The objective of this work was to understand the suitability and favorability of saline reservoirs for storing hydrogen and ensuring the maximum amount of hydrogen would be withdrawn. We used the Sacramento basin in California to demonstrate the feasibility. We carried out several numerical simulation studies to understand key factors affecting the storage and withdrawal of hydrogen. We combined the results from the numerical modeling to develop a screening and ranking set of criteria for hydrogen storage in saline reservoirs. We then used the screening and ranking set of criteria to rank the formations in the Sacramento basin. We studied five formations in the Sacramento Basin. The numerical simulation study showed that to optimize storage and withdrawal of hydrogen, steeply dipping reservoirs up to 15 degrees, reservoirs with low pressures, reservoirs with good porosity (above 20%), and reservoirs with high permeabilities were most favorable for underground hydrogen storage. This work applies a novel and comprehensive site screening and selection criteria for underground hydrogen storage in saline reservoirs. It builds on a similar work done for carbon dioxide storage and hydrogen storage in depleted gas fields but it considers additional objective of needing to withdraw the stored fluid. The case study in Sacramento Basin can be applied to any other basin.

Geothermal well log analyses consist of utilizing pressure and temperature data measured along the wellbore to predict feed zones, reservoir temperature, and reservoir pressure. Interpreting geothermal production logs can be subjective and require great expertise to achieve repeatability. In situations where there may be several log data, interpreting the logs may be time consuming for quick decision-making processes. This work discusses the implementation of a multi-layered deep learning convolutional neural network to automatically diagnose sets of temperature and pressure well logs. The algorithm achieves results similar to those of a professional engineer. This algorithm enables the interpretation of many well logs in just a few seconds. Data input for this project is synthetic well data that mimics real data. 10,000 datasets were used. The data was split as follows: 9,800 and 200 for training and validation respectively. The algorithm used takes as input three “depth-series” logs of temperature, pressure, and temperature gradient, passes the data through a convolutional neural network including a flat layer and then a fully connected layer with five output variables which are the depths of the feed zones, the reservoir temperature, the reservoir pressure and the depth at which the reservoir pressure is known. The cost function for this model was the mean squared error. The optimizer algorithm used was Adam, and the learning rate had an exponential decay. The algorithm recorded the model state that had the lowest mean absolute validation error. The architecture was implemented in Keras with a TensorFlow backend. The best model found during the process of hyper-parameter tuning was used to predict the reservoir characteristics for the validation and testing data sets. The results show a good match between the predicted data and actual data with a training error of 0.9%, validation error of 2%, and test error of 7%. Future works will involve adding more real data to the training and validation set, increasing the number of feedzones that can be identified, and performing sequential analysis using interdisciplinary data.