evaluation of densified refuse derived fuels for use in ... · pdf fileevaluation of densified...
TRANSCRIPT
EVALUATION OF DENSIFIED REFUSE DERIVED FUELS FOR USE IN PULVERIZED COAL-FIRED
STEAM GENERATORS
NORMAN J. STEVENS and JOHN C. GUILLAUMIN Detroit Edison Company
Detroit, Michigan
ABSTRACT
During the Fall of 1976, the Detroit Edison Company conducted an investigation of the chemical, physical and milling properties of two types of densified refuse derived fuels (d-RDF) blended in various ratios with coal.
This paper discusses these successful tests and the feasibility of preparing a d-RDF which can be processed with coal using existing, unmodified coal handling equipment and fired in conventional pulverized coal-fired steam generators.
THE GENERIC FUEL CONCEPT
A growing number of companies in the United States are currently producing, or claim the capability to produce, a combustible material from processed municipal solid waste. These materials, or refuse derived fuels (RDF), are suitable fuel supplements for utility boilers. Several utilities have burned refuse derived fuels processed from MSW with varying degrees of success. Among these are Wisconsin Electric Power at their Oak Creek Station and Union Electric at their Merimac Station.
However, in both cases, conversion to refuse firing required modification of the boiler, installation of a conveying system and enclosed storage facility, and a pneumatic feed system to propel the RD F to the combustion zone of the boiler [1].
A number of firms have carried the development of refuse derived fuels one step further and
491
produced densified RDF (d-RDF). While the primary interest in d-RDF testing to date has been in its potential as a supplemental fuel for stoker fired boilers [2] , a broader range of applications has been the subject of some discussions [3].
Recent tests of d-RDF by Detroit Edison indicate that d-RDF may also have a broad application as a supplemental fuel for pulverized coalfired boilers and may hold the key to implementing resource recovery technology without modification of existing power plant systems.
The primary objectives of Detroit Edison's study are to determine if a densified refuse derived fuel is capable of being (I) blended and milled with coal, and (2) burned in a pulverized coalfired boiler, without further investment in boiler modifications and handling systems and the associated high cost of boiler outage during system retrofit. Further studies will investigate the emissions occurring during the burning process. If the d-RDF can be further developed for open storage, the principle of a generic fuel supplement for pulverized coal-fired boilers will have been successfully demonstrated.
With increasing public sentiment in favor of recycling solid waste materials and a demonstrated minimal cost to convert an existing plant to refuse fuel firing, a generic coal supplement could provide electric utilities with an attractive supplemental fuel source.
PROGRAM DESCRIPTION
A test program to demonstrate the technical
and economic feasibility of producing a generic
refuse derived fuel supplement for pulverized coal
fired boilers was developed at Detroit Edison in
1975.
Phase I of the test program involved chemical
analyses of d-RDF and coal blends and evaluation
of off-site milling of various blends of d-RDF and
coal. The test procedures and results are presented
in this report. All sampling and analyses were conducted employing ASTM methods, where applica
ble [4]. The first phase of the test program, as discussed
in this report, has been designed to assist in deter
mining the technical feasibility of milling d-RDF
with coal in existing coal pulverizers. Based on
the results of these tests, it will be possible to
determine whether the use of d-RDF has the poten
tial of eliminating the need for pneumatic handling
systems and power plant modifications currently
required for supplemental firing of conventional
RDF.
Three types of d-RDF were initially considered
for the test program as shown in Fig. 1 and
described as follows:
1. Briquettes, 1 in. X 1 h in. X 1 � in. (2.54 cm
X 3.81 cm X 4.445 cm), composed of air-classified,
shredded municipal solid waste chemically treated
and pulverized to particles smaller than 20 mesh.
2. Pellets, 3/16 in. diameter X 1/2 in. (0.476 cm diameter X 1.27 cm), composed of wet process
ed municipal solid waste and sludge, dried to 10.4 percent moisture.
3. Cubes, 1 in. X 114 in. X 13,4 in. (2.54 cm X 3.175 cm X 4.445 cm), composed of 1 in. to 3 in.
(2.54 cm to 7.62 cm) pieces of air-classified,
shredded municipal solid waste. The three d-RDFs will be referred to as cubes,
pellets, and briquettes throughout the remainder
of this report.
The primary intent in selection of the three
types of material was to investigate the acceptability of the RDF and not the shape of the d-RDF or method of densifying. However, each RDF had the
demonstrated capability of being formed into the
associated shape and was readily available as a
d-RDF without further technological development.
Invitations to donate d-RDF for the program
were sent to three d-RDF manufacturers and
accepted by firms producing the pellets and
briquettes. Since the probability of successfully
milling the cubes was considered to be very low,
no additional producers of this material were con
tacted when the original invitation to donate the
cubed d-RDF was rejected.
The tests were conducted using 7.5 percent and
15 percent by weight (approximately 5 percent
and 10 percent by Btu) of each d-RDF blended
with Pittsburgh seam coal. Detroit Edison's Engi
neering Research Department performed chemical
and sizing analyses as required. All analyses were
performed in accordance with ASTM D 271, "Standard Methods of Laboratory Sampling and Analysis of Coal and Coke," and ASTM D 410,
"Standard Method of Sieve Analysis of Coal."
1I1111t1l/IIIIIIIII/lll1iiiliilliillllll!!I!!l I !I'III!j!!'!!! 'n!!!! " '! H!!:! !illIlll lllll\lIII!,!" i' illlll"I'" ' \iiill!\\\\\\ill.lllI\\\\\\\IH\1I
�- I ! I ' !" I"' . I ; I \ \ J 2 a .. 5 6 8 9 )0 1 � 12 13 l.<t. to
FIG.l. THREE TYPES OF DENSIFIED REFUSE DERIVED FUELS (d-RDF)
492
The mill performance portion of the test program was conducted by outside firms under contract to Detroit Edison.
