tf/EPA
           United States
           Environmental Protection
           Agency
          Office of Air Quality
          Planning and Standards
          Research Triangle Park NC 27711
EMBReport81-GRA-4
           Air
Letterpress/Off set
Lithographic Industry
Condenser/ESP

Emission Test Report
Lehigh Steck-Warlick
Dallas, Texas

-------
            0 EMISSION TEST REPORT °

       METHOD DEVELOPMENT AND TESTING FOR
    LETTERPRESS/OFFSET LITHOGRAPHIC INDUSTRY
          Lehigh Steck-Warlick Company
                  Dallas, Texas
                   ESED 80/48
                     by

          PEDCo Environmental, Inc,
             11499 Chester Road
           Cincinnati, Ohio  45246
           Contract No. 68-02-3546
            Work Assignment No. 7
                Task Manager
                 Frank Clay
    U.S. ENVIRONMENTAL PROTECTION AGENCY
     EMISSION MEASUREMENT BRANCH, MD-13
 EMISSION STANDARDS AND ENGINEERING DIVISION
RESEARCH TRIANGLE PARK, NORTH CAROLINA  27711
                   May 1981

-------
                            CONTENTS
Figures                                                      ii-1
Tables                                                        iv
QA Element Finder                                              v

1.   Introduction                                            1~1

2.   Summary of Results                                      2-1

     2.1  Emission test results                              2-1
     2.2  Process data                                       2-i:

3.   Process Description                                     3-1

4.   Sampling and Analytical Procedures                      4-1

     4.1  Description of sample locations                    4-1
     4.2  Gas stream flow rate and composition               4-5
     4.3  Particulate and hydrocarbon emission tests         4-5
     4.4  Gas stream isopropanol content                     4-7
     4.5  Process sample analysis                            4-8

5.   Quality Assurance                                       5-1

Appendices

  A  Computer Printouts and Example Calculations             .A-l
  B  Field Data                                              B-l
  C  Laboratory Results                                      C-l
  D  Sampling and Analytical Procedures                      D-l
  E  Calibration Procedures and Results                      E-l
  F  Quality Assurance                                       F-l
  G  Project Participants and Project Log                    G-l
                               11

-------
                             FIGURES

Number                                                      Page
 2-1      Relationship Between Measured Ink Volatile
            Content and Heating Time in Analysis            2-21

 4-1      Inlet Test Site and Traverse Point Locations      4-3

 4-2      Outlet Test Site and Traverse Point Locations     4-4
                               iii

-------
                             TABLES

Number                                                      Page

 2-1      Velocity Traverse Data                            2-2

 2-2      Gas Stream Conditions Measured During
            Particulate and Hydrocarbon Tests               2-4

 2-3      Method 25 Data Summary                            2-5

 2-4      Isopropanol Content of the Emission Stream        2-9

 2-5      Method 5 Data Summary                             2-12

 2-6      Press Operation Summary for the Lehigh
            Steck-Warlick Company                           2-14

 2-7      Press Process Data Summary                        2-16

 2-8      Control Equipment Operation                       2-17

 2-9      Ink Usage at the Lehigh Steck-Warlick
            Company                                         2-19

 2-10     Ink Density and Percent Volatile Content          2-20

 2-11     Residual Solvent Content of Printed
            Signatures                                      2-24
                               IV

-------
                        QA ELEMENT FINDER



     Item



Project Description



Project Organization and Responsibility



QA Objectives



Sampling Procedures



Sample Custody



Calibration Procedures



Analytical Procedures



Data Validation, Reduction, and Reporting



Internal QA Checks



Performance Systems Audits



Preventive Maintenance Schedule



Corrective Action



Quality Assurance Reports to Management
   Page



   1-1



Appendix G



Appendix F, 5-1



Appendix D, 4-1



Appendix B & C



Appendix E



Appendix D, 4-1



Appendix F



Appendix F



Appendix F



Appendix F



Appendix F



Appendix F

-------
                            SECTION 1



                          INTRODUCTION






     During the week of March 23, 1981, personnel from PEDCo



Environmental, Inc., conducted emission tests at the offset



lithographic printing plant of Lehigh Steck-Warlick Company in



Dallas, Texas.  Tests were conducted to determine the efficiency



of a condenser/electrostatic precipitator (ESP) control system



used to control hydrocarbon emissions from the press dryer.



     Hydrocarbon sampling was conducted by means of an Environ-



mental Protection Agency  (EPA) Method 5 sample train with a tee



at the back of the heated filter from which a Method 25 sample



was extracted.  Samples were collected isokinetically, and



Methods 5 and 25 sampling procedures were followed in the opera-



tion of the separate portions of the sample train.  This pro-



cedure was used in order to limit bias that could result from any



aerosol or particulate matter present in the gas streams.



     During the testing, process data were monitored and several



process samples were collected.  Process samples included foun-



tain solutions analyzed for percent isopropanol, process ink



analyzed for density and percent volatile content, and printed



signatures analyzed for the quantity of ink solvent remaining in



the finished product.
                               1-1

-------
     During this week, numerous problems were encountered in the

press operation.  Testing was conducted on only two days, and

both tests were run with the press at a lower than normal

operating speed.  Press speeds during the two tests were 259 and

305 m/min (850 and 1000 ft/min).   Normal operating speed ranges

from 427 to 488 m/min  (1400 and 1600 ft/min).

     Because these tests were run at lower than normal press

speeds, the data are not representative and cannot be used to set

standards for the industry.  While the EPA test was in progress,

a similar test was conducted by Pollution Control Science, Inc.

(PCS).   This effort was sponsored by the Sun Chemical Corporation

and the Graphic Arts Technical Foundation and consisted of

filtered Method 25 samples taken from a single point in the inlet

and outlet ducts.  The samples were taken at the beginning and

end of the EPA runs, which means that there is no corresponding

PCS data for the middle portion of the EPA test.

