1

Characterization of Stack Effect in High-Rise Buildings under

Winter Conditions, Including the Impact of Stairwell Pressurization

Steven Strege and Michael Ferreira

Hughes Associates Inc., Baltimore, MD, USA

Abstract

To characterize the magnitude of stack effect within stairwells and elevator shafts, differential pressure

measurements were taken in fifteen (15) high-rise buildings in four (4) different cities (Cleveland,

Baltimore, Minneapolis, and Philadelphia) during the winter months of January – March, 2013. Test

buildings ranged in height from 143 – 492 ft (44 – 150 m). Outside temperatures during testing ranged

from 10 – 59°F (-12 – 15°C). There was evidence of winter stack present in all buildings tested based on

the differential pressures measured. Under winter conditions, the data suggests that potentially large

quantities of air can migrate, floor-to-floor, via unprotected elevator shafts. Data further suggests

activation of the stairwell pressurization system may cause an increase in vertical air movement via

unprotected elevator shafts. This behavior may impact the movement of smoke floor-to-floor in a fire

event.

The exterior stack force on the building’s envelope does not always translate proportionally to shaft-to-

building differential pressures, as each building is unique. Although a building’s height and climate

(outside temperature) play important roles in determining vertical airflow movement within a building,

other variables such as architectural layout, wind, and ventilation systems may impact shaft-to-building

differential pressures. Hand calculations do not treat the building as a complete system, accounting for all

variables involved. A comparison between measured and hand calculated stack forces show that hand

calculations should be used only for first-order approximations, using conservative building envelope

leakage assumptions.

Keywords

Stack Effect, Smoke Control, Stairwell Pressurization, Elevator Pressurization, Elevator Lobby

2

Introduction

Stack effect is the vertical airflow within buildings caused by the temperature-created density differences

between the building interior and exterior or between two interior spaces [1]. Stack effect in high-rise

buildings under extreme temperature conditions can be a primary driver of floor-to-floor smoke spread

via vertical openings within the building, especially stairwells and elevators shafts.

Pressure differential testing was conducted to characterize vertical movement of air in buildings under

winter conditions. The purpose of the testing was to characterize the magnitude of the stack effect, within

stairwells and elevator shafts, for a variety of buildings, both in type (office, hotel) and height. To

characterize the magnitude of the stack effect, differential pressures measurements were taken in fifteen

(15) high-rise buildings, ranging in height from 143 – 492 ft (44 – 150 m), during the winter months of

January – March, 2013. When possible, data on the impact of stairwell pressurization system activation

on elevator-to-building airflows, was recorded. Measured differential pressures are compared with hand

calculations provided in the Handbook of Smoke Control Engineering [2].

Test Buildings

Differential pressure measurements were taken in fifteen (15) high-rise buildings in four (4) different

cities (Cleveland, Baltimore, Minneapolis, and Philadelphia) during the winter months of January –

March, 2013. Table 1 describes the buildings visited for this study, including: location, building height,

number of floors, year built, occupancy type, exterior wall construction, number of stairs, number of

elevator hoistways, stair door undercut gap, elevator door width, and the approximate elevator door

leakage area. Building names have been omitted from this article to protect the interests of the property

owners.

3

Table 1. Summary of Buildings Tested

As shown in Table 1, the test buildings in this study ranged in height from 143 – 492 ft (44 – 150 m). In

some cases the building height could not be accurately obtained. In Table 1, building height values

shown in italics (with asterisks) are estimated, assuming a floor elevation of 13 feet (4 m), based on the

average floor-to-floor elevation heights of the other test buildings. The floor to floor heights for most of

the test buildings ranged between 11 – 14 feet (3.4 – 4.3 m). Building 3 had unusually tall floor heights,

nearing 19.5 feet (6 m) on average. Outside temperatures measured during testing ranged from 10 – 59°F

(-12 – 15°C).

The type of buildings examined in this study included high-rise residential hotels and office buildings.

The majority of the test buildings (12 of 15) consisted of high-rise residential hotels, due to their ease of

access. High-rise office buildings proved difficult to access for the purposes of obtaining test data, due to

increased security. Generally, most of the buildings in this study had an interior layout consisting of a

central corridor connecting stairwells and an elevator bank with rooms (or offices) located around the

corridor’s perimeter.

