Underfloor Air Distribution (UFAD) Design Guide

Transcription

1 Underfloor Air Distribution (UFAD) Design Guide

2 This publication was prepared under ASHRAE Research Project RP-1064 in cooperation with TC 5.3, Room Air Distribution. ABOUT THE AUTHOR Fred S. Bauman, P.E., is a research specialist with the Center for the Built Environment (CBE) at the University of California, Berkeley. He received his M.S. in mechanical engineering from the University of California at Berkeley. He is an ASHRAE member, member of the Golden Gate Chapter of ASHRAE, and registered mechanical engineer in California. He is a member of Technical Committees 4.7 and 5.3, and of Standards Project Committee R. He served as Chair of TC 5.3, , and Chair of SPC R, He received two Best Symposium Paper Awards from ASHRAE (1992, 1993), and in 1997, received the ASHRAE Distinguished Service Award. He currently leads CBE s research program on underfloor air distribution and task/ambient conditioning, having conducted research in this area since ABOUT THE CONTRIBUTING AUTHOR Allan Daly, P.E., is a principal of Taylor Engineering located in Alameda, California. He received his M.S. in civil engineering from the University of California at Berkeley. He is an ASHRAE member, member of the Golden Gate Chapter of ASHRAE, and registered mechanical engineer in California. His work focuses on HVAC and controls design for commercial and institutional projects. Recent projects include design, analysis, and commissioning of 12 buildings using Underfloor Air Distribution. He and Bauman have taught several workshops together on UFAD design since Allan Daly contributed to this design guide by writing Chapters 7 and 9 and parts of Chapters 11 and 12.

3 Underfloor Air Distribution (UFAD) Design Guide Fred S. Bauman American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc.

4 ISBN American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc Tullie Circle, N.E. Atlanta, GA All rights reserved. Printed in the United States of America Cover design by Tracy Becker. ASHRAE has compiled this publication with care, but ASHRAE has not investigated, and ASHRAE expressly disclaims any duty to investigate, any product, service, process, procedure, design, or the like that may be described herein. The appearance of any technical data or editorial material in this publication does not constitute endorsement, warranty, or guaranty by ASHRAE of any product, service, process, procedure, design, or the like. ASHRAE does not warrant that the information in the publication is free of errors, and ASHRAE does not necessarily agree with any statement or opinion in this publication. The entire risk of the use of any information in this publication is assumed by the user. No part of this book may be reproduced without permission in writing from ASHRAE, except by a reviewer who may quote brief passages or reproduce illustrations in a review with appropriate credit; nor may any part of this book be reproduced, stored in a retrieval system, or transmitted in any way or by any means electronic, photocopying, recording, or other without permission in writing from ASHRAE. ASHRAE STAFF SPECIAL PUBLICATIONS Mildred Geshwiler Editor Erin Howard Assistant Editor Christina Helms Assistant Editor Michshell Phillips Secretary PUBLISHING SERVICES Barry Kurian Manager Jayne Jackson Production Assistant PUBLISHER W. Stephen Comstock

5 Contents Acknowledgments xi Chapter 1 Introduction Purpose of Guide System Description Background Benefits Improved thermal comfort Improved ventilation efficiency and indoor air quality Reduced energy use Reduced life-cycle building costs Reduced floor-to-floor height in new construction Improved productivity and health Technology Needs New and unfamiliar technology Lack of information and design guidelines Gaps in fundamental understanding Perceived higher costs Limited applicability to retrofit construction Problems with applicable standards and codes Cold feet and draft discomfort Problems with spillage and dirt entering UFAD systems 18 v

6 CONTENTS Condensation problems and dehumidification in UFAD systems Applications Organization of Guide 20 Chapter 2 Room Air Distribution Conventional Overhead Mixing Systems Displacement Ventilation and Conditioning Systems UFAD Systems UFAD Room Air Distribution Model Temperature Near the Floor Stratification Height Controlling Stratification 37 Chapter 3 Thermal Comfort and Indoor Air Quality Thermal Comfort Standards Personal Control Thermal Stratification Ventilation Performance Productivity 50 Chapter 4 Underfloor Air Supply Plenums Description Pressurized Plenum Zero-Pressure Plenum Airflow Performance in Pressurized Plenums Dimensional Constraints of the Plenum Plenum Inlets Horizontal Ducting within the Plenum Obstructions within the Plenum Air Leakage Leakage Due to Construction Quality Leakage Between Floor Panels Thermal Performance Thermal Decay Ductwork and Air Highways 66 Chapter 5 Underfloor Air Distribution (UFAD) Equipment Supply Units and Outlets 69 vi

7 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE Types of UFAD and TAC Diffusers Passive Swirl Floor Diffusers Passive VAV Floor Diffusers Linear Floor Grilles Active TAC Diffusers Underfloor Fan Terminals Raised Floor Systems 85 Chapter 6 Controls, Operation, and Maintenance Control Strategies in Pressurized Plenums Supply Air Temperature (SAT) Constant Pressure Variable-Air-Volume (VAV) Controlling Stratification Humidity Control Control Strategies in Zero-Pressure Plenums Individual Outlet Controls Operation and Maintenance Cleaning Considerations in Underfloor Plenums Reconfiguring Building Services Acoustic Performance 97 Chapter 7 Energy Use Air Distribution Energy Air-Side Economizers Extended 100% Free Cooling Extended Integrated-Economizer Free Cooling Climate Factors Cooling-System Efficiency Occupant Thermal Comfort Pre-Cooling Strategies 107 Chapter 8 Design, Construction, and Commissioning Design Phase Construction Retrofit Projects Space Planning Commissioning 116 vii

8 CONTENTS Chapter 9 Perimeter and Special Systems Perimeter System Definition Perimeter System Options Two- or Four-Pipe Constant-Speed Fan Coils Hydronic Heat Pumps VAV or Fan-Powered VAV with Reheat Cooling from VAV Diffusers, Heating from Heating-Only Fan Coil Fan-Powered Outlets Convector or Baseboard Heating Coupled with Central UFAD System Cooling Variable-Speed Fan Coils VAV Change-Over Air Handlers Conference Rooms or Other Special Systems Issues to consider in the Design of Perimeter and Special Systems 132 Chapter 10 Cost Considerations Standard First Cost Components Raised Floor System Slab Modification and Preparation Cleaning and Sealing the Plenum Fire Detection and Sprinkler Systems Design-Dependent First Cost Components UFAD System Design Cable Management Systems Floor-to-Floor Heights Ceiling Finishes and Acoustical Treatment Life-Cycle Cost Components Churn (Reconfiguration) Operation and Maintenance Tax Savings Increased Property Value and Rents Productivity and Health 142 viii

9 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE Chapter 11 Standards, Codes, and Ratings ANSI/ASHRAE Standard : Thermal Environmental Conditions for Human Occupancy ANSI/ASHRAE Standard : Ventilation for Acceptable Indoor Air Quality ANSI/ASHRAE/IESNA Standard : Energy Standard for Buildings Except Low-Rise Residential Buildings ANSI/ASHRAE Standard : Method of Testing for Room Air Diffusion ASHRAE Standard : Measuring Air Change Effectiveness Title-24: CEC Second Generation Nonresidential Standards NFPA 90A: Standard for the Installation of Air-Conditioning and Ventilating Systems Uniform Building and Other Applicable Codes LEED (Leadership in Energy & Environmental Design) Rating System 150 Chapter 12 Design Methodology UFAD vs. Conventional Overhead System Design Building Structure Considerations Building Plan New Construction Retrofit Applications Determination of Space Cooling and Heating Loads Space Cooling Load Calculation Space Heating Load Calculation Determine Ventilation Air Requirements Temperature Control and Zoning Interior Zones Perimeter Zones Other Special Areas Air Distribution System Configuration Plenum Configuration Duct Requirements Determine Zone Supply Air Temperature and Air Flow Requirements 172 ix

10 CONTENTS 12.8 Select and Locate Diffusers Determine Return Air Configuration Select and Size Primary HVAC Equipment Thermal Storage Opportunities 178 Chapter 13 UFAD Project Examples 181 Chapter 14 Future Directions Research Room Air Stratification Underfloor Air Supply Plenums Whole-Building Energy Simulation Model Thermal Comfort Ventilation Performance Field Studies Productivity Studies Cost Studies Design Tools Standards and Codes Building Industry Developments Technology Transfer 188 Glossary 189 References and Annotated Bibliography 207 Index 237 x

11 Acknowledgments The development of this design guide on underfloor air distribution (UFAD) is the result of a cooperative research agreement between the American Society of Heating, Refrigerating and Air-Conditioning Engineers, Inc. (ASHRAE), and the Center for the Built Environment (CBE) at the University of California, Berkeley, for ASHRAE Research Project RP The financial support of both ASHRAE and CBE is gratefully acknowledged. CBE is an NSF/Industry/University Cooperative Research Center whose current sponsors are Armstrong World Industries, Arup, California Department of General Services, California Energy Commission, EHDD Architecture, HOK, Keen Engineering, NBBJ, Pacific Gas & Electric Co., SOM, Steelcase, Inc., Tate Access Floors, Inc., the Taylor Team (Taylor Engineering, Engineering Enterprise, Guttmann & Blaevoet, Southland Industries, Swinerton Builders), Trane, U.S. Department of Energy, U.S. General Services Administration, United Technologies, the Webcor Team (Webcor Builders, Critchfield Mechanical, Rosendin Electric, and C&B Consulting), York International, the National Science Foundation (NSF), and the Regents of the University of California. I would like to thank Allan Daly of Taylor Engineering for serving as a contributing author for this design guide. He has strong practical experience with UFAD systems. Allan was the primary author of Chapters 7 and 9, and contributed sections to Chapters 11 and 12. Technical oversight was provided by ASHRAE Technical Committee TC 5.3 (Room Air Distribution). I would like to express my sincere appreciation for the guidance, constructive comments, and many hours of discussion provided by members of the Project Monitoring Subcommittee (PMS) led by Chair Ken Loudermilk (Trox USA). Other PMS members were Hans Levy (Argon Corp.), Arsen Melikov (Technical xi

12 ACKNOWLEDGMENTS University of Denmark), and Takashi Akimoto (Kanto-Gakuin University). Andrey Livchak (Halton Company) of TC 5.3 also contributed important comments near the end of the review period. I would like to thank Alisdair McGregor (Arup), Robert Shute (The Mitchell Partnership), Jeff Blaevoet (Guttmann and Blaevoet), and Shin-ichi Tanabe (Waseda University), all experts in UFAD technology, for their input at the early stages of this project. Many other individuals have made contributions through their reviews of earlier drafts and generous sharing of ideas and data. I would like to especially acknowledge Mike Critchfield (Critchfield Mechanical), Alf Dyk (E.H. Price Ltd.), Gus Faris (Nailor Industries), Steve Guttmann (Guttmann and Blaevoet), Ralph Hockman (Tate Access Floors), Eric Horn (Webcor Builders), Dan Int-Hout (Krueger), Tim Irvin (York International), Blair McCarry (Keen Engineering), Jim Reese (York International), Dennis Stanke (Trane), Steve Taylor (Taylor Engineering), Dave Troup (HOK), Mark Vranicar (Critchfield Mechanical), and David Wyon (Technical University of Denmark). Several graduate student researchers in the Department of Architecture at UC Berkeley assisted me on this project. I would like to thank Rachel Bannon for her writing and editorial skills, and Jane Lin, Amiee Lee, and Susie Douglas, who produced the majority of the graphics. Many of my research colleagues at UC Berkeley have made valuable contributions through their critical reviews, interest, and enthusiastic support of our UFAD research program. In particular, I would like to thank Tom Webster, my primary co-researcher within the CBE UFAD research program, for our many discussions of UFAD issues that improved our collective understanding of UFAD technology and guided our research directions. I would also like to express my warm appreciation to Ed Arens, Gail Brager, Charlie Huizenga, Cliff Federspiel, Zhang Hui, and David Lehrer, all with CBE. In addition, my thanks go to William Fisk, David Faulkner, and Doug Sullivan of the Indoor Environment Department at Lawrence Berkeley National Laboratory for their technical advice and interest. Finally, I give my love and thanks to Jenny and Rocko for all their support and understanding during the many days, nights, and long hours that I worked on the design guide. xii

13 Chapter 1 Introduction 1.1 PURPOSE OF THIS GUIDE Underfloor air distribution (UFAD) systems are innovative methods for delivering space conditioning in offices and other commercial buildings. Underfloor air distribution derives its name from the use of the underfloor plenum below a raised (access) floor system to supply conditioned air directly into the occupied zone of the building, typically through floor diffusers. The use of UFAD technology is increasing in North America because of the benefits that it offers over conventional overhead air distribution. The purpose of this design guide is to provide assistance in the design of UFAD systems that are energy efficient, intelligently operated, and effective in their performance. This guide also describes important research results that support current thinking on UFAD design and includes an extensive annotated bibliography for those seeking additional detailed information. This guide does not cover conventional overhead air distribution system design procedures in depth but rather focuses on the major differences between UFAD systems and conventional design. For more information on standard heating, ventilating, and air-conditioning (HVAC) design, please refer to other books published by ASHRAE, including the Handbook series [ASHRAE 2000, 2001a, 2002, 2003a], Air-Conditioning Systems Design Manual [ASHRAE 1993], and Designer s Guide to Ceiling- Based Air Diffusion [Rock and Zhu 2001]. Task/ambient conditioning (TAC) systems are a special class of air distribution systems characterized by their ability to allow individuals to have personal control over their local environment, without adversely affecting that of occupants in the surrounding area. A large majority of TAC systems use UFAD with furniture- or partition-based 1

14 CHAPTER 1 INTRODUCTION supply outlets because of the effectiveness of this configuration at providing individual control for nearby occupants. These two closely related air distribution systems share many common features in terms of their design, construction, and operation. This guide also presents preliminary design guidance for TAC systems where available, although applications and experience using this technology are still rather limited. The development of this guide is based on a compilation of available information, including research results from laboratory and field experiments and simulation studies, design experience described in the literature as well as from interviews with practicing engineers, manufacturer s literature, and other relevant guidelines from users of the technology. Despite recent growth in the UFAD market, widespread experience with these systems is still at an early stage, with significant issues the subject of ongoing research. The guidelines presented here are based on the most current and best available data and information. Designers and operators are encouraged to use common sense and good engineering judgment when applying methodologies described in this guide. The guide is intended for use by design engineers, architects, building owners, facility managers, equipment manufacturers and installers, utility engineers, researchers, and other users of UFAD technology. 1.2 SYSTEM DESCRIPTION An underfloor air distribution (UFAD) system uses the open space (underfloor plenum) between a structural slab and the underside of a raised floor system to deliver conditioned air to supply outlets located at or near floor level within the occupied zone (up to 6-ft [1.8-m] height) of the space. Floor diffusers make up the large majority of installed UFAD supply outlets, and throughout this guide, unless otherwise noted, use of the term UFAD system will refer primarily to this configuration. As discussed in Chapter 3, supply outlets can provide different levels of individual control over the local thermal environment, depending on diffuser design and location. Additional details of UFAD systems are presented below. A task/ambient conditioning (TAC) system is defined as any space conditioning system that allows thermal conditions in small, localized zones (e.g., regularly occupied work locations) to be individually controlled by nearby building occupants while still automatically maintaining acceptable environmental conditions in the ambient space of the building (e.g., corridors, open-use space, and other areas outside of 2

15 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE regularly occupied work space). Typically, the occupant can control the perceived temperature of the local environment by adjusting the speed and direction, and in some cases the temperature, of the incoming air supply, much like the dashboard of a car. Although not a requirement, the design of a large majority of TAC systems has involved the use of underfloor air distribution (UFAD). For purposes of presentation in this guide, TAC systems are distinguished from standard UFAD systems by their higher degree of personal comfort control provided by the localized supply outlets. TAC supply outlets use direct velocity cooling to achieve this level of control and are therefore most commonly configured as fan-driven (active) jet-type diffusers that are located as part of the furniture or partitions. Active floor diffusers are also possible. Throughout this guide, use of the term TAC system will refer to a UFAD system featuring active supply outlets with the above-described individual control capabilities. TAC systems that do not employ UFAD, such as desktop systems ducted down from an overhead system, are not covered by this guide. For further information on a complete range of TAC systems, see Bauman and Arens (1996) and Loftness et al. (2002). Figures 1.1, 1.2, and 1.3 present and compare schematic diagrams of a conventional overhead system, UFAD system, and UFAD with TAC system, respectively, for a cooling application in an open-plan office building. Some of the most important advantages of UFAD systems over ceiling-based systems occur for cooling conditions, which Figure 1.1 Conventional overhead air distribution system. 3

16 CHAPTER 1 INTRODUCTION Figure 1.2 Underfloor air distribution system. Figure 1.3 Cutaway of typical office work space showing UFAD with TAC system. 4

17 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE are required year-round in interior office space in many parts of North America. Historically, the approach to HVAC design in commercial buildings has been to supply conditioned air through extensive duct networks to an array of diffusers located in the ceiling. As shown in Figure 1.1, conditioned air is both supplied and exhausted at ceiling level. Ceiling plenums are typically quite deep to accommodate the large supply ducts. Return air is most commonly configured as an un-ducted ceiling plenum return. Often referred to as mixing-type air distribution, conventional HVAC systems are designed to promote complete mixing of supply air with room air, thereby maintaining the entire volume of air in the occupied space at the desired setpoint temperature and evenly distributing ventilation air. UFAD systems are the same as conventional overhead systems in terms of the types of equipment used at the cooling and heating plants and primary air-handling units (AHU). As shown in Figure 1.2, all UFAD systems are configured to use an underfloor air supply plenum to deliver conditioned air directly into the occupied zone, typically through floor outlets. TAC systems use active diffusers that are located as part of the furniture or partitions, although floor-based diffusers are also possible (Figure 1.3). The major features of a UFAD system, with or without TAC supply outlets, are described briefly below. Supply air containing at least the minimum volume of outside air is filtered and conditioned to the required temperature and humidity. It is then delivered by the air-handling unit (AHU) to an underfloor plenum, traveling through a shorter distance of ductwork than for ceiling-based systems. The underfloor plenum is formed by installation of a raised floor system, typically consisting of 2 ft 2 ft (0.6 m 0.6 m) concretefilled steel floor panels. Raised floors used with UFAD systems have typically been installed at heights of in. ( m) above the concrete structural slab of the building, although lower heights are possible. The raised floor system also allows all power/ voice/data (PVD) cabling services to be conveniently distributed through the underfloor plenum (Figure 1.3). Savings associated with these services offset much of the initial cost of the raised floor system. When configuring an underfloor air supply plenum, there are three basic approaches: (1) pressurized plenum with a central air handler delivering air through the plenum and into the space through passive grilles/diffusers, modulated diffusers, and fan-powered termi- 5

18 CHAPTER 1 INTRODUCTION nal units, either used alone or in combination with one another; (2) zero-pressure plenum with air delivered into the conditioned space through local fan-powered (active) supply outlets in combination with the central air handler; and (3) in some cases, ducted air supply through the plenum to terminal devices and supply outlets. The use of pressurized underfloor plenums appears to be the focus of current practice, although zero-pressure plenums pose no risk of uncontrolled air leakage to the conditioned space, adjacent zones, or the outside. Within the plenum, air flows freely in direct contact with the thermally massive slab and floor panels and enters the workspace through diffusers at floor level or as part of the furniture or partitions. Because the air is supplied directly into the occupied zone, floor supply outlet temperatures should be maintained no lower than in the range of F (16-18 C) to avoid uncomfortably cool conditions for the nearby occupants. For TAC supply outlets located closer to the occupant (e.g., furniture- or partition-based diffusers) where the occupant is exposed to diffuser velocity cooling, even warmer supply temperatures may be advisable. UFAD systems are generally configured to have a relatively larger number of smaller supply outlets, many in closer proximity to the building occupants, as opposed to the larger diffusers and spacing used in conventional overhead systems. Outlets that are located within workstations or otherwise near occupants at their work locations are typically adjustable or thermostatically controlled, providing an opportunity for adjacent individuals to at least have some amount of control over their perceived local thermal environment. Fan-driven TAC diffusers can more directly influence local thermal comfort by using increased air movement to provide occupant cooling. Air is returned from the room at ceiling level, or at the maximum allowable height above the occupied zone. This produces an overall floor-to-ceiling airflow pattern that takes advantage of the natural buoyancy produced by heat sources in the office and more efficiently removes heat loads and contaminants from the space, particularly for cooling applications. In contrast to the well-mixed room air conditions of the conventional overhead system, during cooling conditions, UFAD system operation can be optimized to promote some amount of stratification in the space, with elevated temperatures and higher levels of pollutants above head height where their effect on occupants is reduced. 6

19 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE 1.3 BACKGROUND In today s rapidly changing work environment, new factors have emerged that are driving corporate thinking on the type of facility that they will own or occupy. One of the leading drivers is integrated design solutions that provide maximum flexibility to allow facilities to easily adapt to new technologies and new business directions. Secondly, the needs of building occupants are increasingly being recognized as critical in terms of life-cycle cost-effectiveness. Communication, computer, and internet-based technologies enable individual workers to have tremendous control over where, when, and how they work. Advanced and flexible interior furnishings have been developed that can be configured to support a variety of individual and team work patterns. The potential economic benefits of using these and other new building technologies to achieve greater satisfaction within the workforce are known to be very large. These benefits include increased worker productivity, employee retention, reduced operating costs (fewer occupant complaints), and increased market value of facilities. In contrast, HVAC technology has not kept pace with the changing workplace. HVAC approaches have changed little since variable-air volume systems were first introduced 30 years ago. For the vast majority of buildings, it is still standard practice to provide a single uniform thermal and ventilation environment within each building zone, offering little chance of satisfying the environmental needs and preferences of individual occupants (unless, of course, they happen to have a private office with a thermostat). As a result, the quality of the indoor environment (i.e., thermal comfort and indoor air quality) continues to be one of the primary concerns among workers who occupy these buildings. Several documented surveys of building occupants have pointed out the high dissatisfaction with indoor environmental conditions [e.g., Schiller et al. 1988, Harris 1989]. More recently, the Building Owners and Managers Association (BOMA), in partnership with the Urban Land Institute (ULI), surveyed 1,829 office tenants in the U.S. and Canada [BOMA/ULI 1999]. In the survey, office tenants were asked to rate the importance of 53 building features and amenities and to report how satisfied they are with their current office space for those same categories. The following quotes from the report demonstrate the importance of indoor environmental quality and personal control. The most important features, amenities, and services to the responding tenants are related to the comfort and quality of 7

20 CHAPTER 1 INTRODUCTION indoor air, the acoustics, and the quality of the building management s service. Tenants ability to control the temperature in their suite is the only feature to show up on both the list of most important features (96%) and the list of items where tenants are least satisfied (65%). To make an immediate and positive impact on tenants perception of a building, landlords and managers could focus on temperature-related functions by updating HVAC systems so that tenants can control the temperature in their suite or by helping tenants make better use of their existing system. The concept of task/ambient conditioning (TAC) was developed to address many of the problems and concerns outlined above. Just as with task/ambient lighting systems, TAC systems allow ambient air-conditioning requirements to be reduced in noncritical areas. Individually controlled diffusers provide task conditioning only when and where it is needed to maintain occupant comfort. In contrast to the centralized approach described above in which a large zone of the building is controlled by a single wall thermostat, the TAC system concept approaches the optimal solution of providing a collection of many small control zones (e.g., workstations), each under the control of an ideally located and calibrated human thermostat. In addition, by delivering fresh air in the near vicinity of the occupants, TAC systems are more likely to provide improved air movement and preferential ventilation in the occupied zone, as compared to conventional mixing-type air distribution systems. Underfloor air distribution, originally introduced in the 1950s in spaces having high heat loads (e.g., computer rooms, control centers, and laboratories), has proved to be the most effective method for delivering conditioned air to localized diffusers in the occupied zone of a building. In these early installations, the raised floor system was used to handle the large amounts of cables serving the computers and other equipment. By supplying cool air through floor diffusers and returning air at the ceiling, the overall floor-to-ceiling airflow pattern supported the buoyancy-driven air movement and efficient removal of heat loads from the space. The maintenance of thermal conditions within the comfort zone was not a major focus of these early applications as they were primarily concerned with equipment cooling, not people cooling. As a result, the first floor diffusers were not designed to be easily adjustable. 8

21 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE In the 1970s, underfloor air distribution was introduced into office buildings in West Germany as a solution to these same cable management and heat load removal issues caused by the proliferation of electronic equipment throughout the office [David 1984; Sodec and Craig 1990]. In these buildings, the comfort of the office workers had to be considered, giving rise to the development of occupant-controlled localized supply diffusers to provide task conditioning. Some of the first systems in Europe used a combination of desktop outlets (TAC) for personal comfort control and floor diffusers (UFAD) for ambient space control [Sodec 1984; Barker et al. 1987]. To date, UFAD systems have achieved considerable acceptance in Europe, South Africa, and Japan. However, growth in North America was relatively slow until the late 1990s. As with any new and unfamiliar technology, resistance to wider use has been driven by the perceived higher risk to designers and building owners primarily due to a lack of objective information and standardized design guidelines, a lack of well-documented case studies with performance and cost-savings data, and, in the case of underfloor air, the perceived higher first costs of raised flooring. (Most of the cost of access flooring, if not all of it, is amortized by the savings in wiring for electric, power, telephone, and computers, as well as reduced ductwork.) In addition, there are important gaps in our fundamental understanding of UFAD. Key areas where information is lacking are: impact of air diffuser characteristics on stratification, behavior of thermal plumes at solar-heated windows, interaction between thermal plumes and diffuser airflows, ventilation efficiencies, thermal performance of underfloor air supply plenums, and health and comfort benefits. UFAD technology is now in a situation where systems are being designed and installed at an increasingly rapid pace, even before a full understanding and characterization of some of the most fundamental aspects of UFAD system performance have taken place. Although independent market data are not available, estimates from several leading manufacturers of raised flooring and floor diffusers provide the following statistics for the market penetration of raised floors and UFAD systems. In 1995, less than 3% of new office buildings in North America used raised floors, with UFAD considered as a fringe practice. In 1999, 8% of new offices used raised floors with 20%-25% of these including UFAD systems. Prior to the recent economic downturn, manufacturers had predicted that by 2004, 35% of new offices would be using raised floors, with 50% of those using UFAD [Krepchin 2001]. The attainment of these numbers is likely to be delayed, as at the time of writing of this guide, raised floor market penetration is at about 12% 9

22 CHAPTER 1 INTRODUCTION to 15% with about 40% of these using UFAD systems [Hockman 2002]. In terms of previous research, UFAD and TAC systems have attracted the attention of a number of investigators who present data from test chamber studies of several floor diffusers [Barker 1985; Tuddenham 1986; Rowlinson and Croome 1987; Hanzawa and Nagasawa 1990; Arens et al. 1991, 1995; Bauman et al. 1991a, 1995; Fisk et al. 1991; Yokoyama and Inoue 1991, 1993, 1994; Fountain 1993; Fountain et al. 1994; Tanabe 1994; Faulkner et al. 1995; Matsunawa et al. 1995; Tanabe and Kimura 1996; Tsuzuki et al. 1999; Kim et al. 2001; Webster et al. 2002a, 2002b]. Other laboratory studies are reported in the literature describing the performance of TAC desk-based supply diffusers [Arens et al. 1991, 1995; Bauman et al. 1993, 2000b; Faulkner et al. 1993, 1999, 2002; Fountain 1993; Fountain et al. 1994; Tsuzuki et al. 1999; Levy 2002] and partition-based supply diffusers [SHASE 1991; Zhu et al. 1995]. As more underfloor and TAC system installations have been completed in recent years, the experience and knowledge base of these systems have grown. The results of field measurements, occupant surveys, and case studies have also been reported [Wyon 1988; Spoormaker 1990; Hedge et al. 1992; Kroner et al. 1992; Bauman et al. 1993, 1994; Matsunawa et al. 1995; Oguro et al. 1995; McCarry 1998; Webster et al. 2002c; Daly 2002]. Several authors have discussed energy performance, operating characteristics, and occupant issues for UFAD systems in buildings [Tuddenham 1986; Barker et al. 1987; Genter 1989; Arnold 1990; Heinemeier et al. 1990; Sodec and Craig 1990; Drake et al. 1991; Imagawa and Mima 1991; SHASE 1991; Tanaka 1991; Shute 1992; Nagoya University 1994; Matsunawa et al. 1995; Bauman and Webster 2001]. A number of publications have addressed design methods [Spoormaker 1990; Sodec and Craig 1991; Houghton 1995; McCarry 1995; Shute 1995; Bauman and Arens 1996; Bauman et al. 1999a; Bauman 1999; AEC 2000]. In recent years several manufacturers of HVAC systems and components have developed publications and literature addressing UFAD systems [e.g., Trox 1997; York 1999; Int- Hout 2001; Stanke 2001; Argon 2002]. Many design firms specializing in UFAD design now feature project profiles of completed UFAD projects on their web sites. Currently, research on UFAD and TAC systems is ongoing at three university research centers: 1. Center for the Built Environment (CBE), University of California, Berkeley, (including funding from ASHRAE for this design guide). CBE 10

