ctbuh.

org/papers

Title: Elevator Pressurization in Tall Buildings

Author: John H. Klote, Smoke and Fire Consulting

Subjects: Fire & Safety

Vertical Transportation

Keywords: Fire Safety

Vertical Transportation

Publication Date: 2013

Original Publication: International Journal of High-Rise Buildings Volume 2 Number 4

Paper Type: 1. Book chapter/Part chapter

2. Journal paper
3. Conference proceeding
4. Unpublished conference paper
5. Magazine article
6. Unpublished

Council on Tall Buildings and Urban Habitat / John H. Klote

International Journal of High-Rise Buildings
International Journal of
December 2013, Vol 2, No 4, 341-344
High-Rise Buildings
www.ctbuh-korea.org/ijhrb/index.php

Elevator Pressurization in Tall Buildings

John H. Klote
Smoke and Fire Consulting, 19355 Cypress Ridge Terrace, Leesburg, VA 20176, USA

Abstract

During a building fire, smoke can flow through elevator shafts threatening life on floors remote from the fire. Many buildings
have pressurized elevators intended to prevent such smoke flow. The computer program, CONTAM, can be used to analyze
the performance of pressurization smoke control systems. The design of pressurized elevators can be challenging for the
following reasons: (1) often the building envelope is not capable of effectively handling the large airflow resulting from elevator
pressurization, (2) open elevator doors on the ground floor tend to increase the flow from the elevator shaft at the ground floor,
and (3) open exterior doors on the ground floor can cause excessive pressure differences across the elevator shaft at the ground
floor. To meet these challenges, the following systems have been developed: (1) exterior vent (EV) system, (2) floor exhaust
(FE) system, and ground floor lobby (GFL) system.

Keywords: Elevator pressurization, Network modeling, Smoke control, Stairwell pressurization, Tall buildings

During a building fire, smoke can flow through eleva- struction. CONTAM is capable of simulating the pres-
tor shafts threatening life on floors remote from the fire. sures and flows throughout very large and complex buil-
To prevent such smoke flow, many buildings have pre- dings. Because CONTAM is a product of the US National
ssurized elevators or enclosed elevator lobbies, but enclo- Institute of Standards and Technology, it can be downloaded
sed elevator lobbies are not discussed in this article. at no cost (www.bfrl.nist.gov/IAQanalysis/CONTAM/do-
The focus of this article is design of pressurized eleva- wnload.htm). For specific information about using CON-
tors for tall buildings. The design of pressurized elevators TAM for smoke control applications see Chapter 14 of
can be a challenge and it can be more challenging in buil- the Handbook Smoke Control Engineering.
dings that also have pressurized stairwells. This article The primary purpose of CONTAM simulations is to
gives the reasons for this and describes kinds of elevator determine if a particular smoke control system in a par-
pressurization systems that can overcome these challenges. ticular building is capable of being balanced such that it
For extensive information about smoke control, see the will perform as intended. A secondary purpose is to pro-
Handbook Smoke Control Engineering (Klote et al. 2012). vide information to help size the system components such
Chapter 11 of this handbook is about elevator pressuri- as supply fans, exhaust fans, and vents. Design analysis of
zation, but this chapter does not address tall buildings. elevator pressurization needs to include any other pressu-
rization systems operating together as they would during
Design Analysis a building fire.
The flows in CONTAM are calculated as m = CA(2
Network analysis models are often used for design ana- p)1/2 where m is the mass flow through the path (kg/s),
lysis of pressurization smoke control systems. CONTAM C is the flow coefficient (dimensionless), A is the flow
(Walton and Dols 2005) is a network model that was area (m2), p is the pressure difference across path (Pa),
developed primarily for indoor air quality, but it has been and is the gas density in path (kg/m3). The user inputs
so extensively used for analysis of pressurization smoke values of C and A for flow paths, and the CONTAM
control systems that it has become the de facto standard. calculates the pressure differences and the flows. Table 1
This article discusses CONTAM, but it should be recog- lists values of C and A for doors. The flow areas are
nized that analysis with another network model is possi- relatively leaky so that calculated supply air would be
ble. In CONTAM, a building is represented as a network conservatively large, but other values may be appropriate
of spaces connected by flow paths such as gaps around for some projects.
closed doors, open doors, and cracks and holes in con- Table 2 lists some values of C and A for some other
flow paths with various leakage classifications. An engi-

