Audun Botterud

Vladimir Dvorkin, Audun Botterud

While power systems research relies on the availability of real-world network datasets, data owners (e.g., system operators) are hesitant to share data due to security and privacy risks. To control these risks, we develop privacy-preserving algorithms for the synthetic generation of optimization and machine learning datasets. Taking a real-world dataset as input, the algorithms output its noisy, synthetic version, which preserves the accuracy of the real data on a specific downstream model or even a large population of those. We control the privacy loss using Laplace and Exponential mechanisms of differential privacy and preserve data accuracy using a post-processing convex optimization. We apply the algorithms to generate synthetic network parameters and wind power data.

Youssef Shaker, Jun Wen Law, Audun Botterud et al.

Policies focused on deep decarbonization of regional economies emphasize electricity sector decarbonization alongside electrification of end-uses. There is growing interest in utilizing hydrogen (H2) produced via electricity to displace fossil fuels in difficult-to-electrify sectors. One such case is heavy-duty vehicles (HDV), which represent a substantial and growing share of transport emissions as light-duty vehicles electrify. Here, we assess the bulk energy system impact of decarbonizing the HDV segment via either H2, or drop-in synthetic liquid fuels produced from H2 and CO2. Our analysis soft-links two modeling approaches: (a) a bottom-up transport demand model producing a variety of final energy demand scenarios for the same service demand and (b) a multi-sectoral capacity expansion model that co-optimizes power, H2 and CO2 supply chains under technological and policy constraints to meet exogenous final energy demands. Through a case study of Western Europe in 2040 under deep decarbonization constraints, we quantify the energy system implications of different levels of H2 and synthetic fuels adoption in the HDV sector under scenarios with and without CO2 sequestration. In the absence of CO2 storage, substitution of liquid fossil fuels in HDVs is essential to meet the deep decarbonization constraint across the modeled power, H2 and transport sectors. Additionally, utilizing H2 HDVs reduces decarbonization costs and fossil liquids demand, but could increase natural gas consumption. While H2 HDV adoption reduces the need for direct air capture (DAC), synthetic fuel adoption increases DAC investments and total system costs. The study highlights the trade-offs across transport decarbonization pathways, and underscores the importance of multi-sectoral consideration in decarbonization studies.

Yuxiao Liu, Bolun Xu, Audun Botterud et al.

Data-driven models analyze power grids under incomplete physical information, and their accuracy has been mostly validated empirically using certain training and testing datasets. This paper explores error bounds for data-driven models under all possible training and testing scenarios, and proposes an evaluation implementation based on Rademacher complexity theory. We answer key questions for data-driven models: how much training data is required to guarantee a certain error bound, and how partial physical knowledge can be utilized to reduce the required amount of data. Our results are crucial for the evaluation and application of data-driven models in power grid analysis. We demonstrate the proposed method by finding generalization error bounds for two applications, i.e. branch flow linearization and external network equivalent under different degrees of physical knowledge. Results identify how the bounds decrease with additional power grid physical knowledge or more training data.