Sizing of cooling systems is a crucial part of Passive House design, especially given a warming climate and a greater incidence of heatwaves. Undersized cooling equipment in Passive House homes in Aotearoa New Zealand is becoming an issue. With this in mind, we undertook extensive research that assessed the applicability of cooling load numbers in PHPP. The results of this research were presented at the PHINZ conference in Wellington in 2024. In this article we attempt to draw out the implications of this research for Passive House designers in Aotearoa New Zealand.
PHPP by default calculates a single zone model focussed on a 24-hour average cooling load, suitable for very well shaded buildings with high thermal mass and where cooling is supplied for ALL 24 hours of the design day. This doesn’t match how we typically build and cool in NZ. PHPP does not give you the peak cooling load, which is what HVAC engineers would typically calculate using a multi-zone dynamic energy model to size equipment.
Our calculations indicate air conditioners sized according to the PHPP cooling load number with default certification settings, have only 10-27% of the capacity needed to meet the overall building cooling load, depending on the climate zone and building typology*.
For many building typologies, we found that PHPP is useful for calculating overall building cooling load if its settings are adjusted. We’ll walk you through which settings and how. However in most cases the overall building cooling load value is not adequate for selecting cooling equipment on a room-by-room basis, especially when there is a lot of solar gain at the beginning and end of the day.
Our research was designed to find an adequate process for sizing cooling loads. We ran four types of models and compared them to determine how best to use PHPP for this purpose in our design practice. The first model was PHPP with default certification settings. We then tested whether adjusted PHPP cooling settings made for more useful results by checking them against an equivalent single zone dynamic thermal model. Lastly we ran multi-zone dynamic thermal models, which is industry best practice for sizing cooling equipment and routinely used to size HVAC systems in commercial buildings. We modelled three different typologies in three New Zealand climate zones.
What projects are at risk?
Passive House designers can usefully apply a checklist to each project to determine a good approach for sizing of cooling equipment. Note that as certifiers and Passive House designers, Sustainable Engineering Ltd’s team always recommends specifying heat pumps in residential homes. The efficiency of heat pumps and their ability to provide both heating and cooling makes their specification a no-brainer.
Our research suggests PHPP cooling sizing will be sufficiently reliable for residential buildings if the design has:
- a low window-to-floor ratio
- very little east/west glazing
- good summer shading to north windows and
- no unusual peaks in internal loads**.
Even in these scenarios, we still recommend you evaluate whether the default certification settings are appropriate for your design. In most cases the results will be improved by making some adjustments to PHPP’s cooling load settings. Consider project-specific factors such as the amount of thermal mass and the approach taken to cooling. Are you relying on consistent operation of mechanical cooling or is air-conditioning only a top-up on the hottest days of the year?
Factors to consider
It’s important to understand your client and their preferences at the beginning of the design process. What is their experience of what is too hot or too cold? How wide is their band of acceptable indoor temperature? How much time do they spend at home: do they work at home or does a long commute to an office mean the house is unoccupied for 8-10 hours every week day? Lastly, and critically, to what extent are clients happy to be involved in managing internal temperatures; will they night-purge heat and manually open windows to cross-ventilate at appropriate times?
How to adjust PHPP settings
Our adjusted settings are based on the methodology previously presented by Oliver Style, CEO at Praxis Resilient buildings.
Settings to consider adjusting include the following.
- Increase the exterior dry-bulb and dewpoint temperature to match the peak instead of average value.
- Occupancy: is the default occupancy a useful estimate of reality.
- Internal heat gains: equipment and activity
- Cooling setpoint: is 25°C inside appropriate for your client?
- Soiling factor (ie dirty window glass) and temporary shading.
Across all the simulations we ran, adjusting the PHPP cooling load setting better matches industry standard predictions of performance and will help ensure cooling equipment is appropriately sized.
