"Burning Questions," Feature Article, March 2006

burning questions

Combustion research prepares for the more complex fuel supply of the near future.

by Tim Lieuwen and George Richards

the United States has grown accustomed to a reliable and fairly consistent portfolio of fossil energy sources. Over the past decades, transportation, for instance, has relied on domestic and imported crude oil; domestic coal and natural gas have fueled power generation. Fuel oil and natural gas have heated homes.

Some of the consistency that we have taken for granted is changing. Developments are under way to increase imports of liquefied natural gas. Gasification, meanwhile, is a candidate for a clean way to tap the country's vast coal resources. As a result, the coming decade is likely to bring a much greater diversity in composition and properties of gas fuels than American industry has grown accustomed to. And with that diversity will come both opportunities and challenges.

A key challenge today is environmental emissions
requirements, particularly in regard to nitrogen oxides, which are responsible for smog and acid rain. There is also a growing interest in reducing carbon dioxide emissions. Improving efficiency can reduce CO₂ production and save fuel, but in some applications may also increase the output of NO_x. A future challenge will be to devise ways to continue reducing emissions.

Carbon dioxide reductions can be achieved with hydrogen or oxy-fuel power systems, coupled with permanent CO₂ storage underground, or sequestration, but these systems introduce challenges of their own.

The opportunities are for engineers who can meet the challenges.

Gas From Coal

The United States has significant coal reserves. With advanced technology, that coal can be used very cleanly, with levels of emissions comparable to today's natural gas-fired plants. One way to do this is by gasifying the coal, in essence by placing the coal in a large "pressure cooker" to create a gas of mixed hydrogen and carbon monoxide.

Synthetic gas can be used for a variety of purposes, ranging from the production of petrochemicals to generation of electricity in combined-cycle plants, which combine gas and steam turbine cycles. A few IGCC (integrated gasification combined-cycle) plants are in operation using coal. Coal is gasified to fuel gas turbines, whose hot exhaust is captured to make steam to run a second set of turbines.

Depending upon the source of coal and the gasification technology, the composition of the syngas can vary significantly. For example, the hydrogen levels of syngas at current IGCC installations, an important combustion parameter, range from 10 to 60 percent between sites. This range in fuel composition will complicate the design and operation of modern combustors.

A high-resolution simulation of the flame and flow field in a gas turbine combustor shows unsteady vortex/flame interactions.Such calculations promise to provide future combustor designers with important information.

Modern natural gas combustors at power plants are designed for low NO_x production. They are quite different from older models or even from modern aircraft engine combustors. Older systems mix the fuel and air in the combustion chamber, for a robust, stable flame with a wide turndown range—but also one that produces high levels of NO_x and soot.

Modern designs operate in a premixed mode. Fuel and air are mixed upstream of the combustion chamber. While premixing allows for very low emissions, it also introduces a host of operability issues. First, because the mixture can burn before it reaches the combustion chamber, there is a danger of autoignition (much like knock in an automotive engine). In addition, "flashback," where the flame propagates upstream, can occur. In either instance, high-temperature gases entering regions not designed for the heat can damage parts. In addition, these systems are prone to oscillations, referred to as "combustion instabilities," which result in large amplitude acoustic oscillations that can reduce part life.

Field optimization of these systems often involves difficult balancing of a number of performance demands—for example, low emissions, high power, high turndown, high efficiencies, and low pulsation levels. Because these demands are often conflicting, the allowable space of operation is often quite tight so that variations in fuel composition, ambient conditions, or even part wear can degrade performance and increase pulsation levels if the system is not retuned.

Current IGCC installations use older technology that can handle the varying syngas compositions without too much difficulty. However, the variability in syngas composition is problematic for premixed operation, the preferred mode for future systems. A system designed to operate reliably with one syngas with low hydrogen levels may need to be redesigned, or may require additional measures (such as steam injection) to operate satisfactorily with a higher hydrogen-content fuel.
In addition, the problem of acoustic pulsations is very system-specific and difficult to predict. Measures taken to eliminate a combustion instability problem with one syngas fuel may actually exacerbate the problem with another fuel, and vice versa.

Insight Into Combustion

In order to meet these challenges, one thing is clear: We need to better understand the complexities of combustion. Treating the combustor as a black box will not work. For this reason, industry, government, and academic researchers have teamed up in several projects, primarily sponsored by the U.S. Department of Energy, to study advanced combustion.

Exciting progress has been made, but more work remains. For example, one important property of a flame is its propagation speed. The problem is that we have little knowledge of the flame speed of syngas mixtures at the pressures and temperatures of interest. Work in making these measurements, such as that at Princeton University or Georgia Institute of Technology, is filling in these gaps, but many more measurements are needed.

Once the properties are known, analysts will be able to validate and improve chemical kinetic mechanisms, needed for computational simulation of combustors. Similarly, an experimental testbed called the Simval project has been fabricated at the National Energy Technology Laboratory with the purpose of providing data from a subscale system that can be used to validate computational simulations. Once validated, these models can then be applied to other conditions. Computational models, if built on the right physics, offer exciting opportunities for evaluating the performance of a given design-fuel combination, without the need for expensive tests.

Work also is being performed to develop better sensors and controls so that plants are "smarter" and can adapt well to variations—much in the same way as today's automotive engines do. For example, workers at Oak Ridge National Laboratory and Georgia Tech have developed acoustic techniques that "listen" to the flame to monitor its health.

