Loading...
HomeMy WebLinkAboutMaverick Boats Phase 2 Expansion - GEL & GRIND BOOTH PERMIT 2107-0635 Dust Explosibility report for Southeast Building Solutions�FAU SeKE July 31, 2018 Gary Mayfield Report No.: FAII8-0952 Southeast Building Solutions Project No.: DST 6789 104 Mcco Lane Oak Ridge, TN 39830 Tel: (865) 29MI94 E-mail: torawield(alsbsmechanical.com Subject: Dust Explosibility Testing Dear Gary, In response to your mquesS Fauske and Associates, LLC (FAI) conducted an Explosion Severity Test per ASTM E1226, a Minimum Explosible Concentration test per ASTM E1515, a Minimum Ignition Energy test per ASTM E2019, a Minimum Ammunition Temperature (MATT or commonly MM test per ASTM E1491, and a Total Combinable Content Test by DSC/TGA. These tests were conducted on the following sample labeled: 1. Maverick Boats. "Sample Yuf,erelvM On ✓nne2l, mra" The material properties and explosion hazard results are summarized in Tables 1 and 2, respectively. Table 1: Material Properties Mah Moist COnt¢nt Mi Rrllete See =4) Iwn) 52 Maverick Bon6 1.3 79S<I5 am W%<San gm Table 2: Explosion Huard Results P., NAMNAMMEE Material (bar) r- lbv-mlv) (g/m') Mith 1ndeieA`e1 9.54IRS, 201a1UM 50<MEC<60 1<MIE<3 MEC - -59 Es-21 MIT Closed Combmtible Mavenfk BOLA Content �) xtX 460 -64 lavvi FAUEKEI OY116w3oi 323 B750 . FAX:i630l �9e6-s<B1 . E-MAIL. INFOOm"I'KEuerm Pow 1, ofFaus4e asAss Gales. UC -2 of24- The calculated Explosion Severity (ES) and Ignition Sensitivity (IS) numbers for the sample tested are listed in Table 3 based on Pittsburgh Pulverized Coal (Lot#1157) dust (P Zl bar, (dP/dtfa 530 bm/s, MIE (wl ma more) = 140 m3, MEC = 43 g/m3, MIT = 5800C, Panicle Size Distribution = 97% < 95pm [January 20121). The ES and IS values were calculated using the following equations: (P x R),vap„ ES — (P x R)rmawran raw Where: P= Maximum explosion overpressure, Hs..) R= Maximum rate of pressure rise,(dP/dta ) IS IT x E x C pa bu.aa coal IT x E x C)samptr Where: T=Minimum Auto -Ignition Temperatee,(MIT) E=Minimum Ignition Energy, (MITI C=Minimum Explosion Concentration, NEC) Table 3: ES and IS Numbers Material ES is NFPA 499-2008 recognizes a Class 11 dust as having an IS greater than or equal to 0.2 or ES greater than or equal to 0.5 to be appreciable explosion hazards requiring elecMcal equipment soluble for Class H locations. However, please note Nat current NTPA 499-2019 recognizes a Class It dust as having an overpressure greater than, or equal to, I bar in the 20-L screening test. Eased on the both NFPA criterions, this sample is classified as a Class 11 dust for National Electrical Code proposes. Please refer to Appendix E in OSHA Directive Number CPL 03-0 YW8 and NPPA 499-2008 for further discussion of the ES and IS numbers. FAI1841152 Pmp.lerary hope." fF uske & a..sociutey uC July 31, 2018 -3 of24- The following paragraphs describe the methods, procedures, and detailed results for the tests conducted atFauske and Associate I.I.C. Material Preparation Per ASTM recommendation, the moisture content of the test sample should not exceed 5 wt.%. The sample was tested "as received" per the client's request The moisture content of the sample was analyzed using a Mettler Toledo M83 Halogen Moisture Analyzer and was determined to be 1.3 wt%. Per ASTM recommendation, the materials for dust explosibility testing should have a panicle size distribution thaz is at least 95% less Nan 200-mesh (75 µm). The sample was tested "as eived" per die client's request The panicle size distribution of the sample was measured using a Malvern MS 30M Panicle Analyzer and was determined to be 99% less than 75 µm (see Appendix A for detailed results). Please note that finer dust under actual process conditions may produce hiaher risk exulosibilityv oarametera Center dust mar produce lower risk explosibility osfon Severity(P... Ke)Tests The test was conducted in accordance with ASTM E1226, "Standard Test Method for Sxplosibilip of Dust Clouds". A 20-L Siwek chamber manufactured by Kahner A.G. of Basel Switzerland was used for the test. Scope This test method is used to determine the deflagration parameters of a combustible dust -air mixture. The parameters measured are the maximum pressure and the maximum role of pressure rise. Data obtained from this lest method provides a relative measure of deflagraton chemctensfics and is also used for the design of protection systems, such as deflagation venting. Terminoloev The Pm„ is the average maximum pressure (above pressure in the vessel at the time of ignition) reached