ASHRAE Standards and CAE Thermal Fluid Simulation
Where ASHRAE standards enter an analysis
Overview
When I run HVAC CFD I'm told to “evaluate it per ASHRAE,” but ASHRAE has a huge number of standards. Which ones actually matter?
From an analyst's point of view they serve four purposes: (1) sources for boundary conditions (outdoor design conditions, heat gains, occupancy), (2) acceptance criteria for results (comfort, ventilation, equipment inlet temperatures), (3) validation of software and models, and (4) calibration against measurements. CFD only gives you velocity and temperature fields; the standards supply the yardstick that turns those fields into pass or fail.
Standards commonly used in analysis
| Document | What it specifies | Use in analysis |
|---|---|---|
| Standard 55 | Thermal environmental conditions for occupants (PMV/PPD, operative temperature, local discomfort) | Acceptance criteria for room CFD results |
| Standard 62.1 | Ventilation for acceptable indoor air quality | Outdoor-air boundary conditions and ventilation-effectiveness checks |
| Standard 90.1 | Energy standard for buildings (performance rating method in Appendix G) | Baseline building and modelling rules for energy analysis |
| Standard 140 | Method of test for building energy simulation programs (BESTEST) | Validating the software you use |
| Guideline 14 | Measurement of energy savings, including model calibration criteria | Judging agreement with measured data |
| Handbook—Fundamentals | Psychrometrics, load calculation methods, design climate data | Source for outdoor conditions, properties and heat gains |
Standards are revised regularly, so always record the edition (year) you used in the report. The values on this page are representative figures to illustrate the approach.
Outdoor design conditions
For cooling design, engineers usually take the 0.4%, 1% or 2% annual values from the Handbook climate data, and for heating the 99.6% or 99% values. 0.4% means the temperature is exceeded for only 0.4% of the hours in a year — about 35 hours. The choice changes supply temperatures and loads in the CFD, so always state which exceedance level you used.
Born from the merger of a heating society and a refrigeration society
ASHRAE — the American Society of Heating, Refrigerating and Air-Conditioning Engineers — was formed in 1959 when a society of heating and ventilating engineers founded in 1894 merged with a society of refrigerating engineers founded in 1904. That is why “heating,” “refrigerating” and “air-conditioning” all appear in its name. Its standards still range well beyond building HVAC, from refrigerant safety to data-centre cooling.
Evaluating comfort (Standard 55) with CFD
PMV and PPD
PMV (predicted mean vote) is an index from −3 to +3 computed from four environmental variables — air temperature, mean radiant temperature, air speed and humidity — plus metabolic rate (met) and clothing insulation (clo). The predicted percentage dissatisfied, PPD, follows from it:
| PMV | PPD | Meaning |
|---|---|---|
| 0 | 5.0 % | The minimum — no environment satisfies everyone |
| ±0.5 | 10.2 % | Edge of the usual Standard 55 comfort range |
| ±1.0 | 26.1 % | One in four dissatisfied |
| ±1.5 | 50.9 % | Half dissatisfied |
| ±2.0 | 76.8 % | Clearly uncomfortable |
CFD packages already include the PMV formula. Doesn't it compute automatically once I have temperature and velocity?
It will compute it, but half the inputs come from outside the CFD. Typical values are 1.0–1.2 met for seated office work and about 0.5 clo in summer and 1.0 clo in winter, and changing those alone easily moves PMV by more than 0.5. The input people most often overlook is mean radiant temperature. In CFD without radiation it is effectively set equal to the air temperature, so cold glazing and solar gains disappear. In a room with a lot of glass, leaving out radiation can flip the conclusion.
Operative temperature
Standard 55 expresses its comfort zone in operative temperature. At low air speed (below 0.2 m/s) it can be approximated as the mean of air temperature $t_a$ and mean radiant temperature $\bar{t}_r$:
Local discomfort and evaluation points
- Vertical air temperature difference: a large difference between ankle (0.1 m) and head (1.1 m) height for a seated person is uncomfortable. Standard 55 uses 3 °C as the limit. It is the first thing to check in CFD of underfloor or displacement ventilation.
- Floor surface temperature: about 19–29 °C for people wearing shoes. Used for floor heating and cooling analyses.
