Conventional Roof Assemblies: An Update to a Six-Year Monitoring Study

Conventional Roof Assemblies:
An Update to a Six-Year
Monitoring Study
Jun Tatara;
Christopher Marleau;
Lorne Ricketts, PEng;
and Graham Finch, PEng
RDH Building Science Inc.
433 Still Creek Drive, Vancouver, BC, Canada
604-873-1181 • lricketts@rdh.com
IIBEC 2020 Virtual International Conve ntion & Trade Show | June 12-14, 2020 R icke tts et al. | 233
Lorne Ricketts is a building science specialist concentrating on new buildings, forensic
investigations, and research work. His broad experience includes enclosure and façade
design consulting, field review, building monitoring and testing programs, energy assessments,
and product testing and development. Ricketts takes an active role in industry education
and has produced numerous technical guidance and research documents in addition to
having spoken at conferences and industry events across North America.
Nonpresenting Authors: Jun Tatara; Christopher Marleau; and Graham Finch, PEng
234 | Ricke tts et al. IIBEC 2020 Virtual International Conve ntion & Trade Show | June 12-14, 2020
ABSTRACT
SPEAKER
A long-term field monitoring study has measured the impacts of membrane color and insulation strategy on the
in-situ performance of conventional roof assemblies. The same roof membrane cap sheet type with three different surface
colors (white, gray, and black) was installed over three different conventional insulation strategies with approximately
the same R-value, creating a total of nine unique roof assemblies on the same building. Sensors were then
installed to monitor key performance indicators for the roofs, including temperature at key layers of the roofs, relative
humidity (RH) within the assemblies, and solar reflectance of the roof membranes.
This paper presents the results of this monitoring work after six years of study. In particular, updates and further
analysis are presented with respect to moisture movement and accumulation within the assembles, insulation
performance, and long-term reflectivity, and soiling of the roofing membranes. Exploratory openings are also to be
performed to confirm results of the monitoring and collect samples for laboratory testing of the roofing materials post
field exposure. The various measurements are analyzed and synthesized to allow for discussion of advantages and disadvantages
of the different membrane colors and insulation arrangements.
IIBEC 2020 Virtual International Conve ntion & Trade Show | June 12-14, 2020 R icke tts et al. | 235
BACKGROUND
Conventional roof assemblies (low-slope roof
assemblies in which the waterproof membrane is
located above the insulation) constitute the majority of
low-slope roof assemblies in North America. The design
of these roof assemblies can have significant impact
on the thermal performance of the roof assemblies
and, consequently, on building energy consumption,
occupant comfort, membrane durability, and assembly
service life. This conventional roof research and field
monitoring study analyzes the performance of nine different
roof assemblies with different membrane colors
(white, gray, and black) and insulation arrangements
(polyisocyanurate, stone wool, and hybrid) for a large
roof with relatively few obstructions located on an
industrial building in the Lower Mainland of British
Columbia. Various sensors were installed within each
of the different roof assemblies to measure indicators
of their performance, including solar reflectance,
material temperatures, RH, and heat flux. This report
provides a six-year update and data analysis from the
beginning of September 2012 to the end of November
2018, and it is an update to previous analysis published
by Finch, 2010; Dell and Finch, 2013; and
Finch, Dell, and Ricketts, 2014.
The solar properties of the sheets as specified by
the manufacturer are provided in Table 1. The three
insulation arrangements are polyisocyanurate (polyiso),
stone wool, and hybrid (stone wool on top of polyiso) as
shown in Figure 1, Figure 2, and Figure 3, respectively.
The combined roof assemblies cover a 40- x 40-ft. area
(1,600 ft²). Figure 4 shows the arrangement of the roof
assemblies on the study building.
FIELD MONITORING RESULTS
The data presented in this study cover the period
from installation in October 2012 to the end of
November 2018. In some cases, data from some of
the sensors are not available due to intermittent malfunctioning
of the installed monitoring equipment, but
these instances are not thought to significantly affect
the findings, and they are discussed further where
appropriate.
