Executive Summary


This is my 11th and last issue as executive editor of SPE Journal. However, before I leave, I would like to highlight and discuss a few issues related to peer review and digital publication.

Over the past 4 years, we have seen a steady growth in the number of submitted manuscripts, and as a result of this, SPE decided to increase the number of executive editors from two to three. As a result, the journal has had four executive editors for a period to ensure a smooth transition. However, SPE Journal would not have been able to keep up the high scientific standard and technical relevance our readers expect without the effort of our many reviewers, associate editors, and a few tireless members of the SPE staff. Many may find it hard to justify and allocate time to review papers for a scientific journal, and, as a result, we have at times struggled to find qualified technical experts who are willing to conduct high-quality reviews in a timely manner.

To make reviewing more attractive, SPE recently started a collaboration with Publons, which enables you to opt in to receive a much deserved public recognition for your volunteering effort. In an attempt to improve the quality of our peer reviews, all reviewers are now asked to take an online course about the reviewing process and what SPE expects in terms of ethics, scientific conduct, and technical quality. SPE has also revised the list of discipline codes used by our editors when searching for qualified reviewers. As a potential reviewer, it is therefore important that you classify your expertise as accurately as possible and supply good keywords that can be matched with specific papers. Once you have been invited and have accepted a reviewer assignment, you should be as diligent as possible, so that we can bring new results to our readers as fast as possible. Our average turnaround time from submission to first response is currently 60 days, but we are trying hard to reduce it further, e.g., by issuing more initial decline decisions. Authors can accelerate the review process by writing good response letters to eliminate the need to consult the original reviewers upon minor changes.

Another of my concerns as executive editor has been how SPE Journal can benefit from the ongoing transformation in scientific publishing. Like many others, the journal went fully digital a few years ago, with issue information presented in HTML and individual technical articles in PDF format. Our article format is nonetheless still traditional and does not yet allow for various kinds of inline dynamic content. Authors can supply such content as supplementary material, along with computer codes, data sets, and other content that would support the important concept of reproducible science, but very few have so far seized this opportunity. I hope that this will change in the future.

For a short period, SPE offered a five-star rating system that enabled readers to rate individual online papers. This was disbanded last year, after receiving much criticism for being biased, unreliable, and too easy to tamper with. Instead, SPE recently started to relate conference and journal versions of the same paper in OnePetro to add more prominence to peer-reviewed content. When accessing a conference paper, you can scroll down to the bottom of the webpage to see if there is an updated, peer-reviewed version of the same paper, and for peer-reviewed papers you can similarly find the previous proceedings version if the paper originates from a conference; see SPE-191699-MS and SPE-191699-PA for an example. This will hopefully help our readers find the most updated content and contribute to improve the impact factors of SPE’s journals, which are used to compare the relative importance of scientific journals. (Impact factors are based on the average number of citations for the published papers in a journal. The impact factors of SPE’s journals are comparably low, since many citations are ‘lost’ when authors cite the conference version of a peer-reviewed paper.) Because of limited staff resources, this bundling will so far only apply to peer-reviewed papers published in 2018 and onward.

Let me also remind prospective authors of the possibility for open access publishing, announced in the editorial of Vol. 21, Issue 4. On the other hand, let me also warn you that SPE is very strict about novelty, and if the plagiarism check we run on all submissions detects that your manuscript has already been published online outside of SPE, it will not be eligible for peer review. SPE follows a similar rule for manuscripts submitted to conference.

I end my short discussion with an appeal to you as a reader: If you want the journal to be relevant in the coming years, it is very important that you contribute by sending us your top scientific results. Although we have not seen many of them lately, SPE Journal is also interested in publishing high-quality review papers and position papers on emerging topics. Likewise, we always need new volunteers that can review papers or be members of our editorial board.

And now to the science: The August issue of SPE Journal brings 29 papers categorized under five topics.


History Matching and Optimization

Li et al. review assisted history matching using design of experiments by means of a comprehensive case study. The authors summarize and discuss best practices of the seven stages of the history-matching process: problem framing, sensitivity analysis, proxy building, Monte Carlo simulation, history-match filtering, production forecasting, and selection of representative models.

