United States
Environmental Protection
Agency
Office of Air Quality
Planning and Standards
Research Triangle Park NC 27711
EMB Report 78-CUS-10
March 1979
Air
Arsenic
Non-Ferrous Smelters

Emission Test Report
Phelps-Dodge Copper
Smelter
Playas, New Mexico

-------
EMISSION TESTING OF PHELPS-DODGE  COPPER SMELTER

              PLAYAS, NEW  MEXICO
                      TO
        ENVIRONMENTAL  PROTECTION  AGENCY

             Contract  #68-02-2812
             Work Assignment  #15
                April  20,  1979


                       By

                 Thomas Rooney
                     TRW
          ENVIRONMENTAL ENGINEERING DIVISION

    One Space Park,  Redondo Beach,  CA  90278

-------
                             CONTENTS

Figures	   iii
Tables	   iii
     1.   Introduction	    1
     2.   Summary and Discussion of Results	    2
     3.   Process Description	    6
     4.   Location of Sampling Points	    7
     5.   Sampling and Analytical  Procedure	    9
6.   Appendices
     A.   Field and Laboratory Data	    14
          1.  Traverse Point Location	    14
          2.  Field Data Sheets	    16
          3.  Analytical  Data Sheets	    23
          4.  Particle Sizing Analysis 	    27
          5.  Meter Box Calibration Data Sheets	    34
     B.   Sample Calculations	    36
     C.   Daily Activity Log	    44

-------
                                   FIGURES

Number                                                                 Page

  1   Converter fugitive emission duct schematic	   3

  2   Arsenic/sulfur dioxide sampling train  	   10

  3   Brinks impactor sampling schematic	   13
                                   TABLES

Number                                                                 Page

  1   Converter Fugitive Emission -
     Arsenic/Sulfur Dioxide Results	   3

  2   Particle Sizing Summary 	  4

  3   Process Sample Analysis 	  5
                                    TM

-------
                                   SECTION 1

                                 INTRODUCTION
     In conjunction with the Environmental  Protection Agency's Program for
developing new source performance standards, TRW performed fugitive emission
tests at the Phelps-Dodge Cooper Smelter located in Playas,  New Mexico.
The test was conducted July 24-27, 1978.

     The process tested was secondary converter hooding system which removed
fugitive emissions from the converter during the slag and  copper blow  cycles.

     The testing consisted of three arsenic/sulfur dioxide tests and three
particle sizing tests which were performed  during  the copper and slag blow
cycles.  The testing location was a seven foot duct located  between the  hooding
system and the stack.  These tests were coordinated with a process  engineer
from the Environmental Protection Agency.

     This report presents the results of the testing program.   The  following
sections of the report contain a summary of the results, description of
sampling points, description of the process, and the sampling procedure  with
the laboratory procedure.  The appendices contain  field data,  laboratory
data, sample calculations, and the daily activity  log.

-------
                                  SECTION  2


                        SUMMARY AND DISCUSSION OF RESULTS
     The results of converter hooding fugitive emission system tests are
summarized in Table 1  and Table 2.  Table 1  consists of Field, Laboratory
and Emission Data.  Table 2 contains particle sizing data from the three
tests.

     During the testing program the following observations and problems
were noted.

     For the first test, twenty-five minutes per sampling point was used to
assure that sampling was done through a complete production cycle.  For
the second and the third test, twenty minutes per sampling point and a
smaller nozzle size was utilized.  After 155 minutes of the third test,
TRW personnel noticed that the AP readings  were abnormally low.  After
checking equipment, the process engineer discovered that the plant opera-
tors inadvertently left the dampers on the system in the open position.  When
the problem was corrected,the AP reading increased to the appropriate reading.
Thus, during 80 minutes of the sampling period of the third test, dilution
air entered the duct which resulted in a non-representative sample.

     During the Data Reduction, the meter volume was back calculated to
account for sulfur dioxide that was removed  by the three 10% hydrogen peroxide
impingers.  The back calculation for sulfur  dioxide was accom-
plished in the following order.  First, parts per million sulfur dioxide
at standard conditions was calculated.  Then parts per million was converted
to a fraction by dividing by 10^.  This number was added to one and the
result multiplied by the volume of gas collected through the dry gas
meter at standard conditions.  The result of multiplication yielded the true
gas volume collected at standard conditions.  Since S02 removal by the peroxide
impingers does not reach the dry gas meter,  corrected values for dry gas meter
volumes (at meter conditions) found on the summary sheets will be slightly
higher than those obtained from the field data sheets.

