6A - Update Memo
October 7111, 2009
TO: Landmarks Board
FROM: James Hewat, Chris Meschuk
SUBJECT: Update Memo
Post WW-II Residential Subdivision Survey and Context
A draft of the context, survey, and management recommendations have been submitted by the
consultant and are currently being reviewed by staff and the Colorado Historical Society. The
information will likely be presented to the Board in December and presentations to the public
will follow.
Valmont Mill and Depot
Staff has completed Colorado Historical Society historic structure assessment (HSA) grant
application for the Valmont Mill. A Depot task force will be convened in the near future to
review the Depot HSA grant application and commence planning for the rehabilitation and re-
use of the building.
Mapleton School Coalition
A request to the Boulder Valley School District to lease the school for use as an early childhood
learning center will be taken up by the School Board on October 13, 2009. Conceptual designs
for the proposed addition and parking were reviewed by the Ldre on August 19th.
2009 CLG Grant Application
The grant to digitize the survey forms and photographs is underway.
New and Pending Land Use. Review Applications
None
Planning Board Calendar
See attached.
Stay-of-Dernolition Status Summary, October 7th, 2009
None
Landmark Applications:
800 Pearl Street designated a landmark, September 1st, 2009
First reading of 114313t1, Street landmark designation, October 6w, 2009
First reading of 115513th Street landmark designation, October 611,, 2009
Articles:
"An Analysis of the Thermal Performance of Repaired and Replacement Windows". Association
of Preservation Technology International Journal. Volume XL, No. 2.
1
An Analysis of the Thermal Performance of
Repaired and Replacement Windows
ROBERT SCORE AND hI~ADFORD S. CARPENTER
Data and analysis of in-situ thermal Introduction use nri l is bcing proposed for a signifi-
monitoring reveal that the repair of As federal buildings that were con- cant modernization projecr. Built in
aging steel windows offers the structed to support the expanding role 1940 and designed by the Chicago
of the U.S. government and the war architectural firm of Holabird and Root,
opportunity to retain historic it is a National Historic Landmark.
the enuringd of the late
building fabric and secure a level of rcacah fort during
thateeir usefu1930sl and
lindves, 19 their The building has nearly 1,200 win-
energy performance that can match caretakers are embarking on rehabilita down along the primary facades and a
or exceed that of modern aluminum- tion and modernization projects to meet single interior light court. The windows
modern and often greatly expanded are constructed of steel shapes and were
framed replacement windows. performance standards. The original installed in a double-hung configuration
igina[
window systems often lack the con- (Fig. 2). The windows are approxi-
struction detailing and other character- mately 54 inches wide and 72 inches
high
istics needed to provide a level of per and have single glazing without
formance acceptable for modern office intermediate muntins. Many of the
space, such as humidification and en- windows exhibit paint flaking and some
ergy performance. Many of these win- surface rusting, while others have more
dove systems have suffered years of significant rusting, particularly at the
base of the jambs (Fig. 3). The windows
neglect and deferred maintenance and
awe their longevity largely to the dura- weareicurrently operable and have counter-
we
of original materials, the robust ghts in concealed weight pockets.
ness of the original construction, and this paper discusses a brief study corn-
layers upon layers of paint. paring two options for treating the
A common and as-yet unresolved windows during the planned repair
issue is the final fate of these windows. Program.
More often than not, building-renova-
tion projects call for the replacement of Performance Requirements and
original windows with modern replicas Design Options
rather than the rehabilitation of the The Lafayette Building is scheduled to
Cr fisting windows often under the guise
undergo a comprehensive renovation,
r - of improving energy performance or including upgrades to the mechanical
e L „ Occupant safety (blast resistance), with
;t , tt and electrical systems, reprogramming
little thought given to the embodied
and renovation of interior spaces, and
;q r - r,','•, energy in the existing windows or the
. renovations to all exterior facades,
whole-life energy commitment of the including upgrading the windows for
rr;,,, new product. This consideration be- blast resistance in accordance with
x' r comes even more crucial when consider requirements of the General Services
ing the demands of achieving LEED Administration (GSA), the building
_ ""o ratings in a renovation project. owner. In order to assist the building
One structure ' currently being consid-
owner in selecting the most appropriate
ercd for such renovation is the Lafayette treatments for the windows, an overall
i juiHim, a federal office buildin Io-
bf g design program for the windows was
sated in downtown Washington, D.C. developed. It identified the performance
g. 1. Southwest elevation, Lafayette Building, (rrg• 1). Originally housing the Export- requirements for the windows, includ-
Vermont Avenue and H Street NW, Washington, Import Bank of the United States and ing the following:
D C. The building's neoclassical exterior has more recently the Department of Vet- . provide blast resistance per GSA
regularly placed, double-hung steel windows, All eran Affairs, the Lafayette Building has
images by authors requirements.
