Year: 2026
Type: article
Abstract: A battery used to smooth solar generation may have less reserve available than its rated power suggests because firming, recovery and reserve share the same stored energy and converter. We examine this interaction using telemetry from a solar plant in a small island developing state, publicly reported as 122 MWp and 92 MWac, over 25 November 2025 to 1 February 2026. A one-minute model propagates state of charge continuously, and each ten-minute reserve activation is assessed separately from its chronological starting state. In the original 56-date analysis, the largest observed ten-minute decline was 74.117 MW. For a 55 MW/15 MWh analytical battery under an idealised zero-latency sampled-information benchmark, the reserve award deliverable in at least 99% of evaluable starts fell from 52.38 MW in reserve-only operation to 21.52 MW after firming and recovery, and to 7.21 MW with actual point-of-interconnection feedback. Enforcing an internal energy-rate ceiling of twice nominal capacity per hour reduced the latter two awards to 2.25 and 0.99 MW; lower tested ceilings failed the firming criterion. Restoring eight additionally usable dates preserved the service-architecture ordering. The results establish the importance of operating envelopes, recovery and information timing in reserve assessment. They are conditional engineering-bound estimates, not demonstrated field-controller capability or battery-health qualification.
Source: Discover Electronics
Author: Curtis Boodoo
Institution: University of Trinidad and Tobago
Language: English
Subfield: Energy Engineering and Power Technology
Field: Energy
Domain: Physical Sciences
Open Access status: diamond
OpenAlex ID: W7225425580