See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/268585429

Static Analysis of Substation Rigid Bus Using the Finite Element Program

Conference Paper · November 2012

DOI: 10.1061/9780784412657.013

CITATION READS
1 1,270

1 author:

Prapon Somboonyanon
AEC Lionstech
13 PUBLICATIONS 14 CITATIONS

SEE PROFILE

All content following this page was uploaded by Prapon Somboonyanon on 13 May 2020.

The user has requested enhancement of the downloaded file.

 Static Analysis of Substation Rigid Bus Using the Finite Element Program

Prapon Somboonyanon1
1
P.E., Senior Civil/Structural Engineer, Burns & McDonnell Engineering Company
Inc., 9400 Ward Parkway, Kansas City, MO 64114; PH (816) 822-3953; FAX (816)
822-3296; email: [email protected]

ABSTRACT

To design a rigid bus system, several design methods are available. The IEEE
Std. 605-2008 design method is widely used but is conservative and has several
limitations. A dynamic analysis approach offers more accurate results but is more
complex and time consuming. The static analysis using a finite element program
design approach was proposed and described in this paper. The IEEE Std. 605-2008
design limitations were first identified followed by design advantages from the
proposed method. Proper design considerations should be accounted for when using
the program to obtain meaningful results. Results were compared between the IEEE
Std. 605-2008 and the proposed design methods using the same applied loads derived
from IEEE Std. 605-2008 design guidelines and the same set of design inputs. Three
case studies were performed, and the results from Case 1 and Case 2 showed the
conservatism in the IEEE Std. 605-2008 design method. Results from Case 3 showed
the impact from overload factors on the same bus arrangement. It was then
concluded that the static analysis using finite element program design approach
offered more accurate results than the IEEE Std. 605-2008. In addition, results
showed that insulator internal forces and bus conductor fiber stresses analyzed by the
proposed method were reduced comparing to the IEEE 605 Std. 605-2008 design
method, which could result in overall construction cost savings due to less insulators,
less bus structures, or less foundations needed.

INTRODUCTION

The design procedure based on IEEE Std. 605-2008 “IEEE Guide for Bus
Design in Air Insulated Substations” for designing a rigid bus system is widely used
by utility companies. This standard provides design guidelines in great detail using a
simplified static analysis approach providing closed-form equations. Although this
design method is simple and easy to implement, it has several limitations and, in
some cases, is too conservative. The primary source that makes the IEEE Std. 605-
2008 design method conservative is short-circuit load. Several papers have been
published to address the conservatism in short-circuit load applied to rigid bus system
using a dynamic analysis approach. For example, Amundsen, Oster, and Malten
(2009) proposed a design method considering rigid bus dynamic response in order to
reduce short-circuit load applied to rigid bus span. Pinkham and Killeen (1971)
provided a design method and simplified formulas to determine dynamic response of
insulators supporting rigid bus. In addition, Iordanescu, Hardy, and Nourry (1987)
presented a dynamic design approach with a finite element technique to determine
stresses and displacements of rigid bus conductors and bus structures. Although the
dynamic analysis design method provides more accurate results that better agree with
experimental data as shown in IEEE Std. 605-2008 Annex F, it is more complex and
time consuming compared to the static design approach. A simplified dynamic
design method may be less complex, but it may not be applicable for all rigid bus
arrangements. Because of the conservatism and limitations in static design approach
based on IEEE Std. 605-2008, and the complexity of dynamic design approach, a
static design approach using finite element program is proposed as an alternative.
The finite element software program used for analysis in this paper is RISA-3D
developed by RISA Technologies, LLC. The main focus of this paper is to address
the design limitations of IEEE Std. 605-2008 and to compare results between the two
design methods.

