← Survey overview · Published May 12, 2026

Methodology

Standardized exterior CFD applied to each facility in the cohort.

This survey applies a single standardized exterior CFD methodology to each facility in the cohort. The same domain, boundary conditions, solver, and reporting format are applied at every site so results are directly comparable across firms.

Scope

Each model resolves exterior airflow and thermal-plume behavior around the building under representative wind and operating conditions: rooftop heat-rejection unit discharge, ground-level generator stack and radiator exhaust during emergency operation, and the resulting plume trajectories, recirculation patterns, and intake exposure. Interior thermal management, life-safety systems, and any aspect of as-built operational performance are out of scope.

Domain and geometry

Cylindrical far-field domain with radius approximately 1 kilometer and ceiling approximately 10 building heights above grade. The cylindrical form permits wind-angle sweeps via a single parameter without remeshing. Coordinate convention: +Y true north, +Z up. Building geometry, rooftop equipment placement, and generator-yard layout are derived from public sources only — published floor plans and data sheets, satellite imagery, and publicly disclosed equipment counts. Neighboring buildings within approximately three building heights are included as bluff-body obstructions. No proprietary drawings or specifications are used.

Atmospheric boundary conditions

Inlet velocity follows a logarithmic atmospheric boundary-layer profile with reference height 10 m, von Kármán constant 0.4, and aerodynamic roughness length tuned to local terrain. Side walls use symmetry, the upper boundary is slip, and the outlet is pressure (0 Pa gauge). For low-wind buoyancy-dominated cases, the inlet is reconfigured as a pressure outlet and side walls switch to pressure outlet. Reference pressure is altitude-corrected via the standard atmosphere. Wind cases include the prevailing summer and winter directions, a low-wind case (~2 mph), and a directional sweep of additional cases per building. Ambient dry-bulb and wet-bulb temperatures are set to ASHRAE 0.4% cooling design values from the nearest reference station.

Heat sources

Rooftop heat-rejection units are modeled with paired mass-flow inlet and velocity outlet boundaries. Outlet discharge temperature is set by an intake-temperature-dependent thermal load, with an optional temperature-dependent derate table where the cooling architecture warrants it, and is stabilized via a running-average intake temperature over the most recent 25–50 iterations. Diesel generator stack and radiator outlets are modeled at manufacturer-typical exhaust gas temperatures and mass flows for the inferred unit class, with intake/exhaust pairing analogous to the rooftop equipment. Where specific equipment is not publicly disclosed, industry mid-range values are used and the sensitivity to that assumption is reported.

Solver and mesh

Siemens STAR-CCM+, segregated solver, steady-state RANS. Realizable k-ε turbulence closure as baseline; the coupled solver with k-ω SST is used for selected low-wind, buoyancy-dominated cases where convergence of the baseline configuration is poor. Air modeled as ideal gas (incompressible formulation) with polynomial-in-temperature specific heat and Sutherland's law viscosity. Polyhedral mesh throughout, with 48 ft base size in the far-field domain, 16 ft base in the rooftop zone, and 2-inch local refinement at generator stack outlets. Refinement blocks bound the rooftop and generator yard. Grid-independence studies confirm resolution adequacy at plume centerline and rooftop surface temperatures.

Outputs

Per building: one headline visualization (plume isosurface or surface temperature heat map), three to four supporting visualizations across wind cases, and a one-paragraph summary identifying the dominant plume behavior observed. Quantitative outputs — chiller-inlet temperature distributions, generator-intake re-ingestion temperatures, maximum recirculation values — are not published; they are shared directly with the named MEP firm of record on request.

Limitations

Exterior CFD built on public-source inputs is a screening-level analysis. Results indicate qualitative tendencies and relative comparisons across cases; they do not represent absolute performance of the as-built facility, do not capture real operational scheduling or controls, and are not a substitute for engagement-grade engineering analysis. No claims are made regarding actual building performance, life-safety, or code compliance.