ventilation efficiencies of desk-mounted task/ambient conditioning systems

Indoor Air 1999; 9: 273–281 Copyright c Munksgaard 1999Printed in Denmark. All rights reserved

INDOOR AIRISSN 0905-6947

Ventilation Efficiencies of Desk-Mounted Task/AmbientConditioning Systems

DAVID FAULKNER1*, WILLIAM J. FISK1, DOUGLAS P. SULLIVAN1 AND DAVID P. WYON2

Abstract In laboratory experiments, we investigated two task/ambient conditioning systems with air supplied from desk-mounted air outlets to efficiently ventilate the breathing zone ofheated manikins seated at desks. In most experiments, the taskconditioning systems provided outside air while a conventionalventilation system provided additional space cooling but no out-side air. Air change effectiveness (i.e., exhaust air age divided byage of air at the manikin’s face) was measured with a tracer gasstep-up procedure. Other tracer gases simulated the release ofpollutants from nearby occupants and from the floor covering,and the associated pollutant removal efficiencies (i.e., exhaust airconcentrations divided by concentrations at manikin’s face) werecalculated. High values of air change effectiveness (∂1.3 to 1.9)and high values of pollutant removal efficiency (∂1.2 to 1.6) weremeasured when these task conditioning systems supplied 100%outdoor air at a flow rate of 7 to 9 L sª1 per occupant. Air changeeffectiveness was reasonably well correlated with the pollutantremoval efficiency. Overall, the experimental data suggest thatthese task/ambient conditioning systems can be used to improveventilation and air quality or to save energy while maintaining atypical level of IAQ at the breathing zone.

Key words Task/Ambient conditioning; Air changeeffectiveness; Pollutant removal efficiency; Ventilation; Indoor airquality.

Received 15 February 1999. Accepted for publication 15 April 1999.

C Indoor Air (1999)

A field study indicated that a larger than expected per-centage of office workers were not satisfied with theirthermal environment while at work, even thoughASHRAE Standard 55 (ANSI/ASHRAE, 1992) thermalcomfort conditions were met (Schiller et al., 1988). Vari-ations in thermal preferences, clothing, and activitylevels are one source of thermal discomfort. While

1Indoor Environment Department, Lawrence Berkeley National Laboratory, 1 Cyclotron Road, MS 90–3058, Berkeley, California, 94720, USA, Tel:π1 510 486 7326, Fax: π1 510 486 6658, e-mail: D–Faulkner/lbl.gov, 2Controls Group, Johnson Controls Inc., Milwaukee, Wisconsin, USA,*Author to whom correspondence should be addressed

some workers prefer to be cooler, other workers innearby workstations prefer to be warmer. Since a singleuniform thermal environment will not satisfy allworkers, some manufacturers and researchers havepostulated that giving workers control of thermal con-ditions at their workstation will increase thermal com-fort and possibly increase worker productivity. Task/ambient conditioning (TAC) is a method for providingoccupants with control of a local supply of air so thatthey can adjust their individual thermal environment.Controlled variables could be the supply-air tempera-ture, velocity, direction, the ratio of room air to mainair handling system supply air, and the radiant tem-perature. TAC systems may provide all or part of theconditioned air to the occupied space.

TAC systems also have the potential to improve ven-tilation at the occupant’s breathing zone because theycan provide supply air preferentially toward thebreathing zone. Supply air from TAC systems usuallycontain a high percentage of outside air, which gener-ally does not contain a high concentration of indoor-generated pollutants. The air supply outlets of currentTAC systems are located at the floor, mounted on thedesk, or incorporated within the workstation par-titions.

Previous research concerning TAC systems hasshown that it is possible to improve ventilation at thebreathing zone (relative to ventilation with perfectmixing) using TAC systems with desk-mounted(Faulkner et al., 1993) or floor-mounted (Fisk et al.,1991; Faulkner et al., 1995) air supply outlets. However,the degree of enhancement has been generally small(∂20–40%) and observed primarily during operatingconditions that are likely to be uncomfortable becauseof high air velocities.

In this paper, to characterize the improvement in

Faulkner, Fisk, Sullivan and Wyon

ventilation at the breathing zone we use two ‘‘venti-lation efficiency’’ parameters. The first is the air changeeffectiveness (ACE), defined as the age of air thatwould occur throughout the room if the air was per-fectly mixed, divided by the average age of air whereoccupants breathe (ASHRAE, 1997). The age of air ata particular location is the average time elapsed sincemolecules of air at that location entered the building.Because the average age of air exiting the room is iden-tical to the age of air that would occur throughout theroom if the indoor air were perfectly mixed (Sandbergand Sjoberg, 1983), the ACE is also the exhaust-air agedivided by the average age of air where occupantsbreathe. A short-circuiting flow pattern increases theroom-air age and causes ACE to be less than unity. Per-fect mixing results in an ACE of unity. Preferentiallyventilating the breathing zone with outside air willcause the ACE to be greater than unity.

