- Email: [email protected]
Suppression mechanism of Al dust explosion by melamine polyphosphate and melamine cyanurate
Journal of Hazardous Materials xxx (xxxx) xxxx
Contents lists available at ScienceDirect
Journal of Hazardous Materials journal homepage: www.elsevier.com/locate/jhazmat
Suppression mechanism of Al dust explosion by melamine polyphosphate and melamine cyanurate Haipeng Jiang, Mingshu Bi, Wei Gao* School of Chemical Engineering, Dalian University of Technology, Dalian 116024, China
G R A P H I C A L A B S T R A C T
A R T I C LE I N FO
A B S T R A C T
Editor: G. Lybertos
The suppression mechanism of melamine polyphosphate (MPP) and melamine cyanurate (MCA) for Al dust explosions is investigated experimentally and computationally. Results show that depending on the concentration of suppressants, the addition of MCA and MPP promotes or suppresses Al dust explosion. For high additive concentration, large agglomerated residues are generated, and condensed phase residues may contain Al particles, MCA or MPP. The chemical composition of condensed phase residues of Al/MCA mixture explosion is mainly Al2O3 and the high boiling products of MEL decomposition (mainly C‒ and N‒containing species). The explosion residues of Al/MPP mixture are composed of Al2O3, high boiling products of MEL decomposition and condensed phosphates. To understand the reasons for pressure enhancement and explosion suppression, a kinetic model considering both gas and surface chemistry of Al particles combustion is developed. The simulations indicate that the high pressure rise is caused by the extra heat released from the exothermic reactions of suppressants and the increase of gas phase products. MPP and MCA can suppress surface reaction by decreasing Al (L) site fraction. Additionally, the vaporization rate of Al particles and the diffusion rate of oxidizers close to the droplet surface are reduced by MPP and MCA addition.
Keywords: Aluminum dust explosions Explosion suppression Detailed suppression mechanism Surface kinetics
⁎
Corresponding author. E-mail address: [email protected] (W. Gao).
https://doi.org/10.1016/j.jhazmat.2019.121648 Received 8 October 2019; Received in revised form 4 November 2019; Accepted 8 November 2019 0304-3894/ © 2019 Elsevier B.V. All rights reserved.
Please cite this article as: Haipeng Jiang, Mingshu Bi and Wei Gao, Journal of Hazardous Materials, https://doi.org/10.1016/j.jhazmat.2019.121648
Journal of Hazardous Materials xxx (xxxx) xxxx
H. Jiang, et al.
1. Introduction
previous study (Jiang et al., 2018). Al dust and suppressants were thoroughly mixed before being stored in a dust container. Then, the 20 L spherical chamber was vacuumed to 0.06 MPa (absolute pressure) and the dispersed air pressure of dust container was set to 2 MPa (gauge pressure). Al/suppressants mixture was dispersed into 20 L chamber by a fast-acting valve. After injection of Al/suppressants mixture through the rebound nozzle, the suspended dust was ignited by a 2 kJ chemically igniter, following a 60 ms delay to create a quasi-uniform dust cloud and allow the initial turbulence to decay. Research indicated that the dust cloud in 20 L chamber was quasi-uniform which formed a concentration gradient, due to the extremely complex multi-phase process (Kalejaiye et al., 2010; Vizcaya et al., 2018), even though the experiments were performed according to the international standard ASTM (2010) and ASTM (2014). In order to obtain reliable results, 5–6 replicate experiments were performed and the values of the explosion parameters were averaged. Melamine cyanurate (MCA, C6H9N9O3) and melamine polyphosphate (MPP, C3H9N6O4P) were selected as suppressants. MCA and MPP dust were supplied from Sichuan Institute of Fine Chemical Industry Research and Design, China. It can be seen from Fig. 2 that the Sauter mean diameters D[3,2] of MCA and MPP were 2 μm and 3 μm, respectively. The thermogravimetric curves of MCA and MPP were shown in Fig. 3. MCA decomposed at approximately 304–428 °C. According to previous studies, HNCO, NH3 and condensed phase products could be generated during the decomposition of MCA (Costa and Camino, 1988; Wang et al., 2009). MCA → C3H3N3O3 + MEL → HNCO + NH3 + condensed phase products. MPP decomposed at approximately 349–900 °C. The decomposition products of MPP were composed of gas‒phase products (H3PO4, HPO3, NH3 and H2O), high boiling product and condensed phosphates (Wang and Yang, 2011). MPP → H3PO4 + NH3 + condensed phase products. Thermal stability and heat absorption capacity of MCA and MPP were evaluated using differential scanning calorimetry (DSC). The quantity of absorbed heat could be measured through integration of the DSC curves. The integration results showed that the amount of absorbed heat during MCA and MPP decomposition was 13.89 kJ/g and 15.453 kJ/g, respectively. Hence, MPP powder had a better thermal stability and could absorb relatively more heat than MCA powder.
The vigorous development of the energy and machinery manufacturing industry has driven the prosperity of metal surface treatment industries, such as finishing and polishing of metal products (mainly aluminum, magnesium and alloy materials). Suspended aluminum dust with small particle size will be dispersed in processing plant. When the concentration of suspended dust in air reaches the explosion limit range, an explosion may occur once the ignition source (such as electric sparks, mechanical sparks, etc.) is encountered. Among metals, aluminum is commonly used in the industry. In recent years, a considerable number of severe aluminum dust explosion accidents have occurred. Explosion suppression is an effective means to mitigate the consequence of aluminum dust explosion. In order to provide a safe manufacturing and processing environment, it is very significant and necessary to clearly and scientifically understand the mechanism of explosion and explosion suppression (Bu et al., 2020; Huang et al., 2019). Many efforts have been made to use different suppressants to mitigate and suppress aluminum dust explosion in the last decades. Dastidar and Amyotte (2002) studied the minimum suppression concentration of solid inhibitor for Al dust in the standard 20 L spherical chamber. Taveau et al. (Taveau et al., 2015) investigated the effect of sodium bicarbonate and potassium bicarbonate on the maximum explosion pressure and deflagration index of aluminum dust in a 4.4 m3 vessels. Chen et al. (Chen et al., 2017) and Jiang et al. (Jiang et al., 2018) studied the suppression capacity of suppressants with different particle size distribution for aluminum dust. However, only a few researchers have studied the detailed suppression mechanism of aluminum dust explosion. As a high reactive metal fuel with large energy density, aluminum dust exerts a higher explosion severity in comparison with organic fuel. In general, a large number of suppressants are required to suppress aluminum dust explosion, which poses significant challenges to traditional explosion suppression methods. The performance of suppressants is a key factor for explosion suppression of aluminum dust. Therefore, the development of high performance suppressants suitable for suppression of aluminum dust explosion has been a significant research topic. N containing flame retardants (such as Melamine cyanurate, MCA) and P‒N containing flame retardants (such as melamine polyphosphate, MPP), as eco‒friendly halogen-free flame retardants, have received considerable attention (Tao and Li, 2018; Yang et al., 2014). Unlike traditional suppressant, which acts through thermal effects and chemical interactions taking place in the gas phase, these flame retardants can play a key role of suppression effect on fuel surface oxidation (Dounia et al., 2018; Jahromi et al., 2003). For MPP addition, a char layer will be produced to impede the solid fuel surface oxidation during the combustion process (Jahromi et al., 2003). MCA operates in a melt-away mode, leading to increased dripping and flame retardancy (Gijsman et al., 2002). Therefore, compared to conventional suppressant, MCA and MPP may show a better performance in suppressing the aluminium dust explosions. In this study, a series of experiments are performed using a standard 20 L spherical vessel to determine the effects of MCA and MPP on the explosion pressure of aluminum dust. The suppression mechanism considering both gas and surface chemistry is first developed. The effect of MCA and MPP on gas and surface reaction of aluminum particles burning is further discussed.
3. Numerical methods 3.1. Initial conditions This work deals with nascent particles by ignoring the presence of the alumina shell. For the calculations of flame temperature and explosion pressure, only detailed reaction kinetics of gas-phase chemical suppression were accounted for. For the case of surface reaction suppression of aluminum particles, both gas and surface reactions of aluminum combustion were considered. Initial pressure and initial temperature were 1 atm and 2300 K, respectively. The mole ratio of N2 and O2 for air was 3.76. To simplify, it was assumed that suppressant particles had been completely decomposed in the flame reaction zone. In the reaction zone, MCA had been decomposed to generate HNCO, NH3 and condensed phase products. MPP had been decomposed to form H3PO4, NH3 and condensed phase products. The condensed phase products of MCA and MPP were not considered. The inerting ratio (α) that used in this part was referred to the molar ratio of suppressant to fuel. To reveal the suppression mechanism of MPP and MCA for aluminum dust explosion, a detailed investigation of Al/suppressants mixture was performed at constant volume.
