ANALYSIS OF ANNUAL ENERGY SAVINGS DUE TO RADIANT BARRIERS

Kenneth E, WilkesOak Ridge National Laboratory

Radiant barriers are receiving increasing attention as an energy conservation measurefor residential buildings, especially for warmer climatesll They are being activelypromoted for use in residential attics, sometimes with exaggerated claims aboutsavings in utility bills that will result from their installationo

In order to provide consumers with factual information that would assist them indeciding upon an investment in a radiant barrier, the Department of Energy, alongwith an industry advisory panel, has developed a Radiant Barrier Fact Sheet. A majorpart of this fact sheet is estimates of energy savings that might be expected fromradiant barriers in various climates.

This paper presents the details of the methodology underlying the energy savingsestimates, and gives a summary of values listed in the Fact Sheet. The energy savingsestimates were obtained from calculations using a detailed attic thermal modelcoupled with DOE-2~lCll A life cycle cost analysis was performed to estimate thepresent value savings on utility fuel costs.. The results show that the fuel cost savingsvary significantly with the level of conventional insulation already in the attic andfrom one climate to another..

INODUCTION

Insulation systems based on reflective surfaces havebeen in existence for many years.. However, interesthas only recently been focussed on the applicationof single sheets of reflective materials in the atticsof residential buildings, an application that is knownas a "radiant barrier" (RB).. RBs may be installed inattics in a variety of configurations.. They may belaid directly on top of existing conventional atticinsulation (the horizontal configuration), attachedto the bottoms of the rafters, draped over the topsof the rafters, or attached directly to the undersideof the roof decking.

Experiments by a number of groups have demon­strated that radiant barriers can be effective inreducing heat flows through the ceiling, especiallyunder summer cooling conditions (see Wilkes andYarbrough 1988)" However, their benefits aresometimes exaggerated in marketing claims. This,along with the wide range in selling prices, has

prompted the U.S" Department of Energy, with theassistance of an industry advisory panel, to developan Interim Radiant Barrier Fact Sheet to serve as asource of unbiased information for the consumer"The modifier, "Interimt

', was included becauseresearch on radiant barriers is not complete andthere will be a need for an updated version.

A major section of the Fact Sheet gives estimates ofcooling and heating energy savings. The methodchosen by the advisory panel for presenting energysavings estimates is the present value of life cycleenergy savings in terms of dollars per square foot ofceiling area. Present-value savings are given for anumber of climates, existing insulation levels, andradiant barrier configurations0 The consumer canthen compare this present value savings with thecost of installing a radiant barrier, which might beobtained as a quote from the installer.. This paperpresents the methodology that was used to develop

Building Equipment and Appliances 1.235

the estimates of present-value savings. Work on theFact Sheet is not yet complete and it does not haveunanimous approval of the advisory panet There­fore, this paper gives the present status of theenergy savings estimates; changes may be madebefore the Fact Sheet is actually issued.

ATIIC/RADIANT BARRIERTHERMAL MODEL

Energy savings estimates were obtained using amodel for the transient performance of atticso Thecurrent model is based on an earlier model that wasdeveloped by Peavy at the National Bureau ofStandards (Peavy 1979), and that was later extendedby Wilkes (1983, 1990a,b). The model includesradiation interchanges among all the surfaces thatface the attic space, using the Stefan-BoltzmannCr4) law, with the idealizations that the surfaces areflat, gray, isothermal, diffusely emitting andreflecting, and have a uniform radiant flux.. Each ofthe surfaces may have a different emissivity (in thispaper, the term "emissivity" is used synonymouslywith the term "emittance"). The model also includesconvection to the ventilation air stream, the flowrate of which is calculated from a combination ofstack and wind pressure effects., Heat transferprocesses at the exterior surfaces include absorptionof solar radiation, convection to the outdoor air,and radiation exchanges with the surroundings.

