BEG

Building Engineering Group

Civil Engineering Department, University of Waterloo, Waterloo, Ontario, Canada N2L 3G1

Driving Rain Loads for Canadian Building Design

For:

Silvio Plescia Canada Mortgage & Housing Corporation

External Research Program

By: John Straube & Chris Schumacher Building Engineering Group

University of Waterloo Waterloo, Ontario Canada 519 888 4015 jfstraube@uwaterloo.ca

October 2005

INTRODUCTION

The University of Waterloo was awarded a grant under the Canada Mortgage and Housing Corporations External Research Program to develop driving rain loads on Canadian buildings. Using detailed hourly wind and rain data, driving rain is calculated. This report summarizes the methods used, the type of data generated, and presents summarized results for use by building designers. A practical design method, partially validated with field data is presented for use.

1.1 The Importance of Driving Rain Loads

The amount of water deposited on the above-grade building envelope by driving rain is larger than any other source of moisture in almost all building types and climates. Both the deposited rain absorbed by the cladding and water penetration of the cladding can be a significant source of moisture for many deterioration mechanisms. Despite the importance of driving rain to building performance, relatively little is known about the magnitude, duration, and frequency of driving rain and driving rain deposition on buildings. Information about the magnitude and nature of the moisture load imposed by driving rain would be very useful and of practical value to many groups, such as: builders and designers of high-rise buildings, by assisting in the selection process for materials and systems, and assessing climate, orientation, height, and other performance related considerations manufacturers of bricks, siding, cladding and windows. The real knowledge of driving rain and its control should be of great value to designers and manufacturers of cladding systems, especially for export to markets in the United States, Asia, and Europe the restoration and repair industry, including product manufacturers, which will be able to make better decisions pertaining to renovation and repair of the huge stock of high-rise residential buildings, researchers and building science practitioners who require an in-depth understanding of driving rain in the natural environment as inputs for computer models so that they can predict wall behavior with more accuracy and confidence. Codes and standards also may benefit from an improved knowledge of driving rain. Both new and prototype cladding and window systems are increasingly being tested for the control of driving rain before they are mass produced and installed. The standards used to define the test protocols were developed at least 30 years ago and embody the understanding of that time. Even with the current understanding of driving rain, it is clear that these standard test methods are not always appropriate for assessing the driving rain control provided by most modern cladding assemblies.
BEG University of Waterloo Page 1 of 15

CMHC

Driving Rain Loads

Oct 2005

RAIN WIND AND DRIVING RAIN

2.1 Rain

The volume of water deposited on a wall system is the result of the interaction between the wind, rain, the building, and the building envelope. The stochastic natures of the wind and rain, and the many variables associated with the building site and geometry makes the prediction of driving rain deposition on a wall a difficult task. Even if the rate of rain deposition on a building surface were known, the relative amounts of absorption, face drainage, and penetration of water will depend on the cladding material, texture, and previous wetting history. This report reviews the nature of rain, and explores the mechanisms involved in wind-rain-building interactions. The rain-wind-building interaction is examined in the following sections, based on previous research by others and supported by experimental measurements of Waterloo, Canada conditions. The interaction is considered below in four parts: the nature of rain, the coincidence of wind and rain at the site, driving rain, and the interaction of buildings, wind, and rain. It is generally accepted that condensation of water vapour onto condensation nuclei creates cloud droplets (aided by cooling as a result of air rising). Cloud droplets are perfectly spherical in shape (because surface tension forces are much larger than aerodynamic or gravity forces) and range from 1 to 50 m in diameter with an average of 10 to 25 m (Barry 1992, Grey 1970, Beard 1976). Such small drops fall so slowly that their movement is governed by the wind currents aloft and they act as suspended particles. The precise mechanism of raindrop formation is not known, but the most widely accepted theory is based on ice accumulation around frozen condensation nuclei (Barry 1992). Hence, this theory predicts that drops can only form within clouds that are below 0 C. Because the vapour pressure at the surface of ice is less than that for water, an ice nuclei quickly grows in size by condensation. If the crystal diameter increases to more than about 0.10 to 0.25 mm, it begins falling. As the ice crystal falls through to warmer air layers, the ice melts and a raindrop forms. If two rain drops in a rainfall touch, they will likely coalesce into a larger drop. Larger drops fall faster and can therefore overtake and absorb smaller drops, further accelerating their fall. However, if the resultant drop becomes too large, aerodynamic forces overcome the surface tension forces holding the drop together and the drop will break apart (both Grey (1970) and Beard (1976) suggest that drops with diameters greater than about 6 to 7 mm are unstable). Both the condensation and coalescence mechanisms are likely at work in the formation of most rain drops. As the rainfall intensity increases, the probability of two raindrops colliding increases and the average raindrop diameter increases. This conclusion is supported by the field measurements of Best (1950), Laws and Parsons (1943) and Marshall and Palmer (1948) but each group developed its own relationship (Markowitz (1976) provides an excellent summary). Later researchers made use of radar readings (Rogers and Pilie, 1962, Caton 1966) and sophisticated electronic drop sizers (Dingle and Hardie1962, Bradley and Stow 1974) to further explore raindrops size distributions. The general

BEG University of Waterloo

Page 2 of 33

CMHC

Driving Rain Loads

Oct 2005

conclusion of this body of research is that the nature of the distribution clearly varies not only with rainfall intensity but also with the type of storm, the cloud height, etc. Figure 2.1 is a plot of the relative probability distribution of raindrop diameter as a function of rainfall intensity given by Best (1950): F() = 1- exp{-( 1.30r 0.232 h

2.245

(1)

where, F() is the cumulative probability distribution of drop diameters for a given rainfall intensity is the equivalent spherical raindrop diameter (mm), and rh is the rainfall rate or intensity on a horizontal plane (mm/m2/h).
0.3 Rainfall Intensity (mm/hr) 0.5 3 10

Relative Frequency

0.2

0.1

0.0 0 1 2 Drop Diameter (mm) 3 4

Figure 2.1: Distribution of raindrop sizes with rainfall intensity While an accurate knowledge distribution of drop sizes may not be available, the physical characteristics of the drops themselves are well understood. Beard (1976) recently collected the available experimental data of raindrop size, shape, and drag coefficient and, based on fluid dynamics principles, developed general relationships for the terminal velocity of raindrops as a function of air pressure and air temperature. These relationships match most available experimental
BEG University of Waterloo Page 3 of 33

CMHC

Driving Rain Loads

Oct 2005

data, including the comprehensive and widely accepted data of Gunn and Kinzer (1949). Figure 2.2 is a plot of this data. Dingle and Lee (1972) developed a simplified equation, accurate to 2.5%, for raindrop diameters of more than 0.3 mm: Vt() = -.166033 + 4.91844 - .8880162 + .0548883 9.20 (2)

where, in mm and Vt() is the terminal velocity of a raindrop in still air (m/s). More precise ( 0.5%) and much more unwieldy equations are available; see Beard (1976).
10

Terminal Velocity (m/s)

6 curve fit from Dingle data from Gunn & Kinzer

0 0 1 2 3 4 Drop Diameter (mm) 5 6

Figure 2.2: Terminal velocity of water droplets in still air Hours of rainfall per year and the distribution of rainfall intensity are important variables for building design. It is likely that a climate with a large number of hours of rainfall per year will impose greater demands on an enclosure than one with few hours. Similarly, the load imposed by an intense rainfall will be greater than for light rainfalls. This has been studied by many, but the data is normally presented in the form of peak rainfall intensity for a specified return period. This format is useful for stormwater management design, but less useful for building enclosure design. Leveleton Engineering (Levelton 2004) recently conducted a study of up to 8 years of hourly data for 12 stations in the lower mainland of BC. Their results for Vancouver airport are shown in Figure 2.3. It is clear that the number of hours with high levels of rain are quite limited, in the order of tens

BEG University of Waterloo

Page 4 of 33

CMHC

Driving Rain Loads

Oct 2005

of hours per year. However, it is also important to note that the Vancouver Airport experiences in the order of 1000 hours of light rain and drizzle, more than any other major Canadian city.

Figure 2.3: Hours of Rainfall per year versus intensity (Levelton, 2005)

2.2 Driving Rain: The Interaction of Rain and Wind

Driving rain is defined as the quantity of rain that passes through a vertical plane in the atmosphere. Driving rain occurs because raindrops which are falling to the ground at their terminal velocity are blown sideways at the speed of the wind at any given height above grade (Figure 2.4).

