Martins, Ana (Nederlandse Aardolie Maatschappij) | Marino, Marco (Nederlandse Aardolie Maatschappij) | Kerem, Murat (Shell Global Solutions International) | Guzman, Manuel (Shell Global Solutions International)
This paper presents the first comparison between two different injection methods for foam assisted gas lift. Useful information for operators and technology developers are also provided concerning chemical selection, testing, and deployment of this hybrid artificial lift technology in the field.
The trials have been conducted in a gas lifted oil well with severe slugging and water cut above 50% (selection criteria as per SPE-184217-MS). The surfactant was delivered through a dedicated capillary injection string during the first trial, and the effects of surfactant concentration and depth of injection were evaluated. During the second trial, the surfactant was injected into the gas lift stream at the surface. Different surfactants were utilised for both trials based on stability concerns and method of injection.
Both trialled injection methods successfully stabilized the well flow, terminating severe slugging while increasing the drawdown and delivering an increase in gross production of circa 200%. These results, together with the downhole pressure data collected during the first trial, confirm that the surfactant starts foaming only at the depth where the lift gas enters the tubing. Injecting surfactant into the lift gas stream required higher concentrations than using a dedicated injection string, difference attributable to the slightly different chemistry, but even at those higher concentrations an anti-foamer injection was not required.
Concerning the response time, the well responded in 30 to 60 minutes with capillary string injection, while 6 to 12 hours were required for injection into the lift gas stream. This suggests that the surfactant probably moves slowly down on the annulus walls as a liquid film rather than travelling in droplets dispersed in the gas phase. Based on the outcome of the two trials, it is concluded that the injection via the lift gas stream is as effective as capillary string injection, at a fraction of the initial costs, with lower maintenance requirements, while still allowing access to the well.
Gao, Guohua (Shell Global Solutions (US)) | Vink, Jeroen C. (Shell Global Solutions International) | Chen, Chaohui (Shell International Exploration and Production) | Araujo, Mariela (Shell Global Solutions (US)) | Ramirez, Benjamin A. (Shell International Exploration and Production) | Jennings, James W. (Shell International Exploration and Production) | El Khamra, Yaakoub (Shell Global Solutions (US)) | Ita, Joel (Shell Global Solutions (US))
Uncertainty quantification of production forecasts is crucially important for business planning of hydrocarbon-field developments. This is still a very challenging task, especially when subsurface uncertainties must be conditioned to production data. Many different approaches have been proposed, each with their strengths and weaknesses. In this work, we develop a robust uncertainty-quantification work flow by seamless integration of a distributed-Gauss-Newton (GN) (DGN) optimization method with a Gaussian mixture model (GMM) and parallelized sampling algorithms. Results are compared with those obtained from other approaches.
Multiple local maximum-a-posteriori (MAP) estimates are determined with the local-search DGN optimization method. A GMM is constructed to approximate the posterior probability-density function (PDF) by reusing simulation results generated during the DGN minimization process. The traditional acceptance/rejection (AR) algorithm is parallelized and applied to improve the quality of GMM samples by rejecting unqualified samples. AR-GMM samples are independent, identically distributed samples that can be directly used for uncertainty quantification of model parameters and production forecasts.
The proposed method is first validated with 1D nonlinear synthetic problems with multiple MAP points. The AR-GMM samples are better than the original GMM samples. The method is then tested with a synthetic history-matching problem using the SPE01 reservoir model (Odeh 1981; Islam and Sepehrnoori 2013) with eight uncertain parameters. The proposed method generates conditional samples that are better than or equivalent to those generated by other methods, such as Markov-chain Monte Carlo (MCMC) and global-search DGN combined with the randomized-maximum-likelihood (RML) approach, but have a much lower computational cost (by a factor of five to 100). Finally, it is applied to a real-field reservoir model with synthetic data, with 235 uncertain parameters. AGMM with 27 Gaussian components is constructed to approximate the actual posterior PDF. There are 105 AR-GMM samples accepted from the 1,000 original GMM samples, and they are used to quantify the uncertainty of production forecasts. The proposed method is further validated by the fact that production forecasts for all AR-GMM samples are quite consistent with the production data observed after the history-matching period.
