Conventional production performance forecast by reservoir simulation is very time-consuming. Another major drawback of such an approach is the huge uncertainties in the forecast results owing to data limitation, especially at the field appraisal and development planning stages. As an efficient alternative to reservoir simulation, a novel life-cycle performance forecast method is developed which integrates dynamic information from analog reservoirs with decline curve analysis. The output of the forecast provides essential information for field development planning. This includes water breakthrough timing, duration of typical development phases (developing, plateau, decline and mature), number of active producers, oil rate and water-cut. Other important dynamic performance indicators, such as water production rate, WOR, cumulative production, remaining reserves, recovery efficiency, URR and ultimate recovery factor, which are related to economic evaluation and facilities design, are also derivable.The prerequisite for application of this method is possession of sufficient dynamic information from a large number of mature or abandoned fields, which can be selected as the analog fields. The analog fields ought to have similar characteristics to the target forecast field in terms of trap, reservoir, fluid properties and initial in-place hydrocarbon volumes. Key parameters for selecting analog fields include sedimentary environment, sandbody type, range of reservoir permeability, oil API gravity, viscosity, possible STOIIP, and drive mechanism.The new method can be applied to: (1) fast-evaluation of new assets -the forecast result can be used as direct input for economic evaluation; (2) uncertainties estimate in decision-making process using the forecasted low-, medium-and high-case scenarios; (3) verification and uncertainty reduction for conventional reservoir simulation by comparison with the analogbased forecast results; and (4) mature asset benchmarking in order to identify new opportunities to improve recovery. Using tailor-developed software, the process of analog forecast can be completed efficiently within minutes. The forecast is based on real-world facts and has taken into consideration of key factors that influence actual field development. Therefore, the forecast by the new method is, to certain extent, more approximate to the real field production history. TX 75083-3836, U.S.A., fax +1-972-952-9435

This paper presents selected effective technologies and best practices to improve oil recovery from mature fields through waterflooding optimization. These technologies are proved practical, applicable and cost-effective. They can effectively facilitate further development of mature fields, which is more important than ever before in the current economical down-turn environment. First, this paper summarizes and discusses several pragmatic approaches to optimize waterflooding of sandstone reservoirs. These methods have proved to be beneficial for expanding sweep, thus increasing ultimate recovery. Next, the authors introduce two methods to evaluate the effectiveness of application of these technologies, which can support both qualitative evaluation of the successfulness of the applications; and quantitative estimate of incremental recovery. Finally, this paper illustrates the best practices of each waterflooding optimization technology and the associated reservoir dynamic performance. Waterflooding optimization aims at expanding volume sweep to recover bypassed oil in undrained areas or remaining oil in poorly swept areas. The main established approaches include (a) zonal water injection, (b) changing direction of fluid flow, (c) subdividing injection-production unit, (d) water shut-off to improve areal sweeping efficiency, and (e) cyclic water injection. These techniques are particularly applicable to multi-layer, vertically and laterally heterogeneous reservoirs at the high water-cut production stage. One best-practice example shows that converting comingled water injection into zonal injection has successfully arrested production decline, and resulted in an 11.8% incremental recovery. In another reservoir with 98% water cut and 54.5% recovery of changing injection direction by modifying well pattern has led to a 10% incremental recovery by. In still another example, subdivision of injection-production unit combined with infill wells has achieved incremental recovery ranging from 9% to 20%. A further case illustrates that, compared with continuous water injection, cyclic injection had led to reduced water production, lowered water-cut and increased recovery, and resulted in an incremental recovery of 1.84% EUR. To conclude, maximizing reservoir sweep efficiency is the core step of waterflooding optimization. The techniques and best practices discussed in this paper possess both technical and economic viability. Effective application of those methods is an essential part of profitable reservoir management of mature fields.

Since the remaining oil distribution is characterized by scatter and complex in multilayer sandstone reservoirs at high water cut stage, a new method is put forward in this paper to predict interwell remaining oil distribution by using cores, logging data, geologic model and production testing data. Firstly, the interwell sandbody prediction model are set up to predicte interwell sandbody thickness and sedimentary microfacies by using geostatistics and stochastic modelling methods. Secondly, the remaining oil prediction models of various reservoirs are set up by combining the expert system and neural network based on the comprehensive analysis of the relationship between watered-out degree and influential factors in various layers of inspection wells and separated layer testing wells. These factors include sedimentary microfacies, sandstone thickness, structural pattern, communicating states of sandbody with surrounding water injection wells, and injection distances, intensity of water injection, injection-production relationship. The above-mentioned models are used to predict the interwell remaining oil distribution of each layer. This method has been successfully applied in predicting the remaining oil distribution before drilling the tertiary infilling wells in Xing 1–3 area of Daqing oilfield. About 100 layers are studied in this pilot area and maps of remaining oil distribution in each layer are worked out. In addition, the distribution map of remaining oil sandstone thickness in this pilot area is also drawn up and the potential area of remaining oil is pointed out. This method has been used in 28 infilling adjustment areas in Daqing oilfield and good results have been received. By checking 208 layers of 3 newly drilled inspection wells in these development blocks, coincidence rate of single layer predicting remaining oil is at the range of 81%~84%. This paper explores an effective way to study interwell remaining oil distribution in multi-layer sandstone reservoirs. Introduction The underground oil and water distribution becomes more complex with the water cut rises in Daqing oilfield, it is more difficult for the oilfield to recover the remaining oil. The main purpose of the oilfield development and adjustment at the high water cut stage is to recognize the remaining oil and to produce them1, the difficult point is to determine the space distribution of the remaining oil. Thus, the study of the remaining oil distribution is the important subject during high water cut development stage. Since there are many complicated factors effect the remaining oil distribution and not all the layers have the monitoring data, there is a great difficulty for studying the remaining oil distribution in the multilayer sandstone reservoirs at high water cut stage. On the basis of analyzing the geological and the development factors which effecting the remaining oil distribution, the waterflooding data in the inspection wells and the data of the separate layer oil test of the recent years are used, combining with expert system and neural network, an effective method of determining the remaining oil distribution in a single layer by using expert neural network technique is put forward which is applied to the remaining oil potential study for the third infilling adjustment at late period of high water cut stage in Daqing oilfield. Setting up an Interwell sandbody prediction Model Based on the sedimentary microfacies study, using all kinds of geological statistical method predict interwell geological parameters, such as the reservoir thickness. Using facies mode predicting method to predict sandbody boundary in different sedimentary microfacies The facies mode predicting method is guided by every sedimentation model and sedimentology and based on a large quantity logging data to reasonably predict plane combination characteristics of all kinds of sedimentary microfacies in the reservoir.

