Long-cycle storages aligned with two-dimensional clustered unit commitment

Two improvements are proposed to the two-dimensional clustered unit commitment to incorporate long-cycle storages. One is a long-cycle storage decoupling method that approaches the annual storage operation in discrete typical days while retaining both seasonal and short-term dispatch of them. The other is an integer -relaxed clustering method that overcomes the deficient representativeness of load and renewable profiles in selecting typical periods which is caused by the participation of long-cycle storages in dispatch. An annual -horizon hourly-resolution unit commitment is hence developed that preserves the mixed-integer linear format of unit commitment and provides a framework for annual scheduling of long-cycle storages.

Automated summary

Two improvements are proposed to incorporate long-cycle storages in the two-dimensional clustered unit commitment. The first improvement is a long-cycle storage decoupling method that models the annual storage operation using discrete typical days while considering both seasonal and short-term dispatch. The second improvement is an integer-relaxed clustering method that addresses the deficiency in representing the load and renewable profiles caused by the involvement of long-cycle storages in dispatch. Consequently, an annual-horizon hourly-resolution unit commitment model is developed, which maintains the mixed-integer linear format and provides a framework for scheduling long-cycle storages on an annual basis.