This paper presents a comprehensive model of the fluid-loss process that occurs in hycirauiic fracturing. The modei treats the three fiow regions in the leakoff process-the filter cake, the invaded zone and the reservoir zone-as a composite system with a rigorous account of the interface between the last two zones. Tbk allows automatic consideration of the variation in pressure drops across the three zones and the inclusion of many variable physical aspects in each region. For the description of the filter cake behavior, effects of the internal filter cake, core length, shear rate, and non-Newtonian behavior of fluids have been taken into account. In the invaded zone, the non-Newtonian behavior of the filtrate, the permeability damage caused by the filtrate invasion, and rock and fluid compressibility have been simulated. Different boundary conditions, such as noflow, constant-pressure, finite, and infinite boundaries, have been incorporated into the model, making it applicable for simulating both laborstory tests and the real leakoff process in fracturing treatments under a variety of conditions. The model has been validated against laborstory and published data. Excellent agreement between predicted results and test data has been achieved. The analyses show that the non-Newtonian behavior of the filtrate represented by the model has a significant effect on leakoff. Leakoff would be considerably underestimated if a Newtonian fluid filtrate, as in the classic theory, were assumed instead of the non-Newtonian system. We also show that the vmiation inpressure umps across me rnree zones may rxduse a strong mmunear relation between leakoff and /h that cannot be explained by the classic theory because constant pressure drops are assumed.

This paper describes the study of the effect of asphaltene precipitation and deposition on the development of the Marrat field using a compositional simulation model with asphaltene modeling facilities. The model enables the simulation of asphaltene precipitation, flocculation, and deposition including adsorption, plugging, and entrainment, and the resulting reduction in porosity and permeability and changes in oil viscosity and rock wettability. A workflow was established in the study to i) characterize the equation of state (EOS) by analyzing the fluid PVT and asphaltene data from the lab; ii) calibrate the asphaltene model input parameters using the core flood experimental data; and iii) incorporate the EOS and the asphaltene parameters into the full field simulation model. The model was used to analyze the effects of asphaltene on various development scenarios, including depletion and water injection. For each scenario, the following were calculated and analyzed: field performance including production of oil, gas, and water; asphaltene behavior including precipitation, flocculation, adsorption, plugging and entrainment; formation damage and the effect of rock wettablity changes. The results show that the formation could be severely damaged near the producers where asphaltene is more likely to precipitate because of lower pressures (below the asphaltene onset pressure) and longer time exposure to a larger amount of oil. Formation damage could be reduced by flushing away the deposited asphaltene using higher flow rates if plugging is not significant. The water injection scenario with higher injection rate and higher production rate with BHP limit above 6,000 psia results in less asphaltene adsorption, more entrainment, and therefore less deposition. This in turn causes less permeability damage and more oil production. Introduction The Kuwait Oil Company (KOC) encountered the problem of asphaltene precipitation in Jurassic production wells located in producing areas in West and South East Kuwait. In West Kuwait (WK), the Jurassic production is primarily from the Marrat field. Out of about 45 wells in WK-Marrat 50% have a history of asphaltene cleanouts. These wells contribute to around 7% of the total oil production from WK which can amount to as much as 50 Mbbl/d. Here the reservoir pressure (around 9,500 psi) is considerably above the asphaltene onset pressure (AOP) (estimated between 2,000 psi to 4,000 psi). Therefore, there is no likelihood of asphaltene deposition in the reservoir. However, during production as the pressure of the produced fluid inside the tubing goes below the AOP, asphaltenes start to precipitate from the crude. Asphaltenes gradually deposit in the tubing, reducing its diameter, and in the process cause production rates to drop; eventually the well completely ceases to flow. Once this has occurred, the tubing in the well must be cleaned out to restore the well to production. The Marrat reservoir lies in the South Eastern part of Kuwait, within the giant Burgan field complex. It is a carbonate reservoir, existing at an average depth of 11,000 to 11,500 ft, subsea. The reservoir fluid is generally light, ranging in quality between 36 and 40 API degrees. The original oil in place is estimated to be two billion stock tank barrels. The original reservoir pressure at discovery was determined to be 9,650 psia. A log with the average properties of the reservoir is shown in Figure 1. The fluid is known to be asphaltenic 1 in nature, as evidenced by the deposition of asphaltenes in the production strings and surface production facilities, requiring periodic cleanup and treatment operations. In South and East Kuwait, all the wells in the Marrat reservoir were closed from 1998 to avoid further drop in the reservoir pressure, which had already dropped close to the AOP (the present reservoir pressure is around 8,400 psi and AOP ranges from 5,500 psi to 6,700 psi). Further reduction in reservoir pressure below the AOP could cause asphaltene deposition in the reservoir itself and result in formation damage. The decision was made to produce the wells only after a comprehensive development plan for reservoir pressure maintenance with implemented water injection.

