Utilization of the industrial revolution 4.0 in oil and gas exploration

From AAPG Wiki
Jump to navigation Jump to search
Wiki Write-Off Entry
GeoWikiLogo.jpg
Author Najoud AlOtaibi and Fayrouz AlEssa
Affiliation Saudi Aramco
Competition 2021 Middle East Wiki Write Off

Accelerating the detection of Oil and Gas supplies is critical for development and growth in countries. Therefore, Oil and Gas companies continually search and explore for new hydrocarbon deposits by collecting vast amounts of subsurface data obtained from well and seismic. The upstream sector is also called exploration and production, which is the first stage of the lifecycle of the industry. The cycle begins with exploration for oil and gas deposits; this is followed by the appraisal stage which is the investigation of the volumetrics of these reserves, then the development stage that is the installation of rigs and drilling the well; finally, the production stage which is the extraction of the oil and gas from the subsurface. In this paper, we aim to conduct an extensive literature review on the adoption and utalization of the fourth industrial revolution in the Oil and Gas sector specifically in the exploration and appraisal stage.

Why use AI?[edit]

The nature of the business in the oil and gas industry deals with a wide array of uncertainty in many aspects. From exploration, field development, and future outlook, we exhibit huge ambiguity and risks in each phase. Therefore, employing data-driven analytics and machine learning algorithms in this field has many advantages in minimizing the associated risks and making better investment decisions. This paper will holistically explore the applications of the fourth industrial revolution on the oil and gas industry and identify further improvement opportunities.

First Stage: Exploration[edit]

The early stages of the exploration phase are composed of the integration of surface and subsurface knowledge, which includes the utilization of well and seismic data that tell us about the subsurface structure, which in turn, we can infer about the reservoir characteristics. Geoscientists use immense data that they have at their disposal to guide engineers on where to drill next. The following are a specific application of AI/ML in oil and gas exploration:

Anomalies Detections[edit]

One of the most important applications of Big Data in exploration is used to analyze subsurface structure, specifically the well and seismic data which are huge datasets that are numerous and acquired over tens of thousands of kilometers square. In parallel as the industry grows, the volume of the data increases. Mohammadpoor and Torabi[1], Desai et al.[2], and Qi et al.[3] draw on an extensive range of sources to assess the use of Big Data, prominently in exploration it is used to extract the underlying trend of the seismic data and well data. The exploration stage requires deep knowledge of the surface geology and subsurface structure, nowadays as a result of years of exploring there are immense data in that domain. Hence, the use of Artificial Intelligence can accelerate the process of exploration. In practice, exploring oil and gas using geological and geophysical data is to detect anomalies in the data and this might indicate the presence of hydrocarbons. Machine learning (ML) is utilized in that arena to accelerate the detection of these anomalies and classify huge data.[4] [5] [6] [7] Hange et al.[8] highlights that ML can exploit the existing data to come up with the results that will add value to the industry.

Well log Correlation[edit]

In exploration, data acquired from well data play a pivotal role in characterizing the reservoirs. In practice, geoscientists or engineers are manually interpreting and describing well data, nowadays with the ML this can be achieved simultaneously, hence, accelerating the exploration phase. Maniar et al.[9] showcased the application of supervised machine learning in the detection of the geological tops signature in well data. However, unsupervised learning is still a challenge to apply in the industry. Other machine learning approaches are used in well log-based reservoir characterization to develop predictive models for water saturation, which is an important variable for the production engineers.[10]

Structure Interpretation and Facies Detection[edit]

