Windows, Baths, and Solar Energy in the Roman Empire

Author(s): James W. Ring

Source: American Journal of Archaeology, Vol. 100, No. 4 (Oct., 1996), pp. 717-724
Published by: Archaeological Institute of America
Stable URL: http://www.jstor.org/stable/506675
Accessed: 24-05-2017 04:04 UTC

REFERENCES
Linked references are available on JSTOR for this article:
http://www.jstor.org/stable/506675?seq=1&cid=pdf-reference#references_tab_contents
You may need to log in to JSTOR to access the linked references.

JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of content in a trusted
digital archive. We use information technology and tools to increase productivity and facilitate new forms of scholarship. For more information about
JSTOR, please contact support@jstor.org.

Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at
http://about.jstor.org/terms

Archaeological Institute of America is collaborating with JSTOR to digitize, preserve and extend access
to American Journal of Archaeology

This content downloaded from 132.248.9.8 on Wed, 24 May 2017 04:04:54 UTC
All use subject to http://about.jstor.org/terms
 Windows, Baths, and Solar Energy in the Roman Empire
JAMES W. RING

Abstract sea."' One can infer from his writings that Seneca
Windows were a prominent feature of Roman archi- regrets this new style and, indeed, it is clear that he
tecture and were especially important in the magnifi- looks back with nostalgia on the days of the Roman
cent bath buildings of the Roman Empire. A growing
Republic when baths were properly and modestly
literature attests to the Romans' use of solar energy in
dark.
heating these large buildings. Edwin Thatcher claimed
in 1956 that the windows in such baths did not require Roman baths of the Early Empire were in the fore-
glazing. In this paper I refute this claim, drawing on front of developments both architecturally and tech-
modern ideas about solar energy, heat transfer, human nologically and thus make a very interesting study
comfort, and the effect of glazed windows to analyze
in their own right. This point has been brought out
one room in the Forum Baths at Ostia. This analysis
is compared with that of Thatcher for the same room. and ably developed by Yegul2 and Nielsen. Vaults,
In window size and solar orientation, this room is typ- domes, and large windows were first found in these
ical of Roman baths in many parts of the empire. The baths, where Greek orders were also first combined
solar science and technology of today is thus compared with Roman vaults. Hypocausts were developed and
with that of the Romans and with that of Thatcher's dayf*
used to heat large rooms and, indeed, to heat the
imposing ensembles of large rooms that the great
INTRODUCTION
imperial baths represented. According to Yegill,
The importance of windows in architecture seems
Seneca speaks of the recent invention of tubuli, or
indisputable. But in its chronological development
hollow walls, which maintain an even temperature
Western architecture shows striking variation
in the inas the highest spaces. This in-
lowest as well
its treatment of windows. For example, medieval
vention also prevents condensation on the walls and
churches of the Romanesque style wereincreases the areadark
typically that radiates heat around the bath-
and lit only by small windows that pierced massive
ers. In these large evenly heated spaces, thousands
of bathersfeatures
masonry walls. One of the distinguishing could be and often were accommodated.
of later Gothic churches was the use
Toof large
supply win-water, extensive aqueduct systems
sufficient
dows made possible by outside buttressing.
were developed.The Furthermore, as Yegill maintains,
Roman Empire prototypes, however, andunlike
is showntheir
in a detailed fashion by D.B. Harden,
Romanesque successors, were often litthe
byRomans
magnifi-
by this time had developed glassblow-
cently large windows. The Roman public baths
ing and were of
producing flat panes of window glass.4
the Early Empire are very good examples
Thus, withof
all this
of these elements in hand, it is not
surprising
anachronism. Seneca, writing in the first century that the Romans would have utilized the
A.D.,
says of these baths: "Nowadays... people regard
radiant energy baths
of the sun to help heat as well as light
as fit only for moths if they have not been
these so arranged
magnificent buildings.
that they receive the sun all day long through the is the Romans' interest in so-
Indeed, so obvious
widest of windows, if men can not bathe and through
lar heating get a their use of large south-facing
coat of tan at the same time, and if they can not
windows look
that in 1956 Edwin Thatcher published a
out from their bath-tubs over stretches of land and
paper in which he claimed that the large windows