Predictions of the model have been compared withceiling heat flows measured in a number of labor­atory and field tests (Wilkes 1988, 1989, 1990a,b).An example is given in Figure 1, which shows dataobtained from a field experiment using full-sizehouses near Knoxville, Thnnessee (Levins andKarnitz 1986).. In general, it has been concludedthe model is capable of predicting cumulative ceilingheat flows to within about 10 percent of the meas­ured values.. This level of ac~uracy was judgedacceptable to qualify the model for use in generatingestimates of the energy savings due to radiantbarriers. there is a need for furtherverification of the model, especially in cold climates.

For annual energy analyses, the attic model wasdriven with hourly weather tapes with the indoortemperature being maintained at a constant valuethat was midway between the heating and cooling

1~236 Wilkes

thermostat setpoints. Hourly ceiling heat fluxescomputed by the model were brought into theDOE-2.1C model using the FUNCTION commandand were substituted for the ceiling heat fluxes thatDOE-2 would normally calculate<> Adjustments forthermostat settings in DOE-2 being different fromthe constant indoor air temperature assumed in theattic model, and for the indoor temperature floatingbetween the heating and cooling setpoints weremade in the SYSTEMS section of DOE-2.. Annualenergy savings due to radiant barriers were obtainedfrom model runs on a typical house, the runs beingidentical except for changes in emissivities of thesurfaces facing the attic space.

ENERGY SAVINGS CALCULATIONMETHOD AND RESULTS

Prototypical House

The energy savings estimates were based on a proto­typical ranch style house, which has been used forseveral studies (for example, Labs et al.. 1988). Thehouse is 55 feet long and 28 feet wide, with a floorarea of 1540 square feet The windows are double­glazed and have an area of 184.8 square feet(12 percent of the floor area), and a door on thesouth wall occupies 19.5 square feet The walls arewood frame with R-l1 insulation and have a solarabsorptance of 0.7. The floor is built over a threefoot crawl space and has R-19 insulation..

The ceiling is constructed with 2X4 wood joists24 inches on center, with 1/2 inch gypsum wallboard..The ceiling is insulated with either R-l1, R-19,R-30, or R-38 fiberglass batt insulation, and thejoists are assumed to be covered with insulation forlevels greater than R-11.. The roof has a pitch of 5in 12, a solar absorptance of 0.9, is unshaded, andhas the ridge oriented in the east-west direction.(Exploratory calculations show that chan.ging theridge orientation to the north-south direction affectsthe load reductions due a RB by only about 5 per­cent) Venting of the attic is through soffit and ridgevents, with a total net free area equal to 1/150 ofthe ceiling area. One-third of the vent area isassumed to be in the ridge vent, and the other two­thirds are assumed to be in the soffit vents,corresponding to typical construction practices

ON TOP OF RmD19 INSU TI NAUGUST 13-20, 1985

o 12 24 36 48 60 72 84 96 108 120 132 144 156

TIME (h)

Ie Comparison ofModel Predictions with Ceiling Heat Fluxes Measured in Full-size House. Positive heatflows are from attic into house.

(personal communication from B. Howard, 1989)..Emissivities of nonreflective surfaces were taken tobe 0..9, and the radiant barrier emissivity was takento be Oe05, except as noted below.. (Emissivities ofvarious RB materials range from about 0.03 toabout 0..08.. An average value of 0,,05 was used here,,)

The thermostat settings were taken to be 78°P inthe summer and 70°F in the winter. During thesummer, vvindow venting was assumed when:(1) opening windows provides enough cooling tokeep the zone temperature between 68 and 78°F;(2) the outside air enthalpy is lower than the insideair enthalpy, and (3) the air-conditioning loadduring the hour can be met totally through naturalventilation at 10 air changes per hour.. Since occu­pants typically would not adjust windows after goingto bed, a time of day schedule was added to keepwindows closed between 11 p.m. and 7 aem.. Duringthe window venting was assumed when theindoor temperatures would rise above 78°R TheseOD~~ratlnf! characteristics are essentially the same as

those used by Labs et at (1988) and Huang et a1.(1987)e The effect of keeping windows closed at alltimes has not been explored"