had a long and storied history of federal
APT BULLETIN: JOURNA_- OF PRESE VA[InN TECHNOLOGY / 40 2, 2009
rxer,,;;; ?I steel would be prepared to SSPC-SP3, treatments proposed for the actual
, the standard specification for power-tool construction in order to provide accu-
-~'+•I cleaning of steel surfaces by the Satiety rate results for comparison (figs. 4
for Protective Coatings (SSPC), and through 7). Mock-ups were installed in
given two coats of acrylic enamel paint. the east elevation of the building at the
1 The exterior light of glass, a single pane eleventh-floor level based upon input
of clear Float glass, would be retained from the owner and design team. The
1 I I I ' where possible. Blast resistance requires location on the east elevation of the
that the original sash be fixed shut. A eleventh floor, a height roughly equal
blast-resistant aluminum-framed storm with the rooftops of surrounding
build- window would be installed on the inte- ing, allowed a relatively unobstructed
n
iI rior face of the window, approximately solar exposure and conditions that vary
3 inches from the face of the glass in the between the diffuse solar radiation of
lower operable sash. The storin window north exposures and the intense solar
would include a single larninate sheet of radiation of south and west exposures.
glazing that includes low-L' lass with a
high solar heat-gain coefficient (SHGC) Performance Monitoring
_ to provide improved passive solar heat
.a gain. The monitoring study was undertaken
in 2006 by 1 Jarboe Architects and
Fig. 2. View of an unrepaired steel-framed Optima 2. Replace the existing windows Simpson Gumpertz Fat Heger, Inc.
window on the east elevation of the Lafayette with new blast-resistant, thermally (S('.H), The purpose of the study was to
Building. Note the loss of paint coatings and broken, aluminum-framed windows document and evaluate the performance
corrosion of the built-up steel frame. with 1-inch insulating glazing. The new of a repaired window and a proposed
windows would closely match the exist- replacement window under similar
preserve the original window sash ing configuration, dimensions, profiles, eXposur-es and to provide direction and
and frames where possible, including and sight lines of the original windows. feedback to the design team for iucor-
original materials., configuration, Original windows would be removed poration into the rehabilitation pro-
from the opening; interior finishes
dimensions, sight lines, and profiles, grain. The monitoring system allowed
as well as the clarity and reflectivity would be removed from the perimeter of for the recording of surface and air
:.I of the original glazing. Any replace- the window; and modifications made to temperatures, as well as relative humid-
ment windows must match existing the perimeter substrate and trim to ity for multiple locations,
window configuration, dimensions, allow installation of mounting clips to The two mock-up windows were
sight-lines and profiles, as well,as the secure the replacement windows to the monitored between IMorch and July
clarity and reflectivity of the original masonry hack-up. Interior finishes 2006. Though the duration was limited
would then be repaired to conceal the
glazing. to just over three months by program
e anchorages. The replacement-window and tenant constraints, sufficient data
.j improve energy performance and insulated-glass unit would include low-E.
reduce air infiltration and water was gathered to compare the perfor-
glass that has a high SIIGC glazing, rtrarrce of the mock-ups over a signifi-
penetration at the windows. which would minimize passive solar
cant range of exterior conditions. This
e provide windows that are easily heat gain. The operable sash would be recorded performance allowed for the
maintainable. fixed shut to meet blast requirements. extrapolation of performance outside o
r e provide a cost-efficient treatment. 1vtore historically accurate steel replace- the range of measured interior and
These performance requirements went windows were not considered for
were used to develop design options, this option as they were cost prohibitive
evaluate the technical options, and then compared to restoring the existing win-
help select the most appropriate treat- doves as described in Option 1 and i ~
ment. Based on the existing conditions offered little thermal-performance irn-
and the design requirements, the follow- provement over the existing windows.
ing two options werc identified:
Option 1. Repair the existing steel Window Mock-ups windows and provide a supplemental 'I'o assist in evaluating the two options,
interior storm window that meets the in-situ mock-ups of both options were
blast requirements and also improves installed in the building, allowing for a
the thermal performance of the existing review of aesthetic impacts, construct-
,~vindow. The window frame would not ability, and cost, as well as testing and
Fig. 3. View of a steel window sill on the eas
be removed from the -,vndow opening monitoring of the thermal perform- elevation of the eleventh floor Note the sign
during repair. The sash would be re- ante.' These mock-ups were con- cant corrosion of tiro steel irame at the sill-u
moved, repaired, and reinstalled; all strutted using the same materials and jamb corner
THERMAL PERFORMANCE OF REPAIRED AND REPLACEMENT WINDOWS
midity, allowing calculation of the
change in enthalpy of the air as it en
tered and exited the chambers.z Formu-
;?>it, 4'i d las for air properties are found in the
ASHRAE Handbook: Fundamentals
S {y;. 2005.