The second ventilation efficiency parameter is thepollutant removal efficiency (PRE). Although rarelyused, the PRE is a more direct indicator than ACE ofthe effectiveness of the ventilation process in control-ling occupants’ exposures to indoor-generated air pol-lutants. We define the PRE as the time-average concen-tration of pollutants in the exhaust air divided by thetime-average concentration where occupants breathe.The PRE is a function of the locations of pollutantsources and the nature of the pollutant emission pro-cess, e.g., emitted with or without momentum. PREsalso depend on the indoor airflow pattern and by theremoval of pollutants by surfaces. If the pollutantsources are distributed uniformly in a space and thepollutant is emitted without momentum, then we ex-pect the values of PRE to be correlated (not necessarilyidentical) with the values of ACE.

Prior research has shown not only improved venti-lation efficiency while using TAC systems, but alsoincreased thermal comfort and occupant satisfaction(Bauman and Arens, 1996; Bauman et al., 1998). Inaddition, energy savings may be realized using TACsystems if they use occupancy sensors to turn off thesystems when workstations are unoccupied. The sav-ings realized from the use of occupancy sensors willdepend upon the amount of time the workstation isunoccupied. Energy savings may also be realizedwhile the building is operating in a cooling mode bymaintaining temperatures away from the worksta-tions warmer than temperatures at the workstations(Borgers and Bauman, 1994; Seem and Braun, 1992).Another potential benefit to using TAC systems isthe cost savings associated with improved workerproductivity and reduced sick leave (Kroner andStark-Martin, 1992).

274

Research ObjectivesThe objectives of this research were to determine theACE and the PREs obtained through the use of twodesk-mounted TAC systems operating in conjunctionwith a conventional (ceiling supply and return) heatingventilation and air-conditioning (HVAC) system.

Task/Ambient Conditioning SystemsThe first TAC system evaluated (Figure 1) is the Per-sonal Environmental Module (PEM) manufactured byJohnson Controls Inc. The PEM contains a thermallyand acoustically insulated mixing box that enclosesfans, dampers, and electronic controls. The damperscontrol the percentages of recirculated room air andventilation system supply air (thus controlling the tem-perature of air supplied). This mixing box is installedunderneath a desk. Air is supplied to the mixing boxfrom a dedicated air-handling unit (AHU) via a flexibleduct that is connected directly to supply ducts or to anunder-floor supply air plenum. Another stream of airenters the mixing box through an air inlet beneath thedesk. After passing through the mixing box, the mix-ture of AHU supply air and room air exits throughtwo desk-top-mounted air supply outlets located at theback corners of the desk. The PEM has a control panelfrom which the following parameters can be changed:air flow rate, percent of room air that is mixed in themixing box with air from the main AHU, volume of awhite noise generator, dimming of a task light, andpower to a radiant heating panel. The air supply out-lets on top of the desk can be rotated 360æ in the hori-zontal direction and contain movable vanes which can

Fig. 1 Air supply from the PEM (solid line) toward occupant,entrained in the thermal plume around occupant (dashed line)

Ventilation Efficiencies of Task/Ambient Conditioning Systems with Desk-Mounted Air Supplies

Fig. 2 Air supplied horizontally from the Climadesk (solid line)toward occupant, entrained in the thermal plume around occu-pant (dashed line)

Fig. 3 Air supplied vertically from the Climadesk (solid line) to-ward occupant, potentially entrained in the thermal plumearound occupant (dashed line)

be rotated ∫30æ in the vertical direction. The normalrange of supply air flow from the PEM is 6 to 71 L/s,although most of our tests utilized supply flow ratesat the lower end of this range.

