Optics and photonics-related technologies have promising potential when it comes to fast, accurate, and in-situ measurements and monitoring of gas turbines and other machinery in the energy sector.
The ability to detect impending problems, whether at a steam, coal, or nuclear power plant or in a natural gas pipeline, is essential for energy efficiency. Non-intrusive and real-time monitoring systems are also critical for preventing leaks, accidents, explosions, and other failures that can have catastrophic consequences for the power supply, workers, and the public.
To make failure prediction and remaining-life assessment efficient and reliable, novel sensors for on-line measurement of machine parameters have become increasingly important.
Most of the commercially available sensors used for optical diagnostics and prognostics operate in the UV to near IR (~2000nm) range. In each spectral domain, the optical sensors can monitor emission, absorption, scattering, or fluorescence.
A laser-based, steam-wetness (quality) sensor can measure the presence of wet steam at the last stage of the turbine bucket, or blade.
Photos courtesy of General Electric
Spectroscopic methods improved significantly after the invention of the semiconductor diode laser in the early 1960s. Pioneered by Ronald K. Hanson and the Hanson Research Group at Stanford University, diode-laser-based diagnostics technology has since found applications in a variety of areas and a versatile range of operating conditions. In particular, tunable diode-laser-based absorption spectroscopy (TDLAS) has proved to be a strong and reliable technique for remote measurements of concentrations and temperatures in harsh environments.
Sensitive measurement capabilities
TDLAS sensors have enabled rapid developments in the field of monitoring and diagnostics for gas and steam turbines. These sensors are mostly based on line-of-sight absorption spectroscopy and provide unique capabilities for fast and accurate in-situ measurements of a variety of important parameters.
Commercially available TDLAS units from Zolo Technologies (USA), CEMTEK Instruments (USA), and TDL Sensors (UK), allow sensitive measurements (up to a few ppm) of CO or CO2, NOx, H2O, and NH3 from the exhaust of gas turbines. This allows for real-time, high-stability measurements, which lead to better emission control through optimization of the DeNOx process.
TDLAS systems are now also frequently used in coal boilers, internal combustion engines, explosion protection systems, and natural gas pipelines for similar applications. A major evolution from single-path measurement has been the 2D tomography of gas species concentration. One such system is Zolo Technologies' ZoloBOSS that maps multiple tunable-diode-laser paths, like a grid, across coal-boiler combustion zone for identifying the burner that needs tuning.
Need for new technology
For combustion diagnostics in gas turbines, a similar 2D-tomography technique would allow for local control of parameters, thus leading to better process optimization. However, the line-of-sight-based diode-laser-absorption technique fails to provide spatially resolved measurements.
Recent research has focused on combining tunable-diode-laser-absorption data with computerized tomography (CT) to determine concentration distribution of chemical species in the combustion zone.
Although optical techniques are non-intrusive and provide fast and accurate measurement of the controlling parameters, they require one or more optical access ports, which can be challenging in a practical application. For example, in grid-like measurements, it is necessary to have multiple ports on the machine, which makes it almost impossible to maintain structural integrity. While this technology has been demonstrated several times in laboratory conditions, successful field implementation in gas turbines has yet to occur.
Laser-based concentration measurements are also applicable when monitoring the performance of steam turbines. Performance monitoring entails continuously evaluating the efficiency of the equipment over time.
Performance calculating procedures for individual turbine sections involve knowing the current and expected turbine section efficiency. The efficiency of low pressure sections of steam turbines cannot be directly determined due to the presence of wet steam in that region. Several reports have pointed out erosion damage of last-stage buckets (or rotating turbine blades) due to wet steam.
To overcome this problem, GE has developed an optical absorption-based steam wetness sensor to directly measure the wetness fraction of saturated steam inside steam turbines.
GE scientist Robert Hall invented the semiconductor diode laser, which is used for the detection of small droplets of water in steam moving at high velocities.
Using optical absorption, two broadband lasers measure the concentration of water and steam individually to estimate wetness. The choices of wavelengths are such that the different phases (water and steam) differ widely to minimize the measurement error.
Combustion temperature test
Another important parameter used as a primary control for the combustion process is the combustion temperature. Temperature values can be measured using TDLAS by studying the absorption line broadening of chemical species.
