ENVIRONMENTAL EMISSIONS FROM A

SUSPENSION FIRED BOILER WHILE BURNING

REFUSE DERIVED FUEL AND COAL MIXTURES

JERRY L. HALL*t, GARY A. SEVERNS*, HOWARD R. SHAN KSt, ALFRED W. JOENSEN*, and

DELMAR B. VAN METER* Engineering Research Institute and Mechanical Engineering Department

Iowa State University* and Ames Laboratory (U.S. Department of Energyt

Ames, Iowa

ABSTRACT

Results are presented of the emission evalua· tions of a suspension fired steam generator operat· ing at the Ames, Iowa power plant while burning mixtures of refuse derived fuel (RDF) and coal. The facilities, test design and sampling procedure are summarized and emission results are given. Emissions of uncontrolled particulates, chlorides, trace elements of copper, lead and zinc all increase consistently with increases in RDF as expected for the composite fuel analysis. Emissions of sulfur oxides and nitrogen oxides decrease with increases in RDF. No discernible trends, within the data scatter, were noted concerning formaldehyde, hydrocarbons, and many of the 19 different trace elements scanned during these experiments. Combustible and noncombustible characteristics of the boiler grate ash became more like the cor· responding flyash characteristics as the amount of RDF was increased. The test data from the City of Ames studies show that the particulate emis­sions are capable of being controlled within al· lowable compliance rates provided the dust col· lectors are suitably designed and operated.

INTRODUCTION

The City of Ames, Iowa has been commercially operating a system for energy and materials reo covery from municipal solid waste (MSW) since November 1975 . The Ames solid waste recovery system consists of a 150 ton/day (136 tid) refuse

processing plant, a 5 00 ton (454 t) Atlas processed refuse storage bin, and the existing Ames municipal power plant. Air classified shredded refuse is pro· duced for use as a supplementary fuel with coal. Ferrous and nonferrous metals are also recovered.

Evaluations of the refuse processing plant and the power plant have been conducted since Feb­ruary, 1976, and are a cooperative venture among the City of Ames, EPA, DOE's Ames Laboratory, Iowa State University, and Midwest Research Institute. Since the Ames system was designed primarily based on the St. Louis experience, these evaluations offered the opportunity to fill tech· nological data gaps as well as confirm selected observations from the St. Louis studies.

Operatio� of the recovery system included burning of the refuse derived fuel (RDF) as a sup· plementary fuel with coal in a suspension·fired steam generator. The supplementary burning of the RDF in stoker boilers was to occur during shutdown of the pulverized coal unit. However, attempts at firing RDF in the existing Ames suspen· sion system revealed that high dropout of unburn· ed wood, cardboard, and large paper prevented continuous burning. Therefore, the suspension· fued bbiler was modified by the addition of bot· tom grates to allow continuous co·firing of RDF and coal. RDF is now burned in the suspension· fired boiler on a regular basis.

497

FACILITIES

Municipal solid waste dumped on the tipping floor of the refuse processing plant is reduced to a

nominal 6 in. (152 mm) size by the first stage shredder. Mter passing the magnetic separator for ferrous removal, the shredded refuse is further reduced in size to a nominal 1.5 in. (38 mm). The shredded material is then air classified into a light combustible fraction (RDF) and the heavy rejects. The heavies are then subjected to further removal of ferrous and nonferrous material.

The RDF is pneumatically conveyed to Unit 7 through either of two 8 in. (203 mm) transport pipelines at an average rate of 4-9 tons/hr (1.0-2.1 kg/s). Injection of the RDF into the furnace of the unit was at a location directly between the pulverized coal injection nozzles except for four runs at 100 percent load.

Table 1 includes descriptive characteristics of the suspension-fired steam generator (Unit No. 7) used for the tests. The unit has a wet bottom ash removal system. An electrostatic precipitator is the type of air pollution control equipment currently installed on this unit.

EXPERIMENTAL DESIGN

For this study it was determined that two major factors could be controlled at various levels. These factors were the steam generator load based on either steam flow generated (or megawatts of power generated) and the amount of RDF based on heat energy input to the boiler. The levels of these factors were chosen to be 60, 80 or 100 per­cent nominal steam generator load, and 0, 10 or 20 percent RDF. To obtain sufficient data for statistical analysis, a factorial experimental design with three replications was devised for the steam generators as summarized in Table 2. Thus the statistical design was 'a 3 x 3 x 3 (3 loads, 3 values of RDF and 3 replications) full factorial experi­ment with 27 runs required to fill the data matrix of this experiment. Additional miscellaneous test­ing was accomplished on this steam generator for purposes of ascertaining compliance with Iowa Department of Environmental Quality rules. Dur­ing these tests the location of the RDF injection point was changed compared to the EPA designed experimental runs. The results of all tests concern­ing this steam generator are contained herein.

To satisfy the objectives of the environmental emissions study, all appropriate input and output streams associated with the operation of the steam generator unit were sampled. A block diagram showing the sample locations of both entering and leaving streams is included as Fig. 1. The tests ac-

complished concerning the environmental study on Unit 7 are summarized in the data matrix for­mat of Table 2 as previously indicated. All inputs to and outputs from the steam generator were evaluated including fuel, combustion air, bottom ash, steam, fly ash and stack gas.

A block diagram showing the sample locations and types of samples collected is included in Fig. 1 . All of the sampling was conducted on a regular basis except the organic species which were sampl­ed on intermittent days as manpower, instrumenta­tion, and equipment would allow.

TESTING AND SAMPLING PROCEDURES

Sampling of effluents was conducted according to EPA prescribed techniques [I ,2,3] . Stack particulate samples were obtained at numerous prescribed points in the smoke stack cross section as shown in Fig. 2. In addition, three stack sampl­ing trains operated Simultaneously while an addi­tional particulate sampling train was located before the particulate collector. Figs. 3 and 4 show the sampling location and sampling points at the col­lector inlet. Input fuel and grate ash were sampled at regular 30 min. intervals throughout the test period, and then mixed to yield a composite sample. Hopper (fly) ash was sampled at the completion of each experimental run. Combustion air to the boiler was monitored by wet and dry bulb thermometry. Steam flow rate, temperature and pressure were also recorded at regular intervals.

The composite coal and RDF samples were analyzed by the Ames Laboratory. Ultimate anal­yses and heating values were obtained by standard ASTM methods. Trace elements in the fuels and ash were determined by x-ray fluorescence [4] (XRF) techniques. Trace elements in the sample train impinger solutions were determined by induc­tively coupled plasma [5] (ICP) techniques.

The size distribution of the particulates after the dust collector was determined via an Andersen type cascade impactor. Particulate samples obtain­ed with the EPA Method 5 sampling train were analyzed via XRF. The impinger solutions front the Method 5 train were analyzed via ICP.

The gases CO2, CO, °2, and N2 in the stack were determined via Orsat techniques. EPA Meth­od 7 was used for evaluation of NO x levels and the EPA Method 6 was used for measurement of S02 . C1 through Cs hydrocarbons were determined by gas chromatography. Several modifications [6,7]

498

TABLE 1 CHARACTERISTICS OF AMES MUNICIPAL POWER PLANT UNIT 7 SUSPENSION STEAM GENERATOR

Manufacturer

Electrical output - MW

Installation date

Pressure temperature

kPa/oC

(psi/ OF)

,

Nominal steam output capacity

kg/h (kg/s)

(lb/hr)

Coal firing equipment

Furnace pressure

Dust collection Equipment

Stack height

meters

(feet)

Heat input at nominal •

capac1ty

MJ/h (kJ/s)

(BTU x 106

/hr)

Coal fired, TPH @ 9540 BTU/lb

Combustion Engineering

35

1968

6205/481

(900/900)

163,600 (45.4)

(360,000)

2 pulverizers

8 nontilting tangential burners

Balanced Draft

American Standard

electrostatic precipitator

61

(200)

460 (0.13)

( 436)

22.9

RDF capability, TPH @ 5000 BTU/lb

(kg/s @ 11,629 kJ/kg)

with 10% and 20% of total 4.4 @ 10% (1.1 @ 10%)

fuel input being RDF 8.7 @ 20% (2.2 @ 20%)

499

ALYSIS FLOW RATE ULTIMATE AN HEATING VAL CHEMICAL AN

UE ALYSIS LEMENTS

TABLE 2 TEST MATRIX FOR UNIT 7 EXPERIMENTAL RUNS

Percent RDF Percent

Load 0% 10% 20%

60% 3 runs --- ---

(1976)

80% 3 runs 2 test runs ---

(1976) (1977)

100% 3 runs --- ---

(1976)

80% 3 runs 3 runs 3 runs (Wyoming (1978) (1978) (1978)

Coal)

100% 3 runs 3 runs 3 runs (Wyoming (1978) (1978) (1978)

Coal)

100% --- --- compliance tests* (Wyoming 4 runs

Coal) (1978)

* RDF injection nozzles relocated to be below the coal injection nozzle.

FILTER PARTICULATE TRACE ELEMENTS

IMP INGER WATER TRACE ELEMENTS

EMISSION RATES OF

PARTICULATE TRACE ELEMENTS

IMP INGER WATER TRACE ELEMENTS

EMISSION RATES OF PARTICULATE & TRACE E ASH SOFTENI NG TEMPERATURE PARTICULATE AND GASEOUS SPECIES

HUMIDITY BAROMETER INTAKE COAL

\ PARTICULATE SIZING

•

,

TEMPERATUR E � AIR / VOLUME FLOW DENSITY ULTIMATE ANA HEATING VALU CHEMICAL ANA

TRACE ELEM ASH SOFTENIN

TEMPERATUR

BOILER • COLLECTOR RDF UNIT STACK \

LYSIS E LYSIS & HOPPER , ENTS ASH " G E FEEDWATER GRATE STEAM -ASH \ FLOW RATE

CHEMICAL ANALYSIS & TRACE ELEMENTS

• SOFTENING TEMPERATURE TEMPERATURE FLOW RATE FLOW RATE FLOW RATE CHEMICAL ANALYSIS & TEMPERATURE

TRACE ELEMENTS PRESSURE SOFTENING TEMPERATURE

FIG.1 SAMPLING LOCATIONS OF ENTERING AND LEAVING STREAMS

5 00

SAMPLING PLATFORM

POINT

A

B

C

0

E

F

IN ,

14.22

24.67

31. 78

37.63

. 42.65

47.25

SAMPLING POINTS 4 In. PORT,

v-.:.c ? -- CAPPED WHEN

NOT SAMPLI NG

WALL

NOT TO SCALE

SAMPLING POINTS

RADIUS POINT

RADIUS

m IN m

0.361 G 51. 32 1.303

0.626 H 55.03 1. 398

0.807 1 58.67 1 .490

0.956 J 62.02 1. 575

l.OB3 K 65.22 1.656

1.200 L 68.15 1. 731

FIG.2 STACK SAMPLING POINTS

RESISTIVITY

PROBE PORT

GAS FLOW •

EXTERIOR

POWER PLANT---...