Characterizing the fracture of reservoir formations is important in determining the flow of oil and gas and eventually the productivity of wells. Therefore, geoscientists delineate the fracture connectivity, orientation and location. El Ouahed et al.[11] applied Artificial Neural Network and Fuzzy Logic to build a 2D fracture intensity map of an oil field in Algeria. Many case studies have emerged from the application of artificial intelligence for structure interpretation. For instance, Obafemi et al.[12] have shown the use of Unsupervised Artificial Neural Network (UANN) in characterizing reservoirs in Deep-water Niger Delta. In addition, from 3D seismic data, one of the main attributes is to extract the seismic facies and this is used to infer about the geological depositional environment, which is mainly used for reservoir characterization. Liu et al.[13] demonstrated seismic facies classification using supervised convolutional neural network (CNN) and a semi-supervised generative adversarial network (GAN). The CNN method is a deep learning algorithm that can take an input image and classify it based on its features (edges, corners, etc) used in mature fields where the well data are immense, whereas the GAN is similar to CNN but used when the well data is limited in new prospects (new fields).[13] [14] [15] [16] [17] In addition, Wang et al.[18] and Cite error: Invalid <ref> tag; invalid names, e.g. too many utilized ML to delineate and predict permeability of tight-sand (unconventional) reservoirs in China, their study has demonstrated its usefulness in the exploration phase and well placement.

Second Stage: Appraisal[edit]

This is the most tedious stage in the oil and gas upstream lifecycle as it includes the proposal and generation of new plays. After the Geophysicists in the exploration team have proposed the first well to be drilled in the area of interest and the discovery of oil and gas has been confirmed. It will then be followed by the appraisal wells, but, as both Di et al.[19] and Ji et al.[20] agree, in order to achieve that goal, first, a 3D seismic survey that has been acquired in the new field must be interpreted. During the appraisal stage, more wells are drilled to collect more real-time data, information and samples from the reservoir[21] in addition to the acquisition of seismic surveys. All previous maps will be updated with the new seismic data and the velocity control available from the newly drilled wells. Based on the revised depth structure maps and seismic attributes that relate to reservoir quality, a specific number of appraisal wells will be placed to determine the extent of the hydrocarbon accumulation and characterize the distribution of the reservoir to better delineate the reservoir and decrease uncertainties to know how to develop the field most efficiently. All will be used in reservoir development studies and will help conclude if pursuing this area is of economic benefit to the company.

Estimation of hydrocarbon in place[edit]

The volumetric calculation is an integral part of the appraisal stage of the field development as it attempts to estimate the field hydrocarbon in place in addition to an insight into the reservoir rock mechanics, which can be achieved by running reservoir simulation after assigning rock properties and taking into account rock fluid flow. The results will impact the hydrocarbon and natural gas production as well as the CO2 storage. Weinzierl et al.[22] applied deep neural networks (DNN) to invert and extract pressure, saturation from poroelastic rock properties as well as rock physics properties. An augmented deep learning cognitive (DL) rock facies classification based on well logs were proposed by Sidahmed et al.[23] as an effort to identify sweetspots.

Reservoir fluid properties[edit]

PVT (reservoir fluid properties) has a significant role in different stages of the hydrocarbon cycle and it is the very first indicator of hydrocarbons presence in the reservoir provided by the mud gas data. For decades, the estimation of reservoir gas and oil properties from mud gas data has been a dream for many engineers and geoscientists as it gives information about the reservoir fluid and rock properties at an early stage with a low cost and continuous in all reservoirs. Yang et al.[24] used an algorithm to train machine learning models on PVT data at specific depth which then applied to Advanced Mud Gas data to predict reservoir fluid properties, specifically gas oil ratio (GOR). Hashemi Fath et al.[25] used multilayer perceptron (MLP) and radial basis function (RBF) for predicting the Gas-Oil Ratio. In addition, Hashemi Fath et al.[25] developed an ANN model for prediction of bubble point pressure. Similarly, Baarimah et al.[26] developed a technique to predict reservoir fluid properties of crude oil systems using artificial intelligence. On the other hand, Cao et al.[27] used the ELM approach to assist in determining rock properties such as porosity and permeability. Also, Zhao et al.[28] integrated SCAL data through machine learning to estimate the relative permeability. Logs make engineers and geoscientists’ lives easier when they contain rock properties data in them as interpreting these data can facilitate in knowing whether the zone being penetrated is hydrocarbon bearing or not.