A schematic of the Simval research project (above) includes a photo of a high-pressure flame as seen through the experimental combustor's optical ports. Catalytic gate SiC devices (below) one day may be used to monitor the hydrogen content of incoming fuels. The metal gate is 100 nanometers thick.

Similar challenges will be posed by a greater diversity of natural gas in the U.S. fuel supply. Because of domestic shortages, there has been a boom in interest in importing liquefied natural gas from Africa, Asia, and South America. Gas composition differs from place to place, so gas from Qatar or Nigeria will not have the same composition as gas from Texas.

On a volume basis, the potential compositional variations in methane, the primary constituent of natural gas, are not substantial, ranging from about 75 to 95 percent. However, offshore sources often contain much higher levels (on a relative basis) of higher hydrocarbons, such as butane or propane. The impact of these variations on properties such as turbulent flame speed or autoignition time are not fully understood, and must be measured to enable future combustor designs to accommodate the widest possible range in fuel composition.

Variable natural gas and syngas pose the same kind of challenges for low-emissions gas turbines, because the devices are usually tightly optimized to meet their ultra-low emissions levels. Fuel composition can change combustion instability characteristics. Unfortunately, we do not understand the combustion process well enough to foresee what the change will be. Combustors are manually tuned to the specific fuel by adjusting various flow splits on the engine.

If the fuel composition remains relatively constant or changes slowly, variations can be dealt with by tuning. The challenge is dealing with swings in composition if these changes occur very rapidly and frequently. One solution being explored is a continuous automated tuning process—as opposed to scheduled manual tuning—that continuously adjusts parameters to optimize performance as the fuel composition, humidity, or ambient temperature changes.

A potential method of accommodating variable fuel composition is to develop technology that can sense and control combustion parameters. For example, a prototype hydrogen concentration sensor is being developed by Michigan State University and the National Energy Technology Laboratory. This sensor is intended to provide a low-cost, rapid measurement of hydrogen concentration in synthetic gas. If implemented in a system, these data can then be fed into a controller, which can make suitable adjustments to the combustor to ensure optimal operation.

Higher energy prices and changes to the regulatory environment have also renewed looks at other fuel sources. For example, the utilization of the gas from landfills due to the decomposition of organic matter is growing rapidly. Combusting these fuels raises interesting challenges because of their low heat content. They can be composed of almost 50 percent of inerts, such as carbon dioxide. Other biomass fuel sources include gasified wood wastes or even gasified chicken waste. Again, a key challenge in such situations is the varying composition of the fuel: The gas produced from one source or gasification process can be quite different from another.

Catch That Carbon

There are also interesting combustion challenges associated with proposed cycles to capture carbon dioxide. CO₂, released during the combustion of any fossil fuel, is a suspected contributor to global warming.

Various studies have shown the potential to capture CO₂ during the coal gasification process, leaving pure hydrogen as a clean fuel. The hydrogen can be used in gas turbines, or supplied to other industrial processes.

Alternatively, syngas, without hydrogen separation, can be burned in an oxy-fuel cycle. In this system, oxygen is supplied by an air separation unit. Burning the fuel with pure oxygen would create a very high-temperature flame. In order to keep the combustion temperatures down, some of the combustion products, CO₂ and H₂O, are recirculated back and mixed with the fuel or air. The water in the exhaust products can be condensed out and the CO₂ pumped into storage in geologic formations such as depleted oil and gas reservoirs, essentially putting the carbon back where it originated. This is commonly referred to as CO₂ sequestration.

In addition, the CO₂ can be used to stimulate the production of marginal oil wells, or to enhance the production of coal bed methane. In these cases, the CO₂, while it is being stored, is useful to enhance production of energy. An example is the Dakota coal gasification plant in North Dakota, where a synthetic natural gas is produced from coal, and the CO₂ from the plant is sent more than 200 miles north to enhance oil-field production in Saskatchewan.

From a combustion standpoint, oxy-fuel systems present new opportunities and issues. After removal of exhaust water by condensation, the remaining exhaust stream is captured for sequestration, leaving no emissions. This simplifies the combustor design, because combustion techniques to avoid NO_x formation are not needed.

Oxy-fuel approaches also can be applied to gas and coal-fired boilers, where increased heat transfer is desirable. However, the oxidizer is no longer free, as oxygen must be supplied from an air separation unit. Consequently, minimizing system costs requires minimizing excess oxygen levels, while maintaining very high combustion efficiency. In contrast, low-NO_x systems burning air typically operate with large amounts of excess oxidizer.

Compared to conventional air-fired combustion, the radiant heat transfer from the hot combustion products to the combustor walls is also a lot higher. This is because the exhaust products are exclusively H₂O and CO₂, both very efficient radiators relative to nitrogen. This may require changes to combustion liner cooling approaches.

A number of market-driven and regulatory forces are motivating a growing diversity in fuel choices and combustion technology. The key challenge for the engineering community is to combust these fuels as has been done over the last decades, but with minimal pollutant levels. Increased understanding of the complexities and intricacies of combustion is enabling these challenges to be met, but a variety of interesting and exciting opportunities remain for continued research and development.

Tim Lieuwen is an associate professor in the School of Aerospace Engineering at the Georgia Institute of Technology in Atlanta and one of the editors of Combustion Instabilities in Gas Turbine Engines. George Richards is focus area leader at the National Energy Technology Laboratory in Pittsburgh and in Morgantown, W.Va.