-------
TABLE 1  CONVERTER FUGITIVE EMISSIONS ARSENIC/S02 RESULTS

RUN NUMBER


! DATE
II STACK PARAMETERS
PST - STATIC PRESSURE, "Ho (MMHG)
Ps - STACK GAS PRESSURE, "He ABSOLUTE (MMHG)
I C02 - VOLUME J DRV
I (>2 - VOLUME !! DRY
SOj - VOLUME 1 DRY
J N2 - VOLUME X DRV
1 H20 - X MOISTURE IN STACK GAS, BY VOLUME
As - STACK AREA, FT (M*)
Mo - MOLECULAR WEIGHT OF STACK GAS, DRY BASIS
Ps - MOLECULAR HEIGHT OF STACK GAS, WET BASIS
Vs - STACK GAS VELOCITY, FT/SEC, (M/S;C)
QA - STACK GAS VOLUMETRIC FLOW AT STACK CONDITIONS, ACFI" (NM /MIN)
Qs - STACK GAS VOLUMETRIC FLOW AT STANDARD CONDITIONS, DSCFM (NM /MIN)
I EA - PERCENT EXCESS AIR
III TEST CONDITIONS
PB - BAROMETRIC PRESSURE, "He (MMHG)
DN - SAMPLING NOZZLE DIAMETER, IN. (MM)
T - SAMPLING TIME, MIN
VN - SAMPLE VOLUME, ACF (M3>
CP - PITOT TUBE COEFFICIENT
TM - AVERAGE METER TEMPERATURE °F (°C)
PM - AVERAGE ORIFICE PRESSURE DROP, "h^O (MMHnO)
VLC - CONDENSATE COLLECTED (IMPINGERS AND GEL), MLS
&p - STACK VELOCITY HEAD "I^O (MMH20)
IV TEST CALCULATIONS
V» - CONDENSED VATER VAPOR, SDCF (NM3)
VM - VOLUME OF GAS SAMPLED AT STANDARD CONDITIONS, DSCF (NM5)
? H20 " PERCENT MOISTURE, BY VOLUME
Ms - MOLECULAR WEIGHT OF STACK GAS, WET BASIS
Vs - STACK VELOCITY, FT/SEC (M/SEC)
I I - PERCENT ISOKINETIC
V ANALYTICAL DATA
A) ARSENIC FRONT HALF
PROBE (MG)
CYCLONE (MG)
FILTER (MG)
ARSENIC FRONT HALF TOTAL (MG)

PF-M, («G/M3)
''MR, (KG/HR)
B) ARSENIC - IMPINGER COLLECTION
IMPINGER #1. 2 (MG)
PPM, (MG/M3)
»/HR, (KG/HR)
IflEittsEBj!i,S,5 (MG)
PPM, MG/M3)
»/HR, (KG/HR)
C) ARSENIC - IMPINGER TOTAL (MG)
PPM, (MG/M3)
#/HR, (KG/HR)
D) TOTAL ARSENIC (MG)
PPM, (MG/M^)
#/HR, (KG/HR)
E) TOTAI SOo IMGJ
PPM
(MG/M3)
#/HR, (KG/HR)
1

ENGLISH
UNITS
7/25/78

- .24
25.60
.2
20.2
.38
79.22
216.7
1.5
35.34
28.97
28.95
49.29
104514.5
69076.1
4.7

25.84
.250
300.
221.95
12
.84
137.9
1.59
.516

2.65
169.97
1.5
28.95
49.29
85.3



-



.1272
.1026

.2832
.1921




.2832
.1921
-
.3654
.2947
-


2649.4671
METRIC
UNITS
7/25/78

-6.10
650.24
.2
20.2
.38
79.22
102.6
1.5
3.28
28.97
28.95
15.02
2960.8
1956.8
4.7

656.34
6.35
300.
6.29
.84
58.8
<0.39
13.11

.08
4.82
1.5
28.95
15.02
85.3

.71

•
1.20

1 .91
.3967
.0466
3.577
.7429
.0872


-
3.577
.7429
.0872
5.4870
1.1396
.1338
49329.6
3845.4857
10244.954
1202.6632
2

ENGLISH
UNITS
7/26/78

- .24
25.55
.2
20.2
.48
79.12
207.5
2.2
35.34
29.01
28.77
59.49
126163.4
83315.9
4.7

25.79
.185
240.
131.19
.84
104.4
.78
.756

2.43
105.98
2.2
28.83
59.44
98.6




-


.0737
.0718

.0062
.0060




.0062
.0060

.0799
.0778

•
-
4012.9187
METRIC
UNITS
7/26/78

-6.10
648.97
.2
20.2
.48
79.12
97.5
2.2
3.28
29.01
28.77
18.14
3574.0
2361 .1
4.7

655.07
4.70
240.
3.72
.84
40.2
19.81
51 .7
19.20

.07
3.00
2.2
28.83
18.12
98.6

.09

-
.60

.69
.2298
.0326
.058
.0193
.0027
.


