PRB COAL DEGRADATION CAUSES AND CURES

By Roderick J. Hossfeld Jenike & Johanson, Inc. 1 technology Park Drive Westford, MA 0188 !"18# #$8."#%.0"00 &voice' #$8."#%.##80 &fa(' r)hossfeld*)enike.co+ Rod H !! ,oal ,o+-.stion Inc. 11/ 0o.th Main 0treet 1ersailles, 23 &voice' 84#.8$".0188 rod5hatt*coalco+-.stion.co+ Introduction 6he coals 7rod.ced in the Po8der 9iver :asin &P9:' are environ+entally and econo+ically attractive to 7o8er co+7anies. 6his ty7e of coal has +ade enor+o.s inroads to 7o8er 7lants d.e to this regardless of any diffic.lties 7lant 7ersonnel +ight e(7erience. 6he diffic.lties in handling and storing P9: coal are d.e to fines generation and s7ontaneo.s co+-.stion iss.es. Many have 8orked on addressing these concerns and ho8 8e can i+7rove o.r .tili;ation of these f.els. &1!10' &9efer to A77endi( A for so+e -asic -.lk solids handling considerations for P9: coal.' P9: coal is e(tre+ely fria-le and 8ill -reak do8n into s+aller 7articles virt.ally inde7endent of ho8 the coal is trans7orted or handled. P9: re7resents the e(tre+es of handling 7ro-le+s< dust is an iss.e 8hen the coal is fine and dry. :.nker and ch.te plugging is an iss.e 8hen the sa+e fine coal is 8et. =nce the coal is e(7osed -y +ining, the degradation 7rocess -egins > the +a)ority of the da+age can occ.r in a very short ti+e, even as short as a fe8 days. 6he e(tent of the degradation that occ.rs de7ends in large 7art on the distance to the 7lant fro+ the +ine, i.e., ho8 long the coal is e(7osed to the at+os7here d.ring trans7ortation. Additional factors s.ch as cr.shed r.n of +ine &,9=M' si;e, and s7ecific handling 7roced.res also i+7act the degradation 7rocess. Additional deco+7osition occ.rs d.ring handling and storage in a 7ile and -.nker, -in or silo. We -elieve that the root ca.se of the degradation is loss of +oist.re that i+7acts the coal -oth +echanically and thro.gh the generation of additional s.rface reaction area, che+ically. 6he co+-ination of the t8o is 8hat +akes P9: coal so diffic.lt to handle. 6his 7a7er foc.ses on the +echanis+s, -oth inherent and e(ternal, that ca.se this ra7id degradation of the coal 7articles, and 7oses so+e of the ?.estions that co.ld lead to 7reventing, or at least retarding the degradation of the 7articles, and th.s avoid the res.lts of the associated 7ro-le+s 8ith handling and storage, s.ch as d.st, 7erceived loss of inventory, -.nker hang!.7s and es7ecially s7ontaneo.s co+-.stion.

What is PRB coal? The Powder River Basin extends from Wyoming into Southeast Montana, with the bulk of the PRB coal being supplied from the Southern Powder River Basin (Wyoming). PRB is classified by the American Society for Testing and Materials (ASTM) as a sub-bituminous A or B (11)coal. Scientists report that these coals have been burning naturally for over 2 million years. Early Native Americans held these coal fired lands to be spiritual. Prehistoric inhabitants of the PRB used porcellanite as weapons and tools. Procellanite is formed from the intensively coal baked shale or siltstone near or in the coal as it burned. You would know this material as slag. The fires are in part caused by the spontaneous combustion of coal. These same properties show up at the plant as hot coal, fires and explosions. While low in sulfur (0.4 to 1.4 LB/MMBTU), PRB coal is also low in heating value (8,000 to 8,800 BTU/LB, on an as received basis for Southern PRB, with one or two mines in the north going as high as 9,400 BTU/LB). Additionally, its friability results in fines and, when dry, the dust (Fig. 1) increases the explosion hazard potential. On the other hand, this same fine coal can be high in moisture content (Fig. 2), which increases its handling difficulty in equipment. Most of the plants currently burning or converting to burn PRB coal have difficulty with these characteristics.