This analysis method produces
repre- sentative heat flows for comparison;
however, the method has some error
because it does not account for the
dynamics of constantly changing bound-
ji ary conditions. Also, the data compar-
e 11 r' zsons produced anomalies when one
window heat flow indicates a thermal
loss or gain while the other window
_ indicates the opposite. The heat-flow
analysis compares only the loss at one
window to the simultaneous loss at the
other window and the gain at one win-
dow to the simultaneous gain at the
Fig. 4. Sil} of the replacement window. Fig. 5. Sill of the repaired window. other window, while ignoring the time
periods when the windows indicated
exterior conditions. Though relatively Roth window mock-ups were instru- opposing flows, which is limited to less
limited in scope and duration, this study mented with surface-temperature sen- than 10 percent of the data where flux
provided valuable real-world perfor- sors (thermocouples) at critical frame approaches equilibrium between gains
mance information, which was used to and glass locations where maximum and and losses. Overall, the effect of these
help guide the design team in the evalua- minimum surface temperatures are two minor simplifications on the overall
tion of potential treatment options. expected to occur, such as at the center comparative thermal analysis was con-
Additional analysis using computer of the glass, the horizontal meeting rail, sidered negligible.
simulation and other analytical tools and perimeter frame locations. Relative-
could be used to further develop perfor- humidity and air-temperature sensors Results
mance characteristics, such as evaluating were also installed within the air cavity The analytical review of the thermal
other exposures and other glazing op- between the storm glazing and the exte performance of the two mock-up win-
tions. rior window to evaluate condensation doves was a complex process with many
Setup and procedure. Surface tempera- potential within the storm cavity (Figs. 8 obvious variables, as well as several that
and 9). Ambient conditions were
tures, ambient conditions, and the heat recorded on the building's exterior and were not so obvious. The main vari-
gain and loss experienced through each interior, including the pressure differen- ables considered were heat gains and
window were measured in order to fully tiat between the interior and exterior losses due to conduction through the
evaluate and compare the thermal per- conditions, using relative-humidity and window frame and solar heat gain
formance of the two mock-ups. A sealed (radiation) through the glass, as these
chamber was installed on the interior temperature sensors and a digital prey were the most significant mechanisms
sure gauge. Data points for accessible for heat transfer through the window
face of each specimen window. The
locations were recorded on a laptop b[ The mock-up
chambers were insulated, and the inte- assembly. while inaccessible data points Y. p apparatus was
rior surface of each chamber (facing the (temperature and relative humidity designed to determine heat gain and
window) was covered with a reflective within th loss, as noted above. However, it was
white coating to minimize unwanted the storm of the rehabilitation not capable of isolating conductive heat
solar heat gain within the chamber. An Window) were recorded on stand alone data loggers. loss or gain from solar heat gain with-
air inlet was installed at the top of the oggers. out additional processing of the data. In
chamber, and an outlet with an electric Heat-flow calculations. Following data order to isolate conductive losses from
fan was installed at the bottom to venti- collection, raw temperature and humid- solar gain, the data indicating heat loss
late the chamber with a known quantity ity data were used to calculate the heat was isolated from the data indicating
of air. The air inlet and outlet tempera- loss or gain through each window. solar gain (i.e., daytime conditions with
tures were measured using thermistors Window heat-flow calculations were solar exposure) where possible. As most
and recorded on a data logger. The made by comparing the temperature of conductive heat-gain opportunities
change in temperature between the inlet the air entering the insulated chamber during the monitoring were during the
and outlet was used to calculate the heat to the temperature of the air leaving. warmer, sunny hours of the day, con-
gain or loss through the window, as The humidity ratio was calculated using ductive heat gain was unable to be
described below. interior temperature and relative hu- isolated from solar heat gain. Therefore,
q
APT BULLETIN. OLRNAL OF ?RESERViVION TECfINJ OLOGY / 10:9, 2009
repaired window. The repaired window
experienced approximately 15 percent
- - - more heat gain during the warm week.
Conductive and radiation heat loss
and gain. Conductive heat loss or gain
occurs through the frame of the win-
dow and was the primary mode of
I 7q: --'a~3 •>f s"";~ r•• I r •;,v ~ i ~ t'*`i energy loss through the window-6•ame
1 q~~ assembly. Radiation heat loss or gain
ir; ''`''i+'•~,' occurs through the glazing assembly
and was the primary mode of energy
ltd t ` d loss through the window. Conductive
x.
T , I and radiation heat losses and gains are
driven by temperature differential
across the window and are easiest to
_ _ measure during cold winter months,
L when the temperature differential acro!
the window is greatest. However, con-
ductive and radiation heat gains can
Fig. 6. Head of the replacement window, Fig. -1. Heed of the repaired window. Note the occur during summer months when
similar sight lines but missing articulations and exterior temperatures are greater than
accessories of the replacement window when the conditioned interior temperature.