2. Experimental Aluminum dust samples (5 μm and 30 μm), the experimental apparatus and methods were similar to our previous study (Jiang et al., 2018). Experiments were performed in a standard 20 L spherical test apparatus (Fig. 1). Experimental procedures were performed according to ASTM (2010) and ASTM (2014), where were also the same with our
3.2. Kinetic mechanism The modeling studies for the undoped and doped Al/air flames were performed at constant volume using SENKIN code and SKSAMPLE code 2
Journal of Hazardous Materials xxx (xxxx) xxxx
H. Jiang, et al.
Fig. 1. 20 L spherical chamber.
Fig. 3. TG‒DSC curves of MCA and MPP.
The alumina condensation is described by reversible reactions, Al2O3c ⇔ Al2O3 (L). Surface chemistry considers species that are adsorbed on the open sites of the particle surface. In this study, species were adsorbed on the open sites Al(L) of the aluminum surface. Surface site species and bulk-phase species were noted using suffixes (S) and (B), respectively. The illustration of an adsorption reaction was shown in Fig. 4. This meant O absorbed on aluminum surface and occupied a free site Al(L) with the generation of O(S). In the reverse reaction, the adsorbate O(S) left the aluminum surface and the bulk aluminum Al(B) became an Al(L). Since the adsorption capacity of the surface of aluminum particles was limited, the number of “site” was also limited, and the number of open sites Al(L) was proportional to Al surface area, so the density unit of open sites Al(L) should be “1/m2”. The total density of open sites Al (L) was 4.42 × 10-9 mol/m2. For bulk species, surface concentration was assumed to be 1. For some surface reaction mechanisms, the surface reaction rate constant can be given by a sticking coefficient
Fig. 2. SEM and particle size distributions of MCA and MPP.
of CHEMKIN software. A detailed description of the gas‒phase mechanism of Al/air flames doped with P‒containing species and NH3 was given by Jiang et al. (Jiang et al., 2019a). For the case of surface reaction suppression of aluminum particles, the attention of modeling study was paid to the surface and gas/surface interactions at the vicinity of an aluminum particle. The gas‒phase mechanism was drawn from previous work on reactions of aluminum (Catoire et al., 2003), NH3 (Li et al., 2013; Konnov, 2009), H3PO4 (Jayaweera et al., 2005; Korobeinichev et al., 2007; Twarowski, 1996) and HNCO (Smith et al., 2019), involving 86 species and 460 elementary reactions. Surface chemistry of aluminum particles combustion was taken from the work of Glorian et al. (Glorian et al., 2016). In order to solve the problem of flame temperature overestimation, a more adequate condensation model is proposed based on the work of Bojko et al. (Bojko et al., 2014). 3
Journal of Hazardous Materials xxx (xxxx) xxxx
H. Jiang, et al.
aluminum particles and tc have the same tendency, as the concentration of the suppressants powder varies. For the case of MPP addition, NH3 and H2O will be produced during the MPP decomposition. NH3 can react with radicals and release heat. When the oxidizer is sufficient, the reaction process of Al combustion may potentially be enhanced by extra heat released from NH3 reaction, which causes a decrease in tb and tc. When the concentration of suppressant is sufficient, the suppression efficiency increases with suppressant concentration increases, resulting in a gradual increase in tc. For MCA addition, HNCO will form during the decomposition of MCA under high temperature atmosphere. MCA → MEL + C3H3N3O3; C3H3N3O3 → 3 (HNCO) (A = 6.30 E7, 1/s; E = 2.84 E4 cal/mol). HNCO can react with O atoms to generate CNO and NH. HNCO + O = CO2 + NH (R430) HNCO + O = OH + NCO (R432) NH and CNO could also consume O atom and O2, such as: NH + O = N + OH (R182) O + NCO = NO + CO (R451) O2 + NCO = CO2 + NO (R455)
Fig. 4. Illustration of an adsorption reaction.
For MCA addition, NH3 will also be generated during the MCA decomposition. However, it can be seen that the generation of HNCO results in the consumption of large numbers of O2 and O atom. tb of 30 μm Al particles is a function of effective oxidizer concentration Xeff (tb = (cdp1.8)/(Xeff p0.1 T00.2) ) and decreases as Xeff decreases (Beckstead, 2005). Clearly, with MCA addition increasing, there is a decreasing Xeff and an increasing tb. Therefore, tc of 30 μm aluminum dust increases considerably with MCA concentration increases. Previous studies demonstrate that the combustion regimes of small and large micro Al particles are completely different (Bazyn et al., 2007). As shown in Fig. 6, tc shows a relatively slight increase as either MCA or MPP is added. For O2, tb of 5 μm Al particles is inversely proportional to the pressure and the oxygen concentration. However, for H2O, tb increase with increasing pressure (Lynch et al., 2009). By increasing MPP concentration, the oxygen concentration decreases, and H2O concentration increases. Hence, tb increases with the pressure and H2O concentration. It is revealed that tb of 5 μm Al dust increases with increasing MCA and MPP concentration. Figs. 7 and 8 demonstrate the maximum explosion pressure Pmax of 30 μm and 5 μm Al dust doped with MCA or MPP. Jiang et al. (2018) explains why the concentration of 5 μm aluminum powder varies from 100 to 300 g/m3. As indicated in the figures, Pmax first increases rapidly
(STICK) in addition to an Arrhenius law similar to gas-phase reactions. This coefficient was defined as a probability that a certain reaction happened when a collision between a given gas-species occurred with the surface. When the sticking coefficient was close to 1, it was usually appropriate to apply the Motz-Wise correction (MWON). Users could turn the corresponding option on by including STICK and MWON in the REACTIONS line of their surface kinetics input file. 4. Experimental results 4.1. Explosion pressure Combustion time (tc) of dust cloud refers to the time between ignition and the Pmax. Burning time (tb) refers to the duration of burning of a single dust particle. As shown in Fig. 5a, for 30 μm aluminum dust, an increase of MCA powder concentration could lead to a gradual increase of combustion time (tc). In contrast, tc firstly decreases rapidly for a low concentration of MPP powder, but then eventually increases considerably as MPP powder concentration increases (Fig. 5b). The dust cloud can be considered as a concourse of single particle dispersed in a gaseous oxidizer. Therefore, we assume that the burning time (tb) of
Fig. 5. Pressure profiles of 30 μm Al dust doped with suppressant powder. 4
Journal of Hazardous Materials xxx (xxxx) xxxx
H. Jiang, et al.
Fig. 6. Pressure profiles of 5 μm Al dust doped with suppressant powder.
sufficient, large agglomerated residues are formed, and suppressant particles appear in explosion residues. It is worth noting that the surface of explosion residues forms a porous structure, as shown in Fig. 10d and e. Since the molecules of both MCA and MPP contain a blowing agent (melamine), a large number of gases products are formed by flame retardant decomposition during the explosion process (Wang and Yang, 2011). Based on the SEM results, it can be inferred that the condensed phase products of MCA and MPP dust explosion can cover aluminum particles surface and hinder its oxidation reaction. Suppressants, aluminum particles, and explosion products agglomerate together to form large porous residues. To reveal the chemical composition and different bonds structure of the mixture explosion residues surface, XPS is performed. Fig. 11 shows Al2p, O1s, C1s and N1s XPS spectra of explosion residues of Al/MCA mixture. In Fig. 11a and b, the peaks at 74.5–75.9 eV are attributed to Al3+ in Al2O3, and the peak at 72.9 eV is active Al0 (Moulder et al., 1995). It can be inferred that aluminum particles contain explosion residues, when the concentration of MCA is sufficient. From Fig. 11c and 11d, the O1s spectrum is divided into two peaks, which is corresponding to the binding energies of O in Al2O3. In Fig. 11e and f, the peaks at 284.9 eV–289.6 eV can be assigned to CeC bond, CeH bond, CeN bond, C]O bond, and C]N bond (Dante et al., 2015). As shown in Fig. 11g, the peaks at 398.8 eV and 400.0 eV are attributed to C]N bond and eNH2, respectively (Wang and Yang, 2011). These Ce and
for a low concentration of MCA and MPP, but then drops significantly as suppressant concentration increases. These results demonstrate that both MCA and MPP addition can enhance or suppress the explosion severity of Al dust. Since the lack of suppressants can increase the risk of accidents, it is extremely important to measure the minimum inerting concentration (MIC) of suppressant. Fig. 9 illustrates the inerting curves as a function of suppressant dust concentration. MIC data of SBC and ABC are adapted from Jiang et al. (2018, 2019b). The concentration of suppressant powder to the right of each inerting curves is sufficient to prevent an explosion. At the concentration to the left of each inerting curves, an explosion occurs. Meanwhile, the explosion is enhanced at the concentration to the left of each enhancement curves (Fig. 9a and b). As shown in Fig. 9c, MCA powder exerts the strongest suppression effect on 30 μm aluminum dust explosion, followed by MPP powder, ABC powder (NH4H2PO4 > 90 %) and sodium bicarbonate (SBC), respectively. For 5 μm aluminum particles, the MIC of ABC powder is the lowest. MCA and MPP deliver almost equivalent suppression performance. By comparison, a large number of SBC particles are required to prevent 5 μm aluminum dust explosion. 4.2. Condensed phase explosion residues As shown in Fig. 10, when the deployment of suppressants is
Fig. 7. The maximum explosion pressure of 30 μm Al dust for various suppressant concentration. 5
Journal of Hazardous Materials xxx (xxxx) xxxx
H. Jiang, et al.
Fig. 8. The maximum explosion pressure of 5 μm Al dust for various suppressant concentration.