     The following individuals were present to observe the sam-

pling program:

     Mr. Frank Clay - EMB Task Manager
     Mr. Theodore Michaelis - Engineering Science Process Officer
     Mr. J.D. Boehlert - Sun Chemical Corporation
     Mr. J.L. Zburovsky - Sun Chemical Corporation

     These data will be evaluated for use in a Control Techniques

Guideline document for the letterpress/offset lithographic print-

ing industry.
                              1-2

-------
                            SECTION 2



                       SUMMARY OF RESULTS






2.1  EMISSION TEST RESULTS



Gas Stream Flow Rate and Composition



     Table 2-1 lists gas stream temperatures and flow rates that



were measured by EPA Methods 1* and 2* prior to particulate



sampling.  Gas stream temperatures at the inlet test site aver-



aged 150°C (302°F) and 161°C (322°F) during the first two days.



After these measurements were taken, it was discovered that a



damper controlling the amount of dilution air added to the dryer



discharge had been inadvertantly left closed.  Plant personnel



corrected the problem, and in subsequent velocity traverses the



recorded inlet temperatures were 73°C (163°F) and 86°C (186°F).



The outlet gas temperature dropped from 42°C (107°F) on the first



day to 30°C (86°F) and 32°C (90°F) on the following days.  This



temperature drop was due to adjustments made on the control



system by a representative of United Air Specialists, who was on



hand to inspect the condenser/ESP operation.



     Gas flow rates were adjusted to standard conditions based on



the assumed moisture contents indicated in Table 2-1.
 Federal Register, Vol. 42, No. 160, August 18, 1977.
                                2-1

-------
                     TABLE 2-1.   VELOCITY TRAVERSE  DATA
Date,
1981
3/24
3/25
3/26
3/28
Time,
24-h
1035
0730
1400
0830
Gas stream temperature
Inlet
°C
150
161
73
86
°F
302
322
163
186
Outlet
°C
42
30
32
32
°F
107
86
90
89
Gas stream flow rate
standard conditions9
Inlet
dscmh
7,000
6,826
9,664
9,893
dscfh
247,200
241,080
341,280
349,380
Outlet
dscmh
10,433
9,778
12,550
12,441
dscfh
368,460
345,300
443,220
439,380
Standard conditions:   760 mm Hg,  20°C (29.92  in.  Hg,  68°F),  adjusted
to dry basis with the following assumed moisture contents:
3/24, 25/81:  Inlet,  3.5%; Outlet,  3.0%;  3/26/81:   Inlet, 3.0%;  Outlet  3.0%;
3/28/81:   Inlet,  3.0%; Outlet,  1.5%.
                                    2^2

-------
     Table 2-2 lists the gas stream flow rate and composition


measured during the particulate and hydrocarbon sampling.  At the


inlet site, the flow rate was measured at 10,275 dry standard


cubic meters per hour (dscmh)  (362,839 dscfh) during Test 1 and


10,296 dscmh during Test 2.  Gas temperatures were 73° and 81°C


(163° and 178°F), respectively.  Orsat analysis at the inlet site


showed 0.4 percent carbon dioxide and 20 percent oxygen.  At the


outlet site, the flow rate was measured at 12,999 dscmh  (459,065


dscfh) for Test 1 and 11,751 dscmh (414,984 dscfh) for Test 2.


Outlet temperatures were 32° and 34°C (89° and 93°F) for the two


tests.  Orsat analysis showed 0.2 percent carbon dioxide at the


outlet site.  The difference between the inlet and outlet flow


rates is due to air leakage around the fan which was located


after the ESP and before the outlet sampling location.  Some


leaks were sealed for the second test.


     All calculations for hydrocarbon and particulate emission


rates were based on flows measured during the test.  These data


are presented in Table 2-2.


Hydrocarbon Emissions from Method 25 Analysis


     Each test at the inlet and outlet of the condenser/ESP


consisted of three separate particulate traverses.  During each


traverse, a Method 25* sample was collected from the back of the


Method 5** filter for hydrocarbon analysis.  Table 2-3 summarizes


the data from the Method 25 sampling.
 *
  Federal Register, Vol. 45, No. 194, October 3, 1980.
**
  Federal Register, Vol. 42, No. 160, August 18, 1977.
                              2-3

-------
                   TABLE 2-2.   GAS STREAM CONDITIONS MEASURED DURING PARTICULATE AND HYDROCARBON TESTS

Test No.
LWI-51
LW0-51
LWI-52
LWO-52

Sample
location
Inlet
Outlet
Inlet
Outlet

Gas stream
temperature
°C
73
32
81
34
°F
163
89
178
93

Gas absolute
pressure
mm Hg
756.8
762.1
750.9
755.7
in. Hg
29.80
30.01
29.56
29.75

Actual
flow rate
acmh
12,402
13,694
12.967
12,660
acfh
437,987
483,606
457,911
447,072
Flow rate at
standard ^
conditions
dscmh
10,275
12,999
10,296
11,751
dscfh
362,839
459,065
363,589
414,984

Composition,
% H20
1.89
1.51
2.90
2.23
% o2
20.0
20.2
20.0
20.2
% co2
0.4
0.2
0.4
0.2
to
     aActual  flow rate at stack temperature and pressure.
     bFlow rate adjusted to standard conditions:  760 mm Hg, 20°C (29.92 in.  Hg,  68°F);  dry basis.

