Experimental Assessment of Two Stage Combustion of High-PVC Solid Waste with HCl Recovery

Energy recovery from solid wastes is, to a large extent, limited by over-large fractions of PVC (polyvinyl chloride) and other organic chlorine compounds, which can cause problems in the operation of an incinerator, combustor or gasifier. However, PVC behaves differently from most other plastic materials. At temperatures in the range 200-400 °C, PVC decomposes into HCl (hydrogen chloride) and a cokes/char-like residue. This residue can then be burned at higher temperatures as any other chlorine-free solid waste-derived fuel. In this work, a laboratory-scale test facility for two-stage combustion of high-PVC solid waste was built and put into operation at Helsinki University of Technology. The facility (40 kW fuel input) contains two fluidized beds: a bubbling fluidized bed (BFB), operating with nitrogen at ∼350 °C, and a circulating fluidized bed combustor (CFBC), operating at 800-850 °C. Most of the chlorine can be removed (recovered) as HCl from the BFB, which leaves a small percentage to be removed from the flue gases after the CFB. At the same time, this reduces the HCl corrosion problems, with a low risk of formation of chlorinated species such as dioxin and furans. The design work was based on process optimization calculations for a 40 MWthermal plant design case using a process simulation program (PROSIM). The design case was scaled down from 40 MW to 40 kW thermal fuel input. A new module for the pyrolysis reactor was built to be used in the simulation of the two-stage combustion process. The simulation work showed a process thermal efficiency of approx. 36 % compared to 33 %, depending on pyrolysis temperature and the PVC content in the solid waste, for a conventional waste incineration plant. The design of the test facility was also based on kinetic data on the de-hydrochlorination of typical PVC, on combustion of chars from PVC and wood, and on information from the literature. Two types of PVC (bottle-grade PVC and sewage pipe PVC) were used as a fuel in a two-stage combustion test facility. The results from both tests were very promising. The chlorine content in the chars taken from the BFB for the bottle-grade PVC was below 0.1 %-wt (from 50.93 %-wt in the PVC), at chlorine-to-carbon mass ratio < 0.001 kg Cl / kg C. This means that a char (which is forwarded to the CFBC) in Class I (less than 0.15 %-wt Cl) of the Finnish classification of solid recovered fuel (SRF) can be produced using this process. The char chlorine content for the sewage pipe PVC was about 5 %-wt (from 53.54 %-wt in the PVC) at chlorine-to-carbon mass ratio < 0.06 kg Cl / kg C. Also, the char sample analysis of sewage pipe PVC 2 for the determination of polychlorinated toxic compounds showed very small amounts of PCDD (0.78 ng/g, 0.155 ng TEQ/g) and PCDF (2.51 ng/g, 0.35 ng TEQ/g). The result shows that the process has the potential to remove most of the chlorine from a fuel input that contains large amounts of PVC chlorine, i.e., amounts that are problematic for other thermal processes. This despite the use of a cheap perforated distributor plate in both reactors (BFB and CFBC), which may have affected the bed material mixing (especially in the BFB) and the quality of the char that was forwarded to the CFBC. The results also show that the total amount of chlorine released with the CFBC flue gases as HCl was less than 6.5 % of the Cl input with the PVC. The test results show new information about the behavior of different types of PVC in a two-stage combustion process. They confirm that the optimal temperature range for operating the BFB during pyrolysis, 340-350 °C, is the optimum temperature range in which to release most of the chlorine from the PVC. It also shows the stabilizer effects of different PVC types on degradation during pyrolysis in the BFB reactor, which affects the quality of the char forwarded to the CFBC. Many tests to examine the PVC behavior when mixed and pyrolyzed or co-pyrolyzed with other fuels like wood (Finnish pine) and Polish coal were performed in this work. The results show that there is an interaction in PVC / wood pine mixtures that affects the degradation of both fuels. A similar interaction also occurs with the PVC / Polish coal mixture.