DESIGNING A SEISMIC ANCHORAGE SYSTEM FOR A NEW STAINLESS STEEL TANK ON A SKIRT
ABSTRACT
This case study outlines ONGUARD’s seismic design approach for a new liquid storage tank on a stand holding about 65,000 gallons of wine. The 7 gauge stainless steel tank is 20.3 ft in diameter and has 27.0 ft tall walls. It sits on a 1.0 ft tall concrete-filled 7 gauge stainless steel skirt. ONGUARD presents two energy-dissipating seismic solutions for the tank on a skirt. The solutions are used to compare two different design approaches, one aligning with API Standard 650 and the other aligning with NZSEE’s “Seismic Design of Storage Tanks” 2009 Guideline.
INTRODUCTION
This study begins with a collection of all the preliminary design information. Next, the possible design approaches are discussed, and two design approaches are identified for analysis. Then the design approaches are analyzed in two design solutions. After the two solutions are compared, the final solution is selected. Lastly, closing remarks note ONGUARD’s preference for design approach.
PRELIMINARY DESIGN INFORMATION
ONGUARD begins every design by gathering preliminary design information including tank construction parameters, site ground motion parameters, and seismic design factors.
TANK CONSTRUCTION PARAMETERS
First, ONGUARD consults with the client to determine what tank construction information is available. For new construction applications, the tank builder provides tank construction drawings which are used to collect design parameters (e.g. Figure 1). Figure 2 is a photograph of tanks built according to the tank construction drawings provided.
Table 1: Tank Construction Parameters
Tank Construction Parameter | Symbol | Value | Units |
Nominal volume | Vtank | 65,000 | gal |
Tank wall height | Htank | 27.0 | ft |
Diameter | Dtank | 20.3 | ft |
Top cone height | Hcone | 2.0 | ft |
Turret height | Hturret | 0.7 | ft |
Skirt height | Hskirt | 1.0 | ft |
Maximum slab height | Hslab | 24 | in |
Tank wall thickness | Twall | 10 | gauge |
Tank floor thickness | Tfloor | 7 | gauge |
Tank skirt thickness | Tskirt | 7 | gauge |
Tank floor slope | [-] | 2.39 | ° |
Figure 1: Liquid storage tank construction drawings provided by the tank builder
Figure 2: Liquid storage tanks built according to the tank construction drawings
For tanks on skirts, ONGUARD designs new anchors to be installed between the stainless steel skirt wall and the slab.
Figure 3: An example of an OG PRO anchor mounted between skirt and slab. The anchor is welded to a backing plate which reinforces the skirt. A hole beneath the anchor is drilled in the slab for anchor stud, which is epoxied into the slab hole and threaded into the anchor coupler.
SITE GROUND MOTION PARAMETERS
Second, ONGUARD uses the tank facility’s street address to lookup site ground motion details. These details are provided by the Structural Engineers Association of California in the form of an OSHPD Seismic Design Map [1], which references ASCE 7-10. If the client is able to provide a geotechnical report, the report may be used in substitution for the OSHPD Seismic Design Map.
Table 2: Site Ground Motion Parameters
Site Ground Motion Parameter | Symbol | Value | Units |
Site Class | [-] | D – Stiff Soil | [-] |
Numeric seismic design value at 0.2 second SA | SDS | 1.691 | [-] |
Numeric seismic design value at 1.0 second SA | SD1 | 0.920 | [-] |
MCER ground motion. (for 1.0s period) | S1 | 0.812 | [-] |
SEISMIC DESIGN FACTORS
Third, ONGUARD references API 650 [2] to determine the seismic design factors for the new seismic tank system including ONGUARD OG PRO ductile tank anchors.
Table 3: Seismic Design Factors from API 650 Section E.5 [2]
Seismic Design Factor | Symbol | Value | Units |
Convective response modification factor | Rwc | 2.0 | [-] |
Impulsive response modification factor | Rwi | 4.8 | [-] |
Seismic use group | I | 1.0 | [-] |
The response modification factors from Table 3 are used to calculate anchor force demands. Anchor force demands are used to specify the anchor yield load for the selected anchor. All other components of the structural system are designed for demands 15% greater than the anchor yield load.
DESIGN APPROACH DISCUSSION
Today’s most prominent seismic tank design standards and guidelines are not consistent in assuming anchor force distribution.
Three anchor force distributions that form the basis of three seismic tank design approaches are identified in NZSEE’s “Seismic Design of Storage Tanks” Guideline [2]. These anchor force distributions are depicted in Figure 2.
Figure 2: Three Possible Anchor Force Distributions depicted in NZSEE’s “Seismic Design of Storage Tanks” Guideline [2]. The anchor force distribution dictates how the seismic forces interact with the liquid storage tank.