spectrum at 530.6 eV, 531.6 eV and 531.2 eV are attributed to the binding energies of O in Al2O3. The peaks at 532.9 eV, 532.4 eV, and 533.3 eV can be assigned to the O in PeO bond, CeO bond, P]O bond and C]O bond. From Fig. 12g, the peaks at 133.9 eV are attributed to
Ne containing species are condensed phase explosion products corresponding to the high boiling products of MEL decomposition. Fig. 12 shows Al2p, O1s, C1s, P2p and N1s XPS spectra of Al/MPP mixture explosion residue. From Fig. 12c and d, the peaks of O1s
Fig. 9. Inerting curves of suppressants for Al dust explosion (adapted from Jiang et al., 2018 and Jiang et al., 2019b). 6
Journal of Hazardous Materials xxx (xxxx) xxxx
H. Jiang, et al.
Fig. 10. Images of explosion residues (Al: 300 g/m3).
5. Enhancement and suppression mechanism
the binding energies of P in pyrophosphate, metaphosphate or phosphate (Moulder et al., 1995). These condensed phosphates of P‒containing species are decomposed from MPP. From Figs. 11g and 12 h, the C]N bond and eNH2 are present in all condensed residue, for high additive concentration. Hence, it is probably the species corresponding to MEL or its decomposition products. The XPS results prove that Al/MCA mixture explosion residues are compose of Al2O3 and high boiling products (decomposed from MCA). Al/MPP mixture explosion residues are mainly Al2O3, high boiling products (decomposed from MEL) and condensed phosphates. Al particles, MCA or MPP may contain explosion residues. Results for condensed phase residues of 5 μm Al/suppressant mixture are similar, so they are not given and plotted.
5.1. Combustion enhancement mechanism Fig. 13 shows the calculated maximum flame temperature of gaseous‒Al doped with MCA and MPP. As can be seen that calculated maximum flame temperature decreases monotonously with suppressants concentration increases (φ refers to the equivalence ratio). Apparently, instantaneous heat release decreases with the addition of Suppressants. Fig. 14 demonstrates the pressure evolution at various inerting ratios. As can be seen that the maximum pressure firstly increases and then decreases with α increase. The trend of the calculated results (Fig. 14) is consistent with those of experiments (Fig. 6). NH3 will generate during the decomposition of MPP and MCA. When the
Fig. 11. Al2p, O1s, C1s and N1s XPS spectra of Al/MCA mixture explosion residues. 7
Journal of Hazardous Materials xxx (xxxx) xxxx
H. Jiang, et al.
Fig. 12. Al2p, O1s, P2p, C1s and N1s XPS spectra of Al/MPP mixture explosion residues.
oxidant is sufficient, the extra heat released by the combustion of NH3 leads to an increase in total heat release and heat release time, but results in a reduction in instantaneous heat release. When the oxidant is insufficient, instantaneous heat release and total heat release decrease with the addition of MCA and MPP. Moreover, in a constant volume container, as the instantaneous pressure rises and the flame temperature decreases, it means an increase in the amount of gas-phase products. It is indeed found in the experiments that large numbers of gasphase products are produced when the explosion is enhanced. To investigate the effect of chemical kinetics on flame temperature and explosion pressure, sensitivity coefficients of flame temperature and explosion pressures are determined for MCA- and MPP-doped aluminum flame as shown in Figs. 15 and 16, respectively. It can be seen from Figs. 15a and 16 a that HNCO + O = NCO + OH (R432), HCN + O = NCO + H (R460) and NH3 + O = NH2 + OH (R163) could reduce the flame temperature and the explosion pressure by consuming O atom. HNCO + M = NH + CO + M (R437), HNCO + H = NH2 + CO (R433) and aluminum oxidation reaction could increase explosion pressure. From Figs. 15b and 16 b, the most sensitive reactions of suppression of flame temperature and explosion pressure are reactions of the P ‒containing components: H3PO4 = HPO3 + H2O (R281) and HPO3+M = PO2+OH + M (R286). It is worth noting that R163 has a positive sensitivity to explosion pressure at α = 0.2 but has a negative sensitivity at α = 0.5 and 1.0. Therefore, NH3 and HNCO can
Fig. 13. Calculated maximum flame temperature of gaseous‒ Al doped with MCA and MPP.
Fig. 14. Explosion pressure of gaseous‒ Al doped with (a) MCA and (b) MPP. 8
Journal of Hazardous Materials xxx (xxxx) xxxx
H. Jiang, et al.
Fig. 15. Sensitivity analysis of flame temperature (ϕ Al = 1.0).
change in the order of suppression effectiveness of different suppressants for 5 μm and 30 μm Al dust explosion is attributed to the difference in the suppression effect of gas-phase reaction and surface reaction. Additionally, the rate of aluminum vaporization can be reduced by these suppressants addition, since the vaporization rate is limited by the number of Al(L). It can be also seen from Fig. 18 that site fractions of AlO(S) and O(S) decrease with the increase of inerting ratio. These results indicate that the diffusion rate of oxidizers close to the droplet surface is reduced by suppressant addition (Glorian et al., 2016).
enhance or suppress explosion pressure. On the basis of these results we concluded that when the concentration of MCA and MPP is low, the oxidizer is sufficient, the extra heat released from the exothermic reactions of suppressant and the increase of gas phase products could cause the pressure to rise. When MCA and MPP concentration was high, the pressure could be reduced by the explosion suppression. 5.2. Suppression mechanism AlO and O are the key radicals in gaseous-Al/air flame. Comparison of AlO and O concentration doped with different suppressants concentrations is used to determine the suppression capacity of MPP and MPP for gas-phase reaction of Al combustion. As shown in Fig.17, the concentration of AlO and O gradully decreases with inerting ratio increases. MCA is found to be more effective than MPP in scavenging flame radicals and suppressing gas-phase reaction of Al combustion. The combustion mode of aluminum particles depends on the composition of the oxidizer and particle size (Sundaram et al., 2016). Surface reactions become significant for small particles (< 10 μm) or large particles burning in H2O and CO2 atmosphere under high pressure (> 5 atm). For Al particles with fixed surface area, the rate of surface reaction depends on the number of active sites. Suppressants can interrupt surface reactions by reducing the number of active sites. Hence, the effect of suppressants on Al(L) site fraction is studied to evaluate the suppression capacity of MCA and MPP for surface reactions. As shown in Fig. 18, Al(L) site fraction decreases with inerting ratio of MPP increases, but slightly decreases with inerting ratio of MCA increases, indicating that MPP has a better suppression effect on the surface reaction of Al combustion. Simulation results reveal that the
6. Conclusions Experimental and numerical investigation are conducted to reveal the suppression mechanism of MCA and MPP for Al dust explosion. The conclusions can be summarized as follows. (1) For low additive concentration, the addition of both MCA and MPP can enhance or suppress Al dust explosion. The addition of MPP decreases the combustion time due to the increase of effective oxidizer concentration. With increasing MCA addition, there is an increase of O2 and O atom consumption, which results in the reduction of the effective oxidizer concentration and the rise of combustion time of aluminum dust explosion. (2) The explosion residues of Al/MCA mixture are Al2O3 and C‒ and N‒ containing species (the high boiling products of MEL decomposition). The explosion residues of Al/MPP mixture consist of Al2O3, high boiling products (decomposed from MEL) and condensed phosphates. When the suppressant is sufficient, a porous
Fig. 16. Sensitivity analysis of pressure (ϕ Al = 1.0). 9
Journal of Hazardous Materials xxx (xxxx) xxxx
H. Jiang, et al.
Fig. 17. Temporal evolution of (a) AlO and (b) O atom in Al/MCA/air (red line) flame and Al/MPP/air (blue line) flame at various inerting ratios. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article).