Corresponding author: John H. Klote

neer cannot exactly know the flow areas in a particular
Tel: +571-258-0355; Fax: + building, but an educated guess can be made that allows
E-mail: [email protected] CONTAM simulations to determine if a specific smoke
342 John H. Klote | International Journal of High-Rise Buildings

Table 1. Flow coefficients and flow areas of doors

Flow Flow
Flow Path
Coefficient Area1 (m2)
Single door (closed) 0.65 0.023
Single door (opened) 0.35 2.0
Double door (closed) 0.65 0.045
Double door (opened) 0.35 3.9
Elevator door (closed) 0.65 0.06
Elevator door (opened) 0.65 0.56
1
Relatively leaky flow areas for closed doors were chosen so
that calculated supply air would be conservatively large.
Figure 1. The US Standard Atmosphere shows temperature
Table 2. Flow areas and flow coefficients of leakage paths and pressure decreasing with altitude.
Leakage Flow Flow Area
Flow Path
Classification Coefficient (m2 per m2)*
with (see Simulation - Airflow Numerics Properties of
Tight 0.5010-4
the help or user manual).
Average 0.1710-3
Walls 0.65
Loose 0.3510-3
Very Loose 0.1210-2
Shaft Temperature
Tight 0.6610-5 Elevator equipment has a typical range of operating
Floor
Average 0.65 0.5210-4 temperature. Often there is no effort to maintain this tem-
(or roof)
Loose 0.1710-3 perature for passenger elevators during building fires, be-
(m2 per m)** cause the elevators are taken out of service. In some loca-
Curtain Tight 0.00061 tions the elevator equipment may need to be maintained
0.65
wall gap Loose 0.0061 to give the fire service the option of using elevators
*This is m2 of leakage per m2 of wall, floor, or roof. during building fires.
**This is m2 of leakage per m length of curtain wall gap. Often supply air to elevator shafts is untreated such that
the temperature in pressurized shafts is close to the out-
Table 3. IBC Pressure difference criteria side temperature. The use of untreated air has the benefit
System Minimum p (Pa) Maximum p (Pa) on minimizing the adverse impact of stack effect. The
Pressurized shaft temperature can be expressed as TS = TO + (TB TO)
25 62 where TS is the temperature in the shaft, TO is the tem-
Elevators
Pressurized perature outdoors, TB is the temperature in the building,
25 87
Stairwells and is a dimensionless heat transfer factor. As with pre-
ssurized stairwells, there has been little research conduc-
control system in the building is capable of being balan- ted on the heat transfer factor for elevator shafts. For
ced such that it will maintain acceptable pressurization. supply air that is conditioned to the building temperature,
For more information about flow areas and flow coeffi- the heat transfer factor is one. For untreated supply air, a
cients, see Chapter 3 of the Handbook of Smoke Control heat transfer factor of 0.15 is suggested as being conser-
Engineering. vative regarding the impact of stack effect.
Acceptable pressurization of elevator and stairwell pre- The following discussions are specifically for systems
ssurization systems happens when the pressure differen- that use untreated supply air, and it is recommended that
ces from the elevator shaft to the lobby are between mini- systems that use treated supply air be analyzed with
mum and maximum pressure differences, and the IBC re- CONTAM.
quirements for these pressure differences are listed in
Table 3. Elevator Piston Effect
Outdoor temperature and wind data are needed for the
design and analysis of smoke control systems, and Chap- The motion of an elevator car causes transient pressures
ter 2 of the Handbook Smoke Control Engineering provi- that are called elevator piston effect, and elevator pressu-
des such data and barometric pressures for 1663 locations rization systems should be designed so that this piston
throughout the world. Temperature and pressure decrease effect does not adversely impact the system performance.
with altitude above sea level as described in the U.S. An analysis of piston effect was developed and verified
Standard Atmosphere (NASA 1976) (Fig. 1). For analysis with experimental data by Klote (1988). Fig. 2 shows the
of smoke control systems for tall buildings, a feature in pressure differences due to piston effect of an ascending
CONTAM can be used to account for density to decrease elevator car in a 15 story building. From this figure it can
 Elevator Pressurization in Tall Buildings 343