ka viewed in conjunction with the repaired window. Since hot-weather conductive and radi
tion heat gains occur during the hottes
conductive performance is best urea- gains of the two windows was smaller part of the day when solar gain is peal,
Syr sured during nighttime conditions (Fig. 11). ing, it was not possible to separate sol:
dr.tring colder temperatures (i.e., night Calculations reveal that the peak and conductive gains through the win-
conditions with no solar exposure and solar heat gain for the repaired window dow, and the data therefore were not
' large thermal gradient across the mock- was only 10 to 25 percent greater than separated in the analysis,
up). Total heat gain and loss through that of the replacement window during The coldest exterior temperatures
• I the windows were then calculated the warmest week. A comparison with were observed during the first week of
('Table 1). the solar gains during the coldest week monitoring. Temperatures were not as
I illustrates the improved conductive heat cold as typical peak winter conditions.-
Solar heat gain. As expected for the gain and loss performance of the re- However, they provided adequate opp(
east facing windows, solar beat gain is paired window mock-up with respect to tunity to measure window performance
the largest source of heat gain for the the replacement window mock-up. Heat during cold weather. Heat-loss data wa
window system and spikes for both loss was calculated to be 4.0 kW for the isolated From the heat-gain data, and tl
i window mock-ups in the late morning replacement window and 1.5 kW for the differences in the daily peak heat loss f
when the windows are most exposed to repaired window during the warmest each of the windows were compared f(
sunlight. Ileat gain tapers off in the week. When daytime solar heating con- this cold week (rig. 12).
afternoon as direct solar exposure ditions (net heat gain for both windows) As expected, hear-loss peaks occur)
decreases due to indirect diffuse solar were considered, a heat gain of 150.6 in the very early morning hours, bcfur
exposure. There was a significant diffea•- kW for the replacement window and solar heat gain begins. The repaired
ence in peak heat gains between the two 169.3 kW for the repaired window was window experienced 15 to 35 percent
mock-up windows (Fig. 10). calculated during the coldest week. The less heat loss through the window tha
Calculations show that the daily peak net heat gainlloss for the week (exctud- the replacement window during the
solar heat gain for the repaired window
' ~,'ttl was approximately 3.5 to 40 percent ing conditions where. one window is coldest week. Heat loss was calculatc(
greater than for the replacement win- experiencing heat gain or loss and the to be 121 kW for the replacement wir
Clow during the coldest week of moni- other window is experiencing the oppo- dow and was 89.5 kW for the repaire
torin In addition, the net net heat gain site) for each window was found to he window during the week. Considerin)
also i . ncluded reductions heat gain due 146.6 kW gain for the replacement daytime solar heating conditions (net
to conductive heat loss through the wrndow and 167.7 kW gain for the heat gain for both windows) only, a h
window unit, which was greater for the
replacement window, as noted below. As Table 1. Heat Gain and Loss Totals for the Two Mock-up Windows over a Testing Peri
expected, when comparing the measured of Approximately 12 Weeks
performance during warmer weather, Specimen _ Net Hcat Loss (kW) Net Heat Gain (kW)
the difference between the peak heat Replacement Window 419.1 994.6
Repaired Window 11___n296.9 ` 1264.8
S
THERMAL PEfRFORMANC.F OF REPAIRED AND REPLACEMENT WINDOWS
gain of 33,1 kW for the replacement
window and 62.1 kW for the repaired w
window was calculated for the coldest
week, a significantly smaller heat flux "
than with conductive losses alone. The
net heat gain or loss for the week for
each window (excluding conditions
r
where the mock-ups are experiencing t=•"
opposing heat flows) was found to be
57.9 kW loss for the replacement win Xi
dow and 27.4 kW loss for the repaired,; him
window. The repaired window expcri-
enced nearly 70 percent less heat loss ~
during the cold week. It is important to
understand that as the exterior tempera
ture drops, the heat loss (and difference
in heat loss) for the windows will in- /
crease and that this difference in perfor-
mance would be expected to become
more pronounced during typical peak
wintertime conditions. Thus, the con- Fig. 8. Interior view of the repaired window prior Fig. 9. View of the replacement-window mock-
ductive gains during hot weather are to installation of the interior chamber. Note the up prior to the installation of the interior cham-
inversely proportional to the losses. thermal and relative-humidity sensors are ber. The temperature sensors are indicated with
indicated with dots. Some sensors are installed dots.
on the repaired window, and others are installed
Discussion on the storm sash, resulting in more sensor
locations than on the replacement window.
The purpose of this study was to com-
pare differences in thermal performance
of the two window mock-up specimens its solar heat-gain performance. For tions existing on the building, a limita-
under identical operating conditions. example, a window with a wide frame tion of. the study scope and budget.
