COMPRESSOR REPAIR, COMPRESSOR DIAPHRAGM DESIGN, STEAM TURBINE EFFICIENCY, BEARING PERFORMANCE AND OFFSHORE INLET AIR FILTRATION
There is no show on the calendar like the Turbomachinery Symposium for bringing together all the key players of our field — a technical gathering that represents the heart of our industry.
This year’s meeting featured in-depth discussions about compressor repair, steam turbine efficiency, subsea reliability, compressor diaphragms, high-performance bearings and offshore inlet-air filtration. Included were contributions from Dresser-Rand (DR), Siemens, GE, Southwest Research Institute (SwRI), GE Lufkin, ExxonMobil, Solar Turbines, Clarcor Industrial Air, Camfil and MAN Diesel & Turbo.
Organized by the Turbomachinery Laboratory (Turbo Lab), part of the Texas A&M Engineering Experiment Station (TEES) & The Texas A&M University System, the symposium comprised 19 lectures, 20 tutorials, 28 discussion groups and 18 case studies as well as 350 exhibiting companies.
“This is the premier conference for turbomachinery and pump professionals developed for the industry and by the industry,” said Dara Childs, Director of the Turbo Lab. “Our goal is to promote professional development, technology transfer, peer networking and information exchange among industry professionals.”
“Market challenges” was the focus of the keynote presentation. Lennart Nilsson, head of technology in the new D-R business within Siemens Power and Gas, began by reminding the audience on just how much change has occurred over past 12 months. In that time crude oil has declined from $100 per barrel to less than $50 and Siemens acquired both D-R and the Rolls-Royce aeroderivative business (RR).
Lennart Nilsson
of Siemens
delivered the
keynote
presentation
“Oil and gas companies are decreasing their investments and reducing CAPEX,” said Nilsson. This means that many greenlighted projects will either be delayed or will have to be revised considerably in order to find a way to do them more cheaply.
“Low oil prices will shut high-cost production technologies out of the market,” said Nilsson. “Long term, if oil prices remain at current levels, a lot of pressure will be put on oil and gas producers to innovate and to find ways to bring costs down via collaboration.” He added that the threshold for economic recovery of around 50% of remaining global oil resources is $60 per barrel.
He cited various oil producers that have announced cutbacks as well as tactical shifts to move their investments to easier projects. They are scaling back efforts in oil sands, the Arctic and deep water to focus on the extraction of more economically sound oil reserves (Figure 1).
Figure 1: Future oil & gas projects are highly dependent on the price of crude oil
“The rotating equipment industry must respond with reduced CAPEX, no more project overruns, higher efficiency, flexibility and the delivery of integrated solutions,” said Nilsson.
This is particularly relevant to Siemens as it just invested a considerable amount on its D-R and RR deals. While this positions the company to compete more broadly in the oil and gas market, the timing of the oil price could not have come at a worse time. But once the company rides out the current market challenge, it has a portfolio now that includes micro LNG technology (LNGo), medium- and large-sized LNG (BOG and EFG compression), and floating LNG. Nilsson reckons the market sweet spot right now for LNG is in the 1- to 3 MTPA category. Siemens also possesses a hermetically sealed compressor for up-stream business especially for sour gases which already has 18,000 hours in operation in an onshore production plant. It can also be used for enhanced oil recovery (EOR) and CO2 injection. But with cash so tight at the moment, Siemens is giving attention to the services space. Nilsson touted his company’s ability to recondition existing rotating equipment to make it suitable for variations in plant capacities, gas compositions and other changing client needs. And regardless of the ups and downs of the economy, he remains bullish about overall prospects.
“Oil and gas is the backbone of the world’s infrastructure and industry, and rotating equipment is the backbone of the oil and gas industry,” concluded Nilsson.
Compressor Development
Technical sessions are always the highlight at the Turbo Symposium. A Tuesday session reviewed the development and testing of a multi-stage, internally cooled diaphragm for a centrifugal compressor. The goal? To reduce by at least 10% the CO2 compression power for carbon capture and sequestration applications.
Experimental demonstration of the design was performed on a full-scale 3 MW, 6-stage back-to-back centrifugal compressor operating in a closed-loop test facility over a range of operating conditions. This pilotscale compression plant at 55,000 lb/hr represents about 5% to 10% of the flow rate of a full-scale coal plant. This resulted in improvements to the diaphragm’s mechanical strength and its manufacturing process.
“Carbon capture was hot five years ago, and although you may not be hearing much about it today, it is a technology that is not going away,” said Jeffrey Moore, Manager of Rotating Machinery Dynamics at SwRI.
