OR WAIT null SECS
© 2023 MJH Life Sciences™ and Turbomachinery Magazine. All rights reserved.
By Drew Robb
Some shows stress engineering talent. Others are all about vendor experience or academic excellence. The Turbomachinery Symposium — one of the year’s premier events — brings together all three. And 2018 was no exception.
Problems, such as compressor fouling, mechanical seal failure, impeller inefficiency, nitrogen compression, piping vibration, and more were covered in detail. John Crane, MHI Compressor, Siemens, Baker Hughes GE, ExxonMobil, Southwest Research Institute, and Air Products, for example, provided experts to discuss such issues and offer practical solutions.
The 47th Turbomachinery and 34th International Pump User’s Symposia (TPS 2018) attracted 365 exhibitors, and 4,750 attendees representing 45 countries. Engineers and technicians, from novice to experienced, chose from a combination of 18 short courses, 18 lectures, 23 tutorials, 24 discussion groups and 32 case studies.
Eric Petersen, the new Director of the Turbomachinery Laboratory, delivered a welcome address on the first morning of the show.
“I am grateful to all of our authors, leaders, delegates, exhibitors and advisers who make TPS a staple event for the industry,” said Dr. Eric L. Petersen, Turbo Lab director. “This is a unique forum where the best in the industry come to share their expertise.”
He laid out the history and services offered by the lab. For example, it conducts high-pressure flame propagation tests for gas turbines (GTs). Models that are validated in the laboratory can then be used in the field to predict real world performance.
“We also use shock tubes and lasers to study high temperature reactions,” said Petersen. “We have unique devices that give us the ability to predict and measure extreme conditions.”
A 37,000 ft2 facility contains 12 vibration-damped test cells. Extreme temperatures, pressures, rotational speeds, loads, and flow rates can be studied at large scale.
In these tests, a shock wave is generated at around Mach 3 to produce high temperatures. Chemical and high-pressure reactions are studied under conditions commonly experienced by GTs. This helps to predict operation at different temperatures and extremes.
Research, education, and professional workforce development are conducted for rotordynamics, aerodynamics, vibration, heat transfer, tribology and bearings, combustion, and advanced materials. This work addresses challenges, such as durability, reliability, emissions, performance, stability, and efficiency.
Students with advanced degrees research industry problems. Texas A & M is ranked seventh in the world in mechanical engineering based on published research, citations, collaboration with industry and awards. Currently, 14 projects chosen by industry are being funded at $50,000 each.
“As the new Lab Director, I will continue to promote investment in state-of-the-art equipment and test rigs to keep up with current technology and measurement techniques,” said Petersen. “We will nurture multi-disciplinary research that is both fundamental and applied.”
Masaki Shakuda, Engineer, Compressor Engineering and Design Section of MHI Compressor Corp. (MCO), discussed fouling of process gas compressors in ethylene plants. Several types of contamination can often foul the compressor flow path, and plant production is significantly lost.
Wash-oil injection is a traditional remedy: oil is injected through nozzles installed on the suction piping and return bend of each stage. However, fouling material has been reported during turnarounds in machines with relatively few years of operation even though wash-oil injection was carried out at the required intervals.
“Fouling disturbed the gas flow and reduced compressor performance,” said Shakuda. “A greater amount of power had to be consumed to maintain the intended ethylene production level. “
MCO and ExxonMobil worked together to find a more effective approach to develop an improved online wash-oil injection system using CFD/FEM analysis. They carried out a flow pattern analysis on conventional systems.
They discovered that oil was not getting to the entire compressor. Researchers considered the option of increasing the number of nozzles. The size, direction and angles of injection were also considered, along with various parameters in order to develop a new system.
The injection flow rate was decided by adjusting momentum ration of gas phase and liquid phase. This varies depending on user process data, and provides the required washing oil flow rate.
“The new design gave the best oil coverage on both the pressure side and suction side,” said Shakuda. “From a range of delivery of 20°, it was improved to up to 40° of the circumferential area.”
Benjamin White, PE, of Southwest Research Institute, delivered a talk on mechanical stress and flow considerations for the piping design of centrifugal compressors. He addressed a range of factors that must be considered in the piping associated with the installation of any new centrifugal compressor system.
Multiple factors must be balanced in the piping design to have an overall successful installation. The piping must be configured and supported in a manner to safely contain the mechanical forces from the internal fluid pressures, as well as the weight of the piping, fittings and valves.
It must not place any unusually high loads on the compressor itself or any piping supports due to thermal expansion, pressure elongation or weight loads. The layout should result in an even flow velocity profile that does not result in detrimental impact to the aerodynamics performance of the compressor.
