The Issue
In this case study, we are looking at the optimisation of a new school canteen concerning natural ventilation located in Singapore. The study aimed to provide the design team with knowledge, which would help to inform design decisions, including answering the following questions:
How can solar heat gain and glare be reduced and the facade optimised for enhanced visual comfort?
How will different types of openings in or below the roof help to increase the airflow and reduce the operative temperature in the occupied zone?
What is the required wind speed or elevated airspeed needed for occupants to feel thermally comfortable in the occupied zone?
Our Approach
For this study, we used IES VE to conduct various simulations for daylighting, thermal comfort and air flow. The following steps outline our approach:
1. Building a simplified model in SketchUp and establishing different design iterations to be studied. The main space of concern was divided into two separate rooms, with a connecting void, to get the temperatures in the occupant zone and the ceiling zone.
2. Import the geometry into IES VE and assign loads and constructions. Generic load schedules were assigned to represent the worst-case scenario.
3. For the air flow and thermal comfort simulations different window opening types and profiles were assigned in IES VE MacroFlo.
4. For the daylighting simulations material properties were assigned in IES VE Radiance.
5. The results were retrieved from IES VE and analysed, refer next section: Our Findings.
Our Findings
The daylighting results were analysed first. The daylighting results would help to guide the design in terms of visual and thermal comfort, to avoid excessive direct sun exposure, which is a source of glare and thermal discomfort. The tables below show an overview of the initial cases, Base and Case 1. Case 2 and Case 3 were added later as the design developed further.
Fig. Simulation cases overview
The initial daylighting and glare results suggested that Case 1 would provide a better and more even daylight distribution while reducing the glare probability. For Case 1 (closed windows), the excessive sunlight drops to 0% and the daylight level is lowered 4-fold to about 500 lux, which is just nice for a school canteen.
Fig. Initial daylighting and glare studies
The operative temperatures in the occupied zone and the air temperatures in the ceiling zone were then analysed. The operative temperature results for the occupied space showed that the temperature will rise with up to 2 degrees above the outdoor temperature. (Similarly is seen in a kampung house with little thermal mass, refer Gregers explanation in the following video).
Fig. Operative temperature and thermal comfort in the occupied zone
ASHRAE 55 and the Adaptive Method for Natural Ventilated buildings were then used to evaluate the operative temperatures. The operative temperature acceptability limit was found based on the prevailing outdoor temperature (28.7°C) in Singapore to establish the upper limits for 80% acceptability without elevated air speed (0.3m/s) and with elevated air speed (0.6m/s, 0.9m/s and 1.2m/s) using fans.
The figure shows that the lowest operative temperatures in the occupied zone can be achieved by implementing the design in cases 1 and 2, with case 2 performing slightly better due to openings placed at a higher location.
Fig. Annual average and maximum operative temperatures
Case 3 was later introduced with an updated roof design, including a longer jack roof. The below figure shows how the different designs will perform on a day with little to average wind speeds. The figure shows that the ventilation performance and air changes during the occupied time mainly rely on the buoyancy effect due to little surrounding wind. Larger openings located higher in the design give optimal natural ventilation and increased ventilation.
Fig. Comparison of Air Changes in Occupied Zone vs Internal Gain and Wind Speed
A closer look at the annual average operative temperature difference between the jack roof with closed top windows and a few/all windows open showed that having all the windows open would be more effective in lowering the operative temperature.
Fig. Annual average temperature difference ΔT (°C)
As shown in the figures and graphs an elevated air speed would be required to meet the 80% acceptability limits for Operative Temperature as per ASHRAE 55. An alternative calculation approach for compliance with the Singaporean Green Mark thermal comfort criteria was also carried out. Green Mark gives a formula for Predicted Mean Vote as follows:
PMV = a + b x DBT + c x WIND
Where the inputs for schools are as follows:
Value of a: -6.805
Value of b: 0.267
Value of c: -0.87
DBT Baseline: Indoor air temperature (°C), 31°C (Schools)
WIND: Indoor wind velocity (m/s)
By applying the wind speed of 0.3m/s, 0.6m/s, 0.9m/s and 1.2m/s it is seen that a wind speed of at least 1.2m/s is required for the occupant’s thermal sensation to be within PMV -0.5 to 0.5, categorized as a thermal sensation of Neutral.
Fig. Green Mark Thermal Comfort PMV estimates and required wind speed for naturally ventilated schools
Our Recommendations
To increase the usability of the roof for future PV installations and to increase the natural ventilation possibilities within the school canteen our recommendations were as follows:
Provide a “clean” roof which will be optimal for future PV installations
Increase the amount of openable top hung windows in the facade (below the roof for a bigger high difference between the inlet and outlet) or jack roof
Install solid or translucent panels above the occupied zone to reduce visual and thermal discomfort in the early morning and late afternoon.
Ceiling fans will be required for achieving thermal comfort. The ceiling fans shall be located near the occupied zone, to avoid pulling potentially hot air from below the ceiling into the occupied zone. The ceiling fans shall be able to provide an elevated air speed of 1.2m/s or more.
Please visit our Resource page about Natural Ventilation for rules of thumb and thermal comfort calculators, which can be used in the initial/early design phases.