.058
.0193
.0027
.7480
.2491
.0353
38570.4
4822.2099
12847.0950
1821.5700
3

ENGLISH
UNITS
7/26/78

- .24
25.55
.2
20.2
1.10
78.50
216.1
2.2
35.34
29.23
28.98
40.54
85957.9
56063.1
4.7

25.79
.185
240.
90.36
.84
114.0
.38
.349

1.63
71.69
2.2
29.05
40.49
99.1




-


.1231
.0807

.0242
.0158

_
-

.0242
.0158
-
.1473
.0965
-

-
6206.6644
METRIC
UNITS
7/26/78

-6.10
648.97
.2
20.2
1.10
78.50
102.3
2.2
3.28
29.23
28.98
12.36
2435.1
1588.2
4.7

655.07
4.70
240.
2.56
12.
.84
45.6
9.65
8.86

.05
2.03
2.2
29.05
12.34
99.1

03

-
.75

.78
.3841
.0366
.153
.0753
.0072
_


.153
.0753
.0072
.9330
.4594
.0438
60003.6
11090.078
29545.642
2817.3692
AVERAGE

ENGLISH
UNITS
7/26/78

- .24
25.57
.2
20.2
.65
78.95
213.4
2.0
35.34
29.07
28.90
49.77
105545.3
69495.0
4.7

25.81
.207
260.
147.83
12.
.84
118-8
.92
36.7
.540

2.24
115.88
2.0
28.94
29.74
94.3




-


.1080
.0650

.BB95
.0713

_
-
.
.0895
.0713

.1975
.1563
-

.
4289.6334
METRIC
UNITS
7/26/78

-6.10
649.39
.2
20.2
.65
78.95
100.8
2.0
3.28
29.07
28.90
15.17
2989.95
1968.7
4.7

655.49
5.25
260.
4.19
12.
.84
48.2
23.28
33.5
13.72

.07
3.28
2.0
28.94
15.16
94.3


.28
"
.85

1.1267
.3369
.0386
1 2627
.2792
.0324

_
.
1.2627
.2792
.0324
2.3894
.6160
.0710
49301.2
6585.9245
17545.897
1947.2008

-------
    TABLE 2.  PARTICLE SIZING SUMMARY
(LOCATION-PHELPS DODGE-PLAYAS,  N.  MEXICO)

LOCATION
CONVERTER
CONVERTER
CONVERTER
TEST
ONE
TWO
THREE
PARTICLE SIZE DISTRIBUTION %
>5p
24.0
19.0
38.0
3-5y
8.0
13.0
6.0
l-3y
19.0
32.0
14.0
< in
49.0
36.0
42.0

-------
                     TABLE 3.
              PROCESS SAMPLE ANALYSIS
       SAMPLE
DATE SAMPLED
AS
Flash Furnace Feed        7/26/78 - 7/27/78    .0132%
Flash Furnace Slag        7/26/78 - 7/27/78    .0032%
Flash Furnace Matte       7/26/78 - 7/27/78    .0051%
Electric Furnace Matte    7/26/78 - 7/27/78    .0088%
Electric Furnace Slag     7/27/78 - 7/27/78    .0120%
Converter Slag            7/27/78 - 7/28/78    .0072%
Converter Blister         7/27/78 - 7/28/78    .0139%

-------
  SECTION  3






PROCESS DESCRIPTION

-------
                                  SECTION 4


                          LOCATION OF SAMPLING POINT
Outlet from Converter Hooding System

     Samples from converter hooding system were taken from a seven foot
diameter horizontal duct located approximately 70 feet above the ground.
The sampling ports on the top and side of the duct allowed for vertical
and horizontal  traverses during sampling.  The nearest upstream flow
disturbance was 7 duct diameters from the sampling location.  The nearest
downstream flow disturbance was greater than ten duct diameters from
the sampling location, where there was a 90° bend.  Twelve traverse points,
six on each traverse, were used.  Sampling was done for twenty minutes per
point to provide sampling through a complete slag  and copper blow cycle.
Figure 1 illustrates the cross-sectional view.