The newly proposed approach for history matching and uncertainty quantification is quite efficient and robust. The DGN optimization method can efficiently identify multiple local MAP points in parallel. The GMM yields proposal candidates with sufficiently high acceptance ratios for the AR algorithm. Parallelization makes the AR algorithm much more efficient, which further enhances the efficiency of the integrated work flow.
Building realistic and reliable subsurface models requires detailed knowledge of both the rock and fluids involved. While the hydrocarbon volume estimation has a profound impact on the viability of a development, next to the saturation height models and free fluid levels the hydraulic communication and permeability have a significant role as well. Compartmentalization could change the field development plan: e.g. increase the well count, necessitate significant change to the well profiles (e.g. extended range drilling), require complex and expensive completion strategy.
When in different parts of the same field different free fluid levels are identified, leading to different fluid contacts for the same rock quality, the lateral hydraulic communication at the field level can be challenged. This aspect is of importance since the hydrocarbon volume distribution has drastic impact on total hydrocarbon recovery. At the same time building and initializing a model based on different free water level positions across the field, zero capillary pressure, is challenging.
Perched water contacts are the result of water entrapment during the hydrocarbon migration that could lead to variability in free fluid levels across a field. The fundamental controls that lead to the perched contacts formation are studied and shown to be the rock quality and relative permeability. Counter-intuitively, the perching effect is not going to feature in poor quality rocks with sub-milli Darcy permeability – the effects would be visible only for a considerable barrier height, with Free Water Level to barrier height of tensto hundred meters.
In addition, realistic heterogeneous models are studied to investigate the heterogeneity effect on perching and on formation pressures. Whilst low permeability is correlated to a wide range of depths where two phases are mobile, the perching controls in high permeability contrast formations are studied.
Using a dynamic modelling route, potential water entrapment occurrence as a result of high permeability contrast is shown, without structural control, i.e. an underlying impermeably zone defining a trap. The main control in such a case is water permeability just as in structurally controlled perching. This work challenges the industry view that model initialization should be performed with buoyancy as an equilibrium driving mechanism. Such a saturation modelling route would lead to discrepancies when compared to using the capillary pressure as a direct input instead of buoyancy.
Ernens, Dennis (Shell Global Solutions International BV and University of Twente) | van Riet, Egbert J. (Shell Global Solutions International) | de Rooij, Matthias B. (University of Twente) | Pasaribu, Henry R. (Shell Global Solutions International) | van Haaften, Willem M. (Shell Global Solutions International) | Schipper, Dirk J. (University of Twente)
D. Ernens, Shell Global Solutions International BV and University of Twente; E. J. van Riet, Shell Global Solutions International; M. B. de Rooij, University of Twente; H. R. Pasaribu and W. M. van Haaften, Shell Global Solutions International; and D. J. Schipper, University of Twente Summary Phosphate-conversion coatings are widely used on (premium) casing connections for protection against corrosion. These coatings provide galling protection in conjunction with lubricant. The friction and wear that occur during makeup and subsequent load cycling strongly influence the sealing performance of the metal/metal seal. An extensive test program was set up to investigate the role of phosphate coatings during makeup and in the subsequent sealing of the metal/metal seal. With pinon-disk, anvil-on-strip, and ring-on-ring tests, the interactions between the substrate, lubricant, and phosphate coating were investigated. A comparison was made between uncoated and coated specimens using base greases and formulated greases: API-modified lubricant and two commercially available yellow dopes. The results indicate a strong influence of the phosphate coating leading to damage-free makeup, low wear, and less dependence on the lubricant for optimal sealing ability. This is attributed to the formation of a hard and smooth dissimilar surface, the ability to adsorb the lubricant, and the generation of a transfer layer on the uncoated countersurface. It is concluded that taking the interaction with phosphates into account could enable lubricants to be tailored for sealing performance, and thus can ease the transition to environmentally friendly rated lubricants. Introduction Phosphate-conversion coatings (Rausch 1990; Narayanan 2005) were initially applied on (premium) casing connections for protection against corrosion during storage. A side effect of the presence of the phosphate coatings was improved galling resistance (Ertas 1992). Phosphate-conversion coatings therefore play an important role in the proper makeup of casing connections and their subsequent sealing performance. The premium connection (Figure 1), and for this paper its metal/metal seal, should be considered as a (tribo)system (Salomon 1974; Czichos and Winer 1978), which is defined as the combination of lubricant (dope), coating, surface finish, and casing-material grade under sliding conditions. The contact conditions are determined by the pin/box interference and the mechanical properties of the pipe material.