during the course of a deflagmtionf the optimum concentration of the dust tested. P. is determined by a series of tests over a large range of concentrations. The pHvdt)m„ is the average maximum value for the rate of pressure increase per unit time reached during the course of a deflagmtun for the optimum concentration of dust tested. It is determined by a series of tests over a large range of concenrcations. The Deflagrafion Index, Ika is the average maximum dPhh normalized to a LO-m'volume. It is measured at the optimum dust concentration. Ks, is defined in accordance with the following cubic relationship: FAI18-0752 Pmp.lerary Isvi e.N fF wske & asociates,, aC July 31, 2018 J of24- where Tes[�Setu _ rrI X Via P = pressure, bar t = time, sec V test chamber volume, m' Sat Deflagration Index, bar -Ms The 20-liter chamber, as depicted in Figure 1, was equipped with m air nozzle to disperse the material dust, and two electrodes for connecting two (2) squib (Sobbe) igniters (2 x 5 U — 10 0 of ignition met") to a voltage source. The chamber was equipped with two pressure transducers to measure the pressure output of an explosion. The pressure -time data was collected by a high- speed dam acquisition system. The test was automated and computer controlled. A more detailed description of the apparatus is available from the mmufecturer. Figure 1: 20-L Sheek Chamber Test Setup Calibration Per ASTM F1226, "Standard Test Method for Explosibiflry ofDmi Ciouds," §10, "Calibmfion and Stmdardimtion" p.p. 10.2, the rest vessel to be used for routine work must be standardized sing dust samples wbase Ks. and Pm„ parameters are known in the I-m' chamber. The Pm„ value for each dust most agree to within t 1010 with the 1-m'value and the Kstvalue must agree To within t20a/a. The Khhner 20-L Siwek chamber test equipment used for testing was calibrated by carrying out the calibrating tests on Niacin USP Special (Nicotinic Acid —28 par mean diameter) provided by the test equipment manufacturer as pan of a bi-annual round robin inter -lab calibration exercise. The dust deflagration data measured in the FA120-L chamber, shown in Table 4, is compared with the dam reported by the manufacturer for other 20-L and I-m' chamber dam. The FAT rest equipment and methodology compare very well m the accepted explosion severity values. FAT 18 0752 Peop.ieraey Pmpe,"fFawke & a..soeiatee, uC July 31, 2018 -5 of24- Table 4: Dust Deflagration Calibration Data Molhurecooteot Meuo Parllele Sb e M tedhl Tested (cot%) (pm) Nladn O9V 9Poa11 Al2 —28 am NICOsoe A¢W roeivM el (30.LI Is"OrCelia l] Ranks P.. Ke' P^^ Ifu era bar v-m/r bar bar-wa 8.It IM 256t IM 821IN/o zs4tlo% TPAIResuN 8.At IG% 252t I0% 8.2t IM 252t IM 10% NU 10% FAIR8.2t Ps' Iso MSp4mI Mar) mm/ae 7.8t IW/ 248t M. Iual 7.9t IM 246tIG% Procedure For each teat, a known amount of material was weighed and then placed in the dust dispersion chamber or on top of the rebound nomle if there was too much that to fit in the dispersion chamber. The ignition source (igniters) was placed in the ester of the chamber. The chamber was sealed and all valves were closed. Tom the chamber was partially evacwted so after addition of dispersing air, the desired nominal pressure in the chamber of one bar absolute was reached prior 0 radiation of the deflagration ten. Then the automated test sequence was initiated using the computer control software provided by the manufacturer. The dust dispersion chamber was pressurized to 21 foods). The dispersion solenoid valve was opened releasing the dust into the 20- L chamber and raising the chamber pressure on I bar(a). After a computer controlled delay time of 60 t 5 on the chemical squib igniters were initiated. The resulting pressure use ured by two piezoelectric pressure transducers. The pressure -time history dam was then reviewed and the maximum explosion pressure and rate of pressure rise was determined. The manufacturer supplied computer program used in review, the pressum-time data applies a correction factor to the maximum measured explosion overpressure 0 account for the energy of the ignition source and the quenching effect of the vessel walls. To determine the maximum explosion pressure and pressure rise rate, successive tests were run at increasing dust concentrations while using the same strength ignition source (10 U). FA118-0752 Pmp.ioary Pmpe.N yF wake & a.rsoeiutex aC July 31, 2018 -6of24- The peak pressure was determined by analyzing the pressure -time curve. The peak pressure rise raw was determinxi by measuring