-------
                                             Traverse  point  locations
Tra-
verse
Point
Loca-
tions
1
2 '
3
4
5
6
Fraction of
Stack I.D.
.044
.146
.296
.704
.854
.956
Distance
From Inside.
Wall (in)
3.66
12.30
24.85
59.15
71.70
80.34
Stack
  From
converter
  hoods
            Figure  1.  Converter  fugitive emission duct

-------
                                  SECTION 5

                       SAMPLING AND ANALYTICAL PROCEDURE
A.  Arsenic/Sulfur Dioxide Sampling

      The sampling train used for arsenic/sulfur dioxide collection consists
of an EPA method 5 train modified by adding two additional  impingers in series
to the four used in the method 5 train.   The first two impingers contained
150 milliliters of distilled water each, the third impinger was empty,  the
fourth,fifth and sixth impingers contained 150 milliliters  of 10% hydrogen
peroxide each.  The seventh impinger contained 250 grams of silica gel.
Sampling train schematic is presented in figure 2.

      Before each test a velocity traverse of the stack was done to determine
the average stack temperature and velocity pressure.   The velocity traverse
was done according to EPA Methods 1 and 2.  A grab sample of the stack  gas
was taken and analyzed with a Fryite apparatus for C02.  Before the first
test at each location the moisture content of the gas stream was estimated
by either condensation in impingers as in EPA Method  4, or by wet and dry
bulb thermometer if the stack gas temperature was below 1200F.

      The arsenic/sulfur dioxide samples were taken at traverse points  at
the center of equal areas within the stack.  The number of traverse points
was determined by the number of duct diameters upstream and downstream  from
the nearest flow disturbances.  The sampling rate was adjusted to isokinetic
conditions using a nomograph which had been set based on the preliminary
velocity traverse data, and moisture estimate.

      The sampling time per traverse point was 20-25  minutes, to assure
sampling during the whole process cycle.

      Leak checks of the sampling train were done at  the beginning of each
test, just before the sampling port change, and at the end of the test.
At the end of each test the sampling train was inspected for cracked or
broken glassware, and to assure that the filter remained intact.

Sample Recovery

      The sampling nozzle and probe liner were rinsed with 0.1N NaOH and
brushed out with a nylon bristle brush with a teflon  tubing handle.  The
remainder of the sampling train was removed to the mobile laboratory.  The
front half of the filter and connecting glassware were rinsed with 0.1N
NaOH and this rinse was added to the nozzle and probe rinse.  The filter was
removed from the filter holder and placed in a polyethylene container,  which
was labeled and sealed.  The first three impinger solutions were measured and
placed in a glass sample container along with a 0.1N  NaOH rinse of the  im-
pingers.  The contents of the fourth, fifth, and sixith impingers were  measured
and placedin a separate glass sample container along  with a distilled
water rinse of the impingers.  The silica gel in the  seventh impinger was
weighed to the nearest 0.5 grams, and regenerated.

-------
                                                                             12
                                                                                  13
                                                                     17
                           Figure 2.  Arsenic sulfur dioxide train
1.   Calibrated Nozzle
2.   Heated Probe
3.   Type S Pitot
4.   Cyclone Assembly
5.   Filter Holder
6.   Heated Box
1.   Ice Bath with Impingers
a.   150 mis H20
b   150 mis H20
           KEY
 c.   Empty
 d.   150 mis  H202
 e.   150 mis  H202
 f.   150 mis  H202
 g.   250 gms. Si lea  Gel
12.   Thermometer
13.   Check Valve
14.   Vacuum Line
15.   Vacuum Gauge
16.
17.
18.
19.
20.
21.
22.
Main Valve
Air Tight Pump
By-Pass Valve
Dry Test Meter
Orifice
Pitot Manometer
Thermometer
                                             10

-------
B.  Analysis

      The samples were analyzed for sulfur dioxide by taking an aliquot of
the hydrogen peroxide impinger solutions and titrating with barium perchlorate
solution and thorin indicator as described in EPA Method 6 (Determination  of
Sulfur Dioxide Emissions from Stationary Sources).