IEEE Std. 605-2008 DESIGN LIMITATIONS

Rigid bus systems are generally used to connect electrical equipment, as well
as transmission structures, in a substation or switchyard. Regardless of bus
arrangements, the system typically consists of rigid bus conductors, insulators, and
bus structures. With an increase of fault currents in substations resulting in higher
short circuit loading, different insulator configurations have been used to support bus
conductors and to transfer loads to bus structures, as shown in Figure 1. Closed-form
equations provided in IEEE Std. 605-2008 are valid only for single insulator
configurations. For double insulator configurations, some designers may use IEEE
Std. 605-2008 assuming that insulators have the twice cantilever strengths. With the
conservatism in the design method, this assumption may still be valid. However, it is
obvious that the design method is not applicable for delta configurations.

Figure 1. Rigid bus system.

Insulator strengths vary by materials such as porcelain, polymer, silicone, etc.

This paper investigated only porcelain insulators. In general, insulator torsion,
tensile, and compression strengths are much higher than cantilever strength. The
IEEE Std. 605-2008 provides closed-form formulas to design rigid bus conductor
based on 1) insulator strength, 2) bus conductor fiber stress and 3) bus conductor
deflection. For the design based on insulator strength, closed-form formulas address
only insulator cantilever strength, excluding insulator torsion, tensile, and
compression capacity. For single- and double-insulator configurations, tensile and
compression loads developed within insulators are relatively low compared to its full
capacity, therefore tensile and compression strength checks might not need to be
checked although it is not recommended. In contrast, for the delta configuration,
tension and compression loads developed within insulators are usually higher than the
other configurations, so tensile and compression strength checks should be
performed. In addition, for some bus arrangements, insulators could experience high
torsional stresses, especially at fixed fittings. Therefore, insulator tensile,
compression, and torsion strengths should always be checked regardless of insulator
configuration.
In a two-level bus configuration, similar to Figure 1, a lower level bus could
experience high stress due to loads transferred from an upper level bus through an
A-frame. The IEEE Std. 605-2008 does not address this problem. Thaik (1995)
described this problem and provided a static design method to determine bus stresses
and deflections. However, Thaik’s proposed method may not be suitable for all bus
arrangements or all insulator configurations.
In addition, rigid bus conductors are made of elastic material, usually
aluminum or copper and are sometimes bent to accommodate change in yard
elevations or phase spacing in bus arrangements. Formulas in IEEE Std. 605-2008
are not suitable for non-parallel bus conductor.

STATIC DESIGN OF RIGID BUS USING FINITE ELEMENT PROGRAM

A static design approach using finite element program is proposed to address

the IEEE Std. 605-2008 design limitations mentioned previously. Rigid bus system
components created in RISA 3-D program are defined as either beam or column
element with dimensions and material properties matching actual components.
Several sources are available for these values including IEEE Std. 605-2008. Figure
2 shows an example of a bus model created in RISA-3D. Several advantages are as
follows:

 The design can be performed in any commercial structural analysis

program. A variety of materials and members are normally available in
those programs, which are applicable for any bus arrangement.
 The design approach can be used with any insulator configuration or bus
arrangement regardless of complexity and size.
 All insulator strength characteristics can be analyzed.
 Stresses and deflections in bus conductors or A-frame members at any
point of a bus arrangement are known.
 Creating a bus model may initially require more time compared to the
IEEE Std. 605-2008 design method or a simplified dynamic method.
 However, once the model is complete, impacts from any change or
modification can be quickly analyzed.
 Analysis of results comparing different bus arrangements, materials, or
even cost could be performed quickly and more accurately.
 The entire substation layout can be designed concurrently, including rigid
bus conductors, insulators, bus structures, and potentially foundations.
 If bus structures and foundations are designed by others, then the design
approach offers more accurate results, resulting in more economical
design for bus structures and foundations.

Figure 2. Example of a substation rigid bus model utilizing RISA-3D.