the slope of the prvssule-timecurve (dP/dt). An initial concentration of 125 g/m' was tested. Then the dust concentration was systematically increased until curves were obtained for bofi (dP/dt),„and P. that indicated an optimum value had been reachcd. If it was indicated that the optimum concentration for (dP/dt)a,,,; or P. was less than 125 ghv, the tested concentration was halved; (60, 30 glmc) until the optimum value was obtained. At least two (2) additional test series were ran at the concentrations where the maximums were found and at one (1) concentration on each side of the maximums. Explosion Severity /A._ Rrl Results The explosion severity test results for the sample tested are summanzed in Table 5. Figures 2 and 3 depict the explosion overpressure and rate of pressure rise as a function of dust concentration for the dust sample tested. The derailed explosibility dam are also presented in Table6. The sample tested was found to be highly explosible producing a Ks, value of 201 bar-Msec, than, classifying it as an S@ class dust since the Ks, value was between 200 bar-Msec and 300 bar-Msec. Mitigation strategies will be required when handling this powder in order to minimize the risk of an explosion. Plese consult NFPA 652, 654, 68, and 69 for guidance on explosion mitigation. Please note that finer dust under actual process conditions may produce banner risk explosibil Table 5: Explosion Severity Test Results Molamre Mean Particle Pe„ (dP/Nj. Ka MaMrIVl Tared ConMn[ give orm) (bar) (bvrh) (banm/v) M.H R Maverick Boazs 13 ]W. <]S 4m ].S a 10'/ 739a lo% 301 a In/. IW%<5lp m FAI18-0752 Pmp./erary Prope.Nlpayske&.15wriates,, uC July 31, 2018 -] of U. Table 6: Detailed Explosion Severity Test Results ssapm: Maeeevk eam ROM empmMe; 16•C 6aramma Pruvw M Ma aderecH Ml r:6au DAearno: July 2s. 2018 12s m 250 nP 500 m 150 N 1000 aP 12M m Run P. dP/dl Ws P. dP/M /s P. dP/dl Ws P. dP/dt /s P. dP/d2 Ws P. dP/M /s 1 3.9 106 65 16 1.5 483 ]C 613 7.1 7" 68 676 2 62 300 ZI 391 "l.3 569 7.0 712 66 out 3 63 310 ]] 681 70 669 7.1 ]CD 6/ 528 Ran ISm R. dP/dt Ws 1 6.6 5<5 2 100 I 250 YO TL 1" 1250 lYo a Figure 2: Explosion overpressure as a function of dust concentration in a 20-L chamber for Maverick Boats FA118-0952 Pmprietap Pmpe,"fFauake & A5smiares uC July 31, 2018 -8 of24- mnn Imo Ix 250 500 750 Imo 1250 Ism usual Figure3: Rate of pressure rise as a function of dust concentration in the 20-L chamber for Maverick Boats MlnLnum Exolosible Concentration MEC) Teats The rest was performed per ASTM E1515, 'Standard Test Method for Ministers Eaplosible Concentration of Combustible Dusts ". A 20-L Siwek chamber, manufactured by Kdhner A.O. of Hanel Swittedand, was used for the teat. Score Minimum F,xplosible Concentration is the minimum concentration of a combustible dust that is capable of propagating a deflagration through a uniform mature of the dust in ad, under the specified conditions of the test. The MEC data obtained Gam this test method can provide a relative measure of the concentration of a dust in the form of a uniform cloud necessary for an explosion. The values obtained by this test method are specific to the sample tested (pmticularly the particle site distribution) and the meshed used and is not to be considered intrinsic material consumers. Test The 20-liter Siwek chamber, as depicted m Figure 1, was equipped with an aG nozzle to disperse the material dust, and two electrodes for connecting one squib (Sobbe) igniter (5 kl stand energy) to a voltage source. The chamber was equipped with two pressure transducers to FAI180752 Pmprietwy PmPe,"yFauske&A..sociare,., uC July 31, 2018 -9 of24- measure the pressure output of an explosion. The pressure -thee data was collected by a high- speed data acquisition system. The test was automated and computer controlled. A more derailed description ofthe apparatus is available from the munufeco er. Procedure For each tat, a knoum amount of material was weighed and then placed in the dust dispersion chamber or m top of the rebound nozzle if there was too much dust to fit in the dispersion chamber. The ignition source (igniter) was placed in the center of the chamber. The chamber was sealed and all valves were closed. Then the chamber was partially evacuated so after addition of dispersing air, the desired normal pressure in the chamber of one bar absolute was reached prior in initiation of the deflagration lest. Then the automated test sequence was initiated using the computer control software provided by the manufacturer. The