Air Infiltration and Compliance Modeling

Author: William Allen, PhD

Energy use for heating and cooling in residential buildings is largely driven by a need to keep the internal temperature either higher (in winter) or lower (in summer) than the outside temperature. It would appear obvious, therefore, that reducing air infiltration by improving the air-tightness of the envelope is a good thing[1]. Reducing the rate of exchange of conditioned indoor air with unconditioned outdoor air should reduce heating and cooling loads, and thus improve the energy efficiency of the building. This article looks at how air leaks are treated in the compliance software, and how changing the leak rates can affect modeled energy use.

CBECC-Res is the free, CEC approved, compliance software that is most commonly used in California for compliance modeling.  Inside CBECC is another piece of software, the California Simulation Engine (CSE). The CSE is the modeling ‘engine’ carries out the calculations for energy use and building efficiency, based on the inputs made to CBECC. Many of the properties of the model which are fixed in CBECC (e.g. the thermostat set points – these cannot be changed by the user), can be changed in the CSE. For compliance purposes, the default air infiltration rate is set to 5ACH50 for the standard design. The distribution of the leaks is laid out in the Alternative Compliance Manual[2]:


Model description

To best isolate the effects of air leakage, the building modeled was chosen to be as simple as possible. Starting with the CBECC example file for a 2100sqft single family home, the model was edited to remove the garage to reduce the number of zones between which air can flow. This allows the effect of changing a particular airflow path to be better understood.  In order to isolate the effect of moving the air and thermal barrier between the roof and the ceiling, the model used either R-38 at the ceiling and R-0 at the roof, or R-0 at the ceiling and R-38 at the roof. The model was not changed to make it comply – the aim of this article is to make broad comparisons, and not specific recommendations for a particular house of climate zone.


Leak area and ACH value

When the attic is modeled in CBECC-Res as a vented attic, the leak area is split between the ceiling and the walls as specified in Table 1. The leak area between the conditioned zone and the attic is equal to the total leak area through the four walls. The attic is modeled with a leak area (more accurately described as a ventilation area) given by the ‘1/600’ rule of thumb, where the ventilation area is one six-hundredth of the floor area. The total leakage area (excluding the attic ventilation) is primarily determined by the ACH50 value entered in the CBECC-Res model. In the CSE, the leak area scales with the ACH50. If the ACH value is doubled, the leak area doubles in the CSE input file. The relative size of the leak areas does not change, so the percentages given in Table 1 are still adhered to. For the house modeled here, the leak areas are summarized in Table 2 for a vented attic with 2ACH50, and 5ACH50. The leak areas are 2.5 times as large in the 5ACH50 case as in the 2ACH50 case, while the attic ventilation area remains unchanged.


Vented vs unvented attic

To model the attic as unvented in CBECC-Res it is only necessary to select ‘Unventilated’ from the dropdown list on the attic properties page in the model (as shown in Figure 1).


Figure 1


The effect of this change on the CSE input file is to move the ceiling leaks (which were previously at the thermal and air barrier between the conditioned and unconditioned spaces) to the roof plane, which is now the thermal and air barrier. The ceiling plane between the living space and the attic is no longer an air barrier and is assigned leak areas ten times those of the attic leaks. The effect of this change is shown in Table 3



Quality Installation of Insulation (QII)

Insulation construction quality can be modeled in CBECC-Res as either ‘Standard’ or ‘Improved’. Selecting ‘Improved’ triggers the inspection requirements for QII and will mean the software uses the full R-values for the insulation used. If ‘Standard’ is selected, the performance of the insulation in the model is downgraded according to the rules shown in Table 4[3]

Table 4

Selecting ‘Improved’ insulation has no effect on the performance of continuous sheathing and so does not affect air leakage rates.