The second TAC system (Figures 2 and 3) is the Cli-madesk manufactured in Sweden by Mikroklimat. TheClimadesk includes a panel attached to the undersideof a conventional desk, connected by a flexible duct toa portable fan-filter unit placed next to the desk. Thefan-filter unit is supplied with air from the same dedi-cated AHU as the PEM. Supply air exits two adjustableoutlets underneath and close to the underside of theworksurface. These outlets are located close to a seated

275

worker’s knees and direct air horizontally above theseated worker’s thighs and towards the torso. In thehorizontal plane, the angle of air supply from theseoutlets is manually adjustable. There is an additional,non-adjustable outlet at the front edge of the desk,which directs air almost vertically upwards, butslightly away from a seated worker, to minimize un-wanted cooling of the torso. A variable proportion ofthe total air flow (0–100%) can be directed to this thirdoutlet, as desired by the occupant. All three outlets arefitted with metal ‘‘honeycomb’’ final sections whosepurpose is to induce parallel flow and reduce entrain-ment of the vertical jet. The maximum primary airflowfrom the Climadesk is 7 L/s. A controllable radiantheating panel, attached to the underside of the desk,can provide heating of the lower part of the body.

Research MethodsAll experiments were performed in a controlled en-vironment chamber (CEC) with a 5.5 m by 5.5 m floorand 2.5 m high ceiling. The CEC resembles a modernoffice space and has provisions for a high degree ofcontrol over the method of ventilation and the indoorthermal environment (Bauman and Arens, 1988). Fig-ure 4 shows the floor plan of the workstations in thechamber. During experiments, two identical TAC sys-tems were operated, one in workstation 2 (WS2) andone in WS3, while a conventional HVAC system sup-plied air through a perforated diffuser located in theceiling. Air was exhausted from the chamber througha ducted ceiling-level return grill. All measurementswere performed under steady state conditions.

Fig. 4 Plan view of CEC with workstations denoted WS1, WS2,WS3, and WS4. All sample points were 1.1 m above the floor.Points 1 and 5 are 3 cm below the tip of the nose; points 2 and 3are immediately above each shoulder; point 4 is 15 cm in front ofthe nose; points 6 and 7 are at the edge of the desk


Tab

le1

Exp

erim

enta

lco

ndit

ions

and

resu

lts

Test

Clim

ades

kor

Air

Supp

lyM

anik

inC

ond

itio

nsfr

omTA

Cpe

r(T

AC

unit

)C

ond

itio

nsfr

omO

verh

ead

Air

Cha

nge

Pol

luta

ntR

emov

alP

EM

Dir

ecti

onP

osit

ion2

Ven

tila

tion

Syst

empe

r(T

AC

unit

)E

ffec

tive

ness

5E

ffic

ienc

ies

%O

ASu

pply

OA

Supp

lyFl

ow3

Flow

3Te

mp

%O

A4

Supp

lyO

AR

oom

Floo

rB

ody

(L/

s)(L

/s)

(C)

Flow

(L/

s)Fl

ow4

(L/

s)Te

mp

(C)

Sour

ce6

Sour

ce7

129W

Clim

ades

kN

AU

prig

ht10

07

719

.30

350

24.1

1.05

0.95

0.98

130

Clim

ades

kV

erti

cal

Upr

ight

100

77

19.1

036

025

.61.

030.

881.

0413

2C

limad

esk

Ver

t&

Hor

iz1

Upr

ight

100

77

19.6

035

025

.11.

061.

001.

2814

5C

limad

esk

Ver

t&

Hor

iz1

Upr

ight

100

77

18.4

031

025

.31.

151.

141.

4713

1C

limad

esk

Hor

izon

tal1

Upr

ight

100

77

20.2

035

025

.91.

371.

151.

3514

0C

limad

esk

Hor

izon

tal1

Upr

ight

100

77

25.8

033

025

.51.

331.

171.

3114

1C

limad

esk

Ver

tica

lL

ean

100

77

19.4

031

025

.11.

731.

551.

4414

2C

limad

esk

Ver

tica

lL

ean

100

77

19.1

031

025

.31.

831.

491.

5214

3C

limad

esk

Ver

tica

lL

ean

100

33

24.4

1231

425

.61.

751.

381.

5214

4C

limad

esk

Ver

tica

lL

ean

100

33

26.1

1231

427

.51.

941.

351.

5813

5P

EM

Par

alle

lU

prig

ht21

378

19.2

017

022

.9N

A0.

921.

1113

6P

EM

Par

alle

lU

prig

ht19

387

19.1

030

022

.7N

A0.

931.

0813

7P

EM

Par

alle

lU

prig

ht20

388

19.3

031

023

.81.

040.

981.

0913

3P

EM

Tow

ard

Upr

ight

100

1010

19.7

033

024

.81.

631.

251.

5413

9P

EM

Tow

ard

Upr

ight

100

99

19.6

034

024

.31.

421.

201.

4613

4P

EM

Tow

ard

Upr

ight

100

1010

19.5

034

024

.2N

A1.

