Академический Документы
Профессиональный Документы
Культура Документы
Silky Wong1; Eric Wey2; Chris Letchford3; Ken Kernaghan4; and Abhijeet
Yesare5
1
Fluor Corp, One Fluor Daniel Drive, Sugar Land, TX 77478. E-mail:
silky.wong@fluor.com
2
Downloaded from ascelibrary.org by New York University on 07/15/15. Copyright ASCE. For personal use only; all rights reserved.
Fluor Corp, One Fluor Daniel Drive, Sugar Land, TX 77478. E-mail:
eric.wey@Fluor.com
3
Rensselaer Polytechnic Institute (RPI), 110 Eighth Street, Troy, NY 12180. E-mail:
letchc@rpi.edu
4
Fluor Canada Ltd., 55 Sunpark Plaza SE, Calgary, AB, Canada T2X 3R4. E-mail:
ken.kernaghan@fluor.com
5
Fluor Daniel India Private Limited, Gurgaon 122002, India. E-mail:
abhijeet.yesare@fluor.com
Abstract
The latest ASCE report “Wind Loads on Petrochemical and Other Industrial
Structures” acknowledges some international documents and resources that guide the
practicing engineer who is involved in reviewing designs or analyzing wind loads on
structures at international Petrochemical facilities. Despite the consensus and
increased uniformity in wind load determination within the profession, engineers are
frequently confronted with situations where multiple country standards are required to
be used for the same project. This presents significant challenges to designers, caution
is required when applying and interpreting multiple codes and standards. This paper
describes a comparison of design wind load calculations for pipe rack structures based
on seven major international codes and standards: ASCE 7-2010 (American),
AS/NZS 1170.2 2011 (Australian and New Zealand), EN Eurocode 1 2010
(European), GB 50009 2012 (Chinese), IS 875 2008 (Indian), NBC 2010 (Canadian),
and SP 20.13330 2011 (Russian). A parametric study of a pipe rack module located in
non-hurricane area supporting piping, electrical cables and equipment is presented.
This simple case study highlights comparative approaches and data among the
selected standards. Recommendations based on the study findings and the application
of standards in wind loads on pipe rack modules are provided.
KEYWORDS
Wind loads, Modular, Pipe racks, Codes, Standards, International, Non-building structures
INTRODUCTION
© ASCE
another country’s wind design codes1 for Front-End Engineering Designs (FEEDs) of
non-building structures for cost saving purposes. However, detail engineering design
must comply with local codes for construction permit approvals by local jurisdictions.
When the results of applying a different code during initial design are significantly
different from that of using the local code, considerable project cost and schedule
impacts will be created. For modular structures, redesigning entire module
configurations may be required to accommodate the design deviations due to various
construction and logistic conditions: Typically, the static design for modular
Downloaded from ascelibrary.org by New York University on 07/15/15. Copyright ASCE. For personal use only; all rights reserved.
structures should be completed based on in-place (operation), land transport (by truck
or Self-Propelled Modular Transporter [SPMT]), and sea transport (by ship or barge)
conditions to design the superstructure and foundation layouts. Structural member
sizes must be designed meet strength, deflection, as well as stability requirements.
This paper illustrates along-wind loads2 exerted at the ultimate limit state on a pipe
rack module located in non-hurricane region with a fairly open terrain during the in-
place condition. To conduct a comprehensive comparison of the wind loads on the
module, seven major international codes and standards were utilized in this study: the
American Society of Civil Engineer’s Minimum Design Loads for Buildings and
Other Structures (ASCE) [1], the Australian and New Zealand Standard (AS/NZS)
[3], the European Standard (EN) [9], the China National Standard (CNS) [6], the
National Building Code of Canada (NBC) [12, 13], the Indian Standard (IS) [4], and
the Russian Code (SP) [16]. The values are then compared to the transportation
analysis case. The recommendations from the authors based on the study results are
given at the end of the paper.
This wind load comparison study is based on a pipe rack module for a processing
facility in Ardmore, Oklahoma (Base Case). Introducing hurricane winds adds a
degree of complexity that may not be accommodated by different wind load codes;
therefore, we considered a non-hurricane case in this pipe rack module study for a
more straightforward comparison. For facilities located in hurricane regions, such as
at specific locations in United States Gulf coast and the Caribbean, the corresponding
design wind speeds are based on site-specific simulation models [17].
Modularization approaches are commonly used to prevent site congestion and ensure
a safer job site with less construction personnel at site. It was assumed that the
module was fabricated and assembled in the Philippines, and transported to the
jobsite by ship and then by SPMT.
The case study pipe rack module is 10.6-m wide, 168-m long, and 19.13 m at its
highest elevation from the finish grade line (at 0 m). The typical bent spacing is 6 m.
Vertical diagonal braces are located at every sixth bay in the longitudinal direction.