* I wish to thank the American Academy in Rome for a Faculty Fellowship in 1993 during which research lead-
help in arranging visits to various Roman sites, including ing to this paper was completed.
Ostia, in 1985; the Corning Museum of Glass and Rakow 1 Sen. Ep. 86, trans. R.M. Gummere, 1986, as quoted by
Library for help in exploring Roman glass in 1991; Todd E Yegill, Baths and Bathing in Classical Antiquity (Cambridge,
Moore of the Physics Department at Hamilton College for Mass. 1992) 40.
his careful reading of the manuscript; Carl Rubino of the 2Yegiil (supra n. 1).
Classics Department at Hamilton for his advice; Fikret Yegiil 3 I. Nielsen, Thermae and Balnea: The Architecture and Cul-
of the University of California at Santa Barbara and Ingrid tural History of Roman Public Baths (Aarhus 1990).
E.M. Edlund-Berry of the University of Texas at Austin for 4 Yegiil (supra n. 1) 363-65; and D.B. Harden, "Domes-
reading an earlier version of this manuscript and advising tic Window Glass: Roman, Saxon and Medieval Studies in
about publication; and Richard de Dear of Macquarie Uni- Building History," in E.M. Jope ed., Studies in Building His-
versity, Sydney, for his advice about thermal comfort con- tory. Essays in Recognition of B.H. St.J. O'Neil (London 1961)
ditions in these baths. Finally, Hamilton College provided 39-63.
717
American Journal of Archaeology 100 (1996) 717-24

This content downloaded from 132.248.9.8 on Wed, 24 May 2017 04:04:54 UTC
All use subject to http://about.jstor.org/terms
 718 JAMES W. RING [AJA 100
ing shortages
of the second-century Forumand expense,Baths the Romans, like
attheO
Greeks before them,
he took to be unglazed, turned to solar heat."8
provide "a strik
stration of the potentialities of the Rom
method and, in DID THE ROMANS USE GLAZING?
extension, of the prin
diant heating. It was
Thethis method
Forum Baths that early
in Ostia were constructed m
rooms possible and,
in the to
second date, we
century A.D. There is nohave n
controversy
them in a modern building. It is evide
about the existence of large windows in several o
Roman engineers the
had a baths.
rooms of these greater
These windows faced confid
the
diant heating than south
we and have and the
hence would intercept a sun's
greate
beam ra-
of what it could accomplish."
diation most of the day, particularly from early after-
This paper sets out
noon to to investigate
near sunset, th
which were the most popular
Thatcher's confidence in
hours for Romans radiant
to bathe. heati
These rooms indeed are
Could there be enough heating
typical of baths provided
built during this period in many parts
and the hypocaust of
to allow nude bathers
the empire.
fortable even though the
Thatcher large
gives attention to thewindows
whole set of large-
were open and unglazed?
windowed roomsWhat do the
but for our purposes m
let us con-
ciples of passive solar heating
centrate on one, room 4, which seems to and
have been the
physiology of heata warm
transfer
room, or tepidarium.tell us
Figures 1 and abou
2 show
Baths and Thatcher's claim?
the southern elevation and north-south section, re-
In addition, I take up a point raised by Yegiul. spectively,
He of this room and its window. The dimen-
suggests that Thatcher has gone too far in his claims
sions are those given by Thatcher.
for radiant heating: The question is whether or not the walls, vault,
In full admiration of the system's potential, I still and floor surrounding the nude bather on all but
doubt if the implications of radiant heating should window side can be maintained at a high-enough
the
be stretched that far. Not only is the evidence for win-
temperature to ensure comfort. Thatcher approaches
dow glass and window frames (both in wood and the issue from the standpoint of the nude bather
metal) from the heated rooms of Roman baths across
the Mediterranean overwhelming, but Thatcher's the-
exposed both to the radiant energy of the sun and
that given off by the surrounding heated room sur-
sis, despite its theoretical possibility, seems to refute
faces, holding that radiant energy, if the walls are
the precepts of simple economic logic. It may be that
by heating the floor, the walls, and the vault to a high
maintained at close to skin temperature, can by it-
degree, sufficient radiant energy could be released
self establish a comfortable temperature. My ap-
to offset the effects of low air temperature on a cold
proach, on the other hand, is to assume that com-
winter day, but why should fuel and energy be wasted
fort will be determined by conditions in the room,
in order to make an open-air hot bath possible when
the same degree of warmth and comfort couldboth be the air temperature and the radiant temper-
achieved with much lower furnace activity and fuel
ature being considered along with air currents, or
consumption in a glazed and well-insulated room?6
convective flows, and the relative humidity of the
This question is also raised by Jordan and Perlin air
in as prescribed by the American Society of Heat-
an article about the use of solar energy in ancient
ing, Refrigerating, and Air Conditioning Engineers
(ASHRAE).9 Comfortable conditions for nude sub-
times.7 They claim that by the first century B.C.,
Rome had to import timber from the fringes ofjects its have been studied carefully in climate cham-
domains, such as the Alpine regions, in part because
bers. An example of one such study, which also shows
of the Roman love of bathhouses- there were 800 the importance of heat transfer at the skin surface,
baths in Rome alone in the third century A.D.- isbutthat of de Dear, Ring, and Fanger."1 Also of im-
also because of the growth of industry and manu- portance is the air flow over the skin. This too has
facture. As they point out, "prices of wood, charcoal,
been studied and reported on by, for example, Fanger
and small firewood rose steeply. To avoid the grow-
and his colleagues."1

