Epoxy vs reinforced metal sleeve

In alignment with the ongoing pipeline integrity management system, the operator planned to excavate the location, remediate the identified defects and reinstate the crossing. Wood Group undertook repair design and repair options studies and recommended the use of epoxy sleeves.

However, the operator’s engineering specifications stated that epoxy sleeves could not be used in areas subject to vibration, but that a Type-A sleeve (a tightfitting reinforced metal sleeve) was acceptable. Because of the anomalies on the outside of the pipe, it would have been difficult to fit a metal-on-metal sleeve, whereas the recommended sleeve has a gap between the metal components that is injected with epoxy to make the repair. 

To justify the preferred repair strategy and conclude whether a deviation from the operator’s engineering specifications was applicable, Wood Group carried out vibration analyses of both the epoxy and the Type-A sleeves.

FEA models

In order to validate the proposed epoxy repair, Wood Group’s vibration specialists produced and compared finite element models of the epoxy and the Type-A sleeves. The models were 10 meters long and consisted of the pipeline, two 3-meter sleeves and one 2-meter sleeve. The three sleeves were spaced 10 millimeters apart to allow them to move independently (see Figure 1).

Figure 1: Pipe, sleeves and epoxy layer

Because the critical failure mode would be the tensile adhesion between the epoxy and the steel pipe, or the inter-laminar shear at the boundary, the models required through-thickness unaveraged stress distributions and were meshed with solid finite elements (instead of shell elements).

The models were analyzed to determine the natural frequencies and mode shapes. Figure 2 has the first mode at 33.5 Hz.

Figure 2: Mode 1 for the epoxy repair

Analysis method

There was no measured data available for the site, so an empirically calculated ground-borne velocity profile was used. Predicting the maximum vibration velocity level required taking into account the type of train, the distance from the track, the train speed, a ‘track quality’ factor and a building amplification factor.

Where:

v = vibration velocity (mm/s RMS)
vT = empirically based value depending on type of train
D = distance from track (m)
D0 = reference distance (m)
B = empirically derived constant
S = train speed (km/hr)
S0 = reference train speed (km/hr)
FR = track quality factor
FB = building amplification factor

A conservative approach was taken, where a freight train traveling at 70 km/h passed over a pipeline buried 3.5 m below. This gave the following velocity amplitude profile (Figure 3):

Figure 3: Velocity amplitude profile

Measured vibration velocity spectra indicated that the typical frequency bandwidth for ground-borne vibration from passing trains was 2-200 Hz, so the profile was applied over that range to the two models.

Figure 4 is a typical stress plot showing the maximum value in the epoxy.

Figure 4: Stress in the epoxy at 33.5 Hz when subjected to the velocity profile

Figure 5 shows the response for the maximum nodal stress in the steel sleeve.

Figure 5: Maximum stress response of a node in the steel sleeve

All stresses for the maximum locations in each material were squared, summated and then square-rooted to give a root mean square (RMS) value. These were then multiplied by a crest factor of 5 (to give a quick idea of how much impacting is occurring in a waveform) and then doubled to provide a peak-to-peak (Pk-Pk) level.

Surprising results

Table 1 shows the results for the three critical locations across the two models.

Table 1: Results summary

	Epoxy Sleeve Stress (MPa)	Epoxy Stress (MPa)	Type-A Sleeve Stress (MPa)
Pk-Pk	18.442	1.064	21.342

The Type-A sleeve stress due to the vibrational loading was 21.342 MPa, which is higher than the 18.442 MPa stress found in the metal epoxy sleeve. Therefore, by the operator’s standards, the metal component of the epoxy sleeve is not considered to be an issue when subject to the magnitude of vibrational loading expected at this site.

For the epoxy sleeve repair, a stress of 1.064 MPa was calculated in the epoxy resin layer. For the resin in question, the limiting value would be the tensile adhesion between it and the steel. For this configuration, the value is 12.61 MPa. Therefore the calculated stress of just over 1 MPa has a large safety factor on the allowable tensile adhesion, especially due to the conservative assumptions used.

