Gas turbines can offer a high power density. They’ve become the preferred drivers for large compressors and very large pumps for a variety of reasons, such as:
Cost. A train driven by a modern gas turbine can provide appreciable capital cost savings compared to one driven by a steam turbine or an electric motor. These savings can amount to 10–20% of the total installed costs when the infrastructure needed for a steam-turbine-driven train (boilers, boiler-feed-water systems, etc.) or an electric-motor-driven one (a large power plant and the massive electric transmissions/distributions) are included.
Equipment count. Steam-turbine and electric-motor drivers add to the equipment count in a major way but a gas turbine doesn’t.
Plot space. The new generation of gas-turbine-driven trains require a smaller footprint than steam-turbine- or electric-motor-driven units.
Figure 1 shows an example of a modern compact ≈30-MW gas turbine for a compressor drive in a chemical processing unit.
AN IMPORTANT TREND
The chemical industry, like other manufacturing sectors, increasingly is opting for aero-derivative gas turbines instead of traditional heavy-frame industrial-type gas turbines. The aero-derivative units offer higher efficiencies, greater flexibility, more compact design, lighter weight, better on-condition maintenance and faster parts replacement. They permit efficient control of speed and torque together with potential to use integrated and intelligent control and operation concepts.
Aero-derivative gas turbines take advantage of advanced aircraft-engine technologies including modular design. This modularity enables replacing major components without removing the gas turbine from its support mounts. Aero-derivative gas turbines use antifriction bearings in contrast to the hydrodynamic bearings found in heavy-frame industrial-type gas turbines.
Today, aero-derivative gas turbines can offer 10–15% greater efficiency than traditional heavy-frame industrial-type units.
In a hot-end-drive gas-turbine configuration, the output shaft is at the gas turbine end, where the exhaust gas can reach high temperatures. In this design, the load transmission should be fitted through the exhaust duct. This raises several potential issues: long output-shaft length, high-temperature effects on the transmission shaft, exhaust-duct turbulence and pressure drop, and maintenance accessibility. Insufficient attention to any of these details could result in power loss, high vibration, shaft or coupling failure, and increased downtime for maintenance.
In the cold-end-drive configuration, the output shaft extends from the front of the air-compressor. Thus, the driven equipment is accessible and relatively easy to service. However, this design also poses serious drawbacks. The inlet-air system should be configured to accommodate the output shaft and the driven equipment. The inlet duct should be turbulence free. The inlet-air system should provide vortex-free, uniform airflow throughout the operating speed range and various operating conditions. A poor design can lead to catastrophic problems. For example, inlet turbulence can induce surge in the axial air compressor that could result in complete destruction of the gas turbine. So, today, a hot-end-drive usually is chosen.
Single-spool integral-output-shaft gas turbines primarily are used to drive an electric generator. This kind of turbine can’t supply the high torque required to start a large pump or compressor under full pressure. However, providing an electric motor with a variable speed drive (VSD) to help during startup can enable use of such a gas turbine in a train with a large pump or compressor.
A split-output-shaft gas turbine could be considered as a single-spool gas turbine driving a free power turbine. The air-compressor-turbine shaft (air compressor and its turbine shaft assembly) isn’t physically connected to the power turbine shaft. Instead, they are coupled aerodynamically. This makes starting easier in a mechanical drive application. The gas turbine can attain self-sustaining operation before it picks up the load of the driven equipment. The power turbine could be designed to operate at the same speed as the driven equipment, potentially eliminating the need for a gear unit and its losses, which typically run 2–4%. This design is limited to the hot-end-drive configuration. In a dual-spool split-output-shaft gas turbine, independent low- and high-pressure air compressors and turbines generate the hot gases that drive the free power turbine.
A gas turbine should be capable of supplying sufficient torque to drive a particular compressor or pump over its whole specified speed range under all ambient air conditions and operating situations. For example, increasing the fuel and air flow rates and anticipating the turbine inlet-guide-vane opening could help for a gas turbine startup.
QUALIFYING A GAS TURBINE
To use a gas turbine in a mechanical drive application, particularly as a compressor driver, demands an extensive qualification process that includes studies and shop tests. Pay specific attention to the following components for a variable-speed-range application:
• axial-compressor rotating and stationary blades;
• turbine rotor and buckets; and
• combustion systems.
Gas turbine components are subject to a number of failure mechanisms such as low-cycle fatigue crack initiation, cycling crack propagation, disc burst and creep, all of which should be evaluated over the speed variation range.
