Inflow Control Devices (ICDs) have been adopted for commercial steam-assisted gravity drainage (SAGD) production for nearly ten years and yet the function they serve is not well understood, and field data evaluating their performance remains scant. Thus, the purpose of the current study is twofold: Firstly, the study derives a simplified analytical model demonstrating how increasing the dP across ICDs acts to improve conformance along a producing lateral. The resulting equation of the analysis acts as a simple rule of thumb for determining an appropriate pressure drop across ICDs to achieve conformance. Secondly, the study evaluates the performance of ICDs that had been installed in four wells, two of which had ICDs installed prior to circulation and two that adopted ICDs later in their lifecycle. The field data shows that ICDs increase production rates and improve conformance along the lateral. These improvements are achieved by an increased drawdown facilitated by the ICDs. This part of the study highlights how early-life results may differ between ICD bearing wells compared to their conventionally completed (slotted liner) offsets, with the improved conformance and ability to develop more challenging reservoir resulting in different oil production profiles and composite SORs.
Objective/Scope: The definition of the locations of new wells in mature fields is a challenging problem especially in contexts of high geological complexity and low data reliability, when running fluid-flow simulations can be extremely difficult. For this reason, we develop an innovative Surrogate Reservoir Model, based on a data-driven process, which combines Machine Learning algorithms with spatial interpolation techniques. We call our approach WIZARD (acronym for: Well Infilling optimiZAtion through Regression and Data analytics). Methods/Procedures/Process: WIZARD is a collection of different data-driven methods for the sake of definition of new infilling well locations on the basis of the expected cumulative oil productions (after a fixed target period) of unexploited areas of the reservoir. The first method, named COSMIC, is used to find a correlation between petrophysical well properties and well productivity through a regression algorithm. The second method, that we call WIZARDMAP, uses spatial interpolation methodologies like K-Nearest Neighbours to estimate input petrophysical well data far away the existing wells and the trained COSMIC model applied to these interpolated data to predict the expected cumulative oil productions in unexploited areas of the reservoir. Finally, predictions of WIZARMAP model are compared with the ones given by another method that we call WIZARDROC, that is a predictive model trained by using only the cumulative oil productions of the existing wells and their locations.
Shoeibi Omrani, Pejman (TNO) | Vecchia, Adrian Luciano (Wintershall Noordzee B.V.) | Dobrovolschi, Iulian (TNO) | Van Baalen, Thijs (Wintershall Noordzee B.V.) | Poort, Jonah (TNO) | Octaviano, Ryvo (TNO) | Binn-Tahir, Huzaifah (Binn-Tahir Consulting) | Muñoz, Esteban (Wintershall Noordzee B.V.)
Reliable forecasting of production rates from mature hydrocarbon fields is crucial both in optimizing their operation (via short-term forecasts) and in making reliable reserves estimations (via long-term forecasts). Several approaches may be employed for production forecasting from the industry standard decline curve analysis, to new technologies such as machine learning. The goal of this study is to assess the potential of utilizing deep learning and hybrid modelling approaches for production rate forecasting. Several methods were developed and assessed for both short-term and long-term forecasts, such as: firstprinciple physics-based approaches, decline curve analysis, deep learning models and hybrid models (which combine first-principle and deep learning models). These methods were tested on data from a variety of gas assets for different forecasting horizons, ranging from 6 weeks to several years. The results suggest that each model can be beneficial for production forecasting, depending on the complexity of the production behavior, the forecasting horizon and the availability and accuracy of the data used. The performances of both hybrid and physical models were dependent on the quality of the calibration (history matching) of the models employed. Deep learning models were found to be more accurate in capturing the dynamic effects observed during production - this was especially true for mature fields with frequent shut-ins and interventions. For long-term production forecasting, in some cases, the hybrid model produced a greater accuracy due to its consideration of the long-term reservoir depletion process provided by the incorporated material balance model.
Prior to 2008, shale gas reservoirs were deemed uneconomical to produce. Hydraulic fracturing and horizontal drilling have shifted this perspective, reducing the flow resistance from the reservoir to the well. Notwithstanding the thousands of shale gas wells currently actively producing around the world, factors controlling the permeability and flow behaviour in shale gas formations are still incompletely understood. A profound understanding of the flow processes manifesting in shale gas reservoirs will contribute to more effective Enhanced Gas Recovery (EGR) schemes, ultimate recovery and accurate gas production forecasting.
