Adequacy of time-series reduction for renewable energy systems

TL;DR

This paper evaluates how different methods of reducing time-series data impact the accuracy of renewable energy system models, focusing on loss of load and system costs, and highlights the importance of implementation strategies.

Contribution

It provides a systematic assessment of time-series reduction methods for renewable energy models, emphasizing the influence of implementation details on model accuracy.

Findings

Implementation as chronological sequences with re-scaled time-steps minimizes loss of load.

Grouped periods require more computational resources due to additional variables.

Adequacy of reduction methods depends on time-series length and implementation approach.

Abstract

To reduce computational complexity, macro-energy system models commonly implement reduced time-series data. For renewable energy systems dependent on seasonal storage and characterized by intermittent renewables, like wind and solar, adequacy of time-series reduction is in question. Using a capacity expansion model, we evaluate different methods for creating and implementing reduced time-series regarding loss of load and system costs. Results show that adequacy greatly depends on the length of the reduced time-series and how it is implemented into the model. Implementation as a chronological sequence with re-scaled time-steps prevents loss of load best but imposes a positive bias on seasonal storage resulting in an overestimation of system costs. Compared to chronological sequences, grouped periods require more time so solve for the same number of time-steps, because the approach requires additional variables and constraints. Overall, results suggest further efforts to improve time-series reduction and other methods for reducing computational complexity.