|Theme||Visible||Selectable||Appearance||Zoom Range (now: 0)|
According to the latest SPE-PRMS guidelines, reservoir simulation may be applied in reserves estimation. However, no clear guidance is available (from SPE or public sources) on how dynamic models should be used. This paper addresses this gap by providing a practical overview of the application of reservoir simulation in reserves evaluations, based on the authors' experience and a literature review. Furthermore, a novel simulation-based reserves estimation method is presented.
SPE released an update of the PRMS reserves estimation guidelines in 2018. It is noted that the most common methodology applied is decline curve analysis (DCA). However, in many circumstances DCA alone is not sufficient for a reliable reserves estimate. To improve the reliability of the estimate, results from reservoir simulation and / or analogue reservoirs could be required. A dynamic model may be used in several ways in reserves estimation: To generate production forecasts based on a consistent physics-based model containing geology, and assuming different subsurface realizations, to establish a range of technical recoverable resources (TRR) To provide an estimate of recovery factor in order to validate the forecasts from other prediction techniques To establish oil/gas decline trends for comparison to the DCA results To provide vital insights into the dynamic behavior of the reservoir that can be further used in DCA, material balance or sector models.
To generate production forecasts based on a consistent physics-based model containing geology, and assuming different subsurface realizations, to establish a range of technical recoverable resources (TRR)
To provide an estimate of recovery factor in order to validate the forecasts from other prediction techniques
To establish oil/gas decline trends for comparison to the DCA results
To provide vital insights into the dynamic behavior of the reservoir that can be further used in DCA, material balance or sector models.
Several articles have addressed the use of reservoir simulation in reserves estimation. It is noted that the main challenge is to generate the range of uncertainties in the TRR, not only the "best technical estimate". Reserves assessors usually apply a combination of several methods to estimate the uncertainty in the remaining oil recovery. A novel reserves estimation method is introduced, which uses results from reservoir simulation on analogous reservoirs.
This paper provides a detailed overview of all the reserves estimating methods that can be applied with help and guidance from reservoir simulation models to determine 1P, 2P and 3P reserves, while honoring SPE-PRMS guidelines. Also, a novel technique is presented that can be used for reserves determination of a large number of small reservoirs. It is shown how this technique was successfully applied during a reserves assessment for a group of fields in the Middle East.
The paper provides a comprehensive yet practical overview of how reservoir simulation may be applied in reserves evaluations.
In addition, a novel method is described. It entails the estimation of the range of TRR's for a number of reservoirs using reservoir simulation and adopting a statistical approach while honoring SPE-PRMS guidelines. The spread in ultimate recovery factors can be combined with the maturity index to establish a technical uncertainty relationship, that can be applied to estimate the reserves of analogous reservoirs.
The PDF file of this paper is in Russian.
North Morgan Belayim is a mature reservoir with more than 40 years of production and injection. Low rate, low pressure and high water cut wells are the main features for the Belayim complex reservoir. Integrated static and dynamic study was conducted across more than 100 wells to have a reliable reservoir description, set a full depletion plan and determine bypassed oil potential.
Pressure and production performance and mapping of water cut, salinity and pressure were analyzed concluding reservoir compartmentalization. That was very consistent with findings from static describtion; Structure, Stratigraphic and formation evaluation components. Estimated recovery factor, RFT data and production logs showed the reservoir requires a new zonation as the current does not explain scattered measured data. Rock typing was performed for indetifying new zonation. Voidage replacement ratio exercise was used to evaluate water flood efficiency.
After the study, the compartmentalization has been supported by stratigraphic findings; the reservoir is found to be composed of two big geological fans while there is no single evidence from structural aspects was found to affect connectivity between sand bodies. The rock typing was very efficient in defining flow units. A total of six distinctive units have been identified instead of three units. This helped a lot in distinguishing between producible vs inproducible zones after considering the permeability cut-off. This was significant in fine-tuning STOIIP and recovery factor values. The better understanding of reservoir helped asset team to change its strategy in such field development not only for optimizing wells’ locations but also for waterflooding management and perforation strategy considering low-permeability and high-permeability zones.
This paper is a good example for integrating static description with dynamic data seeking better reservoir understanding. In addition, this paper proves the criticality of crosschecking different tools in filling gaps and optimizing redevelopmetion options in mature fields.