-------
                    TABLE 2-3.  METHOD 25 DATA SUMMARY3
Test No.
LW-1-1
LW-1-2
LW-1-3
Average
LW-2-1
LW-2-2
LW-2-3
Average
LW-Ae
Hydrocarbon
concentration
as CHd,b ppm
Inlet
630
902
653
728
1319
1897
1104
1440
1739
Outlet
475
657
755
629
350
569
600
506
827
Hydrocarbon mass flow
rate as CH4C
Inlet
kg/h
4.31
6.17
4.47
4.98
9.04
13.00
7.57
9.87
11.92
Ib/h
9.50
13.60
9.85
10.98
19.93
28.66
16.69
21.76
26.28
Outlet
kg/h
4.11
5.68
6.53
5.44
2.74
4.45
4.69
3.96
6.47
Ib/h
9.06
12.52
14.40
11.99
6.04
9.81
10.34
8.73
14.26
Control d
efficiency ,
%
4.6
7.9

69.7
6b.8
38.0
59.9
45.7
 Press speed was less than normal  and  data  should  not  be  used for  standard
 setting.

 Parts per million by volume measured  as  methane.

°Based on  dry standard gas flow rates  measured  during  the test  period  (see
 Table 2-2) and on the molecular weight of  methane:  16 g/g-mole
 (16 Ib/lb-mole)                                    kg     kg

 Control  efficiency is based on mass emission rates  —^Vj	^— x 100.

eLW-A run  as standard Method 25 (single point,  constant   in
 sampling  rate) simultaneous with inlet and outlet runs
 LW-2-3.
                                    2-5

-------
     During Test 1, hydrocarbon concentrations at the inlet test



site were measured at 630, 902, and 653 ppm as methane with an



average concentration of 728 ppm as methane.  The equivalent



hydrocarbon mass flow rates to the condensers were 4.31, 6.17,



and 4.47 kg/h (9.50, 13.60, and 9.8 Ib/h) with an average value



of 4.98 kg/h (10.98 Ib/h).



     At the outlet site, Test 1 concentrations were 475, 657, and



755 ppm as methane.  The outlet emission rates were 4.11, 5.68,



and 6.53 kg/h (9,06, 12.52, 14.40 Ib/h) with an average value of



5.44 kg/h  (11.99 Ib/h).



     Based on mass emission rates, the control efficiencies



indicated by the results are 4.6 and 7.9 percent for the first



and second samples and negative efficiency for the third sample



and the average.  The press was operating at the rate of 259



m/min  (850 ft/min) through most of the test period.



     During Test 2, the press speed was 305 m/min  (1000 ft/min)



and the measured inlet hydrocarbon concentrations were 1319,



1897, and 1104 ppm for an average of 1440 ppm as methane.  The



average value of 1440 ppm corresponds to a hydrocarbon mass flow



of 9.87 kg/h (21.76 Ib/h) as methane.



     At the outlet, the measured concentrations were 350, 569,



and 600 ppm as methane with an -average concentration of 506 ppm.



The average mass emission rate was 3.96 kg/h (8.73 Ib/h).



     Based on the average inlet and outlet mass emission rates,



the average control efficiency for Test 2 was 59.9 percent.  The



control efficiencies for Samples 1, 2, and 3 were 69.7, 65.8, and




                               2-6

-------
38.0 percent, respectively.  Press speeds during both tests



were lower than normal.



     Test LW-A consisted of standard Method 25 samples collected



simultaneously with the third inlet and outlet samples in Test 2.



These samples were collected at a single point in the inlet and



outlet stacks using a constant sampling rate of 85 ml/min.  No



particulate filters were used at either site.  The hydrocarbon



concentrations measured with standard Method 25 procedures were



1739 ppm at the inlet to the condensers and 827 ppm at the outlet



of the ESP.  The control efficiency at 45.7 percent was higher



than the 38 percent measured with Sample LW-2-3, which was taken



from the back of the Method 5 filter.



     If the values of the first test were to be consistent with



the values of the second test, the inlet concentrations of the



first test should be multiplied by a factor of 2.  The data,



however, have been checked and rechecked and samples were re-



analyzed to check results.  The initial values remain unchanged.



     The procedure of collecting Method 25 samples from the back



of a Method 5 filter is an innovation.  Data on the precision and



accuracy expected from this sampling procedure has not yet been



compiled and published by the EPA.  In reviewing the data in



Table 2-3, it appears that the variance in the measured concen-



trations for both tests is about +20 percent.  This variance may



be partly due to the variance in actual stack conditions and



partly due to the inherent precision of the test procedure.



Thus, the negative efficiency indicated by the data in Test 1 is




                               2-7

-------
probably more a reflection of the precision of the test method



than the actual operating conditions.  The actual efficiency is



probably low but positive.  Based on the variance in the data,



the true efficiency could be as high as 10 to 15 percent or as



low as the 5 percent indicated by Tests LW-1-1 and LW-1-2.  The



data from Test 2 show about the same variance as measured effi-



ciency ranges from 69.7 percent to a low of 38.0 percent.



     Appendix B contains field data sheets for the velocity



traverses and Method 25 sampling.  Appendix C contains all



laboratory results.  Appendix D contains a complete description



of the sampling and analytical procedures.



Isopropanol Content



     Isopropanol is used as a roll wetting agent in the press



fountain solutions and as a general solvent for cleaning around



the press.  To determine the amount of isopropanol vented in the



press dryer discharge, samples were collected at each site and



analyzed for isopropanol content.  Table 2-4 gives the results of



these analyses.



     During each particulate test, an evacuated stainless steel



tank was used to collect the samples.  A stainless steel probe



line with an outside diameter (O.D.) of 0.635 cm was used.  A



needle valve was placed between the tank and the probe line to



control sample flow rate.  In analyzing these samples, several



light hydrocarbon peaks were included in the total isopropanol



content (see Section 4.4).  This was based on the assumption that



the light hydrocarbons present resulted from chemical reactions



of the isopropanol in the dryer.