He explained that CO2 capture has a compression penalty as high as 8% to 12% in a typical pulverized coal power plant. Further, pressures of 1,500 psi to 2,000 psi are required for pipeline transport for CO2 reinjection. Based on the study of a 400 MW coal plant, therefore, the typical flow rate is 600,000 to 700,000 lb/hr.
During the study, two concepts stood out: isothermal compression and a liquefaction/ liquid pumping scheme.
With true isothermal compression, heat removal is required during the compression process. The other option of compressing then liquefying and pumping also proved problematic as optimizing the liquefaction temperature had numerous tradeoffs and was equipment intensive.
In the end, SwRI opted for a semiisothermal compression approach in a multistage barrel compressor that cools the diaphragms inside the compressor. A pilotscale test loop consisted of a 6-stage, backto- back D-R Datum D12R6B compressor with a suction pressure of 15 to 25 psi, a discharge pressure of 230 to 260 psi, a compressor casing rated for 1,200 psi, a mass flow rate of 55,000 to 75,000 lb/hr and a power output of 3,000 HP.
The train also included a variable-speed induction motor and a variable-speed drive by WEG, a Lufkin gearbox and a Kopflex torquemeter coupling located at the highspeed shaft.
Its cooling circuit design has small coolant channels with thinner walls. This brought about better heat transfer due to the high coolant velocity, frequent direction changes and a counter-flow configuration.
“We made minimal modification of the gas path,” said Jason Kerth, Technology Commercialization Manager focused on Compressed Air Energy Storage products at Dresser-Rand. He said that the conical flow path features were flattened to simplify manufacturing while maintaining the same number of vanes. Diffuser radius ratios were increased in higher flow coefficient stages to increase cooling surface area. All coolant plumbing was kept internal to the bundle.
Tests revealed that cooled diaphragm technology in a six-stage, back-to-back inline centrifugal compressor reduced power requirements by 3% to 8% compared to traditional intercooling between two compressor sections.
“The flow range is extended by up to 4% with cooling, and up to 9% power savings are realized at design flow when operated as a straight-through compressor with no intercooling,” said Moore.
High-performance bearings
With increasing demand for high-performance gearboxes, larger, faster and more highly loaded bearings are requiring more oil and creating more heat than ever before. This means the lubrication systems must be larger to handle increasing heat loads and oil demands.
“Bearing selection is important as bearings can contribute to more than 50% of the oil flow and power loss of the gearbox,” said Josh Ronan, Design Engineer for GE Oil & Gas’ Lufkin products.
Offshore gearbox applications, in particular, are greatly affected due to space constraints and increased lubrication system size and cost. In an effort to reduce the oil flow and heat load requirements for the gearbox, experimental tests were performed on a “pressure dam” (a standard journal bearing with a shallow pocket machined into it to add oil for cooling and stability), and offset half- and tilting-pad journal bearings, operating at and beyond design speeds and loads.
Offset half bearings allow a wedge of oil to develop for stability, while tilting pad bearings, composed of four pads supported by pivots, act independently from each other. All three bearings were six inches in diameter, had bypass cooling and were not allowed go above 250°F. A shaft speed of at least 14,000 rpm was reached.
Field data collection from a full-speed, full-load string test of a compressor train was used to determine the best bearings for turbomachinery gearboxes. Testing concluded that pressure dam bearings experienced excessive metal temperatures at high speed, while offset halves showed diminished performance at speed. Tilting pad bearings, on the other hand, scored well at flow rates of 40 GPM and at speeds of 8,000 to 16,000 rpm.
“The tilting pad bearings remained the coolest by far at high flow rates and speeds,” said Tom Shoup, Engineering Manager at GE Oil & Gas’ Lufkin products. “They satisfied the standard performance requirement at up to 128 m/sec, whereas offset halves did well up to 112 m/sec and pressure dams up to 96 m/sec.”
Improving steam turbines
What can you learn from ten years of experience in design, manufacturing, experimental validation and field application of steam turbines? A lot, according to Lorenzo Cosi, steam turbine engineering manager at GE Oil & Gas.
In particular, Cosi paid attention to lowpressure (LP) stage design characteristics as they typically set limits to the maximum turbine flow and rotating speed. “The LP stage is often the bottleneck for the maximum rotating speed in mechanical drive,” said Cosi. “The compressor could be running at a higher velocity, but the blades can’t go any faster.” Additionally, they exert an influence on turbine efficiency as the LP stage delivers 20% to 40% of the power.
