The document discusses analyzing stress in a buried branch connection using pipe stress analysis software Caesar II. The initial results using the basic soil model showed unrealistic high stresses at a tee. Further review found the basic soil model had much weaker axial than lateral soil springs. Switching to the more realistic American Lifelines Alliance soil model resulted in stress levels within acceptable limits by providing more balanced soil resistance. The lessons learned were to not rely solely on software and have a deep understanding of its limitations and assumptions.

1. Understanding Piping Stress Analysis
Analyzing Buried Branch
Connections Using Caesar II
Zachary Swartz
June 10, 2015

2. Agenda
1) Introduction to Pipe Stress Analysis
2) Buried Tee Analysis
a) Preliminary Results
b) Determining the problem
3) Solution
4) Lessons Learned

3. Introduction
Pipe Stress Analysis:
• Computer modeling software simulates how
piping system will react under real-world
conditions:
Weight ǁ Temperature ǁ Pressure
• Piping will expand and contract in response to
these conditions

4. • Piping movement:
o creates stress in the pipe
o applies a force to the pipe restraints and connecting
equipment
• Stress levels are checked to ensure they remain
within allowable limits
• Forces on restraints/equipment are checked to
ensure they are not excessive
 Caesar II: Pipe stress analysis software
Introduction

5. New connection to
existing 24” pipeline
Above
Ground
Below
Ground
Above
Ground
Below
Ground
Caesar II Stress Model

6. Modeling Soils in Caesar II
• Caesar II simulates the soil surrounding the
pipeline using springs of varying stiffness
Fig. 1: Caesar II modeling the soil as springs
Stiff SpringWeak Spring
Dense SoilLoose Soil

7. • 2 options for calculating soil springs in Caesar II:
1) Basic Soil Model
2) American Lifelines Alliance (ALA) Soil Model
• Basic Soil Model initially selected:
o Simpler model requiring limited inputs
o Relatively simple theory: “hand calculations”
o Geotechnical report provided soil data to match
Basic Soil Model inputs
Modeling Soils in Caesar II
IMPORTANT! TO BE DISCUSSED

8. Basic Soil Model
• Soil surrounding the pipeline is modeled using
the Basic Soil Model
Fig. 2: Sample
Geotechnical Report
Fig. 3: Basic Soil Model
Input Screenshot
Receive
Geotechnical
Report
Extract Soil Data
Input Data into
Soil Model

9. Caesar II calculated springs
are applied to the model to
simulate the surrounding soil
Above
Ground
Below
Ground
Springs modeled in the lateral
(sideways/vertical) and axial
directions
Caesar II Stress Model

10. Preliminary Results
Buried Tee is overstressed
(Node 770)
INITIAL DIAGNOSIS:
• Pipeline expands, drags the vertical piping with it
• The vertical piping is resisted by the soil packed around it
• Piping breaks at the Tee!
Soil around piping
resists movement

11. Preliminary Results
There’s a stress problem!
Let’s fix it!
a) Make the Tee stronger!
b) Increase wall thickness!
c) Soft padding at the Tee!
$$$$$$$$$$

12. http://thesalesblog.com/wp-content/uploads/2013/11/Screen-Shot-2013-11-11-at-9.38.40-PM.png
On second thought, let me take another look and
get back to you…

13. Preliminary Results
Wait a second…
Detailed stress results at Tee:
o Bending stress > 1 Gpa (1,000,000 kPa)
o Unrealistic! Impossible order of magnitude
o Must be error in the stress model
Fig. 4: Caesar II Stress Results
Tee

14. Reviewing the Model
• Review the soil springs calculated by the Basic
Soil Model:
• Axial spring is much weaker relative to the
lateral springs (Side/Up/Down)
Axial Soil Resistance  WEAK
Lateral Soil Resistance  STRONG
Soil Spring Spring Stiffness (N/cm/mm)
Axial 3
Side 297
Up 297
Down 297
CAUTION!
Fig. 5: Basic Soil Model Soil Spring Summary

15. Fig. 6: Pipeline expansion and Caesar II springs sketch
Reviewing the Model
• Axial resistance (red arrows) = WEAK
• Lateral resistance (blue arrows) = STRONG

16. Fig. 6: Pipeline expansion and
Caesar II springs sketch
Reviewing the Model
• Pipeline expands in the
+ve X direction
• Weak axial resistance =
large pipeline expansion
• Tee moves from original
position χ  χ’

17. 24” pipeline:
• only weak axial springs
affected
No issues
16” branch:
• Pipeline movement pulls
branch in the +ve X
direction
Reviewing the Model
Fig. 6: Pipeline expansion and
Caesar II springs sketch
• strong lateral springs engaged

18. • “Unstoppable force vs.
immovable object”
• Branch must deflect, but to
deflect against strong
lateral springs, a huge force
is required
F = k x
force = Stress
Reviewing the Model
Fig. 6: Pipeline expansion and
Caesar II springs sketch

19. • Basic Soil Model:
o Theory adequate for traditional pipeline analysis
(typically, a single line in a 2D plane)
o Theory translates poorly to unique geometry of
buried branch connection
• Limitation of the software
More realistic soil model is required
More realistic stress results obtained
Secondary Diagnosis

20. Fig. 7: Caesar II Basic Soil Model Fig. 8: Caesar II ALA Soil Model
• American Lifelines Alliance (ALA) soil model used
instead
• Different theory to calculate soil springs
• Additional inputs and information required
ALA Soil Model

21. • More balanced axial/lateral springs obtained
• Recall: lateral springs on branch are a problem
• ALA lateral springs are 27x less stiff than Basic
model
ALA Soil Model
Soil Spring Basic Soil
Spring Stiffness (N/cm/mm)
ALA Soil
Spring Stiffness (N/cm/mm)
Axial 3 11
Side 297 11
Up 297 2
Down 297 28
27x reduction
F = k x
force = Stress
x27x27

22. • Visual comparison of 2 soil models:
Basic Soil Model ALA Soil Model
Basic Model vs. ALA
VERTICAL VERTICAL

23. Stress level OK: 25% of allowable
(12x reduction from previous 294%)
• Less stiff lateral springs  vertical branch can
deflect into the soil with less force required
STRESS IS ACCEPTABLE
ALA Results

24. https://encrypted-tbn2.gstatic.com/images?q=tbn:ANd9GcRskVcSS2Z_SyCYae9zAgQtspvH8dVxPtxkIXnEmdFotzPicEEa

25. • Caesar II Basic Soil Model was initially used:
1) Simpler model
2) Easy to procure and input data for Basic Model
3) Geotech report provided soil inputs in Basic
Model format
• Simple communication issue (or lack thereof)
Client can provide us inputs for either format,
we just have to ask!
Conclusion

26. • ALA Soil Model used after further examination
More realistic soil spring values
• Costly (potentially impossible) modifications were
avoided
Conclusion

27. Lessons Learned
1) Software is an aid, but NOT A REPLACEMENT for
engineering judgement
Engineer must constantly
analyze the results and think,
“does this make sense?”
Engineering is critical thinking,
not “plug and place”
http://www.build-the-body.com/muscle-confusion.html

28. 2) Imperative for the engineer to have a deeper
understanding of the software
 Must understand the
math occurring behind
the scenes
 Software should never be
a calculation “black box”
http://www.chemistry-blog.com/2012/05/04/the-source-code-debate/
Lessons Learned

29. 3) Good reminder that every tool has its limits
 Software can only be as accurate/helpful as
the engineer utilizing it
 “Garbage in, then garbage out”
http://left.mn/2014/02/polymet-knew-now-knew/#sthash.GxGAUAcl.dpuf
Lessons Learned