                              2-8

-------
                                 TABLE  2-4.   ISOPROPANOL CONTENT OF THE  EMISSION STREAM
Test No.
LW-1
LW-2
Isopropanol
concentration ,
Inlet,
ppnf
88
117
Outlet,
ppnJc
24
86
Equivalent
concentration
as methane,
Inlet,
ppmd
264
351
Outlet,
ppnr
72
258
Percent of total
hydrocarbon3
Inlet,
I
36.3
24.4
Outlet,
%
11.4
51.0
Isopropanol
emission rateb
Inlet
kg/h
2.25
3.01
Ib/h
4.98
6.64
Outlet
kg/h
0.7U
2.53
Ib/h
1.72
5.58
ro
vo
       The Isopropanol concentration as methane compared with the average total hydrocarbon concentration as
       methane (see Table 2-3) for the corresponding test number.
       Based on dry standard flow rate measured during the test and on the molecular weight of isopropanol:
       60.09 g/g-mole.
      "Parts per million by volume as isopropanol.
       Basis:  3 moles of methane formed per mole of isopropanol.

-------
     At the inlet site, isopropanol concentrations were 88 and



117 ppm as isopropanol.  For comparison to the Method 25 data,



the equivalent concentrations if measured as methane would be 264



and 351 ppm.  Thus the alcohol content was 36.3 percent of the



total hydrocarbon concentration measured with Method 25 for Test



1 and 24.4 percent for Test 2.  Based on the molecular weight of



isopropanol (60.09 g/g-mole), mass flow rates were 2.26 kg/h



(4.98 Ib/h) for Test 1 and 3.01 kg/h (6.64 Ib/h) for Test 2.



     At the outlet site, isopropanol concentrations were 24 ppm



and 86 ppm as isopropanol.  The equivalent concentrations as



methane would be 72 and 258 ppm.  The mass emission rates of



isopropanol were 0.78 kg/h  (1.72 Ib/h)  in Test 1 and 2.53 kg/h



(5.58 Ib/h) for Test 2.



     During Test 1 at the outlet site,  the needle valve in the



sample train became partially plugged,  thus reducing the sampling



rate.  Although the total sample volume was about the same as the



other samples, the plugged valve and reduced sample rate may have



affected the amount of isopropanol collected.  This may account



for the lower than expected value of isopropanol found in this



test (24 ppm).



     Isopropanol would reduce the control efficiency of the con-



trol system operation because isopropanol has a much lower



boiling point than the ink solvent and would therefore be less



condensible at the condenser operating temperature.  In the



second test where the measured efficiency was positive, the



control efficiency would change from 59.9 percent, based on






                              2-10

-------
average total hydrocarbon content, to 74.0 percent, if the



alcohol content were subtracted from the total hydrocarbon at the



inlet and outlet.  In Test 2 the hydrocarbon emission rates



corrected by subtracting the isopropanol content would be 7.43



kg/h (16.38 Ib/h) at the inlet and 1.93 kg/h (4.26 Ib/h) at the



outlet.



Method 5 Data Summary



     Table 2-5 summarizes the data obtained from the Method 5



sampling train.  In all cases, the amount of material collected



was extremely low.  Method 5 filters were used through both



tests.   At the inlet site, the total filter catch was only 0.8 mg



in 7.42 standard cubic meters sampled (262.030 scf).  At the



outlet site, the filter catch was zero in 6.535 scm (230.768



scf).  Condensible organic material ranged from 11.9 mg for Test



1 to 43.7 mg for Test 2 at the inlet.  At the outlet site, the



condensible organic catches were 22.5 mg Test 1 and 8.7 mg in



Test 2.  Impinger contents were recovered after each test.



     In no case was there sufficient material to ensure an accur-



ate recovery and weighing.  Therefore, the only conclusion indi-



cated by the Method 5 data is that the particulate and aerosol



content of the gas stream was negligible.



     Data in Table 2-5 do not include particulate catch from the



toluene rinse of the front-half glassware and probe.  These



rinses were shipped to the Emission Measurement Branch laboratory



for analysis.  No particulate could be observed by visual inspec-



tion of the rinses.




                               2-11

-------
                     TABLE 2-5.   METHOD 5  DATA  SUMMARY
Test
date,
1981
3/26
3/26
3/28
3/28
Test
No.
LWI-5-1
LWO-5-1
LWI-5-2
LWO-5-2
Sample
location
Inlet
Outlet
Inlet
Outlet
Filterable9
particulate
kg/h
0.001
0.000
0.001
0.000
Ib/h
0.002
0.000
0.002
0.000
Condensible
orqanics
kg/h
0.033
0.086
0.120
0.033
Ib/h
0.073
0.189
0.265
0.072
Condensible
inorganics
kg/h
0.009
0.006
0.013
0.006
Ib/h
0.020
0.013
0.029
0.013
Does not include toluene rinse  of front-half  glassware and probe.
                                   2-12

-------
     All Method 5 field data are presented in Appendix B of this

report.  Values in Table 2-5 were calculated without including

the Method 25 sample volumes.  These sample volumes ranged from 3

to 4 liters per sample and thus had a negligible effect on the

Method 5 calculations.  For comparison, calculations were made

onsite using the Method 25 sample volumes.  These calculations

are shown in Appendix B with the field data sheets.  Method 5

computer calculations are shown in Appendix A of this report.
Laboratory data are presented in Appendix C.  A complete descrip-

tion of the sampling and analtyical procedures is shown in

Appendix D.

2.2  PROCESS DATA
     During each test, the process operation was carefully moni-

tored, and operating parameters were recorded every 10 minutes.

In addition, an attempt was made to measure the volume of ink

dispensed, the quantity of isopropanol used, and the amount of
ink solvent and water recovered from the condenser drain.