Lorentzen et al. present a robust and flexible workflow that uses a combination of acoustic impedance and production data to history match the full Norne Field. The workflow utilizes an iterative ensemble smoother and is based on a sequential formulation that matches production data before seismic data to avoid that the history-matching performance is dominated by the large amount of seismic data. The authors use techniques from image denoising to estimate the unknown spatially correlated noise level in the seismic data, and introduce additional parameters representing rock clay content, which has a significant effect on the seismic data. Code to generate the initial ensemble can be found on GitHub.

Liao et al. study how surrogate models constructed by an adaptive stochastic collocation method can be used to accelerate the Markov-chain Monte Carlo method in Bayesian interference. The authors construct surrogate models using interpolation over nested sparse grids, taking into account the relative importance of different problems in high-dimensional problems, and to propose using a transformation process to improve the accuracy in cases with strongly nonlinear (discontinuous or nonsmooth) relationships between input parameters and output responses.

Parameter uncertainties in ensemble-based methods are usually assumed to be univariate Gaussian random fields but will in reality follow other random distributions. Aanonensen et al. apply the concept of model likelihood and model probability to incorporate several alternative prior models and develop a method based on importance sampling to estimate the corresponding parameters. The authors apply their new method to condition a dynamic reservoir model to 4D-seismic data by tuning the top reservoir surface, given several alternative prior interpretations with uncertainty.

Yao et al. propose a new method to history match multiscale fractures hierarchically to dynamic production data. The method is based on a hybrid approach in which large-scale fractures are represented with an embedded discrete fracture model, whereas small-scale fractures are modeled with a dual porosity/dual permeability model.

Jahandideh and Jafarpour present a framework for stochastic optimization that enables the user to account for uncertainty associated with future drilling operations when optimizing over current decision variables such as well controls and placement. To this end, the authors use a sequential approach in which well locations and controls are repeatedly optimized until improvements in the objective function fall below a user-specified threshold.


Reservoir Simulation Methods

Xu and Sepehrnoori discuss in detail how an embedded discrete fracture model can be implemented for corner-point grids. The key point to this end is a general-purpose geometrical algorithm for computing the intersection between a general polyhedral cell (3D matrix) and a general polygonal cell (2D fracture).

Convergence failure in Newton solvers have traditionally been cured by chopping the timestep but can also be mitigated by preventing the Newton updates from jumping into a region of different curvature. Although unconditionally convergent, such trust-region methods can be overly restrictive. Klemetsdal et al. show how one can detect oscillations in the Newton updates and use this to adaptively turn trust regions on and off, thereby giving a more effective and less restrictive method.


Heavy Oil

Zargar and Farouq Ali introduce an analytical approach that couples heat transfer with fluid flow to evaluate the effect of confinement and well interference on steam-assisted-gravity-drainage (SAGD) performance. The new analytical model uses moving boundary conditions and enables estimation of oil-production rate, steam-injection rate, thermal efficiency, steam-chamber velocity, unsteady temperature profile, heat distribution, and cumulative steam/oil ratio.

The paper by Irani is part of a longer study of subcool control in SAGD producers; the first and third parts are already published. In the second part, the author formulates the effect of localized hot spots on well controls as a virtual skin factor and demonstrates that longer hot spot will make it more challenging to control the steam breakthrough by choking the well at a given rate.

Shi and Babadagli present an experimental investigation of the oil/solvent mixing process during solvent displacement of heavy oil. The authors categorize four different motion modes of mixing zone and interfaces of miscible fluids in a capillary tube to identify the degree of interface development, and quantify the mixing zone, mass flux, and flow behavior using dimensionless parameters.

