The use of pulverised biomass for electrical power generation is of growing importance in the UK as a route to low carbon electricity. It can be used in existing power stations co-fired with coal or 100% biomass firing. However, this use of biomass has led to several major biomass storage or burner feed explosions in recent years. There is minimal information in the open literature on the explosion risks of pulverised biomass, as the fibrous nature of pulverised biomass results in it blocking the injection system of the standard ISO 1 m3 and 20 L spheres. New injection systems for fibrous biomass developed and calibrated for the ISO 1 m3 spherical explosion vessel were used in this research. In addition to the explosion safety data, the experimental methods enabled the measurement of the turbulent spherical flame speed, from which the fundamental laminar burning velocity of the pulverised biomass could be determined, this data is relevant to practical burner design and flame stability.
In dust explosion research the dust concentration has always been reported as g/m3 and not converted to equivalence ratio, Ø. An important feature of the present work was the presentation of the flame propagation properties as a function of equivalence ratio, Ø. This enabled comparison to be made with equivalent burner operating conditions and gas explosions data.
A feature of dust explosions was found, that has rarely been reported elsewhere, and this was that around 50% of the dust that was injected was left as a debris in the vessel after an explosion test. This debris was vacuumed out of the vessel, collected, weighed and analysed. The debris was composed of ash from the biomass that did burn, completely unreacted biomass and partially pyrolysed particles. The mass of the debris was deducted from the mass injected and the actual Ø that the flame propagated through was determined.
Torrefaction is a process involving heating the biomass in an inert atmosphere at about 200°C-300°C, which breaks up the biomass fibres and makes it easier to handle and pulverise. The present work presents the first measurements of the explosion and flame propagation properties of these new biomass materials. The results are compared with the raw biomass from which the torrefied material was derived.
Research was undertaken on the explosion and flame propagation characteristics of a range of raw biomass, torrefied biomass, coal and mixtures of biomass with coal. Fuel characteristics (chemical composition, particle morphology, size distribution) were compared in order to assess the most influential parameters on the reactivity of torrefied and raw biomass. The experimental evidence suggests that pulverised biomass flame propagation occurred in the gas phase, leaving no char residue, indicating that for the biomass that participated in the flame propagation all the mass was burned. Evidence suggested that coal and torrefied biomass flames did result in enhanced char in the debris and that surface reactions through the diffusion of oxygen were part of the flame propagation process.
For minimum explosion concentration measurements the Hartmann tube explosion technique was modified to work repeatably for fibrous biomass and to determine flame speeds. This enabled the most reactive mixture to be determined. The MEC of biomass and torrefied biomass were found to be leaner (Ø=0.2-0.3) than for coal or gaseous hydrocarbons. This supports the conclusion that for the Hartmann equipment all the mass injected must burn, as if only part burned the MEC would be richer. The current methods for determining the MEC in the ISO 1 m3 and 20 L sphere were shown to be invalid as they were based on the injected concentration of dust, with no account taken of the fact that most of it did not burn, so the actual concentration at the lean limit was unknown. More work is required on the reliable determination of MEC.
Torrefied biomass was found to be more reactive than the raw biomass due to the presence of finer particles in the torrefied biomass samples and not due to the material being inherently more reactive. Torrefied, raw biomass and coal samples were found to have KSt values ranging from 60 to 150 barm/s and the maximum explosion pressure ranged between 8 and 9 bar. The mixtures that gave these peak reactivities and pressures was around Ø = 2 – 3, quite different from the peak reactivity of gases at Ø=1.05. The reason for peak reactivity occurring at richer mixtures was addressed as part of the research. Biomass and coal were found to have a similar range of reactivity and peak pressures. Synergistic effects in the reactivity of biomass/coal mixtures were observed with certain fuels and blend ratios. TGA analysis gave indication of such synergistic effects which are likely to occur due to interaction of the fuels during the devolatilisation step. However, no synergistic effects were detected for a mixture containing 50% torrefied biomass.