Figure 2.4: Driving rain

BEG University of Waterloo Page 5 of 33

CMHC

Driving Rain Loads

Oct 2005

The amount of driving rain in unobstructed wind flow can be calculated with reasonable accuracy. The speed at which raindrops fall is a function of the size of the drop. Essentially, as the drop size increases the rain drop terminal speed increases at a decreasing rate. The wind carries the drops along horizontally due to drag. The combination of gravity and wind forces determines the trajectory of the drop, and simple geometry can then be used to assess the amount of rain passing through a vertical plane. Complicating this assessment is the fact that there is a range of raindrop sizes in any rainstorm. Research by various meteorologists can be used to correlate the distribution of drop sizes in a rainstorm with the intensity of the rainfall. Based on this type of approach Lacy [1965] proposed a simple equation relating wind speed and rainfall intensity to driving rain: rv = 0.208 V rh where, rv is the rate of rain passing through a vertical plane (l/m2/h), V is the average wind velocity (m/s), and rh is the average rainfall rate on a horizontal plane such as the ground (mm/m2/h). This equation was based on a mix of his field measurements and some theoretical considerations. Subsequent theoretical work and a considerable amount of field measurement [Straube 1998] has allowed us to extend and generalize Equation 3 to rv = DRF V(z) rh where V(z) is the wind speed at, z, the height of interest (m/s), The proportionality constant in Equation (4), is the ratio of rain on a vertical plane (driving rain) to rain on a horizontal plane (falling rain) and has been defined [Straube & Burnett 1997] as the driving rain factor (DRF). From simple geometry, it can be seen that the Driving Rain Factor is equal to the inverse of the terminal drop velocity: DRF =1/Vt (5) (4) (3)

Field studies at the University of Waterloo [Straube & Burnett 1997], in Germany [Kuenzel 1994] and computer models [Choi 1994] have found that the value for the DRF ranges between 0.20 to 0.25 for average conditions. This is the reason that the simple semi-empirical Lacy equation was so successful. However, DRF does vary considerably for different rainfall intensities and rain storm types. For example, it can range from more than 0.5 for drizzle to as little as 0.15 for intense cloudbursts. The cosine of the angle between the plane of interest and the direction of the wind can be used to account for wind direction on a plane oriented in a specific direction. rv = DRF V(z) rh cos (6)

BEG University of Waterloo

Page 6 of 33

CMHC

Driving Rain Loads

Oct 2005

where is the angle between a line drawn perpendicular to the wall of interest and the wind direction. Experimental work [Straube 1998] has shown that the quantity of driving rain in an unobstructed wind flow can be calculated with an accuracy of better than 10% using Equations 2, 5 and 6, provided the median raindrop diameter predicted by Equation 1 is used.

2.3 The Coincidence of Rain and Wind

As shown, windspeed has a direct effect on the amount of driving rain It cannot, however, be assumed that the wind and falling rain can be modelled as statistically independent events. In fact, there is strong evidence to suggest that the wind-speed probability distribution is somewhat different during rain events. It is also well documented that the wind direction distribution is different during rain events (e.g., Skerlj 1999). Although all climates are not the same, the Surry study (Surry et al 1995) did find many climates in which the predominant wind direction was quite different than the wind direction during rain (for example, see Figure 2.5). The importance of wind speed to the intensity of driving rain can be calculated using the methods outlined above. However, high wind speeds also may generate high stagnation pressures that can drive rain into the cracks and openings of some types of enclosures. Wind speed can be converted to a stagnation air pressure (assuming a temperature of 15 C) using Equation 7. Pstag = 0.6 V2 Where V is the windspeed [m/s] and Pstag is the stagnation pressure [Pa] Small flaws in non-absorptive face-sealed claddings for example, may leak significantly more when exposed to pressure differences [Lacasse et al 2003]. However, brickwork does not leak significantly more under pressures of 50 Pa than 0 Pa [Straube and Burnett 2000]. In most cases the increase in air pressure is less important than the increase in water deposition, and lab testing provides some support for this contention [Lacasse et al 2003]. Hence, although practitioners often observe an increase in rain-penetration control problems in high exposure conditions much of the evidence suggests that these problems are usually due not to an increase in air pressure difference but to an increase in rain deposition. (7)

BEG University of Waterloo

Page 7 of 33

CMHC

Driving Rain Loads

Oct 2005

Figure 2.5: Wind Direction during Rain and During all hours for St Johns NFLD [Surry et al 1995] Results from the Leveleton study referenced earlier [Levelton 2005] for Vancouver, shown in Figure 2.6, demonstrate that the windspeed is faster during rainfall than during all hours. Although the difference shown in the plot is significant for Vancouver (the windspeed during rain is almost twice that during all hours) others (Surry 1985) have found that the increase in windspeed during rain is more modest (in the order of 10%) for a broader range of climate locations (Figure 2.7). Figure 2.8 plots the results of an unpublished detailed study of 25 years (1965-1990) of hourly wind and rain data for Seattle-Tacoma Airport, Washington. The average wind speed is higher by almost 20% during rainfall (4.65 m/s), but is low in absolute terms (an average of 3.9 m/s, or an equivalent pressure of about 10 Pa). This plot also shows that the hourly average stagnation pressures during more than 99% of all rain events are lower than 20 Pa. By comparison, the Levelton study found that the average wind pressure (based on similar hourly measurements as the SeaTac airport data) during rain at Vancouver Airport was just under 20 Pa, and that the peak hourly average (over nine years) was 162 Pa. It should also be noted that the shape of the distribution means that the mean

BEG University of Waterloo

Page 8 of 33

CMHC

Driving Rain Loads

Oct 2005

(average) and median (the value which is more than half of all samples) windspeed are quite different a few hours per year experience high speeds, and this increases the average.

Figure 2.6: Rainfall Intensity versus Windspeed for Vancouver (Levelton 2004) Light bars are during rain. Dark bars are all hours

BEG University of Waterloo

Page 9 of 33

CMHC

Driving Rain Loads

Oct 2005

Figure 2.7: Rainfall Hours for various Canadian locations (Surry 1995)

BEG University of Waterloo

Page 10 of 33

CMHC

Driving Rain Loads

Oct 2005

15%

0.6

Wind Pressure (Pa) 5.4 2.4 9.6

15

22

29

12%
Relative Probability (%)

During Rain All hours

9%

6%

3%

0% 0 1 2 3 4 5 6 7 Wind Speed (m/s)

Figure 2.8: Relative probability distribution of rainfall, wind speed, and wind pressure for Seattle, WA from 25 years or hourly data

2.4 Previous Driving Rain Data Formats

To assess the influence of climate on driving rain exposure, designers have in the past resorted to Boyd's driving rain map of Canada [Boyd 1963] or Grimms map of the United States (Grimm 1982). These maps plot the annual Driving Rain Index (DRI). The DRI is the product of the annual average wind speed and total annual rainfall, that is: DRI = Vwind,average rh,total where V and rh are based on annual statistics. Although useful as broad and relative climate measures, the DRI does not represent the actual quantity of rain deposition. However, as can be seen by comparison with Equation 6, the DRI could be converted to an actual quantity by use of the proper DRF based on hourly values. For lack of better information, these DRI plots assume that both the wind speed during rain and wind direction during rain are the same as those during average conditions. However, this causes some error since as described above, the distribution of wind speed during rain is slightly higher
BEG University of Waterloo Page 11 of 33

(8)

CMHC

Driving Rain Loads

Oct 2005

(typically 10 to 20% higher) than during all hours and the wind direction during rainfall is often quite different than during non-rainy hours [Surrey et al 1995, Skerlj 1999]. Both of these conclusions are climate dependent. A further limitation of the application of DRI plots to building design is that they do not reflect the impact of orientation and exposure. It is clear to practitioners that different localities have different building faces with significantly higher rain deposition and that a bungalow is exposed differently than a high rise building on a hilltop. It is this need for more accurate plots of driving rain that reflect the different climates of Canada, account for the actual distribution of wind speed and rain fall intensity, and account for wind direction during rain, that prompted this research project.

2.5 Previous Driving Rain Research

2.5.1 Field Measurements Driving rain has been investigated by building scientists and meteorologists in the field for many years. Driving rain has been measured under real conditions in Europe for at least 75 years (continuously for more than 50 years at one location we know of) although not usually on building systems. Lacy conducted the seminal English language study in the early 1960s, but his study was mostly aimed at driving rain in the free wind and during average rain events (Lacy 1965). Boyd (1963) generated a Canadian driving rain map in 1963 based mostly on Lacys work, but this dealt only with driving rain in open fields away from buildings. Baker and Robinson of IRC/NRCC produced an excellent and still valid qualitative study of driving rain wetting patterns and staining. Driving rain has been measured on an increasing number of buildings. Early studies (in the early 90s) were focused at IRC/NRCC, who mounted several gauges on a three story building. An array of 15 driving rain gauges were installed on a test hut at the University of Waterloo in 1994, with the express intent of improving our understanding of the distribution of rain. As part of the MEWS study, this author installed arrays of gauges on a 4 storey apartment in Kitchener and a large highbay industrial building in Vancouver in 1999. Five of the same UW gauges were installed by RDH Building Engineering on building retrofits in Vancouver as part of another research project (funded by CMHC, the Homeowner Protection Office and the BC Housing Management Commission). Finally, higher resolution commercial driving rain gauges have been installed on numerous test buildings around North America (North Carolina, Georgia, Washington, Minnesota, Alaska). All of this research, supported by results from the UW testhut facility and MEWS monitoring, suggest that driving rain can be rationally predicted (Straube 1997, 1998). 2.5.2 Modeling Initial research, in Singapore by Choi (1993), and at IRC/NRCC by Karagiozis and Hadjisophocieous (1995, 1997), has attempted to develop predictive computer models of rain wetting with mixed results. Work by others, such as Wisse (1994), and Van Mook (1999)

BEG University of Waterloo

Page 12 of 33

CMHC

Driving Rain Loads

Oct 2005

developed the modelling of driving rain further. Both wind tunnel and computer models however, require validation and calibration with real data. Recent work in Europe (Blocken and Carmeliet 2000) has provided some comparison between field studies and models and hence more confidence that computer models may provide useful data. CMHC also provided support to study driving rain in the Boundary Layer Wind Tunnel at the University of Western Ontario (Inculet and Surry 1995). This provided indications as to the distribution of rain deposition on tall buildings and the impact of overhangs.