201.

4313

8P

EM

Tow

ard

Upr

ight

1515

219

.020

347

24.3

1.17

1.00

1.07

W.A

irin

cham

ber

wel

lm

ixed

wit

hos

cilla

ting

des

kfa

ns.

1.U

nder

des

kno

zzle

spo

inte

din

war

d.

2.U

prig

ht:M

anik

inse

ated

upri

ght

abou

t15

cm(6

inch

es)

from

edge

ofd

esk.

Lea

n:N

ose

ofm

anik

inin

vert

ical

jet

from

Clim

ades

k.3.

Ave

rage

flow

oftw

oop

erat

ing

TAC

syst

ems.

4.Sy

stem

set

for

0%ou

tsid

eai

r,ho

wev

erso

me

outs

ide

air

leak

edin

toth

eve

ntila

tion

syst

em.

5.A

irC

hang

eE

ffec

tive

ness

ΩA

geof

Air

inR

etur

n/A

geof

Air

inV

enti

late

dB

reat

hing

Zon

e.6.

Pol

luta

ntso

urce

son

the

floo

rin

each

offo

urw

orks

tati

ons.

7.P

ollu

tant

sour

ces

onth

ebo

die

sof

each

man

ikin

.

276


The furnished chamber contained sources of heatand air motion typical of real offices, including: over-head lights (with a total power of 500 W of whichroughly 100 W directly entered the chamber); and apersonal computer containing a small cooling fan anda monitor in each workstation (90 W each). A seatedheated manikin was located in both WS2 and WS3with both of the TAC systems. Electric resistance heat-ing elements wrapped around the manikin in WS 3 re-leased 75 W (a typical rate of release of sensible heatby an office worker). The other skin-temperature-con-trolled thermal manikin in WS2 released approximate-ly 100 W. For experimental purposes, each workstationcontained a 15 W particle counter.

Experimental ConditionsThe test variables are listed in Table 1. The test con-ditions were based upon previous tests with TAC sys-tems and anticipated use of these systems. Each work-station with the TAC system and manikin was con-figured identically for each test. With the PEMoperating, the air supply outlets were either pointed to-ward the occupant or parallel to the side walls of theworkstation (see Figure 1). With the Climadesk operat-ing the supply air was either supplied horizontallyunder the desk (parallel to the seated manikin’s thighs)(see Figure 2), vertically upward from the front edge ofthe desk (Figure 3), or an approximately equal air flowin both the horizontal and vertical directions. The totalamount of outside air supplied to the room, nominally10 L/s-occupant, was based upon ASHRAE Standard 62(ASHRAE, 1989). During all but three tests, all of theoutside air was supplied to the room through either thePEM or Climadesk. The conventional overhead venti-lation system recirculated air between 61–71 L/s. Thetotal rate of air supply from the overhead system plusthe TAC system was typical for rooms of this size (4 L/s/m2 or 0.8 cfm per ft2). In three tests with the PEMs op-erating, some room air was mixed with outside air thatwas supplied through the desktop air outlets. For mosttests the air was supplied through the TAC systems atnominally 19æC (except for three tests in which it was25æC). The room temperature was controlled at ∂25æC.Both the supply air temperature and the room tempera-ture are higher than typically found in a conventionallyventilated office space because the TAC systems providelocal cooling. In all but four of the tests, the manikinswere seated upright with their faces located about 15 cmback from the edges of the desks. Four tests were runwith the manikins leaning slightly forward with theirfaces in the vertical air supply jets emerging from theClimadesks.

277

Measurement MethodsAir Change EffectivenessACE was measured using a tracer-gas step-up pro-cedure. After steady state test conditions were estab-lished, sulfur hexafluoride (SF6) tracer gas was injectedat a constant rate (within 1%) into the supply or out-side air duct. Mixing fans inside the HVAC systemductwork ensured thorough mixing of the tracer in thesupply airstream.

Using three gas chromatographs with electron cap-ture detectors (GC-ECD), tracer-gas concentrationswere measured as a function of time during the periodof concentration increase. Concentrations were meas-ured approximately every four minutes at the follow-ing locations: the outside air duct; the supply air duct;the return/exhaust duct; and seven locations withinthe chamber (at the mouth/nose in each workstationand in WS2, 15 cm in front of the nose and immedi-ately above each shoulder). The GC-ECD units werecalibrated after each test with thirteen calibrationgases.