Townsend and Bottum, Inc., a construction contractor headquartered in Ann Arbor, Michigan was contracted to perform the initial grinding tests of the test coal and coal d-RDF blends. Mixtures of dried coal and d-RDF were to be fed to a 372A two-roll Raymond coal pulverizer using air at room temperature at the mill inlet. The pulverizer was set to produce a standard 70 percent minus 200 mesh power plant grind at rated capacity of 3 tons/hr. Observations of mill operation at full load, as well as measurements of production rate, were to be recorded when processing the coal/ d-RDF blends. Samples were to be taken for later sizing and chemical analysis of mill feeds and products.
The Power Generation Group of the Babcock and Wilcox Company, Barberton, Ohio. was contracted to perform further grinding tests on the coal/d-RDF blends. B&W was to arrange with Allis Chalmers to grind the coal and coal/d-RDF blends on an MPS type pulverizer at the Allis Chalmers Process Test Center in Oak Creek , Wisconsin.
The classifier settings on the MPS type pulverizer were to be varied while mill feed rate and mill pressure differential were to be held constant to simulate different mill loadings. These tests were run on a sized, as-received mill feed while controlling the mill outlet temperature to 150 F (66 C). A variety of mill operating parameters were monitored continuously and samples were taken for sizing and chemical analysis.
Plots of mill output and unit energy vs product fineness were to be provided as well as indications of possible explosion or fire hazard when pulverizing d-RDF in a "hot" mill.
Detroit Edison's Engineering Research Department personnel witnessed the grinding tests performed by T &B and Allis Chalmers under this test program.
B&W also was contracted to run Hardgrove and continuous grindability tests as well as burning profIle tests on the fuels and blends at their Alliance Research Center, Alliance, Ohio.
Sufficient quantities of coal and each d-RDF were shipped to each contractor so that representative tests of the coal/d-RDF blends could be conducted.
493
GRINDABILITY TESTS
PROCED U R E
The laboratory scale mill that is used by B&W at its Alliance Research Center to determine continuous grindability indices is shown in Fig. 2. Dried 16 X 30 mesh material is fed into the grinding zone of the mill at approximately 0.7 oz (20 gm) per minute. Pulverized material swept from the grinding zone is gravity-fed through a 1.5 in. (3.81 cm) pipe into an Allen Bradley Sonic Sifter. The pulverized product and the coarse product which requires recycling are each directed to separate compartments at the bottom of the sonic sifter. The product is continuously collected and weighed in a tared polyethylene tray and the coarse coal is collected for periodic introduction into the recycle system.
Mill drive shaft torque is measured by a Lebow reaction-type torque sensor positioned below the mill bowl.
Testing by B&W of selected coal samples over the past years has resulted in a preliminary, empirical correlation of apparent full scale grind ability obtained from field measurements with continuous mill performance. For a coal of unknown
3
u o o o
1 - DRIVE MOTOR
2 - TACHOMETER
e
3 - TOROUE RECORDER
2
4 - TORQUE TRANSDUCER
6 - BALL AND RACE MILL
1
8 - THRUST LOADING WEIGHT
7 - BED DEPTH INDICATOR
8· - RECIRCULATED COAL FEEDER
g - COARSE COAL FEEOER
10 - COAL CONVEYOR
11 - CLASSIFIER
12 - PULVERIZED COAL PROD UCT
13 - SCALE
4
7
10
FIG. 2. SCHEMATIC OF LABORATORY - SCALE CONTINUOUS MILL
9
12
13 I--"i
11
grindability characteristics, testing in the continuous mill and applying the current correlation is intended to provide a grindability index which predicts full scale performance in B&W pulverizers. 'This index is called the Continuous Grindability Index (CGI), having the same units and applied in the same manner as the Hardgrove Grindability Index (HGI). The impetus behind the development of the continuous grindability mill and index was the inability of the Hardgrove method to predict the full scale grindability characteristics of low rank coals. The present correlation is based on data points for a number of specific low rank western coals and this should be kept in mind when applying the results of this method for unusual materials, such as refuse and a coal and refuse mixture.
R ESULTS
The results of grindability determinations on
the coal, d-RDF, and blends are contained in Table 1. The behavior of the coal during grind ability testing was comparable to that expected for this bituminous coal.
TABLE 1 GRINDABILITIES
Sample
Coal
Briquettes
Pellets
l5� Briquettes - 85% Coal
15'1. Pellets - 85% CoaL
GrindabllHy HGI CGr -
59 50
32 56
See Note No. 1 See Note No. 3
63 48
512• See Note No. 3
1. No grinding or new surface area produced.
2. The partially ground, fluffy waste material did not mix well with the 16 X 30 mesh coal.
3. The 16 X 30 mesh feed prepared from the pellets could not be fed into the mill evenly and rapidly plugged the mill outlet and the sonic sifter classifier.
A. B riquettes
The briquettes were readily pulverized and passed through the mill in much the same manner as would be expected for a coal. Since the continuous mill uses mechanical sieving in the classification system, the effect of the d-RDF's lower density on its passage through an air classification system and on the final particle size distribution of the product was not investigated at this time. The higher CGI for the briquettes is a result of the lower torque on the mill drive shaft when grinding
494
the briquettes (approximately one half), and indicates that the briquettes appear easier to grind than the coal alone.
A mixture of 85 percent by weight coal and 15 percent by weight briquettes behaved in essentially the same manner as the coal alone. No material handling problems were observed when handling this blend and both the HGI and CGI values for the coal/d-RDF blend were not significantly different than those of the coal alone.
B. Pellets
The particle size of the pellets could not be scaled down to the 16 X 30 mesh required for testing in the continuous mill without significantly changing its handling properties. It could not be evenly fed into the mill, and rapidly plugged the mill outlet and classifier. Fifteen percent pellets in the coal-pellet blend was sufficient to plug the mill and cause the same problems listed above for the pellets alone. No CGI values were obtained and no HGI value could be obtained for the pellets alone.