The exterior conditions were obtained from the
weather station at Agassiz Airport (Environment
Canada, 2018) and the interior conditions were
obtained from on-site monitoring equipment. Both
the exterior and interior conditions provided in this
Conventional Roof Assemblies:
An Update to a Six-Year
Monitoring Study
Figure 1 – Polyiso roof assembly; two layers of polyiso insulation. Total
nominal R-value (hr·ft²·°F/Btu) = R-21.0 (top layer = R-12.0, bottom layer
= R-9.0)
Figure 2 – Stone wool roof assembly; two layers of stone wool
insulation. Total nominal R-value (hr·ft²·°F/Btu) = R-21.9 (top layer =
R-9.5, bottom layer = R-12.4)
Figure 3 – Hybrid roof assembly; base layer of polyiso insulation and
top layer of stone wool insulation. Total nominal R-value (hr·ft²·°F/Btu)
= R-21.5 (top layer = R-9.5, bottom layer = R-12.0)
Cap Sheet Solar Reflective Solar Thermal Emittance
Color Index (SRI) Reflectance (Infrared)
White 70 0.582 0.91
Gray 9 0.138 0.85
Black -4 0.040 0.85
Table 1 – Roof membrane cap sheet properties specified by manufacturer.
section were used for analysis throughout this report. The average exterior and interior conditions for each month are provided in Figure 5.
EFFECTS OF MEMBRANE CAP SHEET COLOR
Reflected solar radiation from the roof cap sheets was measured using solar radiation sensors (pyranometers) mounted approximately 1 m (3.2 ft.) above the roof surface and pointed downwards.
The total horizontal solar radiation from the sun was measured as well as the reflected solar radiation from the white (at a high and a low point on the roof surface) and gray roofs to determine their in-service solar reflectance. This testing was not performed in accordance with ASTM reflectance measurements typically used for rating of roofing membranes. The technique used in this study is intended to provide a relative measure of reflectance to allow for comparison of the in-service performance of the roofs in the study.
Figure 6 compares the average wintertime solar reflectance of the roof membranes between December 21 and March 20 from 2013 to 2018 (including manufacturer’s ratings). The comparison was made for wintertime due to data loss that occurred in the spring, summer, and fall of 2014 and the summer of 2017. Note that the measurement of the reflectance of the white membrane at a low point on the roof is less than that of the membrane at a high point, which is likely due to the low point being more prone to collecting dirt and becoming soiled faster than the high point.
The difference between the field-measured values and manufacturer’s rated value is expected to be due to the difference in measuring techniques. While the measured reflectance of the white membrane in winter 2013 is similar to the manufacturer’s rated reflectance, the measured gray membrane reflectance is significantly higher than the rated value. This is potentially because the sensor on the gray roof is located relatively close to the white roof and may be measuring some reflected radiation from the white roof or
236 | RickeICKEtts et al. IIBEC 2020 Virtual International ConveVEntion & Trade Show | June 12-14, 2020
Figure 4 – Layout of the nine test roofs on the study building (test area in red dash).
Figure 5 – Graph of monthly average temperature and dew point temperature during the monitoring period from October 2012 to December 2018.
Figure 6 – Average reflectance of membrane between December 21 and March 20 over the six-year monitoring period. Note when the sensors were cleaned (March 2015 and December 2016), indicated with red dashed lines.
other adjacent surfaces. It’s also possible that the gray membrane is simply more reflective than its rating indicates.
Results after three years of monitoring implied a declining trend in the white roof membrane reflectance, assumed to be caused by soiling and weathering of the membrane over time. However, an overall spike in winter 2016 shows that reflectance levels increased, in general, slightly above or similar to the initial values measured for winter 2013. This is almost certainly a result of cleaning of the monitoring equipment, which occurred in March 2015 and December 2016. Therefore, the trend found in the first three years which indicated soiling of the membrane, was likely caused by soiling of the sensor rather than soiling and weathering of the membrane. Results from the continued monitoring period indicate that the reflectance of the white membrane at the higher areas has not reduced significantly over the six-year service life.
Despite uncertainty due to cleaning of the sensors, it is apparent that the white membrane in the lower portion of the roof does become more soiled than at the higher area, and this results in a measurable difference in reflectance. Portable instruments are available that can be taken to the roof to get accurate measurements (Smith, Liu and Paroli, 1998); however, a sample of each membrane was instead removed from the roof to compare with the same type and color of membrane kept from 2012 and stored in a controlled laboratory setting.