     Table 2-6 summarizes the press operation during the test
periods.  Press speed during both tests was slower than normal

because of problems with the press operation.  Test 1 press speed

averaged 261 m/min  (856 ft/min).  During Test 2, the press speed

was constant at 305 m/min (1000 ft/min).  Normal press speed

ranges from 427 to 488 m/min (1400 to 1600 ft/min).  Some down-

time occurred during each test.  Testing was stopped whenever the

press went down and resumed only when the press was back on line

at full operating speed.  Appendix B contains a complete list of

downtime and causes along with the field data sheets.
                               2-13

-------
                        TABLE  2-6.   PRESS OPERATION SUMMARY FOR THE LEHIGH STECK-WARLICK COMPANY
Test
No.
LW-1
LW-2
Date,
1981
3/26
3/28
Measurement
period,
24-h time
1500 - 2010
0845 - 1745
Downtime during
measurement,
min.
40
141
Total
operating
time,
min.
270
399
Signature count,
signatures/
minutes recorded
96,000/245
106,820/413
Press speed
summary
Time period
1500 - 1520
1520 - 1545
1545 - 2010
0845 - 1745
m/min
262
290
259
305
ft/mi n
860
950
850
1000
to
I

-------
     Table 2-7 lists the average values of the press operating



parameters monitored during the testing.  Data were recorded



every 10 minutes during the test, but no significant variations



were observed.  Appendix B presents a complete data list along



with field data.



     During each test, the ambient relative humidity was measured



in the press room with a sling psycrometer.  In Test 1, the rela-



tive humidity averaged 56 percent at 24°C  (76°F).  In Test 2, the



relative humidity was 60 percent at 26°C (79°F).



     At the test site, the condenser cooling fluid temperature,



the air temperature at the outlet of the final condenser, and the



ambient temperature and relative humidity were recorded every 10



minutes during the test.  Table 2-8 lists the average values of



these parameters for each test.  Cooling fluid temperatures were



steady for both tests with a variance of only +1°F through the



entire test period.  The air temperature was also constant, vary-



ing +1°C.



Process Ink



     Process ink is pumped to the press by means of a piston-type



displacement pump.  The piston covers the cross-sectional area of



the ink drum, and ink is forced out of the exit line as the



piston is moved into the barrel.  The displacement of the pump



and the cross-sectional area of the ink drum were measured to



determine the volume of ink used during the test periods.



Samples of each ink color used were obtained from the supplier,



Sun Chemical Corporation, and analyzed to determine the density



and the percent volatile content.



                               2-15

-------
                                             TABLE 2-7.   PRESS  PROCESS  DATA SUMMARY1
Test
No.
LW-1
LW-2
Date
1981
3/26
3/28
Web speed
m/min
261
305
ft/min
856
1000
Web temperature
°C
121
103
°F
249
218
Dryer tern
°C
131
148
perature
°F
269
298
Chill roll water
Temperature
Inlet
°C
17
17
°F
62
63
Outlet
°C
18
19
°F
64
67
Flow rate,
mVmin
0.202
0.204
    Average  values  reported here.   Complete data log is  in Appendix  B.
ro

\->
a\

-------
TABLE 2-8.   CONTROL  EQUIPMENT OPERATION
Test
No.
LW-1
LW-2
Date,
1981
3/26
3/28
Ambient conditions
Temperature
°C
23
21
°F
73
69
% relative
humidity
49
77
Condenser cooling
fluid temperature
Inlet Outlet
°C
28
29
°F
82
84
°C
44
46
°F
111
115
Air temperature at
the outlet of the
final condenser
°C
34
34
°F
93
93
               2-17

-------
     Table 2-9 lists the ink usage for each test.  The mass of



ink used was determined based on the average ink density for all



four colors, 0.985 kg/liter (8.219 Ib/gal).  Density and volatile



content data are listed in Table 2-10.  Ink density varied from



0.974 kg/liter for the yellow and black inks to 1.010 kg/liter



for the blue ink.  This amounts to a variance of only +2.5 per-



cent from the mean value.  The average density of the four ink



colors and the total volume of the four ink colors dispensed were



used to compile the data in Table 2-9.



     For Test 1, the amount of ink dispensed was 95.04 liters



(25.11 gal), which is equivalent to 93.61 kg (206.38 Ib).  A



total of 274 minutes of press operating time was recorded for the



measurement period.  The rate of ink usage was 0.39 kg/min  (0.75



Ib/min).  In Test 2, 152.99 liters (40.39 gal)  of ink were dis-



pensed in 405 minutes of operating time.  The ink usage rate for



Test 2 was 0.37 kg/min  (0.82 Ib/min).



     Volatile content of the ink was determined by means of Pro-



cedure B of ASTM Method D-2367.  In this procedure, the ink is



dispersed in solvent and heated for 1 hour at 110°C to drive off



volatile material.  The sample is weighed before and after the



heating period to determine the percent volatile content gravi-



metrically.  Because this method may not be applicable for all



types of coatings, samples were also heated through extended



periods of 3, 4, 5, 6, and 24 hours to determine the relationship



of heating time to the measured volatile content.  These data are



included in Table 2-10.  Figure 2-1 is a graphical representation



of this same data.             010
                               f*~ J.O

-------
          TABLE  2-9.   INK USAGE AT THE LEHIGH STECK-WARLICK COMPANY
Test
No.
LW-1
LW-2
Date,
1981
.3/26
3/28
Total ink usage, all colors
Volume
liter
95.04
152.89
gal
25.11
40.39
Mass3
kg
93.61
150.60
Ib
206.38
332.01
Total
operating
time,b
mm
274
405
Usage rate
kg/min
0.34
0.37
Ib/min
0.75
0.82
aBased on average ink density of 0.985  kg/liter  (8.219 Ib/gal).
 Total operating time = (measurement  time  period)  -  (downtime in period).
                                     2-19

-------
             TABLE 2-10.   INK DENSITY AND PERCENT VOLATILE CONTENT
Ink color
Density, kg/liter
Ib/gal
% Volatile*9
1 h at 110°C
3 h at 110°C
4 h at 110°C
5 h at 110°C
6 h at 110°C
24 h at 110°C
Blue
1.010
8.428