23 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE has developed a public web site on underfloor air technology ( 2. Center for Building Performance and Diagnostics (CBPD), Carnegie Mellon University (CMU), Pittsburgh, CMU recently completed a stateof-the-art review of Flexible and Adaptive HVAC Distribution Systems for Office Buildings, with funding from the Air-Conditioning and Refrigeration Technology Institute (ARTI) [Loftness et al. 2002]. 3. International Centre for Indoor Environment and Energy (ICIEE), Technical University of Denmark, ICIEE is conducting research on both physical measurements and human response to personalized ventilation, as provided by TAC diffusers. Additional references will be referred to during the discussions presented later in this guide and may also be found in the References and Annotated Bibliography. 1.4 BENEFITS What are the potential advantages that UFAD systems have over traditional overhead air distribution systems? Well-engineered systems can provide the following Improved Thermal Comfort By allowing individual occupants to control their local thermal environment, their individual comfort preferences can be accommodated. In today s work environment, there can be significant variations in individual comfort preferences due to differences in clothing, activity level (metabolic rate), and individual preferences. Recent laboratory tests show that commercially available task/ambient conditioning systems with fan-driven supply outlets (airflow directed at the occupant) provide personal control of an occupant s microclimate over a sizable range up to 13 F (7 C) for desktop outlets and up to 9 F (5 C) for floor-based outlets [Tsuzuki et al. 1999]. These tests measured only sensible cooling rates; total cooling (including latent effects) would be even higher. This amount of control is more than enough to allow the full range of individual thermal preferences to be accommodated. Passive diffusers (diffusers that do not rely on local fans), such as the commonly used swirl floor diffusers in UFAD systems, will not provide this same magnitude of control. However, by being accessible to the occu- 11

24 CHAPTER 1 INTRODUCTION pants, these diffusers can still be effective at influencing the perceived local comfort conditions. For further discussion, see Section Improved Ventilation Efficiency and Indoor Air Quality Some improvement in ventilation and indoor air quality at the breathing level can be expected by delivering the fresh supply air at floor level or near the occupant and returning at the ceiling, resulting in an upward displacement of indoor air and pollutant flow pattern, similar to that achieved in the displacement ventilation systems commonly used in Scandinavia [Nielsen 1996]. Displacement ventilation systems (used for cooling only) typically achieve their improved ventilation performance by supplying 100% outside air at a temperature slightly below comfort conditions and at a very low velocity. Because the supply air has little momentum, buoyancy forces influence the airflow pattern and the supply air spreads out at floor level and then flows upward. Air temperatures and concentrations of some pollutants increase with height in the displacement zone. Because UFAD systems supply air at higher outlet velocities than true displacement systems, greater mixing will occur, diminishing the degree of displacement flow. In addition, the recirculation of indoor air by some underfloor systems will cause mixing of indoor air and pollutants. An optimized ventilation strategy is to control supply outlets to confine the mixing of supply air with room air to just below the standard respiration height (3-5 ft [ m]) of the space. Above this height, stratified and more polluted air is allowed to occur. The air that the occupant breathes will have a lower concentration of contaminants compared to conventional uniformly mixed systems. Recent research has shown that desk-mounted TAC diffusers can provide significantly improved ventilation effectiveness over mixing systems [Faulkner et al. 2002; Melikov et al. 2002]. For further discussion, see Section Reduced Energy Use Energy savings for UFAD systems over conventional overhead systems are predominately associated with two major factors: (1) cooling energy savings from economizer operation and increased chiller COP and (2) fan energy savings. Economizer savings result from increased hours of full or partial economizer operation due to higher return air temperatures (77-86 F [25-30 C] vs. 75 F [24 C] for overhead systems) and the reduction in cooling energy required during economizer operation because of the use of higher supply air temperatures (61-65 F 12

25 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE [16-18 C] vs. 55 F [13 C] for overhead systems). Chiller savings result from using higher chiller leaving water temperatures due to the higher supply air temperatures. However, this benefit is climate dependent; moisture control requirements in humid climates will reduce or eliminate these cooling energy savings. Many designers caution against this approach since it presents the opportunity to lose humidity control if not done carefully. Fan energy savings are associated with two factors: reduced total air volume and reduced static pressure requirements. The stratified floorto-ceiling airflow pattern in UFAD systems allows most convective heat gains from sources above the lower mixed zone (see Chapter 2) of the space to be returned directly at ceiling level and therefore to not be included in the calculation of the air supply quantity (air-side load). The determination of air supply volumes required to maintain a given comfort condition are therefore only based on heat sources that enter and mix with air in the occupied zone. Static pressures are reduced due to the elimination of most branch ductwork, as the supply air flows freely through the underfloor plenum at low plenum pressures (typical pressures are 0.1 in. H 2 O (25 Pa) or less). From a recent analysis of central fan energy use in UFAD systems, the average savings using a variableair-volume (VAV) control strategy over conventional VAV systems can be estimated to be about 40% [Webster et al. 2000]. Due to the common practice of using fan-powered solutions in perimeter zones, the total fan energy savings may be significantly reduced when the energy use of these additional smaller fan units is considered. Characterization of additional energy savings potential is being addressed by ongoing research. For further discussion, see Chapter 7. TAC systems provide additional energy considerations. In terms of fan energy use, the reduced energy consumption of the central AHU must be traded off against the additional energy used by the active (fandriven) supply outlets. If all occupants have access to a TAC diffuser that provides velocity cooling, the entire space can be operated at a higher temperature with potentially significant cooling energy savings Reduced Life-Cycle Building Costs In modern businesses, churn is a fact of life; a 1997 survey found the national average churn rate (defined as the percentage of workers and their associated work spaces in a building, %/year, that are reconfigured or undergo significant changes) to be 44% [IFMA 1997]. The cost savings associated with reconfiguring building services is a major factor in the decision to install access flooring. By integrating a building's HVAC and cable management systems into one easily accessible 13

26 CHAPTER 1 INTRODUCTION underfloor plenum, floor diffusers along with all power, voice, and data outlets can be placed almost anywhere on the raised floor grid. In-house maintenance personnel can carry out these reconfigurations at significantly reduced expense using simple tools and modular hardware. Firms that are more likely to install underfloor systems are also, for the very same reasons, more likely to churn at a higher rate. For further discussion, see Chapter Reduced Floor-to-Floor Height in New Construction Buildings using UFAD have the potential to reduce floor-to-floor heights compared to projects with conventionally designed ceilingbased air distribution. This can be accomplished by reducing the overall height of service plenums and/or by changing from standard steel beam construction to a concrete (flat slab) structural approach. Concrete flat slab construction can take longer than steel beam construction but is preferred for underfloor systems due to thermal storage benefits, as well as reduced vertical height requirements. By placing most building services in the underfloor plenum, it is not uncommon and certainly possible to eliminate the ceiling plenum. For further discussion, see Section Improved Productivity and Health Research evidence suggests that occupant satisfaction and productivity can be increased by giving individuals greater control over their local environment and by improving the quality of indoor environments (thermal, acoustical, ventilation, and lighting). A review of relevant research has concluded that improvements in productivity in the range of 0.5% to 5% may be possible when the thermal and lighting indoor environmental quality is enhanced [Fisk 2000]. These percentages, though small, have a life-cycle value approximating that of the capital and operating costs of an entire building! For further discussion, see Sections 3.5 and TECHNOLOGY NEEDS Despite the advantages of UFAD systems, there exist some barriers (both real and perceived) to widespread adoption of this technology. Resistance to wider use has been driven by the perceived higher risk to designers and building owners primarily due to a lack of objective information and standardized design guidelines, perceived higher costs, limited applicability to retrofit construction, problems with applicable standards and codes, and a lack of well-documented case 14

27 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE studies with whole-building performance and cost-savings data. These barriers are summarized below along with ongoing efforts to address these technology needs New and Unfamiliar Technology For the majority of building owners, developers, facility managers, architects, engineers, and equipment manufacturers, UFAD systems still represent a relatively new and unfamiliar technology. Lack of familiarity can create problems throughout the entire building design, construction, and operation process, including higher cost estimates, incompatible construction methods, and incorrect building control and operation on the part of both facility managers and building occupants. As UFAD technology continues to grow, these problems should become less prevalent Lack of Information and Design Guidelines Although in recent years there have been an increased number of publications on UFAD technology, including some with design methods, there has not previously existed a set of standardized design guidelines for use by the industry. To address this problem, ASHRAE has funded the development of this design guide through ASHRAE research project 1064-RP, thereby making it available to the professional design and engineering community at large. In addition, a public web site on UFAD technology has recently been developed [Bauman et al. 2000a] Gaps in Fundamental Understanding Currently, there exists a strong need to improve the fundamental understanding of several key issues related to energy and comfort performance of UFAD system design. These issues include the following Room air stratification. What fraction of the convective heat sources in the space will rise up as thermal plumes and be exhausted directly at ceiling level and can therefore be neglected in the calculation of the room cooling air quantity? What effect do supply airflow, supply air temperature, and ceiling height have on room air stratification? Although some empirical design methods exist [Loudermilk 1999], an understanding of controlled/optimized thermal stratification is critical to provide designers with a reliable energy-estimating tool as well as a sound basis from which to develop design tools and guidelines. Recent research is providing new information about the impact of various UFAD system design and operating parameters on room air stratification [Webster et al. 15

28 CHAPTER 1 INTRODUCTION 2002a, 2002b; Lin and Linden 2002; Yamanaka et al. 2002]. For further discussion, see Chapter Underfloor air supply plenum. An important difference between conventional and UFAD system design is the heat exchange between the concrete slab, raised floor panels, and the supply air as it flows through the underfloor plenum. If the slab has absorbed heat, particularly from warm return air flowing along the underside of the slab, then supply temperature will increase with distance from the plenum inlet. Energy and operating cost savings, including peak shaving, can be achieved by using the concrete slab in a thermal storage strategy, but further research is still needed to optimize and quantify this effect. For further discussion, see Chapter Whole-building performance. There currently does not exist a whole-building energy simulation program capable of accurately modeling UFAD systems, a subject discussed by Addison and Nall (2001). This is one of the top technology needs identified by system designers. Additionally, whole-building performance data are needed from completed UFAD projects in the form of energy use, indoor environmental quality, occupant satisfaction, comfort, health, and performance, and first and life-cycle (operating) costs to quantify the relative benefits of the technology Perceived Higher Costs The perceived higher cost is one of the main reasons why UFAD and TAC technology has been slow to be adopted by the U.S. building industry. As discussed above, this situation is now changing due to significant savings in life-cycle costs. In general, the added first cost of the access floor may be offset by cost reductions associated with decreased ductwork and cable and wire installation. Projects are frequently sold on the basis that UFAD is an add-on after the choice is already made to install access flooring for its cable management and reconfiguring benefits for high churn businesses. Considered in this light, the first cost of a UFAD system is commonly less than a conventional system. This technology is still in the early stages of adoption and certainly will see cost reductions as volumes increase and more UFAD-specific products become available. For further discussion, see Chapter Limited Applicability to Retrofit Construction The installation of UFAD systems and the advantages that they offer are most easily achieved in new construction. However, the widespread use of underfloor air distribution in renovation work has been restricted by the feasibility of adding a raised floor in the large majority 16

29 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE of buildings having limited floor-to-floor heights. Current practice calls for typical raised floor heights of inches ( m). A recent full-scale field experiment has found that low-height underfloor plenums (8 in. [0.2 m] and lower) can, in fact, provide very uniform airflow performance across a 3,200 ft 2 (300 m 2 ) area of a building [Bauman et al. 1999a]. In cases of major remodeling, substantial cost savings may be achieved through the use of raised flooring. UFAD systems can also be installed at considerable savings and with improved performance as a retrofit in high-ceiling spaces, such as warehouses (see Section ) Problems with Applicable Standards and Codes Since UFAD and TAC technology is relatively new to the building industry, its characteristics may require consideration of unfamiliar code requirements and, in fact, may be in conflict with the provisions of some existing standards and codes. Three ASHRAE standards have direct relevance to UFAD and TAC systems. ASHRAE Standard [ASHRAE 1992] specifies a comfort zone, representing the optimal range and combination of thermal and personal factors for human occupancy. Standard [ASHRAE 2001b] provides guidelines for the determination of ventilation rates that will maintain acceptable indoor air quality. The revised version of Standard 62 is expected to allow some adjustment in ventilation rates based on the ventilation effectiveness of the air distribution system, a feature that may give credit to UFAD and TAC systems. ASHRAE Standard [ASHRAE 1990] is the only existing building standard for evaluating the air diffusion performance of an air distribution system. Currently only applicable to conventional overhead systems, Standard 113 is now being revised to be compatible with UFAD, TAC, and displacement ventilation systems. Local building and fire codes need to be considered early in the design process. Code officials having limited experience with UFAD and TAC systems have been known to create unexpected roadblocks due to misunderstandings or narrow interpretations of code language. However, fundamentally the codes governing underfloor plenums should be no different than those for ceiling plenums. For further discussion, see Chapter Cold Feet and Draft Discomfort UFAD systems are perceived by some to produce a cold floor and, because of the close proximity of supply outlets to the occupants, the increased possibility of excessive draft. These conditions are primarily 17

30 CHAPTER 1 INTRODUCTION indicative of a poorly designed or operated underfloor system. Typical underfloor supply air temperatures are no lower than 61 F (16 C) and usually higher except under peak load conditions. Nearly all office installations are carpeted so that cold floors should not be a problem. Individually controlled supply diffusers allow occupants to adjust the local airflow to match their personal preferences and avoid undesirable drafts Problems with Spillage and Dirt Entering UFAD Systems Concern is sometimes expressed about the increased probability of spillage and dirt entering directly into the underfloor supply airstream and therefore being more widely distributed throughout the occupied space. Most floor diffusers, however, have been designed with catchbasins (e.g., to hold the liquid from a typical soft drink spill). Tests have shown that floor diffusers do not blow more dirt into the space than other air distribution systems [Matsunawa et al. 1995]. In addition, air speeds within the underfloor plenum are so low that they do not entrain any dirt or other contaminants from the plenum surfaces into the supply air. Using furniture- or partition-based TAC supply outlets, it is also possible to design a system without floor grilles Condensation Problems and Dehumidification in UFAD Systems In humid climates, outside air must be properly dehumidified before delivering supply air to the underfloor plenum where condensation may occur on cool structural slab surfaces. While humidity control of this sort is not difficult, given the large surface area of the structural slab in the underfloor plenum, it is important that it be done correctly. If a higher cooling coil temperature is used (allowing an increased chiller efficiency) to produce the warmer supply air temperatures needed in UFAD and TAC systems, the cooling coil s capacity to dehumidify will be reduced. In humid climates, a return air bypass control strategy can be employed in which a portion of the return air is bypassed around the cooling coil and then mixed with the air leaving the coil to produce the desired warmer supply air temperature (61-65 F [16-18 C]). In this situation other system design considerations will dictate whether a conventional cooling coil temperature (producing a coil leaving temperature of 55 F [12.8 C]) or a colder one (e.g., from ice storage) is used. 18

31 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE 1.6 APPLICATIONS UFAD systems are well suited for all office buildings, especially those with open office plans in which adjustable diffusers can allow occupants to individually control their local workstation environments. In high-tech offices and other businesses with extensive use of information technologies and typically high churn rates (e.g., dot-com offices, call centers, trading floors), the flexibility provided by service delivery systems, including cable management, is a great benefit. Because of the significant savings in life-cycle costs for UFAD systems, owner-occupied buildings are strong candidates for application. Other buildings suitable for UFAD systems include schools, television studios, and light manufacturing installations that don t involve spillage of liquids. Any building that already is using a raised floor system for cable distribution or other purposes should consider a UFAD system. An exception would be clean room applications that are designed to return air at floor level. There are other areas in buildings where raised floors and underfloor air distribution are generally not appropriate. These areas include those in which spillage has the potential to occur, such as in laboratories, cafeterias, and shop areas. Bathrooms have often been considered as an area where raised floor systems should be avoided, but there are cases where they have been used successfully. Although requiring a membrane on top of the floor to protect against leaks, plumbing costs can be reduced by simplifying the piping installation. In high ceiling spaces UFAD systems provide good energy-savings opportunities in cooling applications by promoting thermal stratification. Comfort and improved indoor air quality are maintained in the occupied zone near the floor, while allowing increased temperatures and pollutant concentrations to occur at higher elevations in the space. Auditoriums, theaters, libraries, museums, and converted warehouses all make good UFAD applications. In contrast, these types of buildings can present problems for conventional overhead air distribution design. Buildings using UFAD systems located in dry, mild climates will achieve the best energy savings. These are primarily associated with increased economizer operation and increased chiller COP due to the higher supply air temperatures used in these systems. These climates are also more suitable for the implementation of thermal storage control strategies using the concrete floor slabs of the building. Many of these energy benefits are not available in more humid climates. 19

32 CHAPTER 1 INTRODUCTION 1.7 ORGANIZATION OF GUIDE Since this document represents the first extensive design guide on UFAD technology, most readers will benefit from reading, or at least skimming, all of the sections. The primary focus of the guide is on underfloor air distribution, since this technology has by far the most information and design experience from which to develop the guide. When available, preliminary guidance is also provided on the design of the closely related task/ambient conditioning systems that use UFAD. Although the guide touches upon the principles of conventional overhead air distribution for comparison, it does not contain detailed design guidance for these systems. Instead, the reader is referred to other publications for information on standard HVAC system design. The topics selected for presentation in this guide represent areas in which important differences exist between conventional systems and UFAD design. Chapters 2-11 provide detailed background information on one of these major topics by discussing the knowledge and experience gained through previous research and applications. Chapter 12 steps through the entire design process by providing a more concise discussion of the issues and refers to other sections in the guide for additional details. The following is a summary of the material contained in the sections of this guide. Chapter 1, Introduction, defines UFAD and TAC systems and provides background and an overview of current information about benefits and needs of these technologies. Chapter 2, Room Air Distribution, describes and compares three approaches to room air distribution design (overhead mixing, displacement ventilation, and UFAD) to illustrate key characteristics of room air distribution using UFAD systems. Included in the discussion is how room air distribution impacts thermal stratification, airflow requirements, ventilation performance, and indoor air quality. Chapter 3, Thermal Comfort and Indoor Air Quality, discusses how delivering conditioned air in the near vicinity and under individual occupant control can improve thermal comfort and ventilation performance. Chapter 4, Underfloor Air Supply Plenums, discusses current research and design information on configuring and operating underfloor air supply plenums. Chapter 5, Underfloor Air Distribution (UFAD) Equipment, describes the range of UFAD and TAC products that are currently available. 20

33 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE Chapter 6, Controls, Operation, and Maintenance, discusses control strategies for optimal and energy-efficient operation and maintenance issues for UFAD systems. Chapter 7, Energy Use, summarizes the major system design and operation issues that influence the efficient energy performance of UFAD systems. Chapter 8, Design, Construction, and Commissioning, reviews issues associated with the design, construction, and commissioning process for UFAD installations. Chapter 9, Perimeter and Special Systems, presents and illustrates a range of system design solutions for conditioning perimeter and other special zones. Chapter 10, Cost Considerations, introduces key economic considerations associated with first and life-cycle costs of UFAD systems. Chapter 11, Standards, Codes, and Ratings, reviews applicable building standards and codes and discusses their compatibility with UFAD and TAC technology. In addition, a description of the LEED (Leadership in Energy & Environmental Design) Rating System is provided. Chapter 12, Design Methodology, presents a summary of recommended design procedures for UFAD systems. In particular, those areas where UFAD design differs from conventional overhead air distribution design are discussed. Chapter 13, UFAD Project Examples, presents a list of web sites, references and other sources describing examples of UFAD and TAC system configurations. Chapter 14, Future Directions, describes ongoing research and standards development work, as well as recommended future directions within the building industry, addressing UFAD and TAC technology needs. Glossary, defines terminology related to UFAD and TAC technology specifically and to HVAC design in general. References and Annotated Bibliography, provides a complete list of references for all sections as well as other publications related to UFAD and TAC technology for readers seeking additional information. Brief descriptions of the contents of key references are provided. 21

34

35 Chapter 2 Room Air Distribution The movement of air through the conditioned spaces of buildings plays a critical role in the performance of a building s heating, ventilating, and air-conditioning (HVAC) system, directly affecting thermal comfort, indoor air quality, and energy use. Most of the potential performance advantages of underfloor air distribution (UFAD) systems over conventional air distribution system design arise from the fact that conditioned air is delivered at or near floor level, directly into the occupied zone of the building, and is returned at or near ceiling level. In this chapter, three approaches to room air distribution design (overhead mixing, displacement ventilation, UFAD) are described and compared to illustrate the characteristics of room air distribution using UFAD systems. 2.1 CONVENTIONAL OVERHEAD MIXING SYSTEMS Historically, the approach to HVAC design in commercial buildings has been to both supply and remove air at ceiling level (Figure 2.1). Conditioned air is typically supplied at velocities that are much higher than those acceptable for occupant comfort. Supply air temperature may be lower, higher, or equal to the desired room air temperature setpoint, depending on the cooling/heating load. Incoming high-speed turbulent air jets create rapid mixing with the room air so that the supply jet s temperature quickly approaches that of the entire room. As the jet proceeds into the room, it entrains room (secondary) air into the primary air jet, causing it to grow and spread in size and therefore to reduce in air speed. A system of overhead diffusers is designed and operated so that the ceiling-based supply air jets slow to an acceptable air speed (no higher than 50 fpm [0.25 m/s]) before entering the occupied zone (up to 6 ft [1.8 m]) [ASHRAE 2001a; Rock and Zhu 2001]. 23

36 CHAPTER 2 ROOM AIR DISTRIBUTION Figure 2.1 Conventional overhead air distribution system. Often referred to as mixing-type air distribution, conventional overhead systems promote complete mixing of supply air with room air, thereby maintaining the entire volume of air in the occupied space at the desired setpoint temperature and evenly distributing ventilation air. In this system, room air conditions approach those of the return air leaving the room at ceiling level. Mixing-type systems maintain acceptable indoor air quality through dilution of the pollutants in the space with a sufficient amount of outside air. This uniform mixing control strategy provides little opportunity (other than by increasing the number of zones) to accommodate different thermal preferences among the occupants or to provide preferential ventilation in the occupied zone. In open plan offices, even by adding more zones, overhead systems can never allow individual control of local workstation environments. Available data from field measurements indicate that the ventilation effectiveness within the occupied zones of rooms is usually uniform within ~15% [Fisk and Faulkner 1992; Persily 1986; Persily and Dols 1989]. Poorer mixing and a significant short-circuiting airflow pattern can sometimes occur when warm air is supplied at ceiling level for heating [Fisk et al. 1997]. Int-Hout (1998) discusses the proper selection of diffusers for overhead systems to ensure adequate mixing of the ventilation air in the space and occupant comfort. 2.2 DISPLACEMENT VENTILATION AND CONDITIONING SYSTEMS Displacement ventilation (DV) is based on many of the same principles that are also important in the cooling performance of UFAD systems. Extensive research on DV systems has produced a substantial 24

37 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE scientific literature base. It is therefore instructive to review this literature to understand both the similarities and differences between DV and UFAD systems. Skistad (1994), Nielsen (1996), and most recently REHVA (2001) provide good comprehensive overviews of displacement ventilation based on both theoretical and experimental considerations. Recently, ASHRAE has also sponsored research on the application of DV systems in the U.S. [Yuan et al. 1998, 1999]. In cooling operation, DV and UFAD systems deliver cool air into the conditioned space at or near floor level and return it at or near ceiling level. Thermal plumes that develop over heat sources in the room play a major role in driving the overall floor-to-ceiling air motion by entraining air from the surrounding space and drawing it upward. This buoyancy-driven floor-to-ceiling airflow pattern also adapts naturally to locally high heat loads as the stronger thermal plume rising above these larger heat sources entrains additional cooler room air from low elevations in the space surrounding the heat sources. The primary difference between DV and UFAD systems is in the manner in which the air is delivered into the space. The classic DV system delivers air at very low velocities while UFAD (and TAC) systems employ higher velocity diffusers with correspondingly greater mixing. Additional details and discussion of DV systems are presented below. Displacement ventilation has been widely used in Scandinavia during the past two decades, particularly in industrial facilities with high ceilings and high thermal load [Svensson 1989]. The main goal of this method of room air distribution is to provide improved indoor air quality (ventilation performance) in the occupied zone compared to the dilution ventilation provided by overhead mixing systems. DV systems are especially effective when pollutants are associated with heat sources in the space (e.g., people and printers in offices). As air is heated and rises into the region above the occupied zone, some of it exits the space with only partial mixing with the room air. This principle enables the room load to be satisfied with a lower volume of supply air than would be needed if the room was completely well mixed, all other conditions being equal. The upward movement of air in the room takes advantage of the natural buoyancy of heat gain to the space, producing a vertical temperature gradient. This thermal stratification typically creates two characteristic horizontal zones, as shown in Figure 2.2, for an office configuration. The lower zone predominantly contains cool fresh air and the upper zone contains warm, more polluted air. The horizontal interface separating the two zones features changes in gradient for both temperature and pollutant concentrations. Due to commonly occurring disturbances (drafts, people moving, etc.) these 25

38 CHAPTER 2 ROOM AIR DISTRIBUTION Figure 2.2 Displacement ventilation system. changes generally occur more gradually over a region or layer separating the upper and lower zones rather than at a distinct height. This interface has been identified by various researchers using different names, including stratification height, stratification boundary, interface height, and shift point. We will refer to it as stratification height (SH) in this guide. In the classic definition of a DV system, which is applied only for cooling purposes, air is supplied at very low velocity through supply devices located near floor level (the most common are low side-wall diffusers) and is returned near ceiling level. Although possible, it is not necessary, nor is it common practice, to install a raised floor to operate a DV system. Because supply air is delivered directly into the occupied zone, it is introduced at a temperature only slightly (5-10 F [3-6 C]) below comfort conditions. In contrast to both overhead mixing systems and UFAD systems, the incoming supply air has very little momentum, and as the cooler, heavier supply air enters the space, it spreads across the floor in much the same way as water would. The air is heated as it flows across the floor and then is drawn upward, primarily through entrainment by thermal plumes (Figure 2.2). The aim of a DV system is to deliver fresh conditioned air directly to the occupants without unnecessarily conditioning other space heat sources. In European applications, displacement ventilation systems usually supply 100% outside air (no recirculated indoor air) and can therefore achieve improved ventilation effectiveness compared to mixing systems. Due to hot and humid climatic conditions in many parts of the U.S., most displacement ventilation installations in this country use return air. Improved indoor air quality compared to mixing systems can still be 26

39 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE Figure 2.3 Thermal plume from a point source. achieved because of stratification of contaminants and extended hours of economizer operation [Livchak and Nall 2001]. As shown in Figure 2.2, a stratification height (SH) is established that divides the room into two zones (upper and lower) having distinct airflow conditions. Plume theory helps to explain this two-zone airflow pattern. As a thermal plume rises due to natural convection above a heat source, it entrains surrounding air and therefore increases in size and volume, although gradually decreasing in velocity from its maximum just above the heat source (Figure 2.3). The maximum height to which a plume will rise is dependent primarily on the heat source strength and secondarily on the stratification in the room (which decreases the buoyancy of the rising plume). The lower zone below the stratification height has no recirculation. In this region, as described above, the fresh, cool supply air gradually flows across the room like cold water in a thin layer that is typically 4-6 in. ( mm) thick. It is drawn horizontally toward the heat sources where it joins the rising air in the thermal plumes and is entrained vertically upward. Depending on the amount of supply air, above some height in the space the rising plumes will require additional (recirculated) air from the upper part of the room to 27

40 CHAPTER 2 ROOM AIR DISTRIBUTION Figure 2.4 Schematic diagram of major flow elements in a room with displacement ventilation. feed the entrainment. These plumes will expand and rise until they encounter equally warm air in the upper regions of the space. The upper zone above the stratification height is characterized by low-velocity recirculation, which produces a fairly well mixed layer of warm air whose contaminant concentration exceeds that in the lower levels of the space. A key feature of the stratification height in a true DV system is that vertical air motion across the level is due only to the effects of buoyancy. In an idealized configuration in which only heat sources are present, only thermal plumes of sufficient strength will rise into the upper zone. The net result will be that once the warmer and more polluted air enters the upper zone, it will never reenter the lower zone. This principle is the basis for the improved ventilation effectiveness and heat removal efficiency associated with DV systems. In some practical applications (e.g., morning start-up, winter), there will also be sources of cooling present in the space, such as a cold perimeter window. In this situation, the resulting cold downdraft may transport some air from the upper zone back down into the lower zone. Figure 2.4 shows these basic elements in a simplified schematic of a DV system. In the figure, q 0 represents the supply airflow into the room from a low side-wall diffuser, q 1 is the upward moving airflow contained in thermal plumes that form above heat sources, and q 2 is the downward moving airflow resulting 28