1
In this paper, the term “code” is used interchangeably with the term “standard”.
2
Along-wind load – The wind load acting in a plane parallel to the direction of wind but not transverse
to the direction of wind.
© ASCE
The pipe rack module has moment frames in the transverse direction. The module
part of the rack starts at 4.88 m up to 19.13 m from ground. The piperack is supported
by precast columns from 1.02 m to 4.88 m. Four levels of struts are at 7.70 m, 10.44
m, 13.48 m, and 15.62 m above ground respectively. There is no fireproofing
insulation. The total weight of the pipe rack module is approximately 310 metric tons,
which includes weights of process piping, valves, electrical cable trays structural
steel, walkways, and platforms. There are three levels of process pipes with Top of
Steel (TOS) elevations at 6.02 m, 9.07 m, and 11.81 m from ground, respectively.
Utility lines are on the fourth level with TOS elevation at 14.55 m, and electrical
Downloaded from ascelibrary.org by New York University on 07/15/15. Copyright ASCE. For personal use only; all rights reserved.
cable trays and flare lines are on the fifth level with TOS elevation at 17.29 m above
ground, respectively. The Pressure Safety Valve (PSV) deck is located on the sixth
level with TOS elevation at 21.72 m above ground. An isometric view of the pipe
rack module from a Plant 3D design software and RISA-3D structural analysis
software are shown in Figure 1 and Figure 2 respectively.
Figure 2. Model of Case Study Pipe Rack Module for Structural Analysis.
© ASCE
The selected standards start with defining the basic wind speed or pressure. The basic
wind speeds are defined by statistical analysis of wind speed data recorded by local
meteorological observatories. ASCE, AS/NZS, and IS standards define the basic wind
speed as 3-sec gust speed, while other selected standards define it as the mean wind
speed (10-min mean in EN, GB, SP, and 1-hr mean in NBC), at a reference height of
10 m above ground in an open terrain. The basic wind speed or pressure is also
affected by the Mean Recurrence Interval (MRI or return period, which is an inverse
Downloaded from ascelibrary.org by New York University on 07/15/15. Copyright ASCE. For personal use only; all rights reserved.
of the probability that the event is exceeded in any one year) based on the reliability
level of the structure.
In all selected standards, the along-wind loads (F) are determined from a wind
velocity dynamic pressure (q), a shape factor (Cp – pressure, CF – force), a reference
area (A) and a factor to account for dynamic effects (G) as expressed by:
F = qAGCp (1)
ASCE 7-10. The 2010 edition of American national standard for minimum design
loads supersedes previous editions. This standard defines the basic wind speed as the
3-sec gust wind speed at 10 m above the ground in Exposure Category C. The wind
speed maps were developed based on the corresponding return periods associated
with the four Risk Categories varying from 300 to 1,700 years, in order to eliminate
the wind importance factor provided in the previous provisions3 [1]. ASCE 7 defines
a rigid structure as one with a fundamental frequency greater than 1 Hz with a gust
factor (Gf) of 0.85 [2]. Velocity pressure is determined based on the mass density of
air for the standard atmosphere (1.23 kg/m3). This standard does not include the force
coefficients for open frame structures considering shielding effects; instead, Chapter
5 of [2] is the best method to be employed in this study as the detail pipe rack design
information was provided [18].
3
The major change was to move to limit states with a load factor of 1 and basic wind speeds defined
for much longer Mean Recurrence Intervals (MRIs).
© ASCE
GB50009-2012. The Chinese national loading code uses a 10-min mean velocity
pressure proportional to the square of basic wind speed (that is, V2). The wind speeds
are based on a 50-year return period with the air density of 1.25 kg/m3. Force
coefficients for open frame structures are considered as lattice frames with
consideration for shielding effects.
IS 875 (Part 3)–1987 (Reaffirmed 2008). The current Indian Standard for design
loads for building and structures superseded the previous edition, IS: 875-1964. This
standard is based on basic maximum wind speed (that is, peak gust velocity averaged
over 3-sec duration). Air density is taken as 1.20 kg/m3. This standard does not
include the force coefficients for open frame structures considering shielding effects,
and therefore Chapter 5 of [2] is employed.
NBC 2010. The Canadian national model building code is the basis for provincially
enacted building codes. This is soon to be replaced by NBC 2015. The wind loading
provisions have both along-wind and cross-wind calculation mechanisms based on an
hourly mean wind speed. The air density used is 1.29 kg/m3 for determining wind
velocity pressure values. It also includes requirements for dynamic wind response,
such as vortex shedding.
SP 20.13330 2011. The Russian code is an updated version of SNiP 2.01.07-85. The
wind speeds are based on a 10-min mean with 50-year return period. Air density for
wind pressure is not stated in the code. The wind load safety factor is 1.4 for all
structural components. The force coefficient for the open frame structure was
considered as lattice frames with considerations of shielding effects.