Ugueto, Gustavo (Shell Exploration and Production) | Huckabee, Paul (Shell Exploration and Production) | Wojtaszek, Magdalena (Shell Global Solutions International) | Daredia, Talib (Shell Canada Limited) | Reynolds, Alan (Shell Exploration and Production)
It has been widely demonstrated that frac stimulation efficiency and more importantly production, varies significantly between perforation clusters as well as between sleeve entries. Recent trends indicate that many operators are simultaneously increasing the number of perforation clusters or entries while decreasing frac-to-frac spacing. This is done with the expectation that it will lead to more productive wells overall. The purpose of this paper is to investigate some of the aspects that may limit this approach. There are an increasing number of frac diagnostic tools which allow us to get a better understanding of frac placement and production. Unfortunately, there are only few diagnostic tools available today to characterize the near wellbore region (NWR). Fiber Optics (FO) and other downhole measurements can play an important role in providing information about the NWR. In this paper, we share data and examples from wells where the combination of data from Distributed Acoustics Sensing (DAS), Distributed Temperature Sensing (DTS) and downhole gauges is helping us gain insights about this poorly understood region of our unconventional reservoirs. This paper combines DAS, DTS and downhole pressure gauge data to demonstrate the existence of significant near wellbore complexity, both during stimulation and production. We frequently observe changes in DAS signal and pressure during the stimulation of horizontal wells completed via both "Plug and Perf" (PnP) and Cemented Single Point Entry (CSPE) systems. These changes support the existence of significant near-wellbore tortuosity. Furthermore, we show that pressure data from downhole gauges can differ significantly from surface pressure data extrapolated downhole. This can impact the interpretation of Step-Down-Tests, other analytical techniques relying on the surface pressure alone and affecting the calibration of frac models aimed at understanding the NWR. In wells instrumented with a FO cable behind casing, it is possible to use the DTS data during warmback, following stimulation injection to gain insights about frac geometry in the NWR. Such data provides information about the hydraulic frac dimensions created by the stimulation process in both vertical and horizontal wells. During warmback it is easy to distinguish intervals containing hydraulic fractures near the wellbore where the temperature recovery is lagging compared to the unstimulated portions of the well. FO instrumented horizontal wells allow for estimation of the dimensions of the "Frac-Zone" along the wellbore in the NWR where a combination of hydraulically induced longiditunal and vertical transverse fracs exist. Thermal modeling is also presented for selected stages that further support the qualitative interpretation of the DTS. The diagnostics presented help quantify the dimensions of longitudinal and transverse components in horizontal wellbores in the NWR. This paper also highlights the risk of putting perforation clusters or sleeve entries too close to one another. It is clear that the NWR is poorly understood and more information is needed. Understanding the processes that govern the NWR are essential, after all, this is the region where the well and the reservoir interact.
Barnes, Julian R. (Shell Global Solutions International) | van Batenburg, Diederik W. (Shell Global Solutions International) | Faber, M. J. (Shell Global Solutions International) | van Rijn, Carl H. T. (Shell Global Solutions International) | Geib, Sonja (Shell Global Solutions International) | van Kuijk, Sjoerd R. (Shell Global Solutions International) | Perez Regalado, David (Shell Global Solutions International) | King, Tim E. (Shell Global Solutions US) | Doll, Mike J. (Shell Global Solutions US) | Crom, Lori E. (Shell Global Solutions US)
Alkaline/surfactant/polymer (ASP) flooding is an enhanced-oil-recovery (EOR) technique that involves the injection of a solution of surfactant, alkali, and polymer into an oil reservoir to mobilize and produce the remaining oil. There are several pattern-flood pilots in progress or that will soon be executed to evaluate ASP at a scale relevant to commercial-scale application. The quantities of surfactants needed for these pilots and potential future commercial-scale applications are large (hundreds to thousands of tonnes) and necessitate large-scale manufacture using existing processes and plants for the different manufacturing steps. These operate under slightly different process conditions than those used to make the smaller quantity (50 to 400 kg) of the reference blend used to design the formulation in the laboratory. The upscaling of the surfactant production itself is an essential step to enable field-scale implementation of ASP. To ensure and control the quality of the surfactants produced for pilots with Shell interests, a stage-gated quality assurance/quality control (QA/QC) program was designed and executed. The application of the QA/QC process for a high- and a low-active-matter surfactant-blend concentrate (approximately 60% and 20% active, respectively) is used to illustrate the process.