18.77
31.00
31.60
31.90
31.95
32.55
Black
0.974
8.127

19.48
21.85
23.65
25.05
26.05
32.05
Red
0.980
8.178

27.36
34.60
35.35
35.80
36.00
37.35
Yellow
0.974
8.127

27.84
34.30
35.30
36.00
36.55
38.75
Average
all colors
0.985
8.219

23.36
30.44
31.48
32.19
32.64
35.18
aPercent volatile content by weight determined by Procedure B of ASTM Method
 D-2369-81, which calls for a heating period of 1 hour at 110°C.  To deter-
 mine relationship between heating time and measured volatile content, samples
 were also heated through 3-, 4-, 5-, 6-, and 24-hour periods.
                                    2-20

-------



1—
z
i£ UJ
Z h— H—
«— z a;
O 13
O <_>•-•
1 1 1 1 1 1
c£ LLJ 2
«£ t— CO
UJ 
-------
     The measured volatile content increases steadily with in-



creased heating time.  Even after 6 hours at 110°C, measured



volatile content had not reached a steady level as indicated by



the samples heated for 24 hours.



     Laboratory data for the ink samples are shown in



Appendix C.



Residual Solvent Content in Printed Product



     Samples of the printed signatures were collected during each



test and analyzed for residual solvent content.  The paper



samples were extracted with pentane by means of a soxhlet extrac-



tion apparatus, and the extracts were analyzed for ink solvent



content by gas chromography (GC).  Samples of the ink solvent



were used to identify the solvent peaks from the extract.



Unprinted paper was extracted as a blank to identify and elimi-



nate from consideration those peaks resulting from extractables



in the paper.



     Signature samples were collected at the beginning, middle,



and end of each test run.  Each sample consisted of about 25



signatures.  At the PEDCo laboratory, the top and bottom portions



of this sample were discarded.  At least one whole signature



from the center of the sample pack was then shreaded and loosely



packed in a soxhlet apparatus.  No attempt was made to evaluate



the percent ink coverage on the signature nor to separate the



lightly covered areas from the more heavily covered areas.
                                2-22

-------
     Soxhlet extraction data are summarized in Table 2-11.



Samples from Test 1 averaged 0.30 ing of solvent per gram of



paper.  In Test 2, the samples average 0.20 mg of solvent per



gram of paper.



     The paper used on both test days was Champion paper product



weighing 60 Ib/ream.  A ream of paper is defined as 500 sheets



measuring 25 by 38 in.  Thus, the weight of the paper is 1.26 x


  -4       2
10   Ib/in.   During Test 1, 96,000 signatures were run in 245



operating minutes.  Each signature is 42 in. wide by 45.669 in.


                   2
long or 1918.10 in.   Thus, in Test 1, 5682 Ib of paper per hour



were run through the press.  Based on the average solvent content



from Table 2-11, 1.70 Ib/h of ink solvent was retained in the



printed product.  The measured ink usage rate for Test 1 was 0.75



Ib/min or 45.00 Ib/h.  Assuming that 30 percent of the ink



weight is solvent, 13.5 Ib/h of ink solvent was applied to the



paper.  Therefore, about 12.6 percent of ink solvent applied



remained in the printed product after the press dryer unit.



     In Test 2, 106,820 signatures were run in 413 operating



minutes for a rate of 3751 Ib/h of paper run.  Based on the



extraction data, therefore, 0.75 Ib/h of solvent was retained in



the printed product.  The ink usage rate for Test 2 was 0.82



Ib/min or 49.20 Ib/h.  If the ink contains 30 percent solvent,



14.76 Ib/h of solvent was applied to the paper.  Thus, 5.1 per-



cent of the solvent applied remained in the printed product.
                               2-23

-------
       TABLE 2-11.   RESIDUAL SOLVENT  CONTENT OF PRINTED SIGNATURES
Sample I.D.
Test la
No. 1, top line
No. 1, bottom line
No. 2, top line
No. 2, bottom line
No. 3, top line
No. 3, bottom line
No. 4, top line
No. 4, bottom line
Average Test 1
Test 2
No. 1
No. 2
No. 3
Average Test 2
Weight of paper
extracted, g
36.69
35.26
35.62
37.63
35.42
38.99
40.92
42.03
37.82
40.31
39.90
36.92
39.04
Solvent content
mg solvent/
g paper
0.38
0.40
0.24
0.37
0.21
0.23
0.27
0.29
0.30
0.19
0.21
0.21
0.20
Ib solvent/
Ib paper
3.8 x 10"!
4.0 x TO"!
2.4 x 10":
3.7 x 10";
2.1 x 10"7
2.3 x 10":
2.7 x 10":
2.9 x 10~*
3.0 x 10"4
1.9 x 10"!
2.1 x ID":
2.1 x 10"4
2.0 x 10"4
Signatures from the Test 1  product were split after printing  into  two
 product lines.   Thus, top line and bottom line samples  are  of the  same
 signature.  Product in Test 2 was not split.
                                     2-24

-------
     Laboratory data from the study are presented in Appendix C.



A description of the sampling and analytical procedure is con-



tained in Appendix D.



Fountain Solutions



     During the tests, samples of the fountain solutions were



collected to check the concentration of isopropanol.  In addi-



tion, attempts were made to measure the amount of isopropanol



used per test.



     Isopropanol concentration was measured by determining the



specific gravity of the solutions using a hydrometer.  In Test 1,



fountain solution concentrations ranged from 6.7 to 10.0 percent



isopropanol with an average value for the four solutions of 8.2



percent.  The makeup alcohol was 96.9 percent isopropanol.  In



Test 2, the fountain solution concentrations ranged from 6.2 to



10.9 percent isopropanol with an average value of 8.2 percent.



Makeup alcohol for Test 2 was 96.4 percent isopropanol.