dust dispersion chamber was pressurized to 21 bona). The dispersion solenoid valve was opened releasing the dust trim the 20- L chamber and raising the chamber pressure to I bona). After a computer controlled delay time of 60 is 5 ms the chemical squib igniter was initiated. The resulting pressure rise was measured by two piezoelectric pressure tram lacers. The pressure - time history data were then reviewed and the maximum explosion pressure was determined. The manufacturer supplied computer program used in review the pressure-fime data applies a correction factor to the maximum measured explosion over pressure to account for the energy of the ignition source and the quenching effect of the vessel walls. To determine the minimum explosible concentration, successive tests were run at decreasing dust concentrations while using the same strength ignition worse (5 U). An explosion can be differentiated from a'NO GO" by evaluating the increase in pressure (due to airdustmixture explosion) above that of the igniter. If peak pressure was 1 (one) her more than the igniter pressure, the explosion was considered a `GO". The rate of pressure rise pW/dt) is not considered for MEC calculations. The WC is interpolated between the `Yid' and "No Go" concentrations. An initial concentration (usually 100 WW) was tested. The dust concentration was then incrementally increased or decreased mail "GO" and "NO GO" were established. Then the dust concentration was systematically modified by 10 g/m' when the concentration is 100 g/m' or less, or<25%if grater than 1 W g/m' until the MEC ofthe sample was determined. FAI18-0752 Fmp.hsary Fmpe.NgfFauake&15smjates uC July 31, 2018 -10 of 24- Minimum Exnlocible Concentration Test Results The MEC test results for the sample are summarized in Table 9, Ole MEC being the interpolated concentration for a 1 bar explosion pressure. The calculation is made using the lowest concentration for which an ignition has been observed and the highest concentration tested at which 3 subsequent no ignitions have been observed. The detailed explosibility data is also presented in Table 8. The MEC value for the sample was determined to be between 50 g1W and 60 glm' with an interpolated value of 59 glW. It is important to keep dust concentrations below the MEC in order to minimize the risk of a dust explosion. As an illustration, consider a scenario where the sample is deposited as a thin and even layer in an empty room with dimensions of 12m x 12m a 3m. Assuming an approximate bulk density of0.3 g/ml, a dust layer that is 0.6 ram thick is sufficient, when uniformly dispersed, to create an exploslble dust cloud. Please refer to NPPA 652 and 654 for guidance and instructions. Table l: MEC Test Results Moeace: Content Mean ]rtu cle Sf¢ Minbaam Explouble W MECa, Material Tested hrl, x) (pmlWHOP Comenthaunn 52 Mavenck Boats 1.3 79%<75 µm 50<ry¢C<0 59 100%e<500 am Table 9: Detailed MEC Test Results Room Teniperadrc 15T Bemmelne Premwe: M cast Rami,x H,,,uv 90% 0.koFTIn ray 25. 2018 sagas, tease so ism, isgoac wrom, untio,, FA118-0752 Proprietary Proper" IPauake & Mrociutex uC July 31, 2018 -1I of 24- Minimum ImUnon Enerev (MIE) ofa Dust Claudia Air Tests The test was performed per ASTM E2019, 'Standard Test Menhon for Minimum Ignition Energy of a Dust Cloud in Air' using a Mike 3 MIT test apparatus manufactured by Kilmer A.G. of Basel Switzerland. Scope This test method covers the determination of the minimum ignition energy of a dust cloud in yr by a high voltage spark. Data obtained from this test method provides relative data of ignition sensitivity ofa dust cloud. The values obtained are specific ro the sample tested, the method suW and the test equipment mark The values are not to be considered intrinsic material constants. Any change at particle sum, shape, volatility or moisture will change resuhs. The Test E ermusent Calibration The test equipment used for testing was calibrated by carrying out the calibrating tests on Niacin USP Special (Nicotinic Acid —30 pm mean diameter) provided by the Mike 3 test equipment manufacturer As part of a bi-annual round robin inter -lab calibration exercise. As a reference, Minimum Ignition Energy (MIE) values and ranges for the equipment mane fectmer's reference dust, and MIE value determined for the same reference dust by the FAI test equipment used are shown in Table 9. Table 9: Minimum Ignition Energy Test Results on Reference Dust Mrtenal: Niuin USP 5 u it Wimarm, Acid-v rceiv cmure nR no.