Tyrie, Jeb (Bridge Petroleum) | Mulcahy, Matt (Bridge Petroleum) | Leask, Robbie (Bridge Petroleum) | Wahid, Fazrie (Bridge Petroleum) | Arogundade, Olamide (Schlumberger) | Khattak, Iftikhar (Schlumberger) | Apena, Gani (Schlumberger) | Samy, Mohammed (Schlumberger) | Sagar, Rajiv (Schlumberger) | Xia, Tianxiang (TRACS International) | Nyadu, Kofi (WorleyParsons, Advision) | Maizeret, Pierre-David (Schlumberger)
This paper describes the proposed re-development of the Galapagos Field, comprising the abandoned NW Hutton field and the Darwin discovery (Block 211/27 UKCS) which forms a southerly extension. The paper covers the initial concept and analytical evaluation, the static uncertainty model build, the dynamic model history-match, the iterations between static and dynamic modelling, the development subsea and well locations, the optimisation workflow of the advanced Flow Control Valve (FCV) completions in both producers and injectors and the facilities constraints.
The redevelopment plan involved several multi-disciplinary teams. 20 years of production data from 52 wells were analysed to identify the production behaviour and confirm the significant target that provided the basis for the development concept selection. The full Brent sequence compartmentalised stochastic static model was based on reprocessed seismic plus 14 exploration and appraisal wells. Streamlines, uncertainty sensitivities and mostly good detective work honed a history match to RFT, BHP, PLT and oil and water production. P50, P90/P10 models were selected and over 100 FCVs optimised to deliver the profiles against an identified FSPO facilities’ constraints.
Over 1,000 static models were delivered consisting of sheet sands, incised valleys and channels in heterolithic facies overprinted by a depth trend with appropriate uncertainty ranges. The high well count gave a tight STOIIP probabilistic range of 790/883/937 million stb. The early RFTs illustrated extreme differential depletion between Brent zones and subzones of the Ness. To history-match these the dynamic model retained the static model definition in the Upper Ness to capture the thin but extensive shales. The early 18-month depletion and the late steady production-injection phases were simulated separately in prediction mode and matched the Production Analysis estimated ‘future’ production giving confidence to the history matched model. The initial concept development of 4 subsea-centres, to cover the large field area, with an injector in each compartment proved a robust selection. The horizontal wells increase PI where needed and mitigate internal faulting. The optimisation of the FCVs significantly increased oil production from all zones and drastically reduced water injection and production so that the identified FPSO modifications were relatively modest. The final First Stage Field Development Plan consists of 11 producers and 6 injectors across developed and undeveloped areas confirmed robust P50 reserves of 84 million boe.
Robust concept selection allowed for early identification of production units so that constraints and modifications could be accounted for within the economic model.
The Galapagos field re-development plan is an excellent example of how detailed static and fully history matched dynamic models can lay the foundations for new technology like the optimisation of the FCVs to access bypassed reserves using significantly smaller production units with reduced requirements for power, compression, gas lift, pumping pressure, injection and production. In short, they shrank the facilities.
Lawal, Kazeem A. (FIRST E&P) | Okoh, Oluchukwu M. (Nigerian Petroleum Development Company) | Yadua, Asekhame (Nigerian Petroleum Development Company) | Ovuru, Mathilda I. (FIRST E&P) | Eyitayo, Stella I. (FIRST E&P) | Ramaswamy, Sunil (Schlumberger)
Abstract Given sufficient performance and other data, material balance (MB) is a common method of determining the hydrocarbons initially in-place (HCIIP) in a reservoir. The application of this method requires, as a minimum, historic cumulative production (including injection) and average reservoir pressure. However, determination of historic average reservoir pressures would require shut-in of wells, hence production deferments. As an improvement to the classical MB, the dynamic material balance (DMB) method was developed by Mattar and Anderson (2005). Unlike the MB method, direct measurements of average reservoir pressure are not critical to DMB. In its basic form, the implementation of DMB requires historic production rates, flowing bottomhole pressures and cumulative production, thereby eliminating associated deferments. Although DMB has performed satisfactorily in some applications, its overall robustness remains to be fully explored. This paper conducts rigorous sensitivity checks on selected DMB models. Based on insights gained, their relative strengths and weaknesses are highlighted. To keep the problem tractable, detailed simulations are performed on different three-dimensional (3D) multiphase homogenous reservoir models of known HCIIP. Different cases are simulated, generating relevant performance datasets to evaluate DMB. The parametric tests conducted on this undersaturated compressible oil reservoir include (i) constant vs. variable production rates; (ii) rate hysteresis; (iii) vertical vs. horizontal well; (iv) single vs. multiple wells; (v) healthy vs. damaged well; and (vi) variable skin factors, with hysteresis. Within the parameter space examined, simulation results show that DMB performance (e.g. HCIIP) is sensitive to some of the parameters and subsurface realisations investigated. Against this background, some improvements and guidelines are proposed to enhance the capability and performance of DMB as a technique for reservoir surveillance.