     All four fountain solution temperatures were checked every



1/2 hour through a 4-hour period during Test 1.  Temperatures of



the fountain solution ranged from 14° to 17°C  (58° to 63°F) with



an overall average temperature of 16°C  (60°F).  Alcohol use was



determined by measuring the liquid level in the makeup containers



at various times during Test 1.  This proved impractical, how-



ever, because the makeup containers were frequently refilled,



often before the data clerk could take readings.  As a result,



only one valid measurement was recorded for one 71-minute period.



Based on this measurement, 1.49 liters of 96.9 percent





                              2-25

-------
isopropanol were used in 71 minutes.  This corresponds to a usage



rate of 16.5 g/min (0.036 Ib/min).



     During Test 2, the press operators were instructed to fill



several 5-gal containers with makeup alcohol and then to use



these containers to refill fountain solution containers.  Thus,



alcohol use was measured by weighing the full 5-gal containers at



the start of the day, and as they were emptied.  The alcohol in



these containers, however, was also occasionally used for general



cleaning around the press.  Thus, this measurement does not yield



an accurate indication of the amount of isopropanol delivered to



the fountain solutions.  Based on this measurement, a total of



185.36 Ib of isopropanol was used in 359 operating minutes.



Thus, the usage rate for isopropanol was 0.516 Ib/min or 234.3



g/min.



Material Recovered from the Condenser Drain



     To measure the material collected in the condensers, a



decanter system was set up on the condenser drain line.  The



decanter consisted of a large drum with outlet taps at the top



and bottom of the drum.  Ink solvent was collected off of the top



drain and condensed water from the bottom of the drum.  Solvent



and water were collected in tared buckets and weighed to deter-



mined the quantity condensed.



     In Test 1, a total of 25.91 Ib of ink solvent was collected



in 319 operating minutes.  A total of 5.38 Ib of water was col-



lected in the same 319-minute period.  Thus, 4.87 Ib/h of ink



solvent and 1.01 Ib/h of water were condensed.  In Test 2, 47.40





                               2-26

-------
Ib of ink solvent and 0.89 Ib of water were recovered in 400



operating minutes.  The rates of condensation were,  therefore,



7.11 Ib/h for solvent and 0.13 Ib/h for water.



     The Method 25 data for Test 2 indicates that 13.03 Ib/h of



ink solvent were condensed.
                                2-27

-------
                            SECTION 3



                       PROCESS DESCRIPTION






     Testing was conducted on Press No. 64 at the Lehigh Steck-



Warlick Company in Dallas, Texas.  Press No. 64 is a Toshiba



model four-color printing press with a TEC Systems Series 80



dryer unit.  Vapors exhausted from the dryer are vented along



with dilution air to a control system consisting of five water-



cooled condensers in series followed by an ESP with two collec-



tors.



     The printing rolls on the Toshiba press are 106.68 cm  (42



in.) in diameter with a 116-cm (45.669-in.) cutoff.  Normal



printing speed ranges from 427 to 488 m/min (1400 to 1600 ft/min).



The press is designed for maximum operating speed of 549 m/min



(1800 ft/min).



     The inks used in this printing operation are formulated with



a solvent base that consists primarily of C-10 to C-16 straight



chain, saturated hydrocarbons.  Based on solvent vapor pressure



data, it was estimated that the exhaust air from the dryer could



contain up to 3000 ppm hydrocarbon vapor at a dryer temperature



of 132°C (270°F).



     The other source of hydrocarbon emissions from the press



operation is the isopropanol used as a wetting agent in the



fountain solution.  Samples of the four fountain solutions were




                               3-1

-------
collected during the test.  The isopropanol concentration in


these samples ranged from 8 to 11 percent by volume.  Pure


isopropanol (~97 percent) is used to make up the alcohol content


of the fountain solutions and is also used for cleaning the


printing rolls and other equipment around the press.  Press rolls


are cleaned periodically during the day, and some of the vapors


from the isopropanol used for cleaning are also drawn through the


press dryer.


     The TEC Series 80 dryer normally operates in the temperature


range of 121° to 127°C (250° to 260°F).   This is somewhat lower


than the temperature used in most press dryers, but the dryer is


longer than most units and thus provides more drying area.  This


arrangement is employed by the plant to lower the chance of


damaging the paper or print quality through overheating, to pro-


vide more even drying, and to conserve energy by lowering the


heat input to the dryer and reducing the amount of cooling


required in the chill rolls.  During testing, the dryer tempera-


ture was measured at 131°C  (268°F) for the first test and 148°C


(298°F)  for the second.


     The emission control system consists of five condenser units

                                T3
in series followed by a Smog Hog  ESP with two collectors.  Most


of the control equipment was supplied by United Air Specialists,


Inc.


     A 50/50 mixture of ethylene glycol and water is used as the


cooling fluid in the condenser system.  The gylcol-water mix


enters the condenser system at the fourth condenser then flows
                              3-2

-------
through the fifth, third, second, and first condensers in series.



During testing, the inlet temperature of the cooling fluid



averaged 28°C  (83°F), and the outlet temperature after the first



condenser averaged 45°C  (113°F).  These temperatures were read



from thermometers installed on the inlet and outlet lines by



plant personnel.  Cooling fluid flow rate is approximately 132



liters/min. (35 gal/min).



     Hot cooling fluid from the condensers is directed to a heat



exchanger near the press where the heat of the fluid is used to



preheat dryer blower air.  Because the dryer preheater and the



condensers at the control system are part of the same heat ex-



change system, temperatures at the condenser are not variable.