The early definition of the QA/QC program provided a framework with clearly defined stages for upscaling from laboratory- to large-scale production. The definition of analytical and performance-based laboratory experiments with upfront-defined specifications (minimum and maximum values) and repeatability allowed for clear, unambiguous decisions. Correlations between composition and performance that were developed dependent on pilot-scale production were essential to assure the performance of the larger-scale production. Corefloods, used as the ultimate performance check, showed virtually identical performance for pilot-scale prepared surfactants and surfactants from different large-scale batches.
The paper illustrates that consistent industrial-scale production of surfactants for application in chemical EOR (CEOR) is feasible. To ensure the quality of such surfactant requires a detailed QA/QC program. The successful execution of the QA/QC program for the surfactants for the pattern pilots ensures that the produced large-scale surfactant blend performs as the reference blend used to design the formulation.
Hulea, Iulian N. (Shell Global Solutions International)
The process of differentiating between rock volumes based on petrophysical properties and geological indicators is commonly referred to as rock typing. A rock type can be identified by a given porosity – permeability (k-Phi) transform and Saturation Height Model (SHM) in petrophysical space. Rock typing is a useful method by which geological interpretations are combined with petrophysical measurements and translated into physical space for use in subsurface modelling. Efforts here involve utilizing k-Phi transforms as an input to SHM, thereby streamlining the rock typing process and allowing for compatibility with existing rock typing workflows.
A fundamental part of building realistic subsurface models includes populating a geologic framework with petrophysical properties. From a petrophysical point of view, critical properties with significant impact on the modelling outcome (hydrocarbon volumes recoverable and producibility) are hydrocarbon saturation, permeability and relative permeability. Hydrocarbon saturation is an expression of the rock capillarity translated into a mathematical expression from capillary pressure measurements or well log saturations. Permeability is commonly predicted from porosity, via transform equations used to differentiate reservoir rocks of different quality.
Previous work has shown that permeability and SHM used for subsurface modelling are generally consistent. This implies that the number of input parameters to the SHM can be reduced, which can be done by integrating permeability and saturation data (from logs and core measurements). The number of parameters used in the predictive SHM is reduced from six to four. Here, we propose to constrain the entry pressure (responsible for hydrocarbon entry height) by using routine core analysis data. This approach aiming to look at the plug and log scale has significant benefits when a SHM is derived from well log saturation data or when a limited range in properties is sampled by capillary pressure measurements. The workflow allows the use of other permeability sources (like Drill Stem Test if representative of matrix) as starting point in the process of building a SHM via a simplified Brooks-Corey function. The function can be looked at as a hybrid between the Leverett's J and Brooks-Corey function with entry pressure dependency on the permeability to porosity ratio higher than for Leverett's J (0.7 vs 0.5). The result of linking k-Phi transforms to saturation height modelling allows for compatibility with multiple rock typing approaches that utilize different sorts of parameters to define k-Phi, including Lucia's classic rock fabric numbers, a useful benchmark.
Kirschner, David (Shell International Exploration and Production) | McAllister, Eddie (Shell Upstream Albania) | Davies, Christine (Shell Upstream Albania) | Campman, Xander (Shell Global Solutions International) | Duijndam, Bart (Shell Global Solutions International) | Li, Junlun (formerly with Shell International Exploration and Production) | Marquis, Guy (formerly with Shell International Exploration and Production)
An integrated passive-source seismic and magnetotelluric study was executed in the Dinaride fold-thrust belt of central Albania in 2016. Ambient-noise surface waves originating from the North Atlantic and Mediterranean were recorded on a 380-node network that was deployed over a 500 km2 area for three months. Several hundred earthquakes were also recorded on the network. The data was analyzed by Sisprobe (Grenoble) to generate velocity cubes of the subsurface. Additional seismic data was recorded during a two-week deployment of 280 nodes along a 2D line. Magnetotelluric data was acquired at 105 sites and processed by Schlumberger (Milan) to generate a subsurface resistivity volume for two portions of the AOI. The results of these studies provide a three-dimensional image of the fold-thrust belt that helps explore the area.