WALL

• NOT TO SCALE

VANE

UNIT 7 PORT LOCATIONS •

8EFORE PARTICULATE COLLECTOR

,

, •

CROSS SECTION A-A 4ft x 20 ft

SAMPLING PORTS

'-'------ --

\ "'-l._ VANES \

TO LS.P. •

FIG. 3 SAMPLING LOCATIONS AT THE COLLECTOR INLET

o 0

o 0

UNIT 7 SAMPLING POINTS BEFORE THE PARTICULATE COLLECTOR

o o

o o

o 0 0 0 0

o 0 0 0 0

o 0

o 0

o

o b C

o 0 0 0 0 0 0 0 0 0 0 0 0 b

0 0 0 0 0 0 0 0 0 0 0 0 0 .

OIMENSIONS

IN m IN m .

A 9.25 0.235 • 6.0 0.152

B lS.5 0.470 b 12.0 0.305

C 240.0 6.096 C 4S.0 1. 219

FIG.4 COLLECTOR SAMPLING POINTS

501

of the EPA Method 5 train were used to collect samples for analysis of aldehydes and ketones, chlorides and trace elements.

Table 3 is 11 summary of the sampling methods used for the tests of this study.

During a given test, samples were obtained at all locations previously shown in Fig. 1. The input fuel and boiler load were held as constant as possi­ble at their preselected nominal values during a given test run. Each test was typically 4-5 nr in length if no difficulties were encountered. On some occasions tests lasted for more time than the nominal because of various difficulties that arose concerning sampling, boiler operations of refuse feed to the boiler. Also, on some occasions, back­to-back tests were accomplished during a given day. However, such back -to-back tests were those accomplished for compliance testing as required and monitored by the Iowa Department of Environ­mental Quality. Input coal and RDF were sampled at regular � hr intervals throughout the testing period on any given day. These samples were then mixed to yield an appropriate composite sample for the given test.

The stack effluents were generally sampled ac­cording to EPA prescribed techniques [1,2,3] . The preparation of each sample train include: 1. clean­ing of sample train glassware with an appropriate acid wash followed by several distilled water rinses; 2. preparation of chemicals and loading of sample train impingers with appropriate absorbing solutions; 3. weighing and labelling each impinger of each sample train; 4. weighing and loading particulate train filters; 5. checking sample boxes, control boxes, and sample probes for proper opera­tions; and, 6. transporting the sampling equipment to the test site and preparing for the sampling test as scheduled with the City of Ames personnel.

Before operation of any sampling train, leak checks were performed from the sampling train probe tip to insure that such sample train was leak free. No sampling train was operating until it had passed an appropriate leak check with a maximum

allowable leak rate of 0.02 cfm (O.OO l1/s) at a sampling train vacuum of 15 in. of mercury (50.7 kPa). This leak rate was also allowed at a lower vacuum if the vacuum level of the leak check was not exceeded during operation of a particular sampling train during a test. All sample train leak checks during this study complied with this leak rate criterion. After each experimental run, a leak check was also performed to insure that no leaks had developed during the sampling period.

TAB LE 3 SUMMARY OF SAMPLING AND ANALYTICAL PROCEDURES

Item Sampled

Fuel into boiler Coal

• RDF

Combustion · Air

Ash Grate (bottom ash) Collecteq fly ash (hopper ash)

Steam

Fly ash Before particu­late collector Stack

Flue gas · Stack

Sampling Techniques

Hourly samples combined to form single components sample

Thermometry

Grab samples com­bined to form single composite sample

EPA Method S Brinks impactor

Andersen impactor Brinks impactor

Orsat EPA Method 7 EPA Method 6 Modified EPA

Method S Modified EPA

Method S

Modified EPA Hethods S & 6

Modified EPA Method S

Grab sample

Aspiration across column of macro­reticular resin

502

Analytical Procedure

• Trace elements via XRF Ash via ASTM method Moisture via ASTM method HHV via ASTM method

. Wet bulb and dry bulb temperature

Trace elements via XRF or appro­priate ASTM method Ash via ASTM method Moisture via ASTM method HHV via ASTM method Trace elements in grab sample compo­site taken from particulate collector hopper ash

Flow rate Temperature Pressure

Trace elements in material on filter via XRF

Size distribution via impactor Trace elements in impinger solutions via rcp

CO2' CO, O2, N2 NO SOx

A1�ehydes and ketones via method of Carotti and Kaiser Organic acid via ion chromatograph Cyanide via ion selective electrode Phosphorus via spectrophotometer

Chlorides via colorimetric and spectrophotometric analysis of part of impinger solutions from SO train

x

Mercury, arsenic, antimony beryllium via rcp C] and Cs hydrocarbons via cfiromatography

and

gas

Organics (PCB, POM, etc.) via gas chromatograph-mass spectrograph

TIle samples were obtained either isokinetically or proportionally as required, but only after ap­propriate calibrations of the sample train dry gas meter, flow orifice, pilot tube, sampling nozzle and the temperature indicators on the sampling train had been properly ascertained.

After sampling, the sample train impingers and filters were weighed and prepared for laboratory analysis as necessary. The impinger chemicals were transferred to clean reagents bottles, the filters were desiccated to dryness before weighing, and the coal, RDF, and ash samples were appropriately mixed and processed prior to chemical analysis. All of the samples were then transferred to the various analytical groups of the Ames Laboratory-DOE for analysis.

RESULTS

Most of the results are reported in a data matrix format that indicates the control factor of load and per cent RDF on the experimental runs accomplish­ed. The results for each of the three runs are tabula­ted as a cell average and cell standard deviation in the various tables presented in this section of the results. Each data point on the various plots in­cluded in this section of the report is the cell average. Thus, each data point represents the average of three experimental runs unless other­wise noted. It should also be noted that two stand­ard deviations on each side of the cell average would apprOximately correspond to the 95 percent statistical confidence band. This means that there would be about 95 percent probability that the actual value of the measured variable would fall within the range of two standard deviations on each side of the cell average.

Curves and lines drawn through the data points contained on the plots are meant to show general trends in the data. Such curves are not to imply any particular mathematical expression which may (or may not) govern the behavior of a particular measured variable.

AIR, FEEDWATER AND STEAM CHAR­

ACTERISTICS FOR THE EXPERIMENTAL

RUNS

Table 5 includes the average air, feedwater, and steam characteristics for the experimental runs per­formed for the steam generator Unit 7. Both the average and standard deviations are given for each cell, where each cell represents a given nominal

turbine load and a given percent of RDF input on a heat energy basis. From this table it can be ob­served that the steam load is also tabulated on a percent basis which corresponds approximately to a given percent turbine load. The actual RDF heat energy input tabulated shows some slight variation from the nominal values desired for the test set-up. The variation in the amount of heat energy input is indicative of the amount of control of that particular variable experienced during the operation of Unit No.7.

The characterization of the air entering the boiler unit for combustion purposes is given by wet bulb and dry bulb measurements which have been used to determine the relative humidity of the the air entering the boiler. The barometric pres­sure of the air at that turbine location is also tab­ulated. The feedwater temperature is tabulated as well as the pressure and temperature of the steam which is generated in the boiler unit for use in the turbine. The flue gas or combustion products being exhausted from the boiler into the stack are given in terms of flow rate, temperature and pressure as measured at the sampling location in the smoke­stack of the Unit 7 steam generator complex.

CHARACTERISTICS OF INPUT FUELS

The average of heating values and ultimate analysis constit.uents, as well as their standard deviation for both the coal and RDF used during the tests on Unit 7 are listed in Table 6. It should be noted that the ash content of the RDF is higher than that of the coal used during 1978 and that both the heating values and the amount of sulfur in the RDF is lower than that of the coal for the comparison runs made during this study. The significance of these observations is that as the amount of refuse used in the boiler unit is in­creased, an increased amount of ash will be generat­ed due to using refuse. The additional amount of ash was expected to show up partially as fly ash and partially as bottom ash. Consequen tly, as the RDF increased, the amount of particulate emis­sions was expected to increase. This was also in agreement with the previous data obtained on travelling grate stoker Units 5 and 6 during 1976 and 1977 studies on the travelling grate units.

Because the sulfur content of the RDF was lower than in the coal, it was also expected that the oxides and sulfur emitted from the smoke­stack would decrease Significantly with increases in RDF.

503

TABLE 4 SUMMARY OF SAMPLING TRAIN CHEMICALS AND SAMPLING FLOW RATES

Sample Train

Particulate (EPA Method 5 )

Oxides of Sulfur (EPA Method 6 modified and with 5 impingers)

Organic Acid (EPA Method 5 modified with Tenal< plug*)

Aldehyde and Ketone (EPA Method 5 modified)

Trace Element (EPA Method 5 modified with midget impingers)

Impinger Number and Solution or Material

1

200 mL distilled H

20

2

200 mI, distilled H

20

100 mL 100 mL 8 0 % Isoproponal 3% H

20

2

100 mL NaOH

empty or dry

25 mL NaOH

100 rnL NaOH

100 mL NaHS0

4

25 mL ICL

3

empty or

dry

4

200 g­Drierite

100 mt 100 mt 200 g

3% H2

02

NaOH Drierite

100 mL 200 g: NaOH Drierite

100 mL 200 g-NaHS0

4 Drierite

25 mL 10 g. ICL Drierite

* Tenal< plug used to absorb organic vapors from sampled flue gas.

Typical Flow Rate

i/min

Isokinetic

8

8

1

1

Typical Sampling

Time, min

144

30

60

60

60

TABLE 5 AVERAGE AIR, FE EDWATER. AND STEAM CHARACTERISTICS FOR EXPERIM ENTAL

RUNS ON BOI L ER UNIT 7

Nominal Turbine Load (%)

60

80

80

80

80

100

100

100

100

100

RDF (%) Year

° 197 6

o 1976

o 1978

10 1978

20 1978

o 1976

° 1978

10 197 7

10 1978

20 1978

Steama

Load (%)

RDF Heat Input

(%)

Air Wet Bulb

°c

Dry Bulb

°c

� 4 5 .4 0 . 00 12.0 5 .

a 1.9 0.00 6.6 5 .

x 65.2 0.00 3.0 -1.

a 0.7 0. 00 5 .0 4.

x 64.7 0.00 22.0 28 .

a 1.6 0.00 5 .0 4 .

x 65.3 12. 72 19.0 23.

a 0.5 1.4 3 3 . 0 2.

x

a

x

a

x

63 .8

0.8

86.7

1.5

88.5

18 . 9 6 18.0 23.

