Integrated Uncertainty Assessment for Project Evaluation and Risk Analysis

Abstract

Summary The main technical contribution of the study presented in this paper was, by an integrated assessment of uncertainties in geophysics, geology and reservoir engineering, to provide a rational basis with risk analysis for the management of uncertainties in the development of the field and therefore better decision making. The objective of the study was to integrate the uncertainties identified on the Lambda Lower & Upper reservoirs and to quantify their impact on Gross Rock Volume (GRV), Oil Originally in Place (OOIP), recoverable reserves and produc-tion profiles. The work was carried out in five main steps:Determination of the distribution of the GRV.Building of a representative cloud of geological full field models (1000 equiprobable models) integrating geophysical, sedimentary and petrophysical uncer-tainties. Determination of the distribution of OOIP.Sorting and selection of a representative subset of reservoir models to quantify dynamic uncertainties.Modelisation by means of experimental design of the impact of dynamic uncertainties on the representative subset of geological models.Integration of static and dynamic uncertainties to assess statistical distributions of recoverable reserves, production profiles and plateau duration using experimental design technique coupled with multi-variable regression and Monte-Carlo simulations. The following results were obtained:probability distributions of GRVprobability distributions of OOIP for each reservoir (Lower & Upper) and for each zone of the Upper reservoirprobability distributions of recoverable reserves and production profilesprobability distribution of production plateau durationprobability estimation of different models, associated with quantiles 10, 50 and 90 of the OOIP and Np distributionsmeasure of the weight of the main uncertainties on the OOIP and reserves. Introduction Throughout the life of a hydrocarbon reservoir, from discovery to abandonment, a great number of decisions (which development with which recovery mechanism? the sizing for surface installations? …) depend on incomplete and uncertain information. Indeed, the only certain information comes from the cuttings and cores extracted from wells. This information represents only a tiny percentage of the rock volume involved and may itself be compromised by the way samples have been extracted. Any other knowledge of the reservoir comes from indirect measurements, either seismic surveys, logs or dynamic information gathered from well tests or production histories. Thus, because it comes from an interpretation process, any parameter that characterises the reservoir is uncertain. Finally, these uncertainties are case dependent (the reservoir and its heterogeneities, the production mechanism involved or even the type of surface installation…) and, for a given field, they depend on its stage of development (initial appraisal, initial development, complementary development). Therefore major uncertainties affect the decisions. Uncertainties in reservoir characterization The process that leads from the definition of structural maps to the estimation of reserves and production profiles for a given recovery mechanism and a given development scheme can be summed up in a few main stages:Definition of the reservoir envelope: maps and faults.Definition of contacts and nature of fluids.

Dynamic data integration for structural modeling: model screening approach using a distance-based model parameterization

“…Lecour et al [2] applied this method to complex fault network modeling, focusing only on the uncertainty resulting from interpretation. Similar applications are also found in Samson et al [3], Corre et al [4], Charles et al [5], and Holden et al [6]. Methods for addressing structural uncertainty during seismic processing (=migration) are also proposed.…”

“…The uncertainty range associated with the former can be evaluated from the thickness of reflectors (after time-to-depth conversion). The uncertainty because of the latter is evaluated by analyzing the match between seismic and well depths [5], depth maps resulting from several conversion techniques [4], or the depth of a flat spot [4,5]. The magnitude of uncertainty varies locally depending on the distance from well markers.…”

“…The lower-order uncertainties (smaller scale uncertainty) are modeled by stochastically perturbing horizons and faults from the interpretation. The stochastic perturbation of horizons is implemented using JACTA™/GOCAD [3][4][5]12]. Faults are stochastically perturbed using the method of Zhang and Caumon [13].…”

“…The structural uncertainty modeling tool in JACTA™ [3][4][5]12] is designed to perturb the depth of a horizon by directly deforming a stratigraphic grid (i.e., corner point geometry grid, which can be used for flow simulation), which is built from a structural model. The magnitude of the perturbation at each grid node on the horizon surface is modeled with a spatially correlated perturbation field, which is stochastically simulated using a geostatistical technique.…”

“…Put differently, this method reduces structural uncertainty resulting from the limited quality of seismic data by assimilating reservoir production data. The prior structural uncertainty space is built through stochastic perturbation of horizons [3][4][5]12] and faults [13] and from multiple structural interpretations. The automatic history matching is implemented using the method of Suzuki and Caers [11].…”