To demonstrate the effects of choosing between design approaches, ONGUARD designs two possible seismic tank solutions using two design different approaches in this study. They are referred to as Approach A and Approach B.
DESIGN APPROACH ANALYSIS
APPROACH A – ANTISYMMETRIC
This design approach uses the antisymmetric anchor force distribution, which assumes elastic anchor performance as described in both API Standard 650, Welded Tanks for Oil Storage, Thirteenth Edition [2] and AWWA Standard D100-21, Welded Carbon Steel Tanks for Water Storage [4].
Table 4: Approach A Design Results
Approach A Design Result | Symbol | Value | Units |
Overturning moment acting on tank slab | Ms | 3,489 | ft-kips |
Maximum load per foot width acting on anchor ring | w | 8.20 | kip/ft |
Anchor selection | [-] | OG PRO 81 MKIV | [-] |
Number of anchors | n | 26 | [-] |
Design force per anchor (ASD) | P | 20.07 | kips |
Design capacity per anchor | ɸTn | 20.23 | kips |
Stud embedment depth | Hef | 19 | in |
Minimum slab height | Hslab | 21 | in |
Minimum edge distance (between stud and slab edge) | c | 26 | in |
APPROACH B – DUCTILE ELASTIC
This design approach uses the ductile elastic anchor force distribution, which assumes ductile anchor performance as described in other international tank guidelines such as NZSEE’s “Seismic Design of Storage Tanks” Guideline [2].
Table 5: Approach B Design Results
Approach B Design Result | Symbol | Value | Units |
Overturning moment acting on tank slab | Ms | 3,489 | ft-kips |
Maximum load per foot width acting on anchor ring | w | 5.43 | kip/ft |
Anchor selection | [-] | OG PRO 81 MKIV | [-] |
Number of anchors | n | 20 | [-] |
Design force per anchor (ASD) | P | 17.26 | kips |
Design capacity per anchor | ɸTn | 17.31 | kips |
Stud embedment depth | Hef | 16 | in |
Minimum slab height | Hslab | 18 | in |
Minimum edge distance (between stud and slab edge) | c | 13 | in |
RESULTS COMPARISON
Both design approaches result in the same overturning moment acting on the tank slab (Ms) since the magnitude of total force acting on the tank is held constant.
Approach A requires a greater number of anchors (n) since the anchor ring sees a greater maximum load per foot width (w). The anchor ring is the diameter of the ring of anchors that secure the tank to the stand. Since the anchor ring sees a greater force per nominal width in Approach A, it requires a greater number of anchors to keep the tank secured to the stand without causing buckling in the tank wall.
The design force and design capacity per anchor vary since the number of anchors and loading of anchors differs for each design approach.
Approach B has a smaller design capacity per anchor, meaning each anchor and its associated stud experiences less tension in a design-level earthquake. Less tension on the stud means the stud embedment depth can be reduced. The stud embedment depth is related to the slab height and edge distance which are constrained to avoid concrete breakout. Because the stud embedment depth can be reduced, the minimum slab height and minimum edge distance can both be reduced.
SOLUTION SELECTION
Approach B is preferred since it requires a lesser number of anchors, a lower slab thickness, and a smaller edge distance. Less anchors means a reduction in both hardware purchase costs and labor costs for installation. A lower slab height and smaller edge distance means a reduction in slab volume, saving money on concrete costs. The smaller edge distance also allows the liquid storage tanks to be installed closer together, and the slab footprint to be smaller, opening up more floor space to be used for other purposes at the tank facility.
CLOSING REMARKS
Approach B, the approach that uses the ductile elastic anchor force distribution, is the preferred approach of ONGUARD. ONGUARD has found that this approach has the more accurate anchor force distribution for a tank with ductile anchors such as OG PRO MKIV Anchors, and it provides better protection to the tank walls. ONGUARD has found that Approach A overestimates the anchor forces when ductile anchors are used.
ONGUARD recognizes today’s conflicts in seismic tank design approach, so ONGUARD developed its own research to gain understanding in the actual performance of tanks with ductile anchors in real earthquake events. ONGUARD uses this research to back its design decisions, and we hope to see this research adopted more widely in the future.
REFERENCES
[1] “OSHPD Seismic Design Maps”. Structural Engineers Association of California.
[2] “Welded Tanks for Oil Storage, API Standard 650, 12th Edition”. API. 2013
[3] Priestley, M. J. Nigel et al. “Seismic Design of Storage Tanks”. NZSEE. 2009.
[4] “AWWA D100-21 Welded Carbon Steel Tanks for Water Storage”. AWWA. 2021