Appendix A. Supplementary data Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.jhazmat.2019.121648. References ASTM, 2010. E1226-12a. Standard Test Method for Explosibility of Dust Clouds. ASTM, 2014. E1515. Standard Test Method for Minimum Explosible Concentration of Combustible Dusts. Bazyn, T., Krier, H., Glumac, N., 2007. Evidence for the transition from the diffusion-limit in aluminum particle combustion. Proc. Combust. Inst. 31, 2021–2028. Beckstead, M.W., 2005. Correlating aluminum burning times. Combust Explo Shock+ 41, 533–546. Bojko, B.T., DesJardin, P.E., Washburn, E.B., 2014. On modeling the diffusion to kinetically controlled burning limits of micron-sized aluminum particles. Combust. Flame 161, 3211–3221. Bu, Y.J., Li, C., Amyotte, P.R., Yuan, W.B., Yuan, C.M., Li, G., 2020. Moderation of Al dust explosions by micro- and nano-sized Al2O3 powder. J. Hazard. Mater. 381, 120968. Catoire, L., Legendre, J.F., Giraud, M., 2003. Kinetic model for aluminum‒sensitized ram accelerator combustion. J. Propul. Power 19, 196–202. Chen, X.F., Zhang, H.M., Chen, X., Liu, X.Y., Niu, Y., Zhang, Y., Yuan, B.H., 2017. Effect of dust explosion suppression by sodium bicarbonate with different granulometric distribution. J. Loss Prevention Process Indust. 49, 905–911. Costa, L., Camino, G., 1988. Thermal-behavior of melamine. J. Therm. Anal. 34, 423–429. Dante, R.C., Sanchez-Arevalo, F.M., Chamorro-Posada, P., Vazquez-Cabo, J., Huerta, L., Lartundo-Rojas, L., Santoyo-Salazar, J., Solorza-Feria, O., 2015. Supramolecular intermediates in the synthesis of polymeric carbon nitride from melamine cyanurate. J. Solid State Chem. 226, 170–178. Dastidar, A., Amyotte, P.R., 2002. Determination of minimum inerting concentrations for combustible dusts in a laboratory-scale chamber. Process Saf. Environ. Prot. 80, 289–299. Dounia, O., Vermorel, O., Poinsot, T., 2018. Theoretical analysis and simulation of methane/air flame inhibition by sodium bicarbonate particles. Combust. Flame 193, 313–326. Gijsman, P., Steenbakkers, R., Furst, C., Kersjes, J., 2002. Differences in the flame retardant mechanism of melamine cyanurate in polyamide 6 and polyamide 66. Polym. Degrad. Stab. 78, 219–224. Glorian, J., Gallier, S., Catoire, L., 2016. On the role of heterogeneous reactions in aluminum combustion. Combust. Flame 168, 378–392. Jahromi, S., Gabrielse, W., Braam, A., 2003. Effect of melamine polyphosphate on thermal degradation of polyamides: a combined X-ray diffraction and solid-state NMR study. Polymer 44, 25–37. Jayaweera, T.M., Melius, C.F., Pitz, W.J., Westbrook, C.K., Korobeinichev, O.P., Shvartsberg, V.M., Shmakov, A.G., Rybitskaya, I.V., Curran, H.J., 2005. Flame inhibition by phosphorus‒containing compounds over a range of equivalence ratios. Combust. Flame 140, 103–115. Jiang, H., Bi, M., Gao, W., Gan, B., Zhang, D., Zhang, Q., 2018. Inhibition of aluminum dust explosion by NaHCO3 with different particle size distributions. J. Hazard. Mater. 344, 902–912. Jiang, H., Bi, M., Li, B., Ma, D., Gao, W., 2019a. Flame inhibition of aluminum dust explosion by NaHCO3 and NH4H2PO4. Combust. Flame 200, 97–114. Jiang, H., Bi, M., Li, B., Zhang, D., Gao, W., 2019b. Inhibition evaluation of ABC powder in aluminum dust explosion. J. Hazard. Mater. 361, 273–282. Li, B., He, Y., Li, Z.S., Konnov, A.A., 2013. Measurements of NO concentration in NH3‒doped CH4 + air flames using saturated laser‒induced fluorescence and probe sampling. Combust. Flame 160, 40–46. Lynch, P., Krier, H., Glumac, N., 2009. A correlation for burn time of aluminum particles in the transition regime. Proc. Combust. Inst. 32, 1887–1893.
Fig. 18. Site fractions at aluminum surface as a function of inerting ratio in air.
agglomerated residue is formed, and aluminum particles, MCA or MPP may be contained in condensed phase residues. (3) Calculations show that flame temperature decreases monotonously with suppressants concentration increases. The formation of large quantities of gas phase products and the extra heat released from the exothermic reactions of suppressant leads to an increase in pressure. MPP is more effective than MCA for surface reaction suppression, but MCA is more effective for gas-phase reaction suppression.
Declaration of Competing Interest We declare no financial interests.
Acknowledgments The authors gratefully acknowledge the financial support from the National Natural Science Foundation of China (No. 51874066 and No. 51674059), Key Laboratory of Building Fire Protection Engineering and Technology of MPS (KFKT2016ZD01), the Fundamental Research Funds for the Central Universities (DUT16RC(4)04), (DUT18RC (3) 038). The authors would like to thank Dr. Julien Glorian for providing the chemical kinetic models of gas phase and surface reactions.
10
Journal of Hazardous Materials xxx (xxxx) xxxx
H. Jiang, et al.
dust deflagrations. J. Loss Prevention Process Indust. 36, 246–253. Tao, W., Li, J., 2018. Melamine cyanurate tailored by base and its multi effects on flame retardancy of polyamide 6. Appl. Surf. Sci. 456, 751–762. Twarowski, A., 1996. The temperature dependence of H + OH recombination in phosphorus oxide containing post‒combustion gases. Combust. Flame 105, 407–413. Vizcaya, D., Pinilla, A., Amin, M., Ratkovich, N., Munoz, F., Murillo, C., Bardin-Monnier, N., Dufaud, O., 2018. CFD as an approach to understand flammable dust 20 L standard test: effect of the ignition time on the fluid flow. AIChE J. 64, 42–54. Wang, G.J., Yang, J.Y., 2011. Thermal degradation study of fire resistive coating containing melamine polyphosphate and dipentaerythritol. Prog. Org. Coat. 72, 605–611. Wang, Z.Z., Lv, P., Hu, Y., Hu, K.L., 2009. Thermal degradation study of intumescent flame retardants by TG and FTIR: melamine phosphate and its mixture with pentaerythritol. J. Anal. Appl. Pyrolysis 86, 207–214. Yang, H.Y., Song, L., Tai, Q.L., Wang, X., Yu, B., Yuan, Y., Hu, Y., Yuen, R.K.K., 2014. Comparative study on the flame retarded efficiency of melamine phosphate, melamine phosphite and melamine hypophosphite on poly(butylene succinate) composites. Polym. Degrad. Stab. 105, 248–256.
Huang, C.Y., Chen, X.F., Yuan, B.H., Zhang, H.M., Dai, H.M., He, S., Zhang, Y., Niu, Y., Shen, S.F., 2019. Suppression of wood dust explosion by ultrafine magnesium hydroxide. J. Hazard. Mater. 378, 120723. Kalejaiye, O., Amyotte, P.R., Pegg, M.J., Cashdollar, K.L., 2010. Effectiveness of dust dispersion in the 20-L Siwek chamber. J. Loss Prev. Process Ind. 23, 46–59. Konnov, A.A., 2009. Implementation of the NCN pathway of prompt‒NO formation in the detailed reaction mechanism. Combust. Flame 156, 2093–2105. Korobeinichev, O.P., Shvartsberg, V.M., Shmakov, A.G., Knyazkov, D.A., Rybitskaya, I.V., 2007. Inhibition of atmospheric lean and rich CH4/O2/Ar flames by phosphorus‒containing compound. Proc. Combust. Inst. 31, 2741–2748. Moulder, J.F., Stickle, W.F., Sobol, P.E., Bomben, K.D., 1995. Handbook of X-Ray Photoelectron Spectroscopy. Physical Electronics, Minnesota. Smith, G.P., Golden, D.M., Frenklach, M., Moriarty, N.W., Eiteneer, B., Goldenberg, M., Bowman, C.T., Hanson, R.K., Song, S., Gardiner, W.C., Lissianski, V. V., Qin, Z., 2019 GRI-Mech 3.0. n.d. URL http://combustion.berkeley.edu/gri-mech/version30/ text30.html (Accessed 12 February 2018). Sundaram, D.S., Puri, P., Yang, V., 2016. A general theory of ignition and combustion of nano‒ and micron-sized aluminum particles. Combust. Flame 169, 94–109. Taveau, J., Vingerhoets, J., Snoeys, J., Going, J., Farrell, T., 2015. Suppression of metal
11
Contents lists available at ScienceDirect
Journal of Hazardous Materials journal homepage: www.elsevier.com/locate/jhazmat
Suppression mechanism of Al dust explosion by melamine polyphosphate and melamine cyanurate Haipeng Jiang, Mingshu Bi, Wei Gao* School of Chemical Engineering, Dalian University of Technology, Dalian 116024, China
G R A P H I C A L A B S T R A C T
A R T I C LE I N FO
A B S T R A C T
Editor: G. Lybertos
The suppression mechanism of melamine polyphosphate (MPP) and melamine cyanurate (MCA) for Al dust explosions is investigated experimentally and computationally. Results show that depending on the concentration of suppressants, the addition of MCA and MPP promotes or suppresses Al dust explosion. For high additive concentration, large agglomerated residues are generated, and condensed phase residues may contain Al particles, MCA or MPP. The chemical composition of condensed phase residues of Al/MCA mixture explosion is mainly Al2O3 and the high boiling products of MEL decomposition (mainly C‒ and N‒containing species). The explosion residues of Al/MPP mixture are composed of Al2O3, high boiling products of MEL decomposition and condensed phosphates. To understand the reasons for pressure enhancement and explosion suppression, a kinetic model considering both gas and surface chemistry of Al particles combustion is developed. The simulations indicate that the high pressure rise is caused by the extra heat released from the exothermic reactions of suppressants and the increase of gas phase products. MPP and MCA can suppress surface reaction by decreasing Al (L) site fraction. Additionally, the vaporization rate of Al particles and the diffusion rate of oxidizers close to the droplet surface are reduced by MPP and MCA addition.