Lu, Xiaoguang (C&C Reservoirs) | Xu, John (C&C Reservoirs) | Feng, Lijing (N0. 4 Oil Production Company, Daqing, PetroChina) | Yang, Qing (C&C Reservoirs) | Li, Guoqiang (C&C Reservoirs) | Lin, Lihua (C&C Reservoirs)
Abstract The XB Field in China contains more than 100 thin sand units deposited in a non-marine environment, which results in an extremely heterogeneous sandstone reservoir. Comparison with global field analogs of similar reservoir characteristics indicates that the >60% ultimate recovery in the XB Field is much higher than average. This paper reviews the over 50 years of production history and summarizes its development strategies, successful reservoir management practices, key IOR/EOR technologies and lessons learned, which can benefit efforts of maximizing recovery in other reservoirs. This paper begins by summarizing the basic reservoir and fluid characteristics as well as the production performance history. This is followed by a benchmarking analysis focused on reservoir heterogeneity, fluid properties and recovery factor against global analog reservoirs. Finally, the authors highlight the development strategy, key IOR/EOR methods, and integrated reservoir management practices based on fit-for-purpose reservoir and remaining oil characterization studies. The benchmarking study against global analogs shows the XB reservoir to possess much higher heterogeneities and poorer fluid properties than average. The field is expected to achieve an ultimate recovery of more than 60%, which is substantially higher than the average of 36.7% and P50 value of 36% based on global reservoir analog in C&C Reservoirs DAKS. The key IOR methods applied include pressure maintenance through water injection starting at early development stage, infill drilling, and chemical EOR methods. Water injection and infill drilling have helped improve recovery by 30% and 20%, respectively. Water injection optimization has been applied throughout the 50-year production history, focused on by-passed oil or poorly swept areas. Zonal water injection, subdivision of injecting-producing unit, modification injection pattern and cyclic water injection are methods of this category. Other IOR methods, such as horizontal well targeting by-passed oil, profile modification, and fracturing of low permeability reservoir sands also contribute to the high recovery factor. When the field entered the mature production stage, field-wide polymer and ASP flooding have been implemented based on numerous laboratory studies and pilot tests. The chemical EOR application in this field is one of the most successful cases in the world. Polymer flood and ASP flood are expected to achieve incremental recovery factor of 10% and 20%, respectively. The XB Field case suggests that many mature fields in the globe have the potential to further improve their recovery. Most of the technologies discussed in this paper are well established, conventional, inexpensive and readily available. The key point is that detailed reservoir characterization, remaining oil identification and application of lessons learned from global analogs are of prime importance.
Nurliev, D. R. (RN-UfaNIPIneft LLC) | Rodionova, I. I. (RN-UfaNIPIneft LLC) | Viktorov, E. P. (RN-UfaNIPIneft LLC) | Shabalin, M. A. (RN-UfaNIPIneft LLC) | Makeev, G. A. (RN-UfaNIPIneft LLC) | Novikov, N. O. (RN-UfaNIPIneft LLC) | Suleimanov, D. D. (RN-UfaNIPIneft LLC)
The PDF file of this paper is in Russian.
World oil industry current tendency is a transition to development tight reserves. In order to take into account all possible risks related to different kind of geological and technological uncertainties to help make a decision about project economic efficiency, we propose to utilize multivariate simulation for production forecast. Based on actual data from drilled out zones in deep-water part of clinoforms three types of sections were selected. A geological uncertainty those sections is determined by distance from a source of conglomerate. Digital models for each type were created with keeping major morphological features of associated sand bodies. A pool of simulation models was created by variating geological parameters (absolute permeability, net thickness, Corey coefficients at oil relative permeability curves, original oil in place) and technological parameters (fracture half length, fracture height, fracture conductivity). A process of creating multiple models was accomplished by program module RExLab and Rosneft’s program software complex RN-KIM. Latin hypercube algorithm was applied for generating approximately 5000 models for each type of geology. Out of calculated models were selected cases matching actual liquid production and water cut. The selected cases were being used for evaluating technical and economic efficiency of a development pattern within forecast period. As a result of calculation was given Recovery Factor probability distribution function and cumulative NPV per development areas which allow to accomplish an economical project assessment out of P10/P50/P90 perspectives.