41 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE Figure 2.5 Vertical profiles of pollutant concentrations in a room with displacement ventilation. from cool surfaces. In terms of this simplified configuration, the stratification height will occur at a height (y st ) where the net upward moving flow, q 1 q 2, equals q 0. Clearly, an important objective in designing and operating a DV system is to maintain the stratification height near the top of the occupied zone (1.8 m [6 ft]). If the building occupants are in a seated work position, a lower stratification height (e.g., 1.2 m [4 ft]) may be acceptable. Figure 2.5 illustrates how the stratification height influences indoor air quality in the occupied zone for the idealized case of a DV system with only a heat source (person) in the space and a contaminant source (person s breathing) associated with the heat source [Skistad 1994]. The figure shows two typical vertical profiles of pollutants from a person s breathing. Normalized pollutant concentrations (c/c R ) are plotted vs. normalized room height (y/h), where c R is the concentration at the return grille near ceiling level and H is the height of the room. Both profiles demonstrate how a large increase in pollutant concentration occurs at the stratification height, with cleaner, less polluted air in the lower zone and higher pollutant concentrations in the upper zone. Profile A is produced by a lower airflow rate that results in a stratification height (SH-A) somewhat below head height of a standing occupant. By increasing the airflow rate (loads remain constant) the stratification 29

42 CHAPTER 2 ROOM AIR DISTRIBUTION height (SH-B) is raised above head height in profile B, producing improved indoor air quality at the breathing height. It has also been observed that the stratification height can be locally displaced about 0.7 ft (0.2 m) upward around a person [Nielsen 1996]. This represents the entrainment of cleaner air from lower levels in the room by the thermal plume rising around a person up to the breathing height. The characteristic vertical temperature profiles for DV systems will exhibit similar behavior to those shown in Figure 2.5 for pollutant concentrations. In general, temperatures are lower below the stratification height, increase at a higher rate across the stratification height, and are highest in the upper zone of the room. Measurements have demonstrated that this profile is very stable horizontally in a room, meaning that similar temperatures will be obtained at the same height throughout the space. The vertical temperature gradient is also independent of the location of heat sources in the room, as long as the height of the sources remains constant. The gradient is, however, strongly dependent on variations in height of the heat sources. The most efficient heat removal occurs for heat sources located higher in the space, such as overhead lighting fixtures. Unfortunately, when applied to office configurations in the U.S and other locations with high heat load densities (> Btu/h-ft 2 [30-40 W/m 2 ]) and reduced ceiling heights compared to industrial buildings, DV systems cannot satisfy the cooling demand without imposing excessive thermal stratification in the space and overly cool conditions near the floor. This subject is reviewed by Yuan et al. (1999), who suggest that cooling loads as high as 40 Btu/h-ft 2 (120 W/m 2 ) can be handled. However, to accomplish this, very high airflow rates would be required with obvious energy consumption implications. Another consideration in the design of DV systems is that higher airflow rates require larger diffuser inlet areas (to maintain the low inlet velocities). The availability of wall space for standard low side-wall DV diffusers may limit their application to higher airflow rates. One configuration for a floor-supply DV system has been described by Akimoto et al. [1995]. In summary, the stratification height depends primarily on the room airflow rate relative to the magnitude of the heat sources. Increasing the airflow rate or decreasing the cooling load will raise the stratification height, thereby improving indoor air quality and reducing thermal stratification in the occupied zone. On the other hand, decreasing the airflow rate or increasing the cooling load will lower the stratification height, potentially reducing ventilation performance and reducing ther- 30

43 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE mal comfort (due to increased stratification) in the occupied zone while reducing fan energy use (for a given load). 2.3 UFAD SYSTEMS UFAD and TAC systems differ from true DV systems primarily in the way that air is delivered to the space: (1) air is supplied at higher velocity through smaller-sized supply outlets, and (2) local air supply conditions are generally under the control of the occupants, allowing comfort conditions to be optimized. By introducing supply air with greater momentum, UFAD and TAC systems alter the behavior in the lower region of the space compared to DV systems by increasing the amount of mixing, increasing the temperature near the floor, and reducing the temperature gradient, all other conditions being equal. Although still the subject of ongoing research, this altered behavior in the lower region helps to explain why UFAD systems may have the potential to handle higher cooling loads than typical DV systems in spaces with 9- to 12-ft (2.7- to 3.7-m) ceiling heights (e.g., offices). At higher elevations in the room, above the influence of the supply outlets, the overall airflow performance is very similar to that of DV systems. Based on recent experimental results [Webster et al. 2002a, 2002b; Lin and Linden 2002; Yamanaka et al. 2002] and an extension of displacement theory, the following model, consisting of up to three distinct zones in the room, can be proposed to describe the room air distribution for UFAD systems UFAD Room Air Distribution Model Figure 2.6 shows a schematic diagram of typical airflow patterns in an UFAD system in an office environment. The diagram identifies the two characteristic heights in the room that define the three zones in the room: (1) the throw height (TH) of the floor diffusers and (2) the stratification height (SH), similar to that found in DV systems. As shown in the figure, UFAD diffusers typically create clear zones in their immediate vicinity, representing regions within which long-term occupancy is not recommended due to excessive draft and cool temperatures. However, when under direct individual control by the occupant a feature of UFAD and especially TAC systems these local thermal conditions may be acceptable and even desirable for short-term occupancy. There is a price for improving comfort conditions (at high load) as the increased mixing in the occupied zone diminishes the ventilation performance compared to DV systems. In any case, the control and opti- 31

44 CHAPTER 2 ROOM AIR DISTRIBUTION Figure 2.6 Underfloor air distribution system with diffuser throw below the stratification height. Figure 2.7 Comparison of typical vertical temperature profiles for underfloor air distribution, displacement ventilation, and mixing systems. 32

45 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE mization of stratification is crucial to system design and sizing, energyefficient operation, and comfort performance of UFAD systems. Figure 2.7 presents and compares typical vertical temperature profiles for the UFAD room air distribution model, displacement ventilation, and conventional overhead mixing systems. The profiles shown are representative of normal operating conditions and are intended to demonstrate key differences and similarities between the three air distribution methods, as discussed below. The UFAD profile is based on temperatures in a space outside of the direct influence of supply outlets (outside clear zones) and can vary significantly depending on several control factors (see Section 2.3.4) [Webster et al. 2002a]. In Figure 2.7, the nondimensional temperature, or temperature ratio, is plotted vs. room height, where T is the room air temperature as a function of height, T S is the supply temperature at the floor, and T E is the exhaust temperature at the ceiling. The linear profile for DV systems is based on the 50% rule of thumb that applies to rooms of conventional height and normal heating loads [Skistad 1994]. The temperature near the floor is assumed to be halfway between the supply and exhaust temperatures. The DV profile is assumed to join the UFAD profile at the stratification height. As long as the throw height of the UFAD diffusers is below the stratification height, the upper zone is assumed to perform in a similar manner for both of these systems (for the same room load to supply volume ratio). The mixing system profile represents a uniformly well-mixed room with the temperature everywhere equal to the exhaust temperature Lower (Mixed) Zone. The lower mixed zone is directly adjacent to the floor and varies in depth according to the vertical projection of the (floor-based) supply outlets employed. The air within this layer is relatively well mixed due to the influence of high-velocity jets in the vicinity of the supply air outlets. The upper boundary of the lower zone coincides with the elevation at which the supply air reaches a terminal velocity of around 50 fpm (0.25 m/s). For TAC system applications having diffusers with horizontal projections in some cases, the top of this zone will be similarly defined as the height above which the supply outlets have negligible influence on room air movement. The greater mixing in this zone increases the temperature ratio near the floor to about 0.7 and reduces the gradient in comparison to DV systems. The lower mixed zone will always exist, although its height may vary greatly depending on the vertical projection of the supply outlet(s) and the ratio of the space heat load to the supply airflow to the space Middle (Stratified) Zone. The middle stratified zone is a transition region between the lower and upper zones of the room. The 33

46 CHAPTER 2 ROOM AIR DISTRIBUTION Figure 2.8 Underfloor air distribution system with diffuser throw above the stratification height. air movement in this zone is entirely buoyant, driven by the rising thermal plumes around convective space heat sources. The formation of these thermal plumes is uninhibited in this region, as air movement is not affected by supply air jets. The vertical temperature gradient in this zone tends to be greatest, approaching that for DV systems. The middle stratified zone only exists when the throw height of the supply outlets is below the stratification height, or upper boundary of the room, whichever is lower Upper (Mixed) Zone. The upper mixed zone is composed of warm (contaminated) air deposited by the rising heat plumes within the space. Although its average air velocities are generally quite low, air within this zone is relatively well mixed as a result of the momentum of thermal plumes penetrating its lower boundary. This zone is analogous to the upper zone found in spaces served by DV systems (compare Figures 2.2 and 2.6). Its bottom boundary, the stratification height, is primarily a function of the ratio of the space heat load to the supply airflow rate. As discussed below, if jets from the supply outlets penetrate into this zone, its depth (or even existence) may be affected, although if properly controlled this may be a secondary effect (Figure 2.8). In cases where the supply airflow rate is equal to or greater than the volume of the heat plumes generated within the space, the upper mixed zone will not form and the space may be modeled as a two-zone model, consisting only of the lower mixed and middle stratified zones. 34

47 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE Figure 2.9 Nondimensional temperature near the floor vs. room airflow rate. Experimental UFAD data taken from Webster et al. [2002a], DV results from Mundt [1990]. Tf = temperature near the floor; Ts = supply air temperature; Te = exhaust air temperature Temperature Near the Floor As shown in Figure 2.7, the greater mixing provided by turbulent supply outlets used in UFAD systems increases the temperature near the floor compared to DV systems (for the same supply air temperature and volume). This effect is shown more clearly in Figure 2.9, which plots the nondimensional temperature near the floor as a function of overall room airflow rate, where Tf is the temperature near the floor, Ts is the supply temperature at the floor, and Te is the exhaust temperature at the ceiling. The measurement heights for Tf are in the range of 3-4 in. (0.1 m). Experimental data for both swirl and variable-area floor diffusers are taken from Webster et al. [2002a]. The curve for DV systems is based on a large number of measurements in different rooms [Mundt 1990]. The results for UFAD systems show that the nondimensional temperature near the floor remains close to a constant level of 0.7 over a fairly wide range of airflow rates. For DV systems with minimal mixing by the supply diffusers, however, the nondimensional temperature near the floor gets relatively cooler (closer to the supply air temperature) as room airflow rate increases. This helps to explain the potential advan- 35

48 CHAPTER 2 ROOM AIR DISTRIBUTION tage that UFAD systems have over DV systems when trying to maintain comfort at higher heat loads. For the same room airflow rate, DV systems will need to use a higher supply air temperature than UFAD systems to avoid overly cool temperatures near the floor. Assuming all conditions are the same (heat load, supply airflow, and temperature), DV systems will produce higher stratification in the occupied zone compared to UFAD systems. The only way for DV systems to avoid excessive stratification at high heat loads is to increase the room airflow rate, a subject discussed by Yuan et al. (1999). While suggesting that cooling loads as high as 40 Btu/h-ft 2 (120 W/m 2 ) can be handled by DV systems, Yuan et al. also state that this requires sufficient space for large supply diffusers (often impractical in office configurations), and that the energy consumption will increase significantly Stratification Height In the same manner as for DV systems, the stratification height plays an important role in determining thermal, ventilation, and energy performance. Convective heat sources occurring at or above this height will rise up and exit the space without mixing into the lower zone, enabling lower airflow rates to be used for design load calculations compared to overhead mixing systems. Despite the existence of supply diffusers blowing higher velocity air into the occupied zone (primarily vertically for floor diffusers and horizontally for desk and partition diffusers), the stratification height is predominantly determined by the overall room air supply volume relative to the strength of heat sources in the space, and not (within limits) by the vertical throw of the diffusers [Nielsen 1996]. If the vertical throw is equal to or less than the stratification height (Figure 2.6), the only airflow crossing it will be due to buoyancy effects, similar to DV systems. In the limit as throw and the amount of mixing are reduced, UFAD systems tend to approach the operation of DV systems. If the diffuser throw is close to the stratification height or already exceeds it, the cooler supply air will penetrate into the warmer upper layer before dropping back down into the lower region and bringing warm air down with it (Figure 2.8). Although still a subject of ongoing research, recent results indicate that as long as the diffuser throw does not penetrate too far into the upper zone (up to 7 ft [2.1 m] in a 10-ft [3-m] high room), relatively similar comfort conditions will be produced in the occupied zone in comparison to diffusers with lower throws [Webster et al. 2002a]. The amount of air brought down influences the temperatures in the lower region and can also increase the stratification height, but this is a secondary effect. Higher throws that penetrate above the stratification 36

49 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE height will result in slightly warmer temperatures and less gradient in the lower region, all other conditions being constant. Ongoing research will investigate the effects of diffuser throw on thermal performance in greater detail [Lin and Linden 2002]. In the limit when a very strong supply air jet penetrates far into the upper zone, it is possible to disrupt the stably stratified airflow pattern in the space. For example, previous laboratory experiments [Bauman et al. 1991a; Fisk et al. 1991] demonstrated that when a fan-driven floor supply module was operated at higher air supply volumes, the cool supply jets were able to reach the ceiling, thereby minimizing stratification and producing close to uniform ventilation conditions. This operating strategy of providing a well-mixed space would reduce or eliminate the potential improvements in energy and ventilation performance described above. To avoid eliminating a stably stratified space with UFAD systems, maximum vertical throws of diffusers should be limited to no closer than 2-3 ft ( m) from the ceiling. To achieve optimal performance, it is recommended that diffuser throw heights be closer to the head height of occupants in the space Controlling Stratification Recent laboratory experiments have investigated the thermal stratification performance of UFAD systems using floor diffusers [Webster et al. 2002a, 2002b]. Figure 2.10 shows the impact of variations in total room airflow on stratification for swirl diffusers operating in a simulated interior space with total heat input of 18 Btu/h-ft 2 (56 W/m 2 ) and a supply air temperature of 64 F (18 C). The figure illustrates how stratification increases when room airflow is reduced for constant heat input. The figure also demonstrates how a control strategy might be approached to optimize stratification performance. At the highest flow rate of 1 cfm/ft 2 (5 L/s/m 2 ), the temperature profile exhibits only a small amount of stratification with a head-foot temperature difference of 1.3 F (0.7 C). This would represent a case where the space is being over-aired. On the other hand, at the lowest flow rate of 0.3 cfm/ft 2 (1.5 L/s/m 2 ), the head-foot temperature difference has increased to 6.8 F (3.8 C), exceeding the limit of 5 F (3 C) specified in ASHRAE Standard 55 [ASHRAE 1992]. This temperature profile demonstrates the sensitivity to changes in airflow rate, although it is highly unlikely that a system with cooling loads of this magnitude would be operated at such a low airflow rate. To improve energy performance (reduce airflow) while maintaining thermal comfort (avoiding excessive stratification), the middle profile at a flow rate of 0.6 cfm/ft 2 (3 L/s/m 2 ) may be a reasonable target, as it has a head-foot temperature difference of 37

50 CHAPTER 2 ROOM AIR DISTRIBUTION Figure 2.10 Effect of room airflow variation at constant heat input, swirl diffusers, interior zone. Figure 2.11 Effect of supply air temperature variation at constant heat input, interior zone. 38

51 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE 3.2 F (1.8 C). The difference between the middle and first profiles also demonstrates that despite a 40% reduction in airflow rate, the temperature in the space only increases by about 1 F (0.5 C) up to a height of nearly 4 ft (1.2 m). Figure 2.11 shows test results from Webster et al. [2002b] where supply air temperature (SAT) was varied over the range of F (16-19 C) for constant heat input (19 Btu/h-ft 2 [59 W/m 2 ]) and room airflow rate (0.5 cfm/ft 2 [2.7 L/s/m 2 ]) for a simulated interior space. As shown, the temperatures of the profiles increase or decrease with the change in supply air temperature but retain approximately the same shape. Resetting SAT may be advisable in combination with adjustments in total room airflow to achieve optimal comfort conditions throughout the occupied zone. 39

52

53 Chapter 3 Thermal Comfort and Indoor Air Quality Thermal comfort and indoor air quality are two of the leading factors in determining the success of a building s HVAC system performance. As described in Chapter 2, the traditional design solution in the vast majority of commercial buildings has been to use an overhead air distribution system that attempts to maintain close to uniform temperatures and ventilation air throughout the conditioned space. In this arrangement, supply diffusers are positioned at regular intervals on the ceiling, far from the individual occupants, and, particularly in open plan offices, each control zone will contain several diffusers and a significant number of occupants. This control strategy provides little opportunity to satisfy different thermal preferences among the building occupants (Figure 3.1) or to provide preferential ventilation in the occupied zone. In contrast, underfloor air distribution (UFAD) systems deliver conditioned air directly into the occupied zone of the building close to the occupant. UFAD and, in particular, TAC systems provide an opportunity for individuals to have some amount of control over their local environment (Figure 3.2). This section focuses on TAC systems, due to their strong potential for improved thermal comfort and ventilation performance over other system configurations. UFAD systems with floor diffusers will provide most of the same benefits, although generally at a lower level than TAC systems. Thermal comfort research and standards are discussed and recent ventilation research is also briefly reviewed. Test results are presented that define the occupant cooling performance of three TAC diffusers (two desk-based and one floor-based). The potential benefits and implications of personal comfort control and improved indoor air quality, including increased worker productivity, are discussed. 41

54 CHAPTER 3 THERMAL COMFORT AND INDOOR AIR QUALITY Figure 3.1 Conventional overhead air distribution system. Figure 3.2 Underfloor air distribution system. 42

55 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE 3.1 THERMAL COMFORT STANDARDS Current comfort standards, see ASHRAE Standard [ASHRAE 1992] and ISO Standard 7730 [ISO 1994], specify a comfort zone on the psychrometric chart, representing the optimal range and combinations of thermal factors (air temperature, radiant temperature, air velocity, humidity) and personal factors (clothing and activity level) with which at least 80% of the building occupants are expected to express satisfaction. These standards are based on a large number of laboratory studies in which subjects (primarily university students) were asked to evaluate their comfort in steady-state environments over which they had little or no control. The standards were developed for mechanically conditioned buildings typically having overhead air distribution systems designed to maintain uniform temperature and ventilation conditions throughout the occupied space. Given the high value placed on the quality of indoor environments, it is rather astonishing that a building s HVAC system can be considered in compliance with thermal comfort standards and yet provide a thermal environment with which up to 20% of the building population will be dissatisfied. This is, however, exactly the case in the conventional one-size-fits-all approach to environmental control in buildings. The primary scientific justification for this seemingly low level of occupant satisfaction is clearly revealed in the large body of thermal comfort research on human subjects in a laboratory setting. These tests, which form the basis for the ASHRAE Standard 55 comfort zone, demonstrate that on average at least 10% of a large population of subjects will express dissatisfaction with their thermal environment, even when exposed to the same uniform thermal environment considered acceptable by the majority of the population. In practice, the standard uses a 20% dissatisfaction rating by adding an additional safety factor of 10% dissatisfaction that might arise from locally occurring non-uniform thermal conditions in the space (e.g., stratification, draft, radiant asymmetry). Furthermore, there is an ongoing debate about the degree of relevance of laboratory-based research for occupants in real buildings, where the range of individual thermal preferences will likely be even greater (see discussion below). The bottom line is that a conventionally designed HVAC system using overhead air distribution may result in a surprisingly large number of occupants who are not satisfied with the thermal environment. Air velocity is one of the six main factors affecting human thermal comfort. Because of its important influence on skin temperature, skin wettedness, convective and evaporative heat loss, and thermal sensa- 43

56 CHAPTER 3 THERMAL COMFORT AND INDOOR AIR QUALITY tion, it has always been incorporated into thermal comfort standards on a not to exceed basis. In ASHRAE Standard 55, there are two recommendations for allowable air velocities in terms of (1) minimizing draft risk and (2) providing desirable occupant cooling [Fountain and Arens 1993]. The elimination of draft is addressed by placing rather stringent limits on the allowable mean air speed as a function of air temperature and turbulence intensity (defined as the standard deviation of fluctuating velocities divided by their mean for the measuring period). As an example, the draft risk data (representing 15% dissatisfaction curves) for a turbulence intensity of 40% (typical of indoor office environments with overhead mixing systems) would restrict the mean air speed to 24 fpm (0.12 m/s) at 68 F (20 C) and 40 fpm (0.2 m/s) at 78.8 F (26 C). Although still under debate, the draft risk velocity limits in Standard 55 appear to be most suitable for eliminating undesirable air movement under cooler (heating mode) environmental conditions, a more frequent situation in European climates. In warmer climates, such as those frequently found in the U.S., air motion is considered as highly desirable for both comfort (cool breeze for relief) and air quality (preventing stagnant air). ASHRAE Standard 55 allows local air velocities to be higher than the low values specified for draft avoidance if the affected occupant has individual control over these velocities. By allowing personal control of the local thermal environment, TAC systems satisfy the requirements for higher allowable air velocities contained in Standard 55 and have the potential to satisfy all occupants. 3.2 PERSONAL CONTROL One of the greatest potential advantages of TAC systems over conventional overhead systems is in the area of occupant thermal comfort, as individual preferences can be accommodated. In every work environment, there are significant variations in individual comfort preferences due to differences in clothing, activity level (metabolic rate), body weight and size, and individual preferences. In terms of clothing variations, if a person reduced their level of clothing from a business suit (0.9 clo) to slacks and a short-sleeved shirt (0.5 clo), the room temperature could be increased by approximately 4 F (2 C) and still maintain equivalent comfort. As an example of the variations in activity level that commonly occur, a person walking around continuously in an office (1.7 met) will experience an effective temperature of the environment that is approximately 3 F to 5 F (2 C to 3 C) warmer than that 44

57 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE Figure 3.3 Test configuration for manikin experiments of local cooling from TAC diffusers (D = desktop, U = underdesk, F = floor); perspective view. for a person sitting quietly at a desk (1.0 met), depending on clothing level. Among the floor or furniture-mounted diffusers that are commercially available, active (fan-powered) diffusers are the most effective at providing a wide range of control, particularly jet diffusers that deliver air with a directional component (compared to swirl diffusers). Recent laboratory tests have investigated the occupant cooling capacity of several desk-based and floor-based fan-powered jet supply outlets [Tsuzuki et al. 1999; Bauman et al. 1999b; Bauman et al. 2000b]. Results are shown for three fan-powered TAC diffusers, pictured in Figure 3.3: (1) two desktop diffusers, (2) underdesk diffuser, mounted under the desk surface in the kneespace, at the front edge of the desk [Levy 2002], and (3) floor jet diffuser, featuring four grilles mounted in one raised floor panel. Test results are presented in Figures in terms of prediction models for whole-body sensible cooling rates ( EHT) as a function of maximum air velocity near the person and room-supply temperature difference. The results shown in all cases are for a fixed supply air 45

58 CHAPTER 3 THERMAL COMFORT AND INDOOR AIR QUALITY Figure 3.4 Sensible whole-body cooling rates, EHT ( F), for two desktop jet diffusers blowing air toward a person seated in front of desk. Results applicable to average room temperatures of 72 F to 79 F (22 C to 26 C), room-supply temperature differences of 0 F to 13 F (0 C to 7 C), and supply velocities of 55 to 370 fpm (0.28 to 1.89 m/s). Velocity measured in front of chest of test manikin. direction toward the person seated at the desk. Note that occupants can typically adjust the cooling rate from TAC diffusers such as these by changing both the airflow rate and supply air direction. All three models provide good fits to the test data with R 2 in the range of EHT, or the change in equivalent homogeneous temperature [Wyon 1989], represents the amount of whole-body cooling provided by a diffuser, compared to still-air conditions at the same average room temperature. By presenting results in terms of the air velocity measured where the diffuser air jet hits the person, the results can also be applied to supply outlets that deliver air from generally the same direction (desktop, underdesk, or floor). The results indicate that for the range of test conditions investigated, these outlets can provide personal cooling control of equivalent whole-body temperature over a sizable range: up to 13 F (7 C) of sensible cooling for desktop-mounted outlets, up to 7 F (4 C) of sensible cooling for underdesk outlets, and up to 9 F (5 C) of sensible cooling for floor-based outlets. This amount of control is 46

59 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE Figure 3.5 Sensible whole-body cooling rates, EHT ( F), for underdesk jet diffuser blowing air toward a person seated in front of desk. Results applicable to average room temperatures of 79 F to 82 F (26 C to 28 C), roomsupply temperature differences of 0 F to 13 F (0 C to 7 C), and supply velocities of 60 to 570 fpm (0.31 to 2.90 m/s). Velocity measured 1 ft (0.3 m) in front of diffuser. clearly more than enough to allow individual thermal preferences to be accommodated. The results presented in Figures are examples of how cooling performance changes as a function of velocity, temperature, and diffuser configuration. Please refer to manufacturers for product-specific performance data. In addition to sensible cooling, evaporative cooling rates caused by air motion over a person with wet skin can be significant. For a person having a typical skin wettedness of 0.20, evaporative heat loss can more than double the sensible whole-body cooling rates shown in Figures Swirl diffusers have not been tested under these same test conditions, but they will not provide as much direct occupant cooling as the jet-type diffusers described above will. Swirl diffusers are designed to provide rapid mixing with the room air and thus minimize any highvelocity air movement, except within a small imaginary cylinder (approximately 3 ft [1 m] in diameter) directly above the floor diffuser. 47

60 CHAPTER 3 THERMAL COMFORT AND INDOOR AIR QUALITY Figure 3.6 Sensible whole-body cooling rates, EHT ( F), for fanpowered floor jet diffuser blowing air toward a person seated approximately 1 m (3 ft) to the side. Results applicable to average room temperatures of 72 F to 79 F (22 C to 26 C), room-supply temperature differences of 0 F to 13 F (0 C to 7 C), and supply velocities of 50 to 240 fpm (0.25 to 1.21 m/s). Velocity measured near arm of test manikin on side toward diffuser. Unless an occupant chooses to move within this cylinder, often referred to as the clear zone, room air velocities will be less than 50 fpm (0.25 m/s). As further support for the benefits of providing personal control, recent field research has found that building occupants who have no individual control capabilities are twice as sensitive to changes in the temperature of their environment compared to occupants who do have individual thermal control [Bauman et al. 1998; de Dear and Brager 1998]. What this indicates is that people who know they have control are more accepting of and in fact prefer a wider range of temperatures, making it easier to satisfy their comfort preferences. Research in this area has led to a proposal for an adaptive model of thermal comfort (based on field observations in naturally ventilated buildings) that will be added to the newly revised Standard 55 when it is released to augment the laboratory-based predictive models currently in widespread use [de Dear and Brager 1998; Brager and de Dear 2000]. 48

61 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE Recently, laboratory studies have investigated the human response to personalized ventilation and individual control, as provided by deskmounted TAC diffusers [Kaczmarczyk et al. 2002; Zeng et al. 2002]. The promising performance benefits demonstrated by these studies provide good reasons for further research to more accurately quantify the impact of providing personal control in TAC systems. 3.3 THERMAL STRATIFICATION Thermal stratification results in the air temperature at head level being warmer than at ankle level. ASHRAE Standard 55 specifies a maximum allowable vertical air temperature difference of 5 F (3 C) between heights of 67 in. (1.7 m) and 4 in. (0.1 m) [ASHRAE 1992]. However, as discussed by Wyon [1994a], the research upon which this recommendation is based was carried out in a test chamber in which each subject was given individual control of the average space temperature (thermal gradient was maintained constant). The implication is that if people do not have individual control, they may report local thermal discomfort at a higher rate than expected, even when exposed to a stratified environment within Standard 55 specified limits, a finding observed by Wyon and Sandberg [1990] for a displacement ventilation system. UFAD and TAC systems that provide personal control, however, can be expected to achieve the same low proportion of dissatisfied (5%) as in the original thermal gradient experiments. 3.4 VENTILATION PERFORMANCE Research to date has shown that UFAD and TAC systems using floor diffusers can provide modest increases in ventilation performance compared to overhead mixing systems [Fisk et al. 1991; Yokoyama and Inoue 1991, 1994; Faulkner et al. 1995; Tanabe and Kimura 1996]. Oguro et al. (1995) describe a field study in which the performance of an underfloor air-conditioning system on one floor was compared to the performance of a ceiling-based air-conditioning system on another floor of the same building. In this field study, airborne particle concentrations were significantly lower for the underfloor air conditioning system. Laboratory experiments with desktop-based diffusers have shown that the ventilation efficiency can be improved significantly in comparison to mixing-type air distribution at the worker s breathing level in the occupied zone when the percent of outside air is high and when supply air is directed towards the work location at a low velocity to reduce mixing [Faulkner et al. 1993, 1999]. Faulkner et al. [2002] found that the air change effectiveness at breathing level produced by 49