" E.D. Thatcher, "The Open Rooms of the Terme del ch. 8.

Foro at Ostia," MAAR 24 (1956) 169-264.
6 Yegfil (supra n. 1) 383.
7 B. Jordan and J. Perlin, "Solar Energy Use and Litiga-
tion in Ancient Times," Solar Law Reporter 1.3 (1979) 583-94.
8Jordan and Perlin (supra n. 7) 587.
9 ASHRAE Handbook, 1977 Fundamentals (New York 1977)
1( R.J. de Dear, J.W. Ring, and P.O. Fanger, "Thermal Sen-
sations Resulting from Sudden Ambient Temperature
Changes," Indoor Air 3 (1993) 181-92.
11 P.O. Fanger, A.K. Melikov, H. Hanzawa, andJ.W. Ring,
"Air Turbulence and Sensation of Draught," Energy and Build-
ings 12 (1988) 21-39.

This content downloaded from 132.248.9.8 on Wed, 24 May 2017 04:04:54 UTC
All use subject to http://about.jstor.org/terms
1996] WINDOWS, BATHS, AND SOLAR ENERGY IN THE ROMAN EMPIRE 719

grade level

10 ft
I I I I I I I

Fig. 1. Southern elev

The neutral environment.

problem then On the contrary, they
beco expect
of heat to feel hot.
from theThus, strictly speaking,
hypocau we are not deal-
and from the surfaces of the room to the outside ing with the usual comfort scale but rather with
mainly through the large window. These processesone biased toward the hot end. For example, the
determine the temperatures of the surfaces and thus seven-point ASHRAE scale (cold, cool, slightly cool,
the convective drafts and air temperature. Relative neutral, slightly warm, warm, hot) would allow using
humidity, of course, depends on the vapor pressure ofonly the upper point in the warm rooms of the
water in the space. In room 4, which apparently hadbaths. According to Thatcher, the temperatures of
no pools or baths, the relative humidity would not the walls and floors of these rooms were -400 C
have been particularly high, perhaps about 50%. The (or - 1000 F).12 In fact, this is in the range of evap
temperature and convective flows were thus the pri- rative regulation (sweating) that is adjacent to but n
mary determinants of human comfort in this room.part of the comfort zone, i.e., these are, according
Naked bathers do not expect or want a thermally to ASHRAE, "uncomfortable" conditions.

