ASHRAE Journal - September 2010 - 63

Review of Previous UFAD Design Tools

Development of Design Tool

Only a limited number of cooling airflow design methods have Figure 1 shows a flow diagram of the calculation process of been described for UFAD systems. Most have adopted concepts the design tool. The approach taken for the design tool developfrom the closely related stratified environments produced by ment was to focus on accounting for the differences between displacement ventilation.3 standard overhead (OH) mixing systems and UFAD systems. In this regard Loudermilk4 proposed a method based on the The tool does not calculate the UFAD cooling load from scratch, separation of the conditioned space into two distinct horizontal but instead uses as an input the cooling load calculated for the layers, a lower zone containing primarily cool fresher air, and same building under design with an OH system. In this way, an upper zone containing warm more polluted air. He assumed4 using a familiar load calculation tool the designer can account that the height of the lower zone was equal to the vertical throw for such factors as building shell construction, orientation, height of the floor diffusers, while in a subsequent similar ap- and climate. proach described by Bauman,5 the height was set equal to the As shown in Equation 1, the tool transforms the design cool“occupied zone.” ing load calculated for an OH system (CLOH),6,7 into the design In both methods, the determination of the required cooling cooling load for a UFAD system (CLUFAD) using a correlation airflow quantity is based on the assignment of convective heat equation for UCLR, “UFAD cooling load ratio.” gains occurring above the separation height into the upper CLUFAD = CLOH ×UCLR (1) zone, allowing this portion of the space cooling loads to be isolated from the lower zone where they would impact comfort The total UFAD cooling load is then split into three fracconditions. tions; supply plenum (SPF), zone or room (ZF), and return Several recognized limitations exist in these UFAD cooling plenum (RPF). It is the fraction (ZF) of the cooling load reload design tools. First, all previous methods were based on maining in the room that is used to determine design cooling adding up the convective portion of design heat sources to airflow rates, as a function of user inputs for diffuser supply represent the instantaneous cooling load in the zone. Com- air temperature, diffuser type and number, room setpoint putation of cooling load is temperature, and other key complicated by the radiant parameters. The diffuser Return Plenum RPF exchange between surfaces supply air temperature is and thermal mass in the a function of the plenum Overhead UCLR Zone ZF UFAD zone.6 Accounting for the configuration, plenum in(Mixing) time delay associated with let temperature, and the the absorption of radiant fraction (SPF) of cooling Supply Plenum SPF heat transfer by thermal load assigned to the supply mass and subsequent rere- Figure 1: Flow diagram of design tool showing transformation from cooling plenum. lease by convection into the load calculated for an overhead mixing system into a UFAD cooling load, The four transformations zone is a major challenge and divided between the supply plenum, zone (room), and return plenum. shown in Figure 1 were in cooling load calculadeveloped by conducting tions. The sum of all space instantaneous heat gains at any a matrix of design-day, EnergyPlus (v3.1)1 simulations of a given time does not necessarily (or even frequently) equal three-story prototype office building. EnergyPlus performs the cooling load for the space at that same time.7 For these a fundamental heat balance calculation and contains UFADreasons we have used EnergyPlus with its fundamental heat specific algorithms that have been validated based on extensive balance calculations in the development of the new design laboratory testing.11,12 tool described later. The simulation study investigated the following nine paramSecond, the cool underfloor air supply plenum in multistory eters: floor level (ground, middle, top), zone (interior, north, buildings in combination with room air stratification produces south, east, west perimeter), structure type (light, medium, several new heat transfer pathways leading to significant heavy), plenum inlet temperature, internal heat load, climate quantities of heat entering the underfloor plenum. Dynamic (seven locations), plenum configuration, window-to-wall ratio, and steady-state modeling research has shown that on aver- and presence of carpet. Therefore, the design tool can be applied age 20% to 40% of the total room cooling load is transferred to a wide range of building types and climates. Full details are into the supply plenum, leaving only about 60% to 80% that described by Schiavon, et al.9 must be removed by airflow through the room.8,9 The amount Previously, it was thought that total cooling loads for UFAD of heat entering the underfloor plenum directly influences the and OH systems are nearly identical, meaning that UCLR design cooling airflow rate and the occupants’ thermal com- would equal 1. However, recent energy modeling research has fort. Finally, although there is reliable data describing how demonstrated that they are different.9 As an example, Figure to split heat gain into radiant and convective components,10 2 shows a comparison between the predicted cooling load there is no research-based guidance on how to assign loads profiles for OH and UFAD systems for five zones of a middle to the upper and lower zones of the room. floor for a Baltimore, summer design day. The internal and
September	2010	 ASHRAE	Journal	 63
ASHRAE Journal - September 2010

Table of Contents for the Digital Edition of ASHRAE Journal - September 2010