%) Mao Rroele R6ane,GRo 11 RSiftnuln 1�) FAI M6, Rush, M MI%F.391 MII�3a M10(Fd0 IL (�) hmI b .LLIS -3a4m 0.5cM1E4r ICM1@c3 ICF®c3 ICM1Ea3 ZIaAO 2 7ro30 py_E6 Es=L4 Pn-1.4 Ps-35 Test Se The Mike 3 Woman, as depicted in Figure 4, is equipped with a dust dispersion cup, a dispersion male (mounted at the center of the dispersion cop), a compressed air source (2 bar) and a pair of electrres momted m the rem glass cylinder wall (one a fixed high voltage electrode, the other a moving ground electrode). The discharge circuit consists of a high voltage supply (15 kV or 11 kV), interchmgeable capacitors, and inductance source (I m1q A more detailed description of the apparatus is available from the manufum er. FAI184)752 Pnnvie,ary v ope.N fF coke & A5 xiatex aC July 31, 2018 -12 of 24- Figure 4: MIKE 3 Minimum Ignition Energy (MIE) teat setup The Mike 3 apparatus was powered up and allowed to stabilize for thirty, minutes before tearing. The compressed air supply was turned on and adjusted to 9 bar(g). Then using the condonable acquisition software supplied with the apparatus, the mew levels for file various discharge levels were tested for opembility (1000, 300, 100, 30, 10, 3, and 1 M). Once successfully completed the apparems was deemed ready for dust tests. For each test, a desired amount of dust material was placed in file cup around the dispersion nozzle. The electrode gap was set at b rem. Using the computer contropdam acquisition software the desired energy level was selected, and the program initiated. The dust was then dispersed as a dust cloud inside the chamber. After a set delay time the computer controlled spark would discharge and observations would be made for flame ignirioWpmpagation. The difference between an ignition, "Go" and a non -ignition, "No Go" was evidenced by the appearance of a fireball in file chamber. Once "GoMo Go" spark energy was found for a particular concentration, the procedure was repeated for higher and lower dust concentrations. The spark energy was reduced in steps at the given dust concentration until the dust cloud no longer ignited in any of ten (10) tests at a given energy. The procedure was repeated at different dust concentrations, and ignition delay times, =fit the lowest (minimum) ignition energy value as found. To calculate the capacitor stored ignition energy (stored yr the capacitor by an applied voltage) the following equation was utilized: FAI18 0752 amiviem y rmpe.N fF wske & esxiates, aC July 31, 2018 -13 of24 W tone=2C(Vi— e Where: WSroaED =capacitor stored ignition energy, Joules C = capacitance of the discharge circuit capacitor, Famds Vl = voltage to which the capacitor is charged, Volts V = college on the capacitor after discharge, Volts The M1E is then reported as being greater than the highest energy that was not able to ignite the dust cloud and lower than the lowest energy that was able to ignite the dust cloud (Energy for Non -Ignition < ME < Energy for Ignition). In situations where the lowest energy the tester is capable of producing (1 mJ) igniten the dust being reared, no additional testing is required end the ME is reported as being <t m3. For the purpose of comparison between different apparatus, one statistical ME value (Es) instead of the energy range (E1, E2) may be used. This single statistically interpolatd value (Es) can be estimated by the use of the probability of ignition m follows: (weex l4e l)\ E, = 10 ` (( )E �I // Where: I[E2]=numbs oftests with ignition at the energy E2. (NI+I)[E2] =total number oftens at the energy M. Below is an example calculation: Ignition Energy DUALOMir�gm L1-L obvmber Probability 300 WO 900 12W 1500 JO M I I I M 3of5 10 NI M M I I bola toaaa�a0aH 54LH )` fi,=30 / =1]mJ Where: I = ignition of dust N1= no ignition of dust m 10 vials FA118-0752 Pmp.letary vrope.NfF nke&.15swunes uC July 31, 2018 -14 of 24- Minimum leninon Rnerev (MIE) Test Rest; The MIE test results for the material tested sre summarized in Table 10. The data is represented graphically in Figure 5. The open circles in the figure we dust concentrations that did not ignite at a given ignition energy. The solid squares (if present) represent concentrations Nat ignited. The vertical For above the solid square (if present) is proportional to the number of experiments conducted before the test material ignited. The blue mark (1) represents dam 6om tests conducted with an ition time delay of t20 ors between dust dispersion and spark discharge; while the red represents dam from tests with an ignition time delay of90 ors. Teem with