For instance, vibration bands of H2O between 1.3 and 1.41 μm can provide accurate temperature measurements as they are readily accessible by commercially available diode lasers. However, the selection of line pairs becomes tricky when this technology is used to measure temperature at high-pressure combustion zones.
At pressures above 5 atmospheres, the vibration bands start broadening, and this can cause the spectral features to overlap. This leads to an increase in inaccuracies in the temperature measurement. Some researchers have suggested a multiple-fixed-wavelength technique, instead of a scanned-wavelength strategy, to solve this problem.
System design challenges
In addition to careful selection of absorption bands, measurement of temperature at the combustion zones offers other challenges, such as designing the "pitch" and "catch" optics. Thermal and density variations in these zones cause extensive beam steering and distortion.
Several articles published by the Hanson Research Group highlight the use of large area detectors, high-quality anti-reflective coatings, and wedge-shaped windows for this purpose.
Another challenge with the conventional TDLAS "pitch" and "catch" technique for combustion temperature measurement is the optical accessibility to the combustion zone. Patents presented by Zolo Technologies Inc. and Siemens AG propose using a single optical access port for both "pitch" and "catch" and using the laser reflection from the inner wall of the combustion zone for temperature measurement. So far, a practical demonstration of this technique in gas turbines has not been shown.
Since their development in the 1940s, gas turbines have come a long way in terms of efficiency and sophistication. Major turbine manufacturers such as General Electric and Westinghouse, began by developing turbines for military aviation applications. As gas turbines proved their capability in generating power, the industry split and "power generation" became a separate gas-turbine sector.
Sensors for gas-turbine combustion diagnostics must operate around rotating parts and under harsh conditions.
Photo courtesy of General Electric
Since then, R&D advances, fuel availability, and environmental concerns have consistently driven the growth of gas turbines. Today, gas and combined-cycle power plants are prime movers of the power-generation market.
Along with advancement of material science, the development of cooling technology for gas turbines has allowed machines to operate at higher temperatures, thus producing higher efficiencies. For example, a major focus for R&D during the 1980s involved steam injection into the gas turbine combustor from the Heat Recovery Steam Generator (HRSG) to boost the power output by raising the firing temperature. Higher temperatures, however, also made machines more vulnerable and prone to accidents.
This increased demand of machine-failure prognostics capability, so that possible failures could be predicted and remedial actions taken to prevent major safety incidents and financial losses.
Sensors of the future
Fast, reliable, and accurate prognostics are the ultimate goal for monitoring the health of machines in the energy sector. The ability to detect an impending problem and provide an estimate of remaining time to critical failure is critical for safety, reliable operation, and the prevention of catastrophic failure.
For an accurate and real-time estimation of remaining life, these optical sensors show promise. In the years to come, optics-based monitoring will be a major player in solving some of the toughest problems in the power-generation industry.
- Chayan Mitra, Rachit Sharma, Sandip Maity, and Sameer Vartak work at the Micro and Nano Structures Technology Lab at GE Global Research in Bangalore, India. SPIE members Mitra and Maity are senior scientist and lead scientist, respectively, at the lab. Sharma is a research scientist and Vartak a lead scientist. Their research is focused on the development of optical sensors for energy applications.
Problem with wet steam
Steam can exist in two forms, wet and dry.
The wet steam has water droplets in it in suspension.
If we provide additional heat to the steam, the temperature will remain constant until all water is evaporated. It is then that the temperature increases above saturation temperature and we get superheated steam or dry steam.
In dry steam, we will not see any water droplets in suspension.
Steam at atmospheric pressure has little practical use (apart from humidifiers), as we cannot transport it from one point to another through a pipe under its own pressure. As the pressure is increased, the temperature at which the steam and water co-exists decreases.
In steam turbines used for power generation, even the smallest amounts of suspended water droplets can hit the turbine blades like a bullet, therefore eroding the blades.
The technology to detect these small droplets of water in steam moving at high velocities was enabled by semiconductor diode lasers invented by GE scientist Robert Hall in 1962.
Have a question or comment about this article? Write to us at firstname.lastname@example.org.