5 .27 2.0 5 .

0.00 04.0 -7 .

0.00 11.0 9 .

0. 00 23.0 28.

a 1.3 0.00 3 . 0 3 .

x 88 . 6 10.07 21.0 15 .

a 1.6 0.17 1.0 2.

x 8 6. 9 13.52 22.0 27.

a 1.2 1.95 2.0 2.

x 8 8 .6 20.80 23.0 28 .

a 1.6 1. 18 3 .0 3 .

ReI Hum (%)

3 3 .

1.

4 7.

10 .

Bar Pres

kPa

Feed Water Temp

°c

9 7.20 18 1.

0.44 1.

9 7 .80 195 .

0.78 2.

Steam Temp

°c

4 7 9 .

2.

4 7 6.

2.

60. 9 7.68 197 . 4 7 6.

8. . 0.58 2. 2.

69.

17.

5 7 .

12.

4 3 .

4 .

64 .

6.

5 3 .

10.

64 .

10.

62.

9 .

98.09 196.

0 .44 1.

98.09 194 .

0 . 30 1.

9 8 .00 206.

0 .7 5 2.

9 7 .68 209.

0 . 58 1.

9 8 . 00 209 .

0 . 00 2.

97.92 209.

0 .14 2.

9 7 .68 209 .

0.20 2.

4 77.

1.

4 7 6.

1.

480.

1.

486.

2.

4 77.

2.

4 79 .

3 .

485 .

4 .

Steam Pres

kPa

58 67.

4 .

5888.

18 .

604 8 .

14 5 .

6151.

21.

6034.

152.

5929.

17.

5896.

34 .

5895.

Stack Flow CMS

stp

Stack Temp

oK

3 0 . 08 4 34 .

0.22 4 .

3 7. 62 4 4 1 .

2.85 9 .

38 .79 454 .

2.31 4 .

3 7.76 4 68 .

0 . 59 3 .

3 7 .9 1 481.

0.38 4 .

50 .75 4 5 3 .

0. 18 7.

4 5 .44 4 69 .

0 . 9 6 5 .

4 1.96 4 4 9 .

O. 11. 15 4 .

5 9 17 . 42.92 4 68 .

4 1. 2.34 6.

6234 .

4 8 .

4 3 .53 4 7 9 .

1.59 13 .

a Re 163,600 kg steam/h

b x and a represent the mean and + the standard deviation, respectively

504

Stack Pres

kPa

96.89

0.45

9 7.40

0 .76

97 . 0 6

0 . 5 7

9 7 . 4 6

0 . 4 6

9 7 .4 7

0.29

9 7 .68

0 .61

9 7 .34

0 . 4 4

9 7 . 3 8

0 . 01

9 7 .33

0.10

97. 11

0.18

TABL E 6 VALUES OF COAL AND RDF CHARACTERISTICS AS FIRED

Quantity Coal RDF

Number of Samples 1976

mean Std Dev

(x) Number of Samples

Heating Value (HHV) kJ/g 22.42

(BTU/1b) (9648

Moisture (%) 16.67

Ash (%) 12.98

Carbon (%) 53.96

Hydrogen (%) 3.42

Sulfur (%) 3.27

Chlorine (%) 0.03

Oxygen (%) 9.66

CHARACTERISTICS OF BOTTOM

AND FLY ASH

(0) 14

0.13

57)

3.91

2.30

2.81

0.65

0.85

0.01

1.52

Tables 7 and 8 are tabulations of the combusti­ble and noncombustible constituents of the bot­tom ash and the fly ash respectively from Unit 7. The interesting feature of this data is that the char­acterizations of fly ash and bottom ash in terms of mineral matter and carbon are Significantly dif­ferent when coal is the only fuel burned in the boiler. However, as the fuel mixture changes from 0 - 20 percent RDF, the characteristics of the fly ash and bottom ash become nearly identical.

Another feature of the data in Tables 7 and 8 is that tabulations are given for operation of Unit 7 both before bottom dump grates were installed and after the bottom dump grates were installed . A comparison of this data shows that the bottom grates were highly Significant in retaining unburn­ed combustible material in the furnace region so that the combustible portion of the RDF could continue to burn once it has fallen through the

505

1978 1978

mean Std Dev mean Std Dev

(x) (0) (i) (0) 18 12

23.6 0.52 13.02 0.83

(10156 224) (5602 359)

16.6 1.2 24.04 3.06

9.74 2.23 13.09 2.72

56.6 1.5 30.66 2.92

'4.01 0.19 4.51 0.44

2.79 0.81 0.32 0.05

0.21 0.12 0.35 0.15

9.08 2.26 27.04 2.87

suspension zone to the grate of the boiler. Thus, the dump grates were instrumental in helping derive the available heat energy from the fractions of RDF that had fallen through the suspension zone of the boiler before complete combustion had occurred. For example, with 1 0 percent RDF at 1 00 percent load as much as 35.4 percent of the ash was carbon dropping to the bottom pit of the boiler and being lost to the system before the installation of the dump grates. After the installa­tion of the dump grates the loss of carbon amount­ed to about 1.5 percent of the ash. A similar char­acteristic shows up for the carbon in the fly ash but the degree of this loss in carbon is much less than with the bottom (grate) ash.

Another interesting feature of the data is that the mineral ash content of both the bottom and fly ash became more nearly the same (on the order' of 97 percent) following installation of the dump grates. Prior to the dump grate installation the per­centage of mineral content in the bottom ash was significantly lower, mainly because of the add i-

Parameter (%)

Carbon

Hydrogen

Sulfur

Chlorine

Mineral

Carbon

Hydrogen

Sulfur

Chlorine

Mineral

TABLE 7 ANALYSIS OF BOTTOM ASH FROM BOI L ER UNIT 7

Prior to Installation of Dump Grates; 1976,1977

60% Load 80% Load 100% Load 0% RDF 0% RDF 0% RDF 10% RDF

7.51 (4.90)a

0.87 (0.56)

2.58 (1.12)

0.01 (0.01)

89.0 (4.33)

0% RDF

4.66 (0.88)

0.20 (0107)

1.07 (0.87)

0.01 (0.00)

94.8 (0.82)

5.46 (1.24) 5.53 (0.95) 35.4

0.61 (0.20) 0.49 (0.15) 3.83

3.59 (1. 40) 2.90 (3.95) 0.75

0.00 (0.00) 0.00 ·(0.01) 0.18

90.3 (0.98) 91.1 (4.76) 59.9

After Installation of Dump Grates; 1978

80% Load 10% RDF

2.10 (0.28)

0.23 (0.02)

1.12 (0.99)

0.02 (0.01)

96.6 (0.82)

20% RDF

3.11 (0.72)

0.37 (0.08)

0.31 (0.04)

0.02 (0.02)

96.2 (0.80)

0% RDF

6.62 ( * 0.38 * )

8.98 * ) 0.03 ( * ) 84.0 ( *

100% Load 10% RDF

1.49 (0.27)

0.18 (0.04)

1.12 (0.71)

0.02 (0.01)

97.2 (0.51)

(3.42)

(0.55)

(0.06)

(0.04)

(4.06)

20% RDF

1.85 (1.21)

0.21 (0.12)

0.34 (0.08)

0.02 (0.01)

97.7 (1. 41)

a Values in parentheses are ± one standard deviation

TABL E 8 ANALYSIS OF FLY ASH FROM BOILER UNIT 7

Prior to Installation of Dump Grate 1976, 1977

Parameter 60% Load 80% Load 100% Load

(%) 0% RDF 0% RDF 0% RDF 10% RDF

Carbon 0.79 (0.19) 0.95 (0.27) 1.87 (1.15) 4.68 (0.43) Hydrogen 0.27 (0.08) 0.60 (0.25) 0.61 (0.28) 0.07 (0.02) Sulfur 1. 52 (0.25 ) 1. 35 (0.28) 1. 35 (0.18) 1.02 (0.12) Chlorine 0.00 (0.00) 0.00 (0.00) 0.01 (0.02) 0.00 (0.00) Mineral 97.4 (0.47) 97.1 (0.57) 96.2 (1. 47) 94.2 (0.57)

After Installation of Dump Grate 1978

80% Load 100% Load

0% RDF 10% RDF 20% RDF 0% RDF 10% RDF 20% RDF

Carbon 1.85 (0.55) 2.43 (0.35) 2.54 (0.05) 1. 92 (0.78) 2.41 (0.49) 2.40 (0.40) Hydrogen 0.10 (0.02) 0.11 (0.01) 0.17 (0.05) 0.10 (0.02) 0.11 (0.01) 0.11 (0.02) S ulfur 0.70 (0.34) 0.69 (0.13) 0.86 (0.14) 1.02 (0.51) 0.82 (0.21) 0.83 (0.13) Chlorine 0.01 (0.01) 0.01 (0.00) 0.03 (0.01) 0.01 (0.01) 0.02 (0.01) 0.02 (0.01) Mineral 97.3 (0.55) 96.8 (0.46) 96.4 (0.09) 97.0 (0.39) 96.6 (0.59) 96.6 (0.30)

a Values in parentheses are + one standard deviation

506

tional carbon remaining in the RDF which had dropped into the bottom hopper of the boiler.

Further analysis of th6 samples of RDF indicat­ed that the ash fusion temperatures were from 60-lODe lower than those for coal. The RDF also has a high sodium content which had a significant detrimental effect on fouling index as compared to burning coal only. Thus some difficulty wi th slagging and fouling was anticipated and observed during these tests.

EMISSIONS

Results for the major effluents of interest from the emission tests are presented in Table 9. Each number tabulated in Table 9 represents the average of three ind�pendent runs unless otherwise noted. The standard deviations are also given in Table 9 as the values within parentheses. Standard devia­tions are measures of the variations in the experi­mental results caused by the uncontrolled factors in the experiment as well as uncertainties in the experimental measurements and the analytical analyses.

