Efficient Prediction of SAGD Productions Using Static Factor Clustering

Abstract

OH sands have great amount o f reserves in the world with increasing commercial produc tions. Prediction of reservoir performances o f oil sands is challenging mainly due to long simulation time fo r modeling heat and fluids flows in steam assisted gravity drainage (SAGD) operations. Because o f accurate modeling difficulties and limited geophysical data, it requires many simulation cases o f geostatistically generated fields to cover uncer tainty in reservoir modeling. Therefore, it is imperative to develop a new technique to analyze production performances efficiently and economically. This paper presents a new ranking method using a static factor that can he used fo r efficient prediction o f oil sands production. The features vector proposed can reflect shale barrier effects in terms o f shale length and relative distance from the injection well. It preprocesses area that steam chamber bypasses, and then counts steam chamber expanding an area cumulatively. K-means clustering selects a few fields fo r fu ll simulation run and they will cover cumula tive probability distribution function (CDF) o f all the fields examined. Accuracy o f the prediction is high when cluster number is more than 10 based on cases o f cluster number 5, 10, and 15. This technique is applied to fields with 3%, 5%. 10%, and 15% shale fra c tion and all the cases allow efficient and economical predictions o f oil sands productions

Efficient Assessment of Reservoir Uncertainty Using Distance-Based Clustering: A Review

“…The conner point grid system is used to compare the dissimilarity of structure models [12]. In an oil sands reservoir, a feature vector is defined by the shale length and the relative distance from the injection well because these parameters are important to determine steam chamber and future production [13].…”

“…To solve this problem, the K-means algorithm repeats two procedures by fixing either w ij or b j [8]. The Algorithm 1 consists of the following steps [61]: Make k clusters by assigning data points to the closest centroid Recalculate the centroid of each cluster by the mean of the data in the cluster UNTIL the centroids do not change K-means clustering has been extensively applied to group reservoir models in petroleum engineering [13,16,45,48]. Since reservoir models usually have high dimensionality, some researchers have attempted to perform the algorithm on the featured plane using feature extraction methods, such as PCA and singular value decomposition [5,24,25].…”

“…After calculating all the distances, we can decide a winning node, which has a minimum distance value. Then, we only update the weights of a winning node by Equation (13)…”

“…This study, however, assumed that steam could not detour the shale barriers and would extend to a 180 degree radius. To overcome this limitation, the steam is spread within a 90 degree radius and bypassed in the form of a triangle when it encounters shale barrier [13]. These techniques used static distances that do not require dynamic simulation.…”

Flow Characteristics of Steam and Gas Push in the Presence of Heat Thief Zones Overlying Oil Sands Deposits

“…To ensure practical applicability, additional discussions are required, e.g., reservoir heterogeneity, various flow directions in 3D reservoirs, the chamber interferences, and the optimum wellpad system consisting of several wellpairs [4][5][6]. The reservoir heterogeneity, i.e., the variation of the rock property and the facies models according to the location, should be considered since it is able to influence the flowing characteristics of the displacing fluid, as well as the entire performance [24][25][26][27]. An impermeable barrier, e.g., the shale layer that exists in the bitumen deposit, is able to impede the water influx into the steam chamber, while it also interrupts the optimum enlargement of the steam chamber in terms of the energy efficiency.…”