DESIGN CONSIDERATIONS IN THE PROGRAM

When creating a bus model, in addition to defining proper dimensions and

material properties matching an actual arrangement, other design considerations
should be accounted for. For example,

 Boundary conditions are critical and could have significant impacts on the
results. Without bus structures in a bus model, fixed support can be used
at the base of the insulators. At the top of insulators, where the bus
conductor is supported, this boundary condition varies depending on bus
fitting types. Three common bus fittings used in most substations include
1) fixed fitting, 2) slip fitting, and 3) expansion fitting. Expansion fitting
allows bus conductors to move horizontally parallel to the conductor to
accommodate thermal expansion, but unlike slip fitting, horizontal
movement is limited, approximately two inches. The actual value varies
from one manufacture to another. Defining boundary conditions for fixed
and slip fittings are straightforward, but care should be taken for
expansion fitting.
  As also recommended in IEEE Std. 605-2008, allowable stress of bus
conductors with welded connections are reduced by 50%, therefore
locations of welded connections should be considered in the model.
 When applying short-circuit load in the model without bus structures or
with one-phase bus structures, the direction of short-circuit load on each
conductor phase does not matter as each phase is independent. However,
direction of short circuit load in the model with three-phase bus structures
becomes important as each phase is connected by three-phase bus
structures. Proper direction of load will provide more accurate results to
all components of the system.
 According to ANSI C37.32-2002 standard, Table 4 provides the preferred
mechanical loadings for high-voltage switches. The loads transferring to
switches should be accounted for in the model and should be kept within
the limit.
 Selected load combinations used in a bus model analysis will also have
impacts to results. The impact was investigated and is described in the
case study section.

DESIGN INPUTS

Although different bus arrangements are analyzed in this paper to compare

results in different aspects, all basic design inputs are kept the same in terms of
physical and electrical properties, as well as applied loads. Wind, ice, and short-
circuit loads were derived based on IEEE Std. 605-2008 Clause 11. The design
inputs are as follows:

 All rigid bus conductors are made of “Aluminum 6063-T6.” Rigid bus is
5” Schedule 80, and A-frame is 2-1/2” Schedule 80.
 Damping wire is used and weighs 1.094 lbf/ft in addition to rigid bus
weight.
 Phase spacing, D, is equal to 12 feet, and the system fault current, Isc, is
50,000 A with assumed X/R value of 20. System frequency, f, is taken as
60 Hz. Support flexibility factor, Kf, is equal to 1.0, assuming the system
includes three-phase bus structures.
 Wind speed is 100 mph for extreme wind load case. Ice with concurrent
wind load case is also analyzed with 40 mph wind speed and 0.75” radial
ice. Exposure category is C, and the effective height at which the wind is
being evaluated, z, is 25 feet.

With the above design inputs, computed loads are as follows:

 Bus conductor weight with damping wire = 8.282 lbf/ft

 Extreme wind load on bus conductor = 9.529 lbf/ft
 Ice load on bus conductor = 5.890 lbf/ft
 Concurrent wind load on bus conductor = 1.936 lbf/ft
  Short circuit load = 55.852 lbf/ft
- Computed fault clearing time, Ta = 0.053 second
- Computed decrement factor, Df = 0.927

CASE STUDIES

Case 1: Insulator cantilever strength study

Using design inputs defined in the previous section and additional design
inputs shown below, results from Case 1 showed the conservatism in the IEEE Std.
605-2008 design method. Additional design inputs are as follows:

 The insulator is 80” high with average diameter of 10.5”. Cantilever,

tensile, torsion, and compression strengths are 2,450 lb, 20,000 lb, 60,000
lb-in, and 60,000 lb, respectively.
 All bus fitting heights are 5” high.
 The bus arrangement is two equal spans with fixed fittings in the middle
and slip fittings at both ends.
 The maximum effective bus span, LE, is taken as 5L / 4 for two continuous
spans with P-C-P support conditions, Table 18 of IEEE Std. 605-2008.
 The overload factor for wind load is 2.5, K1, and 1.0, K2, for the short-
circuit load.

To compare results for Case 1, allowable span based on the insulator

cantilever strength was first computed using formulas provided in IEEE Std. 605-
2008 Clause 12.3. The computed span is equal to 21.74 feet. At this span length, the
insulator experienced 100% capacity usage. Then, a bus model, as shown in Figure
3, was created and analyzed using the same design inputs with two equal spans of
21.74 feet to obtain actual stresses within insulators.

