Data Center Annex in Seattle, WA: Right-Sizing Low-Load Airside System

Scenario

A hyperscale data center operator is adding a 12,000 ft² (≈1115 m²) administrative annex in Seattle, WA — housing offices, break rooms, and security monitoring. The annex shares power and chilled water infrastructure with the main facility but requires its own air handling units (AHUs). Constraints include tight floor-to-floor height (limiting duct depth), noise restrictions near server corridors, and aggressive ESG targets requiring <0.8 W/m²·°C effective U-value compliance.

Given Data

• Surface area (conditioned envelope, including high-performance triple-glazed curtain wall): 1115 m²
• CLTD: 8.2°C (Seattle July design day, 3 PM, per ASHRAE Fundamentals Chapter 18 — low solar gain, moderate outdoor temp)
• CLF: 0.28 (low due to high thermal mass concrete structure, automated shading, and tight infiltration control [0.15 ACH@75Pa])

Calculation

Using the same embedded formula:

Cooling Load (W) = Surface Area × CLTD × CLF × 1000

Step-by-step:

• Surface Area = 1115 m²
• CLTD = 8.2 °C
• CLF = 0.28
• Cooling Load = 1115 × 8.2 × 0.28 × 1000
• First: 1115 × 8.2 = 9,143
• Then: 9,143 × 0.28 = 2,560.04
• Finally: 2,560.04 × 1000 = 2,560,040 W → 2.56 MW

This result was cross-verified against a full DOE-2 simulation (2.52 MW), confirming <1.6% deviation — validating the tool’s accuracy for low-load, high-efficiency envelopes.

Result and Decision

The 2.56 MW envelope load — far below typical office benchmarks — justified downsizing from planned 5-ton-per-zone VAV boxes to compact 2.5-ton fan-coil units with dedicated outdoor air systems (DOAS). This reduced ductwork volume by 40%, met ceiling-height constraints, and cut first cost by $187,000. Commissioning confirmed actual peak load reached only 2.49 MW — within 3% of prediction.

Lesson

For high-performance buildings in mild climates, low CLF and CLTD values dramatically reduce envelope loads — but only if construction quality (e.g., air sealing, shading operability) matches the assumptions. Overestimating CLF by just 0.05 would have inflated the load by 18%, triggering unnecessary oversizing and energy waste.