Infiltration was calculated using the Sherman­Grimsfud model for average residential constructionin a typical suburban area with low buildings andtrees within 30 feet (Sherman and Grimsrud 1980)..Internal loads were taken to be 55,100 Btu per day,which corresponds to 3..2 people, 1 kWh per squarefoot lighting, and average appliance levels. Hourlyinternal load profiles were taken from a scheduledeveloped by the California Energy Commission(CEC 1984)"

If the radiant barrier performance parameter waschosen to be the percentage reduction in energyusage, then many of the above assumptions wouldbe critical.. By concentrating on differences in energyusage due to radiant barriers, and by later normaliz­ing the results to a square footage basis, theseassumptions should not be as critical, since they

Building EqUipment and Appliances 1$237

where w == 1.98 [0.0219 + 702005 X 10-5t2]112 (2)

to five years) .. A statistical analysis of these datayielded the following 95 percent confidenceintervals:

NRH = 19202 [0.596 - 1/(1 + O.61/e)] (3)

NRC = 162.3 [O~8318 - 1/(1 + O~182/e)] (4)

where NRH and NRC are the normalized heatingand cooling load reductions (as percentages of theload reductions for a clean horizontal RB) and e isthe radiant barrier emissivitye

Data taken by the Thnnessee Valley Authority withtest cells having horizontal radiant barriers thatwere artificially dusted with "Arizona test dust" donot show such a rapid decline in performance asgiven by NRC (Hall 1988)0 For these data, thenormalized load reductions appear to follow anapproximately linear variation with emissivity.

(1)e = Oa8 .. 0077 exp(-O.11127t ± w)

where e is the emissivity and t is the time in yearsoThe mean equation (i.e., w = 0) has been con­strained to pass through 0.03 at time zero, and tohave a long-time asymptote of 0.8. These interceptand asymptote values were determined from labora­tory studies of artificially dusted samples of RBsthat had emissivities of 0.03 before dusting and 0.8after a large amount of dusting. After this analysiswas performed, an additional house was examined,in which a radiant barrier had been installed forabout 23 yearsa The emissivity was found to be 0.75to 0..80, in good agreement with the projected curveo

The models have been run with various assumedlevels of emissivity for the horizontal RB. Loadreductions obtained from these calculations werenormalized with respect to the load reductions froma clean horizontal radiant barrier, with the resultsshown in Thble 2" The model predicts that the per­formance drops off significantly with increasingemissivity, but the normalized reductions do notvary significantly with climate. Empirical equationsfit to these results are

should nearly subtract out. For example, considertwo houses of equal floor area and similar construe..tion, except that one is a single story house and theother is a two story house.. Heating and coolingloads due to the roof will be a smaller percentage ofthe whole house load for the two story house.Consequently, the percentage reduction in energyusage due to a RB will be smaller for the two storyhouse. However, the difference in energy usage persquare foot of ceiling should be similar for the twohouses.

Load Reductions for Horizontal RB

Calculations were performed using TMY weathertapes for 26 cities and a California Climate Zonetape for Riverside, California. Figure 2 shows anexample of the hour-by-hour ceiling heat fluxespredicted by the attic model (because of plottingsoftware limitations, somewhat less than 8760 hoursare shown).. The sign convention used is that heatflows from the attic into the house are taken to bepositive.. This figure clearly shows the significantreductions in peak positive heat fllL"'t{es due to theradiant barrier.

The results of the DOE-2.. 1C model were annualheating and cooling loads on the house. By usingdifferences between similar runs with and without aradiant barrier, the load reduction due to a radiantbarrier was obtainecL Heating and cooling loadreductions for a clean radiant barrier applieddirectly on top of existing attic insulation are givenin Thble 10 These results show that both the heatingand cooling load reductions are greatest with lowerlevels of attic insulation, as would be expectede Alsoas expected, the cooling load reductions are greaterfor the warmer climates, and the heating load reduc-tions are for the colder climates~

For radiant barriers applied directly on top of atticinsulation, account was taken of the seeminglyinevitable fact that dust will accumulate, resulting inan increasing emissivity and hence a decreasingperformance.. D$ w.. Yarbrough (private communica­tion, 1990) has collected the available data onemissivities of radiant barriers that have beeninstalled in attics for various amounts of time (up

1s238 Wilkes

PREDICTED HOURLY CEILING HEAT FLOWS

MIAMI, R..19 INSULATION

NORS

~ CLEAN RS TOPINSULATION

2.5

2

1.5

~;::

.c"""-:J 0.5@,X:J....J au.