PARTICULATES

The variations of the uncontrolled particulates, controlled particulates and electrostatic precipitator efficiency with RDF are shown in Fig. 5, 6 and 7 respectively. From Fig. 5 it can be observed that the uncontrolled emissions generally increase with RDF except for the 100 percent load data using coal only. Otherwise all the runs show Significant increases in particulate emissions as the amount of refuse derived fuel increases to the steam generator. It is also apparent from this plot that the initial data obtained using coal only in this boiler in 1976 and 1977 indicates a reverse trend in terms of particulate emissions. The expected particulate would be higher at 100 percent load than at 60 percent load as was the case for the 1978 data. The reason for this reversed trend in the 1976 . data is believed to be related to difficulties noticed in operation of the particulate collector on Unit 7 during 1976 and 1977. The precipitator was exten­sively repaired between the 1976/77 experiments and the 1978 experiments. However, it should be emphasized that the scale for the emissions is

TABLE 9 SELECTED EMISSIONS FROM BOILER UNIT 7

Prior to Installation of Dump Grates 1976, 1977

Parameter Units 60% Load 80% Load 100% Load .

0% RDF 0% RDF 0% RDF 10% RDF

Particulates lb/l06

BTUb

(controlled) lb/l0

6BTU Particulates

(uncontrolled) lb/l0

6BTU Oxides of Sulfur

SOx 6

Oxides of Nitrogen lb/IO BTU NOx

lb/lO�BTU Chlorides Formaldehyde lb/l0

9BTU

Methane lb/IO BTU

0.23

9.05

2.61

0.32

5.14 4.56 0.00

(0.07)a

0.35 (0.12) 0.60 (0.09) 0.53 (0.12)

(1. 02) 7.49 (1.72) 8.26 (0.05) 8.35 (0.30)

(0.40) 2.88 (0.70) 3.70 (0.16) 2.88 (1.l4 )

(0.03) 0.26 (0.09) 0.35 (0.02) 0.27 (0.04)

(3.75) 13 .6 (8.42) 28.14 (6.91) 7.65 (5.05) (5.58) 20.9 (44.0) 5.49 (4.58) 60.0 (52.6) (O.OO) 0.00 (O.OO) 0.00 (O.OO) 0.00 (O.OO)

---- ----------------- ------------------------ - ---- -----------------------------------------------------------

Parameter Units

Particulates lb/l06

BTU (controlled)

lb/l06

BTU Particulates (uncontrolled)

lb/l06

BTU Oxides of Sulfur SOx

6 Oxides of Nitrogen lb/IO BTU

NOx lb/lO�BTU Chlorides

Formaldehyde lb/l09

BTU Methane Ib/10 BTU

After

0% RDF

0.21 (0.05)

6.54 (1. 33)

3.42 (0.14)

0.39 (0.02)

10.7 (1. 77) 8.37 (14.0) 5.30 (2.65)

Installation of Dump Grates

80% Load

10% RDF 20% RDF

0.37 (0.09) 0.37 (0.07)

7.63 (0.63) 8.21 (1.21)

2.84 (0.16) 2.33 (0.63)

0.33 (0.02) 0.33 (O. 03)

50.9 (35.8) 93.7 (8. 96) 12. (207.) 0.77 (0.42) 6.07 (1. 58) 3.77 (0.30)

a values in parentheses are + one standard deviation

1978

100% Load

0% RDF 10% RDF

0.42 (0.2l) 0.44 (0.07)

7.93 (3.58 ) 7.28 (0.53) .

3.30 (2.07) 2.33 (0.49)

0.31 (0.04) 0.26 (0.01)

7.65 (1. 88) 58.4 (31. 9) 0.19 (O. 33) 1.44 (O. 72) 3.35 (0.93) 4.58 (1.44 )

b 6 ' to convert from Ib/10 BTU to micrograms/Joule, multiply values in the above table by 0.430

507

20% RDF

0.53 (0.09)

7.47 (0.53)

1.93 (0.5l)

0.26 (0.03)

28.6 (9.35) 0.42 (0.19) 2.47 (0.58)

4.00,--------------------, EFFECT OF RDF ON UNCONTROLLED EMISSIONS

3.80

3.60

� 3.4 .... � �

•

�

� -

� � - 3.10 '" w

" w � �

Ii! ...

� 3.00 u

!5 o '" 60 PERCENT LOAO D '" 80 PERCENT LOAD c. '" 100 PERCENT LOAD

OPEN SYM80L '" 1978 DATA SHADED SYMBOL", 1976 OR 1977 DATA

1.8

1.60 '-__ -'-__ -" ___ ,.L-__ -,J-___ ,L-...J o 4 8 11 16 10

REFUSE DERIVED FUEL HEAT INPUT. PERCENT

FIG.5 UNCONTROLLED PARTICULATE EMISSIONS

:::

o EFFECT OF RDF ON CONTROLLED EMISSIONS

0." r-

0.10 1-

L------"" 0.1

g:' 0.16 •

�

i!j 0.14 -� '" -

� 0.12 " w � �

li! o. ...

8 0.08

0.06

0.04

0.01

o '" 60 PERCENT LOAD D '" 80 PERCENT LOAD c. '" 100 PERCENT LOAD

OPEN SYMBOL'" 1978 DATA SHADED DYMBDL '" 1976 or 1977 DATA FLAGGED SYMBOL'" 1978 DATA AFTER RELOCATION OF RDF NOZZLE

O.OO !,-__ --:-__ -!: ___ ""=-__ -,! ___ -:! o 4 B 1

REFUSE DERIVED FUEL HEAT INPUT. PERCENT

FIG.6 CONTROLLED PARTICULATE EMISSIONS

98.0,---------------------,

97.0

96.0

� u z w -u -

t:: 95.0 w

� " ... u w � � "

u 94.0

93.0

91.0

o '" 60 PERCENT LOAD D '" 80 PERCENT LOAD c. '" 1 00 PERCENT LOAD

OPEN SYMBOL'" 1978 DATA SHADED SYMBOL", 1976. 1977 DATA

91.0 �--+--_!:__--_;';;_--_;';,_--f.:__J o 4 8 11 16 10

REFUSE DERIVED FUEL HEAT INPUT. PERCENT

FIG. 7 ELECTROSTATIC PRECIPITATOR EFFICIENCY

significantly expanded and that all the emissions with the 100 percent coal as the only fuel are with­in 2.8 to 3.9 mg/J of heat energy input.

The controlled emissions generally increase with increases in RDF. This result was expected since the amount of ash in the RDF was propor­tionally larger than that of the coal. For the 100 percent coal runs (0 percent RDF) the decrease in the emissions at 8 0 and 100 percent load for the 1978 data is a result of repair of the electro­static precipitator which occurred late in 1977. Difficulty was experienced with the electrostatic

. precipitator during 1976 and 1977 because some of the plate retainers in the precipitator had failed, rendering some of the precipitator plates to be ineffective during the test runs accomplished in 1976 and 1977. This is one reason why the emis­sions for the 60, 8 0 and 100 percent load in 1976 and 1977 appear to be Significantly higher than

508

the emissions for the corresponding loads in 1978. Thus, the data obtained in 1978 is much more representative for the usual performance of Unit 7. Furthermore, the data of 1978 show very con­sistent trends in the direction anticipated based on the fuel input analysis.

At 100 percent load additional runs were made in 1978 for Iowa Department of Environmental Quality compliance checks with federal and state regulations. Some of these runs were accomplish-

ed after the location of the RDF injection point was placed below that of the coal injection nozzles in the boiler. The specific locations of the coal nozzles were at both the 1 32.5 and 1 29 .2 ft (43 .7 and 42 .6m) level while the RDF was injected at the 1 27 . 1 ft. (42 m) level. For comparison the dump grate level in the boiler was at the 1 00.5 ft. (33 .2 m) level so the suspension zone for the RDF was slightly less than 27 ft . (8.9 m). The RDF injection point for the emission data presented in Fig. 6 is all with the RDF injection point located between the two coal nozzles except for the runs where the data point plotted is flagged. The importance of the location of the RDF injection point is dramat­ically shown in Fig. 6 and indicates that the parti­culate emissions can be reduced significantly by the proper location of the RDF injection point.

It should also be noted that the particulate mat­ter collected during the runs with RDF often contained several pieces of black flaky material which had passed through the electrostatic pre­cipitator and into the sample trains used for col­lection of particulate matter. The black flaky sub­stance would stick to sides of houses and auto­mobiles and was quite disconcerting to several citizens who forwarded complaints. Relocation of the RDF injection point in the boiler eliminated the problem almost completely as well as lower­ing significantly the particulate effluent from the smokestack . Lowering of the amount of particulate effluent from the smokestack because of the re­location of the RDF injection point resulted in the Iowa Department of Environmental Quality not applying emission offsets to the operation of boiler Units 5 and 6 at the Ames Power Plant.

The effect of RDF on the electrostatic precip­itator collector efficiency is shown in Fig. 7 . From this figure it is clear that the efficiency drops con­sistently with increases in RDF. These trends are very consistent for the data obtained in 1 978 and showed the electrost�tic precipitator efficiency to be higher at 80 percent load than at 1 00 percent load. The effect of the repair between 1 977 and 1 978 data is also apparent in this figure . For ex­ample, for the coal only runs, the collector ef­ficiency changed from 93.4 - 94.4 percent ef­ficiency at 1 00 percent load. At the 80 percent load the efficiency increased from 94.9 - 96.8 per­cent thus demonstrating the dynamic effect the repair of the precipitator had on its performance .

Figs. 8 and 9 show the particulate size distri­bution as determined via an Andersen cascade impactor for the 80 and 1 00 percent load runs.

From these figures it is apparent that at 1 00 per­cent load the sizes for coal only, 1 0 and 20 per­cent RDF are all about the same. At 80 percent load the differences were significant. From these figures it is apparent that particle size increases with increases in RDF.

Figures 1 0 and 1 1 show the particulate size distri­bution at the collector inlet and in the stack determined by a Brinks cascade impactor. Compar­ison of these two figures reveals the size distri­bution retained by the electrostatic precipitator. The Brinks and Andersen size distribution results for the stack particulate do not agree because the sampling probes and calibrations are different for each type of cascade impactor. Thus, comparisons can only be made in a relative rather than an absolute sense and only for a given type of cascade impactor.

OXIDES OF SULFUR

The oxides of sulfur (SOx) emitted from the boiler decrease significantly with increases in RDF as shown in Fig. 1 2 . This decrease amounted to about 50 percent for boiler loads of both 80 per­cent and 1 00 percent in going from 0 to 20 percent RDF. Thus, an advantage of using RDF with coal is that relatively high sulfur coal can be used while still meeting EPA regulations.