It is visually evident in Figure 7 that the aged membranes have darkened over time, with the exception of the black membrane. Further laboratory testing of aged membranes is scheduled, and results will be discussed is a separate paper. Interestingly, while changes in reflectance for the white membrane are often attributed to soiling, visual observation seems to indicate that the most significant contributor to surface reflectance of the membranes is degranulation.
Periods of snow accumulation were also found to slightly affect the measured results; however, the average yearly accumulation on the ground between 2013 and 2019 was only 18.5 cm (7.3 in.). During periods when the roof was covered in snow, the overall reflectance was found to increase for all cases. In addition, the insulative properties of the snow were shown to maintain the cap sheet temperature at approximately 0°C (32°F).
A total of nine sensors were installed to measure the roof membrane cap sheet temperature—one for each of the different roof assemblies. The daily maximum and minimum roof membrane cap sheet temperatures were determined using hourly data obtained from the sensors. The plot provided in Figure 8 shows monthly average membrane temperature by membrane color, and the dots indicate the maximum and the minimum temperature that the
IIBEC 2020 Virtual International ConveVEntion & Trade Show | June 12-14, 2020 R RickeICKEtts et al. | 237
Figure 8 – Graph of monthly average membrane temperatures and maximum/minimum membrane temperature for each month by membrane color.
Figure 7 – Roof membrane samples comparing original 2012 (bottom) vs. aged 2019 (top) for black, gray, and white membrane colors.
membrane experienced each month. The values provided for each membrane color are the average of the values for all of the different insulation arrangements (i.e., “white” is the average of polyiso only, stone wool only, and hybrid insulation with white roof membrane cap sheet).
Figure 8 reinforces the large impact that the roof membrane color has on roof membrane temperatures over the monitoring period. This impact is most significant during the summer months when roof membrane temperatures reach their annual maximums. For example, in July 2014, the maximum surface temperature of the black membrane was observed to be 51.2°C (92.2°F) above the maximum ambient temperature, while the maximum temperatures of the gray and white roofs were 42.3°C and 27.7°C (76.1°F and 49.9°F) higher than ambient temperature, respectively. During colder ambient temperatures with less incident solar radiation, the difference in temperature between roof membranes is less significant. In December 2014, for example, the minimum surface temperatures for the black, gray, and white membranes were 5.5°C, 5.1°C, and 5.2°C (~9°F) colder than the minimum ambient air temperature, respectively.
Measurement data also show that membrane color has a more significant impact during the day, and that nighttime temperatures are relatively similar for each of the three membrane colors. This finding is consistent with the relatively significant difference in the solar reflectance but similar emittance for the three membrane colors (see Table 1).
Figure 9 shows the inward and outward heat flux (flow) through the roofs for each of the different roof membrane colors. This figure shows a slight trend of increasing inward heat flow for all membrane colors over the course of the six-year monitoring period. While the cause of this trend is not immediately apparent based on roof membrane weathering and degradation of solar reflectance, it is theorized that this trend is likely primarily due to changes in the thermal performance of the polyiso insulation as it ages. This chart also indicates that
238 | RickeICKEtts et al. IIBEC 2020 Virtual International ConveVEntion & Trade Show | June 12-14, 2020
Figure 9 – Graph of monthly average daily energy transfer by roof membrane color.
Figure 10 – Annual HDD from 2013 to 2018, based on membrane cap sheet color.
Figure 11 – Annual CDD from 2013 to 2018, based on membrane cap sheet color.
typically there is more heat flow inward through the black roofs than through the other assemblies due to the more extreme temperatures experienced on the surface of these roofs.
Another method for assessing the potential energy implications of different roof membrane colors is to use roof surface temperatures to calculate cooling degree days (CDD) and heating degree days (HDD), similar to the values calculated using ambient exterior air temperatures. HDD and CDD provided in this report are calculated with a base temperature of 18°C and 10°C (64°F and 50°F), respectively. The comparisons were made between degree days calculated using ambient temperature and roof surface temperature. The roof surface temperatures used to determine degree days for each membrane color are averaged for the different insulation arrangements with the same roof membrane colors. Calculating HDD and CDD using the roof surface (sol-air) temperature instead of ambient (exterior air) temperature allows for a more accurate indication of heat flows through the assemblies. Note that CDD are presented as negative values in this report.
