Author: Kraig S. Shipley, P.E., Piping & Fired Heaters, Principal Engineer I; Benjamin M. Lopez, Staff Engineer I

Introduction

Steam hammering is a phenomenon that occurs in steam piping systems which do not adequately control the level of condensate, thus resulting in a condensate level increase. The high-velocity steam flowing over top of the rising condensate creates turbulent flow that creates a wave of condensation in the pipe, which eventually builds to a height such that the ID of the pipe is filled with condensate. This slug of condensate then propels at the higher velocity of the steam down the line, accumulating more condensate as it flows. The momentum of this large slug of condensate hits an elbow or pipe cap, resulting in a large dynamic slug force on the pipe. Such a large force from a steam hammer event can displace the piping system enough to slide the pipe shoes many inches off the structural members on which they rest. Once a shoe has slid off the structure, the shoe acts as a directional limit stop and prevents the pipe from recoiling back to its original position. This article presents a steam hammer analysis case study that leverages E²G’s SIMFLEX-IV software.

How can the Owner-User ensure the piping is fit for service?

E²G | The Equity Engineering Group, Inc. was contracted to answer this exact question on a steam header that experienced a steam hammer event. To frame up the required analytical assessments, let’s start by summarizing what has occurred from normal operating condition to the steam hammer upset condition for this steam header:

• The line is put in service and starts-up: 
  • Typical Sustained (Pressure + Weight) and Expansion (Thermal) stresses were assessed from the original design piping flexibility analysis which shows ASME B31.3 Process Piping Code (B31.3) compliance.
• The steam hammer event occurs:
  • Creates a Force large enough to displace the header primarily axially downstream but with some lateral displacement.
  • The steam header shoes fall off the structural member support downstream at some locations allowing the pipe to drop vertically and result in loss of some support locations.
• The steam hammer event is over:
  • The steam header now wants to elastically recoil back upstream to the normal position but encounters the dropped shoe supports acting as single direction axial restraints preventing the steam header from moving back upstream to its normal position.
• Is the piping system Fit-For-Service in the current restrained operating condition?
  • System now has unexpected axial restraints along with missing vertical supports (those that still have gaps between pipe and structure) where the pipe shoes slid off structural members.
  • A piping flexibility analysis is required to quantify both the Sustained (Pressure + Weight) and Expansion (thermal) stresses with the new piping support configuration.
• Returning the system to original design requires a shutdown:
  • Need to quantify the stresses and support reaction loads at the ambient shutdown case based on the new axial restraints and support activity.

E²G performed a piping stress analysis for the steam hammer event using E²G’s software SIMFLEX-IV to determine the stresses in the current displaced operating condition and what stresses will be present during the shutdown scenario. The as-designed support conditions were also assessed to verify that the original design complied with B31.3.

Inspection to Determine Effect on Piping System

Rather than working through a complex dynamic analysis with numerous unknowns and imposing a calculated time history force to a piping model for the steam hammer event, E²G chose a less costly, simplified approach that could be executed within days to determine if the system can remain in operation. We simplified the approach by applying a static force at the end of the steam header and tuning this force to achieve the measured field displacements. This removes all the complexity of trying to accurately calculate the steam hammer force, and instead quantifies the force from the piping model behavior and compares it to what was observed in the field.

After the steam hammer event, extensive inspection was performed, and all noticeable displacements were measured. The measured maximum displacements were then used for tuning our model forces to mirror the field displacements, and then quantifying the resultant piping stresses and support reaction loads. The biggest challenge during inspection is quantifying the displacement in the piping shoes at support locations. This is because the location of the pipe shoe resting on the support prior to the event is typically not known. The best way to determine displacements during this event is to look at wear marks on the structural members from the pipe shoes, prior inspection photographs, and any pipe shoes which displaced enough to drop off the structural members.
For this assessment, we assumed shoes without photos prior to the event to be installed at the center of the structural members, thus allowing the dimension for maximum displacement to be calculated. Without definitive displacement measurements, a sensitive study based on various displacements was (and should be) performed.

Using SIMFLEX-IV for Piping Stress Analysis

E²G’s SIMFLEX-IV piping stress analysis software was used to calculate sustained and expansion stresses along with support reaction loads for the following load cases:

• Case 1 – As Designed
  • Used to show compliance of as-designed piping system
• Case 2 – As-Is Hot Condition
  • Used to impose displacement occurring during steam hammer event
• Case 3: As-Is Hot Condition – Temporary Solution (Install Teflon Slide Plate)
  • Used to resolve overstresses in Case 2
• Case 4: As-Is Hot Condition – Displacement Sensitivity
  • Used to provide perspective on calculated stress sensitivity to displacement input
• Case 5: As-Is Hot Condition – Displacement Sensitivity with Teflon Slide Plate
  • Used to provide perspective on calculated stress sensitivity to displacement input
• Case 6: Shutdown Condition      
  • Used to determine stresses and reaction loads during a shutdown case with the new axial restraints from support movements (shoes sliding off)
• Case 7: Shutdown Condition for Temporary Solution (Install Teflon Slide Plate)
  • Used to determine stresses during shutdown with new axial restraints from support movements (shoes sliding off) with new Teflon slide plate installed