2.6 Testing and Standards related to Driving Rain

One of the uses of driving rain data is to develop and interpret building enclosure component or assembly test results. A brief review of the rain deposition rates and standard tests is therefore useful. The ASTM Standard E331 "Standard Test Method for Water Penetration of Exterior Windows, Curtain Walls, and Doors by Uniform Static Air Pressure Difference" is widely used. This standard employs a uniform static air pressure difference of at least 137 Pa (equivalent to a windspeed of about 15 m/s or over 50 km/h) during the application of a water spray with a minimum rate of 203 mm/hr. The test lasts 15 minutes and any points of water leakage are recorded and described. For masonry walls, ASTM Standard E514 Standard Test Method for Water Penetration and Leakage Through Masonry is used. This method imposes loads of 138 l/m2/hr and a pressure of 500 Pa (equal to a wind speed 28 m/s or 100 km/h) for at least 4 hours. The CSA A-440 Window standard incorporates Driving Rain Wind Pressures (DRWP) as a method of selecting windows to resist driving rain. These pressures were developed by Welsh, Skinner, and Morris (1989) using an hourly rainfall threshold of 1.8 mm/h and five and ten year return periods. Although no driving rain intensities were predicted, air pressures for many Canadian locations were provided. These pressures are the highest hourly average pressures that would be experienced in 5 or in 10 years. The worst hour may be of interest to an enclosure component such as a window, which has no moisture storage capacity. Even so, a leak every 5 to 10 years may not be a problem, since the typically small amount of leakage would quickly evaporate. It is more likely that repeated leaks, many per year, over several years, will result in more serious enclosure deterioration than a single leak of somewhat larger magnitude. Hence, information regarding more common wind pressures and driving rain would be useful. In brief, test standards apply rates of water of about 100 to 200 l/m2/hr, at pressures equivalent to a stagnation pressure generated by a windspeed of 15 to 28 m/s. These test values can be compared to the results generated later in this report.

DRIVING RAIN DATA

Hourly weather files were purchased from the Meteorological Service of Canada, Environment Canada. Of the thousands of monitoring locations, we selected those that provided:

BEG University of Waterloo

Page 13 of 33

CMHC

Driving Rain Loads

Oct 2005

1. hourly observations 2. more than ten years of data (to provide a reasonable basis for annual average statistics) 3. wind speed, wind direction, and rainfall weather elements 4. data collection after 1965 (since this is when the anemometers of many stations were moved to provide consistent open wind measurements without the sheltering effect of obstructions), and 5. a range of major population centres and a broad geographic distribution. These restrictions reduced the available number of stations to less than 100, well distributed across Canada. However, it was found that there is no set of data available that combine hourly rain and wind. Hence, it was necessary to combine the CWEEDS (Canadian Weather Energy and Engineering Data Sets) files (which contain wind speed and direction) and the Environment Canada HLY03 data set (which contain hourly rainfall data). In some cases the hourly rainfall data was not available but an hourly weather flag (which indicates weather conditions, not quantitative rainfall) was. In many cases, data beyond 1989 was not available. To account for hours in which no rain was reported (in the HLY03 files) but hourly observations were (in the CWEEDS files), the Atmospheric Environment Service (AES 1984) equivalent rain rates were used: 1.8 mm/hr for light, 5.1 for moderate and 13.0 for heavy rain. Driving rain data were calculated using equations 1, 2, and 5 and the hourly weather data files created. Each of the cities data from 1965 to 1989 (25 years) was stored in a data file which occupied approximately 9 MB. Eighteen major Canadian cities were investigated in detail. These are listed and described in the Table 3.1. The data calculated for each year 1965-1989 included: 1. Annual rainfall and number of hours of rainfall 2. Annual driving rain from each direction of a 16 point compass 3. Total Annual Driving Rain 4. Annual driving rain on a wall facing each direction of a 16 point compass 5. Peak annual driving Rain and Direction 6. Windspeed vs direction during all hours and during rainy hours Driving rain data was calculated from the data files using

BEG University of Waterloo

Page 14 of 33

CMHC

Driving Rain Loads

Oct 2005

rv () = DRF(rh) V10 rh

(8)

where as before, rv () is the driving rain passing through a vertical plane in the free wind for 16 different compass points () at 22.5 intervals rh is the hourly total rainfall DRF(rh) is the DRF as a function of the hourly average rainfall and V10 is the hourly average windspeed at 10 m above grade. Appendix B contains the plots of driving rain versus direction for each city. All of the annual data is presented in Appendix D.
Location Calgary Charlottetown Edmonton Fredericton Montreal North Bay Ottawa Prince George Quebec Saint John Saskatoon Shearwater St. John's Sydney Toronto Vancouver Victoria Winnipeg Province AB PE AB NB QC ON ON BC QC NB SK NS NF NS ON BC BC MB WBAN Stn. No. 25110 14688 25142 14670 4770 4705 4772 25206 4708 14643 25015 14633 14506 14646 94791 24287 24297 14996 Climate ID 3031093 8300400 3012205 8101600 7025250 6085700 6106000 1096450 7016294 8104900 4057120 8205090 8403506 8205700 6158733 1108447 1018620 5023222 Time Zone -7 -4 -7 -4 -5 -5 -5 -8 -5 -4 -6 -4 -4 -4 -5 -8 -8 -6 Elevation 1079 57 677 23 30 369 126 676 75 107 501 145 141 60 176 5 70 240 Latitude 51.1 46.25 53.32 45.92 45.5 46.37 45.45 53.88 46.8 45.32 52.17 44.88 47.52 44.63 43.67 49.25 48.65 49.9 Longitude -114.02 -63.13 -113.58 -66.62 -73.62 -79.42 -75.62 -122.67 -71.38 -65.88 -106.68 -63.5 -52.78 -63.5 -79.63 -123.25 -123.43 -97.23

Table 3.1: Cities Investigated in Detail Error! Reference source not found. plots the annual rainfall for three different cities for the years 1965-1989. It can be seen that the amount of rainfall can vary significantly from year to year. The data in Appendix D shows that the range of annual driving rain deposition varies by a factor of about 2 to over 4 depending on the year. Hence, although we often use average values, it should be clear that rainfall varies significantly from year to year. For most of the cities studied, twice as much rainfall was recorded during the wettest year than the driest year over the study period (1965-1989). However, annual average values still are likely the most useful design statistic, with a safety factor applied to account for the change in weather from year to year. The most important annual statistics are summarized in Table 3.2.

BEG University of Waterloo

Page 15 of 33

CMHC In Open Field Annual Rainfall City Calgary, AB Charlottetown Edmonton, AB Fredericton, NB Montreal, QC North Bay, ON Ottawa, ON PrinceGeorge,BC Quebec, QC Saint John, NB Saskatoon, SK Shearwater, NS St Johns, NF Sydney, NS Toronto, ON Vancouver, BC Victoria, BC Winnipeg, MB mm/yr 340 1110 411 1072 768 788 814 447 1026 1144 284 1177 1168 1189 718 1164 831 440

Driving Rain Loads On Building Driving Rain Total mm/yr 359 1344 387 867 645 641 674 273 781 1399 324 1326 1742 1630 587 1028 627 480 Average (on wall) kg/m /yr 94 406 109 262 194 196 202 83 243 431 92 411 532 501 177 325 196 139
2

Oct 2005 During Rain Annual Windspeed Rainy during Hours Rain hrs 240 676 295 597 539 510 502 394 570 618 206 616 753 713 459 1012 770 273 m/s 5.9 6.7 4.7 4.1 4.6 4.8 4.4 3.4 4.1 6.3 6.1 5.8 7.8 6.9 4.4 3.9 3.0 6.0

on worst Orientation kg/m /yr 214 653 230 376 261 272 315 173 519 836 140 778 934 970 284 804 414 199
2

Worst wall Direction NW SE WNW ENE E SSE ENE SSW ENE SE SE SE SE SE NW SE SE SE

Table 3.2: Driving Rain Statistics 1965-1989

BEG University of Waterloo

Page 16 of 33

CMHC

Driving Rain Loads

Oct 2005

1600 1400 1200

Annual Rainfall [mm]

Edmonton Toronto Vancouver

1000 800 600 400 200 0

19 65

19 67

19 69

19 71

19 73

19 75

19 77

19 79

19 81

19 83

19 85

19 87

Figure 3.1: Annual Rainfall versus year for selected Cities

BEG University of Waterloo

Page 17 of 33

19 89

CMHC

Driving Rain Loads

Oct 2005

1100
Annual Average Rainy Hours (based on 1965-1989) .

1000 900 800 700 600 500 400 300 200 100 0
Sa sk at oo n C al ga r W in y ni pe Ed g Pr mo n in to ce n G eo rg To e ro n N or to th Ba y O tta w a M on tre Q al ue be Sh c ea rw a Sa te r in tJ o Fr ed hn e C ha ricto rlo n tte to w Sy n dn St ey Jo hn s Vi ct o Va ri nc a ou ve r

Figure 3.2: Average Number of Rainy Hours per Year (1965-1989) As mentioned in the introduction, the stagnation pressure and Driving Rain Index have been used as important measures of assessing rain penetration. Both measures are listed in Table 3.3 along with the number of rainy hours. It can be seen that the stagnation pressure during rainy hours ranges from only 6 Pa in Victoria, BC to as high as 36 Pa in St Johns, NF. Although this is a large range, the annual driving rain also varies from under 300 to over 1700 l/s/m2/yr. This table also shows that the average increase in windspeed during rain relative to average annual rainfall is 20%. The Driving Rain Index (DRI) which is still used in the Canadian Masonry Connectors Standard to assess the need for corrosion resistance. The ratio of annual total driving rain (DR) to DRI averages 232, and has a coefficient of variation of about 12%. Hence, the DRI is a reasonable predictor of the total amount of driving rain, although the scatter is in the order of 25%.

