Prior art — US Patent US20180240021A1

Analysis of Prior Art for US Patent Application US20180240021A1

A review of the citations for U.S. Patent Application Publication No. US20180240021A1 (granted as US11087221B2) reveals several key prior art references that were considered during its examination. These references relate to the use of computational models, simulations, and data analysis for predicting and optimizing hydrocarbon production from reservoirs. Below is an analysis of the most relevant citations and their potential impact on the claims of the '021 application.

Key Prior Art References and Potential Anticipation of Claims:

The core of the invention in US20180240021A1 lies in its method of using a machine-learning-based "classification model" to generate a "probabilistic estimate" of a target well's performance, which then informs the decision to drill the well. The independent claims (1, 8, and 15) broadly cover this method, the system that performs it, and a storage device containing the instructions for it. The analysis below focuses on how prior inventions might anticipate these core concepts.

1. US Patent No. 9,043,188 B2 - "System and method for forecasting production from a hydrocarbon reservoir"

Full Citation: US Patent No. 9,043,188 B2, "System and method for forecasting production from a hydrocarbon reservoir," assigned to Chevron U.S.A. Inc.
Filing Date: September 1, 2006
Brief Description: This patent describes a method for forecasting hydrocarbon production by creating a "proxy model" from a limited number of detailed reservoir simulations. This proxy model, which can be a response surface or a neural network, is then used to rapidly estimate production outcomes for various operational scenarios without running full, time-consuming simulations for each case. The system can be used to optimize field development plans.
Potential Anticipation of Claims:
- Claim 1 (Method) & Claim 8 (System): The '188 patent discloses a method and system that align closely with the initial steps of the '021 application's claims. It involves receiving well and reservoir parameters ("operational scenarios"), forming a predictive model ("proxy model" or "neural network"), and using it to estimate production rates. The '188 patent's "proxy model" can be seen as analogous to the "classification model" in the '021 application. While the '188 patent focuses on "forecasting" and the '021 application on a "probabilistic estimate" for "classification" (good/bad), the underlying process of using a computationally derived model to predict well performance is substantially similar. The distinction may lie in the specific nature of the output (a probabilistic classification vs. a production forecast), but a strong argument for anticipation or at least obviousness could be made.
- Claim 15 (Data Storage Device): The teachings of the '188 patent would inherently anticipate a data storage device containing instructions to carry out its described method, thus overlapping with the scope of claim 15.

2. US Patent No. 9,910,938 B2 - "Shale gas production forecasting"

Full Citation: US Patent No. 9,910,938 B2, "Shale gas production forecasting," assigned to Schlumberger Technology Corporation.
Filing Date: June 20, 2012
Brief Description: This invention details a system for forecasting shale gas production by generating a large number of "realizations" (possible geological models) and running simulations on them. It then uses techniques like principal component analysis and clustering to analyze the simulation results and identify patterns that correlate with production outcomes. This allows for a probabilistic forecast of production.
Potential Anticipation of Claims:
- Claim 1 (Method) & Claim 8 (System): The '938 patent describes a workflow that includes generating simulation results and then using statistical analysis and clustering (an unsupervised learning method mentioned in the '021 application's abstract) to understand and predict well performance. This process of forming a data-driven model from simulation results to make a "probabilistic forecast" is a core element of claims 1 and 8. The use of clustering to group similar outcomes is particularly relevant to the "classification" aspect of the '021 application.
- Claim 15 (Data Storage Device): Similar to the '188 patent, the system described in the '938 patent would be implemented via software on a storage device, making claim 15 potentially anticipated.

3. US Patent No. 10,138,717 B1 - "Predicting well performance with feature similarity"

Full Citation: US Patent No. 10,138,717 B1, "Predicting well performance with feature similarity," assigned to Novi Labs, LLC.
Filing Date: January 7, 2014
Brief Description: This patent discloses a method for predicting the performance of a new, un-drilled well by identifying "analogous" or similar existing wells from a database. It calculates a similarity score based on various features (geological, completion, etc.) between the proposed well and the existing wells. The performance of the most similar existing wells is then used to predict the performance of the new well. This is framed as a machine learning approach.
Potential Anticipation of Claims:
- Claim 1 (Method) & Claim 8 (System): This reference is highly relevant. It explicitly describes a machine learning-based system for predicting the performance of a "target well" before it is formed. The process involves receiving parameters for the target well, using a model based on "feature similarity" (analogous to the '021 application's classification model built from training data), and generating a prediction. The '717 patent's approach is a form of pattern recognition, which is a key element of the '021 application. While it may not explicitly use the term "probabilistic estimate" in the same way, the use of multiple analogous wells to generate a prediction inherently contains a probabilistic element. This patent presents a strong challenge to the novelty of the core method and system claims.
- Claim 15 (Data Storage Device): The '717 patent's method is designed for computer implementation, therefore anticipating the substance of claim 15.