. The Stybarrow development came online at an initial rate of 80,000BOPD in November 2007.
Due to the lack of significant aquifer support, water injection was planned from start-up for pressure maintenance. Acquisition of a variety of data types have enabled key subsurface challenges to be addressed both before and during production. Structural and stratigraphic complexities influence connectivity and therefore must be fully evaluated in order to achieve optimal sweep. A feasibility study concluded that Stybarrow would be a good candidate for 4D seismic monitoring. Two monitor surveys were acquired and, along with other reservoir surveillance techniques, have been used to refine the geological model.
The first monitor survey at Stybarrow was recorded in November 2008. The results of this survey were in agreement with prior 4D modelling and supported the drilling of a successful development well in the north of the field. A second monitor survey was recorded in May 2011, three and a half years after first oil and at 70% of expected ultimate recovery. This survey is currently being analysed to determine if sweep patterns have changed.
The 4D surveys have proven to be an important tool for understanding subsurface architecture and dynamic fluid-flow behaviour. The results of both 4D seismic surveys have provided significant contributions to understanding the dynamic behaviour within the reservoir to facilitate optimal reservoir management.