BEG University of Waterloo

Page 18 of 33

CMHC

Driving Rain Loads During Rain Hours of Avg Rain Windspeed m/s 240 5.9 676 6.7 295 4.7 597 4.1 539 4.6 510 4.8 502 4.4 394 3.4 570 4.1 618 6.3 206 6.1 616 5.8 753 7.8 713 6.9 459 4.4 1012 3.9 770 3.0 273 6.0 All hours Avg Pressure Pa 21 27 13 10 13 14 12 7 10 24 22 20 36 28 12 9 6 22 Windspeed m/s 4.3 6.3 3.6 3.6 4.0 3.7 3.8 3.0 3.8 4.9 4.9 4.3 6.6 5.5 4.1 3.2 2.7 4.8

Oct 2005

City Calgary, AB Charlottetown, PE Edmonton, AB Fredericton, NB Montreal, QC North Bay, ON Ottawa, ON Prince George, BC Quebec, QC Saint John, NB Saskatoon, SK Shearwater, NS St Johns, NF Sydney, NS Toronto, ON Vancouver, BC Victoria, BC Winnipeg, MB

Driving Rain Index 2 m /s DRI/DR 1.46 246 7.01 192 1.48 263 3.87 224 3.07 210 2.90 221 3.12 216 1.36 201 3.94 198 5.57 251 1.40 232 5.06 262 7.76 224 6.52 250 2.93 200 3.71 277 2.22 282 2.11 227

DRF 0.18 0.18 0.20 0.20 0.18 0.17 0.19 0.18 0.19 0.19 0.19 0.19 0.19 0.20 0.19 0.23 0.25 0.18

Table 3.3: Driving Rain Analysis 1965-1989 The Driving Rain Factor (DRF) is the inverse of the average raindrop fall velocity. It has been calculated for all rainy hours. It varies from a low of 0.17 in North Bay to a high of 0.25 in Victoria. The higher the value, the smaller the average size of the rain drops. Choosing an average value of about 0.2 will provide answers within about 10% of a more detailed method. Figure 3.3 plots the number of annual hours of rainfall in numerous categories of rainfall intensity (in mm/hr). This plot shows that there are significant differences between the nature of rainfall in different climates. For example, Saint John, NB experiences 9 hours per year of rainfall intensities over 10 mm.hr, whereas Vancouver, BC experience no hours. Vancouver BC experiences over 500 hours per year of light drizzle (less than 1 mm/hr) whereas Saint John, NB experiences just over 100 hours of the same conditions.

BEG University of Waterloo

Page 19 of 33

CMHC

Driving Rain Loads

Oct 2005

1000

0-1 5-7
Rainfall hours per year in each category

1-3 7-10

3-5 >10

100

10

1 St Johns Edmonton Prince George Vancouver Toronto Winnipeg Ottawa Saint John

Figure 3.3: Hours of Rainfall in Different Rainfall Intensity Categories

DRIVING RAIN ON BUILDINGS

The driving rain data presented above is a meteorological measure of a broad geographical area. Building designers are, however, concerned with the amount of rain that is deposited onto the face of a building. The values given in the plots of directional driving rain can be modified to provide estimates of the quantity of driving rain deposition on a building. The first step is to assess the influence of building shape and aerodynamics, and the second is to correct for building height and topography.

4.1 Building Shape and Aerodynamics

When wind encounters a building, streamlines and pressure gradients form around it. While it is clear that driving rain is re-directed by these streams of air (since the droplets are carried with the wind), accounting for this effect is difficult. A simple and practical approach for estimating rain deposition on buildings is to divide driving rain deposition into the effects of the undisturbed wind (the free wind driving rain, a climate and site specific value) and the effect of the aerodynamics and geometry of the local building. A linear

BEG University of Waterloo

Page 20 of 33

CMHC

Driving Rain Loads

Oct 2005

factor, the rain deposition factor (RDF), can be used to transforms the rate of driving rain in the free wind (i.e. outside of the region disturbed by a building) to the rate of rain deposition on a particular building (Straube 1998). Section 3 presented driving rain data for the free wind. For a particular orientation and spot on the building face, the free wind driving rain values can be modified as: rvb = RDF DRF V(z) cos ( ) rh where rvb is the rain deposition rate on a vertical building surface(l/m2/h), RDF is the Rain Deposition Factor, the ratio of rain in the free wind to rain deposition on a building, which accounts for the effect of building shape and size on rain deposition. DRF is the Driving Rain Factor, which accounts for interaction of the wind and rain in the undisturbed wind, is the angle between the normal to the wall and the wind direction, and V(z) is the windspeed in m/s at z meters above grade. Driving rain data on a standard building face was calculated from the data files using: rbv () = DRF(rh) RDF V10 rh cos ( ) for all +/-90. where all variables are as before. This equation was used to calculate the driving rain on a nominal building at 10 m above grade, in flat, open terrain, with an RDF =1. The data for 18 Canadian cities is presented in Appendix C and the numerical data for each year (1965-1989) is tabulated in Appendix D. As expected, the variation in driving rain is similar year-to-year as rainfall. (10) (9)

BEG University of Waterloo

Page 21 of 33

CMHC

Driving Rain Loads

Oct 2005

1250

Edmonton

Toronto

Vancouver

Maximum Driving Rain on Worst Wall [kg/m2/yr]

1000

750

500

250

19 65

19 67

19 69

19 71

19 73

19 75

19 77

19 79

19 81

19 83

19 85

19 87

Figure 4.1: Driving Rain Deposition for Selected Cities (1965-1989)

BEG University of Waterloo

Page 22 of 33

19 89

CMHC

Driving Rain Loads

Oct 2005

1000

Maximum Driving Rain on Worst Wall [kg/m 2/yr]

C

800

600

400

200

Figure 4.2: Average Annual Driving Rain on Wall facing Worst Orientation (1965-1989) The typical range of pressures and driving rates applied during tests has been reviewed in Section 2.6. It is interesting to compare the predicted rain deposition rates and pressures from the measured data in the field. The cumulative driving rainfall intensity and cumulative wind speed distribution during rain over 1.8 mm/hr was calculated for each city. Some of the results have been extracted for important population centres in Table 4.1. It should be clear from these data that driving rain deposition rates of more than 10 l/m2/hr are exceedingly rare in most climates (that is, they occur for about 1 hour per year). Even in extreme climates, such as Sydney, NS such rates are exceeded only about 10 hours per year. The worst hourly rain deposition rate over 25 years for all climates tend to be in the order of 20 to 40 l/m2/hr.

Table 4.1: Summary Annual Driving Rain and Windspeed Statistics

BEG University of Waterloo

Pr

a Ed lga in mo ry, ce nt AB G on Va eor , A nc ge B ou , B C Vi ver ct , B W or C in ia Fr nip , B ed eg C e , Sa ric MB t o in t J n, N St ohn B Sh Jo , N ea hns B rw , N a F Sy ter, N dn NS or ey th , Ba NS O y, tta O w N C To a, ha ro O rlo nt N tte o, O M tow N on n, tr P Q eal E u , Sa eb QC e sk c, at Q oo C n, SK
Page 23 of 33

CMHC
(mm/hr) 0 1 3 5 7 10 Peak (m/s) 0 1 3 5 7 10 Peak

Driving Rain Loads

Cumulative Driving Rain Distribution, Worst Plane (hrs) 138.3 178.7 125.5 110.8 65 70.8 48.6 41.4 18.7 18.9 12.1 9.2 6.3 5.9 4.6 3.3 2.7 2.3 2.2 1.4 1.1 0.9 1.1 0.4 28.6 16.8 39.4 35.3

Oct 2005

164.7 76.4 22.3 7.5 3.2 1.3 30.9

594.9 307.2 56 9.5 2.8 0.6 15.9

711 373.4 152.5 65.4 30.6 10.7 34.8

149.4 47.7 3.6 0.4 0.2 0 11.2

Cumulative Wind Speed Distribution (hrs, during Rainfall > 1.8 mm/hr) 43.7 54.6 37.3 28.9 51 142.4 42 52.4 36.4 28.2 50.2 128.8 36.2 43.2 31.4 23.1 43.2 73.7 26.3 29.9 23.2 16 31.2 31.8 15 15.8 14 8.5 17.9 12.5 5.2 4 4.3 2.4 4.7 2.6 24.7 19.7 20.6 18.6 21.4 15