~I -0.5

01

..1.5

..20 1000 2000 3000 4000

TIME (h)

5000 6000 7000 8000

2~ Predicted Hourly CeilingHeat Fluxes with and Without Clean RadiantBarrier on Top ofR-19 Insulationin Miami. Positive heat flows are from attic into house.

The approach used for the Fact Sheet is to presenta range of savings for horizontal radiant barriers..The lower end of the range is obtained using theupper level of the confidence interval for emissivityversus time coupled with the model predictions ofnormalized performance versus emissivity. Theupper end of the range is obtained using the lowerlevel of the confidence interval for emissivity versustime coupled with a linear variation of normalizedperformance versus emissivity"

Load Reductions for Roof RBs

Load reductions for radiant barriers applied nearthe roof were estimated by running the model withvarious levels of effective or average emissivities forthe underside of the root The inside surfaces of thegables were also assigned low emissivities (0.05)0Predicted load reductions, normalized to the values

for a clean horizontal radiant barrier are given inThble 3.. In general, the normalized values do notvary significantly with climate or insulation level,but do vary considerably with roof emissivity andalso vary somewhat between the heating and coolingseasons"

For radiant barriers attached to the bottoms of theroof rafters, an effective roof emissivity of 0.08 wasused. This value represents a simple area averagingof emissivities of the radiant barrier and a smallamount of exposed wood roof deck. For radiant bar­riers draped over the tops of the rafters or attacheddirectly to the roof deck, an effective roof emissivityof 0.11 was used" This value accounts for additionalexposed rafter surfaces.. Interpolating in Thble 3 foran emissivity of 0.08 gives normalized cooling andheating load reductions of 78 percent and 88 per...cent (These load reductions are about 18 percent

Building Equipment and Appliances 1.239

~~

Tabl

e1.

.A

nnua

lLoa

dR

educ

tions

Due

toC

lean

Hor

izon

talR

adia

ntB

arri

eri\

)~ 0

Hea

tin

gL

oad

Red

uct

ion

,B

tu/f

t2C

oo

lin

gL

oad

Red

uct

ion

,B

tu/f

t2

~ ~R

-11*

R-1

9R

-30

R-3

8R

-11

R-1

9R

-30

R-3

8(I

)