Single zone models
It’s important to understand that PHPP is a single-zone model that assumes a single temperature throughout the building. Its cooling load results are an average of the condition of the entire building as if all the air inside the building were mixed to a single temperature at all times. Clearly this does not match the reality of a family home or apartment. Multi-zone modelling (as was done for the dynamic thermal modelling), on the other hand includes the internal layout and treats each room as a separate zone. For the purpose of comparison with PHPP, we created a single zone dynamic thermal model by removing all the internal walls. We wanted to investigate how closely the cooling load in the single zone dynamic model matched that of the PHPP file with adjusted settings.
Fig 2
The graph above has the single zone dynamic model set to 100%. It compares this to the results from PHPP with default certification settings and PHPP with adjusted settings, across all simulations. The bars represent the average across all results, and the error lines indicate the minimum and maximum results. As can be seen, the certification settings massively underestimated the cooling load; even with adjustment, the results still varied between 75% and 116%. The Figure below outlines how the adjusted PHPP compares to the single zone dynamic model for the three climate zones modelled.
Fig 3
Multi-zone models
Multi-zone dynamic modelling produces the most useful cooling load results. This is particularly so when whole-house averages disguise significant variations in the solar gain profile of different rooms. Expect this to happen if glazing on east and west faces is not shaded, allowing low angle sun to penetrate deep into the building at the beginning and end of the day
Fig 4
The example above compares single-zone and multi-zone dynamic models. This simulation is the large single-family residence in Wellington. It shows how the overall building peak (from the single-zone model) occurs at 6:50pm, whereas the timing of the peak cooling load in the multi-zone model varies from 9am to 7pm for different rooms.
This zonal analysis shows the limitations of single zone calculations and the averaging that happens across the building. The single zone model is not representative of all rooms. Consideration needs to be given to what rooms are used at which times of the day.
Understanding zonal cooling loads can help designers make decisions about the floor plan, giving consideration to which rooms are occupied at what times of the day and which are most important.
Fig 5
The graph above shows the same simulation, a large single-family home in Wellington. It shows the degree of variation of cooling load intensity on a room-by-room basis and compares it to the other modelling methods that use a single zone. As you can see, it is very difficult to predict cooling loads unless multi-zone dynamic thermal cooling modelling is used.
The inadequacy of results using the PHPP certification settings is very evident. Adjusted settings in PHPP brings the results closer to the results of single-zone dynamic thermal modelling but the limitations of averaged results is also apparent.
The cooling load for the worst-effected bedroom specifically is >120W/m². Using PHPP certification settings to specify cooling equipment would result in a grossly under-sized cooling in this room. Understanding the zonal dynamics would assist informed decisions about cooling system design, not just the size of equipment. Is a multi-split or ducted system called for? Where are the heads located? Would investing in automated external shading considerably reduce reliance on mechanical cooling?
When dynamic thermal modelling is called for
Early in the PHPP modelling, consider using an adjusted PHPP to produce an overall cooling load number you can use to indicate size and design cooling equipment.
Recognise the design characteristics and climates that raise red flags. These factors below, individually and especially in combination, suggest specialist dynamic thermal modelling is recommended.
- Unshaded glazing to the east and west
- large areas of glazing relative to wall area and
- occupant patterns of use that will increase internal heat gains (working from home, lots of visitors staying for extended periods, significant computing equipment or other plug load, regular use of the oven even during summer etc).
If one of our team is providing consulting support on your design, we will flag when we think multi-zone dynamic thermal modelling is needed. You’re also most welcome to call us if you have any concerns. Dynamic thermal modelling for a residential project starts at about $2K. That’s much less costly than dissatisfied clients and retrofitting cooling equipment down the track.
Appendix – How to adjust PHPP
The following screenshots show where to find the relevant worksheets and cells in PHPP if you wish to adjust the settings.
*Note that this underestimation is unique to cooling. The calculation method in PHPP is very suitable for estimating heating loads. But when summer sun penetrates deep into the house through east or west facing doors and windows, the solar gain immediately translates to an increase in cooling load, an impact not captured by PHPP’s calculations.
**This list does not apply in the case of non-residential buildings that have wildly fluctuating occupancy rates—schools and offices are obvious examples.