164.8 155 131.3 110.9 89.9 58.3 22.5

31.6 25 16.8 8.5 2.6 0.3 13.3

4.2 Wind Exposure Corrections

The figures and plots presented here must be modified by the use of linear factors to account for building geometry/shape, height, upstream roughness, and topographic features. 4.2.1 Building Shape Building shape and overhang effects can be accommodated through the use of different Rain Deposition Factors. The RDF is a function of the building's shape, aerodynamics, the wind's angle of attack, raindrop diameter, and wind speed. In general, since most rain events are of low intensity and have a similar distribution of raindrop sizes, average values of RDF measured in the field over a range of driving rain events have shown somewhat consistent results (Straube 1998) and have been shown to be capable of predicting a special case on a 15 minute basis (Straube 1997). It is interesting to note that recent computer modelling back by field measurements (Blocken and Carmeliet 2004) have recently provided strong evidence that a single RDF factor independent of rainfall intensity is a reasonable approximation over a broad range of rainfall intensities. Again, this conclusion would need to be confirmed in hurricane regions because of their different pattern of rain and wind. The most important dependencies are with the building shape and aerodynamics, and hence these are investigated below. Other than Straube (1997, 1998) and Schwarz (1973) the literature contains few references of simultaneous measurements of driving rain in the environment and the driving-rain deposition on a building. However, when results from the literature of field measurements (Schwarz 1973, Frank 1973, Sandin 1988, Flori 1992, Henriques 1992, Kuenzel 1994, Straube 1998, Blocken and Carmeliet 2000), wind tunnel tests (Inculet 1994), and computer modeling (Choi 1994, Karagiozis et al 1997, Blocken and Carmeliet 2000) are evaluated in terms of calculated RDF there is a reasonable agreement of results. Until the results of further research are available, the RDF values given in Figure 4.3 (compiled from the references listed) are recommended. An RDF of 1.0 appears to be appropriate near the upper corners of blunt-edged, rectangular buildings. Over small areas with

BEG University of Waterloo

Page 24 of 33

CMHC

Driving Rain Loads

Oct 2005

unusual or complex geometries (eg. notches in the elevation of a building), much higher RDF values may be more realistic. An RDF of much less than one is likely over most areas of most buildings, but the upper corners of high rise buildings could have limited areas of much greater than one. Peaked roofs and overhangs redirect airflow up and over the building at a distance further from the facade (Figure 4.4) and can thereby have a significant effect on rain deposition (regardless of the building size). This has been shown at the one-storey UW Beghut (Straube 1998), unpublished measurements conducted for MEWS (on a 4 storey apartment) and in wind tunnel studies of a highrise apartment building (Inculet and Surry 1995). For example, adding a 1.5 m wide canopy to a multi-storey building will result in a lower RDF value and can, in theory, be an effective and economical means of improving rain control. Similarly, a peaked roof not only leaks less than a lowslope roof, it may also reduce the amount of driving rain on walls by deflecting the wind (Straube 1998).

Figure 4.3: Rain deposition factor (RDF)

BEG University of Waterloo

Page 25 of 33

CMHC

Driving Rain Loads

Oct 2005

Figure 4.4: Influence of Overhangs on Wind Around Buildings To aid building designers in applying the RDF factor to the driving rain data presented in Section 3, the data was processed to create plots of driving rain deposition at 10 m above grade with an RDF of one, for each of 16 compass points. This data is presented in Appendix C. This data allows users to read the driving rain, cosine-corrected for the orientation of any specific wall, and then multiply this value by height, exposure, terrain, and building shape factors. It is clear from Equation 9 that rain deposition will increase directly with wind speed. Hence, building facades exposed to high wind speeds are also exposed to higher levels of driving rain. Equation 9 is only accurate if the wind speed at the height of interest is applied and the influence of terrain and site features is accounted for in the velocity used. Weather data files provide the unobstructed wind speed at 10 m above grade. Wind speeds increase with height above grade and are accelerated close to the ground on hilltops, and will be higher in exposed conditions (open country) than for buildings protected by other houses, trees or by hills. To correct for height, the type of upwind terrain roughness, and sharp terrain features, such as hills and escarpments, the wind speed, used in driving rain calculations, must be modified. 4.2.2 Height and Upwind Roughness The fact that wind speed increases rather rapidly with height means that the driving rain exposure of tall buildings is much higher than for low-rise buildings. Although there is only one study of rain deposition on tall buildings that measured rain deposition over the height of a building (Schwarz 1973), the results support the approach taken below. Other studies on tall buildings (van Mook 1999, Lacy 1965) have been undertaken but the data is too sparse to capture any correlation between the vertical rain deposition rate and the wind speed gradient. The standard approach used to correct wind speed with height is provided in the National Building Code of Canada (NBCC 1996). The wind speed at any height, V(z) can be found from:
BEG University of Waterloo Page 26 of 33

CMHC

Driving Rain Loads

Oct 2005

V(z) = V10 (z/10)

(11)

Where V10 is the standard wind speed 10 m above grade normally reported by weather stations (m/s), z is the height above grade (z) and is the exposure exponent, shown in Table 4.2.

Wind speed

10 m standard height

Height (m)
Open Country

Suburban

City Center

Figure 4.5: Different Upwind Exposure Types

Exposure

A. Open country: level terrain or open water with few obstructions 1/7= 0.14 and only scattered buildings, trees or other obstructions B. Suburban: urban areas, wooded terrain, or centers of small = 0.25 towns C. City center: large city centers, with heavy concentration of tall 3/8= 0.36 buildings, at least half over 4 stories. Table 4.2: Exponent for Different Upwind Exposure

Exponent

The increase in velocity with height has been calculated for the three different exposure classes and the results summarized in Table 4.3 and plotted in Figure 4.6.
Height (m) 1 3 5 Open Country 0.72 0.84 0.91 Suburban 0.56 0.74 0.84 City Center 0.44 0.65 0.78

BEG University of Waterloo

Page 27 of 33

CMHC 7 10 20 30 50

Driving Rain Loads 0.95 1.00 1.10 1.17 1.25 0.91 1.00 1.19 1.32 1.50 0.88 1.00 1.28 1.49 1.78

Oct 2005

Table 4.3: Wind Velocity Mulitpliers (EHF) for Different Heights and Exposures This exposure and height factor (EHF) can be used directly to adjust the amount of driving rain predicted by Equation 9.

BEG University of Waterloo

Page 28 of 33

CMHC

Driving Rain Loads

Oct 2005

50

40

Exposure: City Center Open Country

Height Above Grade, h [m]

Suburban 30

20

10

0 0.5 0.8 1.0 1.3 1.5 1.8 2.0 Exposure and Height Factor (EHF)
Recommended multiplication factors to apply to above: Sheltered: 0.5 if buildings or obstruction of building height are within a distance equal to twice the building height Hilltop or Cliff top: See speed-up correction factor in text

Figure 4.6: Correction Factors for Different Exposures and Heights (EHF) 4.2.3 Hills and Escarpments It is well known that wind speed increases as it is forced over hills and escarpments. The higher wind speeds result in higher driving rain deposition and higher wind pressures on buildings, and experience suggests that more rain penetration problems occur. The speed-up factor can be

BEG University of Waterloo

Page 29 of 33

CMHC

Driving Rain Loads

Oct 2005

calculated with the aid of the NBCC as well. Figure 4.7 defines the variables needed to use the following equation to find the wind speed corrected for topography (NBCC 1996): TF = {1 + Smax {1 |x| / (k L) } e(-az / L) }

(12)

Where TF is the topography wind speed correction factor, H and L are topographic geometric values defined in Figure 4.7 (m), x is the distance from the crest (m), z is the height above the crest (m) k, Smax and a can be found in Table 4.4.

Figure 4.7: Speed-up Factor Definitions (NBCC 1996)

Hill Shape Two dimensional ridges Two-dimensional escarpments Three-dimensional symmetrical hills Smax 2.2 H / L 1.3 H/L 1.6 H/L a 3 2.5 4 k Windward x<0 Leeward x>0

1.5 1.5 1.5

1.5 4 1.5

Table 4.4: Parameters for topography factor calculations (NBCC 1996)

BEG University of Waterloo Page 30 of 33

CMHC

Driving Rain Loads

Oct 2005

To provide a sense of the size of the TOF has been Table 4.5.
Slope (degrees) (H/2 / L) 7 14 27 36 45 60 66 75 for a L=30 m ridge height, z, above crest (m) 3 6 10 1.19 1.14 1.10 1.35 1.27 1.19 1.62 1.49 1.35 1.84 1.66 1.48 2.06 1.85 1.62 2.59 2.29 1.96 2.89 2.54 2.15 3.61 3.15 2.64 for L=30 m Escarpment 30 1.01 1.03 1.05 1.08 1.10 1.18 1.22 1.35 3 1.12 1.23 1.42 1.57 1.74 2.13 2.36 2.91 6 1.09 1.18 1.34 1.47 1.61 1.94 2.13 2.61 10 1.07 1.13 1.25 1.35 1.46 1.73 1.88 2.28 30 1.01 1.03 1.05 1.08 1.10 1.17 1.22 1.34

Table 4.5: Topography Factor Example Calculations The Building Research Establishment (BRE 1984) has also published a simplified method of accounting for wind accelerating over different types of terrain and provides more detailed information than the NBCC method.

