Ronald K. Hanson, Professor, Mechanical Engineering; Jay B. Jeffries, Senior Research Engineer; Xin Zhou, Xiang Liu, Kent H. Lyle, Graduate Researchers, Stanford University

This project strives to develop laser absorption diagnostics into robust, real-time sensors to aid the combustion engineer in designing fuel-efficient, clean-burning combustors of the future.

Absorption spectroscopy has long been an important tool for combustion diagnostics measurements. Our long-term goal will be to combine these sensors into control systems for early identification of combustion instabilities and eventually to develop sensors with sufficient reliability for closed-loop active combustion control. We envision that advanced combustion systems with such active control will lead to substantial improvements in fuel efficiency and air quality.

This project is poised to tackle three fundamental sensing challenges: near-real-time measurement of the temperature of combusting mixtures, rapid and inexpensive measurement of key species in combustion exhaust, and diagnosis of novel hydrocarbon energy-conversion devices.

Real-time temperature measurement will enable combustion engineers to control the combustion process and therefore extract the most work possible from a reacting mixture. Emerging laser technology combined with high-speed laser control and data processing has provided an opportunity to develop such a control system.

As new combustion systems are developed and emissions requirements become ever more stringent, there is a growing need for robust and inexpensive exhaust gas monitoring. Not only can real-time exhaust composition information ensure environmental compliance, it can provide the combustion control system with valuable performance feedback. This is particularly true of highly strained combustion systems such as those proposed by C.F. Edwards and C.T. Bowman.

Fuel cells are an energy technology with significant efficiency benefits, but relatively little mass-production experience. As this technology matures, there will be a need for low-cost fuel-quality and performance sensors.

Combustion Temperature: Tunable diode lasers in the near-infrared have recently been developed for the telecommunications industry and are available at wavelengths overlapping the water vapor absorption bands near 1400 and 1800 nm. These devices offer the opportunity for robust real-time gas temperature monitors. Temperature is determined from the ratio of absorbance on transitions originating from different ground ro-vibrational states of H₂O. Multiplexed-wavelength sensors have been demonstrated which combine several diode lasers for simultaneously probing multiple absorption transitions (Figure 1).

However, when the absorptions are less than 1%, the analysis of direct absorption data becomes more difficult and time consuming, which limits the time response when used as a control sensor. In addition, the temperature is determined by the ratio of two small signals and thus can have a large uncertainty. Wavelength modulation strategies are currently being investigated to ameliorate both of these problems. First, wavelength modulation yields inherently a zero baseline signal, greatly reducing the data reduction time and uncertainty. Second, wavelength modulation has long been used for sensitive detection of trace concentrations of species; the typical minimum detectable absorbance for direct absorption is 10^-4 while optimized limits for wavelength modulation absorption spectroscopy are often reported as low as 10^-6.

Exhaust Monitoring: Fundamental vibrational transitions in several combustion molecules occur in the mid-infrared from 3-10 µm. These fundamental transitions have significantly stronger absorption than the overtone vibrational transitions in the near-IR. For important non-hydride species such as CO or NO, the absorption strength declines roughly two orders of magnitude for each overtone order. Thus, mid-IR absorption sensors are attractive for sensitive detection of minority combustion species such as the pollutants NO and CO. New laser sources based on quantum cascade architecture are emerging that provide laser light in the mid-IR from compact, room-temperature diode sources.

A new approach is needed for detection of hydrocarbon fuel molecules, as their rich ro-vibrational spectrum is comprised of many individual lines blended into unresolved features extending beyond the scan range of a single diode laser. This difficulty is exacerbated at the elevated pressures of many combustion systems. Recently we have sought to obviate this difficulty with a new diagnostic strategy, namely the use of differential absorption, and promising results have been obtained on relevant alkanes using vibrational overtones near 1.7 microns. We envision that this new sensing approach may develop into a general strategy for sensitive detection of species with broad, unstructured absorption spectra typical of hydrocarbon fuels.

Fuel Cell Sensors: Optical surface diagnostic strategies of sum (difference) frequency generation and infrared absorption have the needed species selectivity for fuel cell research, but their spatial resolution is limited by the wavelength of the light. It is concluded that a hybrid diagnostic combining optical methods with a tapered fiber probe has the potential to make the in situ, species-specific and spatially-resolved measurements on the catalyst-doped membrane surface needed for fundamental understanding of fuel cell chemistry/transport. Advanced sensors also have the potential to solve several practical problems for power plant operation. One problem identified is the "peak-shaving" practice of adding air and propane to natural gas during periods of high demand in New England. Although the mixture maintains constant BTU fuel for standard combustion facilities, the air load can be 10-20% which reduces efficiency of the MCFC electrical production while increasing the temperature of the fuel cell stack. Diode laser sensors for O₂ may provide an effective means of monitoring the natural gas purity.