62 CHAPTER 3 THERMAL COMFORT AND INDOOR AIR QUALITY Figure 3.7 Local air motion improves the perceived air quality. an underdesk diffuser supplying 100% outside air ranged from 1.4 to 2.7. Recently, Melikov et al. (2002) have conducted an extensive series of experiments using a breathing manikin to investigate personalized ventilation for five different desk-mounted supply outlets. They report ventilation effectiveness values of 1.3 to 2.4. These values are higher than reported for commercially available task ventilation or displacement ventilation systems. Even if it is difficult to measure large improvements in ventilation performance for UFAD systems using rapidly mixing floor diffusers, it is generally accepted that by delivering fresh supply air near the occupants and giving them some amount of control, the occupants will perceive an improvement in indoor air quality (Figure 3.7). 3.5 PRODUCTIVITY Research evidence suggests that occupant satisfaction and productivity can be increased by giving individuals greater control over their local environment. In one of the first widely publicized productivity studies of TAC systems, Kroner et al. (1992) analyzed routinely collected worker performance data for an insurance company both before and after moving from an older conventional office building into a new 50

63 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE headquarters building having underfloor air distribution, with each workstation equipped with the desktop diffuser and personal control system described earlier in this section. They concluded that this desktop TAC system was responsible for a 2.8% increase in worker productivity. In a review and analysis of previous research, Wyon (1996) estimates that even under the conditions of thermal neutrality, the provision of individual control of local cooling and heating equivalent to ±5 F (±3 C) can improve group work performance by 2.7% for thinking tasks, 7% for typing tasks, 3.4% for skilled manual tasks, and 8.6% for the speed of individual finger movements. This is because the average neutral temperature cannot satisfy all occupants, whose individual thermally neutral points vary substantially. If the room temperature is raised above thermal neutrality, Wyon estimates the performance improvement to be significantly higher. Another more recent review of relevant research has concluded that improvements in productivity in the range of 0.5% to 5% may be possible when the thermal and lighting indoor environmental quality is enhanced [Fisk 2000]. In a recently completed intervention study, researchers investigated the relationship between ventilation rates and work performance in a call center operated by a health maintenance organization (HMO) [Federspiel et al. 2002; Fisk et al. 2002]. The call center was housed in an open plan office building served by a conventional overhead air distribution system. The study is significant in that the productivity data used in the analysis were obtained from the automated call distribution system operated by the HMO. The agents at this call center perform knowledge-based work by receiving inbound calls from clients and responding by providing medical advice. The handling of each call involved two tasks: talk and wrap-up. Among the results from this study, agents were found to be 16% slower at wrap-up when the measured room temperature was greater than 77.8 F (25.4 C). The provision of local cooling under individual control by TAC systems would be expected to reduce this observed negative impact of elevated room temperatures. 51

64

65 Chapter 4 Underfloor Air Supply Plenums The use of an underfloor plenum to deliver conditioned air directly into the occupied zone of the building is one of the key features that distinguish underfloor air distribution (UFAD) systems from conventional ducted overhead air distribution systems. In the design of underfloor air supply plenums, the primary objective is to ensure that supply air at the required quantity and conditions (temperature and humidity), and containing at least the minimum amount of ventilation air, is delivered wherever it is needed on the floor plate of the building. This process differs from fully ducted designs because as the air passes through the plenum, it can come in direct contact with thermally massive materials (concrete slab and floor panels) that will transfer heat to or from the supply air, depending on the temperature difference and flow rate. In some configurations, the amount of air reaching the desired locations may be influenced by plenum inlet conditions, plenum height, obstructions within the plenum, and leakage from the plenum. These and other design and performance considerations for underfloor plenums are discussed below. 4.1 DESCRIPTION An underfloor plenum is the open service distribution space between a structural concrete slab and the underside of a raised, or access, floor system (Figure 4.1a). As shown in the photo in Figure 4.1b, the raised floor platform is made up of 2 ft 2 ft (0.6 m 0.6 m) steel panels filled with concrete-like material (other compositions and finishes are available). The floor panels are attached and supported at each corner with a screw into the head of an adjustable pedestal that is glued to the concrete slab. Although not shown in Figure 4.1, horizontal stringers between pedestals and sometimes additional diagonal seismic 53

66 CHAPTER 4 UNDERFLOOR AIR SUPPLY PLENUMS Figure 4.1a Schematic diagram of raised (access) floor system. Figure 4.1b Typical installation of a raised floor system on a concrete slab, forming an underfloor plenum. bracing may be added for plenums of greater height (usually above 18 in. [0.45 m]). Underfloor plenums have been used for years as an access route for power, voice, and data cabling. In this arrangement, the cables are installed using modular connections to outlet boxes located in floor panels or system furniture. By providing easy access to make changes to the modular cabling system (by temporarily removing floor panels) and by enabling floor outlets to be located anywhere on the floor plate (by relocating, removing, or adding panels), cabling services can be 54

67 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE Figure 4.2 Installation of raised floor system in open plan office. reconfigured at great savings. When underfloor air distribution is added to an underfloor cable management system, creating a truly integrated service plenum, the same flexibility afforded the cabling system is now available to the HVAC system. Although raised floor plenum heights can be as low as 5 in. (127 mm) for modular cabling systems alone, when combined with UFAD systems, typical plenum heights are in. ( m). Figure 4.2 shows a typical installation in an open plan office with some of the floor panels removed to reveal the underfloor plenum. As shown, carpet tiles are the most common finished floor covering in office environments. Chapter 5 further discusses raised floor and carpet tile products. When designing an underfloor air supply plenum, there are three basic approaches to distributing air through it: 1. Pressurized plenum with a central air handler delivering air through the plenum and into the space. 2. Zero-pressure, or neutral, plenum with air delivered into the conditioned space through local fan-powered (active) supply outlets in combination with the central air handler. 3. In some cases, ducted air supply through the plenum to terminal devices and supply outlets. The designs that are installed often end up as hybrid solutions, including some combination of the above configurations. This guide 55

68 CHAPTER 4 UNDERFLOOR AIR SUPPLY PLENUMS will focus on the first two approaches, as guidelines for fully ducted air distribution systems are well established Pressurized Plenum In pressurized plenums, the central air-handling unit (AHU) is controlled to maintain a small, but positive pressure in the underfloor plenum relative to the conditioned space. Typical plenum pressures fall in the range of in. H 2 O ( Pa). To date, pressurized plenums have been the most commonly installed UFAD configuration. In most practical situations, pressurized plenums can maintain a very constant plenum pressure across a single control zone [Bauman et al. 1999a]. This allows any passive diffuser of the same size and control setting (typical damper opening) located in the zone to deliver the same amount of air to the space. However, airflow performance can be impacted by uncontrolled air leakage and when floor panels are removed for access to the underfloor plenum. See Sections 4.2 and 4.3 for further discussion. When the supply air flows freely through the underfloor plenum, heat exchange with the structural mass (concrete slab and raised floor panels) may influence supply temperature variations as a function of distance traveled through the plenum, as well as other thermal performance issues. These are discussed in Section Zero-Pressure Plenum In zero-pressure plenums, the central AHU delivers conditioned air to the underfloor plenum in much the same manner as with pressurized plenums, but in this case, the plenum is maintained at very nearly the same pressure as the conditioned space. Local fan-powered (active) supply outlets are required to supply the air into the occupied zone of the space. To date, several zero-pressure plenum designs have been installed, but they have not enjoyed the same amount of market penetration as pressurized systems. In terms of airflow performance, fan-powered outlets provide improved control of the supply airflow rate compared to passive diffusers. Active diffusers are well-suited for task/ambient conditioning (TAC) system applications in which occupant control is a key design objective. The removal of floor panels in zero-pressure plenums will not impact airflow performance. Similarly, zero-pressure plenums pose no risk of uncontrolled air leakage to the conditioned space, adjacent zones, or outside. The advantages of no leakage and improved local airflow control must be traded off against several factors. Fan-powered supply outlets 56

69 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE may have a cost premium compared to passive diffusers used in pressurized plenum designs. In terms of energy use, although central fan energy consumption will be reduced compared to that for a pressurized plenum, this savings will be offset by the energy consumed by the large number of small local fans. However, if a pressurized plenum leaks at a high rate, this can also lead to excessive fan energy use. Another consideration with local fan-driven units is the possibility of increased noise levels, although underfloor systems are generally rated as being quieter than conventional overhead systems. Zero-pressure plenums share many of the same thermal performance issues as pressurized plenums. See Section 4.4 for further discussion. 4.2 AIRFLOW PERFORMANCE IN PRESSURIZED PLENUMS A series of experiments [Bauman et al. 1999a] were conducted in a full-scale (3,200 ft 2 [300 m 2 ]) pressurized underfloor plenum test facility to assist in the development of design guidelines for acceptable airflow performance, plenum size, plenum inlets, and the effects of obstructions (i.e., cables, ductwork, equipment) within the floor cavity. Results of these tests are presented in the subsequent sections Dimensional Constraints of the Plenum One of the objectives of the aforementioned study was to identify the minimum plenum height at which acceptable air distribution throughout the plenum could be expected. Tests (see Figures 4.3 and 4.4) performed on plenums varying from 3 in. (75 mm) to 8 in. (205 mm) in height indicate that good distribution within the plenum can be achieved with plenum heights as low as 4 in. (100 mm). However, lowheight plenums (approaching 4 in. [100 mm] deep or lower) should be limited to applications where space airflow requirements do not exceed 1 cfm/ft 2 (5.1 L/(s m 2 )), as outlet distribution variances in excess of ±10% were experienced. In the figures, airflow data are presented in terms of delivered airflow ratio vs. distance from the plenum (fan) inlet, which was located at one end of the plenum. The delivered airflow ratio is defined as the measured airflow normalized by the amount of airflow that would be delivered if it were uniformly distributed across the plenum. In other words, a perfectly uniform distribution of air delivery through all floor outlets would yield a delivered airflow ratio of 100% at all distances from the inlet. 57

70 CHAPTER 4 UNDERFLOOR AIR SUPPLY PLENUMS Figure 4.3 Distribution of delivered airflow for 8-inch (205-mm) plenum. Figure 4.4 Distribution of delivered airflow for 4-inch (100-mm) plenum. 58

71 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE The maximum footprint of the plenum can be determined in relation to the feasible number (and size) of inlet locations. Plenum inlet sizes and locations are discussed in the following sections Plenum Inlets The maximum practical distance between the point where conditioned air is injected into the open underfloor plenum and its point of discharge into the space is generally determined by: 1. The degree of thermal decay experienced by the air as it moves to the supply outlet. 2. The residence time of the conditioned air within the open floor cavity. While resident within the underfloor plenum, the conditioned air is subject to heat transfer from the building slab, as well as the room (by means of the raised floor panels). This thermal transfer rate, discussed in greater detail in Section 4.4, generally limits the distance through which the conditioned air may travel according to its maximum allowable temperature rise. Although additional research is needed in this area, designers familiar with underfloor system design typically employ as a guideline a F temperature gain per linear foot of travel ( C/m), resulting in a maximum practical distance of ft (15-18 m) between the plenum inlet and point of discharge into the space Horizontal Ducting within the Plenum Horizontal ductwork and air highways (discussed in greater detail in Section 4.4.2) may be used to bridge the distance between the point of injection into the plenum and the farthest supply outlet. If employed, the velocities in these conduits should be limited to a maximum of 1,200-1,500 fpm (6-7.5 m/s). Outlets can be located along the length of the duct (or air highway) to optimally allocate the air within the plenum. The discharge velocity through these smaller outlets should, however, be limited to 800-1,000 fpm (4-5 m/s). The placement of balancing dampers in these discharge outlets should also be considered to avoid variances in the plenum distribution Obstructions within the Plenum Tests with obstructions located within the plenum airflow indicated that acceptable airflow performance (static pressure variations of ±10% or less) was experienced as long as a 3-in. (75-mm) clear space was maintained perpendicular to the airflow through the plenum (see Figure 4.5). 59

72 CHAPTER 4 UNDERFLOOR AIR SUPPLY PLENUMS Figure 4.5 Distribution of delivered airflow for different obstructions in 8-inch (205-mm) plenum (1.5 cfm/ft 2 [7.6 L/ (s m 2 )]). 4.3 AIR LEAKAGE While air leakage is not an issue for zero-pressure plenum designs, evidence from completed projects using pressurized plenums indicates that uncontrolled air leakage from the plenum can impair system performance [Daly 2002]. If this leakage occurs across the building envelope it will directly impact energy use. If the leakage occurs within the building it may or may not impact energy use depending on where the leakage takes place. In any case, it is highly recommended to minimize leakage from the plenum and, when it is unavoidable, to account for the leakage airflow rate in the operation of the system, as discussed further below. There are two primary types of uncontrolled air leakage from a pressurized underfloor plenum: (1) leakage due to poor sealing or construction quality of the plenum and (2) leakage between floor panels. A third type of leakage occurs when floor panels are removed for access to the plenum, but this is usually temporary Leakage Due to Construction Quality It is important that proper attention be given to the sealing of edge details all around the underfloor plenum, including window-wall con- 60

73 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE Table 4.1: Air Leakage Through Gaps Between Floor Panels (cfm/ft 2 ) [L/(s. m 2 )] Plenum Pressure Carpet Tile Configuration (in. H 2 O [Pa]) None Aligned Offset 0.05 [12.5]* 0.68 [3.5] 0.29 [1.5] 0.14 [0.7] 0.1 [25]** 0.96 [4.9] 0.41 [2.1] 0.20 [1.0] *measured; **estimated Figure 4.6 Airflow and leakage in a pressurized underfloor air supply plenum. nections to the slab, interior walls, along pipe chases, stair landings, elevators, and HVAC shaft walls during the construction phase of the project. Even if this is done, the integrity of a well-sealed underfloor plenum must be preserved over the lifetime of the building, as subsequent work can easily lead to new penetrations. If this is not done carefully, these types of leaks will be the most difficult to locate and fix later in the project. In most cases, designers can expect to encounter leakage losses of 10% to 30%, depending on quality of construction. See Section 8.2 for further discussion of this construction issue Leakage Between Floor Panels Leakage between floor panels, as depicted in Figure 4.6, is a function of the raised floor panel type and installation, carpet tile installation, and pressure difference across the plenum. Despite the relatively low pressures ( in. H 2 O [ Pa]) used in pressurized plenums, the large floor surface area makes this leakage an important consideration for design and operation. Table 4.1 lists air leakage data from recent tests conducted on a typical raised floor installation [Lee and Bauman (In press)]. Results represent the leakage through gaps between floor panels only, as all other 61

74 CHAPTER 4 UNDERFLOOR AIR SUPPLY PLENUMS Figure 4.7 Two different modes of carpet tile installation on raised floor panels: aligned (left) and offset (right). White lines indicate the floor panel edges. gaps in the plenum were sealed during the tests. Measured data are shown for a plenum pressure of 0.05 in. H 2 O (12.5 Pa) and for three different modes of floor covering: none (bare floor panels), aligned carpet tiles, and offset carpet tiles. No adhesive was used to install the carpet tiles during these tests so reported leakage values will be slightly conservative. As shown in Figure 4.7, aligned carpet tiles occur when the size and edges of the carpet tile match those of the floor panel (2 ft 2 ft [0.6 m 0.6 m]). Offset carpet tiles occur when the carpet tile is shifted over so that the edges are not aligned. The floor panels tested represent a design that is known to have the lowest leakage of most commercially available models. Experiments have shown that air leakage will vary approximately as the square root of plenum pressure [ASTM 2000]. Based on this relationship, the air leakage values for a plenum pressure of 0.1 in. H 2 O (25 Pa) can be estimated and are listed in Table 4.1. Please refer to manufacturer s test data for more precise data on air leakage rates for specific raised floor configurations. The magnitude of air leakage from a pressurized plenum shown in Table 4.1 is surprisingly high. The results indicate that the layer of carpeting plays an important role by significantly reducing air leakage rates between floor panels. The performance of a UFAD system with bare floor panels would be severely compromised if no additional means of sealing between panels were installed. Placing carpet tiles across the gaps between floor panels (offset mode) reduces the air leakage rate by 50% compared to aligned carpet tiles. Even with carpeting in place, the results suggest that minimizing leakage from other parts of the underfloor plenum should have a high priority. 62

75 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE Figure 4.8 Thermal decay in an underfloor air supply plenum. 4.4 THERMAL PERFORMANCE Thermal processes within the underfloor plenum and the surrounding thermal mass are known to have an important impact on the effectiveness of the plenum as part of the building s air distribution system. These processes include (1) heat transfer between the slab and the plenum air, (2) heat transfer between the floor panels and the plenum air, (3) variations in plenum air temperature with distance traveled through the plenum, and (4) thermal storage performance of the slab and floor panels. While the delivery of an adequate amount of air through the plenum can be quite reliable, it is more difficult to predict the thermal performance of underfloor plenums. Ongoing research is aimed at developing a thermal model for underfloor plenums as part of a wholebuilding energy simulation code [Bauman et al. 2000a]. Additional work is needed to develop design tools for practicing engineers Thermal Decay An important design consideration is to limit the amount of variation in supply air temperature, often referred to as thermal decay, to acceptable levels for diffusers that are located farther away from the plenum inlet. Figure 4.8 shows a schematic diagram of the most commonly occurring form of thermal decay in underfloor plenums. In cooling operation, cool supply air enters the plenum on the left. As it travels through the plenum it is warmed by heat transfer from the floor panels on the top (heat conducted from the room) and from the concrete slab on the bottom (heat conducted from the return air plenum for the floor below). Based on ongoing research [Bauman et al. 2000a], simplified thermal modeling [York International 1999], and the results of field measurements [Fukao et al. 2002], current estimates for the range of expected temperature increase with distance traveled through the plenum (for typical slab temperatures and airflow rates) are approximately 63

76 CHAPTER 4 UNDERFLOOR AIR SUPPLY PLENUMS Figure 4.9 Plan view of plenum airflow patterns: (a) without inlet vanes, (b) with inlet vanes F/ft ( C/m). Applying this estimate in practice is complicated by several variables, including the following: (1) the fact that the air may not travel in a straight line between the plenum inlet and the diffuser, (2) the number and location of plenum inlets employed, (3) the temperature difference between the plenum air and the slab and floor panels, and (4) the existence of return air directly entering the underfloor plenum. These are discussed briefly below. The example shown in Figure 4.9 indicates that it is possible to create an overall airflow pattern through the plenum zone that increases the distance and length of time that the air travels through the plenum before reaching some of the diffusers. Installing plenum inlet vanes to more uniformly spread the airflow across the full width of the plenum could be a simple solution if temperature measurements at different diffusers in the zone indicate that this is a problem. Adding additional plenum inlets or an air highway (see Section 4.2.3) is another approach for improving the thermal uniformity of the airflow distribution within the plenum. However, this must be traded off against the available access points for plenum inlets and the additional cost of the required ductwork. The amount of heat transferred through the floor panels and slab will have a direct impact on thermal decay in the underfloor plenum. 64

77 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE Recent laboratory experiments report heat transfer rates through a typical raised floor with carpet tiles in the range of W/ft 2 ( W/m 2 ) [Webster et al. 2002a]. Since room and plenum air temperatures are controlled, this floor heat transfer will remove heat from the room, thereby reducing the zone airflow requirements. As the heat is transferred to the supply air, it will still appear as a load at the system level. The heat transfer contribution through the concrete slab can be of similar magnitude to the floor heat transfer rate. Warm return air flowing along the underside of the slab structure on the next floor down will be the main source of heat driving this heat transfer process. Where heat loads are high and the room air is allowed to stratify, the resulting elevated return air temperatures will increase the load on the slab above. On the other hand, on ground floors with slabs-on-grade, single-story buildings, and lightly loaded buildings, heat transfer through the slab will be correspondingly reduced. Conductive heat gain from sunlit facades is another potential contributor to thermal decay. Large amounts of heat collected by the building skin and transferred to the directly coupled building slab can result in a significant rise in adjacent plenum air temperatures. In practice, plenum air temperature decays of as much as 10 F (6 C) have been observed in the outer 4 ft (1.2 m) of the plenum near these sunlit walls. As such, direct routing of plenum supply air through diffusers within a few feet (less than 1 m) of such facades is discouraged. The employment of fan-assisted (cooling) terminals discharging into a common (insulated) duct minimizes this problem by (1) drawing the supply air from several feet (~2 m) inside the plenum and (2) insulating the discharged air from the floor slab and warm façade. Chapter 9 discusses the advantages and liabilities of various control strategies for perimeter spaces. In some plenum configurations, return air may directly enter the plenum without being ducted back to the AHU where it can be easily mixed with the primary supply air. The most typical pathway for this to happen is in zero-pressure plenums with active diffusers, where room air may enter through passive equalizing floor grilles, when airflow demand at the fan-powered outlets exceeds the primary air quantity entering the plenum, creating a slightly negative-pressure plenum. In pressurized plenums, it is unlikely that room air will reenter the plenum through passive diffusers. If this occurs, the major concern is the impact of the warm return air on supply temperatures. The net effect may be unexpectedly high supply air temperatures, giving the impression that thermal decay is severe, when in fact the source of the temperature rise is the unmixed return air entering the plenum. 65

78 CHAPTER 4 UNDERFLOOR AIR SUPPLY PLENUMS Figure 4.10 Reduced thermal decay in an underfloor air supply plenum with pre-cooled thermal mass. It is difficult to predict the impact that the various factors described above have on thermal decay in an underfloor plenum. To date, most designers with experience in UFAD design are using a rule-of-thumb limiting the maximum distance from the plenum inlet to the farthest diffuser to about ft (15-18 m) in pressurized plenum designs. In zero-pressure plenum configurations with return air recirculating directly into the underfloor plenum, a previously used solution is to deliver the primary air at more frequently spaced intervals throughout the plenum. Shute (1995) recommends distribution intervals for this situation of no greater than 30 ft (9 m). Nighttime precooling of the thermal mass in the plenum can also partially offset some of the thermal decay by allowing higher plenum inlet temperatures to be used. As illustrated in Figure 4.10, when the exposed thermal mass in the plenum has been cooled to a lower temperature, the magnitude of the thermal decay will be reduced (compare to Figure 4.8). If properly implemented, this thermal storage control strategy allows the building to act as a fly wheel during daytime cooling periods, thereby saving energy and reducing energy costs (Section 7.5) Ductwork and Air Highways The use of ductwork and air highways to distribute supply air through parts of the underfloor plenum is another common method for controlling temperature variations in the plenum. In fact, early UFAD designs tended to be very conservative and resembled ducted overhead air distribution systems placed under the floor. These early designs featured a considerable amount of ductwork delivering air to temperature control zones that were defined by underfloor partitions. This approach made the thermal performance of the UFAD system easier to predict but cluttered up the plenum with ducts, partitions, and other HVAC equipment. In recent years, it has become recognized increasingly that 66

79 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE it is desirable to the extent possible to minimize the amount of installed ducts, air highways, and other HVAC-related components. The resulting open plenum can more easily serve as a highly flexible and accessible service plenum. While ductwork can isolate airflow from the thermal decay described in Section 4.4.1, air highways can still be influenced by heat transfer from both the slab and floor panels. Although no research to date has addressed this issue, designers should be aware that thermal decay may occur in long runs of air highways. 67

80

81 Chapter 5 Underfloor Air Distribution (UFAD) Equipment Due to growing interest in UFAD systems in the U.S., several new products have been introduced in recent years and this trend is expected to continue. In this section we briefly describe some of the products currently available, including both UFAD and TAC diffusers, underfloor fan terminals, and raised floor systems. Not all products are included as this is intended to provide an overview of the range and types of equipment obtainable. Product listings are provided for information only and do not constitute a recommendation or endorsement. Another recent review of UFAD and TAC equipment is provided by Loftness et al. [2002]. It is recommended that you contact the equipment manufacturers directly to obtain the most up-to-date product information. 5.1 SUPPLY UNITS AND OUTLETS Types of UFAD and TAC Diffusers Figure 5.1 is a schematic diagram showing five possible locations and types of supply diffusers that can be located within a typical workstation. All diffusers that are positioned near an occupant s work location should be controllable to some extent by the occupant. The most common occupant controls are velocity (volume) and/or supply air direction. Floor diffusers are installed as part of a standard UFAD system. As shown in Figure 5.1, diffuser #1 is a round swirl floor diffuser and #2 is a rectangular jet floor diffuser. Other floor diffusers are available as discussed further below. The most effective TAC diffusers are local fan-driven, jet-type diffusers that are located on the furniture in close proximity to the occupant. This configuration makes their controls easily accessible, allowing occupants to optimize their personal 69

82 CHAPTER 5 UNDERFLOOR AIR DISTRIBUTION (UFAD) EQUIPMENT Figure 5.1 UFAD and TAC diffuser locations in a workstation. comfort by controlling the quantity of air supply being directed toward them. In Figure 5.1, diffuser #3 is a desktop diffuser, #4 is an underdesk diffuser, and #5 is a partition-based diffuser. Refer to Chapter 3 for performance data on effective cooling rates for three different TAC diffusers. In all areas outside of workstations, the same floor diffusers shown in Figure 5.1 can also be used. Most manufacturers provide both passive and active diffusers. Passive diffusers are defined as air supply outlets that rely on a pressurized underfloor plenum to deliver air from the plenum through the diffuser into the conditioned space of the building. Active diffusers are defined as air supply outlets that rely on a local fan to deliver air from either a zero-pressure or pressurized plenum through the diffuser into the conditioned space of the building. Diffusers can be configured as constant air volume (CAV) or variable air volume (VAV). Although most diffusers have some form of manual control of supply volume (by controlling a damper), only those that provide automatic adjustment of the supply volume can be classified as true VAV diffusers. Additionally, in perimeter zone applications where automatic control is needed to respond to rapidly changing loads, linear floor grilles 70

83 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE Figure 5.2 Cutaway photo of passive swirl floor diffuser [Trox 2002]. are frequently installed, typically ducted from a fan-coil unit to provide cooling and heating when needed. Several ceiling-mounted supply outlets that allow some amount of personal adjustment have also been developed, but these are not included in this design guide. Please refer to Loftness et al. [2002] for information on these products Passive Swirl Floor Diffusers Figures show four different passive swirl diffuser designs that are commonly installed in UFAD systems. More models are available for this than for any other type of diffuser. The swirling airflow pattern of air discharged from this round floor diffuser provides rapid mixing of supply air with the room air up to the height of the vertical throw of the diffuser. Although the discharge pattern for most swirl diffusers is not adjustable, nearby occupants have limited control of the amount of air being delivered by rotating the face of the diffuser (Figures 5.2 and 5.3) or by opening the diffuser and adjusting a volume control damper (Figure 5.4). Different models and sizes are available, but the maximum flow rate for passive swirl diffusers operated at typical underfloor plenum pressures is cfm (40-47 L/s) at 0.08 in. H 2 O (20 Pa). Most models are equipped with a catch basin for dirt and liquid spills. Figure 5.4 shows the various components that are assembled to install a complete floor diffuser. Grilles are supported in the predrilled 71

84 CHAPTER 5 UNDERFLOOR AIR DISTRIBUTION (UFAD) EQUIPMENT Figure 5.3 Passive swirl floor diffuser [Nailor 2002]. Figure 5.4 Passive swirl floor diffusers [Price 2002]. 72

85 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE Figure 5.5 Passive variable-area (VAV) diffuser [York 2002]. hole through the carpet tile and raised floor panel with a trim ring on top and a retainer, or mounting ring, below. Two grille designs are shown in Figure 5.4. The one on the right with radial slots produces the standard swirl discharge pattern. The two on the left feature a combination grille with part (radial slots) producing the same swirl discharge pattern and part (circular slots) producing more of an inclined jet discharge pattern. The directional characteristics of the jet discharge allow an occupant to control the amount of air blowing toward them (for increased cooling) or away from them by rotating the grille Passive VAV Floor Diffusers Figure 5.5 shows a variable-area diffuser that is designed for VAV operation. It uses an automatic or manual internal damper to adjust the active area of the diffuser in order to maintain a nearly constant discharge velocity, even at reduced supply air volumes. At maximum flow, the diffuser is designed to deliver 150 cfm (71 L/s) at a plenum pressure of 0.05 in. H 2 O (12.5 Pa). This passive diffuser does not require a local fan, but 24-volt power is needed for the thermostatically controlled damper motor. Air is supplied through a slotted square floor grill in a jet-type airflow pattern. Occupants can adjust the direction of the supply jets by changing the orientation of the grille. Supply volume is controlled by a thermostat on a zone basis or as adjusted by an individual user. Figures 5.6 and 5.7 present two different approaches for converting swirl diffusers to VAV diffusers. Both require control power to auto- 73

