|Dynamic Fuel Injection for Flameless Combustion – |
A Retrofit Option for
Victorian Power Stations?
|by Quanrong Fan, Independent Researcher, Fansmelt|
|Retrofitting power stations to improve fossil fuel combustion efficiency can be a low-cost, and rapidly implemented option to reduce carbon emissions. In recent times, the drive for higher efficiency has led to the introduction of high temperature (supercritical / ultra-supercritical) boilers. However high temperature combustion leads to the formation of nitrogen oxides (NOx), and costly NOx scrubbers are required to prevent air pollution. The high temperatures also require the use of exotic materials, further increasing the installation and maintenance cost of the boiler. For these and other reasons it is generally not possible to retrofit high temperature burners to existing boilers.|
Australia’s current brown coal power stations use standard boiler materials and relatively low temperature combustion, however this comes at the expense of much lower efficiency. Recently there has been interest in so-called MILD or 'flameless' combustion, which by using lower oxygen levels and re-cycled flue gas can reduce NOx production and deliver up to a 30% improvement in thermal efficiency under lab conditions. This article explores how dynamic fuel injection, combined with flameless combustion (which is the subject of a BCIA-supported PhD project – see article MILD Combustion of Pulverised Brown Coal), might be an answer to enabling high efficiency retro-fit for Australia’s power stations.
Quanrong Fan has had success in working with steel-making manufacturers to test a dynamic injection lance for retro-fit applications, and approached BCIA to seek partners in the power sector interested in developing this technology – please contact BCIA if you wish to receive further information.
|Conventional combustion – with a visible flame – has been perceived as natural phenomenon since ancient times and was unchallenged for use in high temperature industries until the arrival of flameless combustion. For flameless combustion, the injection nozzles for fuel and air are arranged so that flue-gas near the burner can be entrained to dilute the fuel and air. Flameless combustion is thus conducted under lower O2 concentration, and within the entire furnace instead of by the formation of bright flame near the burner (see Figure 1 below).|
Flameless combustion can deliver up to an order of magnitude reduction in NOx production. With combustion occurring over the entire furnace, there are no hot spots available in the furnace. NOx production is proportional to the peak temperature of combustion – while it takes a few seconds to produce a substantial amount of NOx at about 1,900K, it only takes a few milliseconds in a flame at a temperature of 2,300K. Added advantages of flameless combustion include reduced thermal stress on the burner and furnace, and reduced burner noise.
|Figure 1: Comparison of flame combustion and flameless combustion by Milani and Wunning.|
|Milani and Wunning(1) defined flameless combustion by a recirculation ratio of Kv which is the mass flow rate of recirculated flue-gas with respect to that of the fuel and air.|
|The ratio Kv depends on the burner design for flue-gas recirculation involved in the air / fuel dilution. When a limited quantity of flue-gas is recirculated for the dilution, combustion takes place with a visible flame attached to the burner as shown in the left photo of Figure 1 (above). To achieve flameless combustion, the burner needs to entrain a large amount of flue gas for significant dilution of the oxygen, thus flameless combustion can be established within the entire volume of the combustion chamber as shown in the right picture of Figure 1 (above).|
Research groups around world have used various terms for flameless combustion such as Flameless Oxidation FLOX, High Temperature Air Combustion HTAC and MILD combustion. To date, all these approaches for flameless combustion have been associated with the use of a stationary burner for the injection of the fuel and air.
Flameless combustion using a stationary burner is constrained by the mixing pattern of flue-gas recirculation, and a burner with one injection nozzle has been used for the illustration in Figure 2 (below). In the stationary burner as shown in Figure 2A (below), fresh fuel / air is injected into the middle of the symmetrical flame, and there is limited space for the flue-gas to be entrained for the air dilution before the combustion.
A fundamental game-changer for the mixing pattern, leading to highly efficient dilution, is proposed by means of dynamic fuel injection using a moveable burner.
|In dynamic injection, mixing is related to the movement of the burner. This firing mode is demonstrated with the burner movement to the right as shown in Figure 2B (below), where the flame is pushed to the left to form an asymmetrical flame due to the drag force of the flue gas. Dilution of fuel / air takes place above the nozzle, where there is 90° angle between the streams of flue-gas and fuel / air, and the extent of mixing of the two streams is related to the movement speed of the burner.|
Flameless combustion in a dynamic injection mode can be expressed by a mixing ratio, Km, which is the ratio of the mass flow rate of flue-gas with respect to that of fuel and air.
|The mixing pattern of dynamic injection is no longer an entrained recirculation of flue gas, but a direct mixing between two streams with a contact angle. The amount of flue-gas involved in dilution is now proportional to the movement speed of the burner. With a slow movement of the burner, only a small amount of flue-gas is available for dilution, and flame combustion can be established as shown in Figure 2B (below). As the movement speed of burner increases, flameless combustion is expected due to the increase in the amount of flue-gas available for direct mixing with the air before the combustion. |
Figure 2C (below) displays the burner changing the moving direction towards the left side; the burning flame is now located on the right side of the burner under the drag force of the flue-gas. The contact mixing of two streams can be seen above the nozzle on the left side of the flame, where mixing dilution of fuel and air can be achieved before combustion.
With continuous movement of the burner, the burning flame is under the drag force of the flue gas at all times. At the lower part of the flame near the nozzle, arriving flue-gas will heat up and dilute the fuel / air, and flameless combustion can be achieved above the self-ignition temperature; near the top of the flame, heat energy generated by the combustion will be removed by the stream of lower-temperature flue-gas.
While further experimentation is required, it is expected that flameless combustion with dynamic injection will have three advantages:
- Improved mixing pattern and dilution of the fuel / air before combustion;
- Improved dilution of combustion due to fuel injection by the dynamic burner across the volume of the furnace, and from different angles; and
- Improved transfer of heat energy to the flue-gas.
|It is expected that flameless combustion with dynamic injection might be obtained at a lower value of the mixing ratio Km due to the contact dilution, and the temperature is expected to be more homogeneous in the combustion chamber due to the dynamic flame.|
The investigation of the flame and flameless combustion by means of the dynamic burner should expand our knowledge and understanding of the burning efficiency, burner design, combustion dynamics and NOx formation. Such a design with related benefits may be well suited for retrofit to improve the efficiency of existing boilers for Australia’s power stations.
|Figure 2: A. Stationary burner with flue recirculation as the mixing pattern; B. Dynamic injection with burner moving to the right direction;|
C. Dynamic injection with burner moving to the left direction.
|(1) Ambrogio Milani, J. G. Wünning, "Flameless Oxidation Technology”, Advanced Combustion and Aerothermal Technologies, NATO Science for Peace and Security Series C: Environmental Security, 2007, pp 343-352.|