4.3 Driving Rain Loads for Design

To apply all of the foregoing, a designer needs to 1. find the driving rain load on a building, rbv, for the nearest climate region (based on the 18 available plots) and specific orientation (within 22.5). 2. assess the building geometry (RDF) 3. consider the upwind exposure and height of interest on the building (EHF), and 4. apply a topography (TOF) factors if needed. Given these values the design load can be found from: Design Driving Rain Load = rbv RDF EHF TOF where the annual driving rain load is in kg/m2 or l/m2. The same factors are used to find the driving rain load from the worst direction, the annual total, or other values. 4.3.1 Example As an example of using the data and methods described, consider two different walls in Toronto, one facing east and one west. Inspection of the Toronto driving rain graph in Appendix C shows a free wind driving rain quantity of 130 mm/yr for the west and 275 mm/yr for the east. If one considers a bungalow wall 2 m above grade, sheltered by closely spaced houses in a suburb, the 150
BEG University of Waterloo Page 31 of 33

(13)

CMHC

Driving Rain Loads

Oct 2005

mm/yr would be modifying by a factor of 0.7 (from Figure 4.6) and a further reduction factor of 0.5 (from the note on sheltering). If the bungalow had a peaked roof with a 300 mm overhang, an RDF of 0.5 would capture the highest rain values. The result would be a driving rain total of 130*0.7*0.5*0.5= 23 mm per year, which is equivalent to 23 liters per m2 per year. For an east facing wall on the top floor of a 50 m tall blunt edged (RDF=1.0) condominium in a suburban exposure, Figure provides a correction factor of 1.5. Using and RDF of 1 for the top corners, the driving rain deposition would be predicted to be 275 * 1.5 * 1.0 * 1.0 = 412 mm per year or 412 l/m2/year almost 20 times as much rain as the sheltered low-rise bungalow wall facing west. This example demonstrates the very significant influence of exposure and orientation in a climate. It should also be noted that the combination of high exposure and choice of building shape (high RDF) in a low driving rain climate (such as Edmonton) can result in much more rain deposition than a sheltered low rise building in a high driving rain climate (such as Vancouver).

CONCLUSIONS

Hourly weather data from several sources has been combined to create weather files of hourly wind, wind speed, and rainfall for 18 cities for the period of 1965 to 1989 inclusive. This data was used to calculate climate statistics and driving rain. A methodology for calculating the driving rain load on buildings has been presented and explained. It is clear that the many complexities of real buildings and sites will render the methodology suggested here inaccurate. However, the method should allow one to generate a reasonable estimate, and to compare one site and building to another on a relative scale. Also, as has been shown, rainfall and wind vary considerably from year to year. Hence, the practical need for a high level of accuracy is questionable. These data and methodology, although partially validated through field measurements, should be used with care and professional judgment. Many more field measurements of tall and complex building shapes are needed, as are measurements in a range of exposure conditions. Sincerely,

Dr John Straube Assistant Professor Dept of Civil Engineering and School of Architecture University of Waterloo

BEG University of Waterloo

Page 32 of 33

CMHC

Driving Rain Loads

Oct 2005

REFERENCES

AES. Software Implementation for Climatological Ice Accretion Modeling Project, Internal Report to Energy and Industrial Applications Section, Atmospheric Environment Service, Canadian Climate Center, Toronto, Canada, 1984. Barry, R.G., and Chorley, R.J., Atmosphere Weather and Climate, 6 ed., Routledge, New York, 1992. Beard, K.V., "Terminal Velocity and Shape of Cloud and Precipitation Drops Aloft", J. of the Atmospheric Sciences, Vol. 33, 1976, pp. 851-864. Best, A.C., "The Size Distribution of Raindrops", Quart. J. Royal Meteor. Soc., Vol. 76, 1950, pp. 16-36. Blocken, B., Carmeliet, J. A Simplified Approach for Quantifying Driving Rain on Buildings, Proc. Of Performance of Exterior Envelopes of Whole Buildings IX, Clearwater, Dec. 2004. Blocken, B., Carmeliet, J., Driving Rain on Building Envelopes I: Numerical estimation and Fullscale Experimental Verification, J. of Thermal Envelope and Building Science Vol 24, No. 1, 2000, pp. 6185. Boyd, D.W., Driving-Rain Map of Canada, DBR/National Research Council, TN 398, Ottawa, 1963. Bradley, S.G., and Stow, C.D., "The Measurement of Charge and Size of Raindrops: Part II. Results and Analysis at Ground Level", J. of Appl. Meteor., Vol 13, Feb 1974, pp. 131 - 147. BRE, The Assessment of Wind Speed over Topography. Digest 283, Building Research Establishment, Garston, U.K, March 1984. Caton, P.G., "A study of the raindrop-size distributions in the free atmosphere", Quart. J. Royal Meteor. Soc., Vol 92, 1966, pp. 15-30. Choi, E.C.C., Simulation of Wind-Driven Rain Around a Building, Journal of Wind Engineering and Industrial Aerodynamics, Vol 46-47, 1993, pp. 721-729. Choi, E.C.C.. Determination of the wind-driven-rain intensity on building faces. Journal of Wind Engineering and Industrial Aerodynamics, Vol 51, 1994, pp. 55-69. Dingle, A.N. and Hardy, K.R., "The description of rain by means of sequential rainddrop-size distributions", Quart. J. Royal Meteor. Soc., Vol. 88, 1962, pp. 301-314. Dingle, A.N., and Lee, Y., "Terminal Fall Speeds of Raindrops", J. of Appl. Meteor., Vol 11, August 1972, pp. 877 - 879. Flori, J-P., Influence des Conditions Climatiques sur le Mouillage et le sechalge d'une Facade Vertical. Cahiers du CTSB 2606, September, 1992. Frank, W., Entwicklung von Regen and Wind auf Gebaeudefassaden, Verlag Ernst & Sohn, Berichte aus der Bauforschung, Vol 86, 1973 pp. 17-40. Grimm, C.T., "A Driving Rain Index for Masonry Walls", Masonry: Materials, Properties, and Performance, ASTM STP 778, J.G. Borchelt, Ed., American Society of Testing and Materials, 1982, pp. 171-177. Gunn, R, and Kinzer, G.D., "The terminal velocity of fall for water drops in stagnant air", J. Meteor., Vol 6, 1949 pp. 243-248.

BEG University of Waterloo

Page 33 of 33

CMHC

Driving Rain Loads

Oct 2005

Handbook on the Principles of Hydrology, ed. Donald M. Grey, National Research Council of Canada, Ottawa, 1970, pp. 2.2-2.3. Henriques, F.M.A.. "Quantification of wind-driven rain - an experimental approach", Building Research and Information, Vol. 20, No. 5, 1992, pp. 295-297. Inculet, D. and Surry, D.. Simulation of Wind Driven Rain and Wetting Patterns on Buildings. A report published by the CMHC, February 1995. Karagiozis, A., and Hadjisophocieous, G.,Wind-Driven Rain on High-Rise Buildings, Proc. of the BETEC/DOE/ASHRAE Thermal Performance of Building Envelopes VI, Dec. 1995, pp. 399-406. Karagiozis, A., Hadjisophocieous, G., and Cao, S., "Wind-Driven Rain Distributions on Two Buildings", J. of Wind Engineering and Industrial Aerodynamics, Vol 67 and 68, 1997, pp. 559-572. Knzel, H.M.. Regendaten fr Berechnung des Feuchtetransports, Fraunhofer Institut fr Bauphysik, Mitteilung 265, 1994. Lacasse, M. A., OConnor, T.J., Nunes, S., Beaulieu, P. 2003. Experimental Assessment of Water Penetration and Entry into Wood-Frame Wall Specimens - Final Report. MEWS Consortium, IRC/NRCC,Ottawa, February, 2003. Lacy, R.E., Driving-Rain Maps and the Onslaught of Rain on Buildings, Proc. of RILEM/CIB Symposium on Moisture Problems in Buildings, Helsinki, (Building Research Station Current Paper 54, HMSO Garston, U.K) 1965. Laws, J.O., and Parsons, D.A., "Relation of raindrop size to intensity", American Geophys. Union Trans., No. 24, pt. 2, 1943, pp. 453-460. Levelton Engineering, Wind-Rain Relationships in Southwestern British Columbia Final Report. Research report for CMHC, Ottawa, August 2004. Markowitz, A.M., "Raindrop Size Distribution Expressions", J. of Applied Meteorology, Vol. 15, 1976, pp. 1029-1031. Marshall, J. S., and Palmer, W. M., "The distribution of raindrops with size", J of Meteor., Vol. 5, Aug. 1948, pp.165-166. NBCC, Users Guide -to National Building Code of Canada 1995 Structural Commentaries (Part 4). Canadian Commission on Building and Fire Codes, National Research Council, Canada,1996. Rain Penetration Control Guide. CMHC Report, Ottawa, 2000. Rogers, R.R., and Pilie, R.J.,"Radar Measurements of Drop-Size Distribution", J. of the Atmospheric Sciences, Vol 10, Nov. 1962, pp. 503-506. Sandin, K., The Moisture Conditions in Aerated Lightweight Concrete Walls. Proc. of Symposium and Day of Building Physics, Lund University, Swedish Council for Building Research 1988, pp. 216220.. Schwarz, B., Witterungsbeansphruchung von Hochhausfassaden. HLH Bd. 24, Nr. 12, 1973, pp. 376-384. Skerlj, P.F., A Critical Assessment of the Driving Rain Wind Pressures Used in CSA Standard CAN/CSAA44-M90. M.A.Sc. Thesis, Faculty of Engineering Science, University of Western Ontario, London, Canada, 1999. Straube, J.F. Burnett, E.F.P., "Rain Control and Design Strategies". J. of Thermal Insulation and Building Envelopes, July, 1999, pp. 41-56.
BEG University of Waterloo Page 34 of 33