BURNING PROFILES
PROCED U R E
A schematic diagram of the apparatus used for determining burning profIles of fuels is shown in Fig. 3 [5]. Briefly, this system permits derivative thermogravimetric analysis (dTGA) of oxidation of a fuel under controlled conditions continuously and simultaneously measuring the rate of weight
BALANCE " �tH�� ISOLATION f--I
���I�W OERIV, COMP,
- ! RATE
WEIGHT
LOSS
BURNING
PROFILE
FURNACE TEMPERATURE
A - P'U.TlNUM SAMP'lE CAUCleLE H - BALANCE tEAM
t _ VYCOR TUIE I _ RID!!A WEIGHT
•
C - TH EAMOCOU'LES � - .. ICAOMETER NUll AOJUSTMfNT
o - SPLIT CHUT FURNACE It - DC 0 T DISPLAC£MfNT
!! - TAIU TO' TA4NSOUC£A
" - TAAE WEIGHT l _ MAGNETIC OAMI"(R
G _ IENOIX FREE-FLEX ,",VOT ,. _ NICHROME SUSI'ENSlO"
0.1:104 INCH OIAMIElER
FIG. 3. SCHEMATIC OF DERIVATIVE THERMOGRAVIMETRIC ANALYSIS SYSTEM
-..
change and the temperature of the heated sample. The Burning ProfIle, which is a graph of the rate of sample weight loss as a function of increasing temperature, is then used to evaluate the combustion of the fuel. In the case of solid fuels, 300 mg of -60 mesh material is spread in a uniform layer across the bottom of the sample crucible and heated at the rate of 27 F /min (15.6 C/min).
30
28 -
26 �
z 24 r--
::::t 22 I-....
C> ::::t 20 c-
-
V) 18 l-V)
9 1 6 I-'i: C> 14 I--
w � 12 e-lL. 0 10 I-w � 8 -<1 a::
6 -
4 -
2 -
� 0
(C) 0 100
I I
(F) 32 200
I
. I
I I I
300 I
600
I I I I I I 500 700 900 1100 (C)
1000 1400 1800 2000 (F)
FURNACE TEMPERATURE
FIG. 4. BURNING PROFILE FOR COAL FROM MONROE PLANT OF DETROIT EDISON
30
28
26
z 24 -
::::t 22 ....
� 20 -
V) V) 9 r :x: Q w � lL. 0 w r <1 a::
18
16
14
12
10
8
6
4
2
o
l-I-l-I--
-
-
-
-
-
-
-
-
I-/,\. (C) 0 100
(F) 32 200
I
300 I
600
I I
500
1000
I I I
700
1400
FURNACE TEMPERATURE
I I
900 1100 (C) I I I
1800 2000 (F)
FIG. 5. BURNING PROFILE FOR MUNICIPAL SOLID WASTE BRIQUETTES
495
Burning profIles for each of the five samples tested are shown in Figs. 4 through 8. The percentage d-RDF and coal refer to percent on a Btu basis and not on a weight basis. Standard burning profIles for coals ranging in rank from Anthracite to Lignite are shown in Fig. 9 [6] .
30
28 -
26 -
z 24 --
::::t 22 -....
C> 20 -::::t
-
V) 18 -V) 9 16 -
'i: 14 -C>
-W � 12 -
lL. 10 0 -
w 8 r -
<1 a::
6 -
4 -
2 -
/.\.J 0 (C) 0 100
(F) 32 200
I I 300 500
600 1000
I i I I I 700 900 1100 (C)
1400 1800 2000 (F)
FURNACE TEMPERATURE
FIG. 6. BURNING PROFILE FOR MUNICIPAL SOLID WASTE PE LLETS
I
I I 700
1400
FURNACE TEMPERATURE
I I 900 1100 (C)
1800 2000 (F)
FIG. 7. BURNING PROFI LE FOR MIXTURE CONTAINING 10 PERCENT MUNICIPAL SOLI D WASTE BRIQUETTES AND 90 PERCENT DETROIT EDISON COAL PREPARED
ON A BTU BASIS
z -
:::;: .....
Cl :::;:
R ESULTS
The Burning Profiles for the briquettes and for the pellets (Figs. 5 and 6) indicate that they should be very easy to burn. The temperature range for start of oxidation [392-482 F (200-250 C)] is low, and the material burns very rapidly, exhibiting a high combustion rate at a relatively low temperature without burnout occurring at a temperature lower than for any coal (Fig. 9).
Burning Profiles for the mixture of briquettes and coal and the mixture of pellets and coal (Figs.
30
28 I-26 I-24 I-22 I-20 I-
-
en 1 8 I-en
g 16 I-� 14 Cl I--
w :;= 1 2 l-t... 0 10 I-w f- 8 l-e:( a:
6 I-4 I-2 I-0
(Cl 0 100 I I
(Fl 32 200
I I I I " i\ [ I 300
600
500
1000
700 I I
1400
FURNACE TEMPERATURE
900 1100 (Cl
1800 2000 (Fl
FIG. 8. BURNING PROFI LE FOR MIXTURE CONTAINING 10 PERCENT MUNICIPAL SOLID WASTE PELLETS AND 90 PERCENT DETROIT EDISON COAL PREPARED ON A
2 0 z
8 I--
::E I 6 I-.....
"
::!' 14 I--
en en 121-9 tj: 101-" -
� 81-
100
BTU BASIS
CODE COAL M -
...... , ...
