01-10-2016, 10:49 AM
1457170397-3.4.2.Hanson071454652873286.pdf (Size: 292.3 KB / Downloads: 39)
Investigators
Ronald K. Hanson, Woodard Professor, Mechanical Engineering; Jay B. Jeffries,
Senior Research Engineer, Mechanical Engineering; Xin Zhou, Xiang Liu, Dan
Mattison, Hejie Li, and Adam Klingbeil, Graduate Researchers, Mechanical Engineering
Abstract
A new paradigm of optical absorption sensors has been investigated. Our research
has focused on developing diode laser absorption sensing and demonstrating that it has
potential to provide the immediate feedback needed to optimize the combustion process
in practical environments. Such sensors could be used to improve combustion efficiency
of fossil fuels and thus reduce CO2 greenhouse gas emissions. For the next fifty years,
the improvement in fuel economy will be an important tool to reduce these harmful
emissions. The research reported here confirms that these sensors have the ability to
improve the fuel economy of existing combustion devices as well as enable new highefficiency
combustion strategies.
Multiple new sensor strategies have been developed and demonstrated, all based on
tunable near-infrared (NIR) diode lasers and absorption spectroscopy. We highlight
results using three sensor technologies: 1) a fiber-coupled, wavelength-multiplexed
sensor for gas temperature, H2O, O2 and/or CO suitable for monitoring and controlling
practical combustors (e.g. IC-engines, large-scale power plant combustors), 2) a robust,
rapid-time-response gas temperature sensor suitable for physics-based combustion
control strategies using a control variable proportional to local heat release, and 3) novel
strategies for fuel monitoring needed in active combustion control of instabilities as well
as monitoring and controlling the unburned hydrocarbons in combustor exhaust. Our
long-term goal is to incorporate these new sensor technologies into closed-loop control
strategies to maximize combustion efficiency and/or minimize key emissions (UHC, CO,
and NO) and to suppress lean-blow-out (LBO). These sensors thus offer great promise
for monitoring and control of combustion and energy conversion technologies of the
future.
Introduction
The objective of our research has been the development of advanced sensors that
have the potential to: (1) minimize the environmental impact of energy conversion via
control of combustion-generated pollutants such as NO, CO and unburned hydrocarbons;
(2) reduce greenhouse gas (CO2) by improving combustion efficiency; and (3) monitor
the fugitive emissions from greenhouse gas sequestration efforts. These novel sensors
could enable a new generation of strategies for active monitoring and control of
combustion and energy conversion technologies of the future.
Our research seeks to develop and apply a new class of optical sensors, based on
absorption of light from tunable diode lasers that enable in situ measurements of
temperature and gaseous composition, in real time and in a variety of research-oriented
and practical energy-conversion systems. These sensors have potential to enable
exploratory research on new energy conversion concepts, to expedite the pace of
development of new combustion technologies with reduced pollutant and greenhouse
emissions, and to facilitate gains in performance in existing combustion systems. In
addition, the real-time capability of these sensors will enable explorations of new,
unsteady energy-conversion schemes with the potential for reduced emissions through
real-time control.
The societal impact of improvements in efficiency of existing fossil-fuel
infrastructure can be very substantial. According to the International Energy Agency’s
2002 World Energy Outlook, renewable energy sources (including hydro) provided about
5% of the world’s energy use in 2000, and they project the growth in renewables to
barely keep-up with the growth in demand providing 6.5% in 2030. The energy market
penetration by hydrogen is even smaller, with an estimate that hydrogen technology
might provide 0.01% of the world’s electricity in 2020, which might grow to 1% in 2030.
The tiny penetration of the renewable and hydrogen fuels into the US marketplace means
that incremental improvements to existing fossil fuel use will have a bigger immediate
impact than huge changes in renewable or hydrogen fuel use. Thus, if we can develop
sensors that would reduce fossil fuel use by 5% it would be an equivalent impact as a
doubling of the use of renewable fuels. Similarly a 1% improvement in fossil-fuel
efficiency is predicted to have the same reduction on the atmosphere loading of CO2 as
the doubling of the predicted use of hydrogen fuels in 2030. For the next thirty to fifty
years, these incremental improvements in fossil-fuel utilization will have the biggest
impact on atmospheric CO2 release. The laser-based sensors investigated in this GCEPsponsored
work have the potential to enable this reduction in greenhouse gas emission.
This report highlights three important accomplishments: First, a compact, rapid-timeresponse
wavelength-multiplexed TDL sensor has been developed for in situ
measurement of temperature in an internal combustion engine. The sensor is based on
absorption of light by water vapor (naturally present in air and in re-circulated exhaust
gas). The work (collaborative with the University of Michigan and the Combustion
Research Facility at Sandia National Laboratory) is motivated by the critical importance
of temperature in the development of next-generation combustion strategies, such as
HCCI (homogeneous charge, compression ignition) for internal combustion engines. The
results from these successful demonstration experiments were presented at the 31st
International Symposium on Combustion in August of 2006[publication list #6], the
world’s most prestigious forum to report advances in combustion science and
engineering.
