Search

Pore-scale images and core-scale imbibition experiments suggest that hydrocarbon and water tend to flow through their own pore network, referred to as "stratified flow". The objective of this paper is to investigate the occurrence of these phenomena at reservoir scale by analyzing flowback and post-flowback production data. Understanding multiphase flow regimes is the first step for selecting proper relative permeability curves and developing representative multiphase rate-transient models. Another objective of this paper is to investigate the harmonic decline (HD) behavior of water-oil ratio (WOR) versus cumulative water production volume and how it could be employed to compare load recovery performance of different wells. We analyzed flowback and post-flowback production data of six multi-fractured horizontal oil wells completed in Eagleford Formation. The proposed data-driven methodology involves using multiphase diagnostic plots of rate-normalized pressure, rate decline, and WOR. We applied this methodology to i) investigate the relationship between water and hydrocarbon at early production time; ii) model WOR with respect to cumulative water production; and iii) evaluate how fracturing/completion design parameters affect well performance. The results show three key findings: i) During early production time, we observe independent flow regimes (stratified flow) of water and oil indicating their production under different drive mechanisms. Water is produced from an apparently closed tank comprising induced fractures and the surrounding stimulated matrix, and oil is produced independently at a significantly lower rate due to oil influx from matrix into fractures. ii) After jet-pump installation, we observe coupled flow of water and oil indicating their production under similar drive mechanisms provided by the pump. iii) Semi-log plot of WOR versus cumulative water production shows HD trend that is relatively less sensitive to operational changes compared to water rate-decline plots.

Summary Flowback data from seven multifractured horizontal tight oil/gas wells in Anadarko Basin show two separate regions during the single-phase water production. Region 1 shows a dropping casing pressure, and Region 2 shows a flattening casing pressure. This paper investigates the flowback behavior of the two regions, and illustrates how flowback data can be interpreted to estimate effective fracture pore volume, and to investigate its relationship to completion-design parameters. We construct diagnostic plots to understand the physics of Regions 1 and 2. Region 1 represents pressure depletion in fractures, and Region 2 represents the hydrocarbon breakthrough into the effective fracture network. The results of our analyses indicate that the duration of Region 1 depends on initial reservoir pressure and hydrocarbon type. We apply a previous flowback model (Abbasi et al. 2012, 2014) on Region 1 to estimate effective fracture pore volume, and also propose a procedure to estimate fracture compressibility by use of diagnostic-fracturing-injection-test (DFIT) data. The results suggest that the estimated effective fracture pore volume is very sensitive to fracture compressibility, and is generally larger than the final load-recovery volume, and less than the total injected-water volume. The results also suggest that most of the effective fractures are unpropped, and host the nonrecovered fracturing water. We investigate the relationship between the estimated effective fracture pore volumes and completion-design parameters, including total injected-water volume, proppant mass, gross perforated interval, and number of clusters, by use of the Pearson correlation-coefficient method. The results show that total injected-water volume, gross perforated interval, and the number of clusters are among the key design parameters for an optimal fracturing treatment. Higher total injected-water volume and closer cluster spacing generally lead to a larger effective fracture pore volume.

The inter-salt argillaceous dolomite reservoirs in the central region of China contain large abundance of oil resources with ultra-low permeability and porosity. However, the oil wells in this area show a very quick reduction with the decline of formation pressure. This article aims to investigate the main possible reasons that affect oil well productivity in the target oilfield. This study begins with analysis of capillary microscopic model, core stress sensitivity experiments, and non-Darcy percolation experiments. The impact of effective stress on permeability and porosity of the reservoir was revealed in this article. The novel productivity model and productivity evaluation model which couples stress sensitivity and threshold pressure gradient were proposed. The analysis of capillary microscopic model shows stress sensitivity of permeability to be much greater than that of porosity during the process of depressurization. The core stress sensitivity experiments results indicate that permeability and effective stress show index relationship while porosity and effective stress show binomial relationship. Damage rate and recovery rate of permeability and porosity were put forward to describe the degree of influence of stress sensitivity on permeability and porosity. The models were used to investigate the factors that affect single well productivity for the target oilfield. Application of the proposed model to this tight oilfield indicates that, the degree of influence of stress sensitivity is much greater than that of threshold pressure gradient. In addition, the greater the stress sensitivity coefficient and threshold pressure gradient are, the greater the productivity reduction will be.

In this study, we propose a new method for estimating average fracture compressibility (cf¯) during flowback process, and apply it on flowback data from thirty multi-fractured horizontal wells completed in Eagle Ford, Horn River, Montney and Woodford formations. We conduct complementary diagnostic flow regime analyses and calculate cf¯ by combining a flowing material balance equation with rate-decline analysis.We observe two production signatures during flowback: (1) single-phase water production followed by hydrocarbon breakthrough and (2) immediate production of hydrocarbon with water. Water rate-normalized-pressure plots show pronounced unit slopes, suggesting pseudo-steady state flow. Water decline curves follow a harmonic trend during multiphase flow; from which we forecasted ultimate water production as an estimate of initial fracture volume. The cf¯ estimates are within the range of 7 – 200 × 10-6 psi-1, and are generally lower than the values previously estimated using Aguilera’s type curves and DFIT data. Also, cf¯ estimates for deeper dry gas wells are relatively higher than those for shallower oil wells.

Hydraulic fracturing combined with horizontal drilling is the key to unlocking vast unconventional reservoirs. However, understanding the relationship between completion-design parameters (CDPs) and fracturing efficiency remains challenging. The objective of this paper is analyzing flowback data to 1) evaluate the change in effective fracture volume (Vef), and 2) investigate the existence of correlations between CDPs and Vef. We analyze flowback data and CDPs of 22 gas and oil wells completed in the Eagle Ford formation. First, we estimate ultimate water recovery and initial effective fracture volume ( V ef i ) using harmonic-decline (HD) model. Second, we estimate the loss in Vef ( V loss ) during flowback using a new iterative approach that accounts for dynamic fracture porosity and compressibility. Third, we introduce a new parameter to characterize fracture closure rate (FCR) during flowback. FCR represents how fast Vef shrinks during flowback. Then, we present a dimensionless type curve to evaluate FCR and investigate its correlation with CDPs. Finally, we conduct a multivariate analysis to correlate CDPs with V ef i and FCR. The developed correlation between V ef i and CDPs shows that the gross perforated interval (GPI) has the most significant effect on V ef i . The proposed correlation between FCR and CDPs demonstrates the significant effect of proppant concentration on fracture closure during flowback.

The proposed dynamic prediction model is based on the occurrence and migration law of coalbed methane and on the pseudo-steady state diffusion model. The model integrated multi-phase flow theory and material balance equation. It is established by considering not only desorption and diffusion but also crossflow. Moreover, the calculation results show the rationality of the method used. The research analyzes the effects of crossflow factor and diffusion coefficient on development effectiveness. The effect of crossflow factor is obvious in the late production stage but not in the early stage. The diffusion coefficient affects gas production during the exploitation stage.

The potential environmental impact of hydraulic fracturing is a significant public concern. A key question is: How far can fractures go beyond the well? We analyze microseismic, pressure interaction (frac-hit and interference), and tracer data recorded during fracturing of and production from 52 wells in the Horn River Basin. The results show that the wells communicate through fracture-like pathways which allow water, gas, and proppant flow. The lateral and vertical well communications occur at distances above 1 km and up to 130 m, respectively. Well communication may last for over 1.7 years possibly due to proppant flow, which prevents complete fracture closure during flowback and post-flowback processes.