The SPE Europe Energy GeoHackathon will kick off with boot camps beginning 21 October.
This year, the challenge will be to use a data set from Croatia. The goal is to use machine learning to predict reservoir properties for two new wells using well and seismic data from six other wells. This information will then be the basis of a geothermal field development plan, with the goal of generating electricity for a town of 50,000 people.
The event is organized by volunteers from SPE sections in Europe and from international technical sections moved by the desire to make the energy transition happen in a "datafied" and sustainable way.
Four weeks of bootcamp sessions on data science and geothermal energy will begin on 21 October. Data scientists and industry experts will deliver weekly sessions to equip participants with the relevant technical knowledge to understand the challenge and be ready for the GeoHackathon.
The boot camp sessions are
- Croatia: the geothermal energy front: the past, present, and future
- Introduction to the challenge: data review, handling and interpretation
- Geothermal reservoirs: geology basics
- Geothermal reservoirs: deciphering the rocks through petrophysics
- Leapfrog: building a conceptual model
- Geothermal energy: heat-flow quantification
- Geothermal field development plan: the approach
- Geothermal energy: project economics
The hackathon is driven by the idea that data-driven prediction and machine-learning classification of reservoir properties from seismic data and well logs can enable solutions for
- Creating computationally efficient workflows to interpret a large volume of data obtained from seismic campaigns
- Identifying new features and facies using unbiassed artificial intelligence workflows contributing to geothermal field developments
- Speeding up the interpretation of seismic data
Part 1 of the hackathon, which will be 60% of the event, will focus on data analytics tasks. It will involve reviewing well and seismic data and initializing the neural net. Data from the six wells will be used to train the model. That the neural net has been correctly trained will be confirmed before it is then used to predict the reservoir properties for two new wells.
Part 2, the remaining 40%, will be spent creating a geothermal development plan. Using the data gathered earlier, plus cost and heat data, a plan will be created to develop a geothermal reservoir for power generation. The plan will consider
- How many wells need to be drilled and determine what the production is?
- How much electricity can be generated based on predicted production rates?
- How much does it cost and is it economic?
Participants completing all the boot camps sessions and the hackathon challenge will be awarded with a digital certificate of training by SPE.
Find out more and sign up here.