Arsenic-Analysis

      1.  Filter - warm filter and loose particulate matter with 50 ml
0.1N NaOH for about 15 minutes.  Add 10 ml concentrated HMOs and bring  to
boil for 15 minutes.  Filter solution through no. 41  Whatman paper and
wash with hot water.  Evaporate filtrate, cool,  redissolve in 5 ml  of 1:1
HN03, transfer to a 50 ml  volumetric flask and dilute.

      2.  Probe Wash and Impinger Solns -  These should be combined and a
200 ml sample withdrawn.  Add 10 ml concentrated HN03 and evaporate to  a
few milliliters.  Redissolve with 5 ml  1:1 HN03  and dilute to 50 mis.  A
reagent blank should be carried through the same procedure.  The resulting
blank solution should be used in the dilution of standards to matrix match
samples and standards.

      3.  All the samples  prepared above should  be screened by air/acetylene
flame.  The filter samples may require dilution  with 0.8N HNOv  Impinger
solutions containing more  than 25 mg/1  of arsenic should be diluted since
linearity decreases dramatically above that level.

      Since an entrained hydrogen flame provides about five times as much
sensitivity as the air/acetylene flame, a matrix check of a sample in a
hydrogen flame should be carried out by the method of standard additions,
and compared with a value  obtained from matrix matched standards in a
hydrogen flame.  If values are comparable (+5%)  the air entrained hydrogen
flame value should be used.

     Due to high concentrations of copper on the filter an air/acetylene
flame should always be used to dissociate any AsCu compounds stable in
the cooler hydrogen flame.

      4.  For samples below the lmg/1 level, hydride generation is necessary.
An appropriate aliquot of digested sample in 0.8N HN03 containing less  then
about lOug of arsenic is chosen (some screening  may be necessary).   Five mini-
liters of concentrated H2S04 is added to the sample which is then placed on  a
hot plate until S03 fumes  fill the flask. A reduction in volume to about 5 ml
or less may be necessary.   This step removes HN03 which causes a violent
                                       11

-------
reaction when the reducing agent is added resulting in poor reproducibility
and lowered senitivity by producing 12, N02 and possibly other species.

      One millilHer of 30% KI and 1 ml of 30% SnCl2 are added to the sample,
the former to act as catalyst in hydride formation and the latter to reduce
all the arsenic to As+3.  The sample is then diluted to about 15 ml, and  15 ml
of concentrated HC1 is added.  Powdered Zn (or NABH4) is then added, the
reaction vessel is immediately closed and the nitrogen or argon carrier
flow initiated.  A peak should be produced within a few seconds.

C.  Particle Sizing

      The size distribution of the particulates was estimated with a
Brinks six stage impactor. Figure 3 is a diagram of the Brinks impactor
sampling system used.

Sampling Procedure

      The Brinks impactor was introduced into the gas stream through the
sampling  port with the nozzle facing the flow of gas.  The sampling pump
was turned on and the pressure drop across the impactor adjusted with the
by-pass valve.  The pressure drop  across the impactor was read from the
mercury manometer.  The pressure drop is proportional to the flowrate
through the impactor, and to the particle sizing cutoffs of each stage.

      Sampling time at each location varied according to grain loading in
the particular duct being sampled.  The impactor plates were inspected after
each test and the sampling time altered on the succeeding test to optimize
the amount of particulate sampled.  Sampling for too long results in carry-
over from one stage to the next, while sampling for too  short a time can
result in insufficient particulate on one or more of the stages for accurate
analysis.

Analysis

      The impactor plates and filters had been dessicated to a constant weight
before the tests, and tare weights taken.  After the test the same procedure
was used to get the final weights of the impactor plates and filters. The
difference between the tare weight and final weight is the weight of parti-
culate collected.

      The cummulative percentage of the total particulate catch which was
collected in each stage was plotted on semi-log graph paper against the
size cutoffs for each stage.  The resulting best fit straight line is the
estimated particle size distribution of the collected particulatesJ
1 Brink, J.A. "Cascade Impactor for Adiabatic Measurements," Industrial
  and Engineering Chemistry, Vol.-5, No. 4, April  1958,  page 647
                                        12

-------
BRINKS  IMPACTOR

      47 MM  GLASS  FIBER FILTER
               SHUT OFF
                 VALVE
•BY PASS  VALVE
         MERCURY
        MANOMETER
                        PUMP
     DRY GAS
      METER
                                                ORIFICE
                                                 MANOMETER
  Figure 3.  Brinks impactor particle sizing system schematic
                            13

-------