METHOD DEVELOPMENT TEST REPORT NO.  2

   U.S.  STEEL - GENEVA COKE WORKS
     COKE OVEN BATTERIES 3 AND 4
            PROVO, UTAH
           SEPTEMBER 1981
            PEDCo ENVIRONMENTAL

-------
      METHOD DEVELOPMENT TEST REPORT NO. 2

         U.S. STEEL - GENEVA COKE WORKS
           COKE OVEN BATTERIES 3 AND 4
                   PROVO, UTAH

                 SEPTEMBER 1981
                 Prepared by

          PEDCo Environmental, Inc,
             11499 Chester Road
           Cincinnati, Ohio  45246
           Contract No. 68-02-3546
                 Task No. 5
                  PN 3530-5
                Prepared for

    U.S. ENVIRONMENTAL PROTECTION AGENCY
         EMISSION MEASUREMENT BRANCH
 EMISSION STANDARDS AND ENGINEERING DIVISION
RESEARCH TRIANGLE PARK, NORTH CAROLINA  27711
          John Brown, Task Manager

-------
                            CONTENTS

                                                            Page

1.0  Introduction                                           1-1

2.0  Test Procedure                                         2-1

3.0  Summary of Results and Statistical Analysis            3-1

Appendices

  A  Field Data Sheets                                      A-l
  B  Project Participants                                   B-l
  C  Test Method 109 Part C (January 1981 Draft)            C-l
                               11

-------
                        1.0  INTRODUCTION






     On September 22-24, 1981, PEDCo Environmental, Inc., person-



nel participated in a series of method development tests at U.S.



Steel Corporation, Geneva Works, Provo, Utah.  The purpose of the



test program was to document the performance of EPA Test Method



109, part C, in determining coke oven door area emissions when



traverses were conducted along the yard and from the bench.



     Results of an earlier test program, designed to determine



the applicability of EPA Method 109, Part C at coke batteries



with coke side sheds, revealed that there was a statistically



significant difference in the number of door area leaks detected



when traverses were conducted along the bench as opposed to,



along the yard.



     This test program was designed to quantify and document any



difference in the number of door area leaks detected by traversing



along the bench as opposed to along the yard.  Batteries 3 and 4



at Geneva Coke Works were selected for this test work.  These



batteries do not have sheds and are similar in construction,



performance, and door leakage history to Batteries 7, 8, and 9 at



Clairton Coke Works which were tested in an earlier program.
                              1-1

-------
     This report presents results of the test series and



describes the test procedures.  The appendix contains a copy of



the field data sheets, list of project participants, and a copy



of the regulation.
                                1-2

-------
                      2.0  TEST PROCEDURE






     EPA Test Method 109, Part C details procedures for deter-



mining emissions from coke oven doors.  These procedures requires



the observer to traverse each side of the battery from ground



level.  For safety reasons, the method recommends the traverse be



conducted outside of the pusher machine and quench car tracks.



Figure 1 presents a diagram of a coke oven battery showing



observer positions for the traverses.  Figure 2 presents a plan



diagram of Batteries 3 and 4 and illustrates observer traverse



position during this test program.



     This test program was designed to quantify and document any



difference in the number of door area leaks detected by tra-



versing along the bench as opposed to along the yard.  In an



effort to obtain accurate data and to minimize the bias created



by process and observer variance, the following test procedures



were used.



     All leak observations made during this development test were



conducted to include door area leaks as defined in the method.



