Infiltration and Natural Ventilation Model for Whole-Building Energy Simulation of Residential Buildings: Preprint

The infiltration term in the building energy balance equation is one of the least understood and most difficult to model. For many residential buildings, which have an energy performance dominated by the envelope, it can be one of the most important terms. There are numerous airflow models; however, these are not combined with whole-building energy simulation programs that are in common use in North America. This paper describes a simple multizone nodal airflow model integrated with the SUNREL whole-building energy simulation program. The required inputs for infiltration are taken from blower door test results and the geometry of the openings for natural ventilation. The flow exponents and coefficients for infiltration and natural ventilation can be input or left to the default values. Control of the natural ventilation openings can be controlled with a time scheduled and the indoor/outdoor temperature difference. The mass-flow rate equations are written in terms of the pressure at the base of each zone. The pressure on each surface is a combination of stack and wind effects added to the zone base pressure. The resulting set of mass balance equations are solved using a Newton-Raphson iterative method with a variable relaxation coefficient. The relaxation coefficient is adjusted each iteration depending on the speed of the convergence. The iterations are stopped when the mass balance in each zone converges to a specified tolerance. The model exhibits good numerical behavior, with no singularities and only a few instances of nonconvergence in some building simulations with no thermal mass and large leakage areas. The infiltration model compares favorably to the LBL infiltration model. The natural ventilation model also compares well with another natural ventilation model and with measured results. The current infiltration model and whole-building simulation program form the basis of a new residential energy-auditing tool. Introduction This paper presents a multizone infiltration/natural ventilation model that is simple, flexible, and driven by weather conditions. This model was added to the SERI-RES version 1.0 (SERI-RES 1988) whole-building energy simulation program as part of work to upgrade the program to a new version called SUNREL (Deru et al. 2002). The model is verified with analytic calculations and compared with other infiltration and natural ventilation models. There have been numerous building airflow models developed; a survey by Feustel and Dieris (1992) reviewed 50 models. The most recognized single-zone model is the LBL model (Sherman and Grimsrud 1980, Sherman and Modera 1986), which has been extensively verified with measured data. The most widely recognized multizone models in the United States are CONTAM (Dols 2001; Dols and Walton 2002) and COMIS (Feustel and Rayner-Hooson 1990, Feustel 1999). Both of these programs were designed to be used as stand-alone programs; however, these programs can also be combined with building energy simulation programs. Kendrick (1993) reviewed such coupling strategies and surveyed nine programs. COMIS has recently been used in combination with a whole-building energy simulation program (Huang et al. 1999). Another stand-alone model that has been integrated into a whole-building energy simulation program is MIX (Li et al. 2000). Most whole-building energy simulation programs have very simplified infiltration models because it is usually a small load in large buildings. However, for small buildings, infiltration can be one of the most important terms in the energy balance equation, and it is one of the most difficult to model accurately. One of the difficulties in simulating infiltration is relating physical measurements of the leakage to a mathematical model. Ideally, the size, location, and characteristics of all leakage areas in the building should be known; however, this information is impossible to quantify. The most common measurement technique for infiltration in small buildings is the blower door test. This method is simple and results in an approximation of the effective leakage area (ELA), but it yields very little information concerning the geometry or spatial distribution of the cracks. Another major obstacle is modeling the pressure distribution