Ages of air (t) were determined from the SF6 tracerdata via the equation

t Ω1

C (tend) etend

0

[C(tend) ª C(t)]dt (1)

where C(t) is the tracer-gas concentration at the pointin question, C(tend) is the steady-state concentration atthe end of the step-up, and t is the time elapsed sincethe start of tracer-gas injection. The ACE was deter-mined from:

ACE Ωtreturn

tbl(2)

where treturn is the age of the return/exhaust air andtbl is the average age of air at the two occupied breath-ing level measurement locations.

Pollutant Removal EfficiencyFor the measurements of PRE, three different perflu-orocarbon tracer-gases were used to simulate sourcesof indoor-generated pollutants. To simulate emissionsfrom the floor covering, four passive emitters of meta-perfluorodimethylcyclohexane (C8F16) tracer gas wereplaced on the floor, one in each workstation. To simu-late emissions from occupants, two emitters of per-fluorodimethylcyclobutane (C6F12) tracer gas were at-tached to the manikin in WS2 and two emitters of per-fluoromethylcyclohexane (C7F14) tracer gas wereattached to the manikin in WS3. The emitters on the


manikins were located near the armpits to simulate theemissions of body odors by occupants. All emitterswere placed in the room the day before the test so thatconcentrations would reach steady state before thestart of the test. To measure the average concentrationsof the perfluorocarbon tracer gases, air samples weredrawn from the occupants’ breathing zone and the re-turn/exhaust duct at a constant rate with peristalticpumps and stored in 2-liter sample bags. The sampleswere collected for approximately the same timeperiods as the tracer step-ups. The bags were subse-quently analyzed with a gas chromatograph (GC withelectron capture detector). The tracer emitters and ana-lytical system have been described previously (Fisk etal., 1993; Faulkner et al., 1999).

The PRE for the ‘‘Floor’’ and ‘‘Body’’ pollutants werecalculated from the equations:

PREFloor ΩCReturn

Floor

12 SCBL2

Floor π CBL3FloorD (3)

PREBody Ω 12 SCReturn

Body2

CBL3Body2

πCReturn

Body3

CBL2Body3

D (4)

where the superscript denotes the measurement loca-tion (Return duct, Breathing Level in either WS2 orWS3) and the subscript denotes the location of the pol-lutant source (floor, manikin in WS2 or manikin inWS3). The values of PREBody indicate the efficiency ofthe ventilation process in controlling exposures to pol-lutants emitted by the occupants in the adjoiningworkstation.

Percentage of Outside Air and HVAC Air FlowRates and TemperaturesTo determine the percentage of outside air suppliedeither to the TAC system or to the main HVAC system,one of two methods was used. The first method wasto measure the tracer gas concentrations in the return/exhaust airstream and the supply airstream down-stream of the junction of the outside-air and supply-airducts. The percentage of outside air was determinedfrom the equation:

%OAΩ(1ª Cm/Cr)100% (5)

where Cm is the concentration of tracer gas in the mix-ture of outside and recirculated air and Cr is the con-centration in the return/exhaust air.

To obtain nearly instantaneous values of the percen-tage of outside air, a second method was to use flowr-

278

ate measurements of the outside airstream and thesupply airstream made using Pitot tubes and pressuretransducers. The percentage of outside air was deter-mined from the equation:

%OAΩ(QOA/Qs)100% (6)

where QOA is the flowrate of the outside airstream andQs is the flowrate of the supply airstream.

The flow rates of air in the HVAC system were meas-ured using Pitot tubes and Venturi flow meters withdifferential pressure transducers. The airstream tem-peratures were measured with thermistors. The meas-urement system is described in detail elsewhere (Arenset al., 1991; Fisk et al., 1991).

ResultsMeasurement precisionDuring Test 129W (see Table 1), the chamber air wasmixed vigorously with fans which ideally should pro-duce the same average concentration for each tracer atevery point in the chamber. Consequently, all of theages of air should be identical and all of the ratios ofconcentrations should equal unity. However, due tomeasurement imprecision and errors (and possible im-perfect mixing despite the operation of mixing fans)not all of the measured ages of air and tracer gas con-centrations are equal. Based on these results and theresults of 11 previous ‘‘well-mixed’’ tests within thechamber, the estimated 95% confidence interval forACE during well-mixed tests is 1.02∫0.06.

Since PRE has not been measured during previous‘‘well-mixed’’ tests, we estimated the precision basedupon measurements of all locations within thechamber during Test 129W. Using propagation of erroranalysis, we estimate the 95% confidence interval forPRE to be 0.95∫0.14 and 0.98∫0.14 for simulated pol-lutants from the floor and body, respectively.