The majority of the test buildings’ exterior walls (envelope) consisted of a glass curtain wall or masonry

construction. All buildings had non-operable exterior windows expect for Building 12 and, in-part,

Building 6. Building 12 in Philadelphia (see Table 1) provides an important contrasting data point, where

each hotel room contained exterior balconies with sliding glass doors. The original construction year of

the test buildings range from 1922 – 2011, however most of the older buildings have been renovated and

it was unclear if renovations included substantial changes to the building’s exterior walls.

Most test buildings contained two stairwells (Building 14 had four stairs). Stair door leakages ranged

considerably, with undercut gaps ranging between 0 (gasketed) – 1.0 inch (0 – 2.5 cm). The number of

elevator hoistways in each building ranged from 3 – 10. Measured elevator door leakage ranged from

Building City, State Height (ft) Floors Year Occupancy Type Exterior Wall ConstructionNumber

of Stairs

Number of

Elevator

Hoistways

Stair Door

Undercut Gap

Elevator Door

Width (ft)

Elevator Door

Leakage (ft2)

1 320 25 1991 Residential Hotel 2 4 tight 3.5 0.60

2 419 31 1967 2 3 1/8" 4 0.58

3 430 22 2002 2 8 1/8" to 1" 4 0.58

4 272 22 1922 2 6 1/8" to 1" 4 0.74

5 399 32 2001 2 3 1/16" to 3/4" 3.5 0.51

6 221* 17 2011Glass Curtain Wall, Mostly

Non-Operable Windows

(Operable in l imited rooms)

2 6 1/16" to 1/2" 4 0.59

7 381 31 1983 2 5 1/4" to 1" 4 0.58

8 221* 17 1987 2 N/A 1/16" to 1/2" 4 0.52

9 227 19 1972Glass Curtain Wall

(Non-Operable Windows)2 N/A 1/16" to 3/4" 4 0.52

10 143* 11 1941 2 N/A 1/4" to 1/2" 4 0.74

11 492 33 1932 2 6 tight to 1" 3.5 0.59

12 364* 28 1964Masonry and Sliding Glass

Doors to Exterior Balconies

(Operable)

2 3 1/4" to 1/2" 3.25 0.49

13 394 32 2009 2 6 1/4" to 3/8" 3.5 0.48

14 299* 23 1995 4 10 1/4" to 3/8" 3.5 0.63

15 195* 15 1926 2 6 1/4" to 3/8" 3.25 0.65

Cleveland, OH

Baltimore, MD

Minneapolis, MN

Philadelphia, PA

Masonry and Fixed Glass

Windows (Non-Operable)

Masonry and Fixed Glass

Windows (Non-Operable)

Masonry and Fixed Glass

Windows (Non-Operable)

Masonry and Fixed Glass

Windows (Non-Operable)

Business

Residential Hotel

4

0.48 – 0.74 ft2/door, which is consistent with values reported in Handbook of Smoke Control Engineering

[2].

Measurements

For each building tested, the following measurements were taken (when possible):

1. Differential pressure between stair and building (in. W.G.)

a. On 1 or 2 upper levels (near top of building).

b. On 1 or 2 lower levels (ground floor).

2. Differential pressure between elevator hoistway and building (in. W.G.)

a. On 1 or 2 upper levels (near top of building).

b. On 1 or 2 lower levels (ground floor).

3. Differential pressure between exterior and building at ground floor doors (in. W.G.)

4. Outside temperature at the time of testing (°F)

5. Building temperature at the time of testing (°F)

6. Number of stairwells and elevator hoistways present

7. Stairwell door undercut gaps (in.)

8. Approximation of total leakage at each elevator door (ft2)

A calibrated hand-held differential pressure gauge (TSI model 8702) was used to take all differential

pressure measurements. For many measurement locations, a range of differential pressures were

observed. This was especially true for elevator hoistway measurements, when elevator cars were in

motion (as measurements were taken during normal building activity). Best efforts were made to take

average differential pressure measurements when all elevator doors were closed and elevator cars were

stationary. Likewise, best efforts were made to take stairwell measurements when all stair doors were

closed.

In an attempt to isolate the impact of outside air temperature on shaft-to-building differential pressures,

repeat measurements were conducted on the same building for different outside temperatures. Repeat

measurements were conducted in Buildings 1, 2, 3, and 11.

When possible, data on the impact of stair pressurization on airflow movement within the elevator shafts

was recorded. This was done by taking differential pressures in the elevator shafts (with respect to the

building), on the top occupied floor of the building, with the stair pressurization system both on and off.