For piping surrounding large HP compressors used in refineries and in natural gas transmission, he laid out a recommended sequence. Begin by determining the initial layout for the compressor and associated components. Draw up a process flow diagram, followed by a piping and instrumentation diagram.
Anchor flanges and pipe clamps are often needed for stability, as well as thermal offsets to prevent thermal expansion from influencing the compressor. Pipe diameters should be based on maximum flow velocities.
For example, natural gas pipelines typically operate at 500 to 1,500 psi, with flow velocities for the main piping between 20 and 65 ft/sec. Recycle and anti-surge or blowdown piping, though may have flow velocities up to 120 ft/sec at times.
At low velocity, liquid or settlement can build up at low points. At high-velocity pressure drop, noise, erosion, flow-induced pulsation, and flow turbulence-induced vibration can occur.
“You have to balance various trade-offs between competing objectives,” said White. “There is no obvious standard design for all applications.”
Once piping diameter is calculated, piping thickness should be considered. Internal pressure determines thickness. However, there can be additional concerns related to economics, material temperature limits, suitability of material for certain applications, and so on.
A long list of codes and standards also should be considered. API 617, for example, covers the centrifugal compressor, but not the piping system. In the U.S., most piping systems are based on several ANSI/ASME standards.
However, these codes offer limited guidance on some topics. That is why it is important to do a thermal analysis and a vortex-shedding, induced-pulsation analysis.
The elevated discharge temperatures of compressors cause thermal expansion, as well as shifting loads and stresses. Computer modeling helps to determine the types of piping restraint, their locations, and loads.
“Be sure to look at flange loads on coolers and compressors based on allowable OEM limits,” said White. “The ideal thermal system is flexible enough to allow for thermal expansion without excessive stress but also control vibration.”
That is why you have to balance the findings of thermal and mechanical analyses to find the optimum approach for a specific location. Thermal expansion is a function of pipe length, temperature differential, and the expansion coefficient.
The resulting stress is a function of the amount of thermal growth, pipe size and layout, restraint types and locations, and allowable stress related to the design code, materials, and load case.
Any analysis should consider a variety of temperatures, pressures, differentials, and conditions. Run the analysis with one unit off, then another off, or with ambient condition changes to see how the model is influenced.
For any problems that show up, the usual solution is to change restrain types, locations or quantity to better channel thermal expansion. Further options are to reroute the piping and replace 45° with 90° elbows. The 45° elbows add flexibility and have lower pressure loss, but tend to provide a higher resistance to bending. The 90° elbows also reduce loads in the axial direction by 25% to 50% in some cases, whereas the 45s offer a smoother flow.
Near the compressor itself, anchors and spring supports may be required to reduce loads. But as piping diameter rises, the system gets stiffer and more modifications are needed (pipe stiffness increases exponentially with diameter). This must be done in such a way to allow a certain amount of flexibility.
Compressor flange loads should be minimized. Misalignment of coupling or bearing loads can be issues as well as flange integrity issues. API 617 determines compressor nozzle loads on centrifugal compressors.
“With piping of 16-inch diameter or below, it is fairly easy to meet the criteria and standards, but above that can be challenging,” said White.
When scaling up a system, engineers sometimes make the mistake of doing so linearly. White noted that this results in a different system, and a new analysis is required.
But those are only a few of the factors to consider. Pulsation and vibration are also important. Centrifugal compressors are much less susceptible to vibration than reciprocating compressors. Nevertheless, there are some risks due to machinery-induced pulsation, mixed compression, flow-induced turbulence and piping vibration.
Mixed compression using centrifugal and reciprocating compressors in the same process, for example, can result in pressure and flow fluctuations at the centrifugal compressor. A tee or bend can generate vortices which bring about low frequency vibration.
Common fixes include bigger branches on tees and reducers to lower vortex frequency, valve relocation, or increasing the diameter of the main line.
Small-bore piping is another area that can experience vibration. White’s suggestion was to keep things as short and stiff as possible and avoid fittings with high stress.
“Thermal stress effects on piping should be considered early in the design phase,” he said.
Discussion group at the show
Jack Bagain, Senior Staff Mechanical Engineer at John Crane, covered design aspects and considerations for mechanical seals in midstream pipeline applications. He focused on applications that place significant demand on mechanical seals, particularly those associated with the handling of various fluids through pipelines. These deal with a wide range of conditions, including variable fluid properties, as well as fluctuations in temperature, pressure, and speed.