     The test crew was comprised of four members which were



divided into two teams.  Each team traversed the same side of the



battery simultaneously.  One team traversed along the bench while



the other team traversed along the yard.
                              2-1

-------
                       COKE SIDE
   OBSERVER
   TRAVERSE
     ZONE
                            BENCH
                                  3 DOOR
COKE CAR
 TRACK
                                        PUSH SIDE
                                                 »>USH
                                                 SIDE
                                       BENCH
PUSHER
TRACK
OBSERVER
TRAVERSE
  ZONE
Figure 1.  End view  of coke  oven  battery  showing  recommended traverse  zone.
                                          2-2

-------
NJ

U>
                                                    BENCH
                             BATTERY 3
 COAL
BUNKER
BENCH
                                               GAS OFFTAKE PIPES
                                                                        OBSERVER LOCATIONS
                                                                          CARE:
 GUIDE
  CAR

BATTERY 4
                                                     RECYCLE
                                                       WATER
                                                       TANK
                                                  o
                                                 QUENCH
                                                 TOWER
          '   Figure 2.   Plan  view of Batteries  3  and 4,  U.S.  Steel  Corporation,  Provo,  Utah.

-------
     At the end of the traverse, the teams switched positions and



another traverse was conducted.  Two traverses conducted in this



manner constituted a set.



     At the conclusion of each set, the two most inexperienced



members of the test crew switched teams and the next set com-



pleted.



     The testing procedure for the coke side of battery No. 4



follow the above procedure except that the teams remained the



same for all runs.  The hot coke car tracks extend the full



length of Batteries No. 3 and No. 4 so there is no safe place to



cross the tracks.  Plant safety requirements prohibit crossing



the hot coke car tracks so yard and bench observers could ex-



change positions only by walking to the end of the tracks beyond



the quench tower.  Therefore, a set of the coke side consisted of



two runs from the yard by the same team while simultaneous runs



were being conducted from the bench by the other team.  At the



conclusion of a set, the teams switched positions in preparation



for the next set.  No individual observer switches were made.



     The door area leak observation sets were planned to be



conducted at 15-minute intervals to avoid repetitive recording



of the same leaks.  The actual observation times were dependent



upon battery processing operations.  Whenever possible, the ob-



servation traverses were made when the bench was clear.  In those



instances when the pusher, door, or coke cars blocked part of the



battery from the view of the observers the affected ovens were



not read.
                               2-4

-------
        3.0  SUMMARY OF RESULTS AND STATISTICAL ANALYSIS






     Four observers divided into two teams, counted the number of



door area leaks on coke oven Batteries Nos. 3 and 4.  For Battery



No. 3, 7 sets of observations were made from the Yard and the



Bench on the Pusher Side.  On Battery No. 4, 10 sets of observa-



tions were made from the Yard and the Bench on the Pusher Side



and 13 sets of observations were made from the Yard and the Bench



on the Coke Side.



     Tabulations of the number of door area leaks reported by



each observer are presented in Tables 1, 2, and 3 for Batteries



Nos. 3 and 4, respectively.  A summary of these results is



presented in Table 4.  The average number of leaks observed from



the Bench was higher than that observed from the Yard for 29 -of



the 30 sets.  Leaks observed from the Bench were equal to the



leaks observed from the Yard for 1 set.  For observations made



from the Pusher Side the Bench to Yard ratio of the total leaks



for Battery Nos. 3 and 4 were each 1.2.  Because of concerns over



the safety of the observers, the number of leaking doors on the



Coke Side were counted for Battery 4 only.  For the 13 sets of



Coke Side observations the Bench to Yard ratio was 2.2.



     In order to determine the statistical significance of the



difference between the number of leaks observed from the Yard



and the Bench the data presented in Tables 1, 2, and 3



                               3-1

-------
TABLE 1.   NUMBER OF COKE OVEN DOOR AREA LEAKS
          Pusher Side Battery No.  3
           U.S. Steel, Provo, Utah
Set
No.
1
2
3
4
5
6
7
Yard
Observer
1
5
5
8
8
7
10
10
2
8
9
11
10
12
11
11
3
7
8
11
10
9
10
8
4
7
8
11
10
9
11
8
Average
Ave
6.8
7.5
10.2
9.5
9.2
10.5
9.2
9.0
Bench
Observer
1
8
8
14
14
12
11
11
2
9
6
14
15
12
14
12
3
9
8
11
12
11
13
11
4
10
9
11
14
14
10
12
Average •
Ave
9.0
7.8
12.5
13.8
12.2
12.0
11.5
11.3
                    3-2