@ig. 1 P9: coal > AdryB

@ig. % P9: coal > A8etB

PRB coal is relatively low ranked. This means that the coal is relatively young. Specific ASTM ranking is just a laboratory method for drawing a line in the sand to differentiate different types of coals. It is basically describing the geological process of transforming plant material to anthracite. The first phase of coalification (fossilization) is to preserve the plant material from oxidation. This peat moss like material is still basically plant material. The first coal like material formed is lignite, or brown coal. The coalification process basically squeezes out oxygen and water. As the plant material becomes less like wood and more like oil, the pore structure constricts limiting the water retention capacity of a coal Chemically, as the oxygen content decreases the coal becomes more hydrophobic or water repelling (water and oil dont mix). This water retention capacity is measured using the equilibrium moisture test. Sub-bituminous coals like the PRB coals are the next step in the coal ranking system. Then comes the low ranked Bituminous C type coal. This is the ranking of many Illinois Basin coals. The higher ranked Bituminous B and C coals are generally found in the Appalachian coalfields. Most of the coal tests that ASTM has standardized were written around higher ranked bituminous and anthracite coals. The tight pore structures of these coals limited the amount of inherent moisture they could hold. Typically these high ranked coals have equilibrium 2

moistures of 1-10. The first step of determining coal quality in the lab is to air-dry the sample to near equilibrium with the laboratory humidity levels. This is done to minimize any impact on lab results of additional drying or absorption of water from the air. In high rank coals, the moisture lost in the air-drying step is near equivalent to the surface moisture. The residual moisture is that moisture that is still locked up in the coal after air-drying. The higher ranked coals that ASTM standards were based on possess this well-defined split between the airdried or surface moisture and the residual or near equilibrium moisture. This is not the case for low rank coals like PRB coal. The sponge-like or wood like nature of PRB coals make the split between surface moisture and inherent moisture a rather fuzzy line. Most of the quality differences between PRB coal and the higher ranked coal can be explained by understanding that the PRB coal has this looser pore structure and additional moisture retention capacity. PRB coal also has more oxygen chemically bonded to the coal and this make the coal hydrophilic (water liking). This helps explain why the PRB coal is likely to reabsorb water after it has dried and degraded. Self-heating characteristics Spontaneous combustion of coal is a well-known phenomenon, especially with PRB coal. This high-moisture, highly volatile sub-bituminous coal will not only smolder and catch fire while in storage piles at power plants and coal terminals, but has been known to be delivered to a power plant with the rail car or barge partially on fire. An explosive case study [1] was presented at the PRBCUG (Powder River Basin Coal Users Group) Annual Meeting in Houston, March 2003 that is a case in point. The dust in a tripper room ignited, causing a major explosion at Wisconsin Public Service J.P. Pulliam Generating Station in Green Bay Wisconsin in June 1991. At the time, the plant was burning a 50/50 blend of PRB coal and bituminous coal, and a fire existed in one of the coal bunkers. Dust within the atmosphere of the tripper room was ignited by a minor explosion, or puff, within the bunker, which triggered a massive explosion in the tripper room, blowing out the outer building walls and roof. While fires prior to this were not uncommon with bituminous coal in the bunker, this was the first serious dust explosion.