After analyzing the monitoring data, and numerous small lights separated by During the winter the low elevation
several trends became apparent, and mullions and muntins has less glazing of the sun at midday causes it to shine
they may impact the selection of the area available to capture solar energy. through south-facing windows, in addi-
window-treatment option for the repair By contrast, a window in the same tion to east-facing windows in the morn-
and replacement program. To better rough opening with a thin frame and ing and west-facing windows in the
understand the results of the testing and one large light will have a greater pro- afternoon. The resulting solar gains can
monitoring, it is essential to have a portion of glass-to-frame area and will help reduce heating costs during the
general understanding of window per- allow more sunlight into the building winter. In the summer, when the sun is
formance. The pertinent performance interior. Both mock-up windows had much higher at midday, the angle of
factors and observations are discussed similar construction, with large unob- incidence of the solar radiation is much
below. structed glazing areas and narrow metal sharper. Consequently, more solar radia-
Solar heat-gain performance. The perimeter frames. Therefore, the differ- tion reflects off of south-facing windows
most significant component of heat ence in performance due to frame con- than is transmitted to the interior. Over-
gain is solar heat gain through the figuration and design was small. heating in summer therefore tends to
window glazing. Several factors affect Window orientation. Placement and occur more frequently at unshaded west-
the solar heat-gain performance of facing windows and, to a lesser extent,
orientation of. the window with respect
windows, including the geometry and to compass direction is the root factor at east windows than at windows that
configuration of the window unit (such affecting solar gains. West- and south- face directly south. The desired amount
as the amount of clear window open- facing windows experience the most of summer and winter solar heat gain is
ing), its placement and orientation, the significant gains, but some gain is possi- determined by the design of the mechan-
type of glazing and coatings used, and ble in all directions from diffuse sky ical system for the building. If the pri-
the amount of interior and exterior radiation nary load on the building is heating
shading. Optimizing these characteris- . The majority of the windows during winter, then optimizing solar heat
tics to improve thermal performance at the Lafayette Building are on the
can significantly improve building east, west, and north elevations, with rate gain during winter months should nts
verY few windows facing due south. wind p e rformance requirements.
operating costs and occupant comfort.
Both mock-up windows faced east, and Glazing coatings. The number and type
t n and configuration. The configura- therefore the recorded data represents of glazing layers and coatings will also
Li on and
=eornetry of a window affects
affect one of the four possible orienta the thermal performance Of the
APT BULLETIN, J(JuRNAL 01 PRFSLHVAT,0N TFCHH01_OGY i 40 2, 701J9
Heat Gain 10 Minute Averages -ColderWeather of the exterior glass and avoids a change
,4n 230 in appearance.
390 --1 - ---1-~ 220
320 - T !t0
300 200 Changes in the position and type of
NO - - - ; o low-E coating on the repaired window
.^40 - - - 170
20 40
may also affect the condensation tests-
2 t 140
;P° = - - 130 a lance of the glazing. Although not
120
specifically discussed in this paper, it is
au important to note here that the most
- - - susceptible location for condensation is
a 2C0 5D
1 1 n -J F _ ?
5 :;0 30 the interior face of the exterior glass in
6 20 2
B° 10 o the repaired window. Adding a low-E
20 - - A f_ -I- - - i_ Ti --1- 2g coating with a lower SHGC: to the inte-
40 __T - - - - _ _ _ --1 - _ 70
2a0 ` r -I- ~0 riot storm would likely improve conden
;'0 .;g sation resistance of the exterior glass, as
Pau s would providing a new outer pane of
300 '1111a'i C.'fC:P14-J•t to • - _100
- t1U glass with a low-E coating with a mid-
3.1) 1 - , -12° b
~o w
- ''iw Uoo l'C ~p U,u w FOa "".P m to low SHGC.
00, 1tisp
6,,~C~ 0•
M~ 10PA~'P40"' lpe" 4 .yv~~,40 0" 'Ito
Ti- Other factors. Covering the window
l rr..~waim7 -rze~wnaa.,+;mnrv -oeuac. openings with draperies an curtains of
other shading devices will also reduce
Fig. 10. Heat gain during the coldest week. the amount of solar heat gain transmit-
ted to the building interior. Keeping th<
window coverings open to admit as
windows. For example, a double- of 0.,83, meaning that 83 percent of the much solar gain as possible sunny
glazed, insulated glass unit consisting of solar heat gain is absorbed and transmit days during the winter will improve
ordinary clear glass reduces solar gain ted through the glazing. This difference Performance. This testing did not in-
by approximately 10 percent compared in SHGC: of the two glazing systenis had elude interior window treatments, so
to asingle-glazed window with the a substantial effect on the solar heat their impact on thermal performance
was not quantified; however, it seems
same glazing area. Since both mock-up gain measured in the two mock-ups. likely that they would impact both
windows included two layers of glazing, Adjusting the heat gains to offset the mock-ups similarly.
the glazing is not a differentiating fac- heating load during the winter or to
tor. The addition of glazing coatings minimize the cooling load in the summer Heat-loss performance. Several pro-
have a more significant impact. Low- can be achieved in either window treat- cesses influence the rates of non-solar
F emittance (low-F.) coatings are micro- merit option, in part by the selection of heat gain or loss through window con-
scopically thin metal or metallic-oxide the glazing coating. ponents. These processes follow a basi
layers or a film coating deposited on the If building-load calculations reveal law of thermodynamics: heat energy
surface of the glass (typically on a that summer solar heat gain must be tends to move from warmer areas to
surface within the glazing cavity) to minimized, the solar heat-gain perfor- colder areas. In Washington, D.C., the
reduce the U-factor (the heat transfer malice of the repaired window can be primary flow of heat is from interior
through the glazing) by suppressing improved with some modification. "fhe spaces to the building exterior during
radiation heat flow. least invasive modification irivolves fall, winter, and early spring. The difft
The difference in solar heat gain reducing the low-E coating on the storm ential temperature tends to be lower
between the repaired window and the window to reduce the solar heat gain. A during summer months, when heat fk
replacement window is likely due to the second option for reducing solar gain is from the hot exterior to the cooler,
t location and type of glazing coating on through the repaired window would be conditioned interior. The principal he,
the two window mock-ups, in addition to add a low-E coating on the inside transfer processes in windows are rad
to the configuration of the glass. The surface of the exterior glass. The addi- tion, conduction, and convection. In
replacement window had a low-E coat- tion of a low-E coating would require addition, excessive air leakage can
ing wirh a relatively low solar heat-gain replacement of the original glass with a contribute to the overall heat loss.
