By Masato Nakamura
Department of Earth and Environmental Engineering
Fu Foundation School of Engineering & Applied Science
The dominant process for the recovery of energy from municipal solid wastes (MSW) is mass burning on a moving grate. Also, the leading grate technology is the Martin reverse acting grate that was modeled in this study. The drying, volatilization, and combustion phenomena during the travel of the solid waste fuel over the grate depend on the physical and chemical properties of the diverse particles in MSW, the design of the grate and combustion chamber, and the distribution of air flowing through the bed of solids. On the average, it takes one hour for particles to traverse a 10-meter long grate. Therefore, the bed behaves essentially as a fixed bed moving downwards, by gravity because of the inclination of the grate. In the case of the Martin grate, the reverse action motion of the grate results in some upward flow and consequent mixing of the materials within the bed.
This study of the flow and mixing phenomena on a reverse acting grate included both theoretical and experimental research. In order to characterize the heterogeneous particle behavior, a two-dimensional stochastic model of particle mixing within the bed on the reciprocating grate was developed. The model was calibrated and validated by means of a full-scale physical model of the reverse acting grate, using tracer particles ranging from 6 – 22 cm in diameter. This size range was based on a quantitative analysis of the size and shape factor distributions of MSW samples collected from different locations in New York City. It was found that different particle sizes result in different residence times because of the Brazil Nut Effect (BNE). In the BNE, larger particles rise to the surface while smaller particles migrate to lower depths of the bed where the reciprocating bars push them backward against the main direction of the MSW flow. The motion of the reverse acting grate (whose speed ranged from 15 to 90 reciprocations/hour) increases the mean residence time of small and medium particles by 69 % and 8%, respectively, while decreasing the mean residence time of large particles by 19%. Also, within this speed range, the mixing diffusion coefficient of each particle size was determined.
Download Thesis (pdf)