Figure 3. Tripper room after explosion at WPS J. P. Pulliam Generating Station

The coal properties that affect spontaneous combustion risk include: Moisture content of the coal or how much drying and rewetting it has been exposed to. Friability or how much size degradation occurs. Particle size or exposed surface reaction area. Rank, PRB coal contains a high percentage of reactive components that tent to decompose as a coals rank increases to Bituminous and Anthracite. Pyrite concentrations greater than two percent (PRB coal is low in pyrite so at least it is not impacted by this (portions of this section from (12)(Kim 1977) These properties primarily influence the rate of heat generation during the self-heating of coal. Since most of the combustible matter in coal is carbon, when coal stored in an atmospheric environment, the carbon slowly oxidizes to form carbon dioxide and carbon monoxide. PRB coal is also one of the highest hydrogen content coals. The oxidation reaction of hydrogen forms water. The production of both water and carbon gases in the coal will not help the situation. These reactions produce heat; since coal is a relatively good insulator, much of this heat is trapped, increasing both the temperature and the rate of oxidation. Depending on how the coal is stored, heat production may substantially exceed heat loss to the environment, and the coal can self-ignite. Self-heating occurs when the rate of heat generation exceeds the rate of heat dissipation. Two mechanisms contribute to the rate of heat generation, coal oxidation and the adsorption of moisture. The reactivity of coal is a measure of its potential to oxidize when exposed to air. The mechanism of coal oxidation is not completely understood. The coals minimum Self Heating Temperature (SHT) is sometimes used as a relative indication of its reactivity. There are various methods used to determine a coals minimum SHT, but they all require running a test in real time and monitoring the temperature of the coal as any reaction occurs. These tests are typically a relative measure of a coals propensity to self-ignite. In general, a coals reactivity increases with decreasing rank. The moisture content of a coal is also an important parameter in the rate of heat generation of the coal. Drying coal is an endothermic process, in which heat is absorbed, and the temperature of the coal is lowered. The adsorption of moisture on a dry coal surface is an exothermic process, with a heat producing reaction. If it is partially dried during its mining, storage, or processing, coal has the potential to readsorb moisture, thus producing heat. Therefore, the higher the moisture content of the coal, the greater the potential for this to occur. The most dangerous scenario for spontaneous combustion is when wet and dry coal are combined; the interface between wet and dry coal becomes a heat exchanger(13) (Smith and other 1991). If coal is either completely wet or completely dry, the risk is substantially reduced. In general, the moisture content of coal increases with decreasing rank. For example, PRB coal has a higher inherent moisture content than bituminous B type coal. The oxidation of pyretic sulfur is also a heat producing reaction. The heat generated can cause the temperature of the surrounding coal to increase, thus increasing the rate of oxidation. Also, as it oxidizes, the pyretic sulfur expands, causing coal degradation to occur. Reactions involving reactive components of the PRB coal, can also be a source of heat generation. Friability and previous oxidation of the coal are also important factors in the self-heating process. 4

The friability of the coal is a measure of the coals ability to break apart into smaller pieces. This exposes fresh coal surfaces to air and moisture, where oxidation and moisture adsorption can occur. Previous oxidation makes coal more friable. Although the oxidized matter is less reactive, the porous nature of the oxidized coal makes the coal more susceptible to air and water leakage when exposed to higher pressure differentials, such as in a pile or bunker. Causes of degradation @ro+ the ti+e it leaves the +ine, P9: coal starts to degrade. 6he +ost dra+atic res.lt of this can -e fo.nd -y o-serving the to7 s.rface of an o7en railcar delivering P9: coal to the 7lant fro+ the +ine. 6he large 7articles have distinct cracks and 8ill shatter into s+aller 7ieces 8hen dro77ed fro+ a height of only feet. Particles contained dee7er in the -ed of coal 8ithin the railcar do not a77ear to -e si+ilarly affected. 6he root ca.se of this degradation of P9: is the drying and res.ltant cracking and 7article si;e degradation and o(idation . 6here are +any varia-les that are 7otential contri-.tors to attrition of P9: coal once it is e(7osed to air< a+-ient te+7erat.re &heat', +oist.re &addition and loss', co+7action, i+7act &dro7 height', inter7article +otion &d.e to general handling', and ti+e. It is felt that all of these i+7act the total degradation 7rocess, ho8ever the loss of +oist.re a77ears to do+inate the 7rocess. 6he +oist.re that contri-.tes to the 7ro-le+ of s7ontaneo.s co+-.stion co+es fro+ h.+idity and fro+ other P9:. Ce8 P9: added over old P9: see+s to create +ore heat at that interface.6he fine 7articles ty7ically have a higher total +oist.re content co+7ared to the coarse 7articles, d.e to their larger s.rface area 7er .nit vol.+e. 0ee @ig.re / -elo8 for si;e verse s.rface area relationshi7. Test program 6he test 7rogra+ investigated the infl.ence of one varia-le on the degradation 7rocess > time. In a relatively dry environ+ent as ti+e 7roceeded the coal dried o.t. As the coal dried o.t it cracke dand -roke do8n into s+aller 7artices 6he role that 7article si;e 7lays in this effect can -e investigated -y e(7osing large and s+all 7articles to a controlled environ+ent &te+7erat.re and +oist.re' and +onitoring the 8eight loss &+oist.re or volatiles loss' over ti+e. 68o different ty7es of tests 8ere r.n. Test Program 1. 6he first test si+7ly allo8ed large P9: coal 7articles to sit at a+-ient conditions &inside a -.ilding' for days 8hile 7hotos 8ere taken. Test Program 2. The second test placed both fines (-1/4 in. particles) and coarse (3 inch particles) in an environmental chamber at controlled conditions (72F and approximately 45% Relative Humidity, with some excursions) for 16 days.