100.0,-----------------,

10.0 ...----10% ROF

CUMMULATl VE PCT LESS THAN 050 FIG. 8 PARTICULATE SIZE DISTRIBUTION FOR

80 PERCENT LOAD

509

V> % 0 '" u -:IE:

0 �

0

� N -V>

� -' u -... '" « �

100.0.-----------------------------------------,

10.0

20% ROF

1.0

10% RDF

PARTICLE SIZING 100% LOAD UNIT 7 ANDERSEN 1978

O� ROF

0.1L-_����������� -���-=_� 2 5 10 15 20 30 40 50 60 70 80 85 90 95 98

CUMMULATIVE PERCENT LESS THAN D50

FIG.9 PARTICULATE SIZE DISTRIBUTION FOR 100 PERCENT LOAD

V>

i5 '" u -:IE:

0 �

0

W N -V>

w -' u -

:;; « �

1DO.OI,----------------------

1D.0

1.0

PARTICLE SIZING UNIT 7 100% LOAD 1978 8RINKS BPC

0% RDF

10% RDF

0.1L-__ ����������������--��� 2 5 10 15 20 30 40 50 60 70 80 85 90 95 98

CUMULATIVE PCT LESS THAN D50

FIG. 10 PARTICULATE SIZE DISTRIBUTION AT THE COLLECTOR

V> %

51 u -:IE:

0 �

0

� N -V>

� -' u -... '" « �

.,

100.0,.----------------------,

0% RDF ___ ..,..",/ 10.0 ./'"" ... 2D% RDF

10% RDF

PARTICLE SIZING UNIT 7 100% LOAD 1978 8RINKS STACK

O. 1 '=---;L--,l;;�,...f.,.......,f;;_-+.::-t;,,._7.;_;�+n_.f<-+n-*""� 2 5 1 0 15 20 30 40 50 60 70 80 85 90 95 98

CUMMULATIVE PCT LESS THAN D50 .

FIG.11 PARTICULATE SIZE DISTRIBUTION IN THE STACK

1.6K----------------,

EFFECT OF RDf ON SOx EMISSIONS

1.2

1.0

..... 0> "-

•

V> i5 0.8 -V> o '" 60 PERCENT LOAD V>

o '" 80 PERCENT LOAD -:IE:

o '" 100 PERCENT LOAD w

• OPEN SYM80L '" 1978 DATA fii 0.6 SHADED SYIIlOL '" 1976 or 1977 DATA

0.4

0.2

0.0 '--___ '--___ '--__ ---,'::-__ --,'--__ ---''--' o 4 8 12 16 20

REFUSE DERIVED FUEL HEAT INPUT, PERCENT

FIG.12 OXIDES OF SULFUR EMISSIONS

510

OXIDES OF NI TROGEN

The oxides of nitrogen (NOx) generally decrease with increases in RDF at all boiler loads as shown in Fig. 1 3 . The decrease was generally in the range of 1 0 to 20 percent and somewhat dependent on boiler load as the RDF was increased up to 20 percent. The NOx emissions generally decreased less for the 1 978 data than for the 1 976-1 977 data. This may represent better operation of the boilers and better control of the combustion zone tem­peratures for the experimental runs of 1 978.

..., ...... Ol ::>.

vi z: 8 In In ::E ....

x 0 z:

0 . 1 8

0 . 1 6

0 . 1 4

0 . 1 2

0 . 1 0

0 . 08

0 .06

0 . 04

0 . 02

EFFECT OF RDF ON NOx

E M I SS I ONS

BOILER UN IT 7

o '" 60 PERCENT LOAD o '" 80 P E RCENT LOAD {). '" 1 00 PERCENT LOAD

OPEN SYMBOL ", 1 978 DATA SHADED SYMBOL ", 1 976 OR 1 977 DATA

O. DO '=-_ __,!---�-__::';;_-___::':_-__,� o

REFUSE DERI VED FUEL H EAT I N PUT . PERCENT

FIG. 1 3 OXI DES OF NITROGEN EMISSIONS

CHLORIDES

The chloride emissions for the suspension fired boiler increased linearly and significantly with in­creases in RDF as shown in Fig. 14 except for the 1 00 percent load, 20 percent RDF data point. The boiler experienced as much as a 1 0 fold increase in chloride emissions as the RDF increased from 0-20 percent for all boiler loads in 1 978. The chlor­ides in the stack emissions are believed to come

5 1 1

from the chlorinated hydrocarbons in the RDF. The chlorides drop for the 1 976- 1977 data be­cause of the dropout of RDF into the bottom hopper since the bottom grates were not installed at this time.

..., ...... Ol c

vi z: 8 In In ;: .... .... c

;:;: a ...J ::t: U

40 .0

3 5 . 0

30 .0

2 5 . 0

2 0 . 0

1 5 . 0

EFFECT O F RDF ON CHLORIDE E M I S S I ONS

o '" 60 P E RCENT LOAD o '" BO PERCENT LOAD {). '" 1 00 PERCENT LOAD

OPEN SYMBOL '" 1 978 DATA SHADED DYMBOL '" 1 976 . 1 977 DATA

D . O � __ L-__ L-_�� __ � __ � o 4 8 1 2 1 6 20

REFUSE DERIVED FUEL H EAT INPUT . PERCENT

FIG. 1 4 C HLORI DE EMISSIONS

A L DEH YDES A ND KETONES

No conclusive trends concerning aldehydes and ketones (reported as formaldehyde) emissions could be observed as illustrated in Fig. 1 5 . Formal­dehyde emissions depend on the constituents of the RDF such as wood chips, leaves and other cel­lulose fiber. No control of the quantity of these items in the RDF was attempted. Consequently the emissions due to the burning of cellulose fiber were as variable as the constituency of the RDF.

METHANE

The unburned (C1 ) hydrocarbons (methane) emitted from the stack increased and then de­creased as the RDF increased to 1 0-20 percent respectively as shown in Fig. 1 6. The level of me­thane detected is very low as was expected.

60

50

o "" 60 PERCENT LOAO o ... 80 PERCENT LOAO " 0, 100 PERCENT LOAO

OPEN SYMBOL ; AFTER INSTALLATION OF 01M' GRATES , 197B

HADED SYMBOL ; BEFORE I NSTALLAT ION OF OUMP GRATE . 1 9 7 6 , 1 977

? 40

30

20

1 0 EFFECT O F REFUSE DERIVEO FUEL ON ALOEHYOE AND KETONE EMISSIONS ANALYSEO ANO REPORTEO AS FORMALDEHYDE

0�0----�======�====�1�2�==�16====�2�0� REFUSE OERIVED FUEL HEAT INPUT, PERCENT

FIG, 1 5 ALDEHYDE AND KETONE E M ISSIONS

2 . Br-------�------�---, EFFECT OF RDF ON METHANE EMISSIONS

2 . 0

1 . 6

� 1 . 2 o "" 60 PERCENT LOAD o ... 80 PERCENT LOAO " "" 100 PERCENT LOAD

OPEN SYt'80L "" 197B DATA � SHADED SYMBOL "" 1976 OR 1977 OATA

! O . B

0 . 4

0 . 0��==t=====!===:'---},----.,J;.-------,!;� REFUSE DERIVEO FUEL HEAT INPUT . PERCENT

FIG, 1 6 M ET HANE EMISSIONS

TRA CE ELEMENTS

A series of 1 9 trace elements were sampled from all input and all output streams associated with the operation of steam generator Unit 7 , Table 1 0 lists the trace elements detected in the input fuels of coal and RDF used during the test of this parti­cular study, The elements selected for analysis are

5 1 2

listed by rank order in Table 1 0 where the ranking had been determined by the concentration given in parts per million, In addition the standard devia­tions which go along with the average concentra­tions are also listed , Another column in Table 1 0 also shows the amount o f the trace element listed on the basis of mass per unit of energy input to the boiler. The values listed in Table 1 0 are overall averages for both the coal and RDF used during tests performed in 1 978. The trace elements with higher proportions of concentration in coal as compared with RDF are also identified in this table as strontium, beryllium, nickel and germa­nium. The elements that were not detected based on the detection limit of the analytical instru­mentation are also indicated in this table. Ele­ments relatively high in concentration in the RDF are shown in Table 1 0 to be zinc, lead, copper, manganese and vanadium.

Table 1 1 lists the trace elements detected in the bottom ash , hopper ash (fly ash) , and parti­culate matter collected on a fIlter in the sampling train located in the smokestack of Unit 7 . These elements are listed in rank order based on mass of the element per unit of energy input to the boiler in this particular table. The detection limits of the x-ray fluorescence device used to analyze for trace elements in the solid materials collected in the output stream samples is also indicated in Table 1 1 . These numbers are in units of parts per million by weight of the element , except for the stack fIlters which are given in units of micrograms per square centimeter of area over which the particulate is collected on the fIlter. Trace ele­ments of relatively high concentration in the fly ash are zinc, lead, manganese , vanadium and strontium. These elements along with nickel, cop­per and chromium are high in relative concentra­tion in the bottom ash.

These tables are presented so that a comparison can be made with the trace elements scanned in the input fuels. If mass balances are desired , the quantity coming in with the input fuel minus the amount of the individual trace element in the bot­tom ash minus the amount of the individual amount in the fly ash collected in the electrostatic precipitator hopper would give an estimate of the amount of the element leaving the smokestack of the boiler.

Trace elements were also analyzed from the particulates collected in the smokestack and from sampling train impinger solutions to yield addi­tional estimates of the trace elements that were

emitted from the boiler stack. However, these estimates are not as reliable as the estimates found by taking what comes in with the input fuel and subtracting what drops out with both the bot· tom grate ash and the hopper fly ash : Some mass balances have been attempted but are not present· ed here because some of the resulting balances are inconsistent due to the low constituent levels, sampling uncertainty, and analytical uncertainty. Sometimes the uncertainties accumulate to the point where attempts at mass balances yield nega· tive quantities leaving the smokestack. TItis does not mean that there is an accumulation of such a

material in the boiler or in the generating system but is simply an indication that there are un· controlled factors in the experiment which cause variation in the data, that there are experimental errors in the measurements, and that there is error in the analysis of the samples. A combination of the variability due to the uncontrolled factors, the experimental error, the analytical error and the analytical detection limits often overshadow the small amount of the trace element actually in the system. This is especially evident when the average values of the trace elements for a given load and percent RDF are ,.relatively small .