Owing to the micro- and nano-size of pores, transport in shale rocks depends on the pore size and predominantly on pore geometry and tortuosity. To gain new insights into the mechanics of gas production from shale formations, we constructed a geometrically accurate model from an actual shale scanning electron microscope micro-image. Taking into account the pertinent rock and gas parameters (e.g., porosity, permeability, viscosity, etc.) we have determined the gas flow rate, the pressure variations and deduced the production rate at the micro-level.
A non-dimensionalization methodology was developed that permits the comparison between micro-scale modelling results with actual core measurements several orders of magnitude larger in special scale. Normalized micro-scale modelling results compare well with actual core data shedding light on some of the important aspects which govern gas flow: geometry, pressure gradient, compressibility, pore throats, and permeability. Moreover, the cumulative gas production for different gases was shown to improve with an increase in the molecular mass of the gases. Ultimately, our efforts aim to tie theoretical understanding with experimental observations deemed significant for boosting the productivity of gas from shale formations.
Al-Mai, Noura (Kuwait Oil Company) | Al-Shuaib, Muna (Kuwait Oil Company) | Alvarado, Omar (Kuwait Oil Company) | Al-Nesef, Mohammad (Kuwait Oil Company) | Al-Saleh, Alaa (Kuwait Oil Company) | Al-Qahtani, Shaikha (Kuwait Oil Company) | Useche, Marcos (Schlumberger) | Franco, Francy (Schlumberger) | Orjuela, Jaime (Schlumberger) | Wibowo, Arif (Schlumberger) | Prakash, Roshan (Schlumberger) | Gornescu, Bogdan (Schlumberger)
An improved workflow for handling an excess of Hydrogen Sulfide (H2S) in treatment facilities is presented. The objective is for the selection of wells to decrease H2S production by reduction of choke sizes with the lowest possible impact on oil production.
During operational restrictions in a gas treatment plant, it is necessary to reduce the H2S quantity received at the facility. To perform this operation, an approach is to limit the production from wells with the biggest rate production of H2S (volume); another solution is to restrict the wells with the highest concentration of H2S (% molar - composition); However, both scenarios have an impact on oil production. KOC - KwIDF Jurassic, has developed a procedure that considers the derivative of the produced H2S curve as a function of produced Oil to find the optimal scenario.
The proposed method was evaluated using the results from a sequence of monthly allocations that include characteristics such as: facility, number of wells, and fluid compositions. The data was analyzed applying three methodologies that have different objective functions, such as H2S volume, H2S concentration, and a new criterion called H2SOR. As a result, there were three different way to rank the wells to reduce H2S production. A monthly plot of Cumulative H2S Production vs Cumulative Oil Production was prepared to calculate the derivative of the curve, finding the optimal solution at different levels of H2S treatment restriction. Finally, the economic analysis for each method was also calculated.
H2SOR methodology guarantees the best alternative to ensure the restriction of sour gas rates with the lowest impact on oil production. It also proves to be the best solution from the economical point of view.
Unconventional resource development is increasing quickly in many places worldwide. For unconventional resources, multistage completions play a key role for both reservoir performance and well economics, which makes completion optimization a critical technical and commercial decision. This work integrates the reservoir modeling, fracture simulation, production forecast, and synthetic data pool generation via Monte Carlo methods, and it simplifies the final optimization process into a selection from multiple options.
There are many approaches used to optimize completion parameters in shale gas development in the Sichuan basin. Although a trial and error method may work well with an adequate number of wells, this approach is not efficient with few wells because it would take many years to optimize the drilling and completion strategy. Also, such an approach may produce ambiguous results related to high uncertainty due to drilling quality and completion inconsistencies.