Larger compactor saves critical time

After the approval of the epoxy repair, the question was asked as to whether the model could be used to determine the groundwork equipment. As the railway line had to be closed for the work, and the window had to be kept to an absolute minimum, one of the main elements was the reinstatement of the site. This entailed compacting layers of backfill around, and then on top of, the repaired pipe. The larger the compactor, the thicker the layers; therefore the time it takes to complete can be reduced.

The excitation characteristics and footprints of the compactors were used in the epoxy model to show the stress levels for the operation, and on this basis, it was possible to use a larger compactor than originally suggested, saving critical downtime of the railway and pipeline.

The typical standard referenced by the compression industry is API 618. While API 618 is a good starting point, it is focused on low- to medium-speed compressor applications that are directly mounted on concrete foundations.

New guideline developed based on GMRC research

The Gas Machinery Research Council (GMRC) identified this industry gap and started a research project in 2011 to develop a guideline specific to separable reciprocating compressors of greater than 2,000 HP and more than 700 rpm. A committee comprised of OEMs, owners, packagers and engineering companies, led by Norm Shade of ACI Services, was formed to develop this guideline. Site interviews, surveys and committee meetings were used to compile a best-in-class set of design practices and analysis approaches based on hundreds of years of collective industry experience. A new guideline entitled “GMRC Guideline for High-Speed Reciprocating Compressor Packages for Natural Gas Storage and Transmission Applications” was released in 2013 describing the results of this project. A three-day training course is also offered by the GMRC to convey the information in the guideline by classroom lectures, case studies and exercises.

Torsional guideline aims to reduce vibration risks

The GMRC project also identified the need for a guideline specific to managing the risk posed by torsional vibration on separable reciprocating compressors. Torsional vibration of the compressor, coupling, and motor or engine driveline can lead to catastrophic failures that result in a significant expense and risk to public safety. A comprehensive guideline called the “GMRC Guideline and Recommended Practice for Control of Torsional Vibration in Direct-Driven Separable Reciprocating Compressor” was prepared as a result of a further GMRC research project in 2013. The guideline provides measures for managing the risk from torsional vibration from the FEED stage of a project through to commissioning.

New guideline for lower-power applications under way

The success of the guideline for high-power reciprocating compressor applications led to the request for a guideline for lower-power applications. A new research program for requirements of reciprocating compressor packages used in field gas, or upstream applications such as gas gathering, gas lift and gas processing is currently in development as a GMRC research project.

Wood Group’s vibration, dynamics & noise team has been fortunate to participate in all of these research projects. Anyone interested in this topic should read the article linked below or contact the GMRC for a copy of the new guidelines and training courses.

• Read the full article Guidelines aim for better high-speed compressor systems (Pipeline & Gas Journal, Norm Shade)
• Get the GMRC guideline

1. Placement of the recycle valve within a hot or cooled bypass loop and the resulting piping lengths and volume
2. The recycle valve’s characteristic curve vs how much gas volume must be relieved from the discharge side of the compressor

Are OEM specifications sufficient when operating multiple units?

Often times, the compressor OEM develops specifications for the surge control system as part of the overall compressor package.  For a standalone compressor, the sizing of the recycle valve and the design of the surge control system can be relatively simple. However, for larger compressor stations with multiple units sharing common headers, station dynamics must be considered in the surge control system analysis and design. In some cases, one unit’s surge controls can affect the amount of shared gas volume in a common suction header – such as the compressor station shown in the figure below.