Initially, E²G performed the piping stress analysis in an “As-Designed” condition matching the piping isometrics to demonstrate ASME B31.3 compliance of the original piping system design. This helps identify areas of higher calculated stresses that may be problem areas when imposing displacements in the “As-Is” cases. Modeling the “As-Designed” piping system also provides the baseline thermal growth to expect at the support locations.

To mirror the displacement determined in the field from the inspection reports, E²G modeled the piping system with displacements at support locations to ensure the calculated stresses in both the sustained and expansion cases are accounting for this displacement. There is difficulty in applying the displacement as a force in the SIMFLEX-IV model as it will only be applied during the sustained or occasional case load case. For this assessment, we want to calculate the expansion stresses with this additional movement from the steam hammer event. Given this, the steam hammer displacements were applied to the piping, which then allowed the resulting stresses to be accounted for in the expansion stress results.

The initial results of the “As-Is” hot condition and shutdown condition indicated significant elastic overstress (225 ksi vs. 30 ksi) in the expansion case with the additional displacement. The location of the overstress was in a branch piping line connecting into the main steam header. This overstress is caused by friction restraining the branch line from moving upstream with the movement of the main steam header after the steam hammer event. E²G proved this theory by running a piping stress analysis in SIMFLEX-IV using an input friction factor for a Teflon slide plate versus the existing steel-on-steel at the branch connection shoe support locations for both the “As-Is” and shutdown cases. The Teflon (friction 0.1) slide plate resultant stress was significantly lower (about 3.75 times lower) than that with steel-on-steel (friction 0.4). Even with the Teflon slide plates, the B31.3 expansion stress was calculated at 59.97 ksi vs. 30 ksi allowable. However, this 30 ksi is the expansion stress allowable from B31.3. For this one-time steam hammer displacement single cycle event, B31.3 provides the leeway to allow the stress to exceed the expansion stress allowable.

ASME B31.1 3S Limit One-Time Displacements

The current ASME B31.3 paragraph 319.2.1(c) states, “Movement due to earth settlement, since it is a single cycle effect, will not significantly influence fatigue life. A displacement stress range greater than that permitted by para. 302.3.5(d) may be allowed if due consideration is given to avoidance of excessive localized strain and end reactions.” ASME BPVC Section III NC-3611.2(f) provides a 3.0Sc acceptance criterion for an expansion allowable stress of a single cycle event. This same 3.0Sc was adopted by ASME B31.1 in paragraph 102.3.2(2). This guidance has been in ASME BPVC Section III NC for more than 40 years. ASME B31.3 is in the process of balloting a modification to paragraph 319.2.1 to adopt this guidance. TN 20-1039 is the current action item in B31.3 to provide a 3.0Sc allowable for displacement stress range for single cycle events and does not require such events to be combined with other displacement stress ranges. E²G leveraged these reference code requirements and good engineering practice to utilize an expansion limit of 3.0Sc allowable for the steam hammer event.

Is Piping Fit for Service?

The results of the piping flexibility analysis cases indicate that ASME B31.3 compliance for sustained stress is achieved for all cases evaluated. However, ASME B31.3 compliance is not met for expansion stress as a result of the downstream displacement imposed from the steam hammer event in the branch piping. To achieve ASME B31.3 compliance for expansion stresses, Teflon slide plates are required to be installed on the branch piping shoe supports to allow the branch line to move upstream with minimal restraint, thus reducing the expansion overstresses. This reduces the expansion stresses to 59.97 ksi, which is below the 3Sc or 60 ksi limit (3*20 ksi).

Inspection should be performed to determine if excessive localized strain (cracking) occurred during the steam hammer event. The SIMFLEX IV piping stress analysis results were used to set the priority of locations to inspect for potential localized cracking from the strain induced during the steam hammer event. These locations were PT inspected for potential surface cracking, which resulted in no evidence of cracking.

With expansion stresses less than 3Sc for the case with added Teflon slide plates installed, and given that there was no evidence of cracking from the inspection, the piping was considered fit for service until the next shutdown. The piping system will then be moved back into place and any damaged pipe supports renewed. A new axial stop was designed and will be installed on the steam header to prevent such movement in case a future steam hammer event occurs. B31.3 compliance was confirmed with the new axial stop. Additionally, engineering operational controls will be implemented to prevent the steam hammer event from reoccurring.

For more information, please submit the form below: