Issue overview

The attached models (single-family home with a crawlspace and an attic) use an airflow network to model the impact of HVAC system distribution leaks. The HVAC system uses a cycling fan to meet the loads, and ventilation air is not provided by the system. The only difference between the cases is the effective leakage ratio input specified in the AirflowNetwork:Distribution:Component:LeakageRatio object, which is used to model air leakage from the crawlspace to the return side of the distribution system. The "baseline2" case assumes virtually no leakage, while the "baseline3" case assumes a high level of leakage.

Simulation results indicate that the high leakage case ("baseline3") uses less heating energy than the no-leakage case when simulated in Boise, ID. This outcome appears counterintuitive, if not incorrect.
The simulation output further reveals that, in the high leakage case ("baseline3"), the infiltration ACH for the only conditioned zone in the model is significantly lower than in the no-leakage case. To some degree, this reduction is expected, as introducing air from the crawlspace into the zone increases the pressure in the zone, thereby reducing infiltration. However, the observed difference in infiltration between the two cases seems disproportionately large and not well correlated with the fan operation (low or no fan operation should show infiltration rate that are similar).

In the figure below, the zone infiltration flow rates for the two models are shown on the top panel—blue represents the case with no leakage, while orange represents the case with leakage. The bottom panel illustrates the conditioned zone's average pressure (PLR-weighted average of zone pressure during on/off cycling operations)—dark green corresponds to the case with no leakage, and light green corresponds to the case with leakage. The chart also displays the total zone pressure for the high leakage case. For the no-leakage case, the average pressure is equal to the total pressure (as shown in the last figure below).

Focusing on the simulation period highlighted within the black rectangle, we observe that the average zone pressures in the two cases are nearly identical. Hence, one would expect the infiltration flow rates to be similar; however, as shown in the top panel, the flow rates differ substantially.

If we plot the infiltration sensible losses during that time, we can observe the same trend: when the average zone pressures are nearly identical, the case with no leakage exhibits higher sensible losses compared to the case with high leakage.

It seems that the infiltration ACH/sensible losses are more closely correlated with the total zone pressure rather than the average node pressure.

Operating System (Multiple choices)

MacOS

Operating System Version

15.7.2

Version of EnergyPlus

25.2

Unmethours link or helpdesk ticket number

N/A

Defect file

baseline2.idf.zip

baseline3.idf.zip