Alb

any

,NY

929

400

193

140

876

409

259

211

Alb

uq

uer

qu

e,NM

931

476

29

923

815

9885

152

242

6A

tlan

ta,

GA60

528

216

313

716

7383

251

640

5B

ism

arck

,ND

1192

513

293

206

706

388

245

191

Ch

icag

o,

IL84

2'37

721

014

496

047

528

422

9D

env

er,

CO98

947

327

723

610

2055

035

729

4E

lT

aro

,CA

792

378

242

197

1232

636

405

351

Ho

ust

on

,TX

387

182

108

8021

6211

2067

252

1K

no

xv

ille

,TN

725

337

206

164

1597

823

517

411

Las

Veg

as,

NV

774

438

277

227

2535

1210

703

539

Los

An

gel

es,

CA73

839

022

718

842

925

616

814

8M

emph

is,

TN63

030

418

016

418

3290

755

544

0M

iam

i,FL

9947

2826

3090

1631

938

727

Min

nea

po

lis,

MN

1062

447

223

154

769

418

257

204

Orl

and

o,

FL27

513

077

6225

7512

9983

266

2P

ho

enix

,A

Z60

632

119

116

233

0815

9594

273

8P

ort

lan

d,

ME

1112

490

253

194

297

120

8262

Po

rtla

nd

,O

R93

742

723

818

655

129

917

814

7R

ale

igh

,NC

741

342

219

162

1440

738

460

359

Riv

ers

ide,

CA8

92

422

248

189

1999

931

556

448

Sac

ram

ento

,CA

821

397

236

192

1592

849

542

445

Salt

Lak

eC

ity

,D

r90

641

522

318

712

8665

140

933

2S

t.L

ou

is,

MO

738

324

169

136

1466

757

479

369

Seatt

le,

lilA

904

364

197

133

223

119

8065

To

pek

a,K

S86

837

921

917

615

2379

051

239

7W

aco,

TX47

722

513

811

923

7111

7571

355

2W

ash

ing

ton

,D

.C.

912

386

212

182

1221

622

386

301

*R-V

a1ue

of

att

icin

sula

tio

n.

Table 2. Effect ofEmissivity on Load Reductions Due to Horizontal Radiant Barriers with R-l1 Insulation

Percent of Clean Horizontal RB Load Reduction

Emissivity Miami , Cooling Minneapolis, Cooling Minneapolis, Heating

0.05 100.0 100.0 100.0

0.10 73.4 74.8 87.8

0.20 48.0 50.8 66.7

0.30 33.5 35.7 51.6

0.40 23.4 24.9 38.5

0.50 16.0 17.4 28.0

Table 3. Load Reductions for Truss Radiant Barriers with Low Emissivity (e = 0.05) GablesInsulation Percent of Clean HRB Load Reduction

City Season Level e = 0.05* e = 0.10 e = 0.15

Miami Cooling R-l1 92.4 68.2 53.8R-19 94.1 70.7 56.8R-30 94.5 73.0 58.2R-38 94.6 71.8 57.9

Minneapolis Cooling R-l1 93.2 70.0 56.2R-19 95.0 74.5 60.7

Miami Heating R-l1 95.4 84.9 74.3R-19 95.9 83.6 69.9R-30 95.3 88.4 76.7R-38 97.5 82.5 72.5

Minneapolis Heating R-ll 96.3 84.2 73.2R-19 95.6 80.4 71.2

*Effective emissivity of underside of roof.

and 7 percent less than would be predicted for aroof emissivity of 0..05,,) For an emissivity of 0,,11,similar normalized factors are 68 percent and 82IV'lVJI."""'.It~I.;'. Cooling and heating load reductions forthe roof RBs were obtained by multiplying the loadreductions for a clean horizontal RB by thesefactors.

An additional factor was applied to the loadreductions for roof RBs to account for the influenceof a roof RB on decreasing attic air temperaturesand hence in decreasing heat gains to air­conditioning ducts that are run in attic spaces. Sincethe ORNL attic model does not include ducts, the

Building Equipment and Appliances 1.241

results of modeling work performed by the DavisEnergy Group were used to obtain adjustmentfactors (Bourne and Hoeschele 1988). Their resultssuggest that the cooling load reductions for atticswith air-conditioning ducts should be greater thanthose without air-conditioning duets by 26%, 37%,48%, and 57% for attic insulation levels ofR-!1, 19,30, and 38, respectively. Heat gains to ducts arerelatively independent of attic insulation level,whereas the heat flow through the ceiling decreaseswith insulation level. Hence heat gains to the ductsbecome relatively more important as the atticinsulation level is increased. Similar adjustmentswere not made for heating load reductions, since theeffects are expected to be smaller.

For roof applications, two values are used for eachcase: one that corresponds to the values calculatedwith the ORNL model and another that is adjustedupwards by the factors derived from the DavisEnergy Group work. The first value applies to atticswith no Aje ducts, while the second value applies toattics with Ale ducts. There is a need for experi­mental verification of the proposed duet adjustmentfactors ..