Air Change Effectiveness and Pollutant RemovalEfficienciesThe ACEs and PREs measured during many of the ex-periments were high (1.3 to 1.9 for ACE; 1.2 to 1.6 forPREFloor and PREBody). Based upon the above analysisfor the well-mixed tests, these results are statisticallysignificant at the 95% confidence level. These resultsindicate that the TAC systems, relative to indoor airthat is thoroughly mixed by other ventilation systems,can substantially increase the effective rate of venti-lation at the breathing zones and reduce pollutant ex-posures.

The highest values of ACE and PRE were measured,


with either of the TAC systems supplying 100% OA atapproximately 7–10 L/s per occupant, with the air sup-ply directed toward the manikin’s face. These valuesoccurred with the supply air temperature either 6æCbelow room temperature or at room temperature. Withthe PEM, the nozzles were pointed toward themanikin’s face and with the Climadesk, the manikinleaned into the vertical air jet exiting the front edge ofthe desk. In experiments with the Climadesk (Tests141–144), the high values of ACE and PRE were verylocalized at the mouth and nose, as measurements ofACE and PRE 15 cm in front of the nose and mouthwere close to unity. High values of ACE and PRE werenot measured with the manikin seated upright and airsupplied through the vertical outlet of the Climadesk(i.e., with the face not located directly in the verticalsupply air jet) see Test 130.

The Climadesk system also produced high ACE andPRE values when the air supply was entirely horizon-tal (Tests 131 and 140), directed toward the manikin’storso from beneath the desk. Tests with a smoke tubesuggest that some of the outside air supplied horizon-tally by the Climadesk was entrained in the thermalplume flowing upward along the body and carriedinto the region of the nose and mouth. Under theseoperating conditions, high values of ACE and PREwere maintained 15 cm in front of the nose and mouth.

With approximately equal amounts of air suppliedvertically and horizontally from the Climadesk, theACE and PRE values were not consistent. Results fromTest 132 indicated little or no improvement in ACE orPRE. Whereas results from Test 145 indicate enhancedACE and PRE as we had expected. Similar to configur-ations discussed above, the improved ACE and PREvalues with this configuration may be highly depend-ent upon the manikin position relative to the Clima-desk outlets and the edge of the desk.

In two tests with the Climadesk, high values of ACEand PRE were maintained when approximately half ofthe total outside air supply was provided by the con-ventional overhead ventilation system. We anticipateda decrease in performance under these operating con-ditions. However, these results were obtained in testswith the manikin’s head located directly in the verticalair supply jet. As discussed above, under these con-ditions the ACE and PRE will vary considerably withsmall changes in the position of the head and an opti-mal location of the manikin’s head may have counter-acted the expected performance decrease.

Figure 5 shows ACE vs. PRE and a linear trendlinefor both the Floor and Body tracers. The values of ACEand PREFloor were reasonably well correlated (RΩ0.92).The correlation of ACE with PREBody was weaker (RΩ

279

Fig. 5 Pollutant removal efficiency versus air change effectivenessfor body and floor sources

0.83). This strong correlation between ACE and PREFlo-

or has been measured previously (Fisk et al., 1997).Since PREs, relative to ACEs, can be measured for awider range of conditions, such as naturally ventilatedbuildings and buildings with non-steady air flow rates(Fisk et al., 1997; Faulkner et al., 1999), PRE measure-ments may turn out to be a practical substitute formore difficult ACE measurements.

Discussion and ConclusionsIn our previous related research, the TAC systems sup-plied air at much higher flow rates. With air suppliedfrom the PEM at flow rates of 19 to 94 L/s, PREs andACEs were significantly above unity only if this airwas 100% outside air directed toward the occupantsface (Faulkner et al., 1993) – a condition that is notlikely to be comfortable (Bauman et al., 1993). Directingthe air away from the face made conditions morecomfortable but resulted in ACEs and PREs close tounity. These results suggest that high rates of air sup-ply from TAC systems may vigorously mix the airwithin the workstation, making it difficult to preferen-tially ventilate the breathing zone. With the lower TACsupply flow rates employed in this current set of ex-periments and the outlets pointed at the occupant,thermal comfort conditions were found generally ac-ceptable (Tsuzuki et al., 1999). One exception may bethe maximum velocity measured (1.34 m/s) at the faceof the manikin while its head leaned into the verticaljet of the Climadesk. Unwanted air movement of 1.34m/s may be unacceptable by some thermal comfort


standards. However, with TAC systems, users canchoose what is acceptable for their own comfort.