- Draught and radiant asymmetry: cold air at the ankles and uneven radiation from cold windows or warm ceilings are also assessed as causes of discomfort.
- Evaluation heights: 0.1, 0.6 and 1.1 m for seated occupants and 0.1, 1.1 and 1.7 m for standing ones. Showing the distribution over the whole occupied zone from the CFD makes the case stronger.
Ventilation, age of air and data centres
Ventilation rate example (Standard 62.1)
The ventilation rate procedure adds a per-person and a per-area outdoor-air rate to get the breathing-zone requirement $V_{bz}$, then divides by the zone air distribution effectiveness $E_z$, which depends on the supply and return layout:
With typical office values $R_p = 2.5$ L/(s·person) and $R_a = 0.3$ L/(s·m²), a 100 m² office for 10 people needs $V_{bz} = 25 + 30 = 55$ L/s (198 m³/h). With a layout where $E_z = 0.8$ — for example warm air supplied and returned at the ceiling in heating mode — the requirement rises to 68.75 L/s (about 248 m³/h).
Checking ventilation effectiveness with age of air
CFD can solve the local mean age of air $\tau$ — the time since the air entered the room — as a scalar with a unit source. Its ratio to the nominal time constant $\tau_n$ (room volume $V$ over airflow $Q$) is an index of ventilation effectiveness:
For the office above (2.7 m ceiling, 270 m³) with 198 m³/h of outdoor air, $\tau_n$ is about 82 minutes. If the CFD gives a mean age of 95 minutes in the breathing zone, $\varepsilon \approx 0.86$, suggesting some outdoor air short-circuits before reaching the occupants. The age-of-air map even shows which supply–return pair is short-circuiting. Seeing what the airflow number alone cannot tell you is exactly where CFD earns its keep.
Data-centre inlet temperatures
ASHRAE Technical Committee TC 9.9 recommends an IT equipment inlet air temperature range of 18–27 °C. Data-centre CFD should check the temperature at the front (inlet) face of each rack against this range, not the room average. With recirculation from the hot aisle, the top servers are often out of range even when the average looks fine.
What matters more than detailed diffuser geometry
Meshing every vane and grille of a ceiling diffuser in room CFD can multiply the model size several times over. Simplifications such as the box and momentum methods proposed by Nielsen and co-workers, which impose only near-diffuser measurements or momentum as boundary conditions, are widely used instead. What matters is checking that the throw and temperature decay a short distance from the diffuser match the manufacturer's data. If they do, the fine geometry has little effect on occupied-zone results.
Calibration and common mistakes
Calibrating to measurements (Guideline 14)
| Data | NMBE | CV(RMSE) |
|---|---|---|
| Monthly | ±5 % | 15 % |
| Hourly | ±10 % | 30 % |
NMBE captures overall bias and CV(RMSE) the scatter. A building energy model can only be called calibrated once it meets these criteria. For CFD, reporting bias and scatter separately for temperatures and velocities at measured points is a convincing way to show how trustworthy the results are.
Common mistakes
| Mistake | Effect | Fix |
|---|---|---|
| Judging comfort from air temperature alone | Radiant effects of glazing and ceilings are missed | Solve radiation and judge on operative temperature and PMV |
| Unit slips between cfm and L/s, ft² and m² | Outdoor-air rates off by around a factor of two | Use SI-edition values or state the conversion (1 cfm ≈ 0.472 L/s) |
| Always assuming distribution effectiveness of 1 | Too little outdoor air, e.g. ceiling supply in heating | Use $E_z$ for the actual layout and confirm with age of air |
| Leaving met and clo at defaults | PMV can shift by more than 0.5 | Choose values for the use and season, and report them |
| Not stating the edition | The basis of criteria can't be traced later | State the standard and year referenced |
Are there tools I can use to get a feel for the numbers?
Try the PMV/PPD thermal comfort simulation for comfort calculations and the cleanroom air-change recovery simulator for outdoor-air rates and contaminant decay. For heat and moisture balances around air handlers use the AHU heat balance simulator and the cooling coil load simulator; for whole-building heat loss, the building heat loss simulator; for data-centre cooling efficiency, the data-centre PUE cooling efficiency simulator. Annual energy analysis is covered in building energy simulation.
Related Topics
Experience the theory firsthand with the interactive simulator for this field
All Simulators