Savings Estimates

Load reductions were converted to energy savings bydividing by equipment efficiencies or coefficients ofperformance.. For the tables of present value savingsin the Fact Sheet, average equipment efficiencies of0..65 for natural gas heating and 21'34 for air...conditioning were used,; The energies were thenconverted to dollars using average fuel prices of$0,,527 per therm (100,000 Btu) for natural gas, and$0,,0786 kilowatt-hour for electricity.. Thesevalues correspond to those used in the developmentof ASHRAE Standard 9O..2P (private communica...tion from D.. Ober 1989)" The present value of lifecycle savings was estimated using a 7 percent real(i.e., over and above general inflation) discount rate,a 25 year life, and national average fuel priceescalation rates (Lippiat and Ruegg 1988).. The fuelprice escalation rates were over and above generalinflation, and averaged about 1.7 percent per yearfor natural gas and 0..16 percent per year for.............................. ..,.,....... ',. Thble 4 gives estimated present valuetables for the horizontal RB and for RBs attachedto the bottoms of the rafters. The Fact Sheet also

1.242 Wilkes

gives present value tables for RBs draped over therafters or attached to the roof deck. For comparisonpurposes, tables for adding extra insulation are alsogiven..

Thble 4 shows that the savings vary significantly withclimate, with larger savings in warmer climates. Thetables also show that savings due to radiant barriersdecrease greatly as the level of insulation isincreased.. For any given situation, an investmentdecision would require comparison of the installedcosts per square foot of ceiling with the presentvalue of lifetime savings per square foot of ceiling..

SUMMARY AND RECOMMENDATIONS

A Fact Sheet on radiant barriers is presently beingdeveloped. Details have been given of the methodsused to estimate the present value of energy savingsdue to radiant barriers, and examples of present­value savings in the Fact Sheet have been presented.The present-value savings show that radiant barriersare more cost-effective for warmer climates. Theyare also more cost-effective when they are used incombination with low levels of attic insulation. Thepresent...value savings for s attached to thebottoms of the rafters are greater than those forRBs applied directly on top of the attic insulation,when account is taken of the effects of dust accumu­lation.. However, the cost of a RB on the rafterswould generally be higher than for a RB on top ofthe insulation (partly because the total area of theroof and gables is greater than that of the ceiling)..Any investment decision would require a compari­son of installed costs with the present value of thelifetime savings.

A number of areas have been identified where addi­tional research is needed. First, validation of theORNL model has been based on data from fieldtests in warmer climates, such as Florida andThnnessee.. There is a need for field experiments incold climates to validate more fully the accuracy ofthe model in predicting heating load reductions I'

All available data suggest that dust will accumulatefairly rapidly on horizontal RBs. The limitedamount of experimental data on the effect ofemissivity on performance of horizontal RBs hasresulted in a wide variation in present value savings..

4~P

rese

ntVa

lue

Savi

ngs

for

Rad

iant

Bar

rier

s

Pre

sen

tV

alu

eS

avin

gs,

Cen

tsp

er

Sq

uar

eF

oo

to

fC

eil

ing

Du

sty

RBo

nA

ttic

Flo

or*

*RB

Att

ach

edto

Bo

tto

ms

of

Raf

ters

**

*

Alb

any,

NY4

-13

2-

61

-3

1-

31

7-1

98

-9

4-

53

-4

Alb

uq

uer

qu

e,NM

5-1

83

-10

2-

61

-5

24

-27

12

-15

8-1

06

-8

Atl

an

ta,

GA5

-17

2-

81

-5

1-

42

1-2

51

0-1

36

-8

5-

7B

ism

arck

,ND

5-1

42

-6

1-

41

-3

18

-20

9-1

05

-6

4-

5C

hic

ago

,1L

4-1

32-

61

-q.

