Tank Info PDF
Tank Info PDF
Tank Info PDF
November, 2003
Introduction:
In the near future, API will publish an appendix to API 650 that will prescribe
requirements for the design of tanks that operate at negative pressure (or vacuum).
This paper discusses:
§ The background on the development of the new appendix. Why and how did it
come about?
§ The scope of the document as defined by the Process Industry Practices group.
§ The technical basis of the appendix. Where do the rules come from?
§ The loading conditions and combinations addressed by the appendix. How is
external pressure combined with roof live load and wind load, e.g.?
§ The equations and formulas that are provided for the design of tank components.
What components are included?
§ A design example to illustrate application of the design procedures. What will a
typical tank subject to a significant vacuum design pressure look like?
§ Guidance for cases where tanks are subject to flooding of the dike area
surrounding the tank.
Background:
The request that API 650 include design rules for tanks with external pressure originated
with the Process Industry Practices (PIP) group, a consortium of numerous petroleum
and chemical companies. In response to PIP’s request, API agreed to develop the
requested rules and incorporate them in an appendix to API Standard 650. An API
Special Task Group was formed to develop the appendix. Thus in late 1998, the Task
Group began work on the appendix.
Fortunately, work on external pressure design of tanks that DuPont had already compiled
and implemented in an internal engineering standard was made available to the Task
Group with DuPont’s permission. This document, together with others that are included
in the References section of the appendix, proved to be a great “starting point” in
formulating and presenting the rules for API 650 tanks.
The PIP request was dated April 15, 1998. The effort has taken about five and on-half
years to this point. The specific Scope of Work defined for the Task Group by PIP was:
“This task group will develop an appendix to the API Standard 650 for external
pressure design. Uniform partial vacuum not to exceed one pound per square inch
gage and partial submergence in near-static liquids shall be considered. Closed
tanks having both supported and self-supporting roofs, anchored and unanchored,
at ambient and elevated temperatures, and of both carbon steel and stainless steel
construction shall be considered. Limitations such as maximum configuration
size and magnitude of external pressure may be imposed for specific geometries,
materials, and loading combinations.”
The technical bases for the provisions in the appendix are derived from several well-
known documents, including:
The format and organization of the appendix intentionally closely resembles that of the
DuPont Engineering Standard. However, closer study will show differences in some
equations. These differences are typically related to nomenclature or safety factors, for
example, to make the equations and results consistent with other provisions of API 650.
The appendix will prescribe rules for designing a tank for a specified design external
pressure and will include rules for combining that specified pressure with other loadings
that may occur simultaneously. A recently approved change to API 650 will prescribe
rules for load conditions and load combinations and will include these rules in a
centralized location in the standard. The new rules are consistent with the philosophy of
loading combinations adopted by the major building codes and are also consistent with
API 650’s level of risk of tank failure. Likewise, the rules in the external pressure
appendix will be consistent with the basic part of API 650 with regard to loading
conditions and combinations.
Thus, starting with a specified design external pressure, roof live or snow load, and wind
pressure (or velocity), the total roof design pressure is calculated as the greater of DL+(Lr
or S) + 0.4Pe or DL +Pe+0.4(Lr or S), where DL = Dead Load. And the total shell
design external pressure is the greater of Pe, excluding wind or W+0.4Pe.
Note that although it is not explicitly stated in the appendix, design external pressure
loading should be combined with any specified seismic loading in accordance with load
combination criteria prescribed in the main body of the standard.
The appendix includes, in both SI and US Customary Units, the equations necessary to
calculate the following:
The appendix will not include prescriptive equations for the design of column and rafter-
supported cone roofs or supported dome roofs. For relatively low specified design
external pressures, where supported cone roofs may be cost-effective, the necessary
equations are available in the standard or its references. For more significant design
external pressures, supported cone or supported dome roofs are generally not cost-
effective compared to self-supporting roofs.
Let us consider what a typical tank design for external pressure might look like. The
appendix includes a design example that illustrates the application of the design
equations for a 75 foot diameter by 48 foot high tank. This tank would probably be
considered large by chemical industry standards, but moderate to small by petroleum
industry standards.
If we assume that process conditions dictate that the tank operates at a vacuum of 0.6
psig, and that the design wind velocity for the location of the tank is 100 mph, the result
of applying the design equations prescribed in the appendix is the tank pictured below.
75 ft DIA.
Pe = 0.6 psig
Wind = 100 mph
Shell Thickness = 0.395 in
Four Intermediate Stiffeners
One Top End Stiffener
All Angle Section Stiffeners
Butt-Welded Dome Roof
Roof Thickness = 0.529 in 48 ft
Roof Dish Radius = 60 ft
In order for this example tank to withstand the specified vacuum loading within safety
factors comparable to API 650 safety factors for other types of loading, the tank roof
needs to thickened considerably and the shell needs to be thickened and stiffened with
four intermediate circumferential stiffeners.
On a tank as large as this one, it was necessary to reduce the dish radius of the dome roof
and increase the thickness beyond the limits of lap-welding.
The thickness of the entire shell was increased to the minimum thickness required for the
lowest shell course for internal pressure loads for convenience in this example. The
appendix allows one to design stiffening arrangements for shells of varying thickness as
well. This allows the most cost-effective combination of shell thickness and stiffening
rings to be developed.
One issue of special importance to the PIP group who requested this appendix was the
case where a tank may be subject to periodic flooding of the dike, impoundment, or
secondary containment area when the tank is empty of liquid. Such an occurrence tends
to uplift the bottom of the tank and to impose hydrostatic pressure against the outside of
the shell. Normally, the amount of liquid, usually primarily water, that collects is not
sufficient to impose significant hydrostatic pressure on the shell of the tank, but the
height of water can be significant enough to “balloon” the bottom of the tank. This is
illustrated in the figure below.
The appendix provides guidance on how to evaluate the design of the tank for this
condition, including such components as the corner weld of the shell to bottom, tank
anchorage, and attached piping and/or bottom fittings such as sumps. One practical
method commonly used to mitigate the risk of damage in this event is to maintain a
minimum liquid level inside the tank at all times that is high enough, considering the
relative specific gravities of the liquids inside and outside the tank, to balance or offset
the pressure causing the bottom to want to balloon. The appendix provides the
equations necessary to calculate this minimum liquid level.
The damage that can be done to tanks by external pressure is similar, whether the
damage is caused by wind pressure or by negative internal gas pressure, as the photos
that follow will illustrate.
Normally, wind damage tends to be less extreme than negative internal pressure because
wind is usually a temporary loading condition whereas vacuum can be sustained over a
significant period of time. Of course there are exceptions to both of these generalities.
Sometimes the damage done can be repaired by simply pulling out the dents. In other
cases, where the shell material has yielded and is plastically deformed, replacement of
some material is necessary.
Since the damage mechanism for vacuum loading is similar to that for wind loading, it
makes sense that the preventative measures in the design for vacuum loading are similar
to those for wind loading, i.e. thickened shells with or without circumferential ring
stiffeners to increase the buckling resistance of the shell.