Surge control system analysis with shared units on common header

One unit causes others to surge, despite correct OEM design

In this recent surge analysis, when Unit 8 underwent an emergency shutdown (ESD) event, the other two units were inadvertently sent into surge by the recycle valves on Unit 8 opening too quickly. One unit’s control system caused the other compressor units to surge unnecessarily, although each individual unit was designed properly by the OEM with sufficient surge controls and recycle valves in place. The only way to understand and design against this dynamic was to model the entire station piping with all three units online at the same time. This approach to integrating the surge model into the compressor station 1-D dynamic flow model allowed the consultant to diagnose the issue and modify Unit 8’s cooled gas recycle valve timing. An acceptable solution was found by delaying the Unit 8 cooled gas recycle valve opening time, to better match the speed descent curve during an ESD and still provide sufficient gas in the common suction header for Unit 6 and 7.

Unprotected surge response of neighboring compressor units to Unit 8 ESD – prior to full surge analysis study

In some cases, the effect of lower suction pressures due to abnormal operations (such as pigging or pipeline blowdown events) may overwhelm the surge control system quickly. Alternatively, pulsating flow and pressure at the inlet of the centrifugal compressor due to nearby reciprocating compressors on a shared header can also cause incipient surge, and require surge control line adjustments or better pulsation controls on the common header. These situations are example cases that traditional OEM surge control system design scopes typically do not include, nor have the capability to model accurately.

Full dynamic surge analysis required to see ‘big picture,’ protect station

Larger scale dynamics of multi-unit pipeline stations must be analyzed in a “big picture” view of a compressor station, which requires a larger flow model of the surge behavior and response of the system. Modeling the station dynamics accurately with multiple units, multiple wheel curves, varying valve opening characteristics and other complications is a challenge that few simulation software applications can handle. Advanced software tools analyze these trickier surge control system effects to assure that the station dynamics and surge behavior of the compressor are modeled in a complete station dynamic flow model, which includes all compressor interactions in a facility.

• Try our free surge sensitivity calculator
• Watch the webinar: What is my risk of surge?

Case study: cold alignment acceptable

An operator had conducted a cold alignment assessment using laser alignment technology on a screw compressor package driven by an electric motor. Under normal operating conditions, the vibration signatures indicated misalignment to be the cause of elevated vibration. The compressor package was shut down for reassessment of the shaft alignment, with results showing that the cold alignment between compressor and driver was acceptable.

Changed operating conditions cause misalignment during run up

The next step was to assess the thermal and dynamic movement of the package during run up and normal operation. The results highlighted significant movement in both horizontal and vertical directions (Figure 1), which was surprising for this type of package and equipment.

Figure 1: Initial thermal and dynamic assessment

The movement of concern in the shaft alignment was a horizontal change resulting from the changed operational and process conditions. Wood Group conducted an inspection and review of the package, and found that the discharge line expanded during operation (Figure 2). This imposed an excessive horizontal force on the non-drive end of the compressor, which in turn caused the coupling to move out of alignment.

Figure 2: The discharge line of this package expanded during operation

Solution

In this case, it was not advisable to compensate for the excessive movement of the discharge line in the alignment settings. Instead, we worked with the operator to review the bracing of the discharge line, which led to reduced thermal and dynamic movement as shown in Figure 3. The improved thermal growth could then be compensated for in the final alignment settings.

Figure 3 - Post bracing thermal and dynamic assessment

There is a high risk that noise in gas compression applications is tonal, or ‘discrete frequency’ noise, which is generated by rotating equipment at a predictable frequency relating to the rotational speed of the shaft and the number of compressor vanes, fan blades, engine pistons or gear teeth. Tonal noise is characterized by pure, spectral tones that occur at a single frequency. Additionally, significant noise can be generated by high flow or pressure drop through a valve, and be transmitted for long distances along piping.

Occupational noise (‘inside the fence’)

Noise risks in the workplace include hearing loss, communication impairment, fatigue and impaired productivity. In this example, tonal, high-frequency noise coming from multiple compressor stations in accommodation and control room units exceeded allowable limits.