1-

31

7-1

98

-10

5-

64

-5

Den

ver

,CO

5-1

52

-7

1-

51

-4

19

-22

10

-12

6-

85

-7

El

To

ro,

CA4

-15

2-

71

-5

1-

41

9-2

21

0-1

26-

85

-7

Ho

ust

on

,TX

5-1

93'

-10

2-

61

-4

23

-28

12

-15

7-1

05

-8

Kn

ox

vil

le,

TN5

-17

2-8

2-

51

-4

22

-25

11

-13

7-9

5-

7L

asV

egas

,NV

7-2

43

-12

2-

72-

63

0-3

61

5-1

99

-12

7-1

0L

osA

ng

eles

,CA

3-

82

-5

1-

31

-2

11

-12

6-

74

-5

3-

4M

emph

is,

TN5

-18

2-

91

-5

1-

42

3-2

71

1-1

47-

96-

8M

iam

i,FL

6-2

33

-12

2-

71

-6

28

-36

15

-20

9-1

37

-10

Min

nea

po

lis,

MN

4-1

32

-6

1-

31

-3

18

-19

8-1

05

-6

3-

4O

rlan

do

,FL

5-2

13

-10

2-

71

-5

26

-32

13

-17

8-1

27

-10

Ph

oen

ix,

AZ

8-2

94

-14

2-

82-

73

6-4

31

7-2

31

0-1

48

-12

Po

rtla

nd

,M

E4

-10

2-

41

-2

1-

21

4-1

56

-6

3-

43

-3

Po

rtla

nd

,OR

4-1

12

-5

1-

31

-2

14

-16

7-

84

-5

3-

4

OJR

ale

igh

,NC

5-1

62-

81

-5

1-

42

0-2

41

0-1

26

-8

5-

7t::

Riv

ers

ide,

CA6

-21

3-1

02

-6

1-

52

7-3

71

3-1

77

-10

6-

8~

Sac

ram

ento

,CA

5-1

83

92

-6

1-

52

3-2

61

2-1

47

-10

6-

85- CO

Salt

Lak

eC

ity

,DT

5-1

62

-8

1-

51

-4

21

-24

10

-12

6-

85

-7

~S

t..

Lo

uis

,M

O5

-16

2-8

1-

51

-4

21

-24

10

-13

6-

85

-7

t::S

eatt

le,

WA

3-

81

-3

1-

20

-1

11

-12

5-

53

-3

2-

2-0

- :3T

op

eka,

KS5

-17

2-9

2-

51

-4

22

-26

11

-13

7-

95-

7CD :;:

,W

aco,

TX6

-21

3-1

02

-6

1-

52

6-3

11

3-1

78

-11

6-

9~ Q

)W

ash

ing

ton

,D

..C..

5-1

52

-7

1-

41

-4

20

-23

9-1

26

-7

5-

6:;:

, 0..

*D

eno

tes

lev

el

of

co

nv

en

tio

nal

att

icin

sula

tio

n..

~ "t)

I*

*N

ote

:V

alu

esre

pre

sen

tra

ng

eo

fsa

vin

gs

du

eto

vari

ati

on

sin

rate

of

du

stin

gan

dto

~ ::su

ncert

ain

ties

ineff

ect

of

du

sto

nh

eat

flo

ws

..0 (1

) rnI

**

*N

ote

:F

irst

valu

eap

pli

es

toh

ou

ses

wit

hno

air

-co

nd

itio

nin

gd

ucts

inatt

ics

..S

eco

nd

valu

e-.t

Iap

pli

es

toh

ou

ses

wit

hair

-co

nd

itio

nin

gd

ucts

inatt

ics

..i\.> ~ CJ

.)

There is a need for additional experimental work todetermine a more accurate relationship betweenemissivity and performance, both for heating andcooling conditionso

Modeling results suggest that roof RBs can have alarge effect on heat gains to air-conditioning ductsthat are run in attic spaces. Experiments are neededto check the adequacy of the adjustment factors thathave been applied to account for duct effectso

ACKNOWLEDGMENTS

This work was supported by the Office of Buildingsand Community Systems, Existing Buildings EnergyEfficiency Research Program, DoS. Department ofEnergy, under Contract No. DE-AC05-840R21400with Martin Marietta Energy Systems, Inc., by theTennessee Valley Authority, and by the ElectricPower Research Institute,. The author wishes toexpress his thanks to the Radiant Barrier SystemsThchnical Panel for their role in developing the FactSheet, and to M. Gettings and H. McLain of ORNLfor their assistance in coupling the attic model withDOE...2 and in performing the DOE-2 runs.