ANTHRACITE I . I _.- ANTHRACITE I .2 -D- LV BITUMINOUS 0.4 _0_ HV BITUMINOUS 1.5
--- SUB BITUMINOUS 12.0 LIGNITE 32.0
300 500 700 900 FURNACE TEMPERATURE,·C
AS FIRED
VM F C - -4.7 84.6 6.4 77.3 16.4 71.0 35.4 48.0 33.4 33.2 27.2 33.4
1100
FIG. 9. COMPARISON OF BURNING PROFILES FOR COALS OF DIFFERENT RANK
ASH
9.6 15. I 12.2 15. I 21.4 7.4
496
7 and 8) indicate that the effects of the d-RDF addition to the coal is approximately additive. While the burnout temperature of the coal is not significantly altered, the ignition characteristics of the mixtures are improved over that of the coal alone. If the uniformity of a mixture of coal and either of these two d-RDFs can be maintained in transit from the pulverizers to the burners, its combustion should be at least as good as the coal alone. However, since the results of these tests are empirical, it will be necessary to monitor the boiler during firing of coal/d-RDF blends to determine accurately the affect of the d-RDF on combustion.
CHEMICAL ANALYSIS
PROCED U R E
Detailed results of chemical analyses performed on samples of coal, d-RDF and coal/d-RDF blends as well as samples of their respective ashes can be found in Tables 2-4. The values contained in these tables were obtained by analyzing pulverizer products as well as laboratory composites of coal and d-RDF.
Table 2 contains the result of proximate and ultimate analyses. The ultimate (moisture free) analyses were performed employing a Carlo Erba Model 1004 Elemental Analyzer.
Table 3 contains the results of fusion temperature analysis performed on ashed samples of the test materials and blends.
R ESULTS
As can be seen from Table 2, the d-RDFs have relatively low sulfur values which result in the blends of coal and d-RDF having sulfur contents 20 percent lower than coal alone. This is offset, to some extent, by the lower heating value of the d-RDF. The lower heating value of the d-RDF is related to its higher oxygen and lower carbon content. This is characteristic of cellulose-based materials. The higher volatile content of the d-RDF results in its low ignition temperature and low burnout temperature as discussed previously. The effects of blending the coal and d-RDF appear to be additive based on the chemical analysis data provided in this table. The ash fusion temperatures of the d-RDF, Table 3, are considerably lower than that of the coal. While ash fusion temperatures of coal/d-RDF blends are lower than those of coal alone, they do not appear to be sufficiently
TABLE 2 CHEMICAL ANALYSIS - COAL. d -RDF & BLENDS
100% 100% 100% As Fired (Proximate) Coal Brig Pellet
Moisture, percent 4.1 4.0 10.4 Ash, percent 13.9 12.3 16.0 Sulfur, percent 2.5 0.5 0.2 Htg Value, Btu/1b 12,291 7,940 6,680 Volatile, percent 36.2 72.6 70.5
Moisture Free (Ultimate)
Carbon, percent 70.10 46.85 37.84 Hydrogen, percent 4.73 5.73 5.39 Nitrogen, percent 1.40 0.31 1.00 Oxygen, percent 6.81 33.80 37.66 Sulfur, percent 2.56 0.51 0.21 Ash, percent 14.40 12.80 17.90
TABLE 3 FUSION ANALYSIS - AS HED SAMPLE
Samp I e
100% Coa I
I DOl 8r j que t tes
100,(, Pel lets
93.5% Coal
7.54 Sri que ttes
B5% COO I 15:4 Briquettes
92.5% Coo I 7.54 Pellets
85"4 Coal
IS,(,Pellets
Ini tla!
2115 (115))
1770 (966)
1940 (1060)
2065 (1129)
2100 (1149)
2095 (1146)
2065 (1129)
rusion Temperature OF {eel Spher i ca t Hemispherical FI u i d
2255 (12)5) 2400 (1)16) 25)0 (I )88)
2040 (1116) 2140 (1171) 2)40 (1282)
2115 (1157) 2200 (1204) 2)20 (1271)
2220 (1216) 2)40 (1282) 2420 (I )2))
2250 (12)2) 2)BO (I )04) 2410 (1321)
2250 (12)2) 2)BO (1)04) 2420 (I )2))
2220 (1216) 2)BO(I)22) 2400 (1)16)
low to cause fouling or slagging problems in the boiler if blends are kept below 15 percent by weight d-RDF.
Table 4 contains the results of spectrographic analysis of ashed samples of the materials and blends included in this study. These results indicate that the concentration of the various constituents of the ash resulting from the laboratory combustion of these coal/d-RDF blends are not significantly different than those of the coal ash alone.
However, only by monitoring the furnace during firing and inspecting the furnace after burning coal/d-RDF blends can the combustion charactersitics and behavior of the fuel ash be accurately determined.
RAYMOND MILL TESTS
PROCEDU R E
Two coal/d-RDF blends of both briquettes and pellets were ground at the Townsend and Bottum
92.5% Coal 85% Coal 92.5% Coal 85% Coal
497
7.5% Brig 15% Brig 7.5% Pellet 15% Pellet
1.8 1.6 6.4 6.9 15.2 13 .5 12.6 14.7
2.1 1.9 2.1 1.7 12,080 11,950 11,850 11,160
36.7 40.2 38.9 48.0
67.43 65.94 65.49 65.49 4.78 4.94 4.68 4.57 1.31 1.15 1.26 1.18 8.84 12.34 12.94 12.03 2.14 1.93 2.13 1. 73
15.50 13.70 13.50 15.80
TABLE 4 SPECTOGRAPHIC ANALYSIS - AS H ED SAMPLE
Spectrographic Analysis, Percent by Weight Constituents Aa:
FeZ03 A1203 HgO Cao TLOZ Na2
0 K20 51°2
lOOt Coal 17 .8 21.7 1.1 8.' 0.7 0.0 2.0 47.7
lOOt Briquettes 9.5 10.5 1.0 11.5 1.1 1.8 1.0 63.6
lOO'%. Pelleta 10.5 15.5 '.2 18.5 2.1 5.8 0.8 42.6
92.51. Coal 7.5 Briquettes 17 ,0 22.5 0.9 5.2 1.0 1.0 1.2 50.6
85'7. Coal 157. Briquettes 18.0 23.5 I.l 8.3 1.1 1.5 1.0 45.5
92.51 Coal 7.5'. Pellets 18.0 21.0 1.0 8.8 0.9 1.0 1.3 47.4
85'1. Cos 1 15t Pelletl 17.0 20.0 1.' 9.0 1.3 1.4 1.4 48.5
Coal Beneficiation Pilot Plant. These blends contained 7.5 and 15 percent d-RDF by weight and 92.5 and 85 percent coal by weight, respectively. Base load data on the coal had been obtained from prior work. The mixtures of dry coal and briquettes were fed into a 372A two-roll Raymond coal pulverizer using air at room temperature at the mill inlet. The Beneficiation Plant pulverizer is set up so that all the pulverizer product from the pulverizer exhauster discharges into a cyclone separator and is, in turn, vented into a dust collector. Mill product samples are taken in the vertical plane of the 12 in. (30.48 cm) horizontal mill exhauster discharge to the cyclone separator using a cyclone type sampler recommended by the pulverizer manufacturer.