LP stage design presents many structural and aerodynamics challenges. For example, said Cosi, the HS family’s largest (31- inch) last-stage blade has a pull load of 700,000 lbs. It operates in the presence of wet steam droplets, tips speeds beyond 1 Mach and larger variations in condenser pressure. This combination of operating requirements makes design of the LP stages the most challenging aspect in steam turbine engineering.
For this reason, OEMs only customize the intermediate-pressure (IP) and high-pressure (HP) stages. For the LP portion, they prefer to design a family, or set of scaled LP stages, and pick those that match the specified compressor speed.
“The compressor’s rotating speed has been increasing as a trend,” said Cosi. “This improves the performance of the whole train, but has triggered the need to design new families.”
The application of the steam turbine also has to be considered. Mechanical drive has a high condensing pressure and a high mass flow, whereas steam turbines used in power generation have a low condensing pressure and lower mass flow (Figure 2).
Efforts to deal with these factors have led to several advances in the design and manufacturing process in recent years. Diaphragms, for example, can now be machined in a solid ring rather than welded together, which improves throat uniformity, and makes them more robust and easier to assemble.
Research also addressed how erosion occurs: steam condenses between the L1 and L2 blades, with water droplets forming at about 3% moisture. Cosi said that these tiny fog droplet sizes (1 micron) do not cause erosion. But they can gather on nozzle surfaces and detach as a water field composed of larger droplets at the trailing edge.
The action of the compressor accelerates the water, causing it to strike the bucket at high velocities. This is primarily why erosion occurs on the suction side. Surface smoothness can be affected, and in worse cases, material can be lost. Such erosion, though, is more of a factor in power generation than when operating in mechanical drive. “I’ve never seen a blade fail due to erosion on the last stage,” said Cosi. “But this type of erosion has a limited impact on efficiency. The amount of erosion seems to be worst in the first year or two and after that it flattens off.”
Mitigation techniques are many. Blade leading edges can be protected by surface hardening or the use of materials such as M152, stellite (non-magnetic and corrosionresistant cobalt alloys) shields or coatings. Water removal systems are another approach, such as end-wall removal features, hollow nozzles with suction capabilities, and heated nozzles.
“It’s smart to find ways to take extra water from the flow field,” said Cosi. “One good way is to use hollow last-stage blades with radial slots to suck out the condensed water field. “However, this makes manufacture of the diaphragm more complex.”
GE has applied these methods to 26 units with the first running since 2009. It has also conducted field performance tests on six units. Efficiency gains of 5% to 6% in LP section can be attained when these techniques are applied to newer steam turbine blades.
Offshore filtration
Air filtration for offshore gas turbines is difficult due to potential damage caused by ingestion of water and dissolved salt particles, as well as dry salt and other dust (Figure 3). These factors, as well as filtration efficiency, dust-holding capacity and capability to prevent water from entering the gas turbine, make it important to select the right type of filter between high-velocity and low- and medium-velocity systems.
For the operator, it is important to understand the impact of the filtration system on engine fouling or hot corrosion, the required maintenance frequency (for water washing and change of filters, for example), and operating conditions at site, said Rainer Kurz, Manager for Systems Analysis at Solar Turbines.
Salt is always a concern, said Dan Burch, Director of Product Marketing at Clarcor Industrial Air. “It be sticky and can combine with dust to adhere to turbine blades.”
Salt is relatively easy to deal with until higher levels of humidity are present, said Burch, at which point it can change into a liquid and leech into the turbine. It remains dry below a relative humidity (RH) of 40%. But at about 75% RH, it is transformed into a liquid.
Offshore oil platforms require protection from more than just salt. Hydrocarbon vapors, oil mist and other forms of particulate are present and can get sucked into the turbine inlet. Burch’s advice is to look for turbine inlet filtration designed for this harsh environment.
The traditional approach has three stages:
- A vane separator to get rid of bulk water and extend the life of the remaining filters
- A coalescing filter to capture salt aerosols and dust
- A vane separator to remove remaining water.
Such an arrangement is compact, lightweight, and good at removing salt and water. However, its level of filter efficiency (up to F7) means there will be many filter change outs.
Figure 3: Example of damage and fouling offshore
Offshore operating patterns affect filtration as well. This includes longer periods between offline washes, extending the time between outages, more running in dry or high-temperature environments, use of more advanced gas turbines and the growing consumption of high-sulfur fuels.
“The latest requirements for offshore filters are increased submicron particle efficiency, more effective salt removal, longer filter lifespan, and a multi-stage, high-efficiency approach,” said Burch.