According to the characteristic of the heterogeneous multi-layer continental sandstone reservoir in Daqing Oil Field, this paper studies a set of stochastic modeling techniques. Multi-step modeling possesses a sound geologic base for predicting various reservoir parameters by microfacies controlling. In order to make a better use of variogram, a single layer is used as modeling unit, resulting in a detailed and reliable model. The depositional microfacies distribution model is improved by man-computer interaction, with detailed consideration of geologists' experiences. Different modeling methods suitable for various types of sandbodies are discussed, and interpretation technique of logging parameters combining petrophysical facies with artificial neural network is developed. The modeling results from unit PII10 in the middle of the northern Block I of Daqing oilfield proves the adaptability of this technique, which provides a new means for identification of the reservoir heterogeneity and uncertainty of understanding. Introduction The Sa-Put-Gao Reservoir in Daqing oilfield is a typical continental multi-layer sandstone reservoir with rapid change of depositional facies in area, and great different properties in various microfacies. Different depositional facies alternate in vertical, result in complex heterogeneity of the reservoir. Over 40 years' development, the oilfield has entered into the later period of high water cut. The heterogeneity of the reservoir is the main cause influencing the recovery. The geologic study of oilfield development is a process of continuous study on the heterogeneity. Reservoir characterization meets with more challenges for improving economic and effective recovery in the later period of oil field development. In recent years, the technique of stochastic modeling is widely used in reservoir characterization, providing a new method for identifying the uncertainty of reservoir characterization with different scales and data resources. According to the depositional characteristics of heterogeneous multi-layer sandstone reservoirs, this paper discusses a method of stochastic predicting modeling based on the logging data of dense wells patterns guided by relative knowledge of sedimentology. Aiming at establishing detailed geologic model of heterogeneous multi-layer sandstone reservoirs, a set of stochastic modeling technique is summarized in this paper, and stochastic models in studied area are constructed. The Depositional Characteristics and Current Data Available of the Heterogeneous Sandstone Reservoirs in Daqing Oilfield Sa-Pu-Gao Reservoir in Daqing Oilfield is a large shallow-water lake basin fluvial-deltic deposit formed in the middle to late period of early Cretaceous. During deposition, the lake basin is very shallow and gentle. Affected by the factors such as tectonic movement, climate change and source supply, alternating layers of sand and shale were formed characterized by a wide distribution deposit, various depositional types, multiple layers, thin single layer, and a large thickness of the whole formation. It has following characteristics. 1. The reservoir distributes vertically with long intervals, multiple layers, and thin single layer. The oil-bearing interval Sae-Put-Gao Reservoir in the north part of Daqing oilfield is 300~500m with more than 100 layers penetrated, and 120~150 single sand layers can be further divided with a thickness of 1~3m. The thinnest layer is only 0.2–0.4m, and the thick distributary channel sandbody is 3~6m. Only a few stacked channels are 10m thick.

Limited data is available for a new discovery at the preparatory stage for the plan of development (POD). Uncertainties exist such as the depositional environment of the reservoir and individual sandbody type, reservoir heterogeneities and fluid properties. This paper presents an innovative and unique method for the qualitative and quantitative evaluation of the influences of these uncertainties on production performance and final recovery factor.The developed methodology enables reservoir engineers and development geologists to conduct uncertainty assessment on production performance and final recovery factors by using a global analogs database and proprietary analytical software tools, including performance forecasting, characterization and attribute cross plotting, etc. The process consists of four steps: (a) selection of the analog reservoirs, (b) performance forecast, (c) uncertainties evaluation, and (d) analysis of key drivers impacting performance and recovery factor. In our case study, Field A is in its POD study stage. Based on available information, reservoir sand in Field A potentially consists of sandbodies of either fluvial channel (FC), distributary channel (DC) or combined fluvial and distributary channels (CFDC). Uncertainty evaluation results indicate that the sandbody of FC, DC and CFDC will produce ultimate recovery factors of 35.9%, 37.9% and 39.7% respectively for the medium-case scenario and 52.4%, 54.3% and 59.2%, respectively for the high-case. The various sandbody types also have a great influence on performances of life-cycle production profile, peak production rate and water-cut variations. Analog analysis of the same sandbody type reveals the impact of three group specific attributes on the ultimate recovery.Compared with the normal approach of conventional geological modeling and reservoir simulation, the analog solution using the proprietary software tool is an efficient process and provides reasonable results.