AI in Petroleum Industry

1. Whether AI would do a better job towards - checking material balance – in the context of understanding and interpreting production data – during the analysis of oil fields under primary production?

Could AI better predict flow performance?

2. Whether AI would do better prediction than Muskat’s modified version of Darcy’s law towards predicting fluid flows in response to pressure-gradient – by looking at the problem – possibly @ pore-scale?

3. Whether the way by which an AI assess the data (such as geological data; geophysical/seismic data; log data; core data; PVT data; production data) is going to be significantly different from the way a reservoir engineer does?

Any scope for an AI to come with an improved prediction of future performance and design production?

Could AI compete with the conceptual skill of a reservoir engineer – who generally efficiently manages in understanding the major uncertainties associated with the field that would probably control and optimize the production scenarios under various operating conditions with the available data?

But, if the total volume of oil-bearing rock exceeds 10 giga m3 (with depth of oil column itself exceeding, say, 100 m; and with temperatures exceeding 100 degrees Cs and pressures exceeding 15,000 psi), then, do we have no other choice than depending on AI?

4. Feasible for AI to conceptualize the reservoir to have two different fundamental entities by incorporating coupled fluid and solid dynamics?

Whether AI would do miracles in deducing the ‘continuous pathways’ as well as ‘the least resistive pathways’ for hydrocarbon production in a complex petroleum reservoir?

Whether AI would be able to provide more insights on the very concept of ‘effective porosity’?

AI could precisely deduce the dip and depths of tilted oil-water and gas-oil contacts?

5. Whether AI would do wonders in ascertaining the ‘presence of an oil field’ in the absence of ‘a well being drilled’ and ‘oil (with significant flow rate) being produced’?

6. What would be the role of AI in estimating STOIIP?

AI would remain very sensitive towards deducing the normally distributed reservoir porosity?

How exactly AI will be able to segregate ‘oil saturation’ from that of ‘hydrocarbon pore volume’?

Any new technique in deducing ‘oil formation volume factor’ by AI?

And, by what means AI will be able to deduce precisely the gross volume of rock with reference to ‘areal-extent and thickness of the oil-field’?

7. Whether AI could do wonders in recovering more oil, while preventing the preferential production of gas?

How efficiently AI would help in maintaining the reservoir pressure above bubble point, while keeping an enhanced driving force to keep the oil flowing;

and

how efficiently AI would be able to direct the injected water/gas   that would be able to displace the oil from the pore spaces of the reservoir leading to increased oil recoveries?

More Suresh Kumar Govindarajan's questions See All
Similar questions and discussions