pressurization needs to operate with a number of exterior

doors open. If the system cannot also operate as intended
with all exterior doors closed, some of these doors may
need to open automatically before the elevators are pre-
ssurized. At locations where the fire service does not prop
open exterior doors, a different approach to open exterior
doors may be appropriate.

Basic System
For the basic system, the building envelope often is not
capable of effectively handling the large airflow resulting
from both elevator and stairwell pressurization. This can
Figure 2. Pressure differences due to piston effect of an result in unacceptable pressure difference at lower floors
ascending elevator car.
as shown in Fig. 3. In this figure, acceptable pressuriza-
tion was only achieved with very leaky exterior walls.
be seen that the pressure difference dropped suddenly as However for some buildings it may be possible to achieve
the car passes the eighth floor where measurements were acceptable pressurization with other exterior walls, but
taken. For elevators in multiple car shafts with car veloci- this should be evaluated with a CONTAM analysis.
ties less than 5 m/s, piston effect should not adversely im-
pact the performance of elevator pressurization. For ele- Exterior Vent (EV) System
vators in single car shafts with car velocities less than 2.5
m/s, piston effect should not adversely impact the perfor- The idea of the EV system is to increase the leakage of
mance of elevator pressurization. the building such that successful pressurization can be
achieved. For an open plan office building, this can be
System Concepts done by the use of vents in the exterior walls. These vents
open automatically on activation of the elevator pressuri-
The simplest elevator pressurization systems consist of zation system so that the pressure differences become
only providing pressurization air supplied to the elevator similar to those of the building with very leaky exterior
shaft at one or more locations. This kind of system is walls (Fig. 3). These vents need to be large enough so that
referred to as the basic system. Other elevator pressuriza- successful pressurization will be achieved. The vents need
tion systems discussed below are: exterior vent (EV) to be larger on transfer floors than on other floors.
system, floor exhaust (FE) system, and ground floor lobby For a building that is not open plan, the EV system needs
(GFL) system. Regardless of the system used, any vent to unobstructed flow paths to the outdoors. These flow paths
the outdoors at the top of the elevator shaft needs to be can consist of ducts above a suspended ceiling from the
closed during elevator pressurization, or the impact of such elevator lobby to vents located in exterior walls. Duct
an open vent must be accounted for in the calculation of penetrations of a fire rated wall may have fire resistance
supply air to the elevator. requirements depending on code requirements.
The design of pressurized elevators can be challenging, To minimize the impact of the wind, the vents in exte-
and it can be more so for buildings that also have pre- rior walls should be located on four sides of the building.
ssurized stairwells. Typically pressurized elevators require The supply intakes need to be located away from the
three or four times as much supply air pressurized stair- vents to minimize the potential for smoke feedback into
wells. The reasons for these challenges are: (1) often the
building envelope is not capable of effectively handling
the large airflow resulting from elevator pressurization,
(2) open elevator doors on the ground floor tend to inc-
rease the flow from the elevator shaft at the ground floor,
and (3) open exterior doors on the ground floor can cause
excessive pressure differences across the elevator shaft at
the ground floor.
In most large cities, the fire service props open exterior
doors when they get to a fire to speed up mobilization,
and the International Building Code (IBC) (ICC 2012)
considers that elevator pressurization is with open exterior
doors. Occupants also open some exterior doors during Figure 3. Typical pressure differences across elevator shaft
evacuation. In this article, it is considered that elevator for a basic pressurization system.
344 John H. Klote | International Journal of High-Rise Buildings

the supply air. These vents may need fire dampers depen-
ding on code requirements.
The pressurization air per floor needed for a shaft in a
building with the EV system can be estimated as
2pEL
V = (1 + )CAEL -------------- (1)