Keywords: Aluminum dust explosions Explosion suppression Detailed suppression mechanism Surface kinetics
⁎
Corresponding author. E-mail address: [email protected] (W. Gao).
https://doi.org/10.1016/j.jhazmat.2019.121648 Received 8 October 2019; Received in revised form 4 November 2019; Accepted 8 November 2019 0304-3894/ © 2019 Elsevier B.V. All rights reserved.
Please cite this article as: Haipeng Jiang, Mingshu Bi and Wei Gao, Journal of Hazardous Materials, https://doi.org/10.1016/j.jhazmat.2019.121648
Journal of Hazardous Materials xxx (xxxx) xxxx
H. Jiang, et al.
1. Introduction
previous study (Jiang et al., 2018). Al dust and suppressants were thoroughly mixed before being stored in a dust container. Then, the 20 L spherical chamber was vacuumed to 0.06 MPa (absolute pressure) and the dispersed air pressure of dust container was set to 2 MPa (gauge pressure). Al/suppressants mixture was dispersed into 20 L chamber by a fast-acting valve. After injection of Al/suppressants mixture through the rebound nozzle, the suspended dust was ignited by a 2 kJ chemically igniter, following a 60 ms delay to create a quasi-uniform dust cloud and allow the initial turbulence to decay. Research indicated that the dust cloud in 20 L chamber was quasi-uniform which formed a concentration gradient, due to the extremely complex multi-phase process (Kalejaiye et al., 2010; Vizcaya et al., 2018), even though the experiments were performed according to the international standard ASTM (2010) and ASTM (2014). In order to obtain reliable results, 5–6 replicate experiments were performed and the values of the explosion parameters were averaged. Melamine cyanurate (MCA, C6H9N9O3) and melamine polyphosphate (MPP, C3H9N6O4P) were selected as suppressants. MCA and MPP dust were supplied from Sichuan Institute of Fine Chemical Industry Research and Design, China. It can be seen from Fig. 2 that the Sauter mean diameters D[3,2] of MCA and MPP were 2 μm and 3 μm, respectively. The thermogravimetric curves of MCA and MPP were shown in Fig. 3. MCA decomposed at approximately 304–428 °C. According to previous studies, HNCO, NH3 and condensed phase products could be generated during the decomposition of MCA (Costa and Camino, 1988; Wang et al., 2009). MCA → C3H3N3O3 + MEL → HNCO + NH3 + condensed phase products. MPP decomposed at approximately 349–900 °C. The decomposition products of MPP were composed of gas‒phase products (H3PO4, HPO3, NH3 and H2O), high boiling product and condensed phosphates (Wang and Yang, 2011). MPP → H3PO4 + NH3 + condensed phase products. Thermal stability and heat absorption capacity of MCA and MPP were evaluated using differential scanning calorimetry (DSC). The quantity of absorbed heat could be measured through integration of the DSC curves. The integration results showed that the amount of absorbed heat during MCA and MPP decomposition was 13.89 kJ/g and 15.453 kJ/g, respectively. Hence, MPP powder had a better thermal stability and could absorb relatively more heat than MCA powder.
The vigorous development of the energy and machinery manufacturing industry has driven the prosperity of metal surface treatment industries, such as finishing and polishing of metal products (mainly aluminum, magnesium and alloy materials). Suspended aluminum dust with small particle size will be dispersed in processing plant. When the concentration of suspended dust in air reaches the explosion limit range, an explosion may occur once the ignition source (such as electric sparks, mechanical sparks, etc.) is encountered. Among metals, aluminum is commonly used in the industry. In recent years, a considerable number of severe aluminum dust explosion accidents have occurred. Explosion suppression is an effective means to mitigate the consequence of aluminum dust explosion. In order to provide a safe manufacturing and processing environment, it is very significant and necessary to clearly and scientifically understand the mechanism of explosion and explosion suppression (Bu et al., 2020; Huang et al., 2019). Many efforts have been made to use different suppressants to mitigate and suppress aluminum dust explosion in the last decades. Dastidar and Amyotte (2002) studied the minimum suppression concentration of solid inhibitor for Al dust in the standard 20 L spherical chamber. Taveau et al. (Taveau et al., 2015) investigated the effect of sodium bicarbonate and potassium bicarbonate on the maximum explosion pressure and deflagration index of aluminum dust in a 4.4 m3 vessels. Chen et al. (Chen et al., 2017) and Jiang et al. (Jiang et al., 2018) studied the suppression capacity of suppressants with different particle size distribution for aluminum dust. However, only a few researchers have studied the detailed suppression mechanism of aluminum dust explosion. As a high reactive metal fuel with large energy density, aluminum dust exerts a higher explosion severity in comparison with organic fuel. In general, a large number of suppressants are required to suppress aluminum dust explosion, which poses significant challenges to traditional explosion suppression methods. The performance of suppressants is a key factor for explosion suppression of aluminum dust. Therefore, the development of high performance suppressants suitable for suppression of aluminum dust explosion has been a significant research topic. N containing flame retardants (such as Melamine cyanurate, MCA) and P‒N containing flame retardants (such as melamine polyphosphate, MPP), as eco‒friendly halogen-free flame retardants, have received considerable attention (Tao and Li, 2018; Yang et al., 2014). Unlike traditional suppressant, which acts through thermal effects and chemical interactions taking place in the gas phase, these flame retardants can play a key role of suppression effect on fuel surface oxidation (Dounia et al., 2018; Jahromi et al., 2003). For MPP addition, a char layer will be produced to impede the solid fuel surface oxidation during the combustion process (Jahromi et al., 2003). MCA operates in a melt-away mode, leading to increased dripping and flame retardancy (Gijsman et al., 2002). Therefore, compared to conventional suppressant, MCA and MPP may show a better performance in suppressing the aluminium dust explosions. In this study, a series of experiments are performed using a standard 20 L spherical vessel to determine the effects of MCA and MPP on the explosion pressure of aluminum dust. The suppression mechanism considering both gas and surface chemistry is first developed. The effect of MCA and MPP on gas and surface reaction of aluminum particles burning is further discussed.
3. Numerical methods 3.1. Initial conditions This work deals with nascent particles by ignoring the presence of the alumina shell. For the calculations of flame temperature and explosion pressure, only detailed reaction kinetics of gas-phase chemical suppression were accounted for. For the case of surface reaction suppression of aluminum particles, both gas and surface reactions of aluminum combustion were considered. Initial pressure and initial temperature were 1 atm and 2300 K, respectively. The mole ratio of N2 and O2 for air was 3.76. To simplify, it was assumed that suppressant particles had been completely decomposed in the flame reaction zone. In the reaction zone, MCA had been decomposed to generate HNCO, NH3 and condensed phase products. MPP had been decomposed to form H3PO4, NH3 and condensed phase products. The condensed phase products of MCA and MPP were not considered. The inerting ratio (α) that used in this part was referred to the molar ratio of suppressant to fuel. To reveal the suppression mechanism of MPP and MCA for aluminum dust explosion, a detailed investigation of Al/suppressants mixture was performed at constant volume.