12 Thatcher (supra n. 5) 190-94.

This content downloaded from 132.248.9.8 on Wed, 24 May 2017 04:04:54 UTC
All use subject to http://about.jstor.org/terms
 720 JAMES W. RING [AJA 100

Yault

suspensura

tubuli hypocaust praefurnium

10ft
ILIIIIIIII

Fig. 2. North-south section of room

We take conditions to be
that temperatures like
of 700 F for thesethose
surfaces will as
Thatcher: outside cause the nude bather to radiate is
temperature heat to them as
just belo
300 F,'1 the interior surfaces of the room are at well as losing heat to them by convection (and con-
1000 F, and the sun is shining in the window duringduction if in contact with them). These will not be
December at 250 British thermal units (BTU)/ft2l/hr,warm conditions for him but rather ones somewhat
assuming a clear sky. Further, we take the temper- on the cool side. The thermal properties of the ma-
ature of the hot gases at the top of the hypocaust terials, i.e., conductivity and coefficients of heat trans-
below the floor to be 4000 F, which is consistent withfer for radiative or convective flow, can be found in
the experiments of Rookl4 with a small hypocaustthe appropriate part of the ASHRAE Handbook.16
in Welwyn, England, although somewhat above the With these parameters known, calculations can
temperature (-3000 F) found by Kretzschmer15 inbe made for the heat flows as shown below in table
experiments in Saalburg, Germany. For the purposes1. The heat flows without and with glass are shown
of the argument here, any temperature up to 4000 F diagrammatically in figures 3 and 4, with figure 3
can be posited. The crucial heat flow is that outshowing inflows and figure 4 outflows. Here we are
through the open window, and if we assume 4000 F concerned with the comparison of flows in versus
we are estimating the maximum heat flow in andflows out, for if they are not equal, the temperature
thus giving Thatcher the best chance of being cor- of the walls and the room will not be constant at,
rect. As does Thatcher, I too assume that the floor andor close to, the desired 1000 E
inside wall surfaces are held at - 1000 F, including Certain caveats about these calculations ought to
the inside of the vault, which, although unheated by be made clear. None of these heat flows can be said
tubuli, by convection and radiation will be at nearlyto be precisely defined. The problems in calculation
1000 F if the walls are also at this temperature. Note are:

1' This is, according to Thatcher (supra 14

n. T.
5) Rook,
182-83,
"The Development and Operati
a quite possible temperature in Ostia in December orJan-
Hypocausts,"JAS 5 (1978) 269-82.
uary, and in my own experience a low but 15 notF.outlandish
Kretzschmer, "Hypokausten," SaalbJb 1
one.
16ASHRAE Handbook (supra n. 9) ch. 11.

This content downloaded from 132.248.9.8 on Wed, 24 May 2017 04:04:54 UTC
All use subject to http://about.jstor.org/terms
1996] WINDOWS, BATHS, AND SOLAR ENERGY IN THE ROMAN EMPIRE 721

sun's beam radiation sun's beam radiation

conduction conduction through glass
from tubuli from tubuli

conduction conduction
from hypocaust from hypocaust

10 ft 10o ft
IU1111111111 I111111111

Fig. 3. Heat flows into

1) 5) I estimate the temperatures are

The
dimensions of the lower sus- n
I use thosepensura (floor)
given surface and the inner
by surfaces of
T the
2) In sometubulicases the
from Rook's and Kretzschmer's m
experiments.
as the upper 6) The solar
partbeam radiation is calculated
of here th
erally I in the same way as it
follow was by Thatcher, agreeing also
Thatcher
3) I assume,with theasmethod used
does by Ring and Hamilton"7
That to
heated. test the performance of a solar classroom at Hamil-
ton College at close to the same latitude as Ostia.
4) The optical quality of Roman glass varies widely
and that used at Ostia in these baths is not known. 7) Convective flows are notoriously hard to cal-
culate, but since such flows do occur in solar houses
A transmission of 50%, assumed here, is probably
a quite conservative estimate. we can expect to achieve order of magnitude results

conduction
though vault
conduction through vault

radiation

convection- flows through glass

loops

radiation

10 ft 10 ft
II!11111111 LLIIIIJlllI

Fig. 4. Heat flows out of roo

17J.W. Ring Research

and A. Institu
Hamilton, "
Hamilton 107-11.
College," Proceedings, Co
and Cooling Systems, Colorado Sprin