V = pressurization airflow to elevator shaft per floor
(m3/s),
C = flow coefficient, dimensionless,
= factor for other elevator shaft leakage (dimen-
sionless), Figure 4. Typical pressure differences across elevator shaft
for a floor exhaust (FE) pressurization system.
AEL = flow area of all elevator doors on the floor (m2),
pEL = pressure difference from elevator shaft to lobby
(Pa), nimize the potential for smoke feedback into supply air.
= gas density in path (kg/m3)
For the EV system, the pressure difference from eleva- Ground Floor Lobby (GFL) System
tor shaft to lobby, pEL, is larger than the minimum pre-
ssure difference but less than the maximum pressure dif- This system has an enclosed elevator lobby on the ground
ference. The factor for other elevator shaft leakage, , de- floor, but the other floors do not have any enclosed ele-
pends on the specific shaft design, construction materials, vator lobbies. This system is the complete opposite of the
and quality of workmanship. A value of = 0.15 seems normal practice of having enclosed elevator lobbies on all
to be a reasonable allowance. floors except the ground floor, but it has the potential for
For example, consider a 27 story building with four ele- successful elevator pressurization.
vator shafts that serve all floors. Each shaft has four ele- An enclosed elevator lobby can reduce the pressure
vator cars. The flow area of an open elevator door is 0.56 differences across the lower floors of the elevator shaft.
m2, and that of a closed elevator door is 0.06 m2. The The GFL system often has a vent between the enclosed
elevators are open on the ground floor, and AEL = 4(0.56) ground floor lobby and the building with the intent of
= 2.24 m2. On floors 2 to 27, the elevator doors are closed, preventing excessive pressure differences across the lobby
and AEL = 4(0.06) = 0.24 m2. The other parameters are C = doors. The lobby doors are the doors between the enclo-
0.65, = 0.12, pEL = 35 Pa, = 1.2 kg/m3. For the ground sed lobby and the building.
floor, V = 12.5 m3/s of supply air. For the other floors, V The pressure difference across the lobby door and the
= 1.33 m3/s. The estimated supply air to each of the four elevator door depend on the area of the vent, and this vent
elevator shafts is 12.5 + 26(1.33) = 47 m3/s. needs to be adjustable to allow for balancing during com-
missioning. The flow area between floors (floor leakage
Floor Exhaust (FE) System and curtain wall gap) is especially important for the per-
formance of the GFL system.
The FE system deals with the building envelope issue
by reducing the amount of supply air used. In the FE References
system, a relatively small amount of air is supplied to the
elevator shafts, and the fire floor is exhausted such that ICC. (2012). International Building Code, International Code
acceptable pressurization is maintained on the fire floor Council, Country Club Hills, IL.
where it is needed. It is common to also exhaust one or Klote, J. H., et al. (2012). Handbook of Smoke Control
two floors above and below the fire floor. Fig. 4 shows Engineering. ASHRAE, Atlanta, GA.
pressure differences across an elevator shaft for a floor Klote, J. H. (1988). An analysis of the influence of piston
exhaust (FE) pressurization system. Because the FE sys- effect on elevator smoke control. NBSIR-88-3751.
National Bureau of Standards, Gaithersburg, MD.
tem only maintains pressurization at some floors, it may
NASA. (1976). U.S. Standard Atmosphere, National Oceanic
need to be approved by the authority having jurisdiction.
and Atmospheric Administration, National Aeronautics
The exhaust can be through a shaft with a fan located and Space Administration, and the United States Air Force.
in a mechanical floor or on the roof, and dampers bet- Avaliable from the National Geophysical Data Center,
ween the shaft and the floors are closed on all floors when Bolder CO.
the system is not operating. On system activation, the Walton, G. N. and Dols, W. S. (2005). CONTAM 2.4 User
dampers open on the floors to be exhausted. The exhaust Guide and Program Documentation, NISTIR 7251, Revi-
inlet needs to be in the elevator lobby or a space open to sed December 14, 2010. National Institute of Standards
the elevator lobby. The outlet of the exhaust fan needs to and Technology, Gaithersburg, MD.
be located away from the inlets of the supply fans to mi-