Objectives/Scope Round-robin tests and laboratory audits have demonstrated that reservoir fluid compositions measurements can be systematically uncertain. To reduce compositional uncertainties, this work uses Machine Learning (ML) algorithms to update the existing measured fluid compositions and applies a compositional adjustment anchored on a decreased number of selected fluids representative of the entire compositional space. The resulting composition will reduce uncertainty in exploration and production models and diminish the risks in field development decisions. Methods, Procedures, Process Statistical error propagation analysis of reservoir fluid compositional measurements were completed to identify the parameters that most affect the overall uncertainty of the compositional analysis data. Advanced ML techniques were then used to perform clustering analysis of 800+ unique fluid compositions from legacy PVT reports. Dimensionality reduction exercise and data clustering utilizing the k-means algorithm were employed to identify parameters that best characterize the reservoir fluid properties to be corrected. The sampling candidate wells were identified to obtain new samples for the selected anchor points fluids that will be subjected to detailed fluids characterization. Results, Observations, Conclusions Systematic error propagation analysis demonstrated that, of the 140 independent parameters measured to determine a reservoir fluid composition (to C30+), the plus fraction concentration and molecular weight measurements uncertainties contributed the most towards the overall composition uncertainty for most black oil fluids. After extraction and collation of available reservoir fluid analysis data, a detailed review was completed to eliminate samples with significant mud filtrate contamination, obvious light end weathering and/or other inconsistencies. Once the dataset was validated, the dimensionality reduction task led to 5 principal dimensions space out of 57 initial possible features. The reduced dimension set included two distinct Cn- molar distribution slopes that best correlated the compositions data. Data clustering analysis was performed to identify the cluster centroids in the reduced dimension space. Wells with the closest compositions to cluster centroids were selected for sampling based on this work and the suitability of target wells for sampling. Further, extensive fluid compositions analysis with physical separation processes will be undertaken on the 50 selected anchor point fluids. Novel/Additive Information An original combination of reservoir fluids domain knowledge and advanced ML techniques have been applied to identify candidate wells to sample with the goal to significantly reduce the uncertainty in measured reservoir fluid composition. This unique workflow for composition corrections will help reduce the uncertainty in the critical techno-economic decisions that depend on fluids characterization data in the oil and gas field development domain.

No abstract

Objectives/Scope: Oil and gas operators use a variety of reservoir engineering workflows in addition to the reservoir, production, and surface facility simulation tools to quantify reserves and complete field development planning activities. Reservoir fluid property data and models are fundamental input to all these workflows. Thus, it is important to understand the propagation of uncertainty in these various workflows arising from laboratory fluid property measured data and corresponding model uncertainty. The first step in understanding the impact of laboratory data uncertainty was to measure it, and as result, ADNOC Onshore undertook a detailed study to assess the performance of four selected reservoir fluid laboratories. The selected laboratories were evaluated using a blind round-robin study on stock tank liquid density and molar mass measurements, reservoir fluid flashed gas and flashed liquid C30+ reservoir composition gas chromatography measurements, and Constant Mass Expansion (CME) Pressure-Volume-Temperature (PVT) measurements using a variety of selected reservoir and pure components test fluids. Upon completion of the analytical study and establishing a range of measurement uncertainty, a sensitivity analysis study was completed using an equation of state (EoS) model to study the impact of reservoir fluid composition and molecular weight measurement uncertainty on EoS model predictions. Methods, Procedures, Process: A blind round test was designed and administered to assess the performance of the four laboratories. Strict confidentiality was maintained to conceal the identity of samples through blind test protocols. The round-robin tests were also witnessed by the researchers. The EoS sensitivity study was completed using the Peng Robinson EoS and a commercially available software package. Results, Observations, Conclusions: The results of the fully blind reservoir fluid laboratory tests along with the statistical analysis of uncertainties will be presented in this paper. One of the laboratories had a systemic deviation in the measured plus fraction composition on black oil reference standard samples. The plus fraction concentration is typically the largest weight percent component in black oil systems and, along with the plus fraction molar mass, plays a crucial role in establishing the mole percent overall reservoir fluid compositions. Another laboratory had systemic issues related to chromatogram component integration errors that resulted in inconsistent carbon number concentration trends for various components. All laboratories failed to produce consistent molecular weight measurements for the reference samples. Finally, one laboratory had a relative deviation for P-V measurements that were significantly outside the acceptable range. The EoS sensitivity study demonstrates that the fluid composition and stock tank oil molar mass measurements have a significant impact on EoS model predictions and hence the reservoir/production models input when all other parameters are fixed. Novel/Additive Information: To the best of our knowledge, this is the first time such an extensive and fully blind round-robin test of commercial reservoir fluid characterization laboratories has been completed and published in the open literature. The industry should greatly benefit from this first-of-its-kind blind round-robin dataset being made available to all. The study provides the basis, protocols, expectations, and recommendations for such independent round-robin testing for fluid characterization laboratories on a broader scale.

PVT data is critical for fluid characterization and EOS modelling, which in turn is the key to estimate the initial hydrocarbon volumes in place, predicting reservoir dynamic behavior and production forecasting. Hence, ADCO gives high importance to have well distributed and high quality PVT data across its different reservoirs. There are various parties involved in planning and execution of PVT studies. There have been problems associated with PVT studies workflow such as poor scope of work, data quality issues, lost historical PVT data that affects the fluid characterization and EOS modelling. In addition, there are inefficiencies due to lack of coordination amongst various activities and parties involved in the PVT studies workflow. In this regard, ADCO embarked on standardizing and automating the entire process from PVT analysis requisition through technical data validation and archiving towards building an integrated PVT e-catalog. In our previous SPE paper (SPE-172832) we discussed our challenges, implementation strategy, and functional design phase of the project including the roadmap. In this paper discussed the progress and achievements made by ADCO towards establishing PVT intelligence Solution and integrated PVT e-catalog. The Solution enables the users to efficiently design, PVT analysis and sampling programs for a variety of fluids types and different studies such as routine, enhanced oil recovery studies, asphalltene studies. It also enables the user to perform consistent and standard validation for the PVT data during project execution The standard scope of work for various PVT studies is predefined based on the industry standards and well defined business logic. The workflow intelligently guides the users to build the right scope of work for PVT analysis and sampling based on the business objectives and reservoir information. The solution has functionality to technically validate the data delivered by the contractors in stages through approval workflows. This is done through a smart quality control tools which is supported by business rules for various tests that enables a semi-automatic quality control for the measured PVT properties. Finally the workflow enables the users to systematically archive quality data in an integrated PVT e-catalog. The e-catalog comprise of PVT database and functionalities for quick search of PVT data across ADCO fields and reservoirs. PVT intelligence enables the users to quickly and independently create high quality PVT sampling and analysis Scope of Work through an automated workflow. The workflow reduces the user effort and uncertainty in manual quality control for the data delivered by the laboratories and resolve data anomalies at the right time. The intelligent functionalities shall improve the quality of the PVT e-catalog which in turn enables better engineering models and calculations for Efficient Fluid Characterization.

No abstract