An innovative workflow is defined in this work that combines reservoir modeling, fracture network simulation, production matching, regression analysis, and Monte Carlo methods. The procedure begins with modeling of the reservoir using the proper geological environment and reservoir properties. Based on this model, the hydraulic fracture network is simulated with varied compl etion parameter sets, including fluid volume, proppant volume, perforation spacing, and stage spacing. Production forecasting is then performed for each of the fracture network simulations, and the result is matched with previous offset well performance. Regression analysis is used to simplify the relationships between the input (completion parameters) and the output (production results). Finally, based on the regression results, a Monte Carlo method is used to generate a large number of input and output pairs creating a type of synthetic completion choice catalog. This catalog provides a pool of completion options, effectively reducing the optimization process to a choice of the best fit-for-purpose options.
A synthetic model based on Sichuan shale gas is used in this study to validate the workflow on a single- well basis. It successfully produced many synthetic simulation results. With the large number of completion parameters—production result pairs—it is easy to filter the results and identify which combinations are preferred in terms of cost and production. This work also demonstrates that optimization is subject to the definition of purpose and duration of the objectives, which can be used as an important evidence to support different strategies.
Application of horizontal drilling and hydraulic fracturing technique has made development of shale gas reservoir successful in the United States during the past decade. Chasing its operational success, researchers have been studying to understand the fundamentals of shale gas production, which will provide valuable information to assist in optimization of shale reservoir development. Unfortunately, the mechanism of shale gas production has not been fully revealed so far, and most reservoir simulation models are adopting the mechanism of coalbed methane production to forecast shale gas development process, which might not be the real case.
In this paper, instead of using numerical simulation model, artificial intelligence and data mining techniques are implemented to study the controlling factors of shale gas production and understand the impacts of reservoir, completion and stimulation parameters in a dynamic manner only according to the field data. A database of Marcellus shale reservoir is generated by integrating information such as well locations, well trajectories, reservoir characteristics, completion, hydraulic fracturing, and production parameters, etc. Neural network models are trained to learn the key performance impacting factors on shale gas production in a dynamic manner, which could assist reservoir management decisions.
The combination of extended-length horizontal drilling and high volume hydraulic fracturing has led to previously unimaginable production increases, yet the recovery potential of unconventional oil and gas resources remains largely unrealized. Recovery factors for unconventional oil and gas wells are typically reported at < 20% in gas shale reservoirs and < 10% in the oil plays.
Neutrally buoyant ultra-lightweight proppants have been demonstrated to effectively provide production from fracture area that is otherwise unpropped and thus, non-contributive with conventional sand/slickwater hydraulic fracturing processes. Production simulations illustrate that treatment designs incorporating neutrally buoyant ULW proppant treatment designs tailored for contemporary unconventional well stimulations deliver cumulative production increases of 30% to over 50% compared to the typical large volume sand/slickwater treatments. Unfortunately, production simulation results may not sufficiently lessen risk uncertainties for operators planning high-cost multi-stage horizontal stimulations. Therefore, several field trial projects using the neutrally buoyant ULW proppant in extended-length horizontal unconventional wells are currently in progress to validate the production simulations.
Since the initial 4-stage fracturing stimulation incorporating neutrally buoyant ultra-lightweight proppant in 2007, deployment has occurred in fracture stimulating hundreds of oil and gas wells spanning multiple basins and reservoirs. Most of the wells are vertical or relatively short lateral wells common to asset development practices predating the unconventional shale completions mania, but many were targeted at the same unconventional reservoirs as the current multi-stage horizontal completions. Several published case histories have documented the production enhancement benefits afforded by the legacy ULW proppant wells, but questions remained as to how those lessons might be correlated to provide engineers confidence in the current production simulations.
Well completion and production information was mined from the various accessible databases for the neutrally buoyant ULW proppant wells. The scope of the legacy data compiled for analysis was limited to the reservoirs common to the current field trials and production simulations, ie. unconventional oil and gas shale reservoirs. Production performance contributions of neutrally buoyant ULW proppant in past applications were compared with the production uplift observed in applications and/or simulated application of neutrally buoyant ultra-lightweight proppant fracturing treatments in current multi-stage horizontal reservoirs.
The lessons learned from this investigation provide the practicing engineer the means to confidently assess production simulation data for multi-stage horizontal unconventional completions incorporating neutrally buoyant ulw proppant in the treatment designs.
