Broad ranges of investment configurations for renewable power systems, robust to cost uncertainty and near-optimality

Achieving ambitious CO2 emission reduction targets requires energy system planning to accommodate societal preferences, such as transmission reinforce-ments or onshore wind parks, and acknowledge uncertainties in technology cost projections among many other uncertainties. Current models often solely minimize costs using a single set of cost projections. Here, we apply multi -objec-tive optimization techniques in a fully renewable European electricity system to explore trade-offs between system costs and technology deployment for elec-tricity generation, storage, and transport. We identify ranges of cost-efficient capacity expansion plans incorporating future technology cost uncertainties. For example, we find that some grid reinforcement, long-term storage, and large wind capacities are important to keep costs within 8% of least-cost solutions. Near the cost optimum a technologically diverse spectrum of options exist, allow-ing policymakers to make trade-offs regarding unpopular infrastructure. Our analysis comprises 50,000+ optimization runs, managed efficiently through multi-fidelity surrogate modeling techniques using sparse polynomial chaos expansions and low-discrepancy sampling.