Simple Tools for Forecasting Waterflood Performance
- Bunyamin Can (Texas A&M University) | C. Shah Kabir (Hess Corp.)
- Document ID
- Society of Petroleum Engineers
- SPE Annual Technical Conference and Exhibition, 8-10 October, San Antonio, Texas, USA
- Publication Date
- Document Type
- Conference Paper
- 2012. Society of Petroleum Engineers
- 4.1.2 Separation and Treating, 2 Well Completion, 5.8.2 Shale Gas, 1.2.3 Rock properties, 3.2.3 Hydraulic Fracturing Design, Implementation and Optimisation, 5.5.8 History Matching, 5.4.1 Waterflooding, 5.6.9 Production Forecasting
- 13 in the last 30 days
- 1,041 since 2007
- Show more detail
- View rights & permissions
|SPE Member Price:||USD 8.50|
|SPE Non-Member Price:||USD 25.00|
Empirical and/or semianalytical tools are frequently applied in most waterflood operations, although grid-based models are also often used. This paper examines the performance of some commonly used tools, such as the water-oil ratio (WOR), Y-function, and Arps. Besides those tools, we introduce a semianalytical approach, which is a modified version of the Y-function formulation. Two other tools that have gained significant traction in unconventional-reservoir performance forecasting, the stretched-exponential decline model (SEDM) and the capacitance-resistance model (CRM), are also used here.
Based on synthetic and field data, the results show that the Arps method is remarkably accurate in all flooding situations, regardless of the underlying physical mechanisms; other published data tend to support this notion. Similarly, both the SEDM and the proposed modified-Y-function method also yield solutions with good accuracy. The latter solutions tend to be pessimistic, however.
Semianalytical and empirical methods for forecasting waterflood performance have been used routinely over many decades, despite the advent of grid-based flow simulations. This popularity stems probably from the speed of generating solutions, particularly those in large and mature floods, where grid-based flow simulations may be tedious. Besides, reserves reporting often necessitate resorting to some type of decline-curve analysis.
Many simple methods have emerged for projecting reservoir performance undergoing waterfloods. Among these, Arps (1945), logarithm of WOR versus cumulative-oil production (Np) plot, and the X-plot are worthy of note. The notion of hyperbolic decline in waterfloods is anchored firmly on many field studies. Studies of Bush and Helander (1968) involving 86 successful waterfloods in Oklahoma, and Ambastha and Wong (1998) describing 78 waterfloods in Western Canada, are cases in point.
Ershaghi and Omoregie (1978) and Ershaghi and Abdassah (1984) proposed the so-called X-plot formulation. This method is premised upon the linear relationship between the logarithm of relative-permeability ratio and water saturation for the intermediate saturation range, and the Buckley-Leverett (1942) equation. Bondar and Blasingame (2002) investigated the suitability of a number of empirical and semianalytical methods. They concluded that the X-plot technique did not offer any advantage over others. In fact, overestimation of oil recovery appeared to be the norm. They also found that plotting reciprocal of fractional-water flow (1/fw) versus Np and reciprocal of oil rate (1/qo) versus Np over instantaneous oil rate (Np/qo) provide results similar to those obtained from the conventional log (WOR) versus Np plot.
Yang (2009) proposed a production-decline analysis tool for waterflood performance forecasting. This method is quite appealing because it is anchored in analytical solution of fractional flow and Buckley-Leverett's 1D displacement theory, and the semilog linear relationship between the oil/water relative-permeability ratio and water saturation.
This study expands upon the scope of Yang's model by improving its performance at high WOR in mature waterfloods. Besides the Arps (1945), Yang, and modified-Yang methods, we used the WOR, CRM, and SEDM (Valko 2009, Can and Kabir 2012) to predict the future waterflood performance. In this study, CRM is used in the same context as we (Kabir and Lake 2011) recently did in forecasting primary performance in unconventional reservoirs, thereby skirting the two-phase flow issue. Note that in a conventional treatment (Sayarpour et al. 2010), an empirical fractional-flow model is used, which applies in mature floods exceeding 50% watercut. Note that the SEDM has been used in the context of primary production from unconventional reservoirs. Using both synthetic and field data, we observed that the WOR method provided the most optimistic solutions, whereas the other methods tend to yield solutions that are well within engineering accuracy.
|File Size||460 KB||Number of Pages||11|