The ability of the air-compressor rotor and stator blades (and assemblies) to operate in the variable-speed range is important, particularly for single-shaft gas turbines, where the air-compressor shaft is directly coupled to the turbine shaft and driven-equipment rotor assemblies. Figure 2 shows an example of a single-shaft gas turbine rotor.
A ping test (a physical test to determine the natural frequencies of an assembly) is useful for assessing the suitability of a gas turbine for a variable-speed application. The test involves instrumenting the machine assembly with measuring devices and then tapping the system. The ping test can serve to validate the modal analysis of gas turbine assemblies.
Usually, an electric motor with a VSD should be used as a starter-helper for a gas turbine in a mechanical drive application. Evaluate the torque ripple effect contributions from the VSD/electric motor in the gas-turbine qualification process, particularly for a single-shaft machine.
The modal analysis often reveals that a resonance may happen in a defined variable-speed range. For example, a resonance could occur between the turbine-section-bucket natural frequencies and one of the excitation frequencies of a rotor when operating in a variable-speed range, thus potentially limiting the range. In a case study, the minimum threshold of a gas turbine’s variable-speed range had to be held to around 90% of the normal gas-turbine speed because of this effect.
The fuel specification in a chemical processing service could differ from that in other applications. This can result in the gas-turbine pressure ratio exceeding limits, or even could lead to unacceptable combustion-dynamics behavior — and, thus, shortened operating life and poor reliability. Generally, any new fuel requires extensive combustion-system testing verification. Some gas fuels available at process plants could give a very weak flame, which, in turn, may cause a combustion blowout or unsuitable combustion dynamics. This kind of problem demands a sophisticated solution. To deal with this issue, several successful designs feature increased dilution area of the combustion liners to achieve a robust flame coupled with sophisticated combustion-system modifications such as redesigned fuel nozzles.
OPERATION AND MONITORING
In sizing and evaluating any gas-turbine-driven unit, always consider the issue of gas turbine deterioration. Typically, gas turbine drivers are sized with a certain percentage — say, 5–7% — added for deterioration. The overall sizing margin (the gas-turbine-produced mechanical power margin compared to the driven-equipment brake power) could be 12–15%. The causes of gas turbine deterioration fall into two major categories:
1. gas turbine performance deterioration; and
2. mechanical degradation.
In general, performance deterioration poses the greater concern. The total performance decline includes both recoverable (for example, by washing) and non-recoverable losses. Some losses only can be remedied by component replacement or repair.
The recoverable performance loss often is caused by airborne contaminant fouling of internal surfaces, particularly in the air compressor. The site environment and operational profile are the main factors that determine the magnitude of recoverable performance loss and the necessary frequency of washing. Periodic washing of a gas turbine could remove around 90% of the overall airborne fouling. Online washing can increase the interval between major washes, which should be done offline. The best approach is to use a coupled schedule for online and offline washings.
Figure 3 shows an example of internals of a modern gas turbine. Such turbines are the most complex pieces of equipment in today’s chemical plants. So, ordinary methods and tools for condition monitoring won’t suffice. While they may identify abnormal vibrations due to malfunction or degradation, they usually won’t provide insights about the root cause.
Advanced vibration analysis methods, which track changes in vibration patterns over time, are better for monitoring a gas turbine. The actual reason for a high vibration could be a missing piece, a connection bolt issue, a crack, a contact between rotating and stationary parts, or an imbalance caused by degradation, fouling, damage, etc. In many cases, the root cause of the vibration problem is deterioration of connecting bolts or degradation in the connection of different components such as a welded joint. For example, a compromised connecting bolt or a broken connection could prompt a gas-turbine rotating assembly to become unbalanced and, thus, generate high vibrations. When high vibration arises, perform a borescope inspection at the first possible opportunity. In one case study, a borescope investigation after a high vibration alarm revealed that the 1st stage rotating buckets were missing trailing edges. In another case study, a borescope inspection scheduled several days after a high vibration measurement showed that the turbine assembly bolts had cracked and the cracks had propagated under persistent stress. This created imbalance-induced abnormal vibration in the rotor.
In some cases of malfunction, the amplitude of the vibration may not change significantly but the phase angle will exhibit more pronounced change. For example, when a connecting bolt begins to crack, the torque on the bolt will drop, relaxing the joint. This relaxation could result in a change in vibration phase angle.
AMIN ALMASI is a rotating equipment consultant based in Sydney, Australia. E-mail him at email@example.com.