Presentation Date: Wednesday, October 17, 2018
Start Time: 1:50:00 PM
Location: Poster Station 14
Presentation Type: Poster
Ernens, Dennis (Shell Global Solutions International BV and University of Twente) | van Riet, Egbert J. (Shell Global Solutions International BV) | de Rooij, Matthias B. (University of Twente) | Pasaribu, Henry R. (Shell Global Solutions International) | van Haaften, Willem M. (Shell Global Solutions International) | Schipper, Dirk J. (University of Twente)
Phosphate-conversion coatings are widely used on (premium) casing connections for protection against corrosion. These coatings provide galling protection in conjunction with lubricant. The friction and wear that occur during makeup and subsequent load cycling strongly influence the sealing performance of the metal/metal seal. Therefore, phosphate-conversion coatings play an important role in the sealing performance of metal/metal seals. An extensive test program was set up to investigate the role of phosphate coatings during makeup and in the subsequent sealing of the metal/metal seal. With pin-on-disk, anvil-on-strip, and ring-on-ring tests, the interactions between the substrate, lubricant, and phosphate coating were investigated. A comparison was made between uncoated and coated specimens using base greases and formulated greases: API-modified lubricant and two commercially available yellow dopes. The results indicate a strong influence of the phosphate coating leading to damage-free makeup, low wear, and less dependence on the lubricant for optimal sealing ability. This is attributed to the formation of a hard and smooth dissimilar surface, the ability to adsorb the lubricant, and the generation of a transfer layer on the uncoated countersurface. It is concluded that taking the interaction with phosphates into account could enable lubricants to be tailored for sealing performance, and thus can ease the transition to environmentally friendly rated lubricants.
Guo, Zhenyu (University of Tulsa) | Chen, Chaohui (Shell International Exploration and Production Incorporated) | Gao, Guohua (Shell Global Solutions US Incorporated) | Vink, Jeroen (Shell Global Solutions International)
Numerical optimization is an integral part of many history-matching (HM) workflows. However, the optimization performance can be affected negatively by the numerical noise existent in the forward models when the gradients are estimated numerically. As an unavoidable part of reservoir simulation, numerical noise refers to the error caused by the incomplete convergence of linear or nonlinear solvers or truncation errors caused by different timestep cuts. More precisely, the allowed solver tolerances and allowed changes of pressure and saturation imply that simulation results no longer smoothly change with changing model parameters. For HM with linear-distributed-Gaussian-Newton (L-DGN), caused by the discontinuity of simulation results, the sensitivity matrix computed by linear interpolation might be less accurate, which might result in slow convergence or, even worse, failure of convergence.
Recently, we have developed an HM workflow by integrating the support-vector regression (SVR) with the distributed-Gaussian-Newton (DGN) method optimization method referred to as SVR-DGN. Unlike L-DGN that computes the sensitivity matrix with a simple linear proxy, SVR-DGN computes the sensitivity matrix by taking the gradient of the SVR proxies. In this paper, we provide theoretical analysis and case studies to show that SVR-DGN can compute a more-accurate sensitivity matrix than L-DGN, and SVR-DGN is insensitive to the negative influence of numerical noise. We also propose a cost-saving training procedure by replacing bad-training points, which correspond to relatively large values of the objective function, with those training-data points (simulation data) that have smaller values of the objective function and are generated at most-recent iterations for training the SVR proxies.
Both the L-DGN approach and the newly proposed SVR-DGN approach are tested first with a 2D toy problem to show the effect of numerical noise on their convergence performance. We find that their performance is comparable when the toy problem is free of numerical noise. As the numerical-noise level increases, the performance of the L-DGN degrades sharply. By contrast, the SVR-DGN performance is quite stable. Then, both methods are tested using a real-field HM example. The convergence performance of the SVR-DGN is quite robust for both the tight and loose numerical settings, whereas the performance of the L-DGN degrades significantly when loose numerical settings are applied.