Shahab D. Mohaghegh – Distinguished Lecturer

Abstract:

Subsurface Analytics is an alternative to traditional reservoir modeling and res. management.

Positively influencing subsurface related decision making is the most important contribution of any new technology. Subsurface Analytics is the application of Artificial Intelligence and Machine Learning (AI&ML) in Reservoir Engineering, Characterization, Modeling, and Management. Applicable to both conventional and unconventional plays, Subsurface Analytics goes far beyond the traditional statistical algorithms that use only production data and fail to take into consideration the important field measurements such as well trajectories, well logs, seismic, core data, PVT, well test, completion, and operational constraints. Subsurface Analytics is the manifestation of Digital Transformation in Reservoir Engineering, Modeling, and Management.

Subsurface Analytics is a new and innovative technology that has been tested and validated in a large number of real life cases in North and Central America, North Sea, Middle East, and Southeast Asia. It has been successfully applied in several highly complex mature fields where conventional commercial reservoir simulators were unable to simultanuously history match multiple dynamic variables for large number of wells. Results and field validations from multiple case studies are included in the presentation.

Subsurface Analytics addresses realistic and useful applications of AI&ML in the upstream Exploration and Production Industry.  The technology has been validated and confirmed for (a) prediction of well behavior under different operational conditions, (b) modeling and forecasting pressure and saturation distribution throughout the reservoir, (c) infill well location optimization for both producers and injectors, (d) choke optimization for production improvement, and (e) completion optimization for production enhancement.

Biography:

Shahab D. Mohaghegh, a pioneer in the application of Artificial Intelligence and Machine Learning in the Exploration and Production industry, is Professor of Petroleum Engineering at West Virginia University and founder of Intelligent Solutions, Inc. He has authored three books (Shale Analytics – Data Driven Reservoir Modeling – Application of Data-Driven Analytics for the Geological Storage of CO2), more than 200 technical papers and carried out more than 60 projects for independents, NOCs and IOCs. He has been featured as the Distinguished Author in SPE’s Journal of Petroleum Technology (JPT 2000 and 2004). He is the founder of SPE’s Petroleum Data-Driven Analytics Technical Section. He has been honored by the U.S. Secretary of Energy for his AI-based technical contribution in the aftermath of the Deepwater Horizon and was a member of U.S. Secretary of Energy’s Technical Advisory Committee (2008-2014). He represented the United States at ISO on Carbon Capture and Storage (2014-2016).

Extracts from URTEC 2790, SPE 2020 Workshop on Rate/Pressure Transient Analysis and Distinguished Lecturer Presentation 2019/20

Lawrence CAMILLERI

With the increasing volatility in the price of oil, even low-cost onshore wells can be economically marginal and not achieve target rates of return under certain oil price scenarios and P10 or even P50 production profile forecasts.  If such market conditions persist, it will become increasingly important that all wells produce to their full potential and that a war be waged on skin and that drainage area be maximised. Where wells are fractured, maximising fracture conductivity and half-length will also be essential. It is the author’s belief that this macro-economic environment will drive an increase in pressure transient analysis (PTA) complemented by rate transient analysis (RTA) because, just as Lord kelvin stated in 1883, what we cannot measure, we cannot improve. 

Traditionally PTA requires build-ups, which is costly, both in terms of intervention and deferred production.  Fortunately, the population of permanent downhole gauges does not cease to increase and this presentation examines how downhole real time pressure and rate data enables PTA in drawdown. Such technology not only provides a source of low-cost transient data, but also the ability to obtain inflow characterisation (mobility, skin, fracture half-length and conductivity and boundary conditions)) within a few weeks of completing a well, thereby confirming the time value of information. 

To illustrate the concepts, PTA and RTA results from a MFHW (Multi Fractured Horizontal Well) case study are reviewed. This is based on data taken from URTEC 2790 paper and is particularly interesting as it demonstrates how high-quality production data can resolve a large number of unknowns. Also the history matching of flowing pressure with an analytical model yielded excellent results despite many discreet changes in flow-rate, which further enhanced confidence in the inflow characterization. Therefore if PTA can be performed in drawdown on a MFHW in unconventional tight formations, it should be relatively easy to achieve on conventional vertical wells. This workflow was applied to over 38 wells in the Bakken to generate a correlation of PV (Pore Volume) and PI ( Productivity Index) as a function of completion and fracture design thereby providing quantifiable feedback on how to drill and complete “winning” wells in a particular geological environment.

As always, economic constraints often stimulate technological creativity, and the author predicts that it will not be long before PTA in drawdown is industrialised and applied to SRP and free flowing wells with the addition of a downhole tools such as a venturis. Can we conclude that, even in a low oil price environment, economic production is possible if production is optimised utilising the right instrumentation and interpretation techniques ?

Biography: 

Lawrence has over 37 years of experience in production operations of which 27 years have been focused on artificial lift  and production optimisation in a variety of roles ranging from field and application engineer to Global Artificial Lift Domain Head for Schlumberger.  He has recently founded his own consultancy business (Camilleri & Associates) focussing on production optimisation.  He has published over 19 SPE conference papers and 5 patents covering all aspects of ESP operation such as inflow characterisation and advanced completions and was SPE Distinguished Lecturer for 2019/2020 on the topic of production optimisation. This body of work is particularly noteworthy as it combines theoretical explanations and field case studies using real time data.

Presented by Distinguished Lecturer:  Roland N. Horne, Thomas Davies Barrow Professor of Earth Sciences at Stanford University

Abstract:

Well monitoring can provide a continuous record of flow rate and pressure, which gives us rich information about the reservoir and makes well data a valuable source for reservoir analysis. Recently, it has been shown that machine learning is a promising tool to interpret well transient data. Such methods can be used to denoise and deconvolve the pressure signal efficiently and recover the full reservoir behavior. The machine learning framework has also been extended to multiwell testing and flow rate reconstruction.

Multiwell data can be formulated into machine learning algorithms using a feature-coefficient-target model. The reservoir model can then be revealed by predicting the pressure corresponding to a simple rate history with the trained model.

Flow rate reconstruction aims at estimating any missing flow rate history by using available pressure history. This is a very useful capability in practical applications in which individual well rates are not recorded continuously. The success of rate reconstruction modeling also illustrates the adaptability of machine learning to different kinds of reservoir modeling, by adjusting features and targets.

Machine learning is also a particularly promising technique for analysis of data from permanent downhole gauges (PDG), given that the massive volumes of data are otherwise hard to interpret using conventional interpretation methodologies.

Biography:

Roland N. Horne is the Thomas Davies Barrow Professor of Earth Sciences at Stanford University, and Professor of Energy Resources Engineering.  He was Chairman of the Department of Petroleum Engineering at Stanford University from 1995 to 2006. 

He is an Honorary Member of SPE, and a member of the US National Academy of Engineering. 

Horne has been awarded the SPE Distinguished Achievement Award for Petroleum Engineering Faculty, the Lester C. Uren Award, and the John Franklin Carl Award.  He is a Fellow of the School of Engineering, University of Tokyo (2016) and also an Honorary Professor of China University of Petroleum – East China (2016).