Our prior research on TAC have also included ex-periments with air supplied from outlets mountedwithin the floor of the workstation (Fisk et al., 1991;Faulkner et al., 1995). ACEs were significantly aboveunity only when the air was directed in a manner thatyielded an upward vertical displacement flow. How-ever, commercially available displacement ventilationdiffusers, which do not allow for individual control,may provide a superior option for obtaining a dis-placement flow pattern.

For the Climadesk unit operating with a vertical airsupply jet, a superior ventilation performance wasachieved with the occupant’s head located preciselywithin the vertical jet of air. Since real occupants willmaintain a variety of postures and locations, the ACEsand PREs achieved in practice may be much closer tounity. Our data and understanding of system perform-ance suggest that supplying air horizontally is themore robust method of assuring high ventilation ef-ficiencies. However, this expectation, based on resultsof only two experiments, needs to be confirmedthrough additional experiments, including some withreal occupants.

Ideally, TAC systems should maintain high values ofACE and PRE while enabling occupants to adjust theirlocal thermal environment and remain comfortable.Concurrent to our measurements, Tsuzuki et al. (1999)made measurements using the thermal manikin inWS2. With the Climadesk or PEM supplying 7–10 L/s,they measured a maximum effective whole body cool-ing equivalent to approximately 1æC. Thus, these twoTAC systems can measurably reduce the effective tem-perature and preferentially ventilate the breathingzone. Use of the radiant heaters integral to each TACsystem increases the range of thermal control. How-ever, if occupants reduce the rate of air supply orchange the air supply direction to increase the effectivetemperature, ACEs and PREs are likely to differ fromthe values measured in these experiments. Reducingthe air supply rate could conceivably increase or de-crease ACEs and PREs. Directing the supply air awayfrom the occupant will most likely reduce ACE andPRE. Changing the air supply temperature without re-circulation of supply air (not an option with the currentdesigns) might increase or decrease ACE or have noeffect. Further research is also necessary to identify thebest methods to maintain high values of ACE and PREwhile enabling occupants to significantly change theirthermal environment.

Overall, the results of this current set of experimentsare very encouraging. The high values of ACE and PRE

280

measured in this study suggest that these two desk-top TAC systems can be used to improve ventilationand pollutant removal at the breathing zone if normalrates of outside air supply are maintained. Alternately,energy savings could be realized while maintaining atypical level of IAQ at the breathing zone by allowingrates of outside air supply to be reduced. For example,the Design Ventilation Rate as recommended in ASH-RAE Standard 62, may be corrected by using the air-change effectiveness. The corrected design ventilationrate is the ratio of the Design Ventilation Rate to theair-change effectiveness (ANSI/ASHRAE 129–1997).Thus, values of ACE from 1.3 to 1.9 would translate toa corrected design ventilation rate 23% to 47% less thanthe ASHRAE Standard 62 Design Ventilation Rate, as-suming the airflow pattern is not significantly changedat the lower flow rates.

AcknowledgementsThis work was supported by Johnson Controls, Inc.and the Assistant Secretary for Energy Efficiency andRenewable Energy, Office of Building Technologies,Building Systems and Materials Division of the U.S.Department of Energy (DOE) under contract No. DE-AC03–76SF00098. We also thank Jim Yi of JohnsonControls, Inc., Fred Bauman and Ed Arens at the Uni-versity of California, Berkeley. In addition, thanks toLeon Alevantis for reviewing a draft of this paper.

ReferencesANSI/ASHRAE (1992) ASHRAE Standard 55–1992, Thermal

Environmental Conditions for Human Occupancy, Atlanta,GA, American Society of Heating, Refrigerating, and AirConditioning Engineers.

ANSI/ASHRAE (1997) ASHRAE Standard 129–1997, Measur-ing Air-Change Effectiveness, Atlanta, GA, American Societyof Heating, Refrigerating, and Air Conditioning Engineers.

ASHRAE (1989) ASHRAE Standard 62–1989, Ventilation for ac-ceptable indoor air quality, Atlanta, GA, American Society ofHeating, Refrigerating, and Air Conditioning Engineers.

Arens, E.A., Bauman, F.S., Johnson, L.P. and Zhang, H. (1991)‘‘Testing of Localized Thermal Distribution Systems in aNew Controlled Environment Chamber’’, Indoor Air, 1,263–281.

Bauman, F.S. and Arens, E.A. (1988) ‘‘The Development of aControlled Environment Chamber for Physical and Subjec-tive Assessment of Human Comfort in Office Buildings’’.In: Kroner, W. (ed) A New Frontier: Environments for Inno-vation, Proceedings of the International Symposium on Ad-vanced Comfort Systems for the Work Environment, Troy, NY,Center for Architectural Research.