This content downloaded from 132.248.9.8 on Wed, 24 May 2017 04:04:54 UTC
All use subject to http://about.jstor.org/terms
 722 JAMES W. RING [AJA 100
Table 1. Heat Flows for Room 4 with Glazed and much more than adequate to provide the outward
Unglazed Windows flow (80,000 BTU/hr) in radiation, conduction, and
convection through the glazed window and in con-
Units of
Direction and 100,000 duction through the vault. The sun alone on sunny
Source of Flow BTU/hr days could provide most of the energy to maintain
100' F temperatures. Indeed, even with the fires re-
Into room
duced on sunny days, there would probably be some
Hypocaust (full heat) 4.0
thermal energy stored in the floors and walls that
Sun (clear day at winter solstice)
Unglazed window 1.3 would maintain the temperature as the sun goes
Glazed window 0.7
down. On days when the sun is obscured by clouds,
TOTALS the hypocaust with a reduced fire, or being on only
Unglazed window 5.3 part of the time, could by itself easily maintain the
Glazed window 4.7
temperature even with the outside temperature at
Out of room (outside its coldest point of the season, i.e., the
temperature =design
300 tem- F
Conduction 0.4 perature of 300 E
Natural convection and radiation
The stored thermal energy, which may come from
Unglazed window 52.5 either the sun or the hypocaust or both, can be
Glazed window 0.4
TOTALS
handled quite easily by the heavy masonry walls,
vault, and floor of this room. With such surround-
Unglazed window 52.9
Glazed window 0.8 ings extra heat in the room can pass readily by con-
duction into the masonry without heating the air
Net Flow (+ = into room)
in the room excessively, i.e., much above 100 .E At
Unglazed window - 47.6
Glazed window + 3.9 night, when the sun is down (even with the fire out),
this stored heat will flow back into the room to offset

the cooling that inevitably will occur. Note that

wooden shutters
using the solar designer's closing the windowNatural
formulas.'" area at night r
than forced convection is assumed, i.e., there is nowould enhance this storage considerably. Such heat
wind blowing. With wind the convective flow is storage is an important element in solar house de-
greater, and could be much greater at high-wind sign and the Romans seem to have incorporated this
velocities. element into their designs as well. The thickness of
Under such circumstances we can expect only tothe floor, or suspensura, is especially interesting in
this regard and it may be that the thickness chosen,
estimate these flows. Even with these rough estimates,
however, some important conclusions can be i.e., 15 in, is not necessary structurally but aids in
reached. The estimates for glazed and unglazed win- long-term heat retention.
dows are given in table 1 above. The details of the Thus, we see that Yeguil is correct in his claim that
calculations follow in an appendix. the Romans would have been wise to use glass in
their bath windows. Furthermore, we see that
RESULTS AND DISCUSSION
Thatcher is probably overly enthusiastic in claiming
such efficiency for radiant heating. Indeed, with
The results of the calculations summarized in table

I and in the appendix indicate that with an openthe windows open, the rooms in these baths could
window the input is - 530,000 BTU/hr while thenot have been maintained at 1000 F and a nude
outflow is - 5,290,000 BTU/hr. In such a case it is ob- bather would soon have become very chilly. Thatche
vious that equilibrium is not possible and that theneglected to consider fully the very great heat flo
100' F surfaces will rapidly cool toward 300 E Theout of the room due to the convective flow throug
same would be true for a nude bather whose skin the open window. He does not completely ignore th
possibility of an air current but claims that such a
temperature normally should be - 930 E Indeed the
outflow would equilibrate with the inflow only when
current would only exist with no wind, and normall
the wall and floor surfaces are within 100 F or less there would be some wind. He then says, "A win
pressure of any but the lowest magnitude would nul
ofOn
thethe
outside
othertemperature,
hand, if the i.e.,
windowat _ is
400 E
glazed thelify the action and set in motion the various air cur
heat from the hypocaust (400,000 BTU/hr) would berents already described."'' These currents, Thatcher

'8J.D. Balcomb ed., Passive Solar Buildings (Cambridge,

'1 Thatcher (supra n. 5) 233-34.
Mass. 1992) 149-52.