The Raymond pulverizer was set to produce a standard 70 percent minus 200 mesh (75 �m) at rated mill output of 3 tons/hr (2727 kg/h). It should be noted that the T&B mill tests employ a dry mill feed and cold air rather than as-received feed and heated air normally used in standard power plant pulverizers. Samples of mill feed and
product were taken for chemical and sizing analyses.
TABLE 6 SIZING AND DISTRIBUTION TESTS.
BRIQUETTES - COAL BLENDS
RESULTS Sizing Tes ts - Hi 11 Outlet
92.5'1 Coal 8S'%. Coal
The purpose of these initial grinding tests was to determine if coal/d-RDF blends of 5-15 percent d-RDF could be processed by a power plant pulverizer at rated capacity and with the required mill product fineness without changing the normal mill operating settings. Results of these test grinds can be seen in Table 5.
� 7.S't Briguettes 157. Bri.guettes
A. Briquettes
('1)
+ 50 mesh 2.1
- 50. +100 mesh 8.3
--100, +200 mesh 22.9
-200 mesh 66.7
Distribution Tests - Mill Outlet
+ 50 mesh 70
•. 50, +100 mesh 60
-100. +200 mesh 30
-200 mesh 20
92.5% Coal 7.51. Briquet tes
30
40
70
80
(%)
2.1
6.4
17 .0
74.5
8st Coal 157. Briquettes
80 20
60 40
30 70
20 80
Grinding performed using the briquetted refuse and coal mixtures went without incident. Removal of the bowl and the classifier inspection ports after each of these two grinds revealed that the mill internals appeared clean and normal. Table 5 •
reveals that the mill load during operation was normal, the mill product was at or near required fineness, and the production rate was satisfactory. No problems were encountered in the handli�g of the mill feed or mill product. Table 6 contains the results of sizing and distribution tests on the mill product from these tests. As might be expected, the d-RDF is distributed primarily in the larger sized fractions of the mill product. Even though the mill product from the 7.5 percent coal/d-RDF blend was not quite the required fmeness, it was termed satisfactory due to the variance in mill production rate (4 percent over rated output) and possible sampling error.
TABLE 7 SIZING AND DISTRIBUTION TESTS.
Screen
+ 50 mesh
PELLETS - COAL BLENDS
Sizing Tests - Mill Outlet
92.5'. Coal 7.57. Pellets
(%) 2.0
- 50. +l00 mesh 10.4
-100. +200 mesh 18. 8
-200 mesh 68.8
B. Pellets
Inspection of the pulverizer internals after processing the 7.5 percent coal/pellet blend revealed that material build-up was occurring on the classifier spiral and might be expected eventually to plug the internal pulverizer classifier entirely. The level and volume of material remaining in the bowl was much greater than occurred for coal alone and might constitute a fire or explosion hazard in a power plant pulverizer. As can be seen in Table 5,
Distribution Tests - Mill Outlet
Screen
+ 50 mesh
- 50, +100 mesh
-100, +200 mesh
-200 mesh
92.5% Coal 7.5'. Pellets
99
50 50
40 60
20 80
TABLE 5 INITIAL TEST GRINDS - COAL/d-RDF BLENDS
100% Coal
Produc t i on Ra te. Ibs/hr (kg/hr) 6001 (2722)
Pulverizer Current, Amps 78
Production Rate, 106Stu/hr( 106 kj/hr) 74.4 (78.5)
Percent Passing 200 Mesh 78
92.5% Coal Z.S'2 Brlguettes
6173 (2800)
75-78
74.5 (78.6)
66.7
498
85% Coa 1 IS� Briguettes
5395 (2447)
70-74
63.2 (66.zJ
74.5
92.5% Coa I 7.5% Pellets
5952 (2700)
77-79
71.6 (75.5)
68.8
85'. Coal 157. Pellets
('.) 26.4
15.7
15.7
42.2
99
857. Coal 15'. Pellets
50 50
40 60
20 80
85% Coa I 15% Pellets
5432 (2464)
58-78
63. I (66.6)
42.4
the pulverizer did run normally and carried a uniform load on the motor throughout the 7.5 percent coal/d-RDF grind. There was no unusual operating noise during the run. It was also observed that the volume of pulverized pellets and coal leaving the mill increased by a factor of 1.5 to 2 over coal alone and might pose some handling problems.
Several problems were encountered when processing the 15 percent pelletized d-RDF blend. The mill loading cycled during this grind from 58 to 78 Amps. The differential pressure across the pulverizer
bowl also cycled along with the amp loading of the motor. Inspection of the internal classifier revealed that the classifier was plugged for 240 deg. and may have eventually plugged the entire 360 deg. The discharge from the cyclone separator to the dust collector plugged during the test grind due to the excessive amounts of oversize, fibrous material escaping the classifier. While no problems were encountered in handling the mill feed, the mill product that was collected was most difficult to move out of the material storage bins as it "ratholed" and would not move Significantly even when using vibrators and air pads.
Table 7 contains the results of sizing and distribution tests on the mill product from the two test grinds employing the pelletized d-RDF. The d-RDF is distributed primarily in the larger sized fractions of the mill product, and the fineness of the 15 percent d-RDF blend is unsatisfactory.