coefficient (SHGC) in the otter pane of pane of laminated low-E glass (the During colder temperatures, ltear is
glass. The low-E coating has a SI-IG(, of thermal stress placed upon the original absorbed by the inside pane of a doul
0.41, meaning that 41 percent of the float glass of the existing windows by glazed window, moves to the cooler
solar heat gain is absorbed and transmit- adding a coating would likely result in outside pane, and is released to the tai
red through the glazing, while the rest is glass fracture). Even though there is a doors. Not only does this heat loss tai!
reflected back to the exterior. The re- diminished benefir to adding low-E place through the glazing by radiatioi
paired window has a low-E coating with coating to the interior storm window also occurs across the spacer materia!
a higher SHGC on the repaired window rather than to the exterior glass, this the insulating glass unit, which scparr
storm. This low-E coating has a SHGC procedure does not require replacement the two glazing layers at their edges
=,~iv1AL r E,=~~'~?iv1ANCE CF REFA FiLii AND REFLACEMLNI WINDOWS
the replacement window only); through Heat Gain 10 Minute Averages - Warmer Weather
the frame of the window by conduction; 360 230
40 - - - - 440
through the movement of air in the 3320 21
00 200
280
space between the two glazing layers by 260 - Iao
convection (more pronounced iii the 2~ ;60
larger air space of the repaired window); _ IV
140
and between the moveable or operable
20 110 -A 100 I 100
frame components by air leakage. Con- o - d0
7o vective losses are typically negligible 0 - - 6°
r .20 _ _ 7-- - - _ _ 40
with respect to other losses and were not .40 30
addressed in this study. ^I„ 20
120 10
Radiative losses and gains. Typically,'
_ - - : o
radiation losses through the window 200 - 1 - ' -50
ago - ~ -t- -6o
glass represent about two thirds of the :2~0 - - :B0
l,--
total heat loss in a standard window. =300 - I-i0°
-320 _ , ox to rix tAfl[rencc5 r9a~r Peakc .110
Because ordinary glass readily emits -340
heat to colder surfaces (i.e., has a high ~le,
6n d1 ;011,40 Oj
emissivity), radiation losses can be Time
reduced by lowering the emissivity of [--ahwWnMnG,rJA -R<peuW..ew<3r4dn -oen-.ra~mj
the glass by installing low-E films.
Placement of the low-E coating in the
Fig. 11. Heat gain during the warmest week,
pane of glass experiencing the greatest
temperature differential will have the
greatest effect on radiation loss through metal-to-metal contact at the sash meet- perimeter of the replacement window
the window. Review of data from this ing rail), thus reducing heat transfer. can be improved by installing spray
study indicates the greatest temperature Air leakage. Window air leakage is a foam in the cavity when the original
differential is across the outer pane of significant contributor to energy costs window is removed. Foam will improve
glass in the repaired window during during both heating and cooling seasons thermal performance of the window
cold weather. Therefore, placement of for most buildings. Air leakage also frame, as well as limit air infiltration.
the low-E coating in the exterior glass affects occupant comfort. Most of the Similar improvements can be made at
will have the greatest impact on radia- air leakage through operable windows the repaired window by installing spray-
tive loss. occurs between the window's sash and foam insulation in weight pockets and
Conductive losses and gains. Conduc- frame or at the meeting rails of a sliding the window perimeter.
tion losses in windows occur primarily sash, as on the replacement window.
