OCR Text |
Show through a CaCl 2 bed. The sample is then split into independent streams to each of the analysers. Each analyzer is connected to a N2 line for purge and zero calibration, and also to a specific standard gas for calibration. 3.4. Description of the gas distribution lines There are 3 different gas lines to the burner : the 02 line, natural gas line and N2 line. Figure 2 is a schematic representation of the experimental set up. 02 and natural gas lines are constituted as follovs : First, a safety solenoid interlock viII automaticaly interrupt gas flov in case of emergency . The solenoid is folloved by a manual throtle v~lve, a rotameter and a pressure gauge. At this point the line is split into tvo streams : 1) a line for the pulsating fraction of the gas flovrate, equiped vith either an solenoid electrovalve vhich is driven by a square signal, or a rotary-plug valve, and 2) a by-pass line used to establish the constant fraction of the flovrate. The separate lines converge back into single lines. Dovnstream the rejoining point of the natural gas line, a nitrogen line is used to adjust the N2 content of the natural gas to the desired value. The natural gas available at the Combustion Research Station varied betveen 0.5 and 5%. The balance required to attain 7% or 15% (tvo levels of N2% vere tested) is thus added to the natural gas through the above mentioned line. Further details are given in section 4.2. 3.5. Pulsation generators Commercial solenoid-type electrovalves vere first used to generate the pulsating fraction of the flov. Eventhough they performed satisfactorily throughout the testing period, they are not designed to vithstand continuous operation at the frequencies required by the technique. A rotary-plug valve vas developed to ensure reliable long term operation. A schematic representation of the rotary-plug valve is shovn on figure 3. The design of the rotary plug determines the ratio OIF and the variable speed electric motor alloved setting the pulsation frequency. This rotary valve responded veIl to sustained operation. Experiments indicate that for similar pulsating combustion conditions the solenoid-type electrovalve resulted in less NOx emissions than the rotary-plug valve. This relatively poorer performance of the rotary valve has been associated to plug design considerations, nov resolved. 3.6 Setting F, 0, C, P and ~ A set of various plugs vas fabricated to test different values of the OIF ratio. To set the C/P ratio, valves V1 and V2 are independently adjusted (see figure 2). Vith the solenoid (or rotary) valve closed and V2 totally open, C is set by opening V1 up to the desired value of the flovmeter. Then the electrovalve is set open (vithout pulsation) and V2 is manually adjusted until the appropiate (C+P) value on the flowmeter is established. Once C/P set, it is possible to by-pass the flowmeter during the experiment in order to prevent damage due to pulsation. From one experiment to another, the total flowrate is identical, (C+P) is then 9 |