Figure 10 shows reduced HDD for the black roof membrane due to increased solar heat gain when compared to the gray or the white roof membrane. Similarly, the lower reflectance of the black roof membrane results in significantly increased CDD (heat flow into roof assembly) as shown in Figure 11. This finding clearly illustrates that energy modeling of buildings based on ambient air temperatures with a lack of accounting for solar absorption on surfaces such as roofs can create significant inaccuracy in the result. In particular, the balance between heating and cooling demand for a building can be significantly altered by roof color.
EFFECTS OF INSULATION ARRANGEMENT
This section evaluates the impact of insulation arrangements on membrane temperature, interior metal deck temperature, and heat flow through the roof assemblies.
The thermal performance of the three insulation arrangements varies with temperature. Polyiso insulation in particular exhibits strongly temperature-dependent conductivity, generally leading to reduced thermal resistance at lower temperatures (Dell and Finch, 2013). Additionally, aging of polyiso has also been shown to reduce its thermal resistance as blowing agents are off-gassed from the cells of the foam plastic insulation (Finch, 2010; Dell and Finch, 2013; BSC, 2013). It should be noted that the polyiso insulation used in these roofs was manufactured before more recent advancements in the blowing agents used to manufacture these products, and thus more modern polyiso products are likely to exhibit a different response.
The plot provided in Figure 12 shows monthly average membrane temperature by insulation arrangements, and the dots indicate the maximum and the minimum temperature that the membrane experienced each month. The values provided for each insulation arrangement are the average of the different roof membrane colors for the similar insulation arrangements. (i.e., “ISO” is the average of polyiso-only roof assemblies with black, gray, and white
IIBEC 2020 Virtual International ConveVEntion & Trade Show | June 12-14, 2020 R RickeICKEtts et al. | 239
The roof surface
temperatures used to
determine degree days for
each membrane color are averaged for the different insulation arrangements
with the same roof
membrane colors.
Figure 12 – Graph of monthly average membrane temperatures and maximum/minimum membrane temperature for each month by insulation arrangement. Note that the ISO average line is positioned directly behind the red ISO-SW average.
roof membrane cap sheet).
While the impact of insulation arrangement on the roof membrane temperature is less significant than the membrane color, insulation arrangement does influence the roof membrane temperature, and the most significant impact is observed in the maximum monthly temperatures. Generally, the hybrid insulation arrangement had the lowest maximum temperatures throughout the year, followed by stone wool. The polyiso-insulated roofs have the highest maximum membrane temperatures. The minimum membrane temperatures are similar for the different roof types, but typically the polyiso-insulated roofs also experience the coldest temperatures.
Daily roof membrane temperatures also show that the roof membrane with polyiso insulation experienced both the highest and the lowest temperatures over the course of a day. The polyiso roof assemblies experienced more extreme roof membrane temperatures as compared to the hybrid roof and the stone wool roofs—likely due to the combination of a difference in thermal mass (heat capacity) of the insulation types, latent energy transfer within the insulation, and the influence of temperature-dependent R-values of the insulation. These reduced maximum and minimum temperatures in the hybrid and the stone wool roofs are consistent with previous findings (Finch, Dell, and Ricketts, 2014) and will likely have a positive impact on the durability and service life of the roof membrane and insulation itself (the rate of deterioration of asphalts is temperature-dependent; lowering the surface temperature of the membrane will extend its performance life).
To further assess the impact of insulation arrangement on the heat flow through the assemblies, Figure 13 plots the inward and outward heat flow through the different insulation arrangements, averaged for each month during the monitoring period. The figure shows a slight trend of increasing inward heat flow for both the polyiso and the hybrid insulation arrangements over the course of the monitoring
240 | RickeICKEtts et al. IIBEC 2020 Virtual International ConveVEntion & Trade Show | June 12-14, 2020
Figure 13 – Graph of monthly average daily energy transfer by insulation arrangement.
Figure 14 – Annual HDD from 2013 to 2018, based on insulation arrangement.
Figure 15 – Annual CDD from 2013 to 2018, based on insulation arrangement.
period, while the stone wool arrangement also indicates some increase, though noticeably less than the other two arrangements. This trend for the polyiso arrangements likely indicates that the thermal resistance of polyiso is decreasing as it ages, which is consistent with previous findings (Finch, 2010; Dell and Finch, 2013; BSC, 2013).
