OCR Text |
Show _ !If"" ... Velocity Vector, X/V Pl .... t .el ' to, ' •• t Plene 3 eon' '(I",.. t aon T Toll Det. l]lIee t I EPA I fll'td Lue ",.. ~: .... ::tt lI Hn • • • • • A' I '" ... lIt .. " .... " . ..... uti ' \ Collbust Ion [nglneerlng Inc KrelSlnger Qevelopllent labor. tory Fig. 4 - Typical Velocity Output Results Concentratton Measurements - Two types of experimentally based tnformation are used extensively in evaluat1ng isothermal furnace flow models. These techn1Ques wh1ch rely on the measurement of tracer gas concentrat10n w1th1n the model are steady state gas mixing and residence time distr1bution. Typlca11y, the 1nput stream of interest ( For example to simulate fuel streams or sorbent streams ) ls tagged with a tracer gas before injection into the model. Measurements are then made to determine the concentration of this gas mixing 1n the model. By compar1ng the tracer gas concentrat10n at any point to the known concentration of the inlet stream, the Quant1ty of material from the tagged input stream can be determined. Present experimental concentration measurement tests hav~ been carried out at KDL using the previously discussed Laser Absorption Spectrophotometer (LAS). A variation based on use of the LAS Is used to determine the residence time distribution of the tagged Input stream within various sections of the flow model. During such a test the laser beam from the LAS Is directed through the flow model and its 1ntensity is measured as a function of time. Generally the tracer gas added to the input stream is rapidly shut off and the time responsel of the tracer concentration in the model is measured. FACILITIES Burner Te~t Facility -The isothermal flow model1ng 0 Industrial sized burner assembl1es has been carried out 1n a burner model1ng facl11ty designed to Simulate the near fteld aerocttnam1cs of one comer of a tangent1ally- fired furnace. The details of a recent series of mode ling tests using this fac111ty are g1ven in the paper ofl1arlon and Towle (6). Th1s test program, which was carried out as the initial phase of a program to develop retroftttable, tangentially-fired, pulverized coal fuel admission assembl1es with 121 'high load turndown capabt llt1es involved the isothermal flow testing of six 'candidate coal nozzle and nozzle tip designs. A single corner of one . elevation of a commercial C-E designed ut1hty boUer was modeled at a geometriC scale of 0.465. As shown 1n Figure 5 ( taken during flow vl~uali~ation tests), the models were constructed prlmarlly of plexiglass to facilitate flow measurement and visual1zation. The burner ~odels included all significant design features lOclud1ng wlndbox tur01ng vanes dampers fuel nozzle, and auxi Ilary air nozzles. Furthermore, all nozzles and tips could be tilted. Simulated furnace walls extend 1.8 m on one side and 0.6 m on the other side, which allowed for instrument access to the burner exit. Alr was supplied to the burner from a 75 kW forced draft fan through a 0.78 by 0.96 m duct. The model windbox mated to this duct and the total air flow was measured with a cal1brated venturi. Fig. 5 - Isothermal Coal Nozzle Flow Model In this Investigation the near field aerodynamics was of primary lnterest, and this reg10n could be modeling in a non-combusting isothermal flow case without special techniques by assuming that combustion has only moderately progressed and that the average jet density has not changed greatly. Thus, based on our previous model1ng experience, geometriC and dynamic similarity only were mainta1ned. Using geometriC simllarity, aJT dimensions of the prototype firing system were linearly scaled, and the scale was selected to be equivalent to the combustion scale tests ( 60 MBtu/h). For drnamic scaling, the Reynolds number in the mode and prototype were equated. |