Reservoir extends[edit]

Estimating the extent of the reservoir is critical before moving to the development stage. As in that stage, it will be determined if the field is worth developing or not and if it is of an economic benefit to the company and can produce a commercial volume. In the earlier stage prior to obtaining core samples at the well location, it is extremely challenging for geoscientists, specifically petrophysicists to attain rock properties such as porosity and permeability. Cao et al.[27] predicted the characteristics of the reservoir parameter distribution from well log data by training Neural Networks and robust Extreme Learning Machine algorithm, which is widely used in the scientific fields to solve classification problems. In the analysis, they have used 4950 samples of 15 wells with core data that yielded an accurate prediction. For the remaining wells, the network model is then implemented to estimate reservoir porosity and permeability.

Well planning and design[edit]

Before drilling wells, a well plan and design are conducted to study the formation properties and set casing design at high-pressure zones. Rashidi et al.,[29] Hussain et al.,[30] and Anemangely et al.[31] (2019) have used advanced ML algorithms to predict the shear properties of the rocks using well data. The results of this prediction according to the authors can be utilized in the well planning process to avoid any well collapse or damage during drilling. Okoli et al.[32] used supervised machine learning algorithms to estimate the downhole vibration on surface drilling parameters to avoid poor drilling accuracy, failures, and non-productive time from wear and tear in the drilling equipment. In addition, Noshi and Schubert[33] did an overview of the role of machine learning in drilling operations that addresses potential drilling problems and the different application of each of the different types of machine learning algorithms which are, supervised learning, unsupervised learning, semi-supervised learning, and reinforcement learning. Ma et al.[34] showcased the importance of utilizing augmented AI to understand the different impact of well spacing optimization on well production performance.

See also[edit]


References[edit]

