![]() |
MNM Service What is it? Latest Edition Search Archive
|
Annual EuroFlam Seminar Programme and abstracts - Vol 2 Contributed by Aristide Mbiock and Peter Roberts IJmuiden, NL, 12th July, 1999 - Ref.: 9906art08 Two weeks ago Monday (MNM Week 26) we reported the very successful 6th Annual EuroFlam Seminar, organised at Cardiff University on Wednesday 23rd and Thursday 24th June. There, Visiting Investigators from the three EuroFlam members, ENEL SPI, IFRF and Cardiff University, met together during the two days to present their reports on the projects that they had carried our during the previous year. This group was augmented by a small number of visiting investigators, who are presently in residence and have carried out a substantial part of their project. Last week (MNM Week 27) we commenced the publication of abstracts of the presentations made.
Abstract In November-December 1997 the IFRF carried out a trial applying this techniques for natural gas combustion. A Nippon Furnace Kogyo Co. burner has been tested, the burner consists of a central injector trough with hot air is injected at high velocity and two NG injectors. The two NG injectors supply the fuel at large distance from the central air. In IFRF furnace N.1 a regenerative burner was simulated using a precombustor to reproduce the combustion air at high temperature. The flue gas of the precombustor were split in two steams, one is injected into the furnace after oxygen addition to keep the O2 level at 20.9 % vol. wet, the second one is vented off after cooling it by two heat exchangers. During the two-week trial the flow field and the mixing have been characterized performing detailed in-flame measurements. Performed measurements:
The flow field shows a strong central air jet creating a large recirculation zone. The NG jets are immersing into hot combustion products and they entrain a large amount of combustion products before mixing with the air. The combustion is not intense and it is taken place in large volume in the furnace. The furnace was operating under conditions similar to a well-stirred reactor. The O2 and temperature fields were uniform all over the furnace. The measured O2 concentration is practically in all the furnace volume around 2-3%. The temperature shows a very flat profile. The peak temperature is substantially suppressed. Despite the high combustion air temperature the in flame measured temperature were as low as 1500 0C. That means that the furnace temperature is 200 0C higher than the preheated air temperature. The resulting very flat temperature profile allows reaching a substantial decrease in the NOX emissions. At this temperature level the thermal NOX formation is almost suppressed. The resulting measured heat flux shows a very flat profile along the furnace. Furthermore the measured heat flux are higher compared to standard gas combustion techniques. It was concluded that this technique represents a very interesting tool for the design of new furnaces matching increase in the efficiency with low NOX and CO emissions.
Abstract The NFK regenerative air heater is simulated using a precombustor that operates using natural gas as fuel (1 MWth). The flue gas from the precombustor is split in two streams. One of them is vented off after cooling by two heat exchangers. Oxygen is added to the other flow so its O2 content is kept about 20,9 % vol. wet. After this O2 addition the "vitiated air" is injected into the furnace. This vitiated air simulates the preheated air of the regenerative burner (~ 800 kg/h, 1300ºC, 20,9 % vol. wet, 80 m/s). The fuel input (HFO or LFO) in the furnace was approximately 580 kW. The flue gas O2 concentration was about 2 % and its temperature was approximately 1200ºC. The results of gas composition measurements show a similar behaviour between natural gas (HTAC97), light fuel oil and heavy fuel oil (HTAC98). Due to the slow combustion the chemistry fields are uniform all over the furnace for the three cases. NOx emissions at the furnace exit were 95 and 215 ppmv. dry for LFO and HFO respectively at a flue gas temperature of 1130 ºC for LFO and 1240 ºC for HFO, while around 80 ppmv. dry of NOx enters the furnace with the vitiated air stream. These very low NOx emissions are explained by the elimination of the peak flame temperature in the furnace. Since this temperature does not exceed 1500 ºC it is considered that the thermal mechanism of NOx formation is suppressed, even with the high temperature of the preheated air coming from the precombustor (around 1300ºC). The homogeneity of temperature field is apparent for all the flames. The whole furnace is filled with products possessing temperatures in the range of 1300-1500 ºC for LFO and 1250-1400 ºC for HFO. The total radiative heat fluxes profiles are very flat and high for all fuels. The heat flux values vary from about 310 kW/m2 for the LFO to about 375 kW/m2 for the HFO. Analysis of the measurements collected during this trial reveals that Excess Enthalpy Combustion works well for LFO and HFO. The NOx and CO emissions are very low. The combustion process of the LFO was very similar to combustion of NG. In these two cases, the whole furnace was illuminated, with no visible flames observed. However for the HFO case it is estimated a very high solid concentration and its flame was always present. It is concluded that the quality of atomisation is the key factor for a successful realisation of the Excess Enthalpy Combustion of heavy fuel oil.
The flow pattern of shear layer in the recirculation of a swirl flame Abstract
Previous investigations (Syred and Beer) showed that for both types of flows have the same flow characteristics. In the flow pattern periodic fluctuations can be seen as a large, three dimensional time dependant instability. This is called the "Precessing Vortex Core", PVC. This feature causes the well-known vibration of combustion chambers at low frequencies. The problem is relevant when gas turbines and for example coal furnaces are put into operation. High-speed flow visualisation of a swirling flame showed clearly the movement of the internal shear layer. So the key position of this movement is the internal boundary to the reversal flow associated with the PVC. The subject of interest is the time dependent flow pattern around the internal boundary region. Therefore the experimental rig is an experimental burner, type moveable block. By co-operation with LEAT University of Bochum this device was given to us for use. The approach enables the continuous of previous studies on this rig. Furthermore we have access to the existing detailed processed data including the statistical evaluation of turbulence based upon 50000 validated Doppler bursts. This background gives the knowledge of the mean velocity profiles and the coordinates of the highest turbulence intensities. We can concentrate therefore our focus on the interesting region at an axial velocity component of 0 m/s itself. To undertake these measurements it is useful to use the non-instrusive LDA-system. Especially for this experimental work we have chosen a 2-D system in forward and back scattering mode to have the possibility to cross check the signals. Considering conditions for the rig, only the parameters swirl number and mass flow are variables. Later on the experimental work will be carried out in hot and cold conditions. The collected data about the flow pattern at this region of the flame will create an interesting data base for advanced turbulence modelling. In any case we will get an improved knowledge and an impression about the movement of the internal shear layer associated with the PVC.
Abstract At the end of the EuroFlam contract the Window had been created using the Qt GUI (Graphical User Interface) toolkit and the main widget using the VTKQGL widgets written by Jan Ehrhardt. A new development contract was then taken to finish some of the higher functions of the program. This contract was split into 4 parts, GUI, UF file format reader, Slice data exploration widget and the Crystal eyes stereo rendering function. The first three of these objectives were finished but the Crystal eyes function remains incomplete at this time. The EuroFlam contract was occupied mostly with learning the basics of C++, TCL, VTK, Qt, OpenGL, DBX, GDB, Object oriented programming theory, the UNIX operating system and related programs. I feel that this project has been invaluable to my professional experience and I have enjoyed my time in Italy enormously, thank you very much for excepting me for this project.
|
||||||||||||||
|
|
Technical comments or suggestions should be sent to: mnm@ifrf.net |
| Page designed and executed by IFRF NET |
|