2. Experimental Aluminum dust samples (5 μm and 30 μm), the experimental apparatus and methods were similar to our previous study (Jiang et al., 2018). Experiments were performed in a standard 20 L spherical test apparatus (Fig. 1). Experimental procedures were performed according to ASTM (2010) and ASTM (2014), where were also the same with our
3.2. Kinetic mechanism The modeling studies for the undoped and doped Al/air flames were performed at constant volume using SENKIN code and SKSAMPLE code 2
Journal of Hazardous Materials xxx (xxxx) xxxx
H. Jiang, et al.
Fig. 1. 20 L spherical chamber.
Fig. 3. TG‒DSC curves of MCA and MPP.
The alumina condensation is described by reversible reactions, Al2O3c ⇔ Al2O3 (L). Surface chemistry considers species that are adsorbed on the open sites of the particle surface. In this study, species were adsorbed on the open sites Al(L) of the aluminum surface. Surface site species and bulk-phase species were noted using suffixes (S) and (B), respectively. The illustration of an adsorption reaction was shown in Fig. 4. This meant O absorbed on aluminum surface and occupied a free site Al(L) with the generation of O(S). In the reverse reaction, the adsorbate O(S) left the aluminum surface and the bulk aluminum Al(B) became an Al(L). Since the adsorption capacity of the surface of aluminum particles was limited, the number of “site” was also limited, and the number of open sites Al(L) was proportional to Al surface area, so the density unit of open sites Al(L) should be “1/m2”. The total density of open sites Al (L) was 4.42 × 10-9 mol/m2. For bulk species, surface concentration was assumed to be 1. For some surface reaction mechanisms, the surface reaction rate constant can be given by a sticking coefficient
Fig. 2. SEM and particle size distributions of MCA and MPP.
of CHEMKIN software. A detailed description of the gas‒phase mechanism of Al/air flames doped with P‒containing species and NH3 was given by Jiang et al. (Jiang et al., 2019a). For the case of surface reaction suppression of aluminum particles, the attention of modeling study was paid to the surface and gas/surface interactions at the vicinity of an aluminum particle. The gas‒phase mechanism was drawn from previous work on reactions of aluminum (Catoire et al., 2003), NH3 (Li et al., 2013; Konnov, 2009), H3PO4 (Jayaweera et al., 2005; Korobeinichev et al., 2007; Twarowski, 1996) and HNCO (Smith et al., 2019), involving 86 species and 460 elementary reactions. Surface chemistry of aluminum particles combustion was taken from the work of Glorian et al. (Glorian et al., 2016). In order to solve the problem of flame temperature overestimation, a more adequate condensation model is proposed based on the work of Bojko et al. (Bojko et al., 2014). 3
Journal of Hazardous Materials xxx (xxxx) xxxx
H. Jiang, et al.
aluminum particles and tc have the same tendency, as the concentration of the suppressants powder varies. For the case of MPP addition, NH3 and H2O will be produced during the MPP decomposition. NH3 can react with radicals and release heat. When the oxidizer is sufficient, the reaction process of Al combustion may potentially be enhanced by extra heat released from NH3 reaction, which causes a decrease in tb and tc. When the concentration of suppressant is sufficient, the suppression efficiency increases with suppressant concentration increases, resulting in a gradual increase in tc. For MCA addition, HNCO will form during the decomposition of MCA under high temperature atmosphere. MCA → MEL + C3H3N3O3; C3H3N3O3 → 3 (HNCO) (A = 6.30 E7, 1/s; E = 2.84 E4 cal/mol). HNCO can react with O atoms to generate CNO and NH. HNCO + O = CO2 + NH (R430) HNCO + O = OH + NCO (R432) NH and CNO could also consume O atom and O2, such as: NH + O = N + OH (R182) O + NCO = NO + CO (R451) O2 + NCO = CO2 + NO (R455)
Fig. 4. Illustration of an adsorption reaction.
For MCA addition, NH3 will also be generated during the MCA decomposition. However, it can be seen that the generation of HNCO results in the consumption of large numbers of O2 and O atom. tb of 30 μm Al particles is a function of effective oxidizer concentration Xeff (tb = (cdp1.8)/(Xeff p0.1 T00.2) ) and decreases as Xeff decreases (Beckstead, 2005). Clearly, with MCA addition increasing, there is a decreasing Xeff and an increasing tb. Therefore, tc of 30 μm aluminum dust increases considerably with MCA concentration increases. Previous studies demonstrate that the combustion regimes of small and large micro Al particles are completely different (Bazyn et al., 2007). As shown in Fig. 6, tc shows a relatively slight increase as either MCA or MPP is added. For O2, tb of 5 μm Al particles is inversely proportional to the pressure and the oxygen concentration. However, for H2O, tb increase with increasing pressure (Lynch et al., 2009). By increasing MPP concentration, the oxygen concentration decreases, and H2O concentration increases. Hence, tb increases with the pressure and H2O concentration. It is revealed that tb of 5 μm Al dust increases with increasing MCA and MPP concentration. Figs. 7 and 8 demonstrate the maximum explosion pressure Pmax of 30 μm and 5 μm Al dust doped with MCA or MPP. Jiang et al. (2018) explains why the concentration of 5 μm aluminum powder varies from 100 to 300 g/m3. As indicated in the figures, Pmax first increases rapidly
(STICK) in addition to an Arrhenius law similar to gas-phase reactions. This coefficient was defined as a probability that a certain reaction happened when a collision between a given gas-species occurred with the surface. When the sticking coefficient was close to 1, it was usually appropriate to apply the Motz-Wise correction (MWON). Users could turn the corresponding option on by including STICK and MWON in the REACTIONS line of their surface kinetics input file. 4. Experimental results 4.1. Explosion pressure Combustion time (tc) of dust cloud refers to the time between ignition and the Pmax. Burning time (tb) refers to the duration of burning of a single dust particle. As shown in Fig. 5a, for 30 μm aluminum dust, an increase of MCA powder concentration could lead to a gradual increase of combustion time (tc). In contrast, tc firstly decreases rapidly for a low concentration of MPP powder, but then eventually increases considerably as MPP powder concentration increases (Fig. 5b). The dust cloud can be considered as a concourse of single particle dispersed in a gaseous oxidizer. Therefore, we assume that the burning time (tb) of
Fig. 5. Pressure profiles of 30 μm Al dust doped with suppressant powder. 4
Journal of Hazardous Materials xxx (xxxx) xxxx
H. Jiang, et al.
Fig. 6. Pressure profiles of 5 μm Al dust doped with suppressant powder.
sufficient, large agglomerated residues are formed, and suppressant particles appear in explosion residues. It is worth noting that the surface of explosion residues forms a porous structure, as shown in Fig. 10d and e. Since the molecules of both MCA and MPP contain a blowing agent (melamine), a large number of gases products are formed by flame retardant decomposition during the explosion process (Wang and Yang, 2011). Based on the SEM results, it can be inferred that the condensed phase products of MCA and MPP dust explosion can cover aluminum particles surface and hinder its oxidation reaction. Suppressants, aluminum particles, and explosion products agglomerate together to form large porous residues. To reveal the chemical composition and different bonds structure of the mixture explosion residues surface, XPS is performed. Fig. 11 shows Al2p, O1s, C1s and N1s XPS spectra of explosion residues of Al/MCA mixture. In Fig. 11a and b, the peaks at 74.5–75.9 eV are attributed to Al3+ in Al2O3, and the peak at 72.9 eV is active Al0 (Moulder et al., 1995). It can be inferred that aluminum particles contain explosion residues, when the concentration of MCA is sufficient. From Fig. 11c and 11d, the O1s spectrum is divided into two peaks, which is corresponding to the binding energies of O in Al2O3. In Fig. 11e and f, the peaks at 284.9 eV–289.6 eV can be assigned to CeC bond, CeH bond, CeN bond, C]O bond, and C]N bond (Dante et al., 2015). As shown in Fig. 11g, the peaks at 398.8 eV and 400.0 eV are attributed to C]N bond and eNH2, respectively (Wang and Yang, 2011). These Ce and
for a low concentration of MCA and MPP, but then drops significantly as suppressant concentration increases. These results demonstrate that both MCA and MPP addition can enhance or suppress the explosion severity of Al dust. Since the lack of suppressants can increase the risk of accidents, it is extremely important to measure the minimum inerting concentration (MIC) of suppressant. Fig. 9 illustrates the inerting curves as a function of suppressant dust concentration. MIC data of SBC and ABC are adapted from Jiang et al. (2018, 2019b). The concentration of suppressant powder to the right of each inerting curves is sufficient to prevent an explosion. At the concentration to the left of each inerting curves, an explosion occurs. Meanwhile, the explosion is enhanced at the concentration to the left of each enhancement curves (Fig. 9a and b). As shown in Fig. 9c, MCA powder exerts the strongest suppression effect on 30 μm aluminum dust explosion, followed by MPP powder, ABC powder (NH4H2PO4 > 90 %) and sodium bicarbonate (SBC), respectively. For 5 μm aluminum particles, the MIC of ABC powder is the lowest. MCA and MPP deliver almost equivalent suppression performance. By comparison, a large number of SBC particles are required to prevent 5 μm aluminum dust explosion. 4.2. Condensed phase explosion residues As shown in Fig. 10, when the deployment of suppressants is
Fig. 7. The maximum explosion pressure of 30 μm Al dust for various suppressant concentration. 5
Journal of Hazardous Materials xxx (xxxx) xxxx
H. Jiang, et al.
Fig. 8. The maximum explosion pressure of 5 μm Al dust for various suppressant concentration.