inductance borer simulate spark discharges from electricaTelectomic so whereas tests without inductance simulate specks from purelideal electrostatic sources. Please note that tests conducted with inductance in the circuit produce species that me more incendive than circuits without inductance —thereby generating a conservative MIE level. The results indicate that the sample tested has a very low MIE value between 1 and and 3 M with n interpolated value of 2.1 ad. Most electrostatic ignition sources have energies less than 1000 or. Common electrostatic ignition sources sm typically less tban 30 mJ (people, small isolated components, etc.). Figure 6 is an illustration comparing various electrostatic discharge energies ro the immutably level of various fuels Grounding and bonding of process equipment should be implemented. Please see NFPA 654, 652,Ti, 68, and 69 for guidance and instructions. Table 10: Minimum Ignition Energy Test Result with Inductance] Ram Tempustux. ITC Bursamia Vreawe: 995 man Rdmiw 54°6 use aria July 212018 Material Tested MdBore Cmteol Meia PUHele9lu MBE Interpolate ME, 6c OAILVO (PM) (m.1) paid) Bmtr 13 52Maverick yrsg5 µm 1<ME<3 2.1 law, <500 um FA1184Y752 Pnw ierary Pmpe,"Y`F^ueke &.15weiates, aC July 31, 2018 -15 of 24- i i j a � O O soo ,too into ,too taao can Figure 5: Ignition Energy as a function of dust concentration in the Mike 3 MIE ]u'Ith inductance] appa etas for Maverick Boats Epuivalent Eneares g 104 104 //� 10 wtg ` loll0 10102 12 table cayaoem.�ne - . ma,.a maw,ee.'w daau. W � 1 � acma.anorsl .., aaoo 1 0 3 101 - 10' f 10 a Erybrl ,raw 104 and g 10' 104 al W Discharge Types Materials Figure 6: An illustration of typical discharge energies compared to ignition energies of various fuel types FA118-0952 Pmpricap, Pnse,"fFamke at4smiane,•, uC July 31, 2018 -16 of 24- Minimum Aumienidon Temperature (MIT) ofa Dust Cloud Text The teat was conducted per ASTM E1491, "Standard Test Method for Minimum Autolgnition Temperature ofLust Clouds" The apparatus used was the BAM Oven manufactured by Kubner A.G. of Basel Switzerland. Scope The test method covers the determination of the minimum temperature at which a given dust cloud will sow ignite when exposed to air heated in a famace at local atmospheric pressure. The data obtained firm this test method can provide a relative measure of dust -cloud auwignitirn temperature. Significance and Use The test data developed from this test method can be used to limit the temperature to which a dust cloud is exposed, thereby mitigating ignition of the cloud Because of the short duration of the test the dam obtained are most applicable to industrial equipment where dust is present as cloud for a short time. The Test Eoutoment Calibration The test equipment used for testing was calibrated by carrying out the calibrating tests on Pittsburgh Bimminous Coal, and the calibration results are presented in Table 11. Table 11: Mmimom Auwignifion Temperature MIT -Dust Cloud Test Results Mesa Permits Size Moisture Content Litervre Data m FM]ms" Material Ye54J gra) (wits) MITurem MlTemaew C Pittsburgh Bituminous Coat <]SPm `S 570 SSO (standard Pefertoce Muennl Lm meivM rarived Procedure The minimum autoigniHon temperature (MITaoa nwa) of the material was determined using a BAM Oven designed at the Bundesanstalt fir Materialforschung and loafing in Berlin, Germany and manufactured by Kuhner A.G. of Basel Switzerland. The BAM Oven apparatus is shown in Figure ]. FA118-0752 Propteraty Propen"Y'F mke & a..sor ams, uC July 31, 2018 -17 of 24- Figure ]: RAM Oven apparatus An initial estimate of the MIT was made by heating the oven to a predetermined temperature (maximum of WO-C) and then switching off the power and allowing the temperature to fall. At intervals of 50-C, as the temperature falls, premeasured dust (1 ml) was dispersed Into the furnace with a blest of air. Observations for the presence, or absence, of flame exiting the rear of the oven were made. After obtaining an estimate of the MIT, the exact value was determined by series of tests at various dust concentrations and at temperatures near the estimate. The selected starting temperature was the lowest temperature for which Flame was observed. The ignition rests were widened by decreasing the test temperature in 10-C increments until flame is no longer observed (within 5 seconds). For this series, the oven was stabilized at the set temperature before each test. For each test, the over was set to the predetermined test temperance. The premessmed dust was placed