86 CHAPTER 5 UNDERFLOOR AIR DISTRIBUTION (UFAD) EQUIPMENT Figure 5.6 Passive VAV swirl diffuser [Trox 2002]. Figure 5.7 VAV floor boot for swirl diffuser [Price 2002]. 74

87 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE Figure 5.8 Placement of linear grilles along exterior glazing in perimeter zones. matically adjust a volume control damper. In Figure 5.6, which is the same grille as Figure 5.2, the circular damper is raised and lowered to change the height of the opening for air to enter from a pressurized plenum. Figure 5.7 is a floor boot that is designed to be mounted on the underside of a raised floor panel containing a round swirl diffuser. The round control damper located in the inlet can be rotated over a range of 90 to adjust the inlet opening from fully open to fully closed. Supply air either enters the boot directly from a pressurized plenum or can be ducted from a fan-powered terminal unit Linear Floor Grilles Linear grilles have been used for many years, particularly in computer room applications. Air is supplied in a jet-type planar sheet, making them well matched for placement in perimeter zones adjacent to exterior windows (Figure 5.8). Although linear grilles may be configured as passive diffusers in a pressurized plenum, care should be exercised to prevent heat transfer from the building façade to the slab when 75

88 CHAPTER 5 UNDERFLOOR AIR DISTRIBUTION (UFAD) EQUIPMENT Figure 5.9 Photo and schematic of linear floor grille with VAV cooling and CAV heating (Price 2002). 76

89 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE Figure 5.10 Desktop TAC supply unit [Johnson Controls 2002]. using them in perimeter zones (see Section 4.4.1). Instead, perimeter zone applications should involve ducting the grilles from fan coil units or some other means of minimizing heat gain from the slab and warm façade. Linear grilles typically have multi-blade dampers that are not designed for frequent adjustment by individuals and are therefore not used in densely occupied office space where some amount of occupant control is desirable. Figure 5.9 shows a recently introduced linear floor grille for perimeter zone applications. The heating inlet (shown) is designed for ducted fan-powered supply air when there is a call for heating. The inlet also features a backdraft damper to prevent air supply when the fan is turned off. On the opposite side of the unit (not shown) is the cooling inlet. Cooling air supply is delivered directly from a pressurized plenum and is modulated in VAV mode by a control damper. The control damper closes down to the minimum opening on a call for heating. Control power is required for the VAV operation of the unit. The unit delivers up to 200 cfm (94 L/s) at 0.1 in. H 2 O (25 Pa) Active TAC Diffusers Figure 5.10 shows a TAC supply unit that features two desktop air supply pedestals, which are fully adjustable for airflow direction, as well as air supply volume by adjusting a control panel on the desk. Air 77

90 CHAPTER 5 UNDERFLOOR AIR DISTRIBUTION (UFAD) EQUIPMENT Figure 5.11 The occupant can control local environmental conditions using a desktop control panel [Johnson Controls 2002]. is supplied from a mixing box that is hung in the back or corner of the knee space of the desk and connected to the two desktop supply nozzles. The mixing box uses a small variable-speed fan to pull air from the underfloor plenum and deliver a free-jet-type airflow from the nozzles. The unit can supply a total of cfm (6-70 L/s) through its two nozzles. See Chapter 3 for occupant cooling performance data for this diffuser. Recirculated air is also drawn from the knee space through a mechanical prefilter. Both primary supply air and recirculated room air are drawn through an electrostatic air filter. As shown in Figure 5.11, the unit has a desktop control panel containing adjustable sliders that allow the occupant to control the speed of the air emerging from the nozzles, its temperature, the temperature of a 200-watt radiant heating panel located in the knee space, the dimming of the occupant s task light, and a white noise generator for acoustical masking. The control panel also contains an infrared occupancy sensor that shuts the unit off when the workstation has been unoccupied for a few minutes. Figure 5.12 shows an active (fan-driven) underdesk TAC diffuser consisting of a panel attached to the underside of a conventional desk, connected by a flexible duct to a portable filter and fan unit placed next to the desk (shown) or in the underfloor plenum. Airflow from two adjustable outlets at the front edge of the desk is used to condition the 78

91 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE Figure 5.12 Underdesk TAC supply unit [Johnson Controls 2002]. occupant s local workspace. One of the supply outlets delivers supply air directly upward into the occupant s breathing zone. The other delivers air toward the occupant s body for cooling purposes. The maximum airflow from the unit is only 15 cfm (7 L/s) because the air is concentrated directly on the occupant. Heating of the lower part of the body can be provided by a controllable radiant heating panel under the desk. Figure 5.13 shows a schematic diagram of another active (fandriven) underdesk TAC diffuser consisting of five 4-way adjustable grilles, similar to a car's dashboard. A fan unit located in the underfloor plenum delivers air through a flexible duct to two outlet locations: (1) the supply grilles (jet-type) mounted just below and even with the front edge of the desk, and (2) in this example, a supply grille located on the backside of the desk. This configuration permits a true task/ambient control strategy to be employed. The total air supply delivered to both supply outlets is thermostatically controlled to maintain overall comfort conditions in the ambient space. The amount of air supplied through the underdesk diffuser can be adjusted by occupants using a damper lever just behind the supply grilles to satisfy their personal comfort preferences. The underdesk diffuser is nominally designed to deliver 0-70 cfm (0-33 L/s) of supply air. See Chapter 3 for occupant cooling performance data for this diffuser. Other configurations using this same TAC control strategy are available (Figure 5.14). 79

92 CHAPTER 5 UNDERFLOOR AIR DISTRIBUTION (UFAD) EQUIPMENT Figure 5.13 Underdesk TAC supply unit [Argon Corporation 2002]. Figure 5.14 Alternative TAC supply outlet configurations [Argon Corporation 2002]. 80

93 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE Figure 5.15 Active TAC floor supply unit [Trox 2002]. Figure 5.15 shows an active (fan-driven) floor supply unit, consisting of two swirl diffusers mounted in a single raised floor panel. A wallmounted thermostat varies the speed of the fan (as shown mounted directly below the floor diffusers) to control the air supply to the space. These units can deliver up to 350 cfm (170 L/s) of supply air at maximum fan speed. They can also be configured with directional (jet type) diffusers that can be adjusted by the occupant to direct the airflow discharge. Partition-based diffusers, mounted in the partitions immediately adjacent to a desk, are another TAC diffuser design. A fan unit in the underfloor plenum delivers air through passageways that are integrated into the partition design to controllable supply grilles (jet-type) that may be located just above desk level or just below the top of the panel. Although uncommon in the U.S., some of these systems have been installed in Japan [Matsunawa et al. 1995]. 5.2 UNDERFLOOR FAN TERMINALS Underfloor fan terminal units are typically used in perimeter zones and other special zones where large and rapid changes in cooling and/ 81

94 CHAPTER 5 UNDERFLOOR AIR DISTRIBUTION (UFAD) EQUIPMENT Figure 5.16 Underfloor fan terminal with two variable-speed fans and hot water reheat [Greenheck 2002]. or heating load requirements occur. Figure 5.16 shows an example of one such terminal unit featuring two variable-speed fans (for added capacity) and a hot water reheat coil. Despite the larger profile of this unit, Figure 5.17 demonstrates how its installation, as well as a standard single-fan terminal unit, is compatible with the 2 ft 2 ft (0.6 m 0.6 m) grid of raised floor support pedestals in an underfloor plenum. The height of fan terminals must also be considered to allow adequate clearance below the raised floor panels. Units with heights as low as 8 in. (200 mm), requiring only a 10-in. (250 mm) high floor, are available. In addition, fan terminal units can be configured to serve a range of operational needs, including constant- and variable-speed fans with heating and cooling coils. Figure 5.18 shows a schematic diagram of one possible fan-powered terminal configuration serving a perimeter zone. The terminal unit with a hot-water or electric heating coil is used in combination with two or more variable-area VAV diffusers (Figure 5.5) to provide both cooling and heating operation. During cooling mode, the fan terminal is turned off, and all diffusers operate in normal VAV mode to deliver the desired amount of cool plenum air (full cooling is shown in Figure 5.18 (a)). During heating operation, the fan terminal is activated, pulling return air from the room through one diffuser and supplying air to the room through the other diffuser. Figure 5.18 (b) shows the diffuser dampers adjusted in full heating position, although an adjustable stop for the damper can be installed to provide minimum ventilation air from 82

95 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE Figure 5.17 Placement of single- and double-fan units between raised floor support pedestals [Greenheck 2002]. (a) Full cooling mode (b) Full heating mode Figure 5.18 Perimeter solution using heating fan terminal with VAV diffusers [York 2002]. 83

96 CHAPTER 5 UNDERFLOOR AIR DISTRIBUTION (UFAD) EQUIPMENT Figure 5.19 Perimeter zone installation of fan terminal unit with VAV cooling and reheat [Trox 2002]. the plenum. Under thermostatic control, room air provides the first stage of heating followed by activation and modulation of the heating coil. This perimeter solution requires no underfloor partitions. Figure 5.19 shows a photo of another possible underfloor fan terminal installation serving a perimeter zone. The fan operates only during periods of reheat and peak cooling demand. Under normal cooling operation, the VAV damper (on the left side of the unit) controls the amount of cooling from a pressurized plenum. The first stage of reheat occurs by closing the VAV damper so that the main source of air is recirculated room air entering through the floor grille at the lower left in the photo. Ventilation air is assumed to enter the space from adjacent overventilated spaces. Under peak heating conditions, an integral (hot water or electric) coil is energized. The terminal discharge at the top of the photo enters a small partitioned plenum zone containing typically 4-6 swirl diffusers (linear grilles could also be used). It is recommended that you contact the equipment manufacturers directly to obtain the most up-to-date information on fan terminal units and applications. Also, see Chapter 9 for further discussion of perimeter and special zone solutions. 84

97 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE Figure 5.20 Installation of a raised floor system creates an integrated service plenum. 5.3 RAISED FLOOR SYSTEMS A raised, or access, floor system is an elevated platform constructed on another floor, typically a concrete slab in a building. In North America, the raised floor platform is made up of 2 ft 2 ft (0.6 m 0.6 m) floor panels that are supported at their corners by adjustable pedestals. The installation of a raised floor system creates a convenient and accessible space that can be used to cost-effectively distribute many building services, including power, voice, data cabling, air-conditioning, fire detection and suppression, and security. Figure 5.20 shows an example drawing of cable and air distribution components in an underfloor plenum. The majority of raised floor systems currently on the market are engineered to meet the concentrated, uniform and rolling loads experienced in a typical workplace environment. For example, galvanized steel-encased lightweight concrete panels combine the tensile strength of steel with the compressive strength of concrete to offer a high degree of rigidity. The lower self weight of both lightweight concrete or steelencased high-density particle board panels reduces deflection even further and makes removing the panels an easier process than that for equivalent standard concrete-filled panels. High quality manufacturing processes enable panels made to very low dimensional tolerances such as ±0.15 in. (3.8 mm). Together with a uniform panel thickness, good edge sealing, and flush-mounted floor diffusers it is possible to achieve a homogeneous floor surface over which loads can be distributed. Currently the vast majority of UFAD installations in office environments use carpet tiles as the finish floor covering. The quality of manufactured carpet tile products has now advanced to the point where attractive and professional installations are possible, suitable for corporate offices. Raised floor systems typically include floor panels and 85

98 CHAPTER 5 UNDERFLOOR AIR DISTRIBUTION (UFAD) EQUIPMENT Figure 5.21 Adhesive is spread on the floor panels and allowed to set up prior to laying the carpet tiles. carpet tiles from different manufacturers, often in different modular sizes. Floor panels are 24 in. (600 mm) square, while carpet tiles are commonly 18 in. (450 mm), although they are also available in 24 and 36 in. (600 and 900 mm). This brings up two important considerations: (1) the method of securing tiles to panels and (2) the issue of alignment versus overlap of the floor panel and carpet tile edges. The majority of carpet tile installations are affixed to floor panels with adhesive (Figure 5.21). It is important to avoid using an excessive amount of adhesive during this process as the flexibility of easy removal of carpet tiles and access to floor panels can be compromised. Too much adhesive also risks bonding adjacent floor panels to each other and gluing the panel screws into their corner holes, both of which can complicate the removal of floor panels. Should the initial carpet tile type need to be replaced, building owners are left with an adhesive residue that must be removed before installation of an alternative tile system. The above issues have the potential to diminish some of the cost savings that raised floor systems offer, as they impede flexibility and ease of installation and maintenance. As an alternative, some manufacturers offer magnetic or dimpled carpet tiles that are indexed to 86

99 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE Figure 5.22 Non-adhesive carpet tile that is held in place with dimples indexed to matching holes on top of floor panel [Tate Access Floors 2002a]. matching holes on top of the underlying floor panels (Figure 5.22). These carpet tiles also exactly match the floor panel in size, allowing the greatest flexibility in repositioning service outlets to anywhere on the floor plate; only one carpet tile must be moved for each floor panel. By comparison, the more common use of 18-in. (450-mm) carpet tiles requires a minimum of 4-6 carpet tiles to be removed to access one floor panel. Another consideration is the choice of aligning carpet tile edges with those of the floor panels (matching one 2-ft. [0.6-m] carpet tile with each floor panel) or offsetting the edges. While the one-to-one match provides the maximum flexibility as discussed above, offset carpet tiles provide an improved seal for air leakage between floor panels from a pressurized plenum (see Section 4.3.2). In addition, some installers claim that offset carpet tiles reduce the chance of frayed carpet tile edges over time during floor panel removal and replacement. All of the above trade-offs must be considered when making a final carpet tile selection. Loftness et al. [2002] provide further discussion of these issues. It is recommended that you contact raised floor and carpet tile manufacturers directly to obtain the most up-to-date information. 87

100

101 Chapter 6 Controls, Operation, and Maintenance The control and optimization of temperatures in the occupied zone and the amount of thermal stratification (during cooling operation) is crucial to system design and sizing, energy-efficient operation, and comfort performance of UFAD and TAC systems. This section presents and discusses (1) recommended control strategies for effective system operation, (2) incorporation of individual control, particularly with TAC systems, to allow occupants to fine-tune their local environment, and (3) operations and maintenance (O & M) issues that differ from issues in conventional system operation. The discussion does not cover all possible control scenarios but is intended to introduce some of the key control strategies that have been frequently used in previously completed projects and explain how they differ from those typically used in overhead mixing-type systems. 6.1 CONTROL STRATEGIES IN PRESSURIZED PLENUMS This section presents control strategies that have been developed and applied in underfloor air distribution (UFAD) installations, the large majority of which have been pressurized plenum designs Supply Air Temperature (SAT) As described in Chapter 1, since air is supplied directly into the occupied zone near floor level, minimum supply outlet temperatures should be maintained in the range of F (16-18 C) to avoid overcooling nearby occupants. For TAC supply outlets located closer to the occupants and used to provide velocity cooling, even warmer minimum supply air temperatures may be advisable. 89

102 CHAPTER 6 CONTROLS, OPERATION, AND MAINTENANCE Constant Pressure A common control method for interior zones maintains constant static pressure in the underfloor plenum to ensure constant volume airflow from each diffuser (similar model and setting). Plenum pressure is maintained by adjusting fan capacity at the air handler. Occupants can make minor changes to local comfort conditions by manually adjusting a diffuser, but such adjustments are viewed as setup adjustments, not operating adjustments. As long as load variations in the zone due to diversity and other occupancy changes are small, and the net impact on plenum pressure by occupant diffuser adjustments is minimal, this strategy results in very nearly a constant-air-volume (CAV) operation and can maintain acceptably comfortable space conditions. In this configuration, one strategy for controlling supply air temperature that has been applied successfully in practice is to use measured return air temperature as a means of maintaining stratification at the desired level. However, even with proper design that promotes stratification at peak conditions, CAV operation can result in a changing environment in the occupied region as load changes. In CAV spaces, constant supply air temperature with decreasing load causes the space temperature profile to shift toward cooler temperatures and become less stratified. In this case, the average occupied zone temperature tends to be a few degrees cooler than the peak load thermostat temperature. Thus, supply air temperature (SAT) reset is recommended. However, the system response time during SAT reset can be significant due to the important impact of the temperature of the thermally massive concrete slab on supply air temperatures. CAV systems become progressively more over-aired as loads decrease from peak conditions, eventually virtually eliminating stratification. If the system is over-designed in the first place, stratification is likely never to be experienced in actual operation, which may explain why many projects in operation today report lack of stratification. Many projects use CAV systems for large interior zones where the perimeter zones are served by supply air passing through the plenum of the interior zone. If these interior systems were conservatively sized compared to actual loads and zone airflow is not properly adjusted during system balancing, then the zone will be over-aired. As discussed in Chapter 4, air leakage from pressurized plenums plus the additional heat loss through the floor surface can provide a substantial portion of the required cooling under part-load conditions. If part-load conditions or over-airing in the interior lead to a significant increase in the SAT, 90

103 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE this may compromise the system s ability to accommodate peak perimeter cooling loads, if they occur simultaneously. In CAV systems, interior zone airflows should be well matched to actual loads; active and robust control of SAT should be employed without starving the perimeter Variable Air Volume (VAV) Perimeter zones typically experience load changes that are greater in both magnitude and frequency than those encountered in interior zones. Automatic VAV control is the preferred strategy in perimeter and special zones (e.g., conference rooms) with these rapidly changing loads. See Chapter 9 for additional discussion of perimeter and special zone design solutions. Under VAV control, as load changes, room airflow and diffuser flow rate will change. Recent tests indicate that VAV operation with constant supply temperature results in a characteristic room air stratification temperature profile that is relatively consistent for moderate changes in load [Webster et al. 2002a]. As described above, a recent trend in perimeter zone designs in pressurized plenum installations is to have the supply air pass through the plenum of the interior zone to serve the perimeter. Some designers are now more frequently considering VAV control in the interior zone as an approach to minimize over-airing (overcooling) and to avoid starving the perimeter zone through SAT reset [Daly 2002] Controlling Stratification The objective of controlled stratification is to minimize energy use (reduce room airflow) while maintaining comfort (acceptable temperatures and stratification in the occupied zone). Overall room air stratification is primarily driven by the balance of room airflow rate in relation to the room cooling load. As discussed in Chapter 2, as room airflow is reduced for constant heat input, stratification will increase. On the other hand, if too much air is delivered to the space, stratification will be reduced, approaching a well-mixed room at the upper limit. Increasing or decreasing the supply air temperature (for constant load and room airflow) does not change the fundamental shape of the stratification profile but simply moves it to higher or lower temperatures. These principles are demonstrated in the example stratification control sequence shown in Figure 6.1. Figure 6.1 shows three different vertical temperature profiles (temperature (T) vs. height (H)). The profiles are not based on measured data but rather are shown for illustration purposes only. The sequence of moving from Profile #1 to Profile #3 is intended to demonstrate how 91

104 CHAPTER 6 CONTROLS, OPERATION, AND MAINTENANCE Figure 6.1 Example sequence for controlling thermal stratification. controlling stratification in UFAD systems requires different considerations than for traditional overhead systems. In this example, stratification is controlled by adjusting airflow and SAT to achieve comfort conditions in the occupied zone for various thermostat settings. The cooling load is assumed to be constant and at its peak value in all three cases. Profile #1 shows a modest amount of stratification (temperature gradient) characterized by supplying too much air to the room. As indicated, the temperature at the thermostat height (shown to be 60 in. [1.5 m]) is TSTAT1, and the average temperature of the occupied zone (from 4 to 67 in. [0.1 to 1.7 m]) is T oz1, avg. Profile #2 represents an alternative design load condition where airflow is reduced. To meet the same thermostat control point (TSTAT1=TSTAT2) with reduced airflow, the SAT also must be decreased. However, due to increased stratification, the average temperature of the occupied zone (T oz2, avg ), which along with the gradient is representative of overall comfort conditions, has been reduced. If the SAT is increased so that the occupied zone temperature is equivalent (T oz3, avg = T oz1, avg ) to that of Profile #1, Profile #3 is produced. This simple example shows how both supply 92

105 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE volume and SAT can be manipulated to create various occupied zone comfort conditions for given thermostat settings. The process of moving from Profile #1 to Profile #3 also demonstrates an approach that could be followed during commissioning to achieve the desired amount of stratification under peak load conditions. More research is needed to more precisely define the relationship between thermal comfort and stratification in UFAD systems. While overhead mixing systems routinely use the thermostat temperature, it is not yet known what the optimal control temperature is for stratified systems (e.g., average occupied zone temperature, weighted occupied zone temperature taking into account the increased sensitivity at head/ neck level, etc.). Readers are cautioned to use this example at their own risk. ASHRAE Standard [ASHRAE 1992] specifies that the amount of stratification in the occupied zone (temperature difference between head and ankle heights for a standing person) be limited to 5 F (3 C). One of the challenges of maintaining comfort in a stratified environment is that the thermostat temperature can no longer be assumed to represent the average temperature in the occupied zone. This is particularly true for thermostats located at a height of 5 ft (1.5 m) near the top of the occupied zone and therefore measuring close to the maximum occupied zone temperature. Because of this, it may be necessary to consider increasing the thermostat setpoint by up to 1-2 F [0.5-1 C] above the desired occupied zone setting under peak load conditions. The recently changed building code requiring thermostats to be installed at a lower 4 ft (1.2 m) height may help alleviate the need to adjust the setpoint as it will be measuring closer to an average temperature in the occupied zone. In any case, a balance between the amount of stratification and average, or representative, occupied zone temperature must be achieved within the limitations of room airflow and SAT available from the system Humidity Control One way to achieve the required higher supply air temperatures while still maintaining humidity control uses return-air face and bypass. Cooling coil temperatures are typically in the range of F (10-13 C) for dehumidification purposes. Only the incoming outside air and a portion of the return air is dehumidified (minimum amount needed for humidity control). The remaining return air is bypassed around the coil, if done at the air handler, and mixed with the cool primary air to produce supply air of the proper temperature and humidity before being delivered directly into the underfloor plenum. The face 93

106 CHAPTER 6 CONTROLS, OPERATION, AND MAINTENANCE and bypass dampers are controlled to achieve the desired supply air temperature as load changes. To save energy, the coil temperature can be varied to control humidity so a greater coil leaving temperature can be used when entering humidity conditions are low. If desired, this configuration also allows a range of coil temperatures to be utilized, including low-temperature air systems with or without ice storage. 6.2 CONTROL STRATEGIES IN ZERO-PRESSURE PLENUMS Although active diffusers may be installed in both zero-pressure and pressurized plenums, this section will focus on control issues associated with zero-pressure plenums with diffusers that are more likely to be adjusted by individual occupants. For additional discussion of task/ ambient conditioning (TAC) system control issues, see Bauman and Arens [1996]. Control strategies for the building's central mechanical system should be well coordinated with the local supply outlets. Since most TAC systems are used for cooling applications, if the general office space is overcooled by the ambient air distribution system control strategy, the cool air provided by the fan-driven local supply units will be unwanted by the occupants. By allowing the overall space temperature to rise (an energy-saving strategy), local cooling can then be used as needed to satisfy individual comfort preferences. Supply volume control of the central air handler requires a different approach with zero-pressure underfloor air distribution systems. Due to the extremely small or nonexistent pressure differentials between the supply plenum and the space, traditional pressure-sensing methods do not provide accurate measurements for control purposes. A CAV-VT approach has been successfully used, but the chance of over-airing or resetting SAT too high under part-load conditions must be carefully guarded against, as described above. An ingenious VAV control strategy proposed by Shute [1992] addresses the need to balance the airflow delivered by the central AHU into the plenum with the airflow leaving the plenum. The approach uses a temperature sensor in a vertical induction shaft directly connecting the return air at ceiling level to the underfloor plenum. Under normal operating conditions with this design, the active floor supply units will be delivering slightly more air to the space than is provided by the central system. So in fact, the plenum will be operated at a slight negative pressure. In this case, the temperature sensor will measure normal room return temperatures as the air is drawn down the induction shaft to mix with incoming primary air. If, however, the temperature in the induc- 94

107 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE tion shaft decreases rapidly, it indicates that the demand for air supply through the floor supply units has been reduced (i.e., fan units have been turned down or off), resulting in the overpressurization of the underfloor plenum. The central air handler can then be throttled down until the reversal in flow direction through the induction shaft is eliminated. A minimum setpoint at the air handler can be used to ensure that sufficient airflow is always supplied to the space. Another task/ambient control strategy described in Section (Figures 5.13 and 5.14) allows individuals to choose the amount of air from an underdesk diffuser for personal comfort without influencing the total amount of air being delivered to the space [Levy 2002]. A fan unit located in the underfloor plenum delivers air to two outlet locations, one under the desk for personal control and one farther away for control of the ambient space. The total air supply delivered to both supply outlets is thermostatically controlled to maintain overall comfort conditions in the room. The individual controls simply divert a portion of this total air quantity to the underdesk diffuser, as desired. A CAV- VT strategy for controlling the central AHU would be the simplest approach with this system. Of course, VAV control strategies could also be applied. TAC system configurations using fan-powered supply outlets provide a convenient means of allowing direct feedback from occupant control actions to improve overall system operation. Advances in direct digital control systems and monitoring capabilities allow this type of solution to be implemented. By monitoring fan speed settings, adjustments can be made to the setpoints for primary supply air temperature, ambient space temperature, and central supply air volume. For example, when a large enough percentage of occupants in the same zone of the building select low fan speeds, indicating that they are too cool, the primary air supply temperature to that zone could be raised. 6.3 INDIVIDUAL OUTLET CONTROLS Due to the importance of individual controls, they should be well designed and convenient to use. While most floor supply units currently on the market are based on manual control (requiring the user to bend down to floor level), it may be advisable to incorporate remote desktop controls to operate the floor units; one such remote-controlled floor unit is described by Matsunawa et al. [1995]. Local supply outlets that are located on a desk or nearby partition provide the best configuration for ease of use as they replicate the familiar environmental control system found on the dashboard of a car. It may also be advisable 95

108 CHAPTER 6 CONTROLS, OPERATION, AND MAINTENANCE to allow switching between local (individual) control and automatic (thermostatic) control as needed. While it is well recognized that building occupants prefer individual control, the extent to which they actually use the local controls, once made available to them, can be surprising. Several field surveys of buildings with operational TAC or UFAD systems have found that a relatively small percentage (10-20%) of the occupants make adjustments to their local controls on a regular basis [Hedge et al. 1992; Bauman et al. 1993, 1998; Webster et al. 2002c]. There are a number of possible reasons for this seemingly limited use of the occupant controls. (1) If the ambient space is well conditioned, there may be little need for individuals to fine-tune their local environment. (2) The design of the controls themselves may not be optimized, making their use difficult and inconvenient. (3) Only an occasional adjustment may be customary because individual preferences are relatively stable. (4) The sense of control may be more important in creating comfort than the actual environmental conditions. (5) The occupants may be unaware that they are allowed to control the air supply outlet in their vicinity (either due to lack of interest by the occupant or by intention of the operations personnel). It is critical to provide clear operating instructions for the TAC supply units to the occupants. Most occupants are unaccustomed to the idea of being able to control their local outlet. 6.4 OPERATION AND MAINTENANCE As with any new building system, building operators should be properly trained to allow the operation and control of the UFAD or TAC system to be optimized. Experience with early projects has demonstrated that UFAD and TAC systems that are operated using traditional control strategies based on mixing air distribution system performance are likely to have deficiencies in energy and operating costs as well as comfort performance. In recent years, as more projects have come online, operating guidelines such as those described above in Sections 6.1 and 6.2 are being developed that allow a wider range of system benefits to be realized. Among the O & M issues that may differ from those encountered with conventional system design are the following Cleaning Considerations in Underfloor Plenums A common concern among occupants of buildings with UFAD systems using floor grilles is that dirt and spillage will more easily enter 96

109 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE the underfloor plenum where it will mix with the air supply stream and be distributed throughout the occupied space. Experience has shown that if the plenum area is thoroughly cleaned at the end of construction, the amount of dust buildup on the slab is not excessive and can be handled through cleaning scheduled as part of normal plenum reconfiguration work. The required frequency of this type of cleaning is estimated to be two to three years depending on the observed rate of buildup. Dirt and particulates that do fall into floor diffusers will be collected by the catch basins beneath the diffusers. These basins should be cleaned as part of the regular maintenance schedule, depending on rate of build-up. Two important considerations are that (1) except near the plenum inlets, air speeds within the underfloor plenum are so low that they do not entrain any dirt or other contaminants from the plenum surfaces into the supply air, and (2) if a spill or other accidental contamination, such as fire, arises that does require cleaning, the accessibility of the underfloor plenum makes this process far simpler and more effective than in the case of overhead ductwork Reconfiguring Building Services As discussed in Sections and , the improved flexibility of raised floor systems provides significant cost savings associated with the reconfiguration of building services. One consideration during the relocation of workstations and furniture in open plan offices is that floor diffusers will often need to be moved (even though it is easy to do so) to accommodate the new positions of the furniture. This may actually add costs compared to overhead HVAC systems, which will typically not be reconfigured (at a potential cost of reduced system performance). Building operations staff will also need to maintain an adequate surplus stock of floor panels and carpet tiles to handle the required reconfigurations in response to the churn rate of the building occupants Acoustic Performance Due to the elimination or minimal use of ductwork in underfloor plenums, the noise generated from the operation of a UFAD system can be substantially less than that from a conventional ducted overhead system. This reduction in commonly found levels of background HVAC noise may create a situation where active sound masking or other acoustic design measures may be required, particularly in open plan offices where lack of sound privacy is a common complaint. 97

