Objectives of the service
Ice-thermal-storage systems in commercial and industrial buildings currently run on static, pre-set schedules loaded once at commissioning. They cannot respond to volatile 15-minute electricity prices or to localised rooftop thermal conditions that dominate chiller efficiency. CoolStorAge closes that gap: an EO-enhanced dynamic optimiser that dispatches the cost-optimal 24-hour operating schedule directly to the customer's Building Management System, cutting cooling-related electricity cost by 20–30%.
Users and their needs
The priority users are owners and operators of large commercial and industrial buildings with installed ice-thermal-storage cooling — shopping centres, hotels, hospitals, office buildings, food-processing plants and data centres. Their operational objective is to maintain comfort or process integrity while minimising cost; today they lack the tools to do that dynamically. The pilot user is a large shopping-centre operator with an installed ice-thermal-storage system, engaged through structured interviews and shadow-mode validation. A secondary user segment is ice-thermal-storage manufacturers and system integrators, for whom CoolStorAge is a sales-support tool that shortens the payback of the underlying hardware investment.
Service/ system concept
CoolStorAge is a cloud-only SaaS with a five-layer architecture: data ingestion (Earth Observation, weather and electricity-market APIs plus the building's own telemetry), a relational data store, a cooling-demand forecast model, a Mixed-Integer Linear Programming optimiser that computes the 24-hour operating schedule, and a secure REST API that delivers the schedule to the customer's on-site controller. The service is hardware-agnostic and deploys over standard industrial protocols (BACnet, Modbus, MQTT). Billing follows a Pay-for-Performance model tied to the measured monthly saving, established against an IPMVP-conformant baseline so the customer can reproduce and audit the figure.
Space Added Value
Standard weather APIs deliver 2-metre air temperature at ~10 km resolution — inadequate for the rooftop conditions that dominate chiller efficiency. CoolStorAge integrates Land Surface Temperature and cloud dynamics from EUMETSAT Meteosat SEVIRI (3 km, 15 minutes), surface energy-balance variables from ECMWF ERA5-Land via Copernicus, and multispectral surface characterisation from Copernicus Sentinel-2 (10 m). The fused EO feed captures rooftop thermal environment, Urban Heat Island effect and solar-gain dynamics at the level of the individual building — features unavailable from any non-space source for historical data. The service is additional step into the energy sector following the ESA BASS activity EOxGrid.
Current Status
The service is at the completion of an ESA Kick-Start activity. The five-layer architecture is implemented and running on cloud infrastructure. All four data-ingestion pipelines — Open-Meteo, Copernicus ERA5-Land, EUMETSAT SEVIRI, Copernicus Sentinel-2 — are operational. The MILP optimiser is implemented in Python/PuLP/CBC. Shadow-mode validation is under way at pilot location. First paying customers are scheduled for the 2027 cooling cycle.
Prime Contractor(s)