through the edges and frames of the Bigger windows tend to leak less air per Conclusions
units and are often expressed in terms unit area than smaller ones. In poorly This analysis shows significant differ-
of U-value, the overall measurement of constructed fixed windows, air leakage ences in thermal behavior between the
conductive heat transfer through the also occurs between the insulated glass repaired-window and the replacement-
window. The thermal-conductance unit and the frame. Windows with the window mock-ups. The repaired win-
characteristics or resistance to heat lowest leakage rates, regardless of type, dow experienced more solar heat gain
transfer, i.e. R-value of the aluminum tend to be fixed windows. Although the during morning and early afternoon
frame of the replacement window and repaired window had a fixed sash, it hours than the replacement window. In
the steel frame of the repaired window, originally was an operable window. turn, the replacement window experi-
also has an effect. Steel is less conduc- The condition of the perimeter con- enced more heat loss through the glass
tive than aluminum and has a higher struction also affects the air-infiltration and frame during evening and early
R-value, thereby reducing the overall resistance of the replacement window. morning hours. Because solar heat gain
U-value of the window. Data from this Air leakage around the replacement can be manipulated (e.g., through the
study indicate that the temperature window can be a significant problem if use of low-E coatings) but heat loss
drop across the frame was different for the windows are carelessly installed in through the frame cannot, the repaired
the two windows. The thermal break the rough opening. Air infiltration at window provides superior heat-loss
installed in the aluminum-framed re- and through the perimeter (frame) of the performance and significantly greater
placement window helps to reduce the window mock-ups was evident during potential for optimizing glazing and
heat loss across the frame; however, the air-infiltration testing, particularly heat-gain performance (particularly for
interior storm window of the repaired around the replacement window. Air the different building exposures) than
window better isolated the steel win- leakage is likely increased by construe- the replacement window. Solar heat
dow frame of the repaired window tion activities to remove the original gains for both windows tended to more
from the building interior (e.g., no window and install the new replacement than offset the heat loss through the
window. Air infiltration around the frame and glazing, a conclusion that
19
API BULLETIN: JOURNAL OF PRESERVATION TECHNOLOGY / 40:2, 2009
Heat Loss 10 Minute Averages -Colder Weather Acknowledgements
,o - r 1W The authors would like to acknowledge our
rC colleagues at Harboe Architects and Simpson
-20
Gumpertz & Heger, Inc., as well as the effort of
-30
ao ' t ,In both D.YI]k4 and theU-S. General Services
so too Administration in the execution of this proiact.
_7~ BO 2
-eo I - - - i° Notes
-100 - so z 1. Both window mock-ups were instrumented
-Ito r ao = with temperature sensors and relative-humidi
120 - 30 3 and air-temperature sensors installed within t.
20 9. o air cavity between the storm glazing and the
-140
so l 6° repaired window. Surface-temperature sensor
Is° - - I I Io were self-adhesive E-type therinocouplcs that
170 Zo uv~ere connected to Veriteq thermocouple log-
V7
eo 30 gers. Each Veritrd logger monitored the rem-
I S ,G l0 35°,S Dil/ci cotes
ao erarore of four thermocouples. Thrrmocou-
6o pies were installed at the following locations
r . eo° zo° °o° rP o° o, rP d° o° o° 6 FP cP
the 40 ,,~~n~~ysY ` 0,~S+ti1°~1., tio~,~,~'°oa1.9110tA, 'p1ry°ch,,0`~ia ym~,j1.` a° ~``ty6° and the rreplacement the frame; window!
Time frame (center), horizontal meeting rail (center
Ne.~~de~~v -fepalnMnCOkeSWJ -ooilato,yv left jamb (upper left), center of glass (lower
light), Thermocouples were also installed on
the cavity face of the storm window at the
Fig. 12. Heat loss during the coldest week. following locations: windowsill frame (center
.I frame of window head (center), left jamb
t (upper left), and center of glass (correspondir
with lower light). Two Dickson D-200 data
may not be applicable to north-facing considerations, the reader should refer loggers were installed within the cavity space
windows and requires further analysis. to the Secretary of the Interior's Preser_ between the repaired window and the storm.
1 The D-200 data loggers recorded air tempera
As solar heat gains vary throughout the nation Brief 13: The Repair and Ther- ture and relative humidity at the lower left ar
year, careful consideration of the com- mal Upgrading of Historic Steel Win- upper right corners of the cavity between the
prehensive building heating and cooling doves, by the U.S. Department of the storm and the repaired window. To record an
bient conditions on the building exterior and
y' loads and mechanical-system require- Interior (available at http:Hw-ww.nps interior, two Vaisala IIMP44 probes and an
ments are needed to optimize the gains gov/history/hps/tps/briefs/brief13.htm)• Omega data-logging pressure box were in-
and losses through the windows, maxi- Any rehabilitation project considering stalled. The pressure box measured the differ
mizing gains when needed while mini- similar window programs should in- ence in interior and exterior ambient pressurt
mizing losses throughout the day and elude careful identification and evalua- The window mock-ups were constructed
site to allow review of numerous features,
the changing seasons. Additional assess- tion of these often competing factors including but not limited to appearance, con-
ment - including the evaluation of conducted in concert with technical structability, cost, and impact to building
other exposures (e.g., north and south), analysis performed by competent profes- tenants, as well as performance. In addition t
thermal-performance monitoring, testing
glazing-coating options, and other sionals so that appropriate options can included air-infiltration testing in accordance
factors - is required to fully develop be evaluated and an optimal solution with American Society for Testing of Matcria
the options and will likely impact over- selected. (ASI'M) E783: Standard'lest Method for tie
all design decisions. This analysis can be Measurement of Air Leakage through Installt
ROBERT SCORE is a architect at Exterior Windows and Doors, as well as wat
achieved with additional mock-ups, project arc penetration testing in accordance with ASf1v
Harboe Architects in Chicago, Illinois, and
careful application of computer simula Ell inati Standard Test Method for nstal
specializes in the restoration of commercial and m urination of Water Penetration of Installed
ed
tion, or a combination of both. cultural properties. He was previously on the
The overall result of this study does Historic Resources Committee of the Chicago Exterior Windows, Skylights, Doors, and
Chapter the AIA and a director of APT and Curtain Walls, Uniform or Cyclic Static A
not diminish the fact that thermal per- o Pressure Differen nce. 'I hough not discussed in
is currently helping to found the Western Great
formance is only a portion of the overall this paper, the repaired-window mock-up
decision process. While the repaired Lakes Chapter of APT. allowed roughly JO percent less air-iufiltratic
window offers superior thermal perfor- BRADFORD S. CARPENTER is a staff engi- leakage than the replacement window, likely
neer in the Washington, D.C., office of SGH. due to the operable sash of the repaired win-
mance, it will also conserve original dow being fixed and sealed shut with sealant
While at SGH, lit! has investigated, designed,
building fabric and minimize material and rehabilitated building envelopes on both and paint coatings, while the replacement
waste by maximizing efficient use of pre- historic and modern structures. He previously window, though fixed shut, relied upon gast.