The next two papers by Richardson et al. study how the diffusivity of gas into bitumen affects the oil-production rate of in-situ recovery processes of heavy oil. Part I examines pressure-decay experiments with a focus on accounting for swelling and how diffusivity depends on solvent content and develops a numerical method that accounts for swelling and varying diffusivity. The authors find that both constant and viscosity-dependent diffusivity can be used to model the mass of gas diffused and the gas-concentration profile in bitumen. However, the constant diffusivity was different for each experiment and mainly depended on the oil viscosity. Part II develops a correlation for the constant diffusivity to the oil viscosity as a tool to quickly estimate gas diffusivity, as well as a correlation of diffusivity to the mixture viscosity for more-rigorous diffusion models.

Takbiri-Borujeni et al. use a spectrum-based upscaling approach to investigate how heterogeneities in rock and fluid properties affect the equations governing SAGD processes, resulting in a new set of stochastic partial-differential equations. The mean equations include auto- and cross-correlations between different perturbed quantities. The upscaled equations can predict flow rate and the shape of the steam chamber and are compared with Butler’s homogeneous model to quantify the effects of heterogeneity in permeability and thermal diffusivity on SAGD efficiency.


Chemical Enhanced Oil Recovery

Åsen et al. use a setup with serially mounted core segments to study the mechanical degradation of synthetic enhanced-oil-recovery polymers at different flow velocities as a function of distance traveled. In linear flow at high shear rates, the authors observe a decline in degradation rate with distance traveled so that the molecular weight could be matched by a power law. In radial flow, emulated by low-velocity reinjection of a polymer solution previously degraded at a higher rate, degradation occurs close to the sandface with only minor degradation deeper in the formation. The experiments confirm previous findings that degradation depends on salinity and molecular weight and show that the higher the share rate and degradation, the higher the fraction of degradation that occurs in the first core segment.

Pu et al. present a systematic study of how new recrosslinkable preformed particle gels (RPPG) can be used to efficiently control conformance to plug open fractures, fracture-like channels, or conduits that exist in many mature oil fields. The RPPG can swell up to 38 times its initial volume and can recrosslink to form a rubber-like bulky material in large-opening features after placement, which significantly enhances plugging efficiency. The authors systematically evaluate the effect of temperature and swelling ratio on recrosslinking time, gel strength after crosslinking, and gel thermostability.

Zhu et al. study a terpolymer-gel system with low-toxic polyethylenimine (PEI) as crosslinker. The authors show that suitable gelation time, gel strength, and thermal stability can be obtained by selecting PEI molecular weight and adjusting terpolymer concentrations. They also show that PEI concentration has less obvious effect than polymer concentration on gelation performance and that very low concentrations of sodium chloride can slightly shorten gelation time. 

Zirconium-based crosslinkers are good choices in high-pressure, high-temperature wells, but suffer from rapid viscosity buildup at high pH and are shear-sensitive. Sokhanvrian et al. investigate the effect of three different ligands (lactate, propylene glycol, and triethanolamine) on the viscosity of crosslinked polymer CMHPG with three different Zr-crosslinkers. Such ligands are attached to the zirconium to delay crosslinking until the fluids pass high-shear environment. The authors also demonstrate that a new aluminum-zirconium dual crosslinker potentially can mitigate fast crosslinking and viscosity buildup at high pH.

Jia et al. explore a simple and safe in-situ generation procedure for foam under surface condition. Foam stability in the wellbore is ensured through double crosslinking with the application of chromium acetate III (Cr3+) and polyethyleneimine (PEI). The authors optimize different foaming agents, gel bases, and the effect of the composition and quantity of the CO2-gas-producing chemicals. They also conduct a macroscopic comparison of the stability and rheological properties of the double crosslinking and the common single crosslinking systems, investigate their stability differences through microscopic research, and perform coreflooding experiments to evaluate working performance.


Fracturing and Fractured Systems

Wu and Sharma study proppant transport through a perforate horizontal casing by use of a combination of computational fluid dynamics and discrete element methods. Their simulations show that proppant intertia strongly influences the proppant transport into a perforation, which results in an increase in the proppant concentration in the slurry as it flows from the heel to the toe side of a plug-and-perforate stage.