BASIC WIND SPEED VALUES WITH DIFFERENT AVERAGING TIMES AND RETURN
PERIODS
© ASCE
transforming wind speeds from one averaging time to another. For example, a 3-sec
gust speed is 1.52 times the mean hourly speed, and a 3-sec gust speed is 1.43 times
the mean 10-min speed.
The selected codes define basic wind speed with various return periods. In this study,
the ratio of the wind speed for any return period to the 50-year return period wind
speed was computed from Equation C26.5-2 in ASCE 7-104:
VT/V50 = [0.36 + 0.1ln(12T)] (3)
where T = the return period in years and VT = T-year return period wind speed.
Downloaded from ascelibrary.org by New York University on 07/15/15. Copyright ASCE. For personal use only; all rights reserved.
Pipe Rack Module Design Parameters. Table 1 shows the wind design parameters
used for this study. The purpose of Case I (54 m/s basic wind speed) is to compare
base shear and overturning moment values without any conversions of differences in
return periods and averaging times of each code. Case II provides results of
considering local codes’ basic wind velocity return periods and averaging times.
4
Formulas for converting wind speeds based on various return periods are specific to the screened
wind speeds recorded in those countries. However, these formulas are not available for some of the
selected codes. To maintain consistency of this study, Equation (3) based on the study by Peterka and
Shahid (1998) [14] was chosen.
© ASCE
In addition, the authors aimed at comparing the wind load results from Case I and
Case II against that obtained from the sea transportation analysis. For the sea
transport forces along the module’s transverse direction, the marine consultant
performed a detailed transportation analysis to provide motion induced transportation
forces based on the voyage from the fabrication yard to the designated offload
terminal near the project site. The transverse sea transportation forces considered in
this study included the horizontal force component due to a roll (or beam sea
condition), the horizontal component of gravity force caused by tilt of the ship deck
during a roll, and the horizontal force component due to heave (which acts in the
barge transverse/vertical direction and concurrently with the rolling motion). Lateral
acceleration coefficients were given and expressed as a function of the distance of
center of gravity above the barge deck. For simplicity, partial wind considerations
were excluded. The vertical and horizontal barge stiffness of 87.56 kN/mm (500
kips/inch) was used.
Table 2A and Table 2B show the results of in-situ analyses of this study, including
normalized base shear and overturning moment comparisons to the Base Case (ASCE
7-10) values. The base shear from the sea transport analysis, which was determined
with a factor of 1.6 for ultimate limit state design per ASCE, is 55% higher than that
calculated from the Base Case.
Using the basic wind speed as 54 m/s for all selected codes yielded the along-wind
normalized base shear mean and coefficient of variation values as 2.74 units and 59%
respectively. Same results can be obtained from the overturning moment values as
listed on Table 2A. The European standard gives the highest base shear and
overturning moment values and they are more than four times as compared to the
values from the American standard. The American standard gives the lowest values.
The large variation of Case I results is within the authors’ expectation, as the selected
codes define different probabilistic methodology in determining the wind speed.
© ASCE
Table 2A. Ultimate Limit State Force Comparisons in Various Wind Codes &
Sea Transport Analysis along Transverse Direction (Case I).
ASCE AS/NZS EN GB NBC SP IS Sea
2010 2011 2010 2012 2010 2011 2008 Transport.
Basic Wind 54 m/s 54 m/s 54 m/s 54 m/s 54 m/s 54 m/s 54 m/s N/A
Speed
Normalized 1.00 1.36 5.24 2.48 4.68 2.32 2.13 1.55
Base Shear
Normalized 1.00 1.36 5.24 2.45 4.66 2.36 2.11 N/A
Downloaded from ascelibrary.org by New York University on 07/15/15. Copyright ASCE. For personal use only; all rights reserved.
Moment
% Deviation 0% +36 % +424% +148% +368% +132% +113% +55%
from Base 0% +36% +424% +145% +366% +136% +111%
Case (ASCE)
Table 2B. Ultimate Limit State Force Comparisons in Various Wind Codes &
Sea Transport Analysis along Transverse Direction (Case II).
ASCE AS/NZS EN GB NBC SP IS Sea
2010 2011 2010 2012 2010 2011 2008 Transport.
Converted 54 m/s 53 m/s 28 m/s 28 m/s 26 m/s 28m/s 40 m/s N/A
Wind Speed
Normalized 1.00 1.29 1.42 0.66 1.13 0.64 1.17 1.55
Base Shear
Normalized 1.00 1.30 1.42 0.65 1.13 0.65 1.15 N/A
Moment
% Deviation 0% +29 % +42% -34% +13% -36% +17% +55%
from Base 0% +30% +42% -35% +13% -35% +15%
Case (ASCE)
From the results in Case II, the mean is 1.04 and coefficient of variation is 29% for
both along-wind normalized base shear and overturning moment. Although the
variation of Case II results are almost half of that obtained from Case I results, it
indicates that each code has considerable variability in the wind load calculation
methodology other than basic wind speed after adjustments of averaging time, return
period, and the ultimate state design factor. Applying Equation (3) to locations
outside of United States is another factor leading to the largely deviated results as the
expression of return period conversion is different from each country region.
