Inertia-aware Unit Commitment and Remuneration Methods for Decarbonized Power System

Abstract

To maintain frequency stability in decarbonized power systems, inertia services from synchronous generators (SGs) and inverter-based resources must be procured. However, designing an inertia-aware system operation poses significant challenges in considering the variability and uncertainty of renewable energy sources (RES) and adopting a remuneration method for inertia provision due to SG commitment variables. To address this research gap, we renovate the inertia-aware chance constraints unit commitment model by incorporating time-coupling constraints for SGs and joint chance constraints for RES uncertainty. We investigate remuneration methods for inertia provision, including uplift, marginal pricing (MP), approximated convex hull pricing (aCHP), and average incremental cost pricing (AIP), applying these to the renovated model. Numerical experiments show that the model enhances frequency stability during a contingency. Among the remuneration methods, only aCHP guarantees revenue adequacy without requiring uplift while maximizing economic welfare. However, the MP requires the highest level of uplift to adequately compensate generation costs, as the price function fails to account for inertia provision.

Figures (11)

• Figure 1: Illustration of the modified 118-bus network from 7904729
• Figure 2: System inertia levels provided by all generation resources (SGs and RES) under different scenarios and renewable penetration levels ($\eta$ = 20–40%). The dashed line denotes the inertia requirement.
• Figure 3: System frequency responses following a generation outage at different times (01:00 and 08:00) and renewable penetration levels ($\eta$ = 20–40%). Each panel compares the Base case without inertia and cases with inertia provision under MP, aCHP, and AIP. RoCoF and frequency nadir values are annotated for each case.
• Figure 4: Number of online SGs and their generation outputs under different scenarios and renewable penetration levels ($\eta$ = 20–40%). Each subplot shows the 24-hour variation of online SGs (top) and the corresponding dispatch levels (bottom). SG indices are sorted in ascending order of marginal cost.
• Figure 5: Box plots of 24-hour energy, reserve, and inertia prices under MP, aCHP, and AIP schemes for different renewable penetration levels ($\eta$ = 20–40%). Each box represents the 25th–75th percentile range with the median line inside.