spectrum at 530.6 eV, 531.6 eV and 531.2 eV are attributed to the binding energies of O in Al2O3. The peaks at 532.9 eV, 532.4 eV, and 533.3 eV can be assigned to the O in PeO bond, CeO bond, P]O bond and C]O bond. From Fig. 12g, the peaks at 133.9 eV are attributed to
Ne containing species are condensed phase explosion products corresponding to the high boiling products of MEL decomposition. Fig. 12 shows Al2p, O1s, C1s, P2p and N1s XPS spectra of Al/MPP mixture explosion residue. From Fig. 12c and d, the peaks of O1s
Fig. 9. Inerting curves of suppressants for Al dust explosion (adapted from Jiang et al., 2018 and Jiang et al., 2019b). 6
Journal of Hazardous Materials xxx (xxxx) xxxx
H. Jiang, et al.
Fig. 10. Images of explosion residues (Al: 300 g/m3).
5. Enhancement and suppression mechanism
the binding energies of P in pyrophosphate, metaphosphate or phosphate (Moulder et al., 1995). These condensed phosphates of P‒containing species are decomposed from MPP. From Figs. 11g and 12 h, the C]N bond and eNH2 are present in all condensed residue, for high additive concentration. Hence, it is probably the species corresponding to MEL or its decomposition products. The XPS results prove that Al/MCA mixture explosion residues are compose of Al2O3 and high boiling products (decomposed from MCA). Al/MPP mixture explosion residues are mainly Al2O3, high boiling products (decomposed from MEL) and condensed phosphates. Al particles, MCA or MPP may contain explosion residues. Results for condensed phase residues of 5 μm Al/suppressant mixture are similar, so they are not given and plotted.
5.1. Combustion enhancement mechanism Fig. 13 shows the calculated maximum flame temperature of gaseous‒Al doped with MCA and MPP. As can be seen that calculated maximum flame temperature decreases monotonously with suppressants concentration increases (φ refers to the equivalence ratio). Apparently, instantaneous heat release decreases with the addition of Suppressants. Fig. 14 demonstrates the pressure evolution at various inerting ratios. As can be seen that the maximum pressure firstly increases and then decreases with α increase. The trend of the calculated results (Fig. 14) is consistent with those of experiments (Fig. 6). NH3 will generate during the decomposition of MPP and MCA. When the
Fig. 11. Al2p, O1s, C1s and N1s XPS spectra of Al/MCA mixture explosion residues. 7
Journal of Hazardous Materials xxx (xxxx) xxxx
H. Jiang, et al.
Fig. 12. Al2p, O1s, P2p, C1s and N1s XPS spectra of Al/MPP mixture explosion residues.
oxidant is sufficient, the extra heat released by the combustion of NH3 leads to an increase in total heat release and heat release time, but results in a reduction in instantaneous heat release. When the oxidant is insufficient, instantaneous heat release and total heat release decrease with the addition of MCA and MPP. Moreover, in a constant volume container, as the instantaneous pressure rises and the flame temperature decreases, it means an increase in the amount of gas-phase products. It is indeed found in the experiments that large numbers of gasphase products are produced when the explosion is enhanced. To investigate the effect of chemical kinetics on flame temperature and explosion pressure, sensitivity coefficients of flame temperature and explosion pressures are determined for MCA- and MPP-doped aluminum flame as shown in Figs. 15 and 16, respectively. It can be seen from Figs. 15a and 16 a that HNCO + O = NCO + OH (R432), HCN + O = NCO + H (R460) and NH3 + O = NH2 + OH (R163) could reduce the flame temperature and the explosion pressure by consuming O atom. HNCO + M = NH + CO + M (R437), HNCO + H = NH2 + CO (R433) and aluminum oxidation reaction could increase explosion pressure. From Figs. 15b and 16 b, the most sensitive reactions of suppression of flame temperature and explosion pressure are reactions of the P ‒containing components: H3PO4 = HPO3 + H2O (R281) and HPO3+M = PO2+OH + M (R286). It is worth noting that R163 has a positive sensitivity to explosion pressure at α = 0.2 but has a negative sensitivity at α = 0.5 and 1.0. Therefore, NH3 and HNCO can
Fig. 13. Calculated maximum flame temperature of gaseous‒ Al doped with MCA and MPP.
Fig. 14. Explosion pressure of gaseous‒ Al doped with (a) MCA and (b) MPP. 8
Journal of Hazardous Materials xxx (xxxx) xxxx
H. Jiang, et al.
Fig. 15. Sensitivity analysis of flame temperature (ϕ Al = 1.0).
change in the order of suppression effectiveness of different suppressants for 5 μm and 30 μm Al dust explosion is attributed to the difference in the suppression effect of gas-phase reaction and surface reaction. Additionally, the rate of aluminum vaporization can be reduced by these suppressants addition, since the vaporization rate is limited by the number of Al(L). It can be also seen from Fig. 18 that site fractions of AlO(S) and O(S) decrease with the increase of inerting ratio. These results indicate that the diffusion rate of oxidizers close to the droplet surface is reduced by suppressant addition (Glorian et al., 2016).
enhance or suppress explosion pressure. On the basis of these results we concluded that when the concentration of MCA and MPP is low, the oxidizer is sufficient, the extra heat released from the exothermic reactions of suppressant and the increase of gas phase products could cause the pressure to rise. When MCA and MPP concentration was high, the pressure could be reduced by the explosion suppression. 5.2. Suppression mechanism AlO and O are the key radicals in gaseous-Al/air flame. Comparison of AlO and O concentration doped with different suppressants concentrations is used to determine the suppression capacity of MPP and MPP for gas-phase reaction of Al combustion. As shown in Fig.17, the concentration of AlO and O gradully decreases with inerting ratio increases. MCA is found to be more effective than MPP in scavenging flame radicals and suppressing gas-phase reaction of Al combustion. The combustion mode of aluminum particles depends on the composition of the oxidizer and particle size (Sundaram et al., 2016). Surface reactions become significant for small particles (< 10 μm) or large particles burning in H2O and CO2 atmosphere under high pressure (> 5 atm). For Al particles with fixed surface area, the rate of surface reaction depends on the number of active sites. Suppressants can interrupt surface reactions by reducing the number of active sites. Hence, the effect of suppressants on Al(L) site fraction is studied to evaluate the suppression capacity of MCA and MPP for surface reactions. As shown in Fig. 18, Al(L) site fraction decreases with inerting ratio of MPP increases, but slightly decreases with inerting ratio of MCA increases, indicating that MPP has a better suppression effect on the surface reaction of Al combustion. Simulation results reveal that the
6. Conclusions Experimental and numerical investigation are conducted to reveal the suppression mechanism of MCA and MPP for Al dust explosion. The conclusions can be summarized as follows. (1) For low additive concentration, the addition of both MCA and MPP can enhance or suppress Al dust explosion. The addition of MPP decreases the combustion time due to the increase of effective oxidizer concentration. With increasing MCA addition, there is an increase of O2 and O atom consumption, which results in the reduction of the effective oxidizer concentration and the rise of combustion time of aluminum dust explosion. (2) The explosion residues of Al/MCA mixture are Al2O3 and C‒ and N‒ containing species (the high boiling products of MEL decomposition). The explosion residues of Al/MPP mixture consist of Al2O3, high boiling products (decomposed from MEL) and condensed phosphates. When the suppressant is sufficient, a porous
Fig. 16. Sensitivity analysis of pressure (ϕ Al = 1.0). 9
Journal of Hazardous Materials xxx (xxxx) xxxx
H. Jiang, et al.
Fig. 17. Temporal evolution of (a) AlO and (b) O atom in Al/MCA/air (red line) flame and Al/MPP/air (blue line) flame at various inerting ratios. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article).