110

111 Chapter 7 Energy Use A commonly cited benefit of UFAD systems is that they save energy when compared to standard overhead (OH) air distribution systems. At present, this claim is difficult to prove quantitatively because of the lack of an energy-modeling tool that properly addresses all of the issues related to energy use with underfloor systems that would allow direct comparison of two simulated systems. To help designers understand the energy impacts of UFAD, this section explains the variety of factors that affect energy use in a descriptive manner. Topics include air distribution, economizer operation, cooling system efficiency, occupant thermal comfort, and pre-cooling strategies. 7.1 AIR DISTRIBUTION ENERGY As described in Chapter 4, the underfloor plenum is a primary air distribution route. Because the use of the plenum eliminates the need for some portion of the ductwork, and because the large size of the plenum creates little restriction to the flow of air, the amount of fan pressure required to deliver air in a building using UFAD can be less than that required in an equivalent OH system. For example, in an OH VAV-reheat system, a typical central fan design might provide 3 in. H 2 O (750 Pa) of pressure to move air through the index run. Typically ½ to 1 in. ( Pa) of this pressure might be required by the VAV box, reheat coil, and downstream lowpressure ductwork to the diffuser. Depending on the specific implementation of an equivalent UFAD system, most of this pressure loss might be eliminated due to the elimination of ductwork to the zones. This chapter was contributed by Allan Daly. 99

112 CHAPTER 7 ENERGY USE This reduction in pressure requirement can result in significant energy savings for a building because air distribution fans account for a large percentage of HVAC energy demand. However, some perimeter system designs (discussed in Chapter 9) can offset central-fan energy savings. A common approach to UFAD perimeter system design employs ducted underfloor fan-powered mixing or VAV boxes. These fans can erode energy savings due to the inefficiencies inherent in small fans and motors. In the example below (Table 7.1), two fan systems are compared both delivering 20,000 cfm to an example small building. In option 1, a single central fan is used. The fan is selected as an airfoil fan with a 30-inch wheel. At 3 in. H 2 O (750 Pa) pressure, this fan uses 13.4 bhp and runs at a combined fan and motor efficiency of 70%. In option 2, the same central fan is run at 2.5 in. H 2 O (623 Pa) of pressure, and four small perimeter fans are designed to provide 0.25 in. H 2 O (62 Pa) of pressure and deliver 3,000 cfm (1,420 L/s) each, simulating two-pipe underfloor VAV fan coils serving 60% of the total airflow of the central fan (the remaining air is assumed to serve interior zones). Typical of manufacturer offerings available now, these underfloor VAV fan coils are listed with electronically commutated motor (ECM) efficiencies. The small fans require 0.3 bhp each. When comparing central fan power, Option 2 exhibits a 13% savings over Option 1. When comparing total fan power, Option 2 shows a 3% savings. The reduced pressure requirements inherent in the underfloor system make up for the decreased efficiency of the smaller fans in this example. For the sake of keeping this comparison simple, the same air quantities are used in both Option 1 simulating an overhead supply system and Option 2 simulating a UFAD system. This design fan power comparison does not capture the annual energy performance of the two systems. Because the small fans handle only the air volumes required by the varying perimeter loads, the annual energy demand of this option will depend on the degree of load variation and the part-load operation of both the central and perimeter fans. A valid comparison of annual energy demand between the two cases could take the form of an hour-by-hour energy simulation of the two systems. Such a comparison is beyond this simple analysis, and while the fan system models exist in computer software programs such as DOE-2 to analyze this case, the physical performance characteristics of zones using UFAD do not yet exist and so more detailed analysis is difficult. 100

113 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE Table 7.1: Comparison of Fan Power for Overhead (Option 1) vs. UFAD (Option 2) Systems Option 1: Central Fan Only Total Airflow 20,000 cfm Fan Type Airfoil Fan Wheel Size 30 in. Design Static Pressure 3.0 in. Motor Size 15 hp Operating Power 13.4 bhp Fan Efficiency 76% Motor Efficiency 92.0% Combined Efficiency 70% Option 2: Central Fan + 4 Perimeter Fans Total Airflow 20,000 cfm Fan Type Airfoil Fan Wheel Size 30 in. Design Static Pressure 2.5 in. Motor Size 15 hp Operating Power 11.7 bhp Fan Efficiency 73% Motor Efficiency 92.0% Combined Efficiency 67% Fan Airflow 3,000 cfm Fan Type Forward Curved Fan Wheel Size 18 in. Design Static Pressure 0.25 in. Motor Size 3 hp Operating Power 0.3 bhp Fan Efficiency 50% Motor Efficiency 75% Combined Efficiency 38% Number of Units 4 Total Power 13.0 bhp Comparison 3% 101

114 CHAPTER 7 ENERGY USE A recent simplified analysis of central fan energy use explored the impact of various design assumptions on the supply fan energy consumption for pressurized UFAD systems vs. overhead (OH) systems [Webster et al. 2000]. The study assumed that the UFAD system configuration allowed the entire supply air volume to be handled by the central air handler (no active, fan-driven outlets) and assumed a 25% reduction in fan static pressure compared to the OH system. The average annual load factor for the combined core and perimeter zones was assumed to be 65% of design load for both systems. The results showed that the average annual fan energy savings using a VAV UFAD system compared to a VAV OH system (both delivering the same amount of air) was about 40%. 7.2 AIR-SIDE ECONOMIZERS Because the operating conditions inherent in UFAD systems are different from OH systems, the circumstances of when and how air-side economizers can be used change from one system type to the other. The two main factors that affect the use of economizers are the supply air temperature (SAT) and the return air temperature (RAT). In general, both the SAT and RAT are higher for UFAD systems than OH systems, though the RAT elevation depends on how much stratification is developed at the zone level. For the sake of comparison, typical OH and UFAD systems will be assumed to have operating temperatures described as follows: SAT Room Setpoint RAT System Type [ F] [ F] [ F] UFAD w/stratification UFAD w/o stratification Overhead (OH) Both increased SAT and increased RAT extend economizer operation. The increased SAT extends 100% free cooling and the increased RAT extends integrated economizer operation. Consider the following simple example of a single room with a cooling load of 22,000 Btu/h (6,450 W) run for each of the three system variations described in the table above. The room setpoint is 75 F (24 C) and exfiltrates 15% of the supply air volume at the room setpoint temperature (i.e., air quantity for building pressurization). In Table 7.2, each group of three lines represents UFAD with stratifica- 102

115 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE CFM supply [ft 3 /min] Table 7.2: Comparison of Sensible Cooling Coil Energy Use for Overhead vs. UFAD Systems Q room [Btu/h] OAT [ F] SAT [ F] Q x [Btu/h] RAT [ F] OA% MAT [-] [ F] Cooling Coil T [ F] Cooling Coil Sensible [Btu/h] 1,180 22, % ,167 22, % ,084 22, % ,180 22, % ,167 22, % ,084 22, % , 729 1,180 22, % , 298 2,167 22, % ,384 1,084 22, % , 113 1,180 22, % ,678 2,167 22, % ,458 1,084 22, % ,650 1,180 22, % ,273 2,167 22, % ,200 1,084 22, % ,021 1,180 22, % ,654 2,167 22, % ,061 1,084 22, % ,452 1,180 22, % ,146 2,167 22, % ,776 1,084 22, % ,809 1,180 22, % ,925 2,167 22, % ,207 1,084 22, % ,525 1,180 22, % ,093 2,167 22, % ,093 1,084 22, % ,

116 CHAPTER 7 ENERGY USE Figure 7.1 Example of sensible cooling energy as a function of outside air temperature. tion, UFAD without stratification, and an OH system, respectively. Figure 7.1 presents these same data graphically. Other assumptions are that the room air supply volume for the UFAD without stratification is twice the volume as for OH and is slightly more than OH for UFAD with stratification. The slight variation in room air-volume requirements results from the assumption that 15% of the supply air exfiltrates at the room setpoint temperature. In the UFAD case with stratification, this 15% effectively decreases the room-air delta T and thus requires slightly more airflow to deal with the load. For the amount of stratification shown, the actual room air-volume required is not known and has been assumed to be twice the OH as a simplification. This example also assumes that the climate has low humidity, so sensible energy is representative of relative energy performance. When an air-side economizer is used in a more humid climate, enthalpy control is advisable. CFM supply is the supply air volume delivered to the room. Q room is the room cooling load. OAT, SAT, RAT have been defined above. MAT is the mixed air temperature. Q x represents the heat exfiltrated from the room. OA% indicated the position of the outdoor air damper. Cooling Coil T indicates the required sensible cooling required, and Cooling 104

117 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE Coil Sensible indicates the net sensible cooling load provided by the HVAC system Extended 100% Free Cooling One hundred percent free cooling (meaning no cooling coil output is required to maintain SAT, not 100% outdoor air [OA%] damper position) only happens when outdoor air temperature (OAT) is less than or equal to SAT. In the OH case, the system is on 100% free cooling up to 55 F (13 C) OAT, corresponding to the OH SAT. In the UFAD cases, 100% free cooling is extended up the 65 F (18 C) OAT, corresponding to the UFAD SAT. Depending on the number of hours in the range of F (13-18 C) OAT in a given climate, this extended 100% free cooling can represent significant energy savings. In effect, the cooling compressors simply do not have to run during these extra 100% economizer hours Extended Integrated-Economizer Free Cooling When the OAT is between a system s SAT and RAT, then the system can take advantage of some free cooling, but it still needs to engage the cooling coil. This situation is called integrated economizer operation because the economizer and cooling coil work together in an integrated manner to maintain the SAT downstream of the cooling coil. As can be seen in Figure 7.1, above 55 F (13 C) and below 75 F (24 C) for the OH system, above 65 F (18 C) and below 75 F (24 C) for the UFAD non-stratified system, and above 65 F (18 C) and below 85 F (29 C) for the UFAD stratified system is where integrated economizer operation happens. In the UFAD cases without stratification, integrated economizer operation is not extended. In the UFAD case with stratification the integrated economizer operation is extended to 85 F (29 C) Climate Factors In the end, which effect dominates to extend economizer operation increased SAT or increased RAT depends on OAT distribution as well as the non-stratified UFAD RAT and building load dynamics. Another key factor that depends on climate affecting UFAD system energy benefits is the extent to which humidity control will drive the system operation. If humidity control concerns dictate that SAT must be low enough to dehumidify supply air, then none of the benefits described here can be captured because the SAT cannot be elevated unless OA humidity is removed in some other manner (i.e., desiccant dehumidification). 105

118 CHAPTER 7 ENERGY USE 7.3 COOLING-SYSTEM EFFICIENCY Cooling system operation at higher temperatures can reduce energy consumption. If cooling-coil leaving air temperatures can be elevated in UFAD systems, then the chilled water temperature serving the cooling coil can also be raised. In chilled water systems like this, or in DX systems delivering warmer off-coil air temperatures, the compressor or compressors serving the refrigerant loops will see lower lifts and consequently run more efficiently and use less energy. This effect is completely dependent on the refrigerant entering and leaving the evaporator coils being warmer, thus reducing the compressor lift; so again, if dehumidification is needed, then this effect cannot be captured. 7.4 OCCUPANT THERMAL COMFORT Recent research suggests that keeping occupied zones comfortable may be in conflict with minimizing energy use (see Section 2.3.4). As discussed in Chapter 2 on room air distribution, a temperature gradient forms in the occupied zone of a space that is developing a stratified room-air profile. This stratification is key to UFAD system dynamics in that it allows the occupied zone of a room to be comfortable, but the unoccupied zone at the top of a room reaches temperatures that would be uncomfortably warm. Energy-efficient system operation relies on a well-stratified room. Unless the room-air temperature difference between the SAT and RAT can be maintained at a high level, more air will be needed to remove the load in a room, and fan energy will correspondingly increase. For example, using the temperatures and systems described in Section 7.2 above, in the OH case a 20 F (11 C) temperature difference was developed between the SAT and the RAT 75 F (24 C) RAT minus 55 F (13 C) SAT. In the UFAD case without stratification, only a 10 F (6 C) room temperature difference was developed. In this case twice as much air is needed to remove the load. In the UFAD case with stratification, again the system was able to generate and maintain a 20 F (11 C) degree room-air temperature difference. However, as the stratified temperature difference across a room develops, so also develops a temperature difference from the bottom to the top of the occupied zone. This occupied zone temperature gradient can adversely affect occupant comfort. As described in Chapter 3, only a 5 F (3 C) variation from ankle to neck is allowed by ASHRAE Standard

119 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE If a strong temperature gradient is needed for energy-efficient system operation, but that gradient generates uncomfortable conditions, then a conflict exists. Surveys and measurements in existing projects suggest that this effect is happening to some extent, but without good mathematical models of room-air temperature dynamics, it s difficult to design systems that will find the best balance between comfort (limited stratification) and energy (increased stratification). The determination of the design cooling air quantity required to maintain comfort in UFAD systems requires a different approach from conventional design methods. In addition to stratification, it can also be affected by heat loss through the raised floor and, in perimeter zones, strong convective plumes that may develop along the building skin. See Section for further discussion. 7.5 PRE-COOLING STRATEGIES Because the slab is typically exposed to the supply-air pathways in UFAD systems, there exists an opportunity for harnessing the thermal mass of the slab in thermal storage or pre-cooling applications. Though this concept is elegant and offers the potential for reduced energy costs and possibly lower cooling peak demands, it can be difficult to implement and proper control requires knowledge of future weather conditions that are obviously difficult to predict. The benefits of this control strategy are furthermore hard to measure. To date, only a few earlier studies of completed projects present anecdotal evidence of the potential energy benefits of this approach [Spoormaker 1990; Shute 1995]. The concept is that, either using free-cooling or compressor cooling at off-peak energy rates, the UFAD system would circulate cool air throughout the building, effectively removing any stored heat from the building and slab and pre-cooling it for the next day. If the next day requires cooling from the time the system is enabled through the peak part of the day, less cooling would need to be generated because the cool storage in the building mass could be used to deal with some of the load. One downfall of this system is that if for some reason there is a period of warmup required in the morning following the nighttime precooling, then the building mass would work against the heating system and more heating energy would be required. This is why many heating designs isolate the warm air supply (e.g., by ducting from a fan coil unit) from the building thermal mass. 107

120 CHAPTER 7 ENERGY USE The potential benefits from this approach are real, but a proven implementation has yet to be studied. More research is needed to address thermal storage performance and support the development of design and implementation guidance. 108

121 Chapter 8 Design, Construction, and Commissioning As underfloor air distribution (UFAD) systems often represent a new approach for many contractors, it is important that members of the design and construction team recognize the differences from the conventional methods to which they are accustomed. In most projects, the implications of the raised floor system and the creation of an underfloor air supply plenum will represent the most significant change. This section presents a number of planning, coordination, and installation issues that should be considered from the beginning of the design phase, through construction, and into occupancy, including the commissioning process. 8.1 DESIGN PHASE The raised access floor platform, which represents a good example of integrated building design, serves multiple functions. It helps create the air supply plenum, conceals and protects cabling and other services, and provides a stable and level walking surface. The physical dimensions of the raised floor system should be considered early in the design process. These include: The existence of a 2 ft 2 ft (0.6 m 0.6 m) grid of pedestals (floor panel supports) across all areas of the underfloor plenum (grid dimensions may differ in raised floor installations outside the U.S.). The finished floor height of the raised floor panels above the concrete slab (typically in. [ m] from top of slab to top of floor panel). All members of the design team must understand the relationship between these dimensions of the underfloor plenum, the size of all 109

122 CHAPTER 8 DESIGN, CONSTRUCTION, AND COMMISSIONING building components that will be placed within the plenum, the placement and requirements for other building services not located within the underfloor plenum (e.g., elevators, access ramps, HVAC shafts), the operating characteristics of the UFAD system, and the requirements for their particular building-related concern. Engaging contractors with some degree of experience in installing and commissioning UFAD systems will be conducive to a smooth installation. It is important that the dimensions of all underfloor plenum components be carefully specified and documented on the approved construction drawings. The maximum allowable width of any fan terminal unit, freestanding ductwork, or other HVAC component when placed between standard pedestals is 22 in. (560 mm) (earthquake design is less). Although it is possible to use a specially fabricated raised floor support structure to span across larger underfloor components or ductwork, this is rarely done in practice and is expensive. All items placed in the underfloor plenum must fit in the clear space beneath the raised floor panels. In new construction, the specified height of the underfloor plenum is often determined by the largest HVAC component that must be contained within the plenum. Keep in mind that floor sag and unevenness of slabs will require some tolerance from precise theoretical dimensions. In addition to the overall size of components, positioning of these within the plenum is important to provide access from above in relation to furniture layouts, as well as to avoid obstructing the route of various other services and equipment within the plenum. The successful employment of UFAD systems requires coordination between all building trades throughout the design and construction process. Successful projects have often allocated some amount of budget to cover the additional effort required for effective coordination. The amount required can be less than $0.10/ft 2 ($1.10/m 2 ) but has proved to be very useful [Vranicar 2002]. Local building and fire code issues should be considered early in the design process. For further discussion, see Chapter CONSTRUCTION Although the number of projects using underfloor air distribution has increased noticeably in the past five years, experience with the installation of this technology is still rather limited within the U.S. building industry. As guidelines have not been available, designers and installers working on these projects have largely developed their own methods and approaches. It is generally accepted that an underfloor air supply plenum can provide benefits during the construction process. 110

123 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE For example, working at floor level to install building services within the underfloor plenum can speed up construction in comparison to the conventional approach of installing most services in the ceiling plenum. On the other hand, there may be some penalty in terms of increased cost or time involved for first-time contractors. As experience is gained, however, it can be expected that new standardized design tools and construction methods will be developed that will provide important cost savings. The following discussion is based on information obtained from designers and installers who have previously worked on UFAD projects. Prior to installation of the raised floor system, the slab must be cleaned and sealed to reduce dust and, if desired, to inhibit bacterial growth. When a well-planned construction sequence is employed, the finished raised floor surface is not installed until after most of the dirtgenerating construction work has been completed. Careful coordination of these activities can help reduce the number of times the slab will need to be cleaned before installation of the raised floor. Any dirt/dust or materials that enter the underfloor plenum prior to occupancy must be removed (e.g., by vacuum cleaner or wet cloth) and the floor cleaned one final time before the internal fit out is completed. The main structural slab, the traditional working platform, will not be available continuously during construction, and therefore a wellcoordinated construction sequence is necessary (see Shute [1995] and McCarry [1995] for earlier discussions of this process). Recent UFAD installations have reevaluated the typical schedule for work involving the fabrication of the underfloor plenum. The following sequence is recommended, as it limits the disruption of having to work on the slab with pedestals in place, prior to the placement of the floor panels. Contractors with experience may modify this sequence or develop their own preferred methods. 1. Thoroughly clean slab surface. 2. Apply any coating or sealant to the slab. 3. Mark the grid of raised floor pedestal locations on the slab surface, but do not install them. This requires careful preplanning and layout of the raised floor grid in relation to all specified underfloor services. By not installing the floor pedestals until after all major building services in the underfloor plenum have been installed on the clean slab surface, contractors can work faster and safer. 4. Install perimeter and other required fan terminal units, other HVAC components, and all required underfloor air distribution ductwork, except air highways and underfloor partitions. 111

124 CHAPTER 8 DESIGN, CONSTRUCTION, AND COMMISSIONING 5. Install underfloor wiring for power/voice/data. Cable runs should be terminated with a coil of extra lengths sufficient to reach all possible locations (floor boxes or partition connections) served by that run. 6. Install all required piping (e.g., hot water supply and return serving perimeter heating coils). Access may be provided along perimeter columns. 7. Verify that all vertical surfaces that are to be located adjacent to the access floor cavity have been adequately sealed according to the floor plenum leakage specification (to be provided to all involved contractors). This includes all junctions of these surfaces with the building slab, penetrations of drywall and other vertical partitions, and any other boundaries with the building slab. 8. Install pedestals and solid raised floor panels. 9. Install air highways and any underfloor partitioning at desired locations. Pay close attention to the sealing of air highways since they are operated at higher pressures than the plenum. Floor panels forming the top surface of air highways should be sealed (taped) around all edges and marked as being permanent (not to be removed even temporarily). See further discussion of plenum sealing below. 10. Determine floor diffuser and power/voice/data terminal locations. In open plan offices, this requires careful preplanning of the locations for partitions and workstation furniture. It also requires consideration of the locations of all major HVAC elements in the underfloor plenum, including fan terminal units, large ductwork, and air highways (if specified). Access to these underfloor components will need to be maintained. All locations are tied to the floor grid originally laid out in step 3 above. Diffusers and cable outlets can then be assigned as desired (e.g., one per workstation, etc.). Large HVAC components should not be located in areas where diffusers will be placed, since nearly all diffusers include baskets and catch basins that hang below the bottom surface of the floor panels into the underfloor plenum. 11. It is preferable to keep solid floor panels in place until diffusers are installed to maintain the raised floor as a safe working platform and to help preserve the cleanliness of the underfloor plenum. Diffusers may be more efficiently installed in precut panels at staging areas. If necessary, install precut floor panels (at locations determined in step 10) by exchanging with an existing solid panel. Install temporary cover plates over the predrilled access holes. 112

125 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE 12. When all dirt/dust-generating construction activity is completed, thoroughly clean the top surface of the raised floor plenum and, if needed, clean the underfloor slab surface at any locations that have accumulated dirt. 13. Install floor diffusers/panel assemblies and power/voice/data terminals. 14. Install carpet tiles according to manufacturer s specifications, cutting access holes for all diffusers, grilles, and power/voice/data terminals. During the construction stage, as-built drawings should be made available, indicating the exact location of services within the underfloor plenum for future access, maintenance and system upgrades. Identification and coordination of trade responsibilities are also considerations during the installation of a UFAD system. While mechanical contractors will typically be responsible for all air distribution ductwork in conventional systems, UFAD designs require that dry wall and/or raised floor contractors be responsible for significant portions of the air distribution system: the underfloor plenum and often the air highways. This is particularly critical in pressurized UFAD systems, where greater care must be taken during construction to seal the underfloor plenum to prevent uncontrolled air leakage [Daly 2002]. As discussed previously, the use of a zero-pressure plenum design can significantly reduce uncontrolled leakage between the plenum and the conditioned space, adjacent zones, and the outside. However, it is advisable in all projects to address the leakage issue, as discussed further below. The installation and sealing (to allowable leakage limits) of sheet metal ducts use well-established methods and are governed by existing building codes and standards. The situation is different for underfloor air supply plenums and air highways. Due to the newness of this technology, applicable codes and standard construction methods have not yet been established. Contractors outside of Division 15 are not accustomed to paying close attention to the sealing of the air distribution path. For example, the sealing of edge details all around the underfloor plenum should address window-wall connections to the slab, stair landings, and HVAC shaft walls. At these locations, other members of the construction team, including the general contractor, may become involved. It is important that the responsible contractors recognize and perform the critical role that proper sealing plays in the effective operation of a pressurized UFAD system. In addition to initial installation, the integrity of a well-sealed underfloor plenum or air highway must be preserved over the course of 113

126 CHAPTER 8 DESIGN, CONSTRUCTION, AND COMMISSIONING all subsequent work within the plenum, even after building occupancy. Sheet metal underfloor partitions used to define separate control zones or air highways can be easily and repeatedly penetrated during installation of other services, such as cabling and plumbing. Seismic bracing, sometimes required for plenums of greater depth (generally higher than 18 in. [0.45 m]), can lead to unsealed openings, and penetrations through exterior walls and along interior structural elements are also commonplace. Specifications should be put in place for the lifetime of the building requiring all such penetrations to be carefully repaired and sealed. Another approach that has been used to reduce uncontrolled penetrations is to pre-install access channels or sealable ports across air highways or through partitions at periodic intervals. Floor contractors generally provide access holes precut in the floor panels for diffusers, power/voice/data terminals, and other outlet boxes. Mechanical contractors should be responsible for all required ductwork (except perhaps air highways) and the installation of all floor diffusers, grilles, fan terminal units, and other mechanical equipment in the underfloor plenum. Trade responsibilities may also be shifted somewhat with regard to the installation of furniture- or partition-based task/ambient conditioning (TAC) systems. Depending on the TAC supply unit design, these systems may be installed by the mechanical contractor or, if well integrated into the furniture and partitions, may become the responsibility of the furniture installers. In the large majority of raised floor systems for office applications, carpet tiles are installed on top of the floor panels. In addition to providing the finished floor surface, the carpet tiles serve a second important purpose by providing a seal over the top of the raised floor installation. Due to the large surface area, leakage through the gaps between floor panels can be significant in pressurized plenum systems. Carpet tiles can reduce the amount of leakage by a factor of 2 or 3 in comparison to bare floor tiles (see Chapter 4 for further discussion). If bare floor panels without carpeting are used, some provision for sealing between floor panels must be provided or large leakage rates can be expected and must be accounted for in the operation of the UFAD system. The installation of carpet tiles raises a number of issues to be aware of, especially as different manufacturers typically supply the carpet and floor panels. Of particular note is the commonly used technique of applying an adhesive to install carpet tiles on the floor panels. Care must be taken to avoid using an excessive amount of adhesive as it may make it difficult to remove carpet tiles to gain access to the floor panels 114

127 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE during subsequent relocation, replacement, or service work. Adhesive accidentally seeping into the underfloor plenum may also damage cable-management components and negatively affect the quality of the supply air. For additional discussion of carpet tile selection and installation, see Section RETROFIT PROJECTS In projects requiring the installation of a new HVAC system within an existing building (e.g., retrofitting), UFAD offers many advantages. Such projects often suffer from limited space for accommodating ducts and other components. By eliminating large overhead ceiling ducts, the total plenum height required in UFAD installations is less than that for ceiling-based systems. Therefore, UFAD is feasible for existing buildings where, due to restricted floor-to-floor/floor-to-ceiling heights, it is necessary to minimize the vertical space occupied by ductwork. In addition, the installation of a raised floor system is less disruptive than that of ducting for overhead systems as the floor can be easily installed, and removed, as an independent platform, leaving relatively few structural scars. This issue is important in buildings where maintaining the integrity of the existing building structure is important for heritage/cultural/structural reasons. Furthermore, installation can be a relatively dry process, once the concrete structural slab has been adequately sealed, minimizing damage to other building elements. 8.4 SPACE PLANNING In partitioned office spaces, consider the relationship between the partition grid and floor grid. It is recommended to offset the partition grid from the floor grid so that partitions do not cover joints between floor panels, thereby preventing access to the underfloor plenum on both sides of the partition. In addition, it is important that underfloor equipment requiring regular maintenance be located in accessible areas, such as corridors, and not underneath furniture and partitions. To be most effective, access to larger underfloor equipment (e.g., perimeter fan coil boxes) should include more than just the floor panel(s) directly above the equipment. Removing a unit for service will be greatly facilitated by providing access to the floor panels surrounding the unit. Designers must consider that, depending on the particular zoning arrangement of a project, fan rooms or access for HVAC distribution may be required at more frequent intervals than with conventional air distribution systems. In addition, for some designs return-air shafts 115

128 CHAPTER 8 DESIGN, CONSTRUCTION, AND COMMISSIONING may be required to be placed directly between the ceiling and the underfloor plenum. These can typically be accommodated around columns or other permanent building elements. At the service core of a building, it has been common practice to omit the raised floor in areas such as restrooms, equipment rooms, stairwells, and sometimes kitchenettes. Generally, a raised concrete core is poured in these areas to accommodate the difference in floor height between the service core and the finished raised floor in the surrounding office areas. At junctions between raised flooring and areas without it, allowances must be made for suitable transitions. More recently, raised flooring has been used in the core areas as well. 8.5 COMMISSIONING A carefully conducted commissioning of a UFAD installation will go a long ways toward ensuring that all building systems are properly applied, installed, and operated, despite the novelty of this technology to some members of the design and construction teams. Commissioning is a systematic process that begins in the design phase and extends through occupancy and the warranty period for the building and uses documentation and verification methods to make sure that the facility meets the design intent and the expectations of the owner and occupants [ASHRAE 1996; Dasher et al. 2002; PECI 2002]. Because UFAD technology is classified as being energy-efficient and green, well-designed systems will tend to be right-sized, not the more common oversized [York 1998]. With less of a safety margin, correct system operation, as verified by commissioning, takes on added importance. Recent research has shown that promoting and maintaining room air stratification is critical to successful design and operation (under cooling conditions) of UFAD systems [Webster et al. 2002a, 2002b]. Overall room air stratification is primarily driven by room airflow rate relative to load. As room airflow is reduced for constant heat input, stratification will increase. On the other hand, if room airflow is increased relative to load, stratification will be reduced, approaching the well-mixed constant temperature profile characteristic of overhead air distribution systems. The objective is to determine the operating point that minimizes energy use (reduced room airflow) while maintaining comfort (acceptable temperatures and stratification in the occupied zone). Because of the important balance between room airflow and heat input to the space, proper and complete commissioning of a UFAD system will require operation and adjustment of the system under peak (or 116