Similar to the analysis of energy implication due to membrane color, the impact of insulation arrangements was assessed by calculating degree days. HDD and CDD were calculated using roof surface temperatures with the same values calculated using ambient exterior air temperatures, as is standard practice. The roof surface temperature used to determine degree days for each membrane color are averaged for the same insulation arrangements with the different roof membrane colors. HDD and CDD calculated using the roof surface temperature instead of ambient exterior air temperature allows for a more accurate indication of heat flows through the assemblies. Again, note that CDD are presented as negative values in this report.
Figure 14 shows reduced HDD for the hybrid insulation roof when compared to the polyiso and stone wool roof assemblies. Naturally, HDD calculated using roof membrane temperature has a strong correlation to daily minimum roof membrane temperature experienced by each insulation arrangement. Roof membrane with polyiso insulation experienced the coldest minimums, while those with hybrid insulation experienced the warmest minimums, and these are reflected on the annual HDD with the coldest roof having the most HDD. The same but opposite correlation between CDD and daily maximum roof membrane does not seem to apply to annual CDD provided in Figure 15. The figure shows slightly increased CDD for the hybrid and stone-wool-only roofs, while the polyiso-only roof—which experienced the highest of the daily maximum temperatures—had the least CDD. These differences in annual HDD and CDD between the insulation arrangement is likely due to a combination of thermal mass and latent heat transfer. Additionally, while there is no significant difference in the CDD by insulation arrangements, this finding clearly illustrates that energy modeling of buildings based on ambient air temperatures with a lack of accounting for solar absorption on surfaces such as roofs can create significant inaccuracy in the result.
MOISTURE MOVEMENT IN
ROOF ASSEMBLIES
Moisture movement within a roof assembly impacts the heat transfer by carrying latent energy with it as it moves within the insulation layers. The RH levels below the insulation (i.e., on top of the air/vapor barrier) over the course of the six-year monitoring period are provided in Figure 16 by insulation arrangement. The values provided for each insulation arrangement are averaged across different roof membrane colors for the similar insulation arrangements.
Figure 16 indicates that there is a seasonal trend in the moisture levels within the assembly at these locations. The seasonal trend is a result of change in the predominant direction of the seasonal vapor drive. During the summer, the top of the insulation is being heated, driving vapor towards the bottom of the insulation, where RH sensors are located. The vapor drive is reversed during the winter.
The trend in average RH throughout the monitoring period also reveals that for all three insulation arrangements, a slight overall year-to-year increase in the RH occurs over the course of the first three years, then appears to stabilize from 2015 onward. This finding likely indicates that it took the roof assemblies approximately three years to reach equilibrium conditions based on the climate to which they are exposed and the initial conditions of the materials of which they were constructed
IIBEC 2020 Virtual International ConveVEntion & Trade Show | June 12-14, 2020 R RickeICKEtts et al. | 241
Figure 16 – Monthly average RH levels below the insulation for each insulation arrangement.
The seasonal trend is a result of
change in the predominant direction
of the seasonal vapor drive. During the summer, the top of the insulation
is being heated, driving vapor towards
the bottom of the insulation, where RH
sensors are located. The vapor drive
is reversed during the winter.
(care was taken to ensure no precipitation entered the roof assemblies during construction).
Figure 17 and Figure 18 plot the RH levels below the insulation layers for different insulation arrangements during sample periods of 48 hours in summer and winter. The daily movement of vapor is more apparent with the stone wool insulation arrangement because stone wool is more vapor-permeable, whereas polyiso is relatively impermeable and consequently restricts the movement of moisture within the insulation. Vapor is driven from the warmer side of the insulation towards the cooler side, and the direction of vapor movement reverses from day to night. The vapor drive is most significant during sunny and hot days in the summer.
Exploratory openings and moisture content field measurements in the fall of 2019 showed that the roof insulation for all assemblies was generally dry and no liquid water was observed. The results of field exploratory openings will be further discussed in subsequent papers.
DISCUSSION AND CONCLUSIONS
Long-term monitoring studies offer unique opportunities to expand our understanding of the performance of building enclosure materials and assemblies. The field monitoring data acquired over the last six years provide insight into the different performance of conventional roof assemblies, depending on insulation arrangement and roof membrane color.