TABLE 1 0 TRACE ELEM ENT CONTENT OF COAL AND RDF USED AS FUEL IN UNIT 7

COAL RDF

Element Level* Element Leve l *

ppm ng/J ppm

S t ront ium** 86 + 2 8 2 . 9 2 + 1 . 15 Zinc 7 6 3 + 345 - - -

Vanadium 83 + 16 2 . 9 2 + . 5 2 Lead 6 1 3 + 289 - - -

Manganese 76 + 2 3 2 . 6 7 + . 79 Copper 5 7 2 + 854 - -

Zinc 66 + 4 1 2 . 39 + 1 . 46 Mangane se 194 + 4 7 -- - -

Beryllium** 37 + 1 2 1 . 55 + 0 . 49 Vanadium 154 + 32 - - -

Lead 36 + 1 3 1 . 26 + . 48 S t ront ium 46 + 1 1 - - -

Tin 20 + 5 0 . 7 1 + . 1 7 Chromium 34 + 8 - - -

Chromium 19 + 7 0 . 74 + . 35 Tin 27 + 8 - - -

Nickel** 18 + 5 0 _ 6 3 + . 1 7 Ant imony 25 + 1 7 - - -

Copper 15 + 3 0 . 52 + 0 . 0 8 Gallium 16 + 3 - - -

Germanium** 5 . 3 + 0 . 9 0 _ 19 + 0 . 04 Nickel 14 + 4 - -

Gall ium 2 . 5 + 0 . 5 0 . 09 + 0 . 0 2 Selenium 8 + 1 - - -

Ant imony BDL+ BDL Cadmium 6 . 4 + 8 . 1 -

Selenium BDL BDL Germanium 1 . 7 + 0 _ 3 -

Thal l ium BDL BDL Thall ium BDL

Mercury BDL BDL Mercury BDL

Arsenic BtL BDL Arsenic BDL

Cadmium BDL BDL Beryll ium BDt

Cobalt BDL BDL Cobal t BDL

* Values listed are overall averages for the coal and RDF used during 1978 tests .

** Trace elements with higher proportions in coal than in RDF

+ BDL s ignifies the element is b elow the analyt ical instrumentat ion det ect ion l imi t .

5 1 3

ng/J

4 . 65 + 2 . 1 3 -

3 . 89 + 2 . 27 -

3 . 58 + 5 . 74 -

1 . 18 + 0 . 33 -

0 . 94 + 0 . 19 -

0 . 28 + 0 . 04 -

0 . 28 + 0 . 04 -

0 . 1 7 + 0 . 06 -

0 . 15 ' + 0 . 11 -

0 . 10 + 0 . 0 2 -

0 . 09 + 0 . 03 -

0 . 05 + 0 . 01 -

0 . 04 + 0 . 05 -

0 . 01 + 0 . 00 -

B DL

BDL

BDL

BDL

BDL

Vl

.....

.j:>.

TA

BL

E 1

1

TR

AC

E E

LE

ME

NT

CO

NT

EN

T O

F B

OT

TO

M A

SH

, H

OP

PE

R F

LY

AS

H A

ND

ST

AC

K F

LY

AS

H

-

Elem

ent '"

(O

P.tec

t ion

Ll

::)its

) PP

M'"

Chro

miu

m

(100

)

Hang

;t.nes

e (7

0)

Zinc

(2

0)

Lead

(2

5)

Nick

el

(20)

Copp

er

(20)

Vana

dium

(2

00)

Stro

ntiu

m (5

) Ga

lliu

m

(10)

Ti

n --

-

Bery

llium

(1

0)

Ant i

mony

--

-

Sele

nium

(5

) Th

all i

um

---

Germ

aniu

m

(10)

He

rcur

y --

-

Arse

nic

(to)

Ca

dmiu

m -

-

Coba

lt

(30)

Bott

om A

sh

Leve

l .

Hax

Max

PPM

ng

/J

2450

5.

62

1200

2.

66

1680

2

.62

1240

1.

68

680

1.5

6

910

1.4'

>

660

1.3

3

750

1.0

3

320

0.58

160

0.1

1

120

0.0

13

230

0.4

36

BOL

BOL

BOL

BOL

.BOL

BO

L

BOL

BOL

BOL

BOL

BOL

BOL

BOL

BOL

Elem

ent·

Av

g ±

Std

!lev

og/J

0.9

13 +

1.4

0 Zi

nc

0.9

50 ±

.80

5.

Hang

anes

e

0.9

64 ±

0.7

90

Lead

0.6

34 ±

0.5

34

Vana

dium

0.2

61 ±

0.3

87

Stro

ntium

0.4

)) ±

0.4

42

Chro

miu

m

0.4

48 ±

0.3

61

Copp

er

0.3

59 ±

0.2

74

Nick

el

0.0

55 ±

0.1

33

Call

ium

0.0

34 ±

0.0

34

Germ

aniu

m

0.0

11 ±

0.0

02

Tin

0.0

96 ±

0.1

17

Bery

lliu

m

BOL

Cadm

ium

BOL

Anti

mony

BOL

Sele

niu

m

BOL

Thal

lium

BOL

Mer

cury

BOL

Ars

enic

BOL

Coba

lt

'" De

tect

ion

11.mIt

s fo

r Bo

ttom

Ash

and

Hopp

er A

sh a

re t

he s

ame

Max

PPM

2250

132

0

1640

1030

1130

404

218

140 86

51

59

120 11

92

BOL

BOL

BOL

BOL

BOL

Hopp

er A

sh

Part

icul

ate

00

St

ack.

Fil

ters

Leve

l lU

emen

t (De

tect

ion

Leve

l Ma

x --

Avg

+ St

d IJe

v Li

mit

s)

I' .. x

Max --"

vg ±

Std

Oev

ng

/J

fiR/J

Mg

/cm

2 lf

g/cm

2 pg

/J

pg/J

10.7

4 4

.717

+ 2

.428

Zi

nc

(0.2

) 82

12

5:0

35.3

5 ±

26.

26

5.8

0 3.

469

+ 1.

178

Lead

(1

) )0

37

:4

18.9

4 ±

13.

19

5.76

2.

994

+ 1

.849

Va

nadi

um

(1)

8 11

.0

7.00

± 1

.44

3.67

2

.744

+ 0

.623

St

ront

luo

(0.1

) 10

D

.7

6.6

6 ±

2.4

0

3.)

) 2

.296

+ 0

.600

Ma

ngan

es.:!

(0.5

) 10

10

,/t

5.4

6 ±

2.1

1

1.44

0

.861

± 0

.259

Ce

rma�l

u",

(0.4

) 1

.7

2.5

9 1

.39

±0

.54

.759

0

.417

+ 0

.189

Co

pper

(0

.2)

2.6

2.69

D

.5±

0.5

8 .4

70

0.3

06 +

0.0

84

Chro

miu

m

(0.5

) 1.

6 2

.43

1.3

3 ±

0.5

5 .3

30

0.1

72 +

0.0

83

Gall

ium

(0

.4)

1 .5

2.2

8 1

.31

± 0

.67

.248

0

.126

+ 0

.039

N

icke

l (0

.3)

1.2

1.6

5 1.

24 ±

0.4

8

.200

0

.096

± 0

.068

Se

len

ium

(0.2

) <0

. 5

0.'7

9 0

.43

± 0

.15

0.3

1 0

.249

+ 0

.057

A

rsen

ic

(0.2

) 80

L BO

L BO

L

0.0

5 0

.019

± 0

.015

Be

ry 1 l

1U"ll

--N

.A.

N.A

. N

.A.

0.3

0 0

.139

+0

.116

Ca

dmiu

m

---

BOL

BOL

BOL

BOL

BOL

Cob

alt

(0.4

) nO

L BO

L BO

L

BOL

BOL

Merc

ury

---

BOL

BOL

BOL

BOL

BOL

Anti

mony

--

BOL

BOL

BOL

BOL

BOL

Tin

---

BOl

BOL

BOL

BOL

BOL

Thal

lium

--

DOL

BOL

BOL

Table 1 2 is the listing of the trace elements detected in the exhaust gas stream from smoke­stack of Unit ·7 . The particular trace elements have been condensed or collected in the solutions in two different sample trains. One solution used was water in the particulate sample train where the flow rate through the train was relatively high at a value of about 0.2 Lis. The other sample train was the "trace element" sample train where the flow rate through the train was relatively low at a value of about 0.02 Lis . The various flow rates through these trains have been previously indicated in Table 4. The solution in the particulate train was water whereas the solutions in the trace elements train were sodium hydroxide (located in an im­pringer used as a prescrubber in a sample train) followed by two impingers which contained iodine monochloride . The analysis of the trace elements used in this particular sample train have been combined to include the total amount collected in the sodium hydroxide prescrubber and both of the iodine monochloride impingers. All of the trace elements found in the solution of the sample

trains have been analyzed by the Inductively Coupled Plasma technique (ICP). The only ele­ments jointly detected in the impinger solutions of both sample trains were cobalt and arsenic. However, other elements consistently detected were mercury, selenium, zinc, copper, manganese and tin .

Table 1 3 is a listing of trace elements detected in the sluice water which was the water used to sluice the bottom ash from the bottom pit of boiler Unit 7. This sluiced material then passes through a pipeline for approximately 1 300 ft. C 400 m) to an ash pond where the sluice water samples and bottom ash samples were collected by the Fiscus-Joensen technique . The levels of strontium, antimony, arsenic and mercury were at the highest relative levels and all at values less than 0 .4 mg/L.