During the process of this study, the authors observed that results based on Chinese
and Russian code are similar. It may be due to the similar underlying theory of
establishing wind load design criteria. In addition, the force coefficient for lattice
frame or structure from the Russian code is significantly lower than all other selected
codes, however, the Russian code requires a wind safety factor of 1.4 for all structural
components. In Case II, although the Russian code shows the lowest along-wind
values and the European standard shows the highest along-wind values, the
transportation case exceeds all other base shear values; this finding concludes that
transportation loads must be considered prior to performing wind load analysis for a
pipe rack module design.
© ASCE
In this paper, the authors presented an onshore piperack module example for a project
in the United States non-hurricane inland region (as the Base Case). Observed
differences using various international codes were described with calculations
conducted. The normalized base shear and overturning moment caused by along-wind
effects were discussed. The following presents our recommendations regarding wind
loads on onshore modules:
Downloaded from ascelibrary.org by New York University on 07/15/15. Copyright ASCE. For personal use only; all rights reserved.
Scaling of Module Steel Tonnages based on Ratios of Wind Speeds Can Lead to
Erroneous Results. It is common for Petrochemical companies to have facilities in
several countries. These facilities are often modularized for reasons of cost, safety,
and speed of construction. The success of a particular facility in one location
provides a legitimate reason to copy the design in another location. If process
conditions remain the same, then the size and ratings of equipment, pipe, and wiring
can remain the same. Environmental conditions, such as wind, however, are typically
not the same at different locations. Differences in governing building codes gives rise
to many problems when incorrect comparisons are made of facilities that are copied
from different countries. Terms used in different building codes may seem similar,
but typically they cannot be practically compared. As shown in this paper, scaling of
steel tonnages based on forces determined from ratios of wind speeds can lead to
erroneous results due to different formulation and factors that are associated with
each code. In addition, various codes have their unique service limit state and
ultimate limit state design requirements for stability, serviceability and strength
checks. For example, material take-off based on member size strength checks may
often lead to underestimation using the Canadian code, as member sizes are more
likely needed to be heavier to fulfill unbraced stability and connection requirements.
© ASCE
and/or construction windows, etc, prior to providing module member sizes from wind
load analysis.
from [2] with absence of nation/local code requirements. However, this approach can
be inadequate in certain countries. When attempting to use a different code as the
basis for FEED in a project, engineers must be very familiar with regional code
requirements, such as load factors, performance limits, specifications for materials
and shapes, etc, as well as the history and background of those requirements.
Future Studies and Research. Design of modular pipe racks is often governed by
transportation analysis (such as imposed forces due to sea acceleration). Studies on
how transportation loads influence the design of a structure governed by wind load
will be of value in contribution to the literature in this field.
ACKNOWLEDGEMENTS
The authors would like to acknowledge the support from Fluor Corporation’s
Professional Publications and Presentations Program. The authors would also like to
express our gratitude to Mr. William Bounds for reviewing this paper. Mr. Bounds is
© ASCE
a senior Fellow in Fluor Corporation and the contributor of Handbook for Blast
Resistant Design of Buildings.
BIBLIOGRAPHY
[1] American Society of Civil Engineers (ASCE). (2010). Minimum design loads for
buildings and other structures (ASCE 7-10), Reston, Virginia.
[2] Task Committee on Wind-Induced Forces of the Petrochemical Committee of the
Energy Division of ASCE. (2011). Wind loads for petrochemical and other
Downloaded from ascelibrary.org by New York University on 07/15/15. Copyright ASCE. For personal use only; all rights reserved.
© ASCE
[17] Vickery, P. (2012). “Design Wind Speeds in the Caribbean.” In C. P. Jones & L.
G. Griffis (Eds.), Proceedings of the 2012 ATC and SEI Conference on
Advances in Hurricane Engineering (pp. 1136–1147). ASCE.
doi:10.1061/9780784412626.099
[18] Wong, S., Sepaha, A., Swamy, N., Amoroso, S., & Naqvi, D. (2012). “Wind
Loads on Non-Building Structures Using ASCE 7-10.” Proceedings of
Structures Congress 2012 (pp. 1438–1451). ASCE.
doi:10.1061/9780784412367.128
Downloaded from ascelibrary.org by New York University on 07/15/15. Copyright ASCE. For personal use only; all rights reserved.
© ASCE