This content downloaded from 132.248.9.8 on Wed, 24 May 2017 04:04:54 UTC
All use subject to http://about.jstor.org/terms
1996] WINDOWS, BATHS, AND SOLAR ENERGY IN THE ROMAN EMPIRE 723

believed, flowed from north to south, or vice versa, it seems, was too sanguine about radiant
Thatcher,
across the room; in the north wall they passed
heating but, nevertheless, the Romans deserve high
through the door, cracks around the door, orfor
praise thetheir use of solar energy. Even Seneca, no
admirer
lunette at the top of the vault, and in the south of conspicuous consumption and easy liv-
wall,
through the upper part of the window. ing, might have admired the frugality that the com-
This assertion seems incorrect because a wind bination of "the widest of windows" and glass panes
blowing in, or eddying through the southern win- in baths demonstrated.
dow, will not nullify the convective effect. Rather it
will cause a change from natural convection to forced
Appendix
convection, changing and distorting the geometry
Calculations of Heat Flows for Room 4
of the convective loop and, as a result, increasing
the mixing of cold and hot air and thus increasing the with and without Glazing
heat loss above that caused by natural convection
A. Heat flows into the room:
alone. The net result would be to set the 52.5 x 105
(i) The heat flow through the suspensura: this is a heat
BTU/hr heat loss calculated above as a minimum
conduction problem where Q, the heat flow per hour,
value and in windy situations to expect this loss istogiven by:
be even greater with the concomitant effect of an
even faster lowering of the bath temperature, more
Q kAAT with k the thermal conductivity of the
concrete slab, Ax the thickness of the
quickly chilling the nude bather.
Note that the natural convective flow calculated slab, A the cross-sectional area, and AT
the temperature difference between
here is only a rough estimate. But it is about 10 times
the top and bottom of the slab.
the inflow so that even if it is overestimated by a fac-
Using British engineering units with
tor of two, it still will be many times greater than the
k = 11 BTU-in/oF-ft2-hr (value used by Thatcher)
inflow. With wind, it will be even greater than cal- A = 1,200 ft2
culated here.
AT = 400 - 100 = 3000 F (using maximum hot
Finally, to return to Yegiil's point, the estimates gas temperature under suspensura)
of heat flows here show not only that nude bathing Ax = 15 in

in Roman baths would not have been possible with- Q = 2.6 x 105 BTU/hr
out glazing but also that with glazing during sunny (ii) The heat flow through the heated walls:
days, the sun with only a little help from the hypo- k = 7.0 BTU-in/oF-ft2-hr (value used by Thatcher)
caust and its furnaces, and hence little wood burned, A = 1,760 ft2 (3 vertical walls)
could have maintained the temperature of these AT estimated to be 200 - 100 = 1000 F
room surfaces at - 100' E Furthermore, on cloudy Ax = 9 in

days the hypocaust with only a low or intermittent Q = 1.4 x 105 BTU/hr
fire would have been able to sustain this tempera-
(iii) The sun's radiant energy in through the window
ture. And even at night the large thermal storage In December in Rome, the sun at noon is only abou
capacity would have kept temperatures from drop- 240 above the horizon and the sun's beam intensity
approximately 250 BTU/hr/ft2. Note that compare
ping very fast so that by the next morning the amount
of heat necessary to return to - 1000 F might havewith 228 at the winter solstice, the noon value on a south
been relatively small. Thus with a normal mix facing
of vertical surface at equinox would be 285 x c
420, or 212 BTU/hr/ft2 at this latitude of 420 N. At th
sunny days, a considerable savings of fuel could be
summer solstice this value would be 116. At noon, r
accomplished even in the depths of winter. At other
gardless of the season, the sun's beam radiation on th
seasons even more savings could be expected. The
surface would always be at its maximum for the day
sun would therefore provide a substantial part of
Note also that these intensity values agree with thos
the heat required. This result, of course, is in accord
used by Thatcher:
with Jordan and Perlin's observations about the in- A = 560 ft2
creasing cost of fuel during this period of rapid Q = 250 A cos 240 = 1.3 x 105 BTUlhr
growth of Roman industry, commerce, manufacture,
or with Roman glass:
and population. Fuel costs would have provided a
Q -the
Thus, 0.65 x 105
total BTU/hr
energy input a 5.3 x 105 BTUlhr, i.e.,
strong incentive for using glazed windows and the
sun's energy.
the sum
x 105 of the above items, or with glass - 4.65
BTUlhr.
In summary, the Romans apparently did display
B. Heat flows out of the room:
considerable know-how in the design of their baths
when judged by the standards and practices of mod-
(iv) Conduction through vault to the outside:
ern science and technology 2,000 years later. k = 7.0 BTU-in/oF-ft2-hr, as in the walls