This latest generation of filters is similar to traditional three-stage filters. But the third stage is replaced with a hydrophobic filter that prevents water penetration while also capturing dust and dry salt down to 0.3 microns. These filters overcome pressure loss by making them 24 inches deep, which also increases the available surface area for filtration (Figure 4).
“These new filters score filtration levels of up to HEPA 11 and 12 compared to F7 for traditional systems,” said Burch. “Their wet salt efficiency is about 10,000 times better and the pressure loss is only nominally worse.”
Low and medium velocity offshore air filtration, generally defined as systems with inlet velocity between 2.5 and 3.5 m/s, was covered by Jim Benson, Senior Product Engineer with Camfil Power Systems, North America.
“As filters collect airborne contaminants, pressure drop increases,” he said. “Therefore, a medium velocity inlet system will have a low-mean operating pressure drop and long filter service life.”
Filter service life
Air velocity also affects the service life of pre-filters and final filters. Overall impact will vary depending on the type or stage of filtration as well as velocity. Filtration devices, such as weather hoods and vane separators, can exert inertial effects on incoming particles and thereby improve efficiency.
These devices are most effective on particles larger than 10 μm, said Benson. However for final stage filters, which use a filter media, the opposite relationship is true. Because sub-micron efficiency is primarily determined by diffusion-filtration mechanisms, this inverse relationship of velocity and efficiency has the largest impact for these sized particles.
As sub-micron particles, such as salts and hydrocarbons, are responsible for compressor fouling, it is critical that submicron efficiency is maximized to reduce the frequency of off-line water washes so that a high level of engine availability is maintained.
“Filters of a medium velocity system operate in an airflow range where high efficiencies are achieved with minimal pressure drop,” said Benson.
End-of-life equipment repair
When you have old equipment that is no longer functioning efficiently and a projected lifespan of the plant that does not justify replacement, what can you do? That was the question facing Karl Edward, Machinery Engineer at ExxonMobil Exploration and Production Malaysia.
He outlined two cases where non-standard equipment repairs were executed to improve reliability while reducing repair costs and downtime. One concerned the repair of an eroded compressor diaphragm using a brush-plating technique. The second was the repair of two failed gearboxes using parts from both to produce one functional gearbox.
“As upstream oil and gas facilities age, it is a challenge to maintain equipment safety and reliability while managing cost pressures and changing production rates that are quite different from design,’ said Edward. “Sometimes at an aging facility, you can find a unique solution that is hard to replicate elsewhere.”
In case number one, plant personnel noticed a drop in compressor performance along with increased vibration. The compressor was removed for an overhaul at which point they observed eroded vanes on the HP inlet diaphragm.
“We had never seen this before and had not prepared to replace the diaphragms as part of the overhaul,” said Edward. “Welding was rejected as it could make the vanes weaker due to the presence of heat.”
The immediate fix was to use a brushplating process, which involved electrolysis of a copper and nickel alloy and deposition onto the vane in order to increase its thickness from 4mm to 8mm. This repair took about two weeks including vibration testing and verification of performance.
The second case involved the repair of two failed gearboxes using parts from both to produce one functional gearbox. These gearboxes, commissioned in 1979, were attached to gas turbine-driven compressors used for reinjection of gas from oil-producing wells. Coupling failures caused bent high-speed gear rotors.
With the oil fields due to end production within two years, there was no way to achieve a return on investment from a gearbox replacement. “We were tasked with finding a fix to get more life out of these gearboxes,” said Edward.
The solution was to mix and match the high-speed gear rotors. A good rotor from one gearbox replaced the bad rotor in another. Tooth contact checks were done to verify at least 80% contact. The result was that the plant successfully reinstated one compressor train.
That left the plant with two bad rotors to deal with. A repair shop verified that the rotors had no hidden additional damage and could be salvaged.
A three-step repair procedure encompassed a hydraulic press to correct gross eccentricities, surface grinding to restore a 1:20 shaft taper profile to the coupling hub area, and full inspection to confirm that no further damage was sustained during the repair.
Rotor balancing
This shortened the shafts by 1 mm, but reduced the bends from 0.20mm to 0.05mm. Rotor balancing and other tests showed the rotors to be in acceptable condition and fit for operation. The rotors ran for the remainder of the facilities lifespan without incident.
“We achieved our goal of two more years of operations with minimal cost or compressor downtime,” said Edward.
The Turbo Lab is holding its next symposium in Singapore from 22 Feb to 25 Feb. For more information on the Asian Turbomachinery Symposium visit http://atps.tamu.edu/