As an example, a noise survey was recently completed at an operator’s four compressor stations. Tonal components in certain high-frequency bands at all assessed stations and a tonal component in the low-frequency band at one compressor station were identified as being significant contributors to overall levels. Detailed sound intensity measurements were analyzed on the compressor unit and engine to identify the exact breakout and provide an engineered solution. In this case, lagging was added in targeted locations on the air inlet and the compressor units. Larger silencers were also added to the engines. This example illustrates that noise solutions can eliminate problematic noise levels for operators.

Environmental noise (‘outside the fence’)

Environmental noise around gas production and transmission facilities can cause annoyance, sleep disturbance, decrease in property values and bad public relations. Since smaller compressor stations can be installed in reasonably close proximity to residential premises, residential noise complaints have become a frequent cause for station modifications.

Engineer undertaking a sound intensity survey to quantify noise emissions and determine sound power levels. Based on survey information, noise control implementation can be targeted and optimized to reduce occupational and environmental noise risks to acceptable levels.

In this example, high noise levels were present during commissioning of a compressor station on a site’s boundary, and non-compliance with the assigned environmental noise limits was identified. The noise consultant called by the operator confirmed that at 63 dBA, noise levels on the boundary of the site were above the assigned limits. Using sound intensity equipment, a scan of the compressors was completed. The exhaust piping was identified as the dominating noise source and the installation of acoustic lagging on the pipes was recommended. A follow-up survey confirmed that the noise was brought down to 50 dBA, ensuring compliance below the assigned levels.

Common causes of noise issues

A large number of noise troubleshooting jobs are due to acoustic ‘weak-spots’ in a package. These acoustic ‘weak-spots’ do not fall into any predictable pattern, and consultants find different reasons for high noise emissions:

Typical noise sources are often due to poor design. This can include:

• Air intake duct bellows or transition pieces
• Uninsulated piping
• Exhaust ducts
• Noisy valves (with poorly selected valve-trim)
• Cooling fans

Untypical noise sources can be caused by location-specific construction or maintenance issues, such as:

• Steel structures
• Enclosure roofs
• Mechanical contact between surfaces
• Fan belts

The best approach to address noise risks and avoid costly changes later is to conduct a noise assessment early in the design stage, and incorporate appropriate controls into package selection. If not covered during design, targeting acoustic weak spots with a detailed survey, and implementing appropriate controls, can mitigate high noise levels in order to meet regulatory requirements for occupational health and the environment.

Learn more about noise management

Vibration clamps are superior to standard designs due to:

• Bolt stretch. Insufficient bolt stretch is a well-known issue causing vibration loosening, resulting in failures. Wood Group’s vibration clamps have longer bolts and sleeves, increasing bolt stretch by three times over conventional clamps.
• Material. An extra-rugged construction ensures reliability even in highly demanding environments. All material is primed for corrosion; additional plating and custom finishes are available.
• Damping. For additional vibration control, Wood Group developed the DamperX™ clamp. This optional feature is ideal for resonant or higher vibration applications because it absorbs vibration energy while reducing stress on the surrounding piping system.
• Thermal expansion. Wood Group developed vibration clamps for thermal applications (eg, where temperature cycling occurs). Specialty lined and slotted clamps are designed with vibration control in mind while allowing for thermal growth (providing both stiffness and flexibility).

Vibration clamps are field proven

Wood Group’s vibration clamps are field proven and have been successfully applied in demanding compressor and pumping applications. Additional support includes:

• Piping vibration design services; main line, auxiliary piping and small-bore piping.
• Pipe stress, transients, including water hammer, and other excitation sources (eg, vibration, acoustics and pulsation).
• Field inspections and troubleshooting.
• Additional vibration control solutions (damping, absorbers).

Clamp loosening makes this vital piece equipment redundant in minimizing vibration impact. Wood Group’s vibration clamp is a more effective approach to avoid fatigue failures, reduce downtime and extend asset lifetime.