REFERENCES

urne, Ro c., and M. A Hoeschele. 1988.."Simulated Performance ofAttic Radiant Barriers inSacramento. It Davis Energy Group, Davis, California95616.

California Energy CommissionG 19840 "ApplicationPackage for Interim Certification of SpaceConditioning Energy Performance CalculationMethods for Residential Buildings<l" CaliforniaEnergy Commission, Sacramento, California,.

II, J<l A 19880 Radiant Barrier Testing to AssessEffects of Dust Accumulation, Attic Ventilation, andOther Key Variables 0 TVA/OPfED&T-88/25,Thnnessee Valley Authority, Chattanooga,Thnnessee 37402&

Huang Y. R. Ritschard, JG Bull, S. Byrne, [D. Wilson, C. Hsui, and D& Foley<l 1987.

Methodology andAssumptionsfor EvaluatingHeatingand Cooling Energy Requirements in New Single­Family Residential Buildings. Technical SupportDocument for the PEAR Microcomputer Program,

1.244 Wilkes

Lawrence Berkeley Laboratory Report LBL-19128,UC-95d, Berkeley, California.

Labs, K, J.. Carmody, Ro Sterling, L. Shen, Y. J.Huang, and D. Parker. 1988. Building FoundationDesign Handbook, ORNL/Sub/86-72143/1, pp. 94-97,Oak Ridge National Laboratory, Oak Ridge,Thnnessee 37831"

Levins, W. E, and M.. A Karnitz.. 1986. CoolingEnergy Measurements of Unoccupied Single-FamilyHouses with Attics Containing Radiant Barriers"ORNL/CON-200, Oak Ridge National Laboratory,Oak Ridge, TN 37831"

Lippiatt, B.. C., and R. 1: Ruegg. 1988. Energy Pricesand Discount Factors for Life-Cycle Cost Analysis1988, Annual Supplement to NBS Handbook 135 andNBS Special Publication 709" NISTIR 85-3273-3,DoS. Department of Commerce..

Peavy, B" A 1979. "A Model for Predicting theThermal Performance of Ventilated Attics." InSummer Attic and Whole-House Ventilation. NBSSpecial Publication 548, pp. 119-1499

Sherman M" H., and D.. L Grimsrud.. 1980.Infiltration-Pressurization Correlationso" SimplifiedPhysical Modeling.. Lawrence Berkeley LaboratoryReport LBL-I0163, Berkeley, California.

Wilkes, K E .. 1983. "Dynamic Thermal Performanceof Walls and Ceilings/Attics.. It Proceedings,ASHRAE/DOE Conference on Thermal Peiformanceof the Exterior Envelopes of Buildings II, ASHRAESP 38, pp" 131-159.

Wilkes, K. E.. 1988. "Thermal Modeling ofResidential Attics with Radiant Barriers.." FifthAnnual Symposium on Improving Building EnergyEfficiency in Hot and Humid Climates, pp& 161-168"

Wilkes, K E" 19890 "Thermal Modeling ofResidential Attics with Radiant Barriers:Comparison with Laboratory and Field Data."Proceedings, Thermal Performance of the ExteriorEnvelopes ofBuildings IV, ASHRAE, pp. 294-311.

Wilkes, K E. 199Oa.. Thermal Model ofAttic Systemswith Radiant Barriers, ORNL/CON..262 (to bepublished), Oak Ridge National Laboratory, OakRidge, Thnnessee 37831.,

Wilkes, K Eo 1990b. Analysis ofReductions in HouseHeating and Cooling Loads with Radiant Barriers..ORNL/CON-281 (to be published), Oak RidgeNational Laboratory, Oak Ridge, Thnnessee 37831..

Wilkes, K E., and D" W. Yarbrough. 1988. RadiantBarrier Research Plan, ORNL/CON-256, Oak RidgeNational Laboratory, Oak Ridge, Thnnessee 37831..

Building Equipment and Appliances 1.245

1.246