Современная тенденция в мировой нефтегазовой отрасли - это переход к разработке коллекторов с запасами нефти, относящимися к трудноизвлекаемым. Для максимального учета возможных рисков, связанных с различного рода неопределенностью геологических и технологических параметров, при принятии решения об экономической эффективности проекта предложено использовать многовариантное гидродинамическое прогнозирование показателей разработки. На основе фактических данных, полученных в уже разбуренных зонах глубоководной части клиноформы, выделены три типа разрезов, геологическая неоднородность которых обусловлена степенью удаленности от источника сноса обломочного материала. Для каждого типа построены цифровые геологические модели с сохранением основных морфологических особенностей сопутствующих песчаных тел. Вариацией геологических (абсолютная проницаемость, эффективная толщина, степень Кори в функции относительной фазовой проницаемости для нефти, начальные запасы нефти) и технологических (полудлина трещины гидроразрыва пласта (ГРП), степень вскрытия пласта, проводимость трещины ГРП) параметров создана выборка моделей для адаптации. Множество моделей создано с помощью программного модуля RExLab и корпоративного программного комплекса «РН-КИМ». С использованием алгоритма «Латинский гиперкуб» сгенерировано около 5000 моделей для каждого типа отложений. Из множества полученных расчетов выбраны модели, обеспечивающие фактическую добычу жидкости и обводненность. Выбранные модели использованы для оценки технико-экономической эффективности системы разработки в прогнозный период. В результате расчетов получены распределения вероятных значений коэффициента извлечения нефти и накопленного чистого дисконтированного дохода, приведенного на гектар площади, которые позволяют провести оценку экономической эффективности проекта.
Abstract Enhanced oil recovery (EOR) applications are strongly recommended and required in Sultanate of Oman to maintain stable oil production levels. Thermal EOR is of particular importance in view of the abundance of heavy, viscous oil within many field clusters in the Sultanate. The selected field for this study is located in South of Oman, where relatively-thin clastics reservoirs at moderate depths are dominant. Reservoir oil is highly under-saturated, 16 API gravity with very low solution gas oil ratio, initial oil viscosity over 800 cP and initial reservoir pressure of about 8,700 kPa. Experience to be gained in this moderate size reservoir has many possible applications in larger reservoirs with similar characteristics within the cluster. The selected reservoir is developed mostly by cold production using horizontal wells with landing points at maximum possible distance above the initial oil-water contact. Few vertical wells were also drilled with thermal compliance for Cyclic-Steam-Stimulation (CSS) applications. Initial cold production rates are in the range of 15 - 35 STm/d/ well at negligible water-cut but within few months the oil rate declines and water-cut increases due to water cresting. CSS applications in thermally-completed vertical wells show encouraging response in terms of steam injectivity (up to 200 mCWE/d) and gain in oil rate (up to 50 STm/d/well) but led to sanding problems in some wells. Field production rate reached a peak two years after start of development then continued to decline with an increase in water-cut till current level of 85%. Cumulative oil production corresponds to only 5.4% of Stock-Tank-Oil-Initially-In-Place (STOIIP). Reservoir pressure indicated a slight decline (about 300 kPa) due to the strong bottom aquifer beneath the relatively thin oil column. Detailed reservoir characterization, modeling efforts for the dipping, truncated structure and optimum development scenarios (under both, cold production and CSS/steamflood applications) are discussed in this paper. Steamflood simulation includes defining the optimum flood pattern and injection scheme as well as quantifying the effect of bottom aquifer with emphasis on utilization of aquifer depletion wells. Predicted performance for the selected development scenarios are compared with common industry practice and analytical methods for similar heavy oil reservoirs. Basis for selecting an optimum location, pattern configuration and injection rate for a steamflood pilot are also included. Simulation results indicate that continued infill drilling using horizontal wells for cold production could achieve an ultimate recovery factor up to only 14% of STOIIP while steamflood (if aquifer depletion wells are successful) could add up to 47% of STOIIP over cold production case. The optimum flood pattern is based on horizontal producers (perpendicular to strike) drilled in the middle of oil column and vertical injectors midway in-between the producers. Effects of pattern size, injection interval, steam injection rate and steam quality on steamflood performance are also included. Cumulative steam-oil ratio is expected to be in the range of 4.0-4.5 mCWE/STm. Economic evaluation results are discussed to show the sensitivity of project economics to oil price and various alternatives of fuel supply for steam generation.
