Data democratization: developing open source data and models

Designing future clean energy systems and the best* pathways to achieve them requires access to high fidelity data and models, with high spatial and temporal granularity. Extensive effort has been made towards understanding our current energy system and modeling future changes: these may include climate and weather models for wind and solar potential across the world, techno-economic and socio-economic models for technology adoption and market proliferation, energy and electricity demand growth models in different customer segments, and synthetic power networks to model electric grids. However, a significant gap still remains, particularly in power systems where real data of network topologies, utility models, and customer data are not widely available due to data privacy concerns and system security. Real energy data is difficult to find, typically not publicly available, and lacks standardization. Further, as we look towards data-driven methods including AI and ML for prediction and real-time operations, representative datasets are needed to support algorithm development and offline testing if we have any hopes to transfer these technologies from research labs to the field. The creation of synthetic datasets with realistic electric grids accompanied with representative load profiles and DER adoption patterns requires considerable effort, but is a valuable and necessary task. 

* In the multidisciplinary world of energy systems, "best" can be measured along multiple axes, with different weights to each factor depending on the application. Generally, best can include cost, carbon reduction or abatement, equitable access, timeline to technology deployment, among other metrics.