Figure 3. Rigid bus model of two equal spans with 21.74 feet in span length
without a bus structure.
 Next, another bus model was created and analyzed using the same bus
arrangement as shown in Figure 3 with the addition of a three-phase bus structure in
the middle, as shown in Figure 4. All steel members supporting the bus arrangement
used in this model are wide flange W8x31 Gr. 36. Results from Phase C insulator at
the middle support for all three configurations were compared and summarized
subsequently in Table 1.

Figure 4. Rigid bus model of two equal spans with 21.74 feet in span length
including a three-phase bus structure.

Insulator Strength Usages

Description
Cantilever Torsion Tensile Comp.
Results from IEEE Std. 605-2008 design
100% n/a n/a n/a
method
Results from the model without a bus
87% 64% 0% 1.3%
structure (Figure 3)
Results from the model with a bus
77% 93% 0% 1.4%
structure (Figure 4)

Table 1. Comparison of results for Case 1.

Table 1 clearly shows the conservatism of the IEEE Std. 605-2008 design
method. Even with the same design inputs, a 13% reduction in insulator capacity
usage was provided from the model without a bus structure. This means designers
could lengthen a span, possibly resulting in a construction cost savings as fewer
insulators would be needed. This reduction in insulator capacity usage is likely from
more accurate calculations in the program accounting for rigidity of the insulator as
well as defined bus fitting conditions. Results from the model using the three-phase
bus structure show a 23% reduction resulting in an even longer span because of bus
structure rigidity consideration in the model. It should be noted that higher
reductions are normally expected from a model with bus structures compared to a
model without bus structures. These results, however, could vary due to several
factors, such as bus arrangements, bus structure configurations and materials, and bus
fittings. Also, results from both models, in Figure 3 and Figure 4, show the insulators
at both ends experienced high torsional stresses. Although stresses were within the
limit, at some point torsional strength could govern the design rather than cantilever
strength for this bus arrangement. As expected, tensile and compression loads in a
single-insulator configuration were relatively low compared to cantilever and
torsional loads.

Case 2: Bus conductor fiber stress strength study

For Case 2, the study was similar to Case 1, which was to observe the design
conservatism, but the focus was on the bus conductor fiber stress rather than insulator
strengths. An allowable span was first computed using formulas provided in IEEE
Std. 605-2008 Clause 12.2. The computed span is equal to 43.35 feet based on
pinned-fixed end conditions, which represents bus conductor stress at 100% capacity
usage. The spans in the two models shown in Figure 3 and Figure 4 were then
modified to 43.35 feet, matching the value from the IEEE Std. 605-2008 design
method. Both bus spans experienced the same maximum bus stress for all three
configurations. Results are summarized and shown below in Table 2.

Bus Conductor
Description
Fiber Stress Usage
Results from IEEE Std. 605-2008 design method 100%
Results from the model (Figure 3 with 43.35-ft span) 70%
Results from the model with bus structure
68%
(Figure 4 with 43.35-ft span)

Table 2. Comparison of results for Case 2.

As shown in Table 2, an approximate 30% reduction in bus conductor fiber

stress was observed from results of the two models shown in Figure 3 and Figure 4
compared to the IEEE Std. 605-2008 design method. Reduction in fiber stress from
the models does not necessarily mean that designers could lengthen span because at
43.35 feet span, insulator strength controls the design. This case study was done to
confirm that the IEEE Std. 605-2008 design method is more conservative than the
finite element program approach. Bus conductor fiber stresses from the two models,
with and without a bus structure, were very close. This is because loads were applied
directly at the bus members. Therefore, when determining stresses within bus
conductors using the same loads, the rigidity of bus structures may not impact the
stress as much as it does to insulators. Again, results between the two models, with
and without a bus structure, could vary due to factors discussed in Case 1.