Measurement Considerations

The driving forces governing air movement in a building include naturally occurring stack effect, wind

effects, fan-powered ventilation systems, and elevator piston effect [3]. It is important to note that shaft-

to-building differential pressures recorded during this study likely included all of these driving forces, at

least to some degree. Although stack effect is expected to be one of the primary contributors, the

magnitude of the other driving forces is not well characterized and should be considered when analyzing

the pressure differential data. Nevertheless, the differential pressures stated in this report are actual

values, measured under normal building activity, regardless of the driving force (or combination of

driving forces) producing the result.

5

Testing in all fifteen building was conducted while the building was under a normal HVAC mode with

elevator lobby doors open. In three of the test buildings (Building 3, 5, and 7), elevator-to-shaft testing

was also conducted with the stairwell pressurization systems activated. It should be noted that stair and

elevator shaft differential pressures during a fire alarm event may be considerably different from those

measured in this study, due to changes in active and passive building systems (ex. stair pressurization and

elevator lobby doors).

Stack Effect Measurement Results

Table 2 provides a summary of the measurements taken at each building, including: outside temperature,

inside temperature, temperature delta, exterior-to-building differential pressure at ground level, stair-to-

building differential pressure on lower/upper levels, and elevator-to-building differential pressure on

lower/upper levels. Positive differential pressure values indicate flow from the shaft (or exterior) into the

building. Negative differential pressure values indicate flow from the building into the shaft (stair or

elevator).

As shown in Table 2, there was evidence of winter stack present in all buildings tested based on the

differential pressures measured. On the lower levels of all buildings, air was observed flowing from the

building into the stairwells and elevator hoistways. Pressure differential magnitudes on the lower levels

for the stairwells ranged from -0.011 to -0.093 in. W.G. (-0.044 in. W.G. average). Likewise, pressure

differential magnitudes on the lower levels for the elevator hoistways ranged between -0.012 to -0.100 in.

W.G. (-0.052 in. W.G. average).

On the upper levels of most buildings (except Building 6 and 7), air was observed flowing from the

stairwells and elevator hoistways into the building. Pressure differential magnitudes on the upper levels

for the stairwells ranged from -0.006 to 0.135 in. W.G. (0.041 in. W.G. average). Pressure differential

magnitudes on the upper levels for the elevator hoistways ranged from -0.008 to 0.140 in. W.G. (0.039 in.

W.G average). In Building 6 and 7, it is likely that pressurized corridors caused air to flow into the stair

and elevator hoistways on the upper levels of the building. A more detailed study was conducted on

Building 7, were stair-to-building pressure differentials were measured on every floor. Air was measured

flowing out of the stairwells into the building on Floors 7 – 18 (middle levels) at differential pressures as

high as 0.012 in. W.G. Air was also measured flowing from the stairs to outside, through roof access

doors, at differential pressures between 0.509 – 0.607 in. W.G.

6

Table 2. Summary of Building Differential Pressure Measurements

All exterior-to-building differential pressure measurements (except Building 14) indicate a winter stack

force was present as air flowed from the exterior into the building on the lower levels. Excluding

Building 14, exterior-to-building pressure differential magnitudes on the ground floor ranged from 0.051

to 0.380 (0.153 average). In Building 14, it is likely that an over-pressurized lobby and/or negative wind

pressures on the side of the building the doors were located on caused air to flow out of the building on

the ground level under winter conditions.

Discussion of Stack Effect Measurements

Results show there is a relatively strong correlation between measured exterior-to-building differential

pressures (at the ground level) and the building’s height and outside temperature. However, the exterior

stack force exerted on the building’s envelope does not always translate proportionally to shaft-to-

building differential pressures. How exterior-to-building pressures translate to shaft-to-building pressures

is a function of the architectural and mechanical features of the building. Architectural features include

Exterior-to-Building

(Ground Level)Stair-to-Building

Elevator-to-

Building

Stair-to-

Building

Elevator-to-

Building

320 12 74 62 0.380 -0.025 -0.021 0.022 0.027

320 28 74 46 0.280 -0.040 -0.016 0.028 0.034

419 12 73 61 0.300 -0.059 -0.070 0.135 0.140

419 28 73 45 0.370 -0.050 -0.088 0.078 0.053

430 28 71 43 0.130 -0.050 -0.100 0.0800.085

(0.170)