Results of test program 1 6he first test si+7ly allo8ed large P9: coal 7articles to sit at a+-ient conditions &inside a -.ilding' for days 8hile 7hotos 8ere taken. 6o see the res.lts of this test, refer to the series of 7hotos of one s.ch 7article in @ig.res /a > /f.

@ig.re /a. P9: coal at start of test.

@ig.re /-. P9: coal after 4 ho.rs.

@ig.re /c. P9: coal after % days.

@ig.re /d. P9: coal after / days.

@ig.re /e. P9: coal after

days.

@ig.re /f. P9: coal after

days.

As is evident from the photos, degradation of the coal starts immediately upon exposure to the environment. In fact, cracks started to appear within one hour after the start of the test. Results of test program 2 Three samples of fines (-1/4 in. particles) and three samples of coarse (3 inch) particles were placed on individual trays and placed in an environmental chamber. The temperature was kept constant at 72F and the Relative humidity at approximately 45% RH (with some excursions) for 16 days. The weight of the samples was monitored and recorded over that time period. The temperature of the coal was monitored also, using a thermocouple, but no change in temperature of the samples was noted. It was not anticipated that any heat would be generated because drying of coal (loss of moisture) is an endothermic reaction. It is likely that even if any small amount of heat were generated due to the slight gain in moisture (exothermic reaction) on one of the large particles, the heat would quickly dissipate because the coal surface area was relatively small compared to the environmental room. The effect of moisture addition on heat generation is a good candidate for further study. The results of this test are shown in Figure 4.
Figure 5. Cummulative % Moisture/Volatiles Loss/Gain for PRB Coal exposed to atmosp ere
(#
Cummulative ,eig t loss or gain *%+ and R- *%+

'# 5# &# "# $# %# # !%# !$# !"# # 5# %## %5#

% Relative Humidity -1/4" fines, as rec 'd moisture = 34% 3" particle1, as rec'd moisture = 23% 3" particle2, as rec'd moisture = 12%

$##

$5#

"##

"5#

&##

)lapsed time * ours+

The moisture values were determined by drying small samples at 107C in a forced convection oven and recording the weight change, until no additional weight loss was recorded. The loss in weight of each sample, divided by its original weight before drying, is denoted as the moisture. The fines, with an as received (starting) moisture of 34%, lost the largest percentage of weight (18%) during the first 24 hours compared to the coarse particles. One coarse particle (particle2) with a starting moisture of 23%, lost only 3% of its weight during the first 24 hours and 9% over 16 days; another

coarse particle (particle 1) with a starting moisture of only 12%, actually gained weight (0.75% in 24 hours, with a net gain of 0.3% over 16 days). The moisture content of the samples at the end of the test is essentially the equilibrium moisture. This was determined to be approximately 10% for the fines and 13% for the coarse particles. As shown in Figure 4, the rate of moisture loss (or gain) decreased well before the test ended. For example, both the fines and coarse Particle2 approached their equilibrium moisture content after 7 days. However, Particle1 still had not reached its equilibrium moisture by the end of the test (16 days). Test conclusions =ne of the si( sa+7les, large 7article%, sho8ed an indication of the 7otential for coal to adsor- +ore +oist.re. If +ore +oist.re, i.e., h.+idity, 8ere availa-le, it is likely that 7article% had the 7otential for adsor-ing +ore +oist.re, setting .7 the conditions for self!heating. As discussed previously, the adsorption of moisture on a dry coal surface is an exothermic process, with a heat producing reaction. A Wyo+ing DniversityEWyo+ing 0tate Feological 0.rvey &(((' st.dy fo.nd that larger, partially dried 7articles 7rod.ce heat as they adsor- +oist.re. Go8ever, as 8as the case 8ith o.r tests, the Wyo+ing st.dy dealt +ostly 8ith dry coal, so data on this effect is li+ited. Preventing and/or retarding degradation 0o+e of the sa+e 7roced.res that are follo8ed to +ini+i;e the 7otential for s7ontaneo.s co+-.stion can -e follo8ed to 7revent 7article attrition< Coal Handling Any 7rocess that can kee7 the coal fro+ additional drying Harger slo8er +oving -elts 0horter free falls, es7ecially in o7en air in 8indy conditions

