OCR Text |
Show Figure 2 shows experimental results that illustrate the fuel rich injection concept. The rich zone stoichiometry was 0.99 and the lean zone stoichiometry was 1.02. Here the selective reducing agent, ammonium sulfate, was injected into the rich zone simultaneously with the burnout air at various temperatures (the comparable excess air injection results in Fig. 1 are shown for reference). The temperatures reported here refer to the point immediately prior to air/agent addition. This approach differs in two ways from the conditions used in Fig. 1. The first is the application of staging. Secondly, the final stoichiometry of the lean zone is 1.02 rather than the 1.25 used in Fig. 1. The fuel rich injection both extends the effective temperature range to lower values and increases the NO removal efficiency. In Fig. 3 the reduction of NO is plotted against the temperature at whi ch the agent was introduced into the furnace. These experiments di ffer from those of Fig. 2 in that the agent was introduced into the rich zone at va r yin g 1 0 cat ion s / temp era t u res rat her t han wit h the burn 0 uta i r • In all cases the burnout air was introduced at 7850 C. Nitric oxide was reduced at all temperatures; however, reduction was 'most favored by low temperatures. These data can be rationalized if high temperature injection is assumed to prov; de the time and temperature necessary to deacti vate a portion of the agent, i.e., convert it to N2. Alternatively, low temperature injection may not provide sufficient opportunity for the thermal decomposition of the solid agents. Note that cyanuric acid becomes a less effective agent below 10750 C. Such behavior would be expected if cyanuric acid sublimation/decomposition became a rate limiting step as temperature was reduced. Fi gure 3 showed the inf1 uence of the temperature at which the agents were injected into the rich zone, with the burnout air injected at a constant temperature. This is contrasted with Fig. 4, which shows the influence of varying the temperature at the burnout point, with the agents injected into the rich zone at 9000 C. The principal observations from Fig. 4 are that the minimum NO emissions are obtained at a lower temperature than overall fuel lean injection (6500C as opposed to 9600C in Fig. 1), that the minimum NO emission is generally lower, and that high NO reduction is obtained over a broader range of temperatures for the sol i d agents. Among the three sol i d agents, cyanuric acid yields the least NO reduction and ammonium sulfate the most. Ammonia injection was characterized by a much narrower active temperature range. Figure 5 describes the influence of rich zone stoichiometry on NO reduction. The solid symbols show the NO concentration at the exit of the lean stage in the absence of any agent. The reduction under fuel rich conditions is due to the staged combustion. The open symbols represent a repeat of this test with the addition of an ammonium sulfate solution spray into the rich zone. The agent causes the overall NO at the exit to show a sharp minimum slightly on the rich side of stoichiometric. Even outside of this minimum, however, the NO reductions remain high. For example, at SRI = 0.9 the final NO relative to that leaving the first zone represents a 60 percent reduction. In this figure, fuel lean conditions were obtained by adjusting the first zone stoichiometry, i.e., the entire system was operated without staging. The reduction noted under lean conditions is due to the 8 |