-------
TABLE 2.  NUMBER OF COKE OVEN DOOR AREA LEAKS
          Pusher Side Battery No.  4
           U.S.  Steel, Provo, Utah
Set
No.
1
2
3
4
9
10
11
12
13
14
Yard
Observer
1
2
2
4
3
3
3
4
2
2
4
2
5
2
6
5
1
3
3
3
2
4
3
4
3
5
4
1
4
3
2
1
3
4
1
2
5
4
2
3
3
2
2
3
Average
Ave
3.0
2.2
5.0
4.0
1.8
3.2
3.2
2.2
1.8
3.5
3.0
Bench
Observer
1
5
4
7
5
2
4
3
2
1
3
2
3
5
5
5
4
5
7
3
2
4
3
3
4
6
3
2
4
5
2
2
4
4
5
3
5
5
2
4
3
3
1
4
Average
Ave
4.0
4.0
5.8
4.5
2.5
4.2
4.5
2.5
1.5
3.8
3.7
                     3-3

-------
TABLE 3.  NUMBER OF COKE OVEN DOOR AREA LEAKS
          Coke Side Battery No.  4
          U.S. Steel, Provo, Utah

Set
No.
5
6
7
8
15
16
17
18
19
20
21
22
23

Yard
Observer
1

1
1

3
3

2
3

1
1

2
1

0
1

2

1
1

3
3
4
1

2
2

1
1

1
2

1
1
3
1
1

1
1

4
1

2
2

1
0

2
2

1
1
4
1
1

1
2


1
3

3
2

2
2

1
1

Average

Ave
1.0
1.0
1.2
3.0
2.5
2.2
2.0
1.8
0.8
1.8
1.8
0.8
1.0
1.6
Bench
Observer
1
4
2

4
5

5
2

4
3

4
6

2
3

2
2
2
4
4

3
3


3
5

4
4

9
8

6
7

3

1
1

5
4

3
3

3
3

4
3

1
1


4

1
1

5
5
4
2

4
4

3
6

2
4

1
2
Average

Ave
3.5
1.0
3.8
4.8
3.2
3.5
3.8
3.5
4.8
6.0
2.8
3.8
1.8
3.6
                     3-4

-------
TABLE 4.   AVERAGE NUMBER OF COKE  OVEN  DOOR AREA LEAKS
                U.S.  Steel, Provo,  Utah
Battery
No.
3
4

Pusher Side
No.
of
sets
7
4
6
10
Date
of
tests
9/24/81
9/22/81
9/23/81

Average
No, of leaks
Yard
9.0
3.6
2.6
3.0
Bench
11.3
4.6
3.2
3.7
Bench
Yard
1.2
1.3
1.2
1.2
Coke Side
No.
of
sets

4
9
13
Date
of
tests

9/22/81
9/23/81

Average
No. of leaks
Yard

1.6
1.6
1.6
Bench

3.2
3.7
3.6
Bench
Yard

2.0
2.3
2.2
                          3-5

-------
were further investigated using Analysis of Variance.  The model

was of the form;


     Yijk = " + Xi + 1i(j) + eijk

where

     y     = the overall mean number of leaks

     X.    = the effect due to the position on the side from
       1      which observations were made  (i.e., Yard, Bench)

     n-/-v = the effect due to the variation of the coking
         •^    process

     e..,  = the random error associated with an observation
       1^     (this is the observer error)

     The results of the analysis of variance of the data for

these  observations are presented in Tables 5, 6, and 7.  Separate

analyses were performed for observations made from the Pusher

Side and the Coke Side, because of the difference in the test

procedures.