Case 3: Study of overload factors used in load combinations

For Case 3, the intent of the study was to compare results of the same bus
arrangement based on different overload factors required per different standards. The
bus arrangement used in this study is shown in Figure 2, which includes single
insulators at end locations, double insulators at mid-locations, delta insulators at
bi-level locations, and bus conductor bents in one location. All fitting types were
used, including fixed, slip, and expansion. For all analysis in Case 3, deflections on
bus members were limited by L/200, both horizontally and vertically, and ice loads
were excluded when checking bus deflections. All analysis included wind loads
applied to the bus model in X- and Z-direction, as well as wind loads applied in a 45º
angle to the bus model. Additionally, the bus model was analyzed under thermal
loads to determine if proper bus fittings were provided. Seismic loads were excluded
from all analysis. Three sets of load factor (OLF) were analyzed, including:

 Set 1 – according to IEEE Std. 605-2008:

- Similar to calculations in Case 1 and Case 2, wind and short-circuit
overload factors are 2.5 and 1.0, respectively, when checking all
insulator strength characteristics (cantilever, torsion, tensile, and
compression). 100% of insulator strengths (no strength reduction
on manufacturer’s published values) were used to compare with
actual loads.
- No overload factors were used on all applied loads when checking
bus conductor fiber stresses and bus deflections. Stresses and
deflections were compared to allowable values.
 Set 2 – according to ASCE 113:
- LRFD methodology was used when checking all insulator strength
characteristics. Overload factors were based on values provided in
ASCE 113 Chapter 3. Per ASCE 113 Chapter 6, Section 6.9.4, a
50% reduction in insulator strengths was applied when comparing
actual insulator loads.
- ASD methodology was used when checking bus conductor fiber
stresses and bus deflections. No overload factor or reduction in
bus conductor strength was used, similar to Set 1.
 Set 3 – according to utility standard practice:
- It is well recognized that the short circuit load derived from
formulas provided in IEEE Std. 605-2008 is conservative. Fault
clearing time value has a significant impact to the short circuit
load. With the conservatism in the short circuit load, as well as
results experienced in actual systems collected over time, many
utility companies have developed their own standards to analyze
rigid bus systems. For comparison purposes only, a set of
overload factors obtained from American Electric Power (AEP)
 Design Standard was chosen. Designers are still responsible for
determining proper values of overload factors when analyzing their
systems.
- Wind and short circuit overload factors were taken as 2.0 and 1.0,
respectively, when checking all insulator strength characteristics.
Actual insulator loads were compared to 100% of the insulator
strengths.
- Similar to Set 1 and Set 2, no overload factors were used when
checking bus conductor fiber stresses and bus deflections.

Overload factors provided above were used in an extreme wind load case
only. In some cases, overload factors under ice with concurrent wind or seismic load
cases may be different. Designers should ensure that proper overload factors are
provided for all load combinations used for the analysis. Overload factors and results
for all Case 3 analysis were provided in Table 3 and Table 4, respectively.
From the results shown in Table 4, it is obvious that insulator cantilever
strength governed the design for this bus arrangement. Overload factors based on
ASCE 113, Set 2, yielded the maximum insulator usage. Overload factors based on
the IEEE Std. 605-2008, Set 1, provided an approximate 25% reduction in insulator
cantilever strength, and an approximate 40% reduction when using overload factors
from the utility standard practice, Set 3. Maximum insulator tensile and compression
usages occurred at bi-level locations. As expected for insulators in delta
configuration, tensile and compression usages were much higher compared to results
from single configuration, as shown in Table 1. Maximum fiber stress and bus
conductor deflection were the same for all three analyses as results were all obtained
using ASD methodology with no overload factors applied. Bus conductor stresses
and deflections under applied loads in this bus arrangement were within the limit. As
shown from results, it is important that proper overload factor values are used to
design rigid bus system as these values could significantly impact overall
construction costs.