Whether there is a risk of shale oil leakage along the depleted wells or multi-level geological fractures during in situ oil shale mining was predicted using the geological data from the Songliao Basin and survey wells in Fuyu county of Jilin province, China. The simulation results obtained employing the Transport of Unsaturated Groundwater and Heat 2 (TOUGH2) software indicate that oil leakage along depleted wells would entail greater risks to the upper aquifer, while the leakage along multi-level geological fractures would involve higher risks to the lower aquifer close to the shale beds. The distributions of shale oil saturation under the original and 1.3-fold formation pressures are slightly different, and the pollution halo somewhat increased only in an aquifer 40-50 m underground. The shale oil saturation in the leaking channel may be 0 during the leaking process.

This study describes an application of a compositional single well simulator to analyse well tests in gas-condensate reservoirs. An important aspect of this application for gas-condensate well tests is accurate fluid property prediction during the multi-phase flow regime, which occurs in the near-well region. The simulator can also be used to understand the impact of liquid drop-out and fracture flow on well productivity. Hydraulic fracturing improves the economics of wells drilled in tight reservoirs. However, the operation involves a significant amount of expenditure. In recent years this technique has also been used to stimulate gas-condensate reservoirs by creating a flow conduit through the condensate banking near the well. Thus, it is crucial to keep a fracture as small as possible. In practice it has been proved that a short, wide fracture can provide much higher production than the traditionally pursued narrow, long fracture. The workflow in this study contains compositional simulation of a single well in a tight gas-condensate reservoir, which is used to generate transient pressure data for well test analysis and interpretation to predict multi-phase flow behaviour, and to analyse productivity impairment due to condensation. Simulation models were then further modified to study the impact of various hydraulic fractures on the well productivity index (PI), which is defined as the ratio of production rate (constant) divided by the pressure drop across the reservoir. PIs for fractured cases are compared with respect to the non-fractured base case. Streamline simulation of the fractured gas-condensate reservoir was also included in the study to allow visualization of the flow profile in and around the hydraulic fracture.

Production history matching in isolation can result in several nonunique realizations of reservoir parameters; using this practice only modifies the reservoir in the drainage region of the measurement wells. In early production and in fields with a low density of producing-well coverage, production history matching is particularly erroneous. By constraining the production history matching with time-lapse seismic data, a faster convergence to a better solution with reduced uncertainty can be achieved. The repeat acquisition of three-dimensional (3D) seismic data, commonly known as "time-lapse" seismic reservoir monitoring or four-dimensional (4D) data acquisition, is based on the fact that reservoir production and injection significantly change the reservoir's fluid saturation, pressure, and temperature. This subsequently alters the acoustic response of the reservoir rock that can be detected using the seismic method. Combined seismic and production history matching was conducted on a small segment of a reservoir model from a North Sea hydrocarbon field, using software which could simulate changes in the reservoir. The selected segment for modeling consisted of a producing well and an injector well. The existence of several repeated 3D high-resolution seismic datasets provided a unique opportunity to carry out a combined time-lapse and production history matching. A quantitative 4D fluid saturation and pressure mapping technique allowed calibration of 4D seismic responses coupled with flow simulator models. The modeled results, using production history matching alone, were poor, predicting low water cut, where in fact, the water cut was as high as 60%. The combination of seismic data and production history matching resulted in a better prediction and quicker convergence to a solution that more closely matched both the calibrated seismic and the production data. The successful results from the integrated and quantitative approach described in this paper demonstrate a potential for the future integration of seismic and reservoir data to deliver improved reservoir production and reservoir management. Introduction Production history matching utilizing seismic been considered by many geoscientists to be the holy grail of subsurface reservoir modeling. However, the results to date are often marred by errors related to data acquisition repeatability, statistical-derived data processing sequences, and a lack of understanding of rock physics. Other limiting factors include the upscaling and cross-scaling of seismic and simulation data. In this paper we describe, an integrated workflow for the quantitative analysis of time-lapse data and its integration with reservoir simulation analysis. The two, high-resolution, broad-bandwidth seismic surveys with excellent repeatability were used in this study. The surveys were acquired 22 months apart from a Jurassic sandstone hydrocarbon reservoir in the North Sea. A calibrated, single-sensor seismic system, with highly accurate acoustic positioning, coupled with steerable streamers was used for acquisition of both baseline and monitor surveys. A great advantage of this system is that perturbations between surveys could be reduced significantly to provide a clearer 4D image for a more accurate interpretation.[1] Furthermore, with better 4D signal-to-noise ratios, the time period between baseline and monitor surveys may be reduced, and reservoirs with only small 4D effects become viable prospects for the time-lapse technique. Using a unique quantitative time-lapse analysis workflow, the two seismic datasets were inverted for saturation and pressure.[2] The seismically derived saturation and pressure were scaled and mapped to the reservoir grid for a combined 4D- and production history-matching analysis using specially designed simulation software.