Coal pile 0ealing the 7ile +ini+i;es air ingress and air +ove+ent in the 7ile. 6his also hel7s 7revent +oist.re loss and si;e degradation. Protecting the 7ile fro+ the 8ind. A stee7er slo7e creates greater 8ind resistance, forcing air into the 7ile. Protecting the 7ile fro+ the 8ind &e.g., 8ind screens' +ini+i;es air +ove+ent thro.gh the 7ile. Mini+i;ing dro7 heights to control drying and attrition d.e to i+7act. Mini+i;ing dro7 heights to control drying and attrition d.e to i+7act. Design for a +ass flo8 7attern &see A77endi( A'. Provide an inerting agent or at+os7here &not reco++ended on a nor+al -asis'

Coal bunker

Considerations for further study T"e effec! of #ois!$re ddi!io% o% "e ! &e%er !io% is &ood c %did !e for f$r!"er s!$dy. P r!icles !" ! re ' r!i lly dried co$ld (e s$()ec!ed !o * ryi%& le*els of i%cre sed rel !i*e H$#idi!y+ ,"ile #o%i!ori%& !"e i%!er% l !e#'er !$re of !"e l r&e co l ' r!icle or (ed of co l fi%es. References Bibliography I1J Dantoin, :., 9. J. Gossfeld, 2. H. McAteeK I+7roving @.el Gandling 8ith P9: ,oal -y ,onverting a :.nker fro+ @.nnel @lo8 to Mass @lo8, Presented at the Llectric Po8er %00" ,onference & L(hi-ition P9: ,oal Dsers Fro.7 Ann.al Meeting, Go.ston, 6M, March 4, %00". %. Pri+er on 07ontaneo.s Geating and Pyro7horicity &D=L!GD:2!1081!#/', D.0. De7art+ent of Lnergy, Washington, D.,. %0484 ". ,hakra-orti, 0.2., A:.lk 0olids Gandling > A+erican Llectric Po8erNs ,oal Pile Manage+ent Progra+B, 14&"', 77 /%1!/%8, J.ly!0e7te+-er, 1##4. /. C@PA 840 A9eco++ended Practices for @ire Protection for Llectric Fenerating Plants and Gigh 1oltage Direct ,onverter 0tationsB, %000 ed. &section 4!/.%, :ins, :.nkers, and 0ilos'. 4. C@PA 84 A:oiler and ,o+-.stion 0yste+s Ga;ards ,odeB, %001 ed. &note< s.7ercedes C@PA 840" A0tandard for P.lveri;ed @.el 0yste+sB, 1##$ ed.' . Hy+an, 9o-ert M. and 1olk+er, John L., OPyro7horicity &s7ontaneo.s co+-.stion' of Po8der 9iver :asin coals > considerations for coal-ed +ethane develo7+ent.O ,oal 9e7ort ,901!1, Wyo+ing 0tate Feological 0.rvey, Hara+ie, Wyo+ing, March %001. $. ,raig, D. A., 9. J. Gossfeld< Meas.ring Po8der!@lo8 Pro7erties, ,he+ical Lngineering, 0e7te+-er %00%, 77. /1!/ . 8. ,raig, D. A., 9. J. Gossfeld< 2ee7ing ,oal @lo8ing, World ,oal, Jan.ary %00", 77. 1$!%% #. ,hakra-orti, 0.2., 9. J. Gossfeld, 9. A. Mesing, 2. 0te77ling< Go8 to Achieve 9elia-le ,oal @lo8 and Maintain Plant Availa-ility, Presented at Po8er!Fen International, =range ,o.nty ,onvention ,enter, =rlando, @H, Dece+-er 11, %00%. 10. Gatt, 9od< 0ticky When Wet, World ,oal, A.g.st 1##$ 11. A06M 1ol.+e 04.04 Faseo.s @.elsK ,oal and ,oke 1%. 2i+, A. F. Ha-oratory 0t.dies on 07ontaineo.s ,o+-.stion of ,oal, D0 :.rea. of Mines I, 8$4 , 7 1" 1##$ 1". 0+ith A. ,. and others Harge!scale st.dies of s7ontani.os co+-.stion of coal, D0 :.rea. of Mines 9=I #"/ , 7"0 1##1