CMHC

Driving Rain Loads

Oct 2005

Straube, J.F. Moisture Control and Enclosure Wall Systems. Ph.D. Thesis, Civil Engineering Department, University of Waterloo, Waterloo, Canada, 1998. Straube, J.F., Burnett, E.F.P. "Driving Rain and Masonry Veneer", Water Leakage Through Building Facades, ASTM STP 1314, R.J. Kudder and J.L. Erdly, Eds., American Society for Testing and Materials, Philadelphia, 1997, pp. 73-87. Straube, J.F., Burnett, E.F.P., Rain Control and Screened Walls, Proc. of the 7th CSCE Building Science and Technology Conference, Toronto, March, 1997, pp. 17-38. Straube, J.F., Burnett, E.F.P.. "Pressure Moderation and Rain Control for Multi-Wythe Masonry Walls". Proc of International Building Physics Conference , Eindhoven, September 18-21, 2000, pp. 179-186. Surry, D., Skerlj, P. , Mikitiuk, M.J., An Exploratory Study of the Climatic Relationships between Rain and Wind, (Final Report BLWT-SS22-1994, Faculty of Engineering Science, University of Western Ontario), CMHC Research Report, Ottawa, February 1995. Van Mook, F.J.. Measurements and simulations of driving rain on the main building of the TUE Proc. Of 5th Symposium on Building Physics in the Nordic Countries, Goteborg, Sweden, August, 1999, pp. 377-384. Welsh, L.E., Skinner,, W.R., and Morris, R.J., A Climatology of Driving Rain Wind Pressures for Canada. Climate and Atmospheric Research Directorate, Draft Report, Environment Canada, Canada, 1989. Wisse, J.A., Driving Rain a numerical study, Proc. Of 9th Symp. On Building Physics and Building Climatology, Dresden, Germany, Sept. 1994.

BEG University of Waterloo

Page 35 of 33

Appendix A: Weather Data sources

Weather data was purchased from Environment Canada. Both CWEEDS and HLY03 databases were accessed. The CWEEDS files are computer data sets of hourly weather conditions occurring at 145 Canadian locations for up to 48 years of record, starting as early as 1953, and ending for most locations in 2001. The primary purpose of these files is to provide long term weather records for use in urban planning, siting and design of wind and solar renewable energy systems, and design of energy efficient buildings. The CWEEDS file and format were used to generate the CWEC (Canadian Weather Year for Energy Calculations) widely used for building thermal simulation. Most of the 21 weather elements such as temperature and wind speed have been abstracted directly from the National Digital Climate Archives maintained by the Atmospheric Environment Service (AES) of Environment Canada in Downsview, Ontario. Nine of the weather elements relate to solar irradiance amounts and have been estimated for each hour for those elements and locations for which observations are not available. The units of Wind speed (field 209) observations are 0.1 m/s. Wind speed is provided in km/h in the AES Climate Archives. The observation is an estimate of the one-minute mean wind speed on each hour for the years before 1985 and a two-minute mean wind speed thereafter. Care is required in assuming wind speed and direction representativeness at nearby locations. Wind speed is sensitive to the height above ground and exposure of the anemometer. Most of the anemometers at the locations in the CWEEDS files are mounted at 10m above ground in a flat, open exposure such as at airport locations. AES anemometers have not always been mounted at 10m above ground, especially before 1975. In general, significant caution is advisable in using wind speeds before 1975. Not only were the anemometers installed at heights other than 10m above ground more frequently before 1975 than after, but the station history files are often ambiguous as to anemometer height and location. Other non-standard anemometer locations, such as on top of aircraft hangers or the air traffic control tower, also occurred at some locations, mainly before 1975. Anemometer exposure is also problematic. Most anemometers are located in a flat, open exposure, especially at major airports. However, some anemometers are located in more exposed locations, such as Cape St. James, BC, which is on an exposed headland on the open coast. Observed wind speeds are not representative of less windy, inland locations, or even other coastal sites not on an exposed headland. Other locations, such as Fort Simpson, NWT, are sheltered by trees, and not representative of nearby, more exposed locations. In general, wind speeds form observing sites are only representative of of other nearby sites if the height above ground and exposures are similar. We also made use of the weather observations in element 78. 078 Occurrence of rain, rain showers or freezing rain

BEG University of Waterloo

Page A1

CMHC

Driving Rain Loads

Oct 2005

0 = None 1 = Light rain 2 = Moderate rain 3 = Heavy rain 4 = Light rain showers 5 = Moderate rain showers 6 = Heavy rain showers 7 = Light freezing rain 8 = Moderate or heavy freezing rain

We also wish to acknowledge the help provided by Maria Petrou, Data Specialist, National Archives Data Management, Atmospheric Monitoring & Water Survey Directorate, Meteorological Service of Canada, Environment Canada

BEG University of Waterloo

Page A2

Appendix B: Climate Statistics including Driving Rain

BEG University of Waterloo

Page B1

CMHC

Driving Rain Loads

Oct 2005

Calgary, AB Driving Rain (mm/yr) N 150 NNW NNE 125 NW 100 75 WNW 50 25 W 0 E ENE NE

WSW

ESE

SW
For an unobstructured plane perpendicular to wind at 10 m above grade.

SE SSW S SSE

Charlottetown, PEI Driving Rain (mm/yr) N 180 NNW NNE 160 NW 140 120 100 WNW 80 60 40 20 W 0 E ENE NE

WSW

ESE

SW
For an unobstructured plane perpendicular to wind at 10 m above grade.

SE SSW S SSE

BEG University of Waterloo

CMHC

Driving Rain Loads

Oct 2005

Edmonton, AB Driving Rain (mm/yr) N 150 NNW NNE 125 NW 100 75 WNW 50 25 W 0 E ENE NE

WSW

ESE

SW
For an unobstructured plane perpendicular to wind at 10 m above grade.

SE SSW S SSE

Fredericton, NB Driving Rain (mm/yr) N 150 NNW NNE 125 NW 100 75 WNW 50 25 W 0 E ENE NE

WSW

ESE

SW
For an unobstructured plane perpendicular to wind at 10 m above grade.

SE SSW S SSE

BEG University of Waterloo

CMHC

Driving Rain Loads

Oct 2005

Winnipeg, MB Driving Rain (mm/yr) N 50 NNW NNE 45 40 NW 35 30 25 WNW 20 15 10 5 W 0 E ENE NE

WSW

ESE

SW
For an unobstructured plane perpendicular to wind at 10 m above grade.

SE SSW S SSE

Montreal, QC Driving Rain (mm/yr) N 150 NNW NNE 125 NW 100 75 WNW 50 25 W 0 E ENE NE

WSW

ESE

SW
For an unobstructured plane perpendicular to wind at 10 m above grade.

SE SSW S SSE

BEG University of Waterloo

CMHC

Driving Rain Loads

Oct 2005

North Bay, ON Driving Rain (mm/yr) N 150 NNW NNE 125 NW 100 75 WNW 50 25 W 0 E ENE NE

WSW

ESE

SW
For an unobstructured plane perpendicular to wind at 10 m above grade.

SE SSW S SSE

Ottawa, ON Driving Rain (mm/yr) N 150 NNW NNE 125 NW 100 75 WNW 50 25 W 0 E ENE NE

WSW

ESE

SW
For an unobstructured plane perpendicular to wind at 10 m above grade.

SE SSW S SSE

BEG University of Waterloo

CMHC

Driving Rain Loads

Oct 2005

PrinceGeorge, BC Driving Rain (mm/yr) N 150 NNW NNE 125 NW 100 75 WNW 50 25 W 0 E ENE NE

WSW

ESE

SW
For an unobstructured plane perpendicular to wind at 10 m above grade.

SE SSW S SSE

Quebec, QC Driving Rain (mm/yr) N 300 NNW NNE 250 NW 200 150 WNW 100 50 W 0 E ENE NE

WSW

ESE

SW
For an unobstructured plane perpendicular to wind at 10 m above grade.

SE SSW S SSE

BEG University of Waterloo

CMHC

Driving Rain Loads

Oct 2005

SaintJohn, NB Driving Rain (mm/yr) N 250 NNW NNE 200 NW 150 WNW 100 50 W 0 E ENE NE

WSW

ESE

SW
For an unobstructured plane perpendicular to wind at 10 m above grade.

SE SSW S SSE

Saskatoon, SK Driving Rain (mm/yr) N 50 NNW NNE 40 NW 30 WNW 20 10 W 0 E ENE NE

WSW

ESE

SW
For an unobstructured plane perpendicular to wind at 10 m above grade.

