Introduction to the design of natural ventilation systems using loop equations.

The design of natural, including passive, ventilation systems assumes one of two genericforms: the nasty design problem where the designer seeks to size ventilation openings givenclimatic conditions and thermal comfort criteria or the nice design problem where the designerseeks to size ventilation openings given climatic conditions, indoor temperature distributions,and specified airflow rates - presumably determined from separate thermal or air qualityconsiderations. The nasty form of design demands consideration of the complex dynamiccoupled interaction of a building's airflow systems, thermal characteristics and airflow andthermal excitations - a challenge that only the most advanced simulation programs have beenable to address and one were few, if any, can claim real expertise at this time. The nice formof ventilation design, on the other hand, is quite tractable and may be approached usingexisting, relatively simple and intuitively direct theory. Yet it is commonly perceived to be ofa nasty character demanding iterative and approximate techniques for its solution.This paper will present an 'exact' approach to the nice design problem that may be consideredto be a more complete formulation of the approximate approach recently published in theCIBSE Application Manual AM 10: 1997 [1]. The approach presented is based on so-calledloop equations that are commonly used in flow network simulation in the hydraulics field buthave been largely ignored in the building ventilation field. It allows direct sizing of a varietyof airflow components and the direct and unambiguous consideration of both stack-driven andwind-driven flows without resorting to simplifying approximations. Yet, the approach isdeveloped in such a way as to enable building designers to identify a full range of feasibledesign configurations so that other, nontechnical design constraints may be included in theprocess of seeking a design solution - an example of such a design scenario will be presented.