Ayan

A few charts and plots might be of interest to our readers pertinent for the formation evaluation side of our journal. Though we mention formation evaluation side, our team also receives substantial reservoir engineering manuscripts as well. Thanks go to SPE staff, who got the basic numbers generated the plots. Fig. 1 shows the subscription numbers to SPE Res Eval & Eng. There is a substantial increase starting from 2005 until 2009, with online subscription increasing steadily. The drop in 2009 reflects relatively tougher times.

Table 1 shows the number of papers published in print and online (started in 2009) and the percentage of papers published within 18 months.

The decreasing percentage figures from 2005 to 2008 is not necessarily caused by the peer-review process, but restricted by the ability to print a "normal" size journal that can be "reasonably carried around." As Fig. 2 shows, the review process accelerated quite bit for the formation evaluation side since the end of 2007. The time to initial decision is now around 16 weeks, despite the significant increase in the incoming manuscripts during 2007 and 2008. The recent uptrend is the transition effect to our new electronic peer-review system and hopefully it will be temporary.

For those who are wondering about the spread of decisions, Fig. 3 gives an outline for the original manuscripts only (not considering revised manuscripts). A reminder: the processes of submitting a paper to a meeting and for peer review are now totally separate, please check SPE.org for the details.

This issue of SPE Res Eval & Eng starts with a paper that introduces a new method to obtain dead-oil viscosity. In the paper, "A Consistent and Accurate Dead-Oil-Viscosity Method," the authors studied 23 existing dead-oil viscosity calculation methods. Using a large PVT database, they propose an improved method that can reduce the average absolute error in the predicted dead-oil viscosity by two- to 13-fold at reservoir and three- to 60-fold at surface conditions. Water injection is the most common method to augment the primary recovery mechanism(s). The paper "Short-Term and Long-Term Aspects of a Water Injection Strategy," reviews the performance of a peripheral waterflood in one giant Middle East carbonate reservoir. The presence of three reservoirs juxtaposed through tree major strike/slip fault zones made the waterflood management more complex and provided surprises, such as water slumping, crossflow, and pressure communication across the zones. Through numerical modeling, the authors show that by maintaining the pressure differentials across the reservoirs, cumulative crossflow can be minimized. In the paper, "Field Applications of Capacitance-Resistive Models in Waterfloods," the authors introduce a spreadsheet based tool that can help engineers manage waterfloods on a daily basis. The method does not require a geological model, neither does it require saturation nor pressure matching. It gives users quick insight on the connectivity of producers and injectors using a simple signal based analytical model and material balance concepts. The method can provide a good understanding of the flood and a base framework before undertaking a more comprehensive numerical modeling approach. In our third paper on waterflooding, "Water Injection Optimization for a Complex Fluvial Heavy-Oil Reservoir by Integrating Geological, Seismic, and Production Data," the authors describe waterflood optimization in a 200-cp heavy oil reservoir at Bohai Bay, China. By timely implementation and update of the flood parameters in a poorly connected system, the pressure decline rate in the field was reduced in two platforms from 26 and 47% to 19 and 14%, respectively. In the paper, "A Simple Analytical Model for Predicting Productivity of Multifractured Horizontal Wells," the authors study the productivity of horizontal wells with multiple transverse fractures with an analytical model, including pseudosteady-state flow for a reservoir of any shape. Their productivity model considers reservoir radial flow, formation and fracture linear flows, and fracture radial flow closer to the horizontal wellbore. Cleanup of wells following mud-filtrate invasion has always raised interest for well test, reservoir, and production engineers. Water blocking, damage variation along the wellbore are phenomena of concern and the problems get worse if the well crosses fractures and high permeability layers. The paper, "Simulation of Dynamic Filtrate Loss During the Drilling of a Horizontal Well With High-Permeability Contrasts and Its Impact on Well Performance" outlines an improved method to model dynamic mud filtrate invasion while drilling a horizontal well. Coupled with a multiphase numerical model, the authors study the extent of invasive damage and its impact on flowback during production. Mud invasion can influence the near wellbore flow and petrophysical properties. In a related paper involving filtrate invasion titled, "Estimation of Dry-Rock Elastic Moduli Based on the Simulation of Mud-Filtrate Invasion Effects on Borehole Acoustic Logs," the authors estimate 2D fluid distributions around the wellbore by honoring resistivity logs. This is followed by computing compressional and shear-wave velocity distributions around the wellbore, which are matched with sonic-log waveform data by tuning dry-rock elastic moduli. In the paper, "Mechanisms of Enhanced Natural Imbibition With Novel Chemicals," the authors investigate natural imbibition mechanisms for intermediate- to oil-wet carbonate rocks. They outline an X-ray-tomography based experimental setup to evaluate capillary and emulsification driven imbibition mechanisms and suggest alternative chemical agents for enhanced oil recovery. A multilayer feed-forward neural network was used to predict the performance of a multi-isotope logging data in the paper, "Implicit Approximation of Neural Network and Applications." In their approach, the authors combined the principle of implicit curves and surfaces with the neural network to describe time vectors series. In the paper "Use of the Beta Distribution To Determine Well-Log Shale Parameters," the authors estimate well-log response using a probabilistic approach, when the number of parameters are greater than the number of measurements. The beta distribution is used to describe the distribution of mineralogical components at a given depth level; clean sand and clean shale Gamma Ray responses for real cases were estimated. Storage of gas in a depleted carbonate reservoir while increasing oil recovery by an additional 5% is described in the paper, "Combined Underground Gas Storage and Increased Oil Recovery in a Fractured Reservoir." The injection of gas leads to gravity drainage of oil from the matrix in the fractured reservoir, which is captured by horizontal wells. Approximately one-third of the injected gas was used as a cushion in the first cycle. The Midale field is a well known fractured carbonate reservoir undergoing full-scale carbon dioxide injection with a wealth of data. In the paper, "Integrated Modeling of the Fractured Carbonate Midale Field and Sensitivity Analysis Through Experimental Design," the authors used static data obtained from cores/logs and well-test data to construct 3D discrete fracture networks. They used response surface modeling to investigate which parameters have the greatest impact on well-test response and found that matrix quality had the strongest influence. Streamline simulation with API tracking is discussed in the paper, "Compressible Streamline-Based Simulation With Changes in Oil Composition." The authors considered compressible flow and used the streamtube volume as the distance coordinate for the transport solution that reduced mass balance errors. In a rather unusual topic, the authors investigate the migration of radionuclide to production wells following nuclear stimulation of a gas reservoir in the paper, "A Geologically Based Markov Chain Model for Simulating Tritium Transport With Uncertain Conditions in a Nuclear-Stimulated Natural Gas Reservoir." Nuclear stimulation technology was tried in the 1960s and 70s at three subsurface test sites in the US. The lack of subsurface data was mitigated by the use of geologically based Markov chain model for the simulation of tritium transport in the reservoir, which is produced in abundance after the nuclear detonation and can be transported both in gas and aqueous phases.

Cosan Ayan, Schlumberger

Fig. 1

Fig. 2

Fig. 3

Table 1