  1. ↑ Mohammadpoor, M., & Torabi, F. (2019). Big Data analytics in oil and gas industry: An emerging trend. Petroleum, August 2018. https://doi.org/10.1016/j.petlm.2018.11.001
  2. ↑ Desai, J. N., Pandian, S., & Vij, R. K. (2020). Big data analytics in upstream oil and gas industries for sustainable exploration and development: A review. Environmental Technology and Innovation, xxxx, 101186. https://doi.org/10.1016/j.eti.2020.101186
  3. ↑ Qi, J., Ha, T., Lyu, B., Lubo-Robles, D., & Marfurt, K. (2020). Seismic attribute selection for machine learning based facies analysis. SEG International Exposition and Annual Meeting 2019, 1853–1857. https://doi.org/10.1190/segam2019-3216007.1
  4. ↑ Lary, D. J., Alavi, A. H., Gandomi, A. H., & Walker, A. L. (2016). Machine learning in geosciences and remote sensing. Geoscience Frontiers, 7(1), 3–10. https://doi.org/10.1016/j.gsf.2015.07.003
  5. ↑ Karpatne, A., Ebert-Uphoff, I., Ravela, S., Babaie, H. A., & Kumar, V. (2018). Machine learning for the geosciences: Challenges and opportunities. ArXiv, 31(8), 1544–1554
  6. ↑ Gharbi, R. B. C., & Mansoori, G. A. (2005). An introduction to artificial intelligence applications in petroleum exploration and production. Journal of Petroleum Science and Engineering, 49(3–4), 93–96. https://doi.org/10.1016/j.petrol.2005.09.001
  7. ↑ Islam, M. S. ul. (2020). Using deep learning based methods to classify salt bodies in seismic images. Journal of Applied Geophysics, 178, 104054. https://doi.org/10.1016/j.jappgeo.2020.104054
  8. ↑ Hanga, K. M., & Kovalchuk, Y. (2019). Machine learning and multi-agent systems in oil and gas industry applications: A survey. Computer Science Review, 34, 100191. https://doi.org/10.1016/j.cosrev.2019.08.002
  9. ↑ Maniar, H., Ryali, S., Kulkarni, M. S., & Abubakar, A. (2019). Machine learning methods in Geoscience. 2018 SEG International Exposition and Annual Meeting, SEG 2018, 4638–4642. https://doi.org/10.1190/segam2018-2997218.1
  10. ↑ Miah, M. I., Zendehboudi, S., & Ahmed, S. (2020). Log data-driven model and feature ranking for water saturation prediction using machine learning approach. Journal of Petroleum Science and Engineering, 194(April), 107291. https://doi.org/10.1016/j.petrol.2020.107291
  11. ↑ 1. El Ouahed, A. K., Tiab, D., & Mazouzi, A. (2005). Application of artificial intelligence to characterize naturally fractured zones in Hassi Messaoud Oil Field, Algeria. Journal of Petroleum Science and Engineering, 49(3–4), 122–141. https://doi.org/10.1016/j.petrol.2005.05.003
  12. ↑ Obafemi, S., Oyedele, K. F., Omeru, T., & Bankole, S. I. (2020). Characterization of deep water turbidite channels and submarine fan lobes using artificial intelligence; Case study of Frem Field deep offshore Niger Delta. Journal of African Earth Sciences, 167(April), 103852. https://doi.org/10.1016/j.jafrearsci.2020.103852
  13. ↑ 13.0 13.1 Liu, B., Li, X., Zheng, G., Ren, Y., Zhang, S., Liu, Y., Gao, L., Liu, Y., Zhang, B., & Wang, F. (2020). Deep-learning-based information mining from ocean remote-sensing imagery. National Science Review, 7(10), 1584–1605. https://doi.org/10.1093/nsr/nwaa047
  14. ↑ Dramsch, J. S. (2020). 70 Years of Machine Learning in Geoscience in Review. Advances in Geophysics, 61(January), 1–55. https://doi.org/10.1016/bs.agph.2020.08.002
  15. ↑ Bergen, K. J., Johnson, P. A., De Hoop, M. V., & Beroza, G. C. (2019). Machine learning for data-driven discovery in solid Earth geoscience. Science, 363(6433). https://doi.org/10.1126/science.aau0323
  16. ↑ Tontiwachwuthikul, P. (PT), Chan, C. W., Zeng, F. (Bill), Liang, Z. (Henry), Sema, T., & Chao, M. (2020). Recent progress and new developments of applications of artificial intelligence (AI), knowledge-based systems (KBS), and Machine Learning (ML) in the petroleum industry. Petroleum, xxxx, 0–1. https://doi.org/10.1016/j.petlm.2020.08.001
  17. ↑ Di, H., Li, C., Smith, S., & Abubakar, A. (2019). Machine learning-assisted seismic interpretation with geologic constraints. SEG International Exposition and Annual Meeting 2019, 5360–5364. https://doi.org/10.1190/segam2019-w4-01.1
  18. ↑ Wang, Z., Gao, D., Lei, X., Wang, D., & Gao, J. (2020). Machine learning-based seismic spectral attribute analysis to delineate a tight-sand reservoir in the Sulige gas field of central Ordos Basin, western China. Marine and Petroleum Geology, 113(October 2019), 104136. https://doi.org/10.1016/j.marpetgeo.2019.104136