430 59 71 12 0.080 -0.031 N/A 0.037 N/A

4 272 16 73 57 0.190 -0.093 -0.093 0.074 0.085

5 399 30 69 39 0.130 -0.038-0.090

(-0.100)0.055

0.030

(0.045)

6 221 34 72 38 0.051 -0.011 -0.012 -0.006 -0.008

7 381 10 77 67 0.140 -0.046 -0.055 -0.002-0.002

(0.005)

8 221 18 72 54 0.120 -0.031 -0.075 0.038 0.030

9 227 18 74 56 0.100 -0.028 -0.048 0.032 0.035

10 143 18 74 56 0.125 -0.014 -0.032 0.010 0.025

492 48 75 27 0.077 -0.085 -0.068 0.035 0.021

492 39 75 36 0.083 -0.078 -0.084 0.040 0.027

12 364 52 75 23 0.077 -0.051 -0.048 0.067 0.080

13 394 51 74 23 0.065 -0.055 -0.033 0.047 0.031

14 299 48 75 27 -0.025 -0.018 -0.025 0.002 0.004

15 195 49 74 25 0.052 -0.040 -0.023 0.015 0.004

143 10 69 12 -0.025 -0.093 -0.100 -0.006 -0.008

492 59 77 67 0.380 -0.011 -0.012 0.135 0.140

339 31 73 42 0.143 -0.044 -0.052 0.041 0.039

Positive dP values indicate Flow Into Building

Negative dP values Indicate Flow Into Shaft

Repeat measurements for different outside temperature

dP values in ( ) indicate stair pressurization active

Upper Levels, dP (in. W.G.)

1

2

3

11

Cleveland, OH

Baltimore, MD

Minneapolis, MN

Philadelphia, PA

Temp.

Delta (F)

Lower Levels, dP (in. W.G.)

Minimum

Maximum

Average

Building City, State Height (ft)Outside

Temp. (F)

Building

Temp. (F)

7

building layout, internal airflow paths and envelope leakages. A building’s ventilation system can also

play a critical role on shaft-to-building differential pressures [3]. The impact of ventilation systems was

not included in this study.

Impact of Architectural Layout

The architectural layout of a building plays a critical role in determining how the exterior stack force on

the building is translated to the interior vertical shafts within the building. The network of airflow paths

within a building acts as a “system” and each building has a unique network. Small changes to the

architectural layout of a building can impact shaft-to-building difference pressures. Take the repeat

measurements of Building 2 for example. Notice as the temperature dropped from 28°F to 12°F, the

exterior-to-building force dropped slightly from 0.370 to 0.300, which is counterintuitive. As expected,

elevator-to-building differential pressures on the upper floors increased with a drop in outside temperature

from 0.053 to 0.140 in. W.G. As expected, stair-to-building differential pressures on the upper floors also

increased from 0.078 to 0.135 in. W.G. However, the primary reason for this substantial pressure

increase, between the two scenarios, is not the change in outside temperature, but is due to a change in the

configuration of a set of corridor double doors that open into an open floor layout. In the lower pressure

scenario, the corridor doors are closed and in the higher pressure scenario, the corridor doors are open.

Opening the corridor doors (which open to a large open floor layout) reduced the flow resistance on that

level, resulting in higher elevator-to-building differential pressures (and vertical airflow movement). This

is analogous to removal of a central corridor. A CONTAM model [4] of Building 2 was constructed and

validates this dynamic. This is just one example of how the architectural layout of a building can play a

critical role in shaft-to-building differential pressure measurements.

The building data set is limited in that all of the buildings tested contained a central corridor connecting

the stairwells and elevator shafts. Although this is a common building layout, other common building

layouts such as buildings having a largely open floor plan were not studied. As previously mentioned,

CONTAM modeling shows these centralized corridors provide airflow resistance (or back-pressure)

decreasing shaft-to-building differential pressures. In comparison, differential pressures at stair/elevator

doors may be much higher on floors with open floor plans as there is less resistance to airflow exiting the

shafts.

Impact of Building Envelope Leakage

A building’s envelope leakage also plays a critical role on shaft-to-building differential pressures. This

dynamic is well documented in the literature [2] and is also supported by the measured data set. For

example, Building 12 had one of the highest measured elevator-to-building differential pressures on the

upper floors, even though the building height was average for the data set (364 ft) and the outside

temperature was relatively warm at 52°F. The likely reason for relatively high shaft-to-building

differential pressures in Building 12 is due to the relatively loose envelope of the building (as each hotel

room has exterior balconies with sliding glass doors).

