At COP 4.6 with 24 MW of chiller power the cooling load would be 110.4 MW, so the condenser would reject 110.4 + 24 = 134.4 MW against 130.5 MW entering the campus — 3.9 MW more heat than the site receives, with no other heat source. For 4.6 to be possible at all, chiller power must be at most 23.30 MW.
Where the 9 MW of pumps and towers sits changes the answer: pump motor heat enters the fluid it pumps, fan motor heat goes to air. The reconcilable range is 4.06 (tight boundary) to 4.44 (all 9 MW into the loops). Neither reaches 4.6.
Base case runs on the stated 4.6 as directed by the organisers; the alternative runs on the implied COP. We show both rather than asserting either.
Industry-standard consensus for centrifugal water-cooled chillers is roughly 1.5–2.0% saved per °F of chilled-water-supply reset, equivalent to about 2.7–3.6% per K — our 2.5%/K assumption sits conservatively inside that band.
BuildingIQ, "HVAC Savings Tip: Chilled Water Supply Temperature Reset." · Eng-Tips professional engineering forum, consensus reply: "1.5% reduction in energy per degree increase in CHWS temp."
Meta's data centre in Odense, Denmark recovers server waste heat through a heat-recovery system that upgrades it to district-heating temperature, supplying roughly 11,000 homes. A comparable heat-pump-on-chilled-water-return retrofit in Hohhot, China cut the district network's electricity draw by about 10% a year.
Munters Data Center Technologies, case study, 2026. · "Towards a Systematic Survey for Carbon Neutral Data Centers," arXiv:2110.09284.
The only public capital benchmarks for data-centre-to-district heat export networks are Danish and use comparable transmission distances (Apple's Viborg facility, 12 km). No Gulf-region installation of this scale and duct type exists yet to confirm the figure, which is exactly why the model tests it hard rather than accepting it.
"Impacts of flexible-cooling and waste-heat recovery from data centres on energy systems: A Danish case study," ScienceDirect, 2023.