Bauman, F.S., Arens, E.A., Benton, C.C. and Zhang, H. (1993)‘‘Localized Comfort Control with a Desk-top Task Con-ditioning System: Laboratory and Field Measurements’’.ASHRAE Transactions, 99, 733–749.

Bauman, F.S. and Arens, E.A. (1996) Task/Ambient ConditioningSystems: Engineering and Application Guidelines, Berkeley,


CA, Center for Environmental Design Research, Universityof California, Berkeley (Publication number CEDR-13–96).

Bauman, F.S., Carter, T.G., Baughman, A.V. and Arens, E.A.(1998) ‘‘Field Study of the Impact of a Desktop Task/Ambi-ent Conditioning System in Office Buildings’’, ASHRAETransactions, 104, 125–142.

Borgers, T. and Bauman F.S. (1994) Section 1: Whole BuildingEnergy Simulations, Localized Thermal Distribution for OfficeBuildings: Final Report – Phase III, Berkeley, CA, CEDR Cent-er for Environmental Design Research, University of Cali-fornia, Berkeley, Report number CEDR-02–94.

Faulkner, D., Fisk, W.J. and Sullivan, D.P. (1993) ‘‘Indoor Air-flow and Pollutant Removal in a Room With Desktop Ven-tilation’’. ASHRAE Transactions, 99, 750–758.

Faulkner, D., Fisk, W.J. and Sullivan, D.P. (1995) ‘‘Indoor Air-flow and Pollutant Removal in a Room With Floor-basedTask Ventilation: Results of Additional Experiments’’,Building and Environment, 30, 323–332.

Faulkner, D., Fisk, W.J., Sullivan, D.P. and Thomas, J.M., Jr.(1998) ‘‘Characterizing Building Ventilation with the Pol-lutant Concentration Index: Results from Field Studies’’. In:Proceedings of IAQ and Energy ‘98, Atlanta, GA, AmericanSociety of Heating, Refrigerating and Air-ConditioningEngineers.

Fisk, W.J., Faulkner, D., Pih D., McNeel, P.J., Bauman, F.S. andArens, E.A. (1991) ‘‘Indoor Air Flow and Pollutant Re-moval in a Room with Task Ventilation’’, Indoor Air, 1, 247–262.

Fisk, W.J., Faulkner, D. and Hodgson, A.T. (1993) ‘‘The Pol-lutant Control Index: A New Method of Characterizing

281

Ventilation in Commercial Buildings’’. In: Seppanen O, Il-marinen R, Jaakkola JJK, Kukkonen E, Sateri J and Vuorel-ma H (eds) Proceedings of Indoor Air ‘93, Helsinki, The 6thInternational Conference on Indoor Air Quality and Cli-mate, Vol. 5, pp. 9–14.

Fisk, W.J., Faulkner, D., Sullivan, D. and Bauman, F. (1997)‘‘Air Change Effectiveness and Pollutant Removal Ef-ficiency during Adverse Mixing Conditions’’, Indoor Air, 7,55–63.

Kroner, W.M. and Stark-Martin, J.A. (1992) ‘‘EnvironmentallyResponsive Workstations and Office Worker Productivity’’.In: Levin H. (ed) Proceedings of Indoor Environment and Pro-ductivity, Baltimore, MD, American Society of Heating, Re-frigerating and Air-Conditioning Engineers.

Sandberg, M. and Sjoberg, M. (1983) ‘‘The Use of Momentsfor Assessing Air Quality in Ventilated Rooms’’, Buildingand Environment, 18, 181–197.

Schiller, G.E., Arens, E.A., Bauman, F.S. and Benton, C. (1988)‘‘A Field Study of Thermal Environments and Comfort inOffice Buildings’’, ASHRAE Transactions, 94, 280–308.

Seem, J.E. and Braun, J.E., (1992) ‘‘The Impact of PersonalEnvironmental Control on Building Energy Use’’, ASHRAETransactions, 98, 903–909.

Tsuzuki, K., Arens, E.A., Bauman F.S. and Wyon, D.P. (1999)‘‘Individual Thermal Comfort Control With Desk-Mountedand Floor-Mounted Task/Ambient Conditioning Systems’’,Center for Environmental Design Research, University ofCalifornia, Berkeley, CA, Indoor Air 99, Edinburgh, Scot-land, in press.

ventilation efficiencies of desk-mounted task/ambient conditioning systems

Documents