Shankar, Vivek (Cairn India - Oil & Gas Vertical of Vedanta Limited) | Beliveau, Dennis (Cairn India - Oil & Gas Vertical of Vedanta Limited) | Jain, Shakti (Cairn India - Oil & Gas Vertical of Vedanta Limited) | Jha, Bhawesh Chandra (Cairn India - Oil & Gas Vertical of Vedanta Limited) | Tiwari, Arjun (Cairn India - Oil & Gas Vertical of Vedanta Limited) | Kumar, Chandan (Cairn India - Oil & Gas Vertical of Vedanta Limited) | Rautela, Chandramohan (Cairn India - Oil & Gas Vertical of Vedanta Limited) | Stanley, Matthew (Cairn India - Oil & Gas Vertical of Vedanta Limited) | Chapman, Thomas (Cairn India - Oil & Gas Vertical of Vedanta Limited)
Abstract Bhagyam is a large oil field in the Barmer Basin of Rajasthan, India. The major producing intervals are shallow Paleocene-aged Fatehgarh sandstones. The paraffinic oil is viscous (20-500 cP) with wax appearance temperature (WAT) only 2-3°C lower than the average reservoir temperature of ~53°C. Bhagyam has been developed with 153 wells in an edge linedrive waterflood and has been producing since 2012. However, performance has been less than projected in the initial field development plan (FDP): not as good as the nearby Mangala and Aishwariya Fatehgarh waterfloods and at the lower-end of that seen in worldwide viscous oil waterflood analogues. Key contributors to Bhagyam's performance were lower than expected initial well productivity and a more rapid rise in water cut than projected. The lower initial well productivities were surprising when compared with experience from Mangala, where initial well productivity was closely aligned with expectations. Simulation models could not replicate Bhagyam performance without numerous major local modifications; hence long-term model predictions were not sufficiently reliable for business planning. Reservoir behaviour was initially attributed to severe heterogeneity and early models used high permeability streaks to match performance. However, saturation logs and selective zonal flow back of a few wells showed that injected water was not confined to select high permeability streaks but was widespread both areally and vertically. The more diffuse nature of water movement suggested a fundamental disconnect between field dynamics and the simulation model description and physics.
The PDF file of this paper is in Russian.
An integrated analytical approach is presented that predicts water breakthrough timing of producing wells in the absence of surveillance (sparse PLTs, reservoir pressure and PTA) and seismic data. It is critical to understand water movement because 20% of the wells in the reservoir have produced water and are now closed for production due to lack of water handling facilities. Results are compared to existing dynamic models which are unable to accurately predict water breakthrough timing.
This approach uses production and tubing head pressure data along with gamma ray logs and wells’ saturation data to provide insights into reservoir properties. It combines geological study along with classical analytical methods like Stiles’ calculations and Rate Transient Analysis (RTA) to understand and predict the movement of water in the reservoir. The approach uses RTA for estimating key reservoir properties and Stiles’ method to calculate water breakthrough timing in the wells. After combining inferences from all data, sweep efficiency calculations are used to corroborate the water breakthrough timingspredicted by dynamic model with analytically calculated timings. Detailed algorithms and workflows are provided.
Findings from geological study of this oilfield in the Middle East and results from Rate Transient Analysis show that water is moving preferentially in certain areas in the reservoir. Combining this concept of preferential water movement with Stiles and sweep efficiency calculations predicts water breakthrough of wells that have broken water within a range of 3 months when compared to history. Converting the conceptual model into a geocircular grid for simulation enables the model to test different geological scenarios and capture uncertainties. The new history-matched models not only predict water breakthrough timing of the wells but also cover a range of geological uncertainties. Better forecast of water production helps calibrate the upcoming water handling facility at the site for improved production from the field.
The novelty of this approach is in the simplicity and speed with which it can solve a complex problem. It overcomes a very common hurdle of data uncertainty and unavailability to answer the pertinent question of water breakthrough timing at each producer well using standard analytical techniques.