This content downloaded from 132.248.9.8 on Wed, 24 May 2017 04:04:54 UTC
All use subject to http://about.jstor.org/terms
724 J.W. RING, WINDOWS, BATHS, AND SOLAR ENERGY IN THE ROMAN EMPIRE
A = 1,880 ft2 Using A = 280 ft2, H = 10 ft, and AT = 701 F as
AT = 100-30 = 700 F above:
Ax = 24 in Q = 6.7 x 106 BTU/hr
Q = 3.8 x 104 BTU/hr This is 25% more than the estimate above. I use
the smaller number in table 1.
(v) The natural convective flow through the open
window: (vi) The radiant heat flow through the window:
Here the convective loop will have hot air exiting Here the Stefan-Boltzman law governs the heat flow
through the top of this large window and cold air com- (the same equation used by Thatcher).
ing in at the bottom. See figure 4. This flow will be caused Q = oAE(T14 - T24)
by the stack effect in which hot, less dense gas is forced with o = 1,730 x 10-12 BTU/hr-ft2-(OF absolute)4
out of the room at the top of the window and, by the A = area of body in question (the window in this
same effect, cold air that is more dense will flow in at case)
the bottom. Solar house designers use the formula given E = 0.9, a factor accounting for emissivities and sol-
below to calculate this air flow.20 id angles subtended by the hot and cold bodies
CFM = 9.4Aeff'HAT as seen through the window. Again this is the
where Aeff is the effective area (ft2) through which value used by Thatcher.
the flow enters and/or leaves the space, and H is theT1 and T2 are 5600 F and 4900 F absolute, respec-
effective height of this area. In this case the Aeff is half tively, for the hot inner surfaces at 1000 F and
the area of the window while H is half the height of outside surfaces (or air) at 30 F.E
the window as half the window is used for outward flow Q = 3.5 x 104 BTU/hr
and the other half for inward flow, or Aeff = 260 ft2, If the window is glazed, a combination of convection,
H = 10 ft (mean height), and AT = 100 - 30 = 700 F radiation, and conduction across boundary layers of air
and: both on the inside and outside as well as the glass itself
CFM = 7.0 x 104 ft3/min can be treated according to the following equation:22
Then the heat transferred will be: Q = UAAT
Q = CFM x 60 x AT x 0.018 where U = 1.10 BTU/hr-ft2-oF, which includes all
where 0.018 is the volumetric specific heat of air in three types of heat flow
A = area in ft2 of window = 560
BTU/ft3 at these temperatures.
AT = 700 F
Q = 5.2 x 106 BTU/hr
Note that another equation for this convection heat Q = 4.3 x 104 BTU/hr
transfer mentioned by Balcomb is that of Weber and
Kearney,2' who arrived at it by similitude modeling and
full-scale testing:
DEPARTMENT OF PHYSICS
Q = 4.6 W(dAT)312 where W is width and d is height
HAMILTON COLLEGE
of opening, which when converted to the vari-
ables used above becomes CLINTON, NEW YORK 13323
Q = 13 A -H (AT)3'2 JRING@ITSMAIL1.HAMILTON.EDU

21) Balcomb (supra n. 18) 149-52. Amherst, Massachusetts, October 1980 (American Section of
21 D.D. Weber and R.J. Kearny, "Natural Convection the International Solar Energy Society, Newark, Del. 1980)
Heat Transfer through an Aperture in Passive Solar Heated 1037-41.

Buildings," Proceedings, 5th National Passive Solar Conference, 22 ASHRAE Handbook (supra n. 9) ch. 11.

This content downloaded from 132.248.9.8 on Wed, 24 May 2017 04:04:54 UTC
All use subject to http://about.jstor.org/terms