Time-Lapse Feasibility Analysis for a Brazilian Offshore Field: Target on Saturation and Pressure Changes Interpretation

Abstract

Time-lapse feasibility study is not only the first step to validate a 4D campaign; it is also an essential tool for the interpretation of the acquired seismic volumes. In this paper we present a practical example of feasibility study for a Brazilian offshore reservoir to illustrate the main issues and benefits of feasibility analysis. Our goal is to emphasize which information may be accessed with the feasibility analysis, as well as the uncertainties and nonuniqueness in time-lapse feasibility and interpretation. Introduction Any seismic interpretation is subjected to uncertainties and may be questionable at some extend. Particularly, time-lapse interpretation is a highly non-unique process, and must be guided by rock physics and seismic modeling. The feasibility analysis, usually done to give a technical validation for a time-lapse seismic survey, is crucial for a consistent interpretation of the different vintages of seismic data. Reservoir simulator output may also be integrated with rock physics relations and seismic modeling tools to aid on time-lapse feasibility and interpretation studies. The first step on feasibility analysis relies on available core data, which ideally may include not only measurements of elastic properties on rock samples, but also on fluid samples. The models based on cores may be classified as a “0-D” modeling. Then, usually in parallel, the next step is to construct some synthetic seismic in one and “one and a half” dimensions, that is, both acoustic seismograms as AVO synthetics. Two and three dimensional synthetic seismic modeling are natural evolutions of the feasibility analysis. Although there are some commercial software to integrate reservoir simulator output to seismic rock properties, Petrobras developed proprietary software that enable the translation from reservoir geo-engineering information’s to seismic properties and synthetics. The purpose is not to compete with other software’s, but to have a tool oriented to our needs and our beliefs. For instance, the parameterization of the elastic behavior of a reservoir may be done in a number of different manners; each one will have its assets and drawbacks. With a proprietary and flexible “translator”, we can access the impact of each formulation and assumption on our analysis. Usually, Gassmann relations are used to estimate saturation changes effects. This requires the inputs of fluid and rock elastic properties. The reservoir pressure variation may affect fluid properties as well as rock frame elastic parameters. This pressure behavior of the rock may be accessed though laboratory measurements. Single phase fluid properties can be estimated with the relations published by Batzle and Wang (1992) or with the recent refinements by Han and Batzle, developed on their DHI Project, at Houston University and Colorado School of Mines. The mixing of fluids is a matter of discussion. It can be done through Wood’s equation, for a uniform fluid mixing. In the case of patchy saturation, one may consider that the fluid modulus value will lie somewhere in between the uniform and the completely patchy predictions (Cadoret et al., 1993; Mavko and Mukerji, 1998). Brie (1995) had proposed an empirical formula to estimate the bulk modulus of fluid mixtures with and additional parameter. We had verified that the Brie approximation can violate the Voigt and Reuss bounds under certain conditions, so it may bring physically incorrect estimations. The construction of synthetics under various possible (and even unusual) scenarios can help in a better qualitative or even semi-quantitative time-lapse interpretation. In fact, some authors refer to quantitative time-lapse analysis, but it may be a too ambitious noun for these studies, since there are too much uncertainties involved in each step we make on this field. In this article we describe some of our efforts on the feasibility analysis for seismic monitoring on a Brazilian offshore field. In spite of the sand quality, that is very sensitive to fluid replacement, this reservoir presents some issues to 4D seismic. Particularly, its reduced thickness and the relatively low API of reservoir oil. Our emphasis is on pressure and saturation changes effects on the seismic behavior of the sands. Feasibility Based on Core Measurements One important issue on pressure and saturation inversion of time-lapse data lies on the intrinsic coupling between the effects of these two parameters on the elastic behavior of rocks. For instance, as pore pressure increases, the modulus of the saturating fluid may increase, but the rock frame bulk modulus may decrease. If water is injected on a reservoir, the increase in water saturation may cause an increase in velocity, but if this injection causes a substantial increase in pore pressure, the velocity may even decrease, depending on the amount of saturation and pore pressure changes. The pressure sensitivity of rocks is being discussed by several authors (e.g., MacBeth, 2004; Vasquez et al., 2005), and may be accessed though laboratory measurements.