129 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE close to peak), as well as partial, cooling load conditions. During preliminary HVAC system commissioning prior to building occupancy, supply air quantities and temperatures can be established according to design estimates. Since standardized design tools based on fundamental research are not yet available, designers will need to proceed cautiously using an empirical approach where they rely on their previous experience with UFAD systems, the experiences of others, or other available information to guide their design decisions. Until the space is occupied and subject to typical cooling loads, however, it will be difficult to verify the proper system operation. Commissioning performed after occupancy (with more typical cooling loads present) will serve as the best approach to achieve the desired system operation. For more discussion, please see Chapter 2 for room air stratification and Chapter 6 for controls and operation. 117

130

131 Chapter 9 Perimeter and Special Systems This chapter discusses a range of system design solutions for perimeter and other special zones. 9.1 PERIMETER SYSTEM DEFINITION Perimeter systems serve a number of functions in commercial buildings. These include the following. Local Heating. Almost all commercial buildings require heat at the perimeter due to the influence of the building envelope. At the same time, this heating demand is often intermittent and during swing seasons can happen early in the same day that cooling is required later. Local Cooling. Perimeter systems are designed to allow wide variation in the amount of cooling that can be provided to a zone to deal with dynamic solar and other envelope loads. The more efficient the envelope that a building has, the more options that are available for perimeter cooling systems. Interior / Perimeter Separation. The perimeter space is defined as the space in a system that is affected by weather and outside conditions. Perimeter systems allow the perimeter to be separated from the interior system to prevent fighting in winter. This is a requirement of virtually all energy codes. Automatic Control. Because of the dynamic nature of perimeter building loads, an important function of perimeter systems is to provide automatic control that can adjust to varying interior conditions. Envelope and other perimeter loads change too quickly and constantly to make manual control an effective option. This chapter was contributed by Allan Daly. 119

132 CHAPTER 9 PERIMETER AND SPECIAL SYSTEMS 9.2 PERIMETER SYSTEM OPTIONS Two- or Four-Pipe Constant-Speed Fan Coils The two- or four-pipe constant-speed fan coil system consists of a fan-coil box located in each perimeter zone. In the two-pipe arrangement, only heating water is provided to the box. In the four-pipe arrangement, both heating and cooling pipes are provided. The diagrams in Figure 9.1 illustrate a two-pipe overhead fan-coil at left and a two-pipe underfloor fan-coil at right. In Figure 9.1 and all subsequent figures in this section, T refers to the room thermostat being used to control the indicated equipment. In both systems illustrated in Figure 9.1, ventilation and cooling air is provided by the central air system. In the case of the overhead twopipe fan coil, the damper under the floor is connected to a segregated portion of the plenum located below a perimeter zone. Cooling and ventilation air is controlled to the zone through an underfloor modulating, pressure-dependent damper. In heating mode, the damper goes to minimum position and the fan in the fan-coil is engaged, recirculating return air to the space. If more heating is required, then the heating coil is engaged. In the underfloor two-pipe diagram, the fan-box intake is fitted with a damper that allows air to be taken from the room in heating mode or from the plenum in cooling mode. As cooling demand varies, either room air or hot-water reheat is used to temper the cooling supply air delivered to the space. In heating mode, the intake damper goes to a minimum position to allow minimum ventilation while the remaining air comes from recirculated room air. Heat is added via the reheat coil as needed. The constant-volume operation of this fan-coil option makes its operation relatively energy inefficient. This system option is also expensive compared to others. Care must be taken in design to address the noise created by these fans. Another variation on the underfloor constant-speed fan-coil ducts 55 F (13 C) air directly to the cooling inlet of the box. The fan-coil then mixes 55 F (13 C) air and room air as needed to maintain comfortable space conditions. The advantages of this option include an ability to deal with high loads and minimized shaft area requirements. A disadvantage is that significant amounts of equipment and ductwork are required under the floor. 120

133 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE (a) (b) Figure 9.1 (a) Two-pipe overhead fan coil; (b) two-pipe underfloor fan coil. 121

134 CHAPTER 9 PERIMETER AND SPECIAL SYSTEMS Hydronic Heat Pumps Similar to the fan-coils described above, a hydronic heat pump can be located either in the ceiling above a zone or in the underfloor plenum below a zone. The heat pump would draw ventilation air from the central system and either harvest or reject heat to a two-pipe heat-pump loop. This is an energy-efficient, though expensive, option. Other issues include maintenance access to the units and the noise generated by the compressors in the heat-pump units VAV or Fan-Powered VAV with Reheat This perimeter system option consists of essentially an overhead VAV system placed under the floor in the raised-floor plenum. In general, this approach does not take advantage of the plenum as a low-pressure air distribution pathway, and the large amount of equipment and ducts placed in the plenum severely limit the flexible use of the plenum space. This system option is usually employed with conventional 55 F (13 C) supply air temperature and can be necessary if envelope loads are high, particularly solar loads. In the case of high envelope loads, the F (17-18 C) supply air temperature typical of UFAD systems may be too warm to effectively remove the loads. This underfloor conventional VAV system using 55 F (13 C) air or colder can deal with high loads. The system efficiency and cost of this option is comparable to standard OH systems, though taken together with the cost of the raised floor it can be an expensive choice Cooling from VAV Diffusers, Heating from Heating-Only Fan Coil This approach changes mode between cooling and heating operation (Figure 9.2). In cooling, thermostatically controlled VAV diffusers modulate to maintain a room-temperature setpoint. In heating, the same diffusers are used in conjunction with a heating-only fan coil. Some diffusers become return inlets for the fan coil by changing the position of their dampers. Other diffusers become heating outlets by changing the position of their dampers. Minimum ventilation is accomplished with a mechanical stop on the heating inlet diffuser. See Section 5.2 for further discussion. Fan coils operate only in heating mode. This system also reduces reheat by heating air from the space rather than heating cool air from the plenum. 122

135 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE Figure 9.2 VAV diffusers with heating-only fan coil Fan-Powered Outlets In a fan-powered outlet perimeter system, a local fan is built into a 2 ft 2 ft ( m) module that can replace one of the standard raised-floor tiles. The fan is controllable by the occupant in the space above, providing a high degree of control. This system is uncommon in new projects, and one manufacturer recently discontinued selling this product in the United States. It can be expensive because there are few manufacturers of this type of outlet. One advantage that this system has is that it is used in conjunction with zero-pressure plenums, which have very low central fan energy use. As described in Chapter 7 on energy use, the perimeter fans being smaller are inherently less efficient, but if they can be implemented to track the loads closely, perhaps their energy use could compare to central system approaches because they completely eliminate any overpressurization of ductwork or plenums Convector or Baseboard Heating Coupled with Central UFAD System Cooling In this option, cooling and ventilation air are provided by the central system. As illustrated in Figure 9.3, perimeter subdivisions of the plenum are created below each zone. A pressure-dependent modulating 123

136 CHAPTER 9 PERIMETER AND SPECIAL SYSTEMS Figure 9.3 Central system cooling with perimeter hot water convector. Figure 9.4 Variable-speed fan coil with reheat. 124

137 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE damper varies the amount of air introduced to the zone based on a space temperature sensor. In heating, the damper goes to minimum position and a convector in a trench that is not open to the plenum, or a baseboard heater located above the raised floor, engages to add heat to the zone Variable-Speed Fan Coils This approach makes underfloor variable-speed fan coils the primary piece of equipment in perimeter zones (Figure 9.4). Where heating is needed, the units are fitted with two-pipe hot water coils or with electric resistance heat. The fan boxes are placed un-ducted at the inlet below each perimeter zone. The variable-speed fan increases or decreases airflow depending on space demands. In heating the fan goes to a minimum speed and the hot water or electric heat is engaged as needed. Typically these fans are controlled with electronically commutated motors, which offer good efficiencies for small fans. This system has the cost advantage of a minimum amount of ductwork and grilles/diffusers. The same diffuser or grille is used in both heating and cooling. Ductwork on the discharge side of the fan-coil needs to be insulated. A disadvantage of this system is the electrical demand and energy consumption of the fan coil, as described in Chapter 7. A further energy disadvantage of this system is that it employs reheat, also as described in Chapter 7. This approach is less flexible than some other perimeter system options. It will be accordingly more costly to reconfigure. Because this system uses fans under the floor located close to building occupants, the noise generated by the fans must be considered in the design and location of these units. Another design consideration is that even with the fan off there is bypass around the fan wheel due to the pressurized floor. This can allow the fan to cycle off during the deadband between heating and cooling since minimum ventilation air can be supplied through the inactive fan. But it can also result in excess cooling if plenum pressures are high (e.g., greater than 0.05 in. H 2 O [12.5 Pa]). Figure 9.5 shows how the fan and heating element (a hot-water coil in this case) are sequenced to provide zone temperature control. Figure 9.6 shows how this type of perimeter system can be implemented as part of an entire building system. 125

138 CHAPTER 9 PERIMETER AND SPECIAL SYSTEMS Figure 9.5 Control sequence for variable-speed fan coil with reheat. Figure 9.6 UFAD system schematic with variable-speed fan coil with reheat in perimeter. 126

139 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE Figure 9.7 Perimeter fan coil units ducted to linear bar grilles. Figure 9.7 shows a plan view of typical perimeter fan coils ducted to supply grilles located at the perimeter of a zone. One variation on this approach uses a small partitioned perimeter plenum instead of ducted supply grilles. This variation is illustrated both schematically (Figure 9.8) and in plan (Figure 9.9). One caution to bear in mind with this approach is that the perimeter plenum becomes pressurized because of the fans and, because it is located directly adjacent to the exterior wall of the building, can potentially become a major source of air leaks to the outdoors. Care must be exercised in thoroughly detailing and sealing the slab and exterior wall connections. Also, the plenum dividers reduce system flexibility VAV Change-Over Air Handlers Another perimeter system option is the use of VAV change-over air handlers, sometimes referred to as variable volume and temperature (VVT) systems. The concept behind this approach is to provide only a single temperature air to an entire building façade through boxes that have a single duct and no reheat coil inside. If the building shape and facades are large enough to justify the cost of the required air handler, then the system overall can be relatively inexpensive. Figure 9.10 shows a system schematically. Figure 9.11 shows how a VAV change-over system would be implemented in an entire building. 127

140 CHAPTER 9 PERIMETER AND SPECIAL SYSTEMS Figure 9.8 Fan coil unit serving partitioned perimeter plenum. Figure 9.9 Plan view of fan coil unit serving partitioned perimeter plenum. 128

141 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE Figure 9.10 Perimeter VVT system. Figure 9.11 UFAD system schematic with VAV change-over system in perimeter. 129

142 CHAPTER 9 PERIMETER AND SPECIAL SYSTEMS Considering the VAV change-over system in detail makes an interesting comparison with the underfloor VAV fan-coil approach. The VAV change-over system has the advantage that it can cost less in some cases due to many inexpensive zones that offset the added cost of required air-handlers. In this system there is also no water piping underfloor, which eliminates the problems associated with leaks. The VAV change-over system can supply lower temperature air to reduce supply air quantity for high glass loads, but there is the danger of drafts. The VAV change-over system is efficient because there are zero reheat losses. It can have lower fan energy in comparison to configurations using small fan-coils with their reduced fan and motor efficiencies. And finally, because the equipment is centralized, there are lower maintenance costs. When compared to the VAV underfloor fan-coil approach, the VAV change-over system has the following disadvantages. It requires larger mechanical rooms for the central equipment. It requires additional small shafts at the building perimeter. It can create conflicts with operable windows. It cannot heat and cool simultaneously. Finally, it requires complex control logic that needs careful commissioning. 9.3 CONFERENCE ROOMS OR OTHER SPECIAL SYSTEMS When conference rooms are located in the interior of a building, they typically are designed with their own zone because of the rapidly and significantly varying loads in these types of spaces. Virtually any of the approaches described above can be adapted for use in conference room or other special space zones. One common approach is the use of a VAV underfloor fan terminal without a reheat coil as illustrated in Figure A single pressurized plenum subdivision is created below the zone and that area is served by the VAV fan terminal. Another common approach is the use of modular active (fandriven) diffuser units as illustrated in Figure A variable-speed fan box is mounted below a single floor panel. Fan speed is thermostatically controlled and integration with an occupancy sensor can allow the fan to remain off during unoccupied periods. Another consideration for conference rooms is to use a fan terminal without heat to supply the conference room with air from adjacent overventilated spaces to reduce the impact of the high ventilation percent on the overall ventilation requirements in accordance with ASHRAE Standard 62. This approach is analogous to using fan-powered VAV boxes (or powered induction units) or exhaust/transfer fans commonly used with conventional overhead systems. 130

143 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE Figure 9.12 Variable-speed fan terminal serving conference room. Figure 9.13 Active (fan-driven) diffusers serving conference room. 131

144 CHAPTER 9 PERIMETER AND SPECIAL SYSTEMS CO 2 sensors can be used effectively to save energy and reduce subcooling in special zones like these as well. 9.4 ISSUES TO CONSIDER IN THE DESIGN OF PERIMETER AND SPECIAL SYSTEMS Some issues to consider when designing perimeter and special systems include the following. The use of plenum partitions for thermal zoning reduces the flexibility of the underfloor space. Using underfloor fans to condition perimeter spaces typically erodes some of the energy benefits of UFAD systems as described in Chapter 7. The maintenance and noise impacts of underfloor fans must also be addressed in a well-designed UFAD system. As described in Chapter 7, some UFAD systems employ reheat in their designs. In most designs, reheat energy losses are small relative to overhead systems because of the warm supply air temperature and the ability to use very low minimum volume setpoints due to the warm air supply from the floor. Still, strategies to reduce or eliminate reheat, such as fan-powered boxes supplying room air rather than plenum air, can be applied effectively to UFAD systems. Take advantage of the natural thermal plume at the skin due to solar radiation, conduction, and infiltration during cooling to reduce supply air requirements. Use blinds and light shelves wherever possible to capture the solar load at the skin. Using the same diffusers for heating and cooling is an effective strategy to reduce cost and floor penetrations. When designing perimeter and special systems, consider the ease and cost of system and equipment reconfiguration. Due to the chimney effect of skin loads and solar with blinds or light shelves plus cooling transmission of the floor, the underfloor system can require the same or even less air to cool the perimeter space with F (17-18 C) air than an overhead system with 55 F (13 C) air. 132

145 Chapter 10 Cost Considerations Objective first and life-cycle costs are crucial to providing a sound basis upon which UFAD systems can be compared to alternatives. As the number of installed UFAD projects has grown in recent years, more examples are available to reveal the cost-effectiveness of these systems. Although costs vary from project to project, lately it has been demonstrated that first costs for these system can be very comparable to conventional overhead design [e.g., Loftness et al. 1999, 2002]. A cost comparison tool containing backup data and information on a wide range of UFAD cost components has recently been developed [Tate 2002b]. Engineers, architects, and contractors are becoming more familiar with UFAD technology as more information becomes available. It is now well recognized by owners and developers that raised floor systems with UFAD significantly reduce costs associated with frequent office reconfigurations. More manufacturers are entering the UFAD market with new products to respond to the increased demand. As the above trends continue, costs can be expected to further decrease. Table 10.1 summarizes many of the cost components that should be considered when evaluating the economic impact associated with the use of a raised floor with (or without) a UFAD system. In the table, these components are segregated according to their expected (positive or negative) contribution to the overall construction costs of the building. For further discussion in this section, the components are divided into the following three groups: (1) standard first cost components, (2) design-dependent first cost components, and (3) life-cycle cost components. Actual cost data are not presented below, as these numbers can fluctuate depending on market conditions. It is recommended that you contact manufacturers, engineers, and installers with experience to obtain the most up-to-date cost information. 133

146 CHAPTER 10 COST CONSIDERATIONS Table 10.1a: Cost Considerations for the Addition of Raised Floor and UFAD Systems: First Costs Addition of Raised Floor System Further Addition of UFAD System Typical Cost Adds Typical Cost Reductions Typical Cost Adds Typical Cost Reductions Basic Structure Costs: Increased column size to support floor Mechanical cores must either be raised or (handicapped) ramping installed Slab-to-slab height increase if space floor-to-ceiling height is to be maintained Cost of the raised floor and premium for carpet tiles (vs. rolled carpet) Basic Structure Costs: No final slab leveling as floor is laser leveled Power/Voice/Data Service Costs: Power/Voice/Data Service Costs: Power wiring uses homerun power modules throughout the space to reduce cabling requirements Floor outlet boxes in each workstation eliminate the need to electrify furniture Modular plugs in outlet boxes reduce the required connection time for PVD services Installation costs are reduced due to the ease of working at floor level Conduit costs may be significantly reduced or eliminated if plenum rated cable is used Basic Structure Costs: Slab must be cleaned (and treated with an antimicrobial agent) prior to floor installation Power/Voice/Data Service Costs: Basic Structure Costs: Slab-to-slab height may be reduced as HVAC equipment and ductwork are removed from the ceiling plenum Removal of HVAC equipment from overhead plenum may eliminate need for false ceiling Power/Voice/Data Service Costs: 134

147 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE Table 10.1a: Cost Considerations for the Addition of Raised Floor and UFAD Systems: First Costs (Continued) HVAC System Costs: HVAC System Costs: HVAC System Costs: Thorough sealing of components/surfaces that compose the underfloor supply plenum Addition of ducts or air highways to ensure proper delivery of the conditioned air through the underfloor plenum Higher diffuser cost due to increased quantity and relatively higher cost ($/cfm) of the outlets Additional smoke detectors for underfloor plenum Special air handlers with bypass HVAC System Costs: Reduction (or elimination) of horizontal (branch) ductwork feeding terminal units Reduction of (rectangular and flexible) discharge ductwork and dampers Reduction of required thermal insulation as supply air passes through an already conditioned plenum Reduced outlet balancing requirements as most diffusers allow occupant adjustment Elimination of radiation dampers on supply outlets Reduction in the number of required terminal units (especially in interior zones) Reduced number of space thermostats and associated wiring as the number of terminal units are reduced Potential reduction in return outlets if false ceiling is eliminated Reduced installation costs as work is done at floor level Possible reduction in air-handling unit size and capacity (where design airflow quantity can be reduced) 135

148 CHAPTER 10 COST CONSIDERATIONS Table 10.1b: Cost Considerations for the Addition of Raised Floor and UFAD Systems: Life-Cycle Costs Addition of Raised Floor System Further Addition of UFAD System Typical Cost Adds Typical Cost Reductions Typical Cost Adds Typical Cost Reductions Utility Costs: Utility Costs: Utility Costs: Utility Costs: Reduced fan operational cost due to lower fan static pressures Possible refrigeration plant operational cost savings due to increased chiller efficiency using warmer return water Extended economizer cycle operation due to higher supply/return air temperatures Maintenance/Operation Costs: Cash Flow Related Intangibles: Maintenance/Operation Costs: Reduced carpet replacement costs resulting from use of replaceable carpet tiles Reduction of workstation relocation and/or service reconfiguration costs due to modular cabling and easily movable PVD service boxes Cash Flow Related Intangibles: Possible accelerated depreciation on access floor and carpet (non-fixed assets) Maintenance/Operation Costs: Cash Flow Related Intangibles: Maintenance/Operation Costs: Reduced failures of control components due to reduction of terminal units Reduced calls to maintenance regarding comfort complaints due to increased level of individual control Cash Flow Related Intangibles: Possible reduction in installation time of HVAC system reduces total construction time and enables earlier occupancy 136

149 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE 10.1 STANDARD FIRST COST COMPONENTS Components labeled as standard are considered to be a required or integral part of the installation of a UFAD system Raised Floor System The raised floor system is the component with the single largest cost increase of a UFAD system over a conventional air distribution system. Like any other building component, the cost of raised flooring can vary depending upon program requirements, location (shipping), union or non-union labor, and size of the project. Since a raised floor system forms an integrated service plenum that serves cabling, HVAC, and other distribution needs, assigning the entire cost of the floor installation to the HVAC system alone is unwarranted. Instead, the cost justification for raised floor systems should be based on the benefits of the entire (HVAC, power, voice, and data) service delivery system. The percentage of new raised floor office buildings using UFAD has increased significantly in recent years and is now near 40% [Hockman 2002]. The added first cost of the raised floor system must be weighed against other first cost savings and the flexibility and reduced costs associated with reconfiguring building services over the lifetime of the building. As discussed below, normally not all of the total floor plate area will be covered by an access floor system on any given floor of an office building Slab Modification and Preparation At the service core of a building, where restrooms, kitchenettes, and elevators are located, it has been common to omit a raised floor. However, raised floors can be used in the core area as well with added benefit and reduced cost. In particular, the cost of installing plumbing, including the setting of traps, has the potential to be reduced. Previously, a raised concrete core was poured in these areas in order to accommodate the difference in floor height between the service core and the finished raised floor level in the surrounding area. The raised concrete core is an expensive unit addition, although it represents a relatively small fraction of a building s floor area. In office buildings, the core consumes anywhere from 3% to 4% of a given floor plate. The HVAC and elevator vertical services typically account for about 4% to 5% of the floor plate. Conversely, raised flooring systems are leveled during installation, eliminating the necessity (and associated costs) of adding a finishing level to the floor. 137

150 CHAPTER 10 COST CONSIDERATIONS Cleaning and Sealing the Plenum The cost of cleaning and sealing the underfloor plenum is directly linked to the scheduling of the overall project (see Chapter 8). This is an add-on cost compared to conventional systems and can become larger than expected if mistakes are made. It is imperative for the project to be extremely well organized, as there is a long list of construction activities that would require a duplicate cleaning of the plenum prior to final placement of the finished raised floor surface Fire Detection and Sprinkler Systems The cost variables of fire safety systems will vary according to the local code requirements. If the raised floor is above a certain height, typically 18 in. (0.46 m), the code may require a sprinkler system. For this reason, many UFAD jobs limit the height of the underfloor plenum to less than 18 in. Local inspectors often require a smoke detection system in the plenum area. The fire safety cost components will not affect all underfloor projects. It will depend on the local jurisdiction. The significance of this category has much more to do with code requirements and interpretation and less with the design of the underfloor system DESIGN-DEPENDENT FIRST COST COMPONENTS Building components whose costs are more likely to change with the choice of a UFAD system are labeled as design-dependent UFAD System Design A very preliminary assessment shows that total costs for HVAC are in the range of $10-15/ft 2 ($ /m 2 ) (~60% core and 40% tenant improvement (TI)), which is roughly 10% of the total building cost. Generally, the core HVAC costs will remain about the same for both UFAD and overhead systems. Therefore, the primary difference will be in TI costs, about $4-6/ ft 2 ($43-65/m 2 ) or 4% of total building costs. This suggests that small differences in HVAC costs may not have a large impact on the overall costs and differences from traditional systems. Other system design-dependent factors that affect TI HVAC costs are described below Diffuser Type. Diffuser costs will be largely dependent on the choice of diffusers for the interior zones of the building, accounting for the majority of air delivery for a given floor plan. The cost of perimeter zone diffusers, often linear grilles or variable-air-volume (VAV) diffusers, makes up a relatively small portion of total diffuser costs. Although plastic diffusers have been the most commonly 138

151 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE installed to date, code officials in some jurisdictions may interpret current fire code language to require metal diffusers because of the fire/ smoke danger with plastic material. This can increase the cost of diffusers by as much as 35%. Also, automatic VAV diffusers and active (fan-driven) diffusers are generally more expensive than passive (manually controlled) diffusers. If a furniture- or partition-based TAC diffuser is selected, installed costs may also be higher. The added first cost and operating costs (fan energy use) of active floor or TAC diffusers must be traded off against the improved personal control of local airflow provided to a nearby occupant System Category. Constant air volume (CAV) vs. variable air volume (VAV) are the two predominant categories used some systems use a combination of the two. The category will impact the types and complexity of terminal devices used in perimeter and core areas. How perimeter zone heating is accommodated will also impact costs, but the cost differences between heating methods may be small unless it includes electrical vs. hot water Underfloor Plenum Ductwork and Partitioning. The extent that air highways and/or ducting and partitioning are used for distribution and zoning must be compared to the typically large amount of ductwork required for overhead air distribution in traditional designs. If a relatively open underfloor plenum configuration can be used, greater savings in the amount of required ductwork can be realized. The range of cost savings associated with the elimination of overhead ductwork is usually enough to offset a significant portion of the added cost of the raised floor system Controls. These costs may not be significantly affected if the basic zoning used for traditional systems is preserved in the UFAD system design Cable Management Systems Access floor systems provide a convenient platform for managing cable systems that meet the demands of modern office space. With the latest trend toward structured cabling, all telecommunications functions power, data, and audio/video are contained within a single wiring infrastructure. Although first costs of structured cabling will be higher than standard cabling, installation costs can be significantly reduced (working at floor level instead of up in the ceiling plenum), resulting in a net savings in overall cabling first costs. The flexibility of these integrated plug-and-play cabling systems makes them well suited for office spaces with high churn rates. In terms of efficiency and low-cost operation/maintenance, when structured cabling is installed 139

152 CHAPTER 10 COST CONSIDERATIONS as part of a raised floor system, in-house personnel using simple tools and standardized connector pieces can easily carry out reconfiguration. By comparison, traditional cabling systems consist of fixed outlets, connections, and long cable runs for which changes usually involve contracting outside labor and considerable disruption within the workplace (see Section on churn). In open plan offices with partitioned workstations, a second cabling cost consideration is the need for electrified workstations with built-in cable management systems. By delivering power, voice, and data cabling directly to virtually any location on the floor plate, raised floor systems can allow non-electrified partitions and furniture to be installed. Although there is a large range in price of workstations and personal furniture, electrified furniture can cost as much as 20% more than non-electrified equivalents Floor-to-Floor Heights With the use of an underfloor system, the heights from slab-to-slab have the potential to be reduced as much as 6 in. to 1 ft (0.15 to 0.3 m) per floor. The amount of reduction is dependent on the structural and plenum design of the baseline conventional building. Concrete flat slab construction can be especially effective at reducing floor-to-floor heights in comparison to standard steel beam construction. This new dimension correlates to a reduction of up to about 7% in vertical structural, thermal, and mechanical components. The reduced area of the curtain wall can be an important cost factor. The savings associated with this component will primarily apply to high-rise office building construction. It can be an important cost consideration in high-rise development where building heights are limited by local building codes Ceiling Finishes and Acoustical Treatment If air distribution and power and data cabling are installed under the floor, it opens up other design options for finishing the ceiling, including the elimination of the suspended ceiling tiles and plenum space above. In most cases, acoustical treatment of some kind will still be needed on the ceiling, particularly with the documented reduced mechanical noise levels for UFAD systems. Designing for acceptable acoustical privacy in open plan offices is challenging enough, and if the masking noise typically available with traditional HVAC design is absent, careful attention must be paid to this issue. 140

153 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE 10.3 LIFE-CYCLE COST COMPONENTS Life-cycle cost components include those building elements whose costs are affected over the lifetime of the building and represent the primary means by which building owners can expect to receive a greater return on their initial investment Churn (Reconfiguration) In modern businesses, churn is a fact of life; a 1997 survey found the national average churn rate (defined as the percentage of workers per year and their associated work spaces in a building that are moved, reconfigured, or undergo significant changes) to be 44% [IFMA 1997]. The cost savings associated with reconfiguring building services is a major factor in the decision to install access flooring. By integrating a building s HVAC and cable management systems into one easily accessible underfloor plenum, floor diffusers, along with all power, voice, and data outlets, can be placed almost anywhere on the raised floor grid. In-house maintenance personnel can carry out these reconfigurations at significantly reduced expense using simple tools and modular hardware. The amount of savings from churn is directly dependent on three variables, whose value may vary from building to building and from organization to organization: (1) annual churn rate, (2) cost savings of moves and reconfigurations per worker (large differences exist between simple moves and moves requiring renovation), and (3) amount of floor area per worker. Firms that are more likely to install underfloor systems are also, for the very same reasons, more likely to churn at a higher rate Operation and Maintenance The primary elements of operation and maintenance costs are: (1) the salaries of operations personnel required to service and maintain the HVAC system and to respond to occupant complaints, (2) replacement costs for equipment, and (3) energy costs. Any differences in commissioning costs between UFAD systems and traditional systems should also be considered. It may be difficult to obtain long-term maintenance cost data for UFAD systems since experience with these systems is limited in U.S. buildings. Some engineers believe that equipment maintenance costs for raised floor-based systems will be slightly higher than those for conventional systems. However, research suggests that the frequency of occupant complaints will be reduced when occupants are given some individual control over their local environment [Bauman et al. 1998]. Most practicing engineers agree that UFAD systems have the potential to save energy in comparison to traditional designs. To date, 141