Objectives/Scope: In order to maximize the recovery of hydrocarbons from liquids rich shale reservoir systems, the cause and effect relationships between production and the stimulation methods need to be clearly understood. In this study, we utilize multivariate regression models to narrow down the variables in flow simulation models and their range. We then use the flow simulation model to understand the fractured well production behavior and field wide well performance in a liquids rich petroleum system in the Duvernay Basin.
Methods, Procedures, Process: Statistical models assume no physical relationship between the model parameters and the response variable, which in this case is produced volumes over a period of time. On the other hand, simulation studies incorporate physical mechanisms of flow to model and predict the production behavior. The simulation models, however, fall short of incorporating all the mechanisms contributing to the production behavior in the complex shale gas reservoir. Thus there is a need for integration of statistical approaches of understanding production behavior along with physics based model and simulation approach. We use the statistical methods to identify the important physical mechanisms that control the production.
Results, Observations, Conclusions: Multivariate linear regression analysis of the 6 month produced volume and its relationship with parameters such as fracture fluid volumes used, proppant weight placed, number of stages fractured provides a model with reasonably good correlation. The 6 month produced volumes correlate with large proppant weights, lower fluid placements and greater density of fracture stages. Use of Random Forests machine learning algorithm on the dataset confirms that the total proppant placed, well length completed with fractures have high importance coefficients. In order to examine the well performance using full physical models, fractured well simulations are performed on particular wells using the trilinear model. The trilinear model predictions are then compared against other production analyses and the regression model results for consistency. The models showed that in the absence of stress dependent permeability, the production forecast was much higher. Thus, stress dependent permeability appears to be an important factor in the modeling and prediction of production from liquids rich shale reservoirs.
Novel/Additive Information: In this study we describe a method to understand the production data from a liquids rich shale reservoir, by integrating multivariate linear regression analysis, machine learning algorithms along with physical model simulations. The results are novel and offer a method to validate either approach to understand cause and effect relationships. This approach may be classified as a new hybrid modeling workflow that may potentially be used to optimize stimulation techniques in liquids rich shale reservoirs.
One of the major challenges associated with the exploitation of unconventional hydrocarbon resources is determining the optimal stimulation design. In this sense, it is necessary to understand how the parameters and variables involved in the completion process impact on production performance; the purpose is to act on such controllable variables and, consequently, maximize production and field development efficiency. Whereas physical driven tools frequently used in the oil industry are very helpful, they always imply a set of assumptions and simplifications regarding the system or phenomenon they try to model; they also require a large amount of unavailable or expensive data to calibrate them. Generally, different combinations of model parameters could explain well production behavior and for each of these solutions the way to optimize completion and development may be different.
Because of these drawbacks, and the big number of unconventional wells available, data-driven workflows have gained popularity in the last years. These models represent an excellent complement to physical driven tools in the attempt to optimize the completion and development strategy in shale plays. Several publications used both parametrical and non-parametrical models in the search of the Holy Grail: a statistical model capable of predicting how stimulation design affects productivity. The aim of this paper is to develop a novel methodology to understand the relation between formation parameters, completion design variables and production performance. An artificial neural network model (ANN) was chosen for this study.
Public production and stimulation data was merged with geological and petrophysical properties maps for almost 13.000 horizontal wells landed in Eagle Ford formation. A back propagation ANN algorithm was trained with this data-set and a cross-validation criterion was used for hyper-parameters optimization. Once the optimal model was selected, a bootstrap algorithm was run to assess for uncertainty in model prediction; these models were trained to determine which part of the input space presented enough data to get a clear signal and in which part the amount of data was not enough to differentiate signal from noise.
ANN models proved to be a fine method for this purpose obtaining R-Squared values between 0.5 and 0.7 for cross-validation sets. Significant relations were observed between production performance and lateral length, true vertical depth, porosity and fracture fluid intensity.
The methodology presented in this paper introduces a novel feature in comparison to previous publications regarding model uncertainty assessment. The coupling of the ANN model with the bootstrap re-sampling technique allowed to better understand which conclusions were statistically significant and which not, a fact that proved to be vital to correctly interpret results. It was demonstrated that such methodology is a good complement to physical-driven models in the aim to comprehend the relation between formation parameters, completion design variables and production performance.