Based on past modeling studies, stair and elevator hoistway connections to ambient may significantly

impact the leakage of a building’s envelope (and shaft-to-building differential pressures). These

connections include; discharge doors, roof access doors, barometric relief dampers and elevator hoistway

vents. The impact of these architectural components was not included in this study.

8

Impact of Stairwell Pressurization on Elevator Shaft Pressures

In Table 2, values in parentheses [()] indicates a differential pressure recorded while the stair

pressurization system was active. As shown in Table 2, for all three buildings when the stair

pressurization system was activated, elevator-to-building differential pressures on the upper floors

increased. The increase in elevator-to-building differential pressures is due to additional air entering the

elevator shafts via leakage air from the pressurized stairwells. The magnitude increase was more evident

in Building 3 compared to Buildings 5 and 7. This may be due to the differences in the architectural

layout of the central corridors and lobby areas on the lower levels. The ground level of Building 3

consists of relatively small/tight elevator lobbies with one set of exterior doors, whereas in Buildings 5

and 7 the lower levels consist of large, open areas with multiple exterior doors (providing more avenues

to relieve leakage air from the pressurized stairwells).

Inherent stack effect holds true if the stairwells (and other vertical shafts) remain at building temperature.

But when the stairwell pressurization systems are activated, unconditioned outside air is supplied to the

stairwells and the temperature within the stairwell begins to transition toward the temperature of the

outside air. As the stair shaft temperature changes over time, the pressure profile in the stair can invert

changing normal winter stack effect to a reverse stack effect in the stair. The opposite is possible under

summer conditions as well. This condition (cold outside, cool stair, warm building) creates competing

stack forces. It should be noted that in all three buildings for this study where the stair pressurization

system was activated, the stairwell temperatures remained relatively warm (>55°F) during pressurization

because the supply air was conditioned (per design) and/or because there were heaters located in the

stairwells. Therefore, the impact of cooling the stairwells during pressurization on elevator-to-building

airflow movement could not be fully characterized in this study. However, a CONTAM model [4] of

Building 2 shows that cooling the stairs with outside air during stair pressurization may produce increased

stack within the elevator hoistways.

The impact of stairwell pressurization on elevator-to-building airflows, in Building 3, was analyzed.

Based on elevator-to-building differential pressure measurements (with and without the stair

pressurization system active) and elevator door leakages, the total quantity of air (potentially containing

smoke in a fire event) flowing onto the uppermost building level from the elevators due to winter stack

was calculated, per Equation 1, for Building 3.

√ (1)

where:

Q = volumetric flow through orifice, cfm

Ae = area of orifice, ft2

= pressure difference across orifice, in. W.G.

K = coefficient, 2610

Results, per Equation 1, are provided in Table 3. As shown in Table 3, as expected winter stack effect

causes air flow to the upper floors of the building via the elevator shafts; approximately 3,500 cfm of

airflow for a nominal eight car shaft. Activation of the stair pressurization system increased elevator-to-

building airflow via the elevator shafts; approximately 5,000 cfm of airflow.

9

TABLE 3. Estimated Airflow Rate from Elevator Shafts to Upper Floor of Building

Due to Winter Stack and Stair Pressurization in Building 3

Calculated elevator-to-building airflows for Building 3, with and without the stairwell pressurization

activate, provides an interesting data point and suggests the impact of stairwell pressurization may be

considerable and warrants further studies.

Field Measurements Compared To Hand Calculations

The measured and calculated stack force exerted on a building’s envelope (which is a function of the

temperature delta and building height) is compared below. Calculated differential pressures are per

equations provided in the Handbook of Smoke Control Engineering [2]. The maximum shaft-to-exterior

differential pressure located at the bottom (or top) of a shaft is calculated per Equation 2 (and assumes

uniform shaft leakage and a neutral plane at mid-height):

(

) (2)

where:

= pressure difference from shaft to outside, in. W.G (Pa)

To = absolute temperature of outside air, °R (K)

Ts = absolute temperature of air inside shaft, °R (K)

h = distance above neutral plane, ft (m)

Ks = 7.64 (3460)

Figure 1 shows a comparison between the measured building-to-exterior differential pressures taken at the

ground level and the hand calculated shaft-to-exterior differential pressure, per Equation 1.