154 CHAPTER 10 COST CONSIDERATIONS energy use data are only available on a project-by-project basis. Energy costs are difficult to predict since no reliable whole-building energy simulation tools are currently available that accurately model UFAD system performance (see Chapter 14) Tax Savings Under some circumstances, raised flooring and other movable components have the potential to qualify as personal property. As a sevenyear property, its cost could be depreciated at a favorable rate compared to standard flooring systems, which would normally be 39-year property. Seven-year property qualifies for double declining balance depreciation, while 39- year property depreciates at a slower rate and over a longer period of time. This potential savings should be investigated carefully and will be largely dependent on tax law interpretations Increased Property Value and Rents It is well documented that office tenants are willing to pay a premium for office space possessing the amenities they prefer. Naturally, market conditions will continuously fluctuate, but in assessing the real premium (if any) paid for a raised floor system, a secondary consideration is the premium that tenants are willing to pay for space with raised flooring. If raised flooring can be directly linked to increased rents and sales prices, the first cost of a raised floor system may be inconsequential by comparison Productivity and Health Research indicates that occupant satisfaction and productivity can be increased by giving individuals greater control over their local environment and by improving the quality of indoor environments (thermal, acoustical, ventilation, and lighting). Improved ventilation and thermal environments, which well-designed UFAD systems can provide, have also been associated with a reduction in the prevalence or severity of adverse indoor health effects [Fisk 2000]. The financial implications of improving productivity or reducing absenteeism caused by illness by even a small amount have the potential to be very large as employee salary and benefits costs typically make up at least 90% of all costs (including construction, operation, and maintenance) over the lifetime of a building. Nationwide, a mere 1% increase in worker productivity would translate into a potential annual cost benefit of $25 billion. In today s competitive world economy, a company s employees make up its most valuable economic assets. Protecting and improving the productivity of these employees will have a strong influence on future investments. 142

155 Chapter 11 Standards, Codes, and Ratings Since UFAD technology is relatively new to the building industry, its characteristics may require consideration of unfamiliar code requirements and, in fact, may be in conflict with the provisions of some existing standards and codes. Applicable standards should be reviewed carefully; revisions and exceptions that are more compatible with UFAD technology will likely be forthcoming as additional research results are obtained. Local building codes and the interpretations of local officials should be considered early in the design process of a building using underfloor air supply plenums. Experience has shown that the first UFAD project in an area governed by an unfamiliar jurisdiction will usually end up establishing the ground rules for code interpretations on future projects. Listed below are brief discussions of the applicable building standards and codes that have important provisions related to the design, installation, and operation of UFAD systems. In addition, a brief description of the LEED (Leadership in Energy & Environmental Design) Rating System is provided ANSI/ASHRAE STANDARD : THERMAL ENVI- RONMENTAL CONDITIONS FOR HUMAN OCCUPANCY [ASHRAE 1992] Earlier versions of Standard 55 were based on the assumption of a well-mixed and uniformly conditioned environment. UFAD systems, however, usually involve greater variability of thermal conditions over both space and time. The effect of providing occupant control has not been fully taken into account, although it is well established that occupants will tolerate greater fluctuations in environmental conditions if they have control over them. The rather strict air velocity limitations 143

156 CHAPTER 11 STANDARDS, CODES, AND RATINGS Figure 11.1 Air speed required to offset increased temperature [ASHRAE 1992]. that were specified in the previous version of Standard 55 were incompatible with the increased local air velocities that are possible with UFAD and task/ambient conditioning (TAC) systems. The current version of ASHRAE Standard 55 [ASHRAE 1992] was revised to allow higher air velocities than the previous version of the standard, if the occupant has control over the local air speed. Figure 3 in Standard (reproduced in Figure 11.1) was added to show the air speed required to offset increases in temperature above those allowed in the summer comfort zone. For example, Figure 11.1 indicates that at equal air and radiant temperatures (tr ta = 0), a local air speed of 150 fpm (0.75 m/s) can offset a temperature rise of about 4.4 F (2.4 C) for a primarily sedentary building occupant wearing 0.5 clo. The figure is based only on sensible heat transfer; total cooling would be expected to be higher if latent effects are taken into account. Standard also specifies allowable air speeds as a function of air temperature and turbulence intensity with the objective of avoiding unwanted drafts when the occupant has no direct local control. At warmer temperatures, however, occupants will desire additional cooling. Increased air movement (and turbulence) is an easy way of achieving such direct occupant cooling. Standard allows these velocity limits, based on turbulence intensity level, to be exceeded if the occupant has control over the local air speed. 144

157 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE 11.2 ANSI/ASHRAE STANDARD : VENTILATION FOR ACCEPTABLE INDOOR AIR QUALITY [ASHRAE 2001b] Standard provides guidelines for the determination of ventilation rates that will maintain acceptable indoor air quality. This most recent version of Standard 62 allows some adjustment in ventilation rates based on the ventilation effectiveness (E v ) of the air distribution system. Mixing-type air distribution systems can at best achieve a perfectly mixed space, defined as having an E v of 1.0, as determined in accordance with ASHRAE Standard 129 (see below). By definition, mixing-type systems cannot provide preferential ventilation (E v > 1), in which some credit could be obtained for improved air change effectiveness at the breathing level in the space. Displacement ventilation systems that deliver supply air at low velocity near floor level and extract air at ceiling level are known to provide improved ventilation effectiveness in the occupied zone (see Chapter 2). This performance characteristic is being addressed more specifically in the newest addendum of Standard 62 [ASHRAE 2003] in which default values for E v are recommended for different air distribution system configurations and modes of operation. These values can and should be used to determine required outdoor air quantities if it is decided to not measure E v directly. The recommended values of E v are (1) 1.2 for displacement ventilation system, (2) 1.0 for an overhead system in cooling mode, and (3) 0.8 for an overhead system in heating mode (known to cause shortcircuiting). UFAD systems are not explicitly addressed since more definitive research on ventilation effectiveness is still needed, but it is expected that E v for UFAD with floor diffusers will be less than or equal to 1.2 but higher than 1.0. Research has shown that E v for personally controlled TAC diffusers can be significantly higher than 1.2 when the supply air is directed toward the occupant s breathing level [Faulkner et al. 2002; Melikov et al. 2002]. It has not yet been determined how to apply these elevated performance numbers for TAC diffusers in Standard 62, since ventilation performance will change when an individual moves away from their local air supply or decides to turn it off. Standard sets minimum ventilation rates for office space and conference rooms at 20 cfm (9.4 L/s) per person and reception areas at 15 cfm (7.1 L/s) per person. In the design and operation of TAC systems containing a large number of occupant-controlled supply modules, some means must be provided to ensure that minimum ventilation rates are maintained, even when people choose to turn off their local air supply. 145

158 CHAPTER 11 STANDARDS, CODES, AND RATINGS 11.3 ANSI/ASHRAE/IESNA STANDARD : ENERGY STANDARD FOR BUILDINGS EXCEPT LOW-RISE RESI- DENTIAL BUILDINGS [ASHRAE 2001c] ASHRAE Standard 90.1 describes requirements for the energyefficient design of new buildings intended for human occupancy. In Section 9.5.2, the prescriptive criteria for zone controls state that there can be no simultaneous operation of heating and cooling systems to the same zone. Some of the unique aspects of UFAD and TAC systems may be in conflict with this requirement. For example, if occupants have control of supply air temperature for heating or cooling from their local diffusers, situations may occur in which some people are requesting heating and others are requesting cooling at the same time within the same zone. In another example, with underfloor air distribution configured to have fan coil units in the perimeter fed from cool plenum air from the interior zone, if there is a call for heating, this will require local reheating of the underfloor supply air to satisfy the heating demand (see Title 24 below for further discussion). These and other relevant situations should be carefully considered as there are exceptions to the criteria described in Standard 90.1 and perhaps subtle differences in the operation of UFAD and TAC systems compared to a conventional overhead air distribution system ANSI/ASHRAE STANDARD : METHOD OF TEST- ING FOR ROOM AIR DIFFUSION [ASHRAE 1990] ASHRAE Standard is the only currently available building standard for evaluating the air diffusion performance of an air distribution system. The current version of Standard 113, however, is based on the assumption of a single uniformly mixed indoor environment, as provided by a conventional overhead air distribution system. This assumption is not appropriate for evaluating the performance of UFAD and TAC systems that deliver conditioned air directly into the occupied zone of the building through supply outlets that are in close proximity to and under the control of the building occupants. UFAD and TAC systems, therefore, not only promote thermal stratification in the space but also may actually encourage other nonuniformities between workstations. Efforts are now underway to revise Standard 113 to include new methods of performance evaluation that are applicable to air distribution systems that deliver air directly into the occupied zone of the building, including UFAD, TAC, and displacement ventilation systems. 146

159 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE 11.5 ASHRAE STANDARD : MEASURING AIR CHANGE EFFECTIVENESS [ASHRAE 1997] ASHRAE Standard describes a test method for evaluating an air distribution system's ability to provide required levels of ventilation air to the building occupants. The results of the tests may be used to determine compliance with ASHRAE Standard 62. If this test method demonstrates that enhanced ventilation effectiveness is provided at breathing level by a UFAD or TAC systems, then credit may be taken by reducing the required outdoor air quantity accordingly TITLE-24: CEC SECOND GENERATION NONRESIDEN- TIAL STANDARDS [CALIFORNIA ENERGY COMMISSION 2001] The CEC Nonresidential Standards (Title-24) defers to applicable ASHRAE standards in most cases. Title-24 does, however, address a few areas that should be taken into consideration in the operation of UFAD systems in California. Title-24 mandates off-hour controls for central HVAC systems and stipulates that the largest sized zone that can be controlled in isolation is 25,000 ft 2 (2,300 m 2 ). In buildings with large floor plates, this size limitation will require that the underfloor plenum be divided into smaller zones using underfloor partitions or other suitable means. Local fire codes may require that the plenum be divided into considerably smaller zones. Title-24 addresses simultaneous heating and cooling, particularly in relation to variable-air-volume (VAV) system operation. When changing over from cooling to heating in a zone, the supply volume must first be reduced to 30% of peak before beginning the heating cycle. This has implications for UFAD system designs that employ an open plenum in which variable-speed fan-coil units in the perimeter draw their primary air from the interior zone of the plenum. On a call for heating in the winter or early morning, fan speeds in these perimeter units will need to be reduced. In addition, it may be difficult to meet this requirement if swirl diffusers are placed in the perimeter zone, since they will not automatically reduce their cooling supply volume in heating mode. The use of electric resistance heating is prohibited according to the prescriptive method in Title-24 for determining a building s allowable energy performance. However, if the alternative computer simulation method is used to predict a building s energy performance for comparison with the Title-24 target energy budget, it may be possible to trade off the use of electric heat with energy savings in other UFAD system 147

160 CHAPTER 11 STANDARDS, CODES, AND RATINGS components (e.g., improved chiller efficiency or increased economizer operation). Title-24 requires thermostatic zone controls with adjustable setpoints. Since TAC systems may maintain temperature differences between locally conditioned zones (workstations) and unconditioned or centrally conditioned areas of the workplace (e.g., corridors), attention should be paid to placing zone controls in representative locations. In general, Title-24 focuses on the effects of overall systems. As a result, the integration between the local and central controls should be carefully considered. The effects of individual thermal preferences on overall air quality and comfort should also be taken into account. Although the current version of Title 24 does not specifically address underfloor air distribution, if enough supporting energy- and cost-saving data can be obtained, UFAD systems could be added to the subsequent revision (three-year cycle) NFPA 90A: STANDARD FOR THE INSTALLATION OF AIR- CONDITIONING AND VENTILATING SYSTEMS [NFPA 1999] NFPA 90A is the most widely used and referenced code in relation to the installation of HVAC systems. This code contains language written several years ago before the widespread introduction of UFAD systems that, depending on one s interpretation, appears to prohibit or restrict the application of underfloor air supply plenums. Selected examples of key language that most frequently come up in the review of an UFAD installation by code officials are described below. In the section titled Location of Air Outlets (Section ), which applies equally to inlets, the code states air outlets shall be located at least 3 in. (7.6 cm) above the floor. This appears to rule out the use of floor diffusers; however, an exception is given as where provisions have been made to prevent dirt and dust accumulations from entering the system. Thus, any floor diffuser without a basket-type device or other means of collecting dirt and debris located underneath the access floor surface would not be acceptable. Where linear grille diffusers, often located in perimeter zones, are specified, an alternative means of collecting dust/dirt must be provided. In addition, outlets located less than 7 ft (2.1 m) above the floor must be protected by a grille or screen through which a ½-in. (1.3-cm) sphere cannot pass. Both the collection device and ½-in. grille spacing requirements are easily satisfied by most commercially available diffuser models, thereby complying with the exception identified in NFPA 90A. To fully 148

161 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE satisfy the intent of the code language to ensure a clean air distribution system, regular vacuuming of the dust/dirt collection devices should be included in the maintenance schedule. In terms of the combustibility of diffusers, Sections and state that air outlets and inlets shall be constructed of non-combustible material or a material that has a maximum flame spread index of 25 and a maximum smoke developed index of 50. There has been considerable debate about the acceptability of diffusers made from polycarbonate materials, which appear to violate the intent of NFPA 90A. For nearly 20 years several established diffuser models have been regularly used in UFAD system installations and yet are made from a plastic material that satisfies the required flame spread index but cannot comply with the smoke index of 50. One argument commonly put forward in defense of plastic diffusers is that the smoke test protocol (ASTM E 84, NFPA 255 requiring that a large 25-ft [7.6-m] sample of the material be burned) cannot reasonably be applied to polycarbonate material. In any event, metal diffusers fully comply with NFPA 90A, and designers should proceed cautiously with the use of plastic materials unless specific exception has been granted by the local building code authority. The combustibility of material in the underfloor plenum is also governed by NFPA 90A in Section The space between the top of the finished floor and the underside of a raised floor shall be permitted to be used to supply air to the occupied area, or return or exhaust air from the occupied area, provided that the following conditions are met: 1. All materials exposed to the airflow shall be noncombustible or limited combustible and shall have a maximum smoke developed index of 50. An exception is given, however, for materials ranging from electrical wires, cables, and optical fiber cables to raised floor panels and fire sprinkler piping. In addition to referencing the codes to which each exempt material must comply, these materials must have a maximum peak optical density of 0.5 or less, an average optical density of 0.15 or less, and a maximum flame spread distance of 5 ft (1.5 m) or less when tested in accordance with the specified test method. Refer to NFPA 90A for additional conditions relevant to the underfloor plenum. In general, placing wires and cables in an air supply plenum is not a problem as long as they are contained in conduit or are rated to be noncombustible. 149

162 CHAPTER 11 STANDARDS, CODES, AND RATINGS 11.8 UNIFORM BUILDING AND OTHER APPLICABLE CODES Local fire codes sometimes place restrictions on the size of open supply air plenums without any smoke breaks in the form of partitions separating the plenum into smaller zones. These fire codes may limit the total area (e.g., less than 3,000 ft 2 [280 m 2 ]) and horizontal dimension in one direction (e.g., less than 30 ft [9 m]) of an unobstructed underfloor air supply plenum. A typical underfloor plenum contains a low level of combustible materials; thus, in certain codes plenums under 18 in. (45 cm) in height do not require sprinklers. The issue of whether sprinklers need to be installed in a plenum is contentious for a number of reasons. First, as electric cabling is typically the only source of fire risk, water is not the best source of fire suppression. Also, if fire/smoke detectors are required by code to be placed within the floor plenum, the question arises as to the effectiveness of standard detection devices within such a low-height cavity. Fundamentally, the codes governing underfloor plenums should be no different than those for ceiling plenums LEED (LEADERSHIP IN ENERGY & ENVIRONMENTAL DESIGN) RATING SYSTEM The United Stated Green Building Council (USGBC) established the LEED rating system with the intent of creating a method to rate the environmental performance of a building. The system works by assigning points to various design and construction process features. Depending on the overall number of points a building earns, it can achieve a Certified, Silver, Gold, or Platinum rating. The LEED rating system consists of five major categories: 1. Sustainable sites 2. Water efficiency 3. Energy and atmosphere 4. Materials and resources 5. Indoor environmental quality In each category, there are both prerequisites and credits. For a building to achieve any level of certification, it must meet the requirements of all the prerequisites. Prerequisites earn no points. Each credit then is assigned a point value or range of point values that can be earned for the building. UFAD systems have relevance in the Energy and Atmosphere as well as Indoor Environmental Quality sections of LEED. In the Energy and Atmosphere section, Credit 1 allows points for optimizing the energy performance of a building. 150

163 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE In the Indoor Environmental Quality section of LEED, UFAD systems can be relevant related to Credit 2 Increase Ventilation Effectiveness. As discussed earlier in this section, UFAD systems may have a higher ventilation effectiveness than overhead systems. Credit 2 requires that the ventilation effectiveness of the installed system be designed to achieve an E v above 0.9 as determined by ASHRAE Standard for measuring air change effectiveness. Compliance is demonstrated through testing or by a narrative and calculations describing how the high-performance system was designed. 151

164

165 Chapter 12 Design Methodology This chapter provides a concise list of issues to be considered, and decisions to be made, during the design process. For more detailed discussions and background information, the reader is referred to other sections of this guide. The focus is on those areas in which the design of UFAD systems differs from conventional air distribution system design. For further reading and design guidance, see Spoormaker [1990], Sodec and Craig [1991], Houghton [1995], McCarry [1995], Shute [1995], Bauman and Arens [1996], Bauman et al. [1999a], Bauman et al. [2000a], and AEC [2000] UFAD VS. CONVENTIONAL OVERHEAD SYSTEM DESIGN UFAD systems are similar to conventional overhead systems in terms of the types of equipment used at the cooling and heating plants and primary air-handling units (AHU). Key differences arise with UFAD systems in their use of an underfloor air supply plenum, warmer supply air temperatures into the room, delivery of air in the near vicinity of occupants (with or without individual control) and the resulting floor-to-ceiling air flow pattern, and the solutions used for perimeter systems. In order to successfully employ a UFAD system, it is essential that the implications of these differences be considered, starting at an early stage in the design process BUILDING STRUCTURE CONSIDERATIONS Building Plan The modularity of all components of raised floor systems can be an advantage in space planning, particularly over large open plan areas. 153

166 CHAPTER 12 DESIGN METHODOLOGY Consider the compatibility of anticipated building plan geometries with the dimensions of the floor grid established by the raised floor system. Raised floor panel dimensions: 24 in. (610 mm) square Underfloor plenum pedestal spacing: same as floor panels, e.g., 24 in. (610 mm) New Construction In new construction, underfloor air distribution has the potential to achieve a reduction in floor-to-floor heights compared to projects with ceiling-based air distribution. This is accomplished by reducing the overall height of service plenums and/or by changing from standard steel beam construction to a concrete (flat slab) structural approach. A single large overhead plenum to accommodate large supply ducts and other building services can be replaced with a smaller ceiling plenum for air return and piping for sprinklers combined with a lower-height underfloor plenum for unducted air flow and other building services [Kight 1992]. Floor-to-floor heights for overhead systems using steel beam construction can also be reduced by using beam penetrations for ducts and other building services. In this comparison, if steel beam construction is used in both designs, floor-to-floor heights should be equal or lower for UFAD buildings. Significantly reduced vertical height requirements can be achieved using concrete flat slab construction, which is usually more expensive than steel beam construction but is preferred for underfloor systems due to thermal storage benefits. In the example shown in Figure 12.1, the underfloor/flat slab configuration allows 10 in. (0.25 m) to be saved in floor-to-floor height compared to overhead/steel beam system design. Even greater savings (up to 22 in. [0.56 m]) can be realized if the ceiling plenum is completely eliminated, exposing the concrete ceilings and providing an opportunity for creative internal design, enhancing daylighting and artificial lighting effects. However, if the conventional suspended acoustic tile ceiling is eliminated, leaving an exposed concrete ceiling or other configuration, careful consideration must be made of the acoustic and/or lighting quality of the space. Designers will also need to consider possible conflicts with local codes (e.g., fire code). High side-wall return is the most common return air configuration for this exposed ceiling design. Table 12.1 presents a comparison of typical floor-to-floor dimensions for a midsize (5-10 stories), high-tech class A office building (assuming a 40-ft clear span between columns). Dimensions are shown 154

167 UNDER FLOOR AIR DISTRIBUTION DESIGN GUIDE Figure 12.1 Comparison of typical floor-to-floor heights for ceiling-based and underfloor air distribution systems. 155

168 CHAPTER 12 DESIGN METHODOLOGY Table 12.1: Comparison of Typical Floor-to-Floor Heights Building Element Structure Concrete Metal deck Ceiling plenum Floor-toceiling Underfloor plenum Total floor-to-floor height Steel beam Fireproofing Steel Beam Construction with Overhead Air Distribution 2.5 in. (65 mm) 2.5 in. (65 mm) 21 in. (530 mm) 2 in. (50 mm) Steel Beam Construction with Underfloor Air Distribution Concrete Metal deck Steel beam Fireproofing 2.5 in. (65 mm) 2.5 in. (65 mm) 21 in. (530 mm) 2 in. (50 mm) Concrete Flat Slab Construction with Underfloor Air Distribution Concrete floor Concrete beam 8 in. (200 mm) 12 in. (305 mm) in. ( mm) 8-12 in. ( mm) 8-12 in. ( mm) 9 ft (2.70 m) 9 ft (2.70 m) 9 ft (2.70 m) 13 ft 1 in.-13 ft 6 in. ( m) in. ( mm) in. ( mm) 13 ft-13 ft 10 in. ( m) 12 ft 4 in.-13 ft 2 in. ( m) 156

169 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE for a conventional overhead system with steel beam construction, and two UFAD system configurations, one with steel beam and one with flat slab construction. Underfloor plenums accommodating both cable/electrical distribution and an UFAD system are often deeper than those employed solely for cable management purposes. However, the additional height required for acceptable airflow performance is not large, based on recent research results [Bauman et al. 1999a]. Underfloor plenum heights are usually determined by largest HVAC components (e.g., fan coil units, terminal boxes, ducts, dampers) located under the floor, requirements for underfloor cabling, and additional clear space for underfloor air flow (usually 3 in. [76 mm] minimum) Retrofit Applications Due to the tremendous size of the existing building stock, retrofit construction will play an important role in the future for the building industry. Projects requiring the addition of an HVAC system often encounter the problem of having limited space for accommodating ducts and other components. Because of the comparable dimensions discussed above, UFAD can be quite feasible in retrofit projects. The most practical retrofit applications will involve (1) buildings with an existing raised floor system (no UFAD), (2) projects where the existing air distribution system (typically overhead) will be renovated, and (3) high ceiling spaces, such as warehouses [Webster et al. 2002c]. The use of raised floor systems in warehouse type buildings can also eliminate problems associated with existing uneven slab surfaces. The biggest challenge with the installation of a raised floor system in an existing building is that stairs, elevators, bathrooms, and other core facilities exist at the original floor level. While elevator stops can be reset, other facilities will usually require steps, ramps, or some other transitional element. The installation of a raised floor system can be less disruptive than that of ducting for overhead systems as the floor can be easily installed, and removed, as an independent platform leaving relatively few structural scars. This issue is important in buildings where maintaining the integrity of the existing building structure is important for heritage/cultural/structural reasons [Guttmann 2000]. Furthermore, installation 157

170 CHAPTER 12 DESIGN METHODOLOGY can be a relatively dry process, once the concrete structural slab has been adequately sealed, minimizing damage to other building elements DETERMINATION OF SPACE COOLING AND HEATING LOADS Cooling and heating loads for a building with a UFAD system are calculated in much the same manner as for a conventional overhead (OH) system. For more information, see Chapter 29 in the 2001 ASHRAE Handbook Fundamentals and Pedersen et al. [1998]. However, the determination of design cooling air quantities must take into account key differences between these systems Space Cooling Load Calculation This section discusses the ways that conventional load calculation methods used for OH systems may be changed to capture performance characteristics of stratified spaces associated with UFAD systems. Although not discussed below, cooling loads can also be affected by heat transferred to the underfloor plenum air supply, either through the slab from the adjacent return air plenum or through the floor panels from the room. Chapter 4 and Section 12.7 address this issue in greater detail Mixing Assumptions for UFAD and OH Cooling Load Calculation. The following load calculation example demonstrates what happens if the assumption of a fully mixed room is applied to a UFAD system. The standard room energy-balance equation for an OH system is as follows: Btu Q = CFM T h cfm F where Q = heat loads in a room, Btu/h, CFM = airflow moving through a room, ft 3 /min, and T = temperature difference between the room setpoint temperature and the supply air temperature, F. The validity of this equation relies on two assumptions that the room is at steady state and that the room is fully mixed. The assumption 158

171 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE that a room is fully mixed is not valid for UFAD systems, and, as such, this simple room energy balance equation cannot be applied here. Consider this example of a room in cooling mode, which sheds some light on the common question asked of UFAD systems, Do UFAD systems need more air than OH systems? In fact, current research indicates that airflow rates are very comparable to overhead systems [Webster et al. 2002a] How UFAD Stratification Affects Loads. Understanding how air becomes stratified in spaces employing UFAD is key to developing a correct cooling load calculation model. As discussed in Chapter 2, the floor-to-ceiling air flow pattern driven by rising thermal plumes in UFAD systems produces a vertical temperature gradient in the space. Unless air supply quantities are exceedingly high, a stratification height is established in the room that divides the room into an upper zone and one or two lower zones (depending on diffuser throw height). In general, the fact that once room air has risen above this stratification height it will not reenter the lower zones represents a fundamental difference from the fully mixed room assumed in OH system 159

172 CHAPTER 12 DESIGN METHODOLOGY load calculations. This principle allows convective heat gains from sources above the stratification height in the room to be exhausted directly at ceiling level and therefore to not be included in the air-side load. In practice, to optimize thermal comfort, ventilation, and energy performance, a good design goal is to maintain the stratification height near the top of the occupied zone (4-6 ft [ m]) above respiration level, depending on whether the primary occupancy is sitting or standing. The fact that the air in the top portion of a room above the stratification height is warmer than the air in the bottom portion is used to the advantage of UFAD systems in that it is primarily only the air temperatures in the lower portion of the room that determine the conditions that affect the comfort of the occupant (see further discussion of thermal comfort in Chapter 3). In the following discussion, this lower portion of the room will be called the occupied zone. Air above the occupied zone can be allowed to warm up beyond what would otherwise be comfortable temperatures. The zone above the occupied zone will be called the unoccupied zone Assigning Heat Gains to Occupied and Unoccupied Zones. Heat loads are physically located in either the occupied or unoccupied zone. For example, a ceiling pendant-mounted light fixture is located in the unoccupied zone. A computer sitting on a desk is located in the occupied zone. Figure 12.2 is a schematic diagram showing some typical loads in an office. The heat from a load is not necessarily allocated only to the occupied or unoccupied zone where the load physically resides. Heat sources must be analyzed based on their convective and radiant components, a subject addressed by Hosni et al. [1999]. Both the location and the convective/radiant split characterizing a specific type of heat load determine where the heat from a load needs to be assigned. Loudermilk [1999] has described a space heat gain analysis using this approach based on empirical estimates. Unfortunately, no researchbased guidance exists to guide the assignment of loads to the occupied and unoccupied zones. This is particularly true for heat sources located near the stratification height (e.g., most desktop computers and equipment). Using the same examples as above, the convective portion of the light fixture can logically be assigned to the unoccupied zone, but a good deal of the radiant portion of that energy needs to be assigned to the occupied zone. In the case of the computer, some amount of both the convective and radiant portions of the load can likely be assumed to be in the unoccupied as well as the occupied zones. Table 12.2 doc- 160

173 UNDERFLOOR AIR DISTRIBUTION DESIGN GUIDE Figure 12.2 Typical loads in an office showing convective and radiant split. Table 12.2: Radiant/Convective Splits for Typical Office Heat Sources Heat Source Radiant Portion Convective Portion [%] [%] Transmitted solar, no inside shade Window solar, with inside shade Absorbed (by fenestration) solar Fluorescent lights, suspended, unvented Fluorescent lights, recessed, vented to return air Fluorescent lights, recessed, vented to return air and supply air Incandescent lights People, moderate office work Conduction, exterior walls Conduction, exterior roof Infiltration and ventilation Machinery and appliances 20 to to