Hong KongReal estate2020 – ongoing

Wharf REIC

Machine-learning optimization of Harbour City chiller sequencing

Wharf REIC reported 61,410 kWh saved in a Harbour City chiller pilot

WorkingCompany's own claimOfficial ↗

Problem

Harbour City operated chiller plants with manual controls. It wanted to reduce electricity consumption at Gateway II and Ocean Terminal Extension.

Approach

Harbour City tested machine learning and cloud-based data analysis in a retro-commissioning pilot. The analysis compared chiller sequences to identify efficient combinations for operation. Wharf REIC reported energy savings and different performance improvements at the 2 locations.

Results

61,410 kWhtotal energy saved by the pilot, according to Wharf REICCompany's own claim[1]

16.6%performance improvement at Gateway II, according to Wharf REICCompany's own claim[1]

3.0%performance improvement at Ocean Terminal Extension, according to Wharf REICCompany's own claim[1]

Takeaway

The same optimization project produced different gains at separate chiller plants.

Sources

Sustainability Report 2020
Wharf Real Estate Investment Company Limited21 May 2021Company publication

Source summaries

[1]Wharf Real Estate Investment Company Limited21 May 2021

Wharf REIC described a machine-learning pilot for Harbour City's chiller controls. It reported 61,410 kWh saved, with performance improvements of 16.6% at Gateway II and 3.0% at Ocean Terminal Extension.

Full details are in the original sources.

Added 1 Nov 2026 · Updated 1 Nov 2026