Stair Pressurization Off Stair Pressurization On

Elevator-to-Building dP (in. W.G.) 0.085 0.17

Number of Elevators 8 8

Area per Elevator Door (ft2) 0.58 0.58

Total Area (ft2) 4.64 4.64

AirFlow rate (cfm) 3,500 5,000

Building-to-Elevator Airflow (Winter Stack)

10

Figure 1. Maximum Building-to-Outside Differential Pressures – Measured vs. Hand Calculations (Eq. 1)

It is important to note that Figure 1 is not a direct comparison. Measured building-to-exterior pressures

were taken within the building’s lobby area on the ground level, not within the stairwell. As shown in

Figure 1, there is a fairly good agreement between measured and calculated build-to-exterior pressures,

but in most cases the measured pressure is lower than the calculated pressure (by 30% on average). This

is expected as the flow resistance (back-pressure) into a stairwell would generally be less than the flow

resistance into the building/lobby, which is commonly maintained at a slightly positive pressure.

Conversely, stairwells are at a neutral pressure. The architectural tightness of the stairwell compared to

the building’s lobby area also plays a role.

The measured and calculated shaft-to-building differential pressure, at the top of the shaft, is compared

below. Calculated shaft-to-building differential pressures are per Equation 3 [2]:

( ⁄ ) (3)

where:

= pressure difference from shaft to building, in. W.G (Pa)

= pressure difference from shaft to outside, in. W.G (Pa)

Asi = per floor leakage area between the shaft and the building, ft2 (m

2)

Aio = per floor leakage area between building and the outside ft2 (m

2)

Figure 2 shows a comparison between measured and calculated shaft-to-building differential pressures at

the top of the stairwells and elevator shafts, per Equation 3. As shown in Figure 2, the measured

pressures at the top of the stairwells and elevators shafts were comparable. In Equation 2, Asi is the

11

leakage area between the shaft and the building and Aio is the leakage area between the building and the

outside. As discussed in Handbook of Smoke Control Engineering, in general, the ratio Asi/Aio varies

from about 1.7 to 7 [2]. A ratio value of 7 would depict a very “tight” building envelope and a value of

1.7 would depict a very “loose” building envelope. Figure 2, shows the calculated maximum shaft-to-

building differential pressure when using an average Asi/Aio ratio value of 4.35. As shown, when

assuming an average Asi/Aio ratio value of 4.35 (which would be a reasonable assumption as a starting

point for the designing engineer) the measured pressures are considerably higher than calculated pressures

(by 250% on average).

Figure 2. Maximum Shaft-to-Building Differential Pressures – Measured vs. Hand Calculations (Eq. 2),

Asi/Aio = 4.35

For hand calculations, using a Asi/Aio ratio value of 2 provides a better match to the measured data, as

shown in Figure 3. However, this value is on the lower end of the range (1.7 to 7) suggested in literature

[2]. Considering that all test buildings (expect for Building 12) consisted of either fixed glass curtain

walls or masonry with fixed windows, the designing engineer may assume the building’s envelope is

“average” and therefore would use an average Asi/Aio ratio (4.35 to 7). This would result in

underestimated calculated pressures differences at the top of the stairwells and elevator shafts.

12

Figure 3. Maximum Shaft-to-Building Differential Pressures – Measured vs. Hand Calculations (Eq. 2),

Asi/Aio = 2.00

There are number of reason why a low Asi/Aio value provides a better match to the measured data,

including:

1. The measured data is impacted by all driving forces of air movement in the building including

naturally occurring stack effect, fan-powered ventilation systems, wind effects, and elevator

piston effect. The hand calculations consider naturally occurring stack effect only.

2. The measured data is impacted by the unique layout and network of airflow paths within each

building. The hands calculations consider a very basic building geometry with a uniform floor

plan, one vertical shaft with uniform leakage connecting all floors, and uniform building

envelope leakage.

Buildings may be initially considered “average” to “tight” based on an engineer’s consideration of the

tested leakage values for a building’s curtain wall system, when actually the envelope leakage is just as

much a function of the ground floor and roof level openings, elevator hoistway vents, barometric relief

dampers in stairs, and the system of ventilation shafts which by default terminates at the building’s

exterior via potentially large outside air (OA) and exhaust air (EA) grilles and louvers.