The in-service reflectance of roofs was measured for gray and white (high and low point) roof membranes in two locations. Measurements over the six-year period generally did not confirm preliminary findings from the first three years of monitoring. It was initially assumed that the decreased reflectance of the white and gray membrane had been caused by soiling and weathering; however, after the sensors were cleaned in March 2015 and December 2016, the reflectance values were shown to have increased in 2015 to nearly initial installation levels from 2012. Because the sensors were prone to soiling in the field, samples of the installed roof membrane have been removed and are scheduled to be tested for their reflectance in a controlled lab setting to accurately measure degradation intensity after a six-year period. Results from the lab testing will be included in a separate paper. However, a preliminary visual comparison of gray and white roof membranes from 2012 and 2019 show that the aged membranes have darkened over time, largely due to degranulation.
The difference in performance between the roof membrane colors was examined using the field monitoring data acquired over the period. Consistent with previously reported results (Finch, Dell, and Ricketts, 2014), roofs with higher solar reflectance were generally found to experience less-extreme temperatures and slower changes in temperature. The impact of membrane color is greatest during sunny and warm periods, and consequently, the impact of the membrane color on CDD was found to be much greater than on HDD (as expected), and in many cases absorption of solar energy will create more additional cooling demand than it will reduce heating demand in the winter, potentially leading to a net energy penalty for using a dark-colored membrane. However, this result is not applicable if the interior space is not air
242 | RickeICKEtts et al. IIBEC 2020 Virtual International ConveVEntion & Trade Show | June 12-14, 2020
Figure 17 – Sample daily RH levels for different insulation arrangements in summer 2017.
Figure 18 – Sample daily RH levels for different insulation arrangements in winter 2017.
conditioned, as is the case for the building on which the monitoring was performed.
The relative performance of insulation arrangements was also evaluated. The monitored measurements reinforced previously reported findings with regard to factors affecting peak temperatures and temperature lag effects (Finch et al., 2014). The data showed a trend of increasing inward heat flow for all insulation arrangements over the course of the monitoring period, and it is most significant with polyiso roofs. This trend for the roofs with polyiso insulation is likely due to aging resulting in decreased thermal resistance, which is consistent with the laboratory measurements of insulation thermal performance (Finch, 2010; Dell and Finch, 2013; and BSC, 2013).
For all three insulation arrangements, a slight overall annual increase in the RH below the insulation was observed from the end of 2012 to mid-2015. However, the RH levels appear to have reached an annual equilibrium from mid-2015 to the end of 2018. It is worth noting that these roofs reached equilibrium not at entirely dry conditions, and that through daily variation from solar-driven moisture, RH levels greater than 100% are measured within the assemblies in some cases, potentially indicating the presence of liquid water. Exploratory openings and moisture content field measurements in the fall of 2019 showed that the roof insulation for all assemblies was generally dry and no liquid water was observed.
REFERENCES
Building Science Corporation (BSC). BSC Information Sheet 502: Understanding the Temperature Dependence of R-Values for Polyisocyanurate Roof Insulation. 2013. Available at: http://www.buildingscience.com.
M. Dell and G. Finch. “Monitored Field Performance of Conventional Roofing Assemblies – Measuring the Benefits of Insulation Strategy.” Proceedings of the 2013 RCI Symposium on Building Envelope Technology. November 13-14, 2013. Minneapolis, MN.
Environment Canada. 2018. Canada Climate Normals. Chilliwack Airport. Available at: http:/climate.weather.gc.ca/climate_normals/.
G. Finch. “Revised R-Values.” Professional Roofing. May 2010. Available at: http://www.professionalroofing.net/.
G. Finch, M. Dell, B. Hanam, and L. Ricketts. “Conventional Roofing Assemblies: Measured Thermal Benefits of Light to Dark Roof Membranes and Alternate Insulation Strategies.” Proceedings from the 2014 RCI Symposium on Building Envelope Technology. March 20-25, 2014. Anaheim, California.
G. Finch, M. Dell, and L. Ricketts. “Conventional Roofs: Measuring Impacts of Insulation Strategy and Membrane Color in Canada.” Proceedings of the 14th Canadian Conference on Building Science and Technology. Toronto, Ontario. 2014.
T. Smith, K. Liu, and R. Paroli. “Field Performance of APP Modified Bitumen Roof Membranes and Coating – The First Six Years.” 10th International Congress Proceedings. pp. 275-302. 1998.
IIBEC 2020 Virtual International ConveVEntion & Trade Show | June 12-14, 2020 R RickeICKEtts et al. | 243