MPS MILL TESTS
A series of tests were conducted to determine the feasibility of processing the two d-RDFs as a blend with coal in a conventional vertical spindle air swept mill (MPS-32). These tests differ from those conducted on the Raymond Mill in that the MPS-32 has an adjustable rotary vane classifier whose setting can be changed external to the mill. The resulting change in the rpm of the classifier simulates different mill loadings when the mill feed is held constant and the mill pressure differential is held between 8-10 in. H20 (203-254 mm H20).
PROCED U R E
Capacity tests on the MPS-32 were run under the following six conditions:
1. 100 percent feed of coal (baseline data) 2. 92.5 percent coal and 7.5 percent, by weight,
briquetted d-RDF
3. 85 percent coal and 15 percent, by weight, briquetted d-RDF
4. 100 percent briquettes at one load point only
5. 92.5 percent coal and 7.5 percent, by weight, pelletized d-RDF
6. 85 percent coal and 15 percent, by weight, pelletized d-RDF
The test procedure consisted of blending and hammer milling the coal and d-RDF to produce a sized feed (nominally 1/8 in.), filling the weigh feeder with the blend, feeding the mill at a constant feed rate through a screw feeder and weighing the product from the mill. Heated air is provided at the inlet and the temperature measured at that point, inside the mill, and at the mill outlet. Mill outlet temperature is variable and controlled by adjusting the inlet temperature. Test runs were made at various classifier settings to obtain a plot of production rate vs particle size of the product. Unit energy is also plotted vs particle size of product. Although the use of fineness as a direct measure of capacity, particularly in a blend test, can be misleading, it is possible to show the relative capacity limits of the mill if it is operated at
499
the normal mill differential pressure band of 8-10 in. H20 (203-254 mm H20). All essential mill, feed, and product parameters were continuously monitored or determined. A sample test data sheet is shown in Table 8.
R ESULTS
A. Briquettes
Both blends employing the briquetted d-RDF were processed without problems. The base capacity of the mill was not affected significantly (Fig. 10). The straight line in this figure and in subsequent figures is an approximation of the base line condition, 100 percent coal, and is plotted to facilitate interpretation of the data for the blends. It was derived from the test grinds on 100 percent coal. It should be noted that the predicted Hardgrove grindability of the coal was 56 HGI based on the data obtained for the baseline condition. The actual HGI for the test coal is 58.
The unit energy required when processing the blends using briquettes was not significantly affected (Fig. 11).
One data point was obtained on grinding 100 percent briquettes. The only significance of this is to show that the MPS-32 did process the material fairly well. Even though the mill power decreased
TABLE 8 ALLIS-CHALMERS MPS-32 ROLLER MILL TEST
MATERIAL COAL TEST NO. 1 SUBMITTED BY BABCOCK & WILCOX CO.
BARBERTON, OHIO TEST NO. 76-136
MILL TABLE SPEED PRESET SPRING FORCE NO. CLASSIFIER BLADES
= 110 RPM = 442 KG/ROLLER
= 12 CLASSIFIER SHROUD DIAMETER = 320 MM F80, 80 % PASSING OF FEED = 4166 MICRONS FEED MOISTURE (WET BASIS) = 2.30 % LOOSE UNIT WEIGHT = 50.0 LB/FT**3 PACKED UNIT WEIGHT = 61.5 LB/FT*- 3
RUN NUMBER 1
CLASSIFIER, SPEED NO. 200 CLASSIFIER SPEED, RPM 185 MILL POWER, KW GROSS 3.38 MILL POWER, NET AT 76% EFF. 2.57 SAMPLE WEIGHT, LB. (WE,T) 57.7 DURATION OF RUN, MINUTES 5.0 PRODUCT % MOISTURE, WET BASIS 0.7 PRODUCT RATE, KG/HR (DRY) 312.4 NET KW HR/DRY METRIC TON 8.22 PRODUCT, % + 75 MICRONS 19.08 80 % PASSING SIZE, MICRONS 53.0
WORK INDEX 6.12 TEMP. MILL INLET, DEG. F. 190.
DEG. C. 88 TEMP. INSIDE MILL, DEG. F. 11500
DEG. C. 46 TEMP. AT EXIT ORIFICE, DEG. F. 109
DEG. C. 43 DELTA PRES. MILL, H20 (AVE) 10.3
(MIN) 10.00 (MAX) 10.5
DELTA PRES. MILL' MM H20 (AVE.) 261 (MIM) 254 (MAX) 268
DELTA PRES. SOLIDS BED, H2O 5.3 MM H2O 135
DELTA PRES. PORTED RING, H2O 5.0 MM H2O 126
AIR F LOW TO MI LL, if /HR 2154 KG/HR 977
AIR FLOW FROM MILL,1f/HR 2371 00 KG/HR 1075
AIR FLOW FROM MILL, HCFM 577 ACTUAL M**3/HR 981
OUTLET, AIR/SOLIDS MASS RATIO 3.44 INLET, AIR/SOLIDS MASS RATIO 3.13 CAPACITANCE, MICROAMPS 20.12
2 3 4
175 150 125 162 139 116
3.75 4.05 4.05 2.85 3.08 3.08
70.0 80.3 80.9 6.0 5.0 5.0 O. O. 0.2
317.5 437.1 439.5 J
8.98 7.04 7.00 26.48 35.44 49.54 83.0' 90.0 99.0
8.64 6.39 6.35 290.':0 295 275 143 146 135 160 175 170
71 79 77 150 150 155
66 66 68 9.2 8.1 8.1 9.0 7.7 7.3 9.5 8.4 8.2
233 206 206 227 196 186 240 212 208
3.9 2.9 2.7 100 74 69
5.2 5.:? 5.4 133 132 137
2108 2098 2074 956 951 941
2278 2270 2244 1033 1030 1018
-
595 593 591 1011 1007 1004
3.25 2.36 2.32 3.01 2.18 2.14
20.81 20.47 19.79
500
DATE 9-24-76
0.80 KG/LITER 0.99 KG/LITER
5 6
225 200 208 185
3.78 4.13 2.87 3.14
50.6 56.9 5.0 6.0 O. O.