in the dust sample tube. The initial test concentration was a measured volume of 1 mi. The tlmt sample tube was then insured into the furnace and the dust dispersed with a blast of no Ignition of the dust was defined as the observation of flame exiting [he flap at the rear of the oven. At the highest temperature for which no flame was observed, at least three dust concentrations were tested (volumes of 05, 1, and 2 ml). The test report lists the MIT as the lowest temperance at which any positive result (flame) was observed over a series of concentrations. Mumnum AubbnRlon Temperature (MITI of Dust Cloud Test Results The Minimmn Autoignition Temperature (MITd�u c w) results for the sample tested are ummariud in Table 12. Figure 8 graphically counties the tabulated test data for the material tested. The MITww need for this sample was determined in be at 460°C. This ignition risk should be evaluated in light of the process enviremment. Please consult NFPA 652, 654, and 70 for guidance and instructions. FAI18 0752 P vivioary Rvpe.N Ipauske & Mswiates aC July 31, 2018 -18of24- Table 12: Minimum Autaignitlon Temperature MIT -Duet Cloud Test Results Material Tested (m) en 52 meU Relatle xumitl' :50% rue orren'. lul .2018 2.5 4Go ONo Go 2.0 E 1.5 j 1.0 0.5 0.0 250 300 350 400 450 500 550 600 650 TemlSeratum(T) Mere 8: MIT Results for Maverick Roats FA118-0752 Pmp.letary Prope, fF wske & A..soeiates, uC July 31, 2018 -19 of 24- Total ConsbustiDie Content Determination The test was conducted by Thermogmvimetric Analysis / Differential Standing Calorimeter Analysis (TGA/DSC) using a TA Instruments SLIT 650 apparatus. The test was conducted per the manufacturer's accepted method. Scone The test method covers a laboratory procedure to assess the presence of enthalpic (exothermic reactions and endothermic reactions) and mass changes using milligram sample sizes, and approximates the temperatures at which these enthalpic changes occur. Setum The TGA/DSC measures mass linalgain and heat flow a various temperatures. A TA Instruments SDT 650ihermogravimetric Analyzer / Differential Scanning Calorimeter (fA Instrument Corporation, New Castle, DE) with a standard configuration (temperature range of room temperature to ION-C) was utilized for this test to determine the endothermic (i.e. melting) and exothermic (i.e. ignition) temperatures of the material in addition to mass loss or gain. The TGA/DSC cell was confined in a ceramic heating cylinder, which was purged with a desired flow rate of purge gas. Procedure In the TA Instrmnents SDT 650 TGA/DSC, a reference and the sample were heated under a specific temperature program rate of 20-C/min. A computerbasedthermal analysis/controller unit controls the tempttature program. The TGA/DSC thermograph, as a function of timdremperanuq was stared and monitored in the computer. Each thermograph was then analyzed to determine endothermic and exothermic peak temperatures and mass loss or gain. In the TGA/DSC, each sample was placed in a 90 µ1 alumina pan. A lid was not used during this analysis. The pan and an empty reference for were placed onto the TGA/DSC balance arm in the cell and heated at a preset rate while the heat flow and mass changes, as a function of tempemhae, of fire sample pan and reference pan were recorded. In this way, the differential heat flow and mass change between the sample and the reference pan was measured. When the test material released thermal energy, a positive heat flow (exothermic release of energy) was recorded. The opposite occurred for absorption of energy, loss of thermal energy, (melting). Sample mass loss was retarded directly. From this information, the onset temperature (when the differential heat flow starts to occur) and peek temperature (when no more heat is liberated or released) can be detenNned. FAI18-0752 Pmp.le,ary PmpernvfF w ke & 15swiates,, aC July 31, 2018 -20 of 24- These reactions, in which the material absorbs or generates heat, are phase changes that equate to complete or partial melfiog/vaporisation (endothermic) or reaction (exothermic). Reactions are considered complete when the peak exothermic temperature is reached. The TGA/DSC tests were conducted coder an air environment. Resu/rs Trom TGWDSC Analysis The resuhs of the TGA/DSC tesfing are presented in Figure 9. The results are discussed fi rther in the figure caption. S,,tw tMilfirg saluuiw DST 8789 h1manak RxM1(M Ratei i) Ew Yp Tem e,eN,e T('q Figure 9: TGA/BSC thermograph for the "Maverick Boats" sample. The Total Combustible