There are additional challenges in the midstream pipeline sector associated with the remote nature of the installations and limited accessibility, he said. That is why equipment monitoring and preventive maintenance support are placing increased emphasis on the criticality of selecting a robust mechanical seal design and associated support equipment.
“Proper seal selection for a pipeline must consider specific gravity, vapor pressure, operating pressure, shaft speed, the presence of dirt or solids, viscosity, temperature, the equipment involved, and the hydrostatic flange pressure rating,” said Bagain. “High-pressure sealing is difficult if a wide range of pressures exist.”
The pressure ratings for mechanical seals are covered in API 682; however, this does not apply to all seals. API 682 defines maximum allowable pressures for different flange classes. The higher the class of the flange, the lower the number of pumping stations required.
Several different materials that can be used in mechanical seals. Carbon graphite is typically metal (antimony or nickel) impregnated carbons. They are good on fluid lubricity, but relatively weak (probably not to be used beyond 1,200 psi).
Silicon carbide can either be reaction-bonded or self-sintered. These seals are hard, wear resistant, and have good mechanical properties. When viscosity is high, they are often deployed.
Tungsten carbine is more common in midstream pipelines. Silicon carbide and graphite composites offer improved dry-run survivability and thermal shock resistance. 316 stainless steels are also found in midstream applications as they are easy to machine. However, duplex stainless steel has increased tensile and yield strength compared to 316.
“For higher pressures, duplex is better,” said Bagain. “In hardware components bigger than 4.33” diameter, 316 deflects under pressure.”
Robert Benton, Jr., Global Expander Technology Manager for the CryoMachinery Department at Air Products, presented on the successful application of nitrogen turboexpanders-compressors to floating and land-based LNG facilities.
“LNG production in the form of large land-based mega-trains and floating production, storage and offloading (FPSO) facilities can be challenging to design,” said Benton.
He included a couple of specific examples: The Ras Laffan LNG project in Qatar uses 6 × 7.8 mtpa LNG trains with four nitrogen expanders per train. Land-based designs have the advantage of being able to accommodate multiple larger skids throughout the site. In Ras Laffain, though, interconnected piping became an issue due to the distance between some skids.
Accessibility, weight, and lack of space are big issues offshore. Ideally, many systems can be consolidated on one skid. Reconfiguration is often needed to improve accessibility.
“It is necessary to work closely with the EPC contractor to understand and chose the best approach for a combination of requirements and to resolve situational concerns,” said Benton.
High-Mach, high-flow-coefficient impellers for LNG plants can pose design problems. Simone Corbo, Lead Engineer for New Product Introduction at Baker Hughes GE (BHGE), has had to deal with such issues as the new generation of LNG plants progress toward larger footprints.
These compression stations are driven by higher-power GTs (more than 100 MW). This increase in power also necessitates an increase in compression gas flow. As a result, the impeller stages in propane compressors require higher specific flow coefficients and a higher Mach number.
He pointed out that the delivered volume of LNG around the world is increasing on average by 5% each year. Currently, 265 mtpa is produced. But it is expected to grow significantly by 2020, primarily driven by demand in Asia, and Europe to a lesser extent.
With production costs rising, BHGE is turning to standardized configurations, larger train capacities, larger GTs and compressors. But design challenges abound: High flow coefficients, high peripheral Mach numbers, impeller stiffness, transonic and supersonic conditions at the leading edge, rotor shock phenomena, and flow distortion within stator components.
“We undertook a study to mitigate shock effects on rotor efficiency and flow distortion at the impeller exit,” said Corbo. “Stator components must also be able to handle non-uniform flows over a wide range of operating conditions.”
BHGE used 3D Computational Fluid Dynamics (CFD) to devise a different blade shape and surface curvature. It included a mixed flow impeller exit, and a vaneless diffuser. This design brought about a 3% increase in polytropic design efficiency and lowered the design head coefficient by 4%.
Michael Huebner, Principal Engineer at Flowserve, detailed best practices for cartridge mechanical seal installations. The reliability of a mechanical seals depends on more than the design of the seal itself, he said. It also depends on the practices used to install the seal into the centrifugal pump or other equipment.
The actual installation may occur in a controlled environment, such as a factory or repair shop. Alternatively, it could be in the field in an installed pump. The steps taken during installation set the foundation for success of seal in operation.
Different pump and seal designs may require different procedures, but there are key elements: inspection, requirements for equipment condition, preparation for installation, installing the seal, setting the seal drive, and more.
The 2019 Turbomachinery & Pump Symposium is set for September 10–12 in Houston. Short courses will be held in conjunction with the symposia on September 9. To learn more about TPS and see photos from this year’s event, visit tps.tamu.edu.