Table 14 is a listing of th.e trends of the trace elements that are in the output streams as they either increase, decrease or remain constant with respect to the amount of refuse derived fuel com­ing in with the input fuel . The plus signs in the

TABLE 1 2 TRACE E LEMENTS COND ENSED OR ABSORBED IN I MP I NGER SOLUT I ONS OF

SAMPLE TRAINS

Trace Element Sample Train ( Impinger ICL) Particulate Sample Train ( Impinger H2O)

Element (Detection* Level Element (Detection* Level Limit) ICL Max Max Ave + Std Dev Limit) Max Max Ave + Std Dev Mg/! Mg/l pg/J pg/J Mg/I Mg/I pg/J pg/J

Cobalt 0 . 0 5 0 . 284 1 741 234 + 403 Cobalt 0 . 002 0 . 19 3 14 . 9 3 . 18 + 4 . 06

Arsenic 0. 40 <0. 400 9 4 3 3 8 5 + 208 Mercury 0. 10 0 . 29 9 14 . 7 4 . 46 + 3 . 04

Zinc 0 . 02 0 . 6 1 7 2 3 3 70 . 6 + 5 5 . 4 Arsenic 0 . 05 0 . 082 8 . 00 5 . 26 + 1. 21

Copper 0 . 01 0 . 09 5 5 5 . 0 2 5 . 6 + 21 . 9 Selenium 0 . 0 4 0 . 060 5 . 7 8 3 . 81 + 0 . 9 8

Manganese 0 . 004 0. 136 45. 3 9 . 5 + 10 . 6 Tin 0 . 04 0 . 2 70 5 . 60 4 . 07 + 0 . 8 7

Beryllium 0 . 001 BDL** BDL BDL Beryllium 0 . 00005 BDL BDL BOL

Cadmium 0 . 10 BDL BDL BDL Cadmium 0. 003 BDL BDL BDL

Chromium 0 . 08 BDL BDL BDL Chromium 0 . 00 3 BDL BDL BDL

Gallium 0 . 2 3 BDL BOL BDL Copper 0. 0005 BDL BDL BDL

Germanium 0 . 30 BDL BDL BDL Gallium 0 . 01 BDL BDL BDL

Mercury 0. 22 BDL BDL BDL Germanium 0 . 01 8DL BDL BDL

Nickel 0 . 1 7 BDL BDL BDL Manganese 0 . 001 BDL BDL BDL

Lead 1 . 80 BDL BDL BDL Nickel 0 . 0 7 BDL BDL BDL

Antimony 0 . 4 2 BDL BDL BDL Lead 0 . 05 BDL BDL BDL

Selemium 0 . 63 BDL BDL BDL Tin 0 . 025 BOL BDL BOL

Tin 0 . 2 3 BDL BDL BDL Stront ium 0 . 0 2 BDL BOL BDL

Strontium 0 . 28 BDL BDL BDL Thallium 0 . 1 BDL BDL BDL

Thallium 7 . 5 BDL BDL BDL Vanadium 0 . 004 BDL BDL BDL

Vanadium 0 . 06 BDL BDL BDL Zinc 0 . 0 2 BDL BDL BDL

* Detec t ion limit of the primary chemical, ICL is given

** BDL indicates below the analyt ical instrument detection limit

5 1 5

table indicate an increase of the trace element with increased amounts of RDF used in the input fuel, whereas the negative sign indicates a decrease in the amount of trace elements with respect in RDF in the input fuel . A zero indicates no detec­table trend within the data scatter. Those elements detected in the stack effluent consistently increas­ing with increases in RDF were manganese , zinc, lead, copper, gallium and chromium.

Table 1 5 lists the estimated trace elements in the ambient air in terms of a stack gas concentra­tion given in units of milligrams per cubic meter (mg/m3) of stack gas. An estimate of the con­centration of the element in the ambient air is obtained from the flue gas concentration by divid­ing the flue gas concentration by a factor of 1 ,000 to yield the ambient air estimate of the element concentration. The estimate of the ambient air concentration is then compared to the threshold limit value (TL V) of the element as given in Ref­erence [8] . The ratio of the estimated ambient air concentration to the threshold limit value has been listed in this table to approximate the relative toxicity of the element scanned ·during the parti­cular tests of this study. Finally, a comparison is

TABLE 1 3 TRACE E LEMENTS DETECTED IN

BOILER SLUICE WATER

Element (Detection Blank Max Ave + Std Oev Limit ) Level Value Value !lglJ mg/l mg/l mgIJ

Strontium 0 . 07 0 . 271 0 . 372 0 . 267 ± 0 . 047

Ant imony 0 . 025 BOL 0 . 233 0 . 069 ± 0 . 061

Arsenic 0 . 05 0 . 083 0 . 100 0 . 074 ± O . OlB

Mercury 0 . 035 0 . 042 0 . 05 0 . 041 ± 0 . 01 7

Manganese 0 . 001 0 . 212 0 . 04 0 . 008 ± 0 . 013

Copper 0 . 0005 BOL 0 . 04 0 . 004 ± 0 . 010

Chromium 0 . 003 BDL 0 . 02 0 . 006 + 0 . 005

Vanadium 0 . 004 BDL 0 . 01 0 . 006 ± 0 . 002

Cobalt 0 . 00 2 0 . 010 0 . 0 1 0 . 003 + 0 . 002

Beryllium 0 . 0005 BOL BOL BOL

Cadmium 0 . 003 BOL BOL BOL

Gallium 0 . 01 BOL BOL BOL

Nickel 0 . 0 7 BOL BOL BOL

Lead 0 . 05 BOL BOL BOL

Selenium 0 . 04 BOL BOL BOL

Tin 0 . 04 BOL BOL BOL

Zinc 0 . 02 BOL BOL BOL

Tha111wn 0 . 10 BOL BOL BDL

Germanium 0 . 01 BOL �DL BOL

made between the data of this study and the pre­dictions listed in Reference [9] from the 8t. Louis data. From the 8t . Louis study it was predicted thaCchromium ( 1 0 .2), copper (7.4), lead (3 .7) and zinc ( 1 .2) would be above the TLY and of relative Significance in the order given. None of the trace elements detected on the stack particulate train filters during these studies have predicted ambient air concentration above the TLY although lead is at a level near 30 percent of the TLY. However, from the trace element sampling train cobalt (5 .3) and arsenic ( 1 .7) are indicated to be above the TLY for the Ames study.

MINOR ELEMENTS

Table 1 6 lists the minor constituents (elements) found in the coal, RDF, bottom ash and fly ash for the 1 978 test runs on Unit 7. The elements listed are aluminum, silicone, sulfur, potassium, calcium, iron and titanium.

FUEL ENRICHMENT FACTORS

Figures 1 7 and 1 8 are plots of fuel enrichment factors of the trace elements to show the relative importance of RDF in causing increased emissions of the particular trace elements scanned in this study. From these figures the elemen ts of zinc, copper, and lead have Significant increases with RDF over coal alone'. Gallium increases slightly with RDF. The reason for copper being relatively high at 1 2 percent RDF and then dropping at 22 percent RDF is not known at the present time. Zinc comes into the RDF from paper where it is used as a fllier material. Lead and copper come . into the RDF from inks used in newsprint and colored pictures. The gallium is felt to be an im­purity that comes into the RDF along with another element such as aluminum.

The fuel enrichment factor (FEF) in Figs. 1 7 and 1 8 is defmed by the relation

5 16

FEF

where

[xl c + RDF [Fel c + RDF

[x1 c [Fel c

[x/Fe1 c + RDF

[x/Fe] c ( 1 )

[x] micrograms of trace element x per Joule of heat energy input

TABLE 1 4 TRENDS OF TRACE E LEMENTS VARIATION WITH RDF INPUT TO THE

STEAM GENERATOR

Antimony

Arsenic

Beryllium

Cadmium

Chromium

Cobalt

Copper

Gallium

Germanium

Lead

Nickel

Manganese

Mercury

Selenium

·Strontium

Tin

Thallium

Vanadium

Zinc

Bottom Ash

+

BDL

BDL

+

BDL

+

+

BDL

+

+

+

BDL

BDL

+

+

BDL

+

+

Hopper Ash

+

BDL

+

+

BDL

+

+

o

+

o

+

BDL

BDL

+

+

BDL

+

+

Stack Particulate

Filter

BDL

BDL

N .A .

BDL

+

BDL

+

+

o

+

o

+

BDL

o

o

BDL

BDL

o

+

Stack Impinger

Water

BDL

o

BDL

BDL

BDL

o

BDL

BDL

BDL

BDL

BDL

+

+

o

BDL

o

BDL

BDL

BDL

+ indicates an increasing trend with RDF increases

o indicates a lack of any definite trend

- indicates a decreasing trend with RDF increases

StacJ< Impinger ICl-NaOH

BDL

o

BDL

BDL

BDL

o

BDL

'BDL

BDL

BDL

BDL

BDL

BDL

BDL

BDL

BDL

BDL

o

Sluice Water

o

o

BDL

BDL

BDL

BDL

o

BDL

BDL

BDL

BDL

BDL

BDL

BDL

o

BDL

BDL

o

BDL

BDL indicates levels were below the detection limit of the analytical instrument

N.A. indicates that no analysis was available

5 1 7

TABLE 1 5 TRACE ELEMENT CONCENTRATION FROM STACK FLUE GAS AS AN ESTIMATE FOR

AMBIENT AIR CONCENTRATION

Concen- S t . Louis* Concen- tration Filter T . E . Flue Gas S t . Louis trat ion from Trace Conc . * Train* eoncen- Conc .

TLV from Filter Element TLV TLV tration TLV mg/m3 mg/m3 train mg /m3 mg/m3

Ant imony 0 . 5 BDL** BDL 0 . 0 22 0 . 044

Arsenic 0 . 5 B DL 0 . 8 7 3 1. 7 4 6 0 . 005 0 . 010

Beryllium 0 . 00 2 BDL BDL 0 . 000 0 . 000

Cadmium 0 . 20 BDL BDL 0 . 00 7 0 . 0 35

Chromium 0 . 05 0 . 00 3 BDL 0 . 060 0 . 511 10 . 220

Cobalt 0 . 1 BDL 0 . 5 30 5 . 300 0 . 085 0 . 850

Copper 0 . 1 0 . 003 0 . 058 0 . 030 0 . 580 0 . 735 7 . 350

Gallium 0 . 00 3 BDL 0 . 003

Germanium 0 . 00 3 BDL 0 . 001

Lead 0 . 15 0 . 04 3 BDL 0 . 28 7 0 . 556 3 . 707

Manganese 5 0 . 01 2 0 . 10 3 0 . 00 2 0 . 0 2 1 0 . 198 0 . 040

Mercury 0 . 05 BDL BDL

Nickel 1 0 . 003 BDL 0 . 00 3 0 . 21 2 0 . 212

Selenium 0 . 2 0 . 00 1 BDL 0 . 005 0 . 001 0 . 005

Stront ium 0 . 015 BDL 0 . 151

Tin 2 BDL BDL 0 . 033 0 . 017

Thall ium 0 . 1 BDL BDL 0 . 000 0 . 000

Vanadium 0 . 5 0 . 016 BDL 0 . 032 0 . 009 0 . 018

Zinc *** 1 0 . 080 0 . 160 0 . 080 0 . 160 1 . 240 1 . 240

* S tack concentrat ions are divided by 1000 .