Content of this sample was estimated to be around 64 wL% FAIIB-0952 Pmprietwy Pmpe.NfFnuske&1smiate,•, uC July 31, 2018 -21 of 24- VOWN I9I19CA5 IDrst Explosibility results indicate that the sample tested had a Ks, value of 201 har-mis; therefore, this dust is classified as an St2 class dust. This material will present am explosion and deflagration hazard risk when dispersed and ignited in air. Secondary explosions may also pose a risk once an initial explosion occurs with the presence of a combustible dust or powder in the area. The Kn and Pm„ values are used in designing plant explosion relief and suppression/ containment systems (see NFPA 68 and 69). Additionally, please consult NFPA 652 and 654 for explosion mitigation guidance and instructions. The WC value for the sample tested was estimated to be 59 g/&. It is important to keep dust concentrations below the MEC in order to mmimiu the risk of a dust explosion. Higher MEC levels reduce the risk of firgitrve dust explosions and explosions in dilute phase pneurvatic transport. Materials which require low concentrations of dust dispersed in sit to create a combustible mixture present a greater explosion hazard. Fugitive dust may also pose a risk once an initial explosion occur. The presence of combustible dust, or powder, m na area acts as an additional fuel for the explosion. It is recommended that concentrations be kept at a minimum (i.e. remove fugitive dust), otherwise explosion venting and inciting or nitrogen suppression may be employed. Please consult NFPA 652 and 654 for guidance and instructions. The results indicate that the sample had an MIE value that is between 1 ad and 3 ml with m interpolated value of 2.1 ad. Most electrostatic ignition sources have energies less Nan 1000 m3. Common electrostatic ignition sion es are typically less thin 30 m3 (goods, small isolated comparisons, era). Grounding and bonding of process equipment should be implemented. The client may also wish to consider inserting the process environment to mitigate an explosion bawd. Please see NFPA 652, 654, 77, 68, and 69 for guidance and instructions. The MITa,, b,e result for the sample tested was at 4d)OC. High temperature process environments should he avoided to minimize the risk of a dust cloud explosion. Please consult NFPA 652, 654, and 70 for guidance and instructions. The TGA SC method demonstrated that the Total Combustible Content of the sample tested was found to be— 64 wt%. Please now that frter dust trader what Process conditions may produce higher risk exolosibility Results from the experimental tests determined that the sample is deemed to he Class II dust for National Electrical Code (NFPA 70) purposes. Please refer to NFPA 499-2017 for further discussion of classification of combustible dusts and hazardous locations for electrical installations in chemical process areas. FA1184)752 P.vivietwy r opem q`F oke & a..s«fates aC July 31, 2018 -22 of24- FAI can assist in interpreting these tests and results and can also facilitate developing explosion mitigation/protection strategies. If you have any further questions regarding the content of this report, please feel free in contact US. Respectfully submitted, Fetishes andAssociates, LLC. Prepared by: Rochelle Andreasen Dust Prail Manager Reviewed and Approved by: Ashok Gurnee Dastida , PhD MBA Vice President, Dust & Flammability Testing and Consulting Services Now Nut the conclusions andrxommrndations in this capon arc Issued on the spxi6c considerations staled and labmamry test meflodologles used These mosidemtlone Include (bon ate net IIMad to) exact ssmple narcosis surd [including article arm disMbution, pancle moRhology, manmrt covert and level of oxidation]; fomwlatlmmpasition heated, conditions of the no, and assumed plant physical puumeets. The conclusions and reascromenclations; may not be applicable for conditions not identicad an those considered Consult local bui ding and Yac roaches, in Winona t0 NPPA 652, 70, 1, 101 anduthroonre a oNPPA nodes, forinaMgione aid gmdeW6 FAI18-0752 Pmpaeray Prope.Nflenuke&a.rseciaex uC July 31, 2018 -23 of24- ADoendu A: Particle Size Analysis Fauske & Associates, LLC AM MISS lll� i a a � Figure Al: Maverick Boats particle sbe distribdou FA118-0752 Pmp.le,aryP pe.NfFwske&A.s«fates uC July 31, 2018 24 of DL ,.4,9FAU S KE DIN a<.xeue r.,,NW: a.ve soraw. nine: IN .pnninnn. urw Figure A2: Maverick Boats particle image portion of barb - non n otism I representation] FA118-0752 Prop xeNy A pert" Ipauske & Maxiates,, UC July 31, 2018