275.4 258.1 10.43 12.16 13.76 23.02 53.0 79.0
9.46 11.03 250 227 121 108 170 180
77 82 160 160
71 71 11.6- 8.8 11.1 8.3 12.0 9.6
296 224 282 210 304 245
6.3 3.1 160 79
5.3 5.7 136 145
1991 2064 903 936
2217 2272 1006 1030
589 603 1000 1025
3.65 3.99 3.28 3.63
20.87 23.69
a:: 1:
by nearly 40 percent and the mill feed was increased 40 percent to hold the required mill pressure differential, the mill did function properly.
Inspection of the mill after the test employing blends of the briquettes as well as 100 percent briquettes showed no hang-up or sticking of the d-RDF at mill temperatures of up to 180 F (82 C). No problems were encountered with fires or explosions during these tests.
B. Pel lets
Mill capacity and unit energy plots for the coal/ d-RDF blends employing the pelletized refuse are shown in Figs. 12 and 13. As can be seen, the pelletized d-RDF could not be handled without re-
-
ducing overall mill capabilities. The blend employing 7.5 percent pelletized
d-RDF was run near the lower limit of normal mill differential levels with the apparent effect of reducing overall mill capabilities. This blend could be handled by the MPS-32.
The blend employing 15 percent pelletized d-RDF could not be handled in the MPS-32 at the normal mill differential range. At a mill pressure differential of 8 in. (203 mm) of water, the mill did not function properly due to the excessive thickness of the grinding bed and subsequent reduction in air flow through the mill. Consequently, the grinding tests were continued at a lower mill pressure differential range of 6-8 in. (152-203 mm) H2 O. This allowed us to grind the 15 percent coal/ d-RDF mixture with significantly reduced mill output and product fineness. Some mill cycling was also experienced (due to changing bed thickness) similar to that observed with the Raymond mill during the test grind of 15 percent pelletized d-RDF/coal blend.
7oo.---------------�-----------,
600
..... 500 C> Q 0 l<: W � 400 a:: Iu
is 300 o a:: Cl.
*0
* 100% BRIQUETTES 0100% COAL II> 9 2.5 % COAL
7.5% BRIQUETTES 85% COAL
Q 15% BRIQUETTES
o
2002LO
---- --
3LO----
4�0
---5�0
--
6J
O--
7�0�80�90�1 00
PERCENT PASSING 200 MESH FIG. 10. MILL PRODUCTION RATE FOR BRIQUETTES/
COAL BLEND RATIOS
501
J: 90 � 80 ::lE � 70
<!l 60 z -
� 50 ct I-Z 40 w u a: w
*
1;1 • 100% 8RIQUETTES o 0100% COAL
92.5% COAL '" 7.5% 8RIQUETTES
85% COAL 1;1 15 % 8RIQUETTES
0.. 30
3L----L---L--L-�-L-L-LJ-��LLLL�20 4 5 6 7 8 9 10 UNIT ENERGY KW-HR/METRIC TON
FIG. 11. MI LL ENERGY REQUIREMENTS FOR BRIQUETTES/COAL B LEND RATIOS
700.---------------_.----------,
600
150L------J-----L---L--��--�� 20 30 40 50 60 70 80 90 100
PERCENT PASSING 200 MESH
FIG. 12. MI LL PRODUCTION RATE FOR PE LLETS/COA L BLEND RATIOS
90 � 80 w ::lE 70 8 N 6 0 C> Z If) 50 If) ct I- 40 z w u a: w
o 100% COAL
II> 92.5 % COAL 7.5% PELLETS
85% COAL [) 15% PELLETS
o
o
'" '"
II> '" '"
o
0.. 30L-__ _L __ _L __ J_��L-�LJ_L�LL��. 3 4 5 6 7 8 9 10 20
UNIT ENERGY KW-HR/METRIC TON
FIG. 13. MILL ENE RGY REQUIREMENTS FOR PELLETS/COAL BLEND RATIOS
DISCUSSION
The demonstrated variability in milling charac- �
teristics is due to the degree to which the municipal solid waste was processed prior to producing the d-RDF. A preference for briquettes or pellets is not indicated, but rather the relative degree of success which can be expected when pulverizing a blend of coal and d-RDF produced from shredded and milled, or shredded and wet processed refuse.
Coal/d-RDF blends containing I S percent by weight briquettes had continuous mill grinding characteristics virtually equivalent to the coal alone . The Continuous Grindability Index (CGI) of the coal is 50, the CGI of the 15 percent coal/ d-RDF blend was 48.
The pelletized d-RDF was not amenable to the size reduction required to prepare it for continuous mill testing. No CGI values could be obtained.
Both the briquettes and the pellets appear relatively easy to burn with low ignition and burnout temperatures . Mixtures containing 1 5 percent by weight of either the briquettes or pellets and 85 percent by weight of coal would have similar burning characteristics to the coal alone. Ignition characteristics of the mixture may be improved as a consequence of the 1 5 percent d-RDF content, as the d-RDF has nearly twice the volatile content of the coal.
While the fusion temperatures of the d-RDFs are lower than those of the coal, the resultant fusion temperatures of the coal/d-RDF blends do not appear to be low enough to seriously effect the boiler. Spectrographic analyses of ashed samples of coal, d-RDFs and coal/RDF blends indicate that while there are differences between the concentrations of some of the constituents in the coal and in the d-RDFs, the ash from coal/d-RDF blends does not differ Significantly from the coal ash alone.
Coal/d-RDF blends containing 7.5 and 1 5 percent by weight briquettes were processed in both a Raymond mill and an MPS type mill without difficulty. The base capacity of the mills and the mill product fineness were not significantly affected.