Second, a novel diode laser sensor using water vapor absorption has been developed
for real-time, in situ measurements of temperature in a swirl-stabilized combustor
relevant to gas turbine engines. If gas turbines can be operated closer to their fuel-lean
limit, their production of pollutant NOx and system maintenance can be reduced.
Although the gas composition and temperature are not uniform along the line-of-sight,
we demonstrate that the low-frequency components in the FFT power spectrum of the
sensor signal provide a control signal that enables new control strategies. The use of
diode laser sensors is a new frontier in combustion control, and the promising, yet
preliminary, results of our successful demonstration experiments were presented at the
31st International Symposium on Combustion in August of 2006.[publication list #7]
Third, we have begun to apply a novel mid-IR laser source to the quantitative sensing
of hydrocarbon fuels. Understanding fuel loading is crucial to optimization of a wide
variety of combustion processes. Liquid hydrocarbons are practical for transportation
uses because of the ability to store and transport large amounts of energy that can be
released in combustion. However, the injection of liquid fuels into combustors produces
a complex two-phase mixture of fuel as a gaseous and liquid aerosol. Light scattering
from the fuel aerosol can make the gas phase measurement difficult. However, during
the past year, in a collaborative effort involving partial support from AFOSR and ARO,
we have begun to investigate differential absorption in the mid-IR as a strategy for gas
phase measurement in the presence of aerosol scattering. Although the these results are
preliminary, they are quite promising, and a paper based on this work was presented at
the 31st International Symposium on Combustion in August of 2006.[publication list #5]
The accomplishments highlighted above serve to illustrate the high potential of smart
optical sensors to impact both research and practice of energy-conversion systems. Our
goal is to reduce greenhouse emissions in two ways: (1) by enabling improved
performance of existing systems, such as stationary power plants and internal combustion
engines, while (2) also contributing to research on new energy-conversion schemes, such
as pulsed combustors and fuel cells, where real-time in situ sensing of critical parameters
will hold a key to proper understanding and optimization of such systems. We hope to
increase the leverage of our research in the future by building partnerships with groups
focused on development of new energy conversion schemes, as well as with groups
seeking to reduce greenhouse emissions through improvements in performance of current
combustion systems.
Background
The Stanford University research on smart optical sensors investigates a sensor
strategy that exploits the use of wavelength-multiplexing to combine the beams from
multiple diode lasers onto a single path as shown in Fig. 1.[for references see papers 16,
18, 19, 23, and 15 in the publication list] The optical absorption signal expected in a
practical combustion application is modeled using laboratory-validated spectroscopic
data. These models enable selection of the optimum molecular transitions from the tens
of thousands of potential candidates. The combination of process and spectroscopic
modeling enables the design of smart absorption-based sensors tailored to the specific
combustion application. This design approach was used to choose optimized laser wavelengths for determining gas temperature from a measurement of the ratio of
absorption in two water vapor transitions. Such laser sensors offer the potential for much
faster measurement rates than was feasible with previous instrumentation.
Results
During our GCEP sponsored project, we have made significant progress on three
different TDL sensor technologies, all with good potential for combustion control
applications: 1) a tunable diode laser (TDL) absorption sensor for crank-angle-resolved
in-cylinder temperature measurements in collaboration with DoE via the Sandia National
Laboratory, 2) a rapid-response gas-temperature sensor for control of a swirl-stabilized
flame in collaboration with the AFOSR and the ONR, and 3) a novel mid-IR sensing
technology for vapor phase fuel in the presence of fuel aerosol interference in
collaboration with the AFOSR and ARO. All three of these sensors have the potential for
smart combustion control targeted at reducing the atmospheric CO2 and NOx load from
conventional combustion sources.
TDL sensing of crank-angle-resolved in-cylinder temperature
Innovative combustion concepts offer the potential of internal combustion engines
with improved efficiency and lower pollutant emissions. The development of real-time
sensors for gas temperature for in-cylinder measurements would provide critical new
tools to perfect these advanced combustion concepts, as temperature is a primary
determining factor in combustion chemistry, e.g. for engine cycles of current interest
based on homogeneous-charge-compression-ignition (HCCI). Figure 2 illustrates our
vision of a fully instrumented research IC-engine, where measurements of fuel, air, and
temperature are made in the intake manifold; fuel air, residual gas, and temperature incylinder,
and unburned fuel and pollutants in the exhaust manifold.
Active control of combustion instabilities using a single tunable diode laser
The drive towards improved fuel economy, reduced pollutant emissions (CO,
NOx, etc.) and increased turbine lifetime has prompted interest in combustors that
operate at very lean fuel/air equivalence ratios. However, fuel-lean combustion is
susceptible to instabilities in the form of thermoacoustic oscillations or blowout.