CONCLUSION

Several design methods are available for designing a rigid bus system. The
IEEE Std. 605-2008 design method is widely used and easy to implement. However,
this design method has several limitations. A dynamic analysis approach is another
alternative providing more accurate results but it is more complex as well as time
consuming. Therefore, static analysis using a finite element program was proposed
as it is fairly simple and can address several limitations in the IEEE Std. 605-2008
design method. Three case studies were analyzed using the same design inputs to
compare results between the IEEE Std. 605-2008 design method and from the
proposed method. As shown from the results in Case 1 and Case 2, it is clear that the
IEEE Std. 605-2008 design method is more conservative. In Case 1, results from the
model showed a 13% reduction in cantilever strength on the same bus arrangement,
and a 23% reduction when including a bus structure. In Case 2, results from both
 Overload Factors for Insulator Checks

Description Extreme Wind Ice with Concurrent Thermal

Load Case Wind Load Case Load Case
DL W SC DL Ice W SC DL T
Set 1 – according to IEEE Std. 605-2008 1.0 2.5 1.0 1.0 1.0 2.5 1.0 1.0 1.0
Set 2 – according to ASCE 113 1.1 1.2 0.75 1.1 1.2 1.2 0.75 1.1 1.0
*Set 3 – according to utility standard practice 1.5 2.0 1.0 1.5 1.5 1.5 1.0 1.5 1.0
* (based on American Electric Power (AEP) Design Standard “Station Standards Bus Design Guideline” with permission from AEP)

Table 3. Overload factors utilized in all Case 3 analysis.

Max. Insulator Strengths Max. Max.

Description Fiber Deflection
Cantilever Tensile Torsion Comp. Stress (in.)
Set 1 – according to IEEE Std. 605-2008 75.2% 34.7% 51.3% 12.9% 53.1% 0.92
Set 2 – according to ASCE 113 94.7% 43.2% 66.2% 17.2% 53.1% 0.92
*Set 3 – according to utility standard practice 68.5% 31.3% 47.5% 12.4% 53.1% 0.92
* (based on American Electric Power (AEP) Design Standard “Station Standards Bus Design Guideline” with permission from AEP)

Table 4. Comparison of results for Case 3.

 models, with and without a bus structures, showed an approximate 30% reduction in
bus conductor fiber stress compared to the IEEE Std. 605-2008 design method.
Additionally, results from Case 3 showed impacts to the design from three different
sets of overload factors. In conclusion, with proper design considerations, proper
design inputs, and proper values of overload factors, the static analysis using finite
element program design approach offers more accurate results, which could, in turn,
result in construction cost savings to a project.

REFERENCES

American Electric Power (AEP). Design Standard. Station Standards Bus Design
Guideline, Columbus, Ohio.

American National Standards Institute (ANSI). (ANSI C37.32-2002). American

National Standard for High Voltage Switches, Bus Supports, and Accessories
Schedules of Preferred Ratings, Construction Guidelines, and Specifications,
New York, New York.

American Society of Civil Engineers (ASCE). (ASCE 113). ASCE Manuals and
Reports on Engineering Practice No. 113. Substation Structure Design Guide,
Reston, Virginia.

Amundsen, T.A., Oster, J.L., and Malten, K.C. (2009). “Analytical techniques to
reduce magnetic force from high fault current on rigid bus.” Proceedings of
the 2009 Electrical Transmission and Substation Structures Conference -
Electrical Transmission and Substation Structures 2009: Technology for the
Next Generation, v 363, p 173-183.

Institute of Electrical and Electronics Engineers (IEEE). (IEEE Std. 605-2008).

IEEE Guide for Bus Design in Air Insulated Substations, New York, New
York.

Iordanescu, M., Hardy C., and Nourry, J. (1987). “Structural analysis and testing of
HV busbar assemblies with rigid conductors under short-circuit conditions.”
IEEE Transactions on Power Delivery, v PWRD-2, n 4, p 1082-1089.

Pinkham, T.A. and Killeen, N.D. (1971). “Short circuit forces on station post
insulators.” IEEE Transactions on Power Apparatus Systems, v PAS-90, n 4,
p 1688-1697.

Thaik, A. (1995). “Beware of high stresses in 'A' frame supported two-level bus
structures.” Proceedings of the 57th Annual American Power Conference.
Part 1 (of 3), April 18, 1995 - April 20, 1995, v 57-2, p 864-869.

View publication stats