Kim et al. investigate stress-dependent permeability of split shale core plugs in the presence of microproppants and develop an analytical permeability model. The model is then used to analyze a series of pressure-pulse-decay experiments of propped shale samples. Amongst others, this provides a quality parameter for the areal distribution of proppant in the fracture, which can be used as a measure of success of the delivery of proppants into microfractures and to design stimulation experiments in the laboratory.

To reduce uncertainties in interpreting diagnostic fracture-injection tests (DFITs), Wang and Sharma derive time-convolution solutions to pressure-transient behavior of a closing fracture with infinite and finite fracture conductivity, which include G-function and square-root-of-time models as special cases. The authors argue that whereas after-closure flow analysis can be used to infer pore pressure reliability, it fails to estimate other parameters correctly. They then propose a new approach to history match the entire duration of DFIT data to estimate formation-flow capacity, without knowing closure stress and the roughness properties of the fracture surface. The new approach is particularly useful in unconventional reservoirs where after-closure radial flow is absent.

Cherivirala et al. discuss the use of complementary metal oxide semiconductors (CMOS) -based energy-harvesting chips as wireless nodes for mapping hydraulic fractures during the shale-gas extraction process. On the basis of tests run in a custom benchtop core-holder chamber that emulates a downhole environment, it is shown that encapsulated CMOS chips are able to harvest electromagnetic energy and thereby operate wirelessly, producing an output signal with a frequency proportional to the harvested power, which is enough to localize chips within the rock inside the core holder.

Hou et al. report the results of true-triaxial hydraulic-fracturing tests conducted on shale outcrop specimens to study the initiation and propagation of hydraulic fractures, while subjected to an alternating fluid injection with guar fluid and slickwater. The research is motivated by the need to develop improved fracturing techniques for deep shale-gas reservoirs.

Sheng et al. develop a novel model for describing induced fractures in which both the spacing and the permeability/porosity of the induced fractures are distributed according to fractal dimensions. A semianalytical Bessel spline scheme is used to solve a corresponding fractal diffusivity equation. The authors compare the solutions of their new model with solutions of former models and use the results to predict production data of a multifractured horizontal gas well in the Barnett Shale.

Guo et al. perform a numerical study to investigate how subsequent injection in parent wells with legacy production affects interwell fracturing interference. The authors solve for the spatiotemporal evolution of pore pressure and in-situ stress to obtain the interwell fracture network consisting of fractures from both the parent and the infill-well completions and capture fracture hits. Simulation results indicate, amongst others, that subsequent injection in parent wells make infill-well fractures grow more transversely, improving completion qualities of infill wells. The authors also present results regarding occurrence of fracture hits and required subsequent injection volumes but emphasize that these are case specific, and that the main contribution of the work lies in the underlying analysis.

Jafarizadeh and Bratvold develop a flexible Markov-decision process for hydraulic refracturing of unconventional oil wells. The process assumes that the optimal refracture or abandonment time depends on the stochastic mean-reverting behavior of prices and the then-current level of production. The paper is accompanied by an Excel Visual Basic implementation of the proposed algorithms.

Fan et al. present a hybrid, experiment/simulation workflow that combines the discrete element method and the lattice Boltzman method with laboratory-measured load-embedment correlation to investigate fracture-conductivity evolution and non-Darcy flow in a proppant-supported hydraulic fracture. The purpose of the research is to provide fundamental insight into basic mechanisms that can be used to optimize proppant-pack conductivity.

Almubarak et al. report a series of experiments conducted to design and tailor raw-seawater-based fracturing fluids so that no scale forms when additives are mixed in or when the fracturing-fluid filtrate is mixed with different formation brines. Promising results are presented for an unconventional and a conventional carbonate well.


Acknowledgments

I would like to thank three persons who recently retired from our editorial board for the service they have provided to SPE and the readers of the journal: Yongqian Fan has been an associate editor since 2012; Arild Saasen has been an associate editor since 2016; and Lincoln Paterson has served 14 years as an editor, including 3 years as executive editor between 2004 and 2006.

As usual, I end my executive summary by thanking all those who have written, reviewed, copy edited, and prepared the papers in this issue. 

Knut-Andreas Lie, SPE J. Executive Editor,
SINTEF Digital / NTNU