Thermoacoustic instability comes from the coupling of heat release to acoustic (pressure)
oscillations, and leads to decreased combustion efficiency and increased pollutant
emissions. Lean blowout (LBO) causes significant safety hazards and risk-based costs,
and reduces engine lifetime and availability. Therefore, practical operations of lowemission,
fuel-lean gas turbine combustors will require a real-time control system to
suppress acoustic instabilities and prevent LBO.
Gas temperature is an important combustion parameter and thus has potential for use
as a control variable in physics-based control strategies. TDL sensors for gas temperature
have the potential to provide control signals to suppress combustion instabilities and
LBO. TDL sensors have better spatial resolution and less sensitivity to background noise
and luminosity than the traditional pressure (microphone) and chemiluminescent
emission sensors, and thus may offer significant advantages. Therefore, we have
investigated the potential for TDL temperature sensor control of combustion instabilities
and LBO in a swirl-stabilized flame used as a laboratory model of a gas turbine
combustor.
A fast, real-time (2 kHz) temperature sensor using a single tunable, fiber-coupled
telecom diode laser is used for the demonstration measurements. The development and
fundamental design rules for this sensor have been discussed in detail.
Mid-IR Sensing of Hydrocarbon Fuels using Wavelength-Tunable Absorption
Understanding the fuel concentration distribution in practical combustors is of
primary significance because it directly affects operating efficiency and regulated
emissions. For this reason we have actively pursued a new mid-infrared (mid-IR)
technology that will enable optical absorption measurements of fuel concentration in
harsh environments. It has long been recognized that hydrocarbon fuels have strong
absorption features in the mid-IR; however, until recently this sensing strategy has been
limited by the lack of mid-IR laser sources. Although previous workers (including
ourselves) have successfully used the 3.39 μm output from HeNe lasers to monitor fuel
concentration, this approach has several limitations. First, absorption measurements with
a single, fixed wavelength are subject to uncertainty and noise from transmission losses
from window fouling, scattering from fuel aerosol or soot particles, and beam steering
from index of refraction gradients in the target gases. All of these problems plague
measurements in practical combustors. In addition, the HeNe laser is quite noisy, and the
3.39 μm laser light can be so strongly absorbed as to severely limit the dynamic range of
the sensor. We have chosen to address these problems using tunable wavelengthmultiplexed
mid-IR laser light enabled by a novel difference-frequency-generation
(DFG) laser.
A new generation of wavelength-tunable mid-IR light sources has recently become
available. Two near-IR diode lasers are mixed in periodically poled lithium niobate
crystals and a laser beam at the difference frequency in the mid-IR is produced.
Wavelength tuning the near-IR laser produces a wavelength-tuned mid-IR beam, and the
intensity noise on this mid-IR beam reflects the very stable operation of the
telecommunications quality of the input near-IR lasers. The tunability allows selection of
a mid-IR sensor wavelength with appropriate absorption strength to optimize the dynamic
range for amounts of fuel expected in the combustor. These laser sources can be
wavelength- or time-division-multiplexed to enable a sensor beam with two colors of
mid-IR light, which enables using differential absorption techniques to minimize the interference losses in transmission from non-absorption sources including beam steering,
window fouling, scattering, etc.
The first demonstrations of a two-color DFG mid-IR laser for differential absorption
of fuel in a heated gas cell, shock-heated vapor, and shock-heated aerosol were presented
at the 31st International Symposium on Combustion.[publication list #5] The success of
these demonstration measurements suggests that this new sensing strategy has significant
promise for fuel sensing in a wide variety of combustion applications.
Conclusions
During this program, the technique of line-of-sight wavelength-multiplexed
absorption sensing using robust tunable diode lasers has been successfully demonstrated
in a wide variety of practical environments including: large-scale coal-fired power plants
(collaboration with Zolo Technologies), firing piston engines (collaboration with DoE via
Sandia National Laboratories), and spatially inhomogeneous model gas-turbinecombustor
flames (collaboration with AFOSR and ONR via the University of
Cincinnati). These collaborations have significantly amplified the impact of the GCEPsponsored
work by enabling tests of sensor performance in practical combustion
environments. In every application attempted, we have been successful. These results
indicate that wavelength-tunable TDL absorption sensors have the potential to enable
new control strategies in a wide variety of practical combustion devices, increasing
overall efficiency and reducing the atmospheric load of combustion-generated pollutants
and products (especially CO2).
In addition to control of practical combustors, we have demonstrated that TDL
absorption sensors are promising for test-bench diagnostics to enable optimization of a
variety of combustion systems including piston engines (collaboration with DoE via
Sandia National Laboratory), gas turbines (collaboration with GE Research Laboratory),
and new combustion technologies.
Finally, a significant advance in has been achieved by extending TDL sensors to fuel
measurements. In collaboration with projects sponsored by AFOSR and ARO, our GCEP
work has led to the first tunable TDL combustion sensing in the mid-IR, thereby enabling
quantitative sensing of hydrocarbon fuels even in the presence of interference from
aerosol scattering.