Appendix A. Supplementary data Supplementary material related to this article can be found, in the online version, at doi:https://doi.org/10.1016/j.jhazmat.2019.121648. References ASTM, 2010. E1226-12a. Standard Test Method for Explosibility of Dust Clouds. ASTM, 2014. E1515. Standard Test Method for Minimum Explosible Concentration of Combustible Dusts. Bazyn, T., Krier, H., Glumac, N., 2007. Evidence for the transition from the diffusion-limit in aluminum particle combustion. Proc. Combust. Inst. 31, 2021–2028. Beckstead, M.W., 2005. Correlating aluminum burning times. Combust Explo Shock+ 41, 533–546. Bojko, B.T., DesJardin, P.E., Washburn, E.B., 2014. On modeling the diffusion to kinetically controlled burning limits of micron-sized aluminum particles. Combust. Flame 161, 3211–3221. Bu, Y.J., Li, C., Amyotte, P.R., Yuan, W.B., Yuan, C.M., Li, G., 2020. Moderation of Al dust explosions by micro- and nano-sized Al2O3 powder. J. Hazard. Mater. 381, 120968. Catoire, L., Legendre, J.F., Giraud, M., 2003. Kinetic model for aluminum‒sensitized ram accelerator combustion. J. Propul. Power 19, 196–202. Chen, X.F., Zhang, H.M., Chen, X., Liu, X.Y., Niu, Y., Zhang, Y., Yuan, B.H., 2017. Effect of dust explosion suppression by sodium bicarbonate with different granulometric distribution. J. Loss Prevention Process Indust. 49, 905–911. Costa, L., Camino, G., 1988. Thermal-behavior of melamine. J. Therm. Anal. 34, 423–429. Dante, R.C., Sanchez-Arevalo, F.M., Chamorro-Posada, P., Vazquez-Cabo, J., Huerta, L., Lartundo-Rojas, L., Santoyo-Salazar, J., Solorza-Feria, O., 2015. Supramolecular intermediates in the synthesis of polymeric carbon nitride from melamine cyanurate. J. Solid State Chem. 226, 170–178. Dastidar, A., Amyotte, P.R., 2002. Determination of minimum inerting concentrations for combustible dusts in a laboratory-scale chamber. Process Saf. Environ. Prot. 80, 289–299. Dounia, O., Vermorel, O., Poinsot, T., 2018. Theoretical analysis and simulation of methane/air flame inhibition by sodium bicarbonate particles. Combust. Flame 193, 313–326. Gijsman, P., Steenbakkers, R., Furst, C., Kersjes, J., 2002. Differences in the flame retardant mechanism of melamine cyanurate in polyamide 6 and polyamide 66. Polym. Degrad. Stab. 78, 219–224. Glorian, J., Gallier, S., Catoire, L., 2016. On the role of heterogeneous reactions in aluminum combustion. Combust. Flame 168, 378–392. Jahromi, S., Gabrielse, W., Braam, A., 2003. Effect of melamine polyphosphate on thermal degradation of polyamides: a combined X-ray diffraction and solid-state NMR study. Polymer 44, 25–37. Jayaweera, T.M., Melius, C.F., Pitz, W.J., Westbrook, C.K., Korobeinichev, O.P., Shvartsberg, V.M., Shmakov, A.G., Rybitskaya, I.V., Curran, H.J., 2005. Flame inhibition by phosphorus‒containing compounds over a range of equivalence ratios. Combust. Flame 140, 103–115. Jiang, H., Bi, M., Gao, W., Gan, B., Zhang, D., Zhang, Q., 2018. Inhibition of aluminum dust explosion by NaHCO3 with different particle size distributions. J. Hazard. Mater. 344, 902–912. Jiang, H., Bi, M., Li, B., Ma, D., Gao, W., 2019a. Flame inhibition of aluminum dust explosion by NaHCO3 and NH4H2PO4. Combust. Flame 200, 97–114. Jiang, H., Bi, M., Li, B., Zhang, D., Gao, W., 2019b. Inhibition evaluation of ABC powder in aluminum dust explosion. J. Hazard. Mater. 361, 273–282. Li, B., He, Y., Li, Z.S., Konnov, A.A., 2013. Measurements of NO concentration in NH3‒doped CH4 + air flames using saturated laser‒induced fluorescence and probe sampling. Combust. Flame 160, 40–46. Lynch, P., Krier, H., Glumac, N., 2009. A correlation for burn time of aluminum particles in the transition regime. Proc. Combust. Inst. 32, 1887–1893.
Fig. 18. Site fractions at aluminum surface as a function of inerting ratio in air.
agglomerated residue is formed, and aluminum particles, MCA or MPP may be contained in condensed phase residues. (3) Calculations show that flame temperature decreases monotonously with suppressants concentration increases. The formation of large quantities of gas phase products and the extra heat released from the exothermic reactions of suppressant leads to an increase in pressure. MPP is more effective than MCA for surface reaction suppression, but MCA is more effective for gas-phase reaction suppression.
Declaration of Competing Interest We declare no financial interests.
Acknowledgments The authors gratefully acknowledge the financial support from the National Natural Science Foundation of China (No. 51874066 and No. 51674059), Key Laboratory of Building Fire Protection Engineering and Technology of MPS (KFKT2016ZD01), the Fundamental Research Funds for the Central Universities (DUT16RC(4)04), (DUT18RC (3) 038). The authors would like to thank Dr. Julien Glorian for providing the chemical kinetic models of gas phase and surface reactions.
10
Journal of Hazardous Materials xxx (xxxx) xxxx
H. Jiang, et al.
dust deflagrations. J. Loss Prevention Process Indust. 36, 246–253. Tao, W., Li, J., 2018. Melamine cyanurate tailored by base and its multi effects on flame retardancy of polyamide 6. Appl. Surf. Sci. 456, 751–762. Twarowski, A., 1996. The temperature dependence of H + OH recombination in phosphorus oxide containing post‒combustion gases. Combust. Flame 105, 407–413. Vizcaya, D., Pinilla, A., Amin, M., Ratkovich, N., Munoz, F., Murillo, C., Bardin-Monnier, N., Dufaud, O., 2018. CFD as an approach to understand flammable dust 20 L standard test: effect of the ignition time on the fluid flow. AIChE J. 64, 42–54. Wang, G.J., Yang, J.Y., 2011. Thermal degradation study of fire resistive coating containing melamine polyphosphate and dipentaerythritol. Prog. Org. Coat. 72, 605–611. Wang, Z.Z., Lv, P., Hu, Y., Hu, K.L., 2009. Thermal degradation study of intumescent flame retardants by TG and FTIR: melamine phosphate and its mixture with pentaerythritol. J. Anal. Appl. Pyrolysis 86, 207–214. Yang, H.Y., Song, L., Tai, Q.L., Wang, X., Yu, B., Yuan, Y., Hu, Y., Yuen, R.K.K., 2014. Comparative study on the flame retarded efficiency of melamine phosphate, melamine phosphite and melamine hypophosphite on poly(butylene succinate) composites. Polym. Degrad. Stab. 105, 248–256.
Huang, C.Y., Chen, X.F., Yuan, B.H., Zhang, H.M., Dai, H.M., He, S., Zhang, Y., Niu, Y., Shen, S.F., 2019. Suppression of wood dust explosion by ultrafine magnesium hydroxide. J. Hazard. Mater. 378, 120723. Kalejaiye, O., Amyotte, P.R., Pegg, M.J., Cashdollar, K.L., 2010. Effectiveness of dust dispersion in the 20-L Siwek chamber. J. Loss Prev. Process Ind. 23, 46–59. Konnov, A.A., 2009. Implementation of the NCN pathway of prompt‒NO formation in the detailed reaction mechanism. Combust. Flame 156, 2093–2105. Korobeinichev, O.P., Shvartsberg, V.M., Shmakov, A.G., Knyazkov, D.A., Rybitskaya, I.V., 2007. Inhibition of atmospheric lean and rich CH4/O2/Ar flames by phosphorus‒containing compound. Proc. Combust. Inst. 31, 2741–2748. Moulder, J.F., Stickle, W.F., Sobol, P.E., Bomben, K.D., 1995. Handbook of X-Ray Photoelectron Spectroscopy. Physical Electronics, Minnesota. Smith, G.P., Golden, D.M., Frenklach, M., Moriarty, N.W., Eiteneer, B., Goldenberg, M., Bowman, C.T., Hanson, R.K., Song, S., Gardiner, W.C., Lissianski, V. V., Qin, Z., 2019 GRI-Mech 3.0. n.d. URL http://combustion.berkeley.edu/gri-mech/version30/ text30.html (Accessed 12 February 2018). Sundaram, D.S., Puri, P., Yang, V., 2016. A general theory of ignition and combustion of nano‒ and micron-sized aluminum particles. Combust. Flame 169, 94–109. Taveau, J., Vingerhoets, J., Snoeys, J., Going, J., Farrell, T., 2015. Suppression of metal
11