Energy use impacts of air sealing

The effect on energy use and compliance margins of air sealing is highly dependent on climate zone: the effect of outdoor air leaking into the building will depend on the temperature difference between outside and inside – if the indoor temperature is 76°F, a cubic foot of outdoor air at 96°F will have twice the effect as a cubic foot of air at 86°F. However, there will be times when the outdoor air temperature is closer to the desired indoor temperature than the indoor air is. At these times, the leaks will be beneficial – similar to the effect of a whole house fan, but on a much smaller scale. The impact of sealing will also vary depending on where the leaks are, and how the building is insulated. To look at how these effects interact with each other, we ran multiple models of the building with a number of changes:

  1. The ACH50 was varied between 0.5 and 7
  2. The attic was modeled as ventilated or unventilated
  3. The roof and ceiling insulation were swapped – one model has an R-38 ceiling and an uninsulated roof, the other has an uninsulated ceiling and an R-38 roof (with the insulation below the roof deck)
  4. The house was modeled in climate zone 7 and climate zone 12


The effect of altering the ACH is similar in all cases:

Figure 2: R-38 roof, R-0 ceiling, CZ12

As the ACH value is reduced from 7 (very leaky) to 0.5 (extremely tight), the energy use initially drops before reaching a low near 2 or 3 ACH. In some cases, the energy use increases as the envelope is further tightened, most likely as a result of losing the leakage at beneficial times of day.

Figure 3: R-0 roof, R-38 ceiling, CZ12

Not surprisingly, the location of the air barrier and the insulation go hand in hand. In Figure 2, with the insulation at the roof, the unvented attic performs better, whereas in Figure 3, with insulation at the ceiling, the vented attic performs better. The performance difference is not the same in the two cases: with an insulated roof, the unvented attic uses 4-5% less energy, but with an insulated ceiling the vented attic uses 2% less.

In the milder climate zone 7, the situation is similar, except that the unvented attic performs better whether the insulation is at the ceiling or at the roof.

Figure 4: R-38 roof, R-0 ceiling, CZ7


Figure 5: R-0 roof, R-38 ceiling, CZ7


Air sealing with spray foam insulation

One of the benefits of closed cell spray foam insulation is the additional air sealing it provides. If a house is built using an unvented attic insulated with spray foam, and a conventionally insulated wall construction, balance between the leaks in the attic and the leaks through the walls will change from the 50/50 balance used by CBECC-Res. While the reduction in the overall leak area can be approximated by lowering the ACH value entered into CBECC, this will not change the ratio of wall to roof leakage. To model that, it is necessary to change the leak areas in the California Search Engine, the modeling software that runs inside CBECC-Res. To gauge the effect of eliminating air leakage in the attic, the models with an unvented attic and insulation at the roof were re-run for climate zone 12 with all the attic leaks removed. This undoubtedly reduces the overall leakage but will reflect the effect of using spray foam insulation in the attic of an otherwise typically sealed building. Figure 6 shows the effect of sealing the attic leaks in the model. The sealed attic is significantly more efficient than the unsealed attic, even in buildings with low overall air leakage.


Cost considerations

Air sealing has three cost considerations: how much it costs to install, how cost effective it is for compliance, and how much it will save on utilities.

The National Renewable Energy Laboratory has published costs for sealing to different levels of air-tightness, based on surveys with builders and contractors. Using these values, we can see how much it costs to tighten the envelope to different levels, shown in Table 5 for the 2100sqft house modeled here

Table 5

For this building in climate zone 12, with a vented attic and an R-38 ceiling, we can directly compare the energy use when the envelope is tightened from 5ACH50 to 3ACH50, which is the most likely change in air-tightness. At 5ACH, the building would use 4939kWh and 352 therms a year. At 3 ACH, it would use 4933kWh and 349 therms. Using costs of 15c/kWh and $1/therm, this would yield annual savings of about $5.

From these numbers, it does not seem that air sealing is a cost-effective way to reduce utility bills. However, it should be looked at as part of a whole building approach: a tighter shell will allow for better control of indoor air quality, modest reductions in energy use can lead to corresponding savings on PV sizing for ZNE or near ZNE buildings, and comfort levels are improved due to smaller changes in temperature (and possibly elimination of draughts). The costs and benefits of air sealing should be considered as a part of a whole-building approach to energy efficiency and comfort


[1] Changing envelope tightness and air infiltration rates will impact indoor air quality. The impact can be positive (if the desire is to keep external contaminants out), or negative (if the desire is to remove internal contaminants. These impacts, and ventilation strategies to address them, are not the topic of this article.

[2] 2016 Residential ACM Reference Manual, Table 3

[3] 2016 Residential ACM Reference Manual, Table 4