Based on the test data, conservative building envelope leakages (i.e. low Asi/Aio ratios) should be used for

hand calculations. This is even more important if the building contains operable exterior openings.

Building 12 for example, is a high-rise hotel with exterior balconies in each room. As shown in Figure 4,

an Asi/Aio ratio of 0.75 was needed to better match calculated pressures with those measured in the field.

13

Figure 4: Maximum Shaft-to-Building Differential Pressures – Measured vs. Hand Calculations (Eq. 2),

Asi/Aio = 0.75

The stack force exerted on a building’s envelope under winter conditions (Figure 1) does not

always translate proportionally for shaft-to-building differential pressures (Figures 2 – 4).

Although a building’s height and climate (outside temperature) play important roles in

determining vertical airflow movement within a building, other variables such as architectural

layout, architectural leakage, wind effects, and ventilation systems should all be considered.

Hand calculations do not treat the building as a complete system; accounting for all variables

involved. Therefore, hand calculations may likely result in inaccurate shaft-to-building

differential pressure predictions. Based on this analysis, hand calculations to predict stack

effect airflows should be used for first-order approximations only, using conservative building

envelope leakage (Asi/Aio ratio) assumptions.

Summary

To characterize the magnitude of stack effect within stairwells and elevator shafts, differential pressure

measurements were taken in fifteen (15) high-rise buildings in four (4) different cities (Cleveland,

Baltimore, Minneapolis, and Philadelphia) during the winter months of January – March, 2013. Test

buildings ranged in height from 143 – 492 ft. Outside temperatures during testing ranged from 10 – 59°F

(-12 – 15°C).

14

There was evidence of winter stack present in all buildings tested based on the differential pressures

measured. On the lower levels of all buildings, air was observed flowing from the building into the

stairwells and elevator hoistways with pressure differential magnitudes ranging between -0.011 to -0.100

in. W.G. (-0.048 in. W.G. average). Similarly, in most buildings (except Building 6 and 7) air was

observed flowing from the stair and elevator hoistways into the building on the upper levels with pressure

differential magnitudes ranging from -0.006 to 0.140 in. (0.040 in. W.G. average).

The shaft-to-building pressure differentials measured in this study suggest potentially large quantities of

air can migrate, floor-to-floor, via unprotected elevator shafts. More air may travel vertically in buildings

than is predicted by the simple hand calculations due to the complex interaction of forces impacting air

flow. Limited data collected, which was supported by CONTAM simulations, suggests that activation

of a stair pressurization system may significantly increase vertical air flow via unprotected elevator

shafts, particularly when unconditioned pressurization air is brought into the stairs and begins to lower

the temperature of the stairs relative to the building’s temperature.

The exterior-to-building stack force, under winter conditions, does not always translate proportionally

for shaft-to-building differential pressures. Although a building’s height and climate (outside

temperature) play important roles in determining vertical airflow movement within a building, other

variables such as architectural layout, leakage, wind effects, and ventilation systems should all be

considered. Hand calculations do not treat the building as a complete system; accounting for all variables

involved. Therefore, hand calculations may result in inaccurate shaft-to-building differential pressure

predictions. A comparison between measured and hand calculated stack forces show that hand

calculations should be used for first-order approximations only, using conservative building envelope

leakage (Asi/Aio ratio) assumptions.

Future studies should look to expand on this data set. Interesting data points would include warm

climates, buildings with operable openings, buildings with open floor layouts, and the impact of stairwell

pressurization systems (especially for systems that provide unconditioned air to the stairwells).

References

1. NFPA 92 (2012), Standard for Smoke Control Systems, National Fire Protection Association, Quincy,

Massachusetts, 2012.

2. Klote, J. H., Milke, J. A., Turnbull, P. G., Kashef, A., Ferreira M. J., (2012), Handbook of Smoke

Control Engineering, American Society for Heating Refrigeration and Air Conditioning Engineers,

Atlanta, GA, 2012.

3. Mowrer, F. W., Milke, J. A., Torero, J. L., (2004) “A Comparison of Driving Force for Smoke

Movement in Building,” Journal of fire Protection Engineering, Sage Publications, London, Vol. 14,

No. 4, 2004.

4. Walton, G. N. and Dols, Stuart (2005), “CONTAM 2.4 User Guide and Program Documentation,”

NISIR 7251, National Institute of Standards and Technology, November 2005.