  19. ↑ Di, H., Li, Z., Maniar, H., & Abubakar, A. (2020). Seismic stratigraphy interpretation by deep convolutional neural networks: A semisupervised workflow. Geophysics, 85(4), WA77–WA86. https://doi.org/10.1190/geo2019-0433.1
  20. ↑ Ji, X., BenHasan, N., Luo, Y., Gashawbeza, E., & Saleh, S. M. (2020). Recognition of salt zones in 3D seismic images using machine learning. 1626–1630. https://doi.org/10.1190/segam2020-3426105.1
  21. ↑ Ben, Y., Han, W., James, C., & Cao, D. (2020). Building a general and sustainable machine learning solution in a real-time drilling system. SPE/IADC Drilling Conference, Proceedings, 2020-March. https://doi.org/10.2118/199603-ms
  22. ↑ Weinzierl, W., & Wiese, B. (2020). Deep Learning a Poro-Elastic Rock Physics Model for Pressure and Saturation Discrimination. Geophysics, 1–55. https://doi.org/10.1190/geo2020-0049.1
  23. ↑ Sidahmed, M., Roy, A., & Sayed, A. (2017). Streamline rock facies classification with deep learning cognitive process. Proceedings - SPE Annual Technical Conference and Exhibition, 0. https://doi.org/10.2118/187436-ms
  24. ↑ Yang, T., Arief, I. H., Niemann, M., Houbiers, M., Meisingset, K. K., Martins, A., & Froelich, L. (2019). A machine learning approach to predict the gas oil ratio based on advanced mud gas data. Society of Petroleum Engineers - SPE Europec Featured at 81st EAGE Conference and Exhibition 2019. https://doi.org/10.2118/195459-ms
  25. ↑ 25.0 25.1 Hashemi Fath, A., Madanifar, F., & Abbasi, M. (2020). Implementation of multilayer perceptron (MLP) and radial basis function (RBF) neural networks to predict solution gas-oil ratio of crude oil systems. Petroleum, 6(1), 80–91. https://doi.org/10.1016/j.petlm.2018.12.002
  26. ↑ 1. Baarimah, S. O., Abdulraheem, A., & Anifowose, F. A. (2014). Artificial Intelligence (AI) techniques for predicting the reservoir fluid properties of crude-oil systems. Society of Petroleum Engineers - International Petroleum Technology Conference 2014, IPTC 2014 - Innovation and Collaboration: Keys to Affordable Energy, 5(October 2015), 3953–3968.
  27. ↑ 27.0 27.1 Cao, Q., Banerjee, R., Gupta, S., Li, J., Zhou, W., & Jeyachandra, B. (2015). Data driven production forecasting using machine learning. Society of Petroleum Engineers - SPE Argentina Exploration and Production of Unconventional Resources Symposium. https://doi.org/10.2118/180984-ms
  28. ↑ Zhao, B., Ratnakar, R., Dindoruk, B., & Mohanty, K. (2020). A Hybrid Approach for the Prediction of Relative Permeability Using Machine Learning of Experimental and Numerical Proxy SCAL Data. SPE Journal, 25(5), 2749–2764. https://doi.org/10.2118/196022-PA
  29. ↑ Rashidi, S., Mohamadian, N., Ghorbani, H., Wood, D. A., Shahbazi, K., & Alvar, M. A. (2020). Shear modulus prediction of embedded pressurized salt layers and pinpointing zones at risk of casing collapse in oil and gas wells. Journal of Applied Geophysics, 183, 104205. https://doi.org/10.1016/j.jappgeo.2020.104205
  30. ↑ Hussain, S. A., Chatterjee, C., Sarkar, S. K., Reyes, A., Majumdar, C., & Das, R. (2019). Predicting horizontal shear slowness- A machine learning approach. Society of Petroleum Engineers - Abu Dhabi International Petroleum Exhibition and Conference 2019, ADIP 2019, 1–12. https://doi.org/10.2118/197128-ms
  31. ↑ Anemangely, M., Ramezanzadeh, A., Amiri, H., & Hoseinpour, S. A. (2019). Machine learning technique for the prediction of shear wave velocity using petrophysical logs. Journal of Petroleum Science and Engineering, 174(November 2018), 306–327. https://doi.org/10.1016/j.petrol.2018.11.032
  32. ↑ Okoli, P., Vega, J. C., & Shor, R. (2019). Estimating downhole vibration via machine learning techniques using only surface drilling parameters. SPE Western Regional Meeting Proceedings, 2019. https://doi.org/10.2118/195334-ms
  33. ↑ Noshi, C. I., & Schubert, J. J. (2018). The role of machine learning in drilling operations; a review. SPE Eastern Regional Meeting, 2018-October(October), 7–11. https://doi.org/10.2118/191823-18erm-ms
  34. ↑ Ma, Z., Davani, E., Ma, X., Lee, H., Arslan, I., Zhai, X., Darabi, H., & Castineira, D. (2020). Finding a trend out of chaos, a machine learning approach for well spacing optimization. Proceedings - SPE Annual Technical Conference and Exhibition, 2020-Octob. https://doi.org/10.2118/201698-ms

gas-oil ratio

Retrieved from "https://wiki.aapg.org/index.php?title=Utilization_of_the_industrial_revolution_4.0_in_oil_and_gas_exploration&oldid=27937"