Efficient proactive transmission planning to accommodate renewables

There is a growing need for tools to help decision makers to proactively plan for transmission infrastructure to accommodate renewables under gross market and regulatory uncertainties. In this paper, we make three contributions. First, we discuss how the current approaches aiming to proactively plan for transmission to accommodate renewables in the US are mathematically inaccurate, particularly with regards to their treatment of uncertainty. Second, improving existing models, we develop a two-stage stochastic network-planning model that takes into account Kirchhoff's laws, uncertainties, generators' response, and recourse investment decisions. Third, for large-scale networks, we demonstrate the use of Benders decomposition, taking advantage of the block-structure of the constraints. Testing our model on a simplified representation of California, we show that there are costs of ignoring uncertainty and that trying to identify robust solutions from a series of deterministic solutions is not necessarily effective, and indeed could result in higher costs than ignoring uncertainty altogether.