SE SSW S SSE

BEG University of Waterloo

CMHC

Driving Rain Loads

Oct 2005

Shearwater, NS Driving Rain (mm/yr) N 350 NNW NNE 300 NW 250 200 WNW 150 100 50 W 0 E ENE NE

WSW

ESE

SW
For an unobstructured plane perpendicular to wind at 10 m above grade.

SE SSW S SSE

St Johns, NF Driving Rain (mm/yr) N 350 NNW NNE 300 NW 250 200 WNW 150 100 50 W 0 E ENE NE

WSW

ESE

SW
For an unobstructured plane perpendicular to wind at 10 m above grade.

SE SSW S SSE

BEG University of Waterloo

CMHC

Driving Rain Loads

Oct 2005

Sydney, NS Driving Rain (mm/yr) N 300 NNW NNE 250 NW 200 150 WNW 100 50 W 0 E ENE NE

WSW

ESE

SW
For an unobstructured plane perpendicular to wind at 10 m above grade.

SE SSW S SSE

Toronto ON Driving Rain (mm/yr) N 150 NNW NNE 125 NW 100 75 WNW 50 25 W 0 E ENE NE

WSW

ESE

SW
For an unobstructured plane perpendicular to wind at 10 m above grade.

SE SSW S SSE

BEG University of Waterloo

CMHC

Driving Rain Loads

Oct 2005

Vancouver, BC Driving Rain (mm/yr) N 500 NNW NNE 400 NW 300 WNW 200 100 W 0 E ENE NE

WSW

ESE

SW
For an unobstructured plane perpendicular to wind at 10 m above grade.

SE SSW S SSE

Winnipeg, MB Driving Rain (mm/yr) N 50 NNW NNE 45 40 NW 35 30 25 WNW 20 15 10 5 W 0 E ENE NE

WSW

ESE

SW
For an unobstructured plane perpendicular to wind at 10 m above grade.

SE SSW S SSE

BEG University of Waterloo

Appendix C: Driving Rain on Vertical Building Surfaces

BEG University of Waterloo

Page C1

CMHC

Driving Rain Loads

Oct 2005

Calgary, AB Driving Rain on Building (kg/m2/yr) N NNW NW 700 600 500 400 WNW 300 200 100 W 0 E ENE NE NNE

WSW

ESE

SW
For wall of a rectangular three-storey building facing direction shown. Consult report for geometric, exposure, height, and topography correction factors.

SE SSW S SSE
Annual average based on 1965-1989

Charlottetown, PEI Driving Rain on Building (kg/m2/yr) N NNW NW 700 600 500 400 WNW 300 200 100 W 0 E ENE NE NNE

WSW

ESE

SW
For wall of a rectangular three-storey building facing direction shown. Consult report for geometric, exposure, height, and topography correction factors.

SE SSW S SSE
Annual average based on 1965-1989

BEG University of Waterloo

CMHC

Driving Rain Loads

Oct 2005

Edmonton, AB Driving Rain on Building (kg/m2/yr) N NNW NW 700 600 500 400 WNW 300 200 100 W 0 E ENE NE NNE

WSW

ESE

SW
For wall of a rectangular three-storey building facing direction shown. Consult report for geometric, exposure, height, and topography correction factors.

SE SSW S SSE
Annual average based on 1965-1989

Fredericton, NB Driving Rain on Building (kg/m2/yr) N NNW NW 700 600 500 400 WNW 300 200 100 W 0 E ENE NE NNE

WSW

ESE

SW
For wall of a rectangular three-storey building facing direction shown. Consult report for geometric, exposure, height, and topography correction factors.

SE SSW S SSE
Annual average based on 1965-1989

BEG University of Waterloo

CMHC

Driving Rain Loads

Oct 2005

Montreal, QC Driving Rain on Building (kg/m2/yr) N NNW NW 700 600 500 400 WNW 300 200 100 W 0 E ENE NE NNE

WSW

ESE

SW
For wall of a rectangular three-storey building facing direction shown. Consult report for geometric, exposure, height, and topography correction factors.

SE SSW S SSE
Annual average based on 1965-1989

North Bay, ON Driving Rain on Building (kg/m2/yr) N NNW NW 700 600 500 400 WNW 300 200 100 W 0 E ENE NE NNE

WSW

ESE

SW
For wall of a rectangular three-storey building facing direction shown. Consult report for geometric, exposure, height, and topography correction factors.

SE SSW S SSE
Annual average based on 1965-1989

BEG University of Waterloo

CMHC

Driving Rain Loads

Oct 2005

Ottawa, ON Driving Rain on Building (kg/m2/yr) N NNW NW 700 600 500 400 WNW 300 200 100 W 0 E ENE NE NNE

WSW

ESE

SW
For wall of a rectangular three-storey building facing direction shown. Consult report for geometric, exposure, height, and topography correction factors.

SE SSW S SSE
Annual average based on 1965-1989

PrinceGeorge, BC Driving Rain on Building (kg/m2/yr) N NNW NW 700 600 500 400 WNW 300 200 100 W 0 E ENE NE NNE

WSW

ESE

SW
For wall of a rectangular three-storey building facing direction shown. Consult report for geometric, exposure, height, and topography correction factors.

SE SSW S SSE
Annual average based on 1965-1989

BEG University of Waterloo

CMHC

Driving Rain Loads

Oct 2005

Quebec, QC Driving Rain on Building (kg/m2/yr) N NNW NW 700 600 500 400 WNW 300 200 100 W 0 E ENE NE NNE

WSW

ESE

SW
For wall of a rectangular three-storey building facing direction shown. Consult report for geometric, exposure, height, and topography correction factors.

SE SSW S SSE
Annual average based on 1965-1989

SaintJohn, NB Driving Rain on Building (kg/m2/yr) N NNW NW 1000 900 800 700 600 500 WNW 400 300 200 100 W 0 E ENE NE NNE

WSW

ESE

SW
For wall of a rectangular three-storey building facing direction shown. Consult report for geometric, exposure, height, and topography correction factors.

SE SSW S SSE
Annual average based on 1965-1989

BEG University of Waterloo

CMHC

Driving Rain Loads

Oct 2005

Saskatoon, SK Driving Rain on Building (kg/m2/yr) N NNW NW 700 600 500 400 WNW 300 200 100 W 0 E ENE NE NNE

WSW

ESE

SW
For wall of a rectangular three-storey building facing direction shown. Consult report for geometric, exposure, height, and topography correction factors.

SE SSW S SSE
Annual average based on 1965-1989

Shearwater, NS Driving Rain on Building (kg/m2/yr) N NNW NW 1000 900 800 700 600 500 WNW 400 300 200 100 W 0 E ENE NE NNE

WSW

ESE

SW
For wall of a rectangular three-storey building facing direction shown. Consult report for geometric, exposure, height, and topography correction factors.

SE SSW S SSE
Annual average based on 1965-1989

BEG University of Waterloo

CMHC

Driving Rain Loads

Oct 2005

St Johns, NF Driving Rain on Building (kg/m2/yr) N NNW NW 1000 900 800 700 600 500 WNW 400 300 200 100 W 0 E ENE NE NNE

WSW

ESE

SW
For wall of a rectangular three-storey building facing direction shown. Consult report for geometric, exposure, height, and topography correction factors.

SE SSW S SSE
Annual average based on 1965-1989

Sydney, NS Driving Rain on Building (kg/m2/yr) N NNW NW 1000 900 800 700 600 500 WNW 400 300 200 100 W 0 E ENE NE NNE

WSW

ESE

SW
For wall of a rectangular three-storey building facing direction shown. Consult report for geometric, exposure, height, and topography correction factors.

SE SSW S SSE
Annual average based on 1965-1989

BEG University of Waterloo

CMHC

Driving Rain Loads

Oct 2005

Toronto, ON Driving Rain on Building (kg/m2/yr) N NNW NW 700 600 500 400 WNW 300 200 100 W 0 E ENE NE NNE

WSW

ESE

SW
For wall of a rectangular three-storey building facing direction shown. Consult report for geometric, exposure, height, and topography correction factors.

SE SSW S SSE
Annual average based on 1965-1989

Vancouver, BC Driving Rain (mm/yr) N 500 NNW NNE 400 NW 300 WNW 200 100 W 0 E ENE NE

WSW

ESE

SW
For an unobstructured plane perpendicular to wind at 10 m above grade.

SE SSW S SSE

BEG University of Waterloo

CMHC

Driving Rain Loads

Oct 2005

Victoria, BC Driving Rain on Building (kg/m2/yr) N NNW NW 700 600 500 400 WNW 300 200 100 W 0 E ENE NE NNE

WSW

ESE

SW
For wall of a rectangular three-storey building facing direction shown. Consult report for geometric, exposure, height, and topography correction factors.

SE SSW S SSE
Annual average based on 1965-1989

Winnipeg, MB Driving Rain on Building (kg/m2/yr) N NNW NW 700 600 500 400 WNW 300 200 100 W 0 E ENE NE NNE

WSW

ESE

SW
For wall of a rectangular three-storey building facing direction shown. Consult report for geometric, exposure, height, and topography correction factors.

SE SSW S SSE
Annual average based on 1965-1989

BEG University of Waterloo