     For Battery No. 3 the average number of door area leaks

observed from the Yard and Bench on the Pusher Side were 9.0 and

11.5 respectively.  The results of the Analysis of Variance  for

these  data are presented in Table 5.  The "Between position"

effect is statistically significant (i.e., p < 0.01).  For these

data the estimate of the observer error  (i.e., variance) of  an

individual observation is 2.01.

     Because of the difference in field procedures for observa-

tions  made from the Coke Side of Battery No. 4, separate

Analysis of Variance were performed for the Pusher Side and  Coke

Side results.  As seen in Table 6, for the Pusher Side, the
                               3-6

-------
TABLE 5.  ANALYSIS OF VARIANCE OF COKE OVEN DOOR AREA LEAKS
                         Pusher Side
                        Battery No.  3
Source of variation
Between position
Between set within
position
Observer error
Total
Degrees
of
freedom
1
12
42
55
Sum
of
squares
70.87
153.0
80.25
304.12
Mean
square
70.87
12.75
2.01

F-ratio
5.56
6.34


P
<0.01
<0.01


                          3-7

-------
TABLE 6.   ANALYSIS OF VARIANCE OF COKE  OVEN  DOOR  AREA  LEAKS
                         Pusher Side
                        Battery No. 4
Source of variation
Between position
Between set within
position
Observer error
Total
Degrees
of
freedom
1
18
60
79
Sum
of
squares
10.52
92.97
51.00
154.49
Mean
square
10.52
5.17
0.85

F-ratio
2.03
6.08


P
>0.10
<0.01


                          3-8

-------
TABLE 7.   ANALYSIS OF VARIANCE OF COKE  OVEN  DOOR AREA LEAKS
                          Coke Side
                        Battery No.  4
Source of variation
Between position
Between set within
position
Observer error
Total
Degrees
of
freedom
1
24
78
103
Sum
of
squares
129.43
70.85
103.25
303.53
Mean
square
129.43
2.95
1.32

F-ratio
43.87
2.23


P
<0.01
<0.01


                          3-9

-------
"Between position" effect is not statistically significant (i.e.,

p < 0.10).  Referring again to Table 4, note that the average

number of leaks observed from the Pusher Side was 3.0 from the

Yard and 3.7 from the Bench.  The estimate of the observer error

for these results was 0.85.

     The results of the Analysis of Variance of the data for the

Coke Side (Battery No. 4) are presented in Table 7.  The "Between

position" effect is statistically significant (p < 0.01).  Refer-

ring to Table 4 the average number of leaks observed from the

Bench was 3.6 compared to 1.6 for the Yard.  The estimate of the

observer error was 1.32.

     The test results for each Battery provide an estimate of the

error of an individual observation.

     Battery No.            Variance          Standard deviation

          3                   2.01                   1.42
          4 Pusher Side       0.85                   0.92
            Coke Side         1.32                   1.15

The difference in observer variance between the three sets of

results is statistically significant (p < 0.05).

     Based upon Bartlett's Test* the variation between these

individual estimates of the error of an individual observation is

statistically significant  (p < 0.05).

     The Bench to Yard ratio of total leaks for Batteries 3 and

4 were 1.2 on the Pusher Side and 2.2 on the Coke Side.  The

Coke Side having a higher Bench to Yard ratio, at this plant, is

a reversal of the trend seen at Batteries 7, 8, and 9 at the
*
 Bartlett, M.S.  Journal of the Royal Statistical Society Supple-
 ment, 4:137(1937).

                              3-10

-------
Clairton works.  Several factors were observed during the testing



of Batteries 3 and 4 which could have an effect on observer per-



formance.  These factors consist of the following:



     Batteries 3 and 4 were much less congested than Batteries 7,



8, and 9.  With more open area around the batteries there ap-



peared to be less interference from coke breeze and dust on the



Pusher Side.



     Steam from the coke wharf provided a constant interference



for the observers when traversing along the Coke Side of the



Battery 4.
                              3-11

-------