** BDL indicates levels b elow the detect ion limit of the analytical instrument

*** values are for ZnC12

5 1 8

Coal

Al Si S K Ca Fe Ti

TABLE 1 6 PERCENTAGE OF MINOR CONSTITUENTS IN COAL, RDF, BOTTOM ASH AND

HOPPER FLY ASH

FUELS - o Summar� of All 1978 Runs

Ave + Std Dev RDF Ave + Std Dev

1 . 29 + 0 . 17 Al .1 . 36 + 0 . 48 1 . 94 + 0 . 25 S1 2 . 92 + 0 . 61 2 . 94 ± 00. 76 S 0 . 38 + 0 . 10 0 . 09 + 0 . 02 K 0 . 32 + 0 . 07 0 . 77 + 0 . 30 Ca 0 . 49 + 0 . 31 1 . 55 + 0 . 35 Fe 0 . 42 + 0 . 08 0 . 05 ± 0 . 00 Ti 0 . 19 + 0 . 04

ASH - Summar� of All 1978 Runs 80% Load 100% Load

Bottom Ash 0% RDF 10% RDF 20% RDF 0% RDF 10% RDF 20% RDF

Al 7 . 43 + 1 . 47 6 . 18 + 0 . 60 Si 14 . 7 + 1 . 5 21 . 0 + 2 . 9 S 0 . 92 + 0 . 49 0 0. 64 + 0 . 41 K 0 . 75 :;: 0 . 08 0 . 87 :;: 0 . 03 Ca 7 . 44 :;: 3 . 40 7 . 59 :;: 0 . 77 Fe 1 7 . 4 + 1 . 7 8 . 9 :;: 3 . 0 Ti 0 . 48 ± 0 . 07 0 . 5 7 ± 0 . 11

Hopper Ash

Al 8 . 5 7 + 1 . 48 9 . 37 + 0 . 83 Si 15 . 70 + 0 . 95 1 7 . 33 + 0 . 76 S 0 . 95 + 0 . 20 0 . 65 :;: 0 . 12 K 0 . 96 :;: 0 . 13 1 . 23 :;: 0 . 06 Ca 6 . 61 + 3 . 26 5 . 55 + 0 . 73 Fe 1 7 . 6 :;: 2 . 9 4 1 3 . 9 0 :;: 2 . 62 Ti 0 . 59 ± 0 . 08 0 . 71 ± 0 . 07

*On1y one ana1yis of 100% Load - 0%

14

1 2

1 0

B

FUEL ENRI CHMENT FACTOR

AMES MUN I C I PAL POWER PLANT

BOILER UNIT 7 BO PERCENT LOAD

REFUSE DERIVED FUEL HEAT INPUT , PERCENT

5 . 43 + 0 . 43 3 . 8* 5 . 23 + 0 . 21 5 . 40 + 0 . 92 20 . 8 + 1 . 3 11 . 0 24 . 25 + 1 . 50 24 . 76 + 2 . 31 0 . 27 + 0 . 10 4 . 8 0 . 6 7 + 0 . 24 <0 . 5 + 0 . 0 0 . 85 :;: 0 . 02 0 . 450 0 . 83 :;: 0 . 02 0 . 95 :;: 0 . 02 7 . 88 :;: 1 . 02 14 . 1 7 . 22 + 0 . 27 7 . 88 + 0 . 21 6 . 07 + 1 . 87 13 . 8 7 . 29 :;: 0 . 89 7 . 26 + 0 . 45 0 . 52 ± 0 . 05 0 . 16 0 . 42 + 0 . 02 0 . 46 :;: 0 . 06

8 . 10 + 1 . 35 8 . 53 + 2 . 26 9 . 23 + 1 . 31 8 . 73 + 1 . 90 16 . 13 + 1 . 91 16 . 47 + 2 . 35 17 . 50 + 2 . 69 1 7 . 30 + 2 . 85

0 . 84 :;: 0 . 15 1 . 11 + 0 . 71 0 . 85 :;: 0 . 26 0 . 84 :;: 0 . 09 1 . 16 :;: 0 . 10 0 . 95 :;: 0 . 17 1 . 16 :;: 0 . 08 1 . 29 :;: 0 . 02 7 . 32 :;: 1 . 39 10 . 22 + 3 . 68 7 . 82 + 1 . 1 7 9 . 57 + 0 . 33 15 . 0 :;: 4 . 42 14 . 8 7 :;: 2 . 80 1 1 . 1 5 :;: 3 . 12 11 . 79 :;: 1 . 91 0 . 65 :;: 0 . 08 0 . 53 :;: 0 . 11 0 . 71 :;: 0 . 05 0 . 63 :;: 0 . 05

RDF data availab le

Pb

24

1 2r------------------------------------,

1 0

a: o t; B

� � � 6 u

� z: '"'

..J

� 4

FUEL ENRI CHMENT FACTOR

AMES MUN IC IPAL POWER PLANT

BOILER UNIT 7 1 00 PERCENT LOAD

Zn

O�O----�----�----��--��--��--�24 REFUSE DERIVED FUEL HEAT INPUT , PERCENT

FIG. 1 7 TRACE E LEMENT FUEL ENRICHMENT

FACTORS FOR 80 PERCENT LOAD

FIG. 1 8 TRACE ELEMENT FUEL ENRICHMENT

FACTORS FOR 1 00 PERCENT LOAD

5 1 9

[Fe 1 micrograms of iron per Joule of heat energy input

C Subscript meaning coal only at the input

C + RDF = Subscript meaning mixture of coal and refuse derived fuel at the input

Iron was chosen as a reference because its value was measured more precisely than any other ele­ment in the samples of these tests. This is mainly because the quantity of iron in the samples is quite high and significantly above the detection limits or resolving ability of the x-ray fluorescence ap­paratus used for trace element analysis. The ratio of the amount of element to the amount of iron was formed to help eliminate or cancel any calibra­tion factors in the analytical apparatus that would be of a multiplicative nature . This was felt to be of importance because many of the trace elements scanned were only slightly above the detection

limits of the apparatus. Thus, elimination of as many uncertainties as possible in the data led to the use of enrichment factor for presentation of the trace element results.

CONCLUSIONS

Refuse derived fuel, in combination with coal, was successfully fired in a suspension fired boiler with no insurmountable problems after bottom grates were installed. The major result of this pro­ject is that the successful burning of the RDF represents a viable technique for conservation of resources as well as helping to keep good farm­land in production instead of using it for landfill.

Some of the significant items which were ob­served during this study were :

1. The combustible properties of the fly ash and the bottom (grate) ash became quite similar as the RDF approached 20 percent. The softening point of t he ash lowered and the fouling index became more detrimental as the RDF was increas­ed in the fuel input. The fouling impact of RDF at Ames has been recently reduced by a process plant modification which included installation of a de­gritter. This helped remove a large proportion of sand-glass grit rna terial.

2. Uncontrolled particulate emissions tend to increase with corresponding increases in the RDF fraction of fuel input. This appears to be a result of both lighter particulates and increases in air flow through the boiler when burning RDF. A dump grate was installed in this unit to facilitate burning

of RDF particles not remaining in suspension . The RDF injection nozzle location in relation to the coal nozzle was found to be important in affecting emissions. The lowest particulate emissions occurr­ed when RDF was injected below the coal injection point.

3. The oxides of nitrogen (NOx) and oxides of sulfur (SOx) both decrease while chlorides increase significantly with increases in RDF. No discernible trends, within the data scatter, were noted con­cerning formaldehyde or hydrocarbon emissions.

Increased emissions of the trace elements zinc, copper and lead corresponded to increases in RDF. Further studies of the trace element emissions are being performed.

This is the first continuously operating system in the United States using the "fluff' or shredded refuse as a supplemental fuel . Thus, the data provided by this paper as well as in some of the referenced reports herein are the only extensive data available for firing RDF with coal in both suspension fired and stoker fired boilers. The results should provide valuable assistance for those planning further facilities.

ACKNOWLEDGMENTS

This project has been supported through EPA Grant R8039030 1 0 and DOE Contract W-7405-Eng-82 . A major portion of the analytical data reported in this paper was provided by Ames lab­oratory-DOE staff supported by the U.S. Depart­ment of Energy, Contract No. W-7405-Eng-82, Office of Health and Environmental Research , Budget Code GK-01 -02-04-3. The following Ames Laboratory staff played a leadership role in the development of analy

.tical methods and in the

direction of the analytical effort : R. Bachman , E. DeKalb, V. A. Fassel , R. J . Hofer , R. N. Kniseley , and J . Richard.

Support has also been received from the American Public Power Association, the Ames Laboratory-DOE, and both the Engineering Re­search Institute and Mechanical Engineering Department at Iowa State University.

REFERENCES

[ 1 1 "Standards of Performance for New Stationary

Sources" Federal Register, Vol . 36, No. 247, Pt. 1 1 ,

December 23, 1 9 7 1 .

[ 2 1 "Standards o f Performance for New Stationary

Sources. Ammendments to Reference Methods," Federal Register, Vol . 41 , No . 1 1 1 , Pt. 1 1 , June 8, 1 967 .

520

(3) "Standards of Performance for New Stationary Sources," Federal Register, Vol. 38, No. 66, Pt. i 1 , April 6, 1973.

(4) Cooper, John A., "Interpretation of Energy­Dispersive X-ray Spectra," American Laboratory, pp. 35-48, November 1976.

(5) Fassel, Vel mer, A., "Quantitative Elemental Analyses by Plasma Emission Spectroscopy," Science, Vol . 202, No. 4364, pp. 183-191, October 13, 1978.

(6) Carotti, A. A. and Kaiser, E. R., Journal of the Air Pollution Control Association, Vol. 22, p . 249, 1972.

( 7 ) Hal l , J. L., et al., "Environmental Evaluation of the Stoker-Fired Steam Generators ( Part III) : Evaluation

of the Ames Solid Waste Recovery System," EPA Special Report (EPA Grant R80390 3010 ) in review, August 1977.

(8) Ananth, K. P., Shannon, l. J., and Schrag, M. G., "Environmental Assessment of Waste-to-Energy Processes Source Assessment Document," U.S. Environ­mental Protection Agency Document EPA-600/7-77-091, Cincinnati, Ohio, August 1977.

(9) Gorman, P. G., et al., "St. Louis Demonstration Project Final Report : Power Plant Equipment, Faci l ities and Environmental Evaluation," Final Report (EPA Contract No. 68-01-1871 ) , October 1977.

Key Words Boiler

Combustion

Emission

Refuse Derived Fuel

521