worked at Newport News Shipbuilding in seals. The repaired-window mock-up had
cxistin . Careful evalu-
g embodied energy comparable water-penetration resistance to
ation of historical and architectural Newport News, Virginia, and for the Architect replacement window.
of the Capitol in Washington, D.C. He can be
significance, as well as the physical reached at BSCarpentcr@sgh-com. 2. Since monitoring did not include barometl
condition of the windows, must also be pressure measurements, calculations include
considered, in addition to future mainte- constant barometric pressure of 101.325 kPI
nance and operation needs. For an in- This assumption carries through calculations
depth discussion of these and outer for both windows and will cancel ont as tht:
e
11 d" the
windows are compared. The humidity ratio in
,,'kg is calculated using interior temperature
„nd relative humidity. The entering- and t r. _ 1 f it
leaving-air enthalpy is calculated in kjikg using the humidity ratio and respective air tempera-
t.:res. The mass of the air flow is calculated
using the specific volume of the discharge air in
inVkg. The data-logger time interval is five
minutes, and all kJ calculations are converted
to watts by multiplying by 1,000 and dividing
h•y 300 seconds. American Society of Heating,
Refrigerating and Air-Conditioning Engineers
I landbook, vol. 1, Fundamentals (Atlanta: ~y
ASHRAF, 2005).?
3. The coldest temperatures recorded were
approximately 32°F (0°C observed over a
span of scvcral hours throughout the first week
of data recording. 'The average peak wintertime I V , r~
Temperatures are typically found by using the Permalac clear coat lacquer
(xterior heating design temperature for Wash-
o 11,t~ protects American heroes
iugton, D.C., which can be found in Table D-1 , Lewis & Clark from the
of the ASHRAF Standard 90.1-2004. The floodwaters of the
exterior heating design temperature of 15"F Mississippi. This installation,
cI n u a) corresponds to he occurrence, which nu- I '"The Captains" Return,'" was
st -eawd by Harry Weber,
means that actual exterior temperatures exceed cnnwneil American sculptor.
this design temperature for all but 0.4 percent
of the year, m: about 1.4 days, during a typical k 'c
year.
i+lr
Bibliography + ~
y i A
American Society of Ileafing, Refrigerating end .
Air-Conditioning Engineers. ASHRAF v s, • t
Standard 55-2004, Thermal Environmental
Conditions for Human Occupancy. Atlanta:
ASHRAF, 2004.
American Society of Heating, Refrigerating aw
Air-Conditioning Engineers. ASHRAF•, Permalac Clear coat lacquers w-.-ae six to ten
Standard 90.1-2004, Energy Standard for
Buildings Except Low-Rise Residential y'(lilrs ; rOtCClI[7n frum Uhl ,31tata:..dosort sizzle. _arctic
Buildings. Atlanta: ASHRAE, 2004.
blast...v.,ind-bowie sand.. and salt spray. That's why
Park, Sharon C. Preservation Brief No. 13: The
Repair and Tbermal Upgrading of Historic sculptors and conservationists from coast to coast insist
Steel Windows. Washington, D.C.: National -
Park Service, 1954. on Permalac®. Especially after one of the short-lived
competitive lacquers has required re-coating.
Permalac is avaifable in matte or satin finish.
Plus there is the new Permalac EF with just 170 grams
per liter of VOCs and Permalac 2K, a two-part coating
system for highly aqueous environments such as fountains.
For more information or to order, contact us at
www.permalac.com or call 215-729-4400.
- a
Permalac is formulated by
f~:Hi1G~,flC `
Peacoc-k
LABORAI ORI ES, INC
1901 S. 54th St.
!''liiladeli`~~a, NA "91 u3
I