A coal/d-RDF blend containing 7.5 percent by weight of pellets could be handled by both mills at normal mill differential levels but with an overall reduction in mill capabilities. Blends containing 1 5 percent by weight pellets could not be handled satisfactorily by either mill. Mill operation was quite erratic and mill outputs were reduced significantly. This blend could not be handled at the
502
normal mill differential range. The effect of variations in the percentage of
pelletized d-RDF on mill operation might require that dual feeder/bunker systems be employed to accurately control the blend ratio to minimize its effect on mill operation.
A test grind of 1 00 percent briquettes was made using the MPS-32 mill and the mill did process the material well.
Apparently, the ease with which the briquettes are pulverized is due to the small particle size of the milled refuse (max of 20 mesh) as opposed to the pellet which is composed of shledded wet processed refuse .
CONCLUSIONS
Based on these preliminary tests, it can be concluded that mixtures of coal and pelletized wet processed d-RDF approaching 7.5 percent by weight d-RDF might be handled successfully in a pulverizer. However, this mixture might eventually plug the classifier, affecting the mill product fineness and the fibrous material will hang on ledges, wedge in cracks, etc., and in a power plant pulverizer system might be a potential fire and/or explosion threat.
It is recommended that d-RDF produced from a milled product be considered for further testing. Blends utilizing the briquetted d-RDF are easier to process and the mill product is easier to handle than those employing the pelletized d-RDF produced from shredded, wet-processed refuse and would allow for variations in the blend percentages without forcing significant reduction or loss of mill equipment load. Some differences in mill differential and power over the range of blends, however, would be expected due to differences in bulk density of the materials.
During the actual firing of coal/d-RDF blends in a utility boiler, the refuse and coal may separate and burn in two distinct zones due to the large differences in particle density and volatility. The ash produced in the two zones would then exhibit the individual characteristics of the coal and of the d-RDF rather than that of the coal/d-RDF together. Only by monitoring the furnace during firing and inspecting the furnace after �urning coal/ d-RDF blends can the combustion characteristics and behavior of the fuel ash be accurately determined.
AC KNOWLEDGMENTS
The authors would like to acknowledge the donation of d-RDF by each of the two suppliers and the cooperation and support of their respective staffs in arranging for transportation of the material to the test sites.
Recognition must also be given to Detroit Edison and contractor personnel who were involved during the testing and analyses.
REFERENCES
[ 1 ) Klumb, D. L. and Brendel, P. R., "Solid Waste as a Suoplementary Fuel in Steam-E lectric Generating Plant
Boilers," presented at the International District Heating Association Annual Conference, Saratoga Springs, New York, June, 1 976.
[2) Wiles, C . C., "Densified Refuse Derived Fuels An Alternative Concept," Proceedings of the 1976 Con
ference on Present Status and Research Needs in Energy
Recovery from Wastes, Hueston Woods State Park, Oxford, Ohio.
[3) Hollander, H. I . , "Parametric Considerations in Utilizing Refuse Derived Fuels in Ex isting Boiler Furnaces," Proceedings of the 1976 National Solid Waste
Processing Conference, Boston, Mass. [4) "Gaseou: Fuels; Coal and Coke; Atmospheric
Analysis," Part 26, ASTM Standards. [5) Wagoner, C . L . , and Duzy, A. F . , "Burning
Profiles for Solid Fuels," ASME 67-WA IFU-4.
[ 6) Moore and E hrler, "Western Coals - Laboratory Characterization and Field Evaluations," ASME 73-WA I
FU- 1 .
. Key Words
Analysis
Boiler
Burning
Fuel
Laboratory
Michigan
Refuse Derived Fuel
)
503
Discussion by
Robert A. Olexsey U. S. EPA
Cinci nnati, Oh i o
The authors describe an alternative concept for use of solid waste as a boiler fuel, namely, the combustion of densified Refuse Derived Fuel (d-RDF) in a pulverized coal utility boiler. The paper describes, in great technical detail, tests conducted by and. through Detroit Edison to determine the compatibility of three different types of d-RDF with conventional coal pulverization equipment, that is, the Raymond and MPS Mills.
The authors describe the tests in excellent fashion and with meticulous detail. However, some clarifying comments by the authors with respect to assumptions and conclusions may be in order.
First, the authors stated that one form of d-RDF referred to as the "cube" was considered to have a low propensity for successfully being milled, and, therefore, no testing was conducted with this material. However, the authors present no data nor qualifying statements to support this assumption. Discarding this "cube" material is a very significant decision since this material is the product that appears to be the type of d-RDF most suitably produced at a "conventional" RDF plant. Therefore, ruling out this material consigns us to working in a very specialized product field.
Second, while the test protocol is otherwise described excellently, I could not find a statement
as to how much coal and how much d-RDF was processed in the testing program. Also, the duration of the pulverization tests is not described. Such information would be extremely helpful in {
assessing the reliability of the data presented. To someone as unfamiliar with ASTM grindability tests as myself, a little more detail on quantities is desirable.
Third,_ the authors, in describing the Chemical Analysis Results, make a statement that could be interpreted as meaning that blends of d-RDF and coal must be kept below 15 percent if fouling and slagging are to be avoided. Since the d-RDF content in the tests described never exceeded 1 5 weight percent, the point at which slagging or fouling will occur cannot really be determined from this data. We must be careful lest the 1 5 percent figure becomes engraved in stone as a sacred point not to be exceeded. Further testing with higher percentages of d-RDF and preferably, as the authors note, in a boiler, is necessary to establish parameter limits for combustion performance.
Finally, perhaps, the conclusions should be tempered with a statement that the data presented is applicable for one type of d-RDF, one coal, and the specific pulverization hardware employed. Extr�polating such data to the general case is risky.
Overall, the'paper is very good. It presents new data, something that does not always occur in discussions of resource recovery. It also addresses a new concept for waste utilization.
504