Experiments conducted by Lehigh Steck-Warlick personnel have



shown that varying condenser temperatures more than +^10 °F results



in fouling and plugging of the condenser coils and reduced



operating efficiency in the ESP.  All testing was conducted under



normal operating temperatures.
                               3-3

-------
                            SECTION 4
               SAMPLING AND ANALYTICAL PROCEDURES

     Tests were conducted at the inlet and outlet of the con-
denser/ESP to determine the hydrocarbon concentrations at each
site and the control efficiency achieved by the condenser/ESP
system.  Because a potential existed for aerosol formation in the
inlet duct, samples were collected isokinetically by means of the
EPA Method 5* sampling procedure.  A Method 25** sample was col-
lected from the backside of the Method 5 filter and analyzed for
total nonmethane organic content.  In addition to the emission
test, samples of the process ink, printed product, and isopropa-
nol fountain solutions were collected for analysis.  The follow-
ing report sections describe the sampling and analytical proce-
dures used in this test.

4.1  DESCRIPTION OF SAMPLE LOCATIONS
     The inlet test site consisted of an insulated 15 by 24 in.
rectangular duct.  Approximately 2-1/2 duct diameters of undis-
turbed flow were downstream from the sample location and about
1/2 diameter upstream.  A total of 25 sample points were used in
each traverse.  Five ports were installed in the duct to provide
 *
**
Federal Register, Vol. 42, No. 160, August 18, 1977.
Federal Register, Vol. 45, No. 194, October 3, 1980.

-------
a 5 by 5 sample point matrix.  Based on the procedures of Method

1*, a 6 by 6 sample point matrix should have been used to provide

36 traverse points.  Studies by Fluidyne, Entropy Environmenta-

lists, Inc., and others indicate that the number of points

sampled to obtain representative data may be less than those
                                123
presently specified in Method 1. '  '   The Method 25 sampling

time is also limited by the size of the sample tank to less than

1 hour.  Thus, to accommodate the Method 25 sample, the number of

sample points were reduced to 25, and each sample point was

sampled for 2 minutes for a total test time of 50 minutes.

Figure 4-1 is a diagram of the inlet test site.

     The rectangular duct at the outlet test site was 20 by 24

in.  The sample location selected was 4 duct diameters from the

nearest downstream flow disturbance and 1 duct diameter from the

nearest upstream flow disturbance.   Four sample ports were

installed.  Traverse points were laid out in a 4 by 5 matrix to

provide a total of 20 traverse points.  Figure 4-2 is a diagram

of the test site and traverse point locations.  Again, the number

of traverse points selected was based on the total test time

allowed by the Method 25 sample train and the rationale presented

above.  Each traverse point was sampled for 2-1/2 minutes for a

total test time of 50 minutes.
if
 Federal Register, Vol. 42, No. 160, August 18, 1977.

 H.A. Hanson, R.J. Davini, J.K. Morgan, A.A. Iverson.  Fluidyne.
 Particulate Sampling Strategies for Large Power Plants Including
 Nonuniform Flow.  EPA-600/2-76-170, June 1976.
2
 Traverse Point Study.  June 1977.  Entropy Environmentalists,
 Inc., P.O. Box 12291, Research Triangle Park, North Carolina
 27709.  18 p.

 P.G.W. Hawksley, S. Badzioch, and J.H. Blackett.  Measurement of
 Solids in Flue Gases.  The British Coal Utilization Research
 Association.  Leatherhead, Surrey, England, 1961.
                               4-2

-------
ROOF LEVEL
              -15 in.
        DRYER DISCHARGE
           SIDE  VIEW
0.8 ft

   SAMPLE PORTS
                                                     -24 in.-
o   o   o   o   o
   SAMPLE  PORTS
                                            ROOF
                                                                    LEVEL
                          DRYER  DISCHARGE
                            FRONT VIEW
 TRAVERSE POINT LOCATION

          24 in.
0
o
o
o
o
o
0
o
0
o
^•^ ^^™
0 '
o
o
0
o
^M— *••
o
0
o
0
o
MM, MH
o
o
0
o
o
                      15 in.
       SAMPLE  PORTS
TRAVERSE
POINT NO.
1
2
3
4
5
LOCATION IN DUCT,
INCHES
2.0
5.0
8.0
11.0
14.0
       Figure 4-1.   Inlet test  site and traverse point locations.
                                  4-3

-------
                ESP
                            1 3/4 ft
                            7 1/4 ft
                                      SAMPLE PORTS
                                      o    o  o   o
                                         -20 in.-
                                        I.D. FAN
                TRAVERSE POINT LOCATION
    20 in.
 o   o    o   o


 o   o    o   o


 o   o    o   o


 0000


 o   o    o   o

- —.,— —— -

  SAMPLE  PORTS
24 in.
TRAVERSE
POINT NO.
1
2
3
4
5
LOCATION IN DUCT,
INCHES
2 5/8
7 3/8
12 1/4
17
21 3/4
  Figure 4-2.  Outlet test site and traverse point locations.
                               4-4

-------
4.2  GAS STREAM FLOW RATE, TEMPERATURE, AND COMPOSITION


     Gas stream flow rate and temperature were measured before


each test using the procedures described in EPA Method 2*.  Dur-


ing each particulate test, an integrated bag sample was collected


and analyzed by orsat analysis to determine the oxygen and


carbon dioxide content.  Method 3* sampling and analytical pro-


cedures were used.


     Gas stream velocities were measured with an S-type pitot


tube and an inclined draft gauge manometer.  Temperatures were


measured with a calibrated thermocouple and digital potentiom-


eter.


     Calculations and results for velocity traverses are shown in


Appendix A.  Equipment calibration procedures and results are


presented in Appendix E.



4.3  PARTICULATE AND HYDROCARBON EMISSION TESTS


     Hydrocarbon and particulate sampling were conducted simul-


taneously by means of a method that combined Methods 5* and 25**.


In conducting these tests, the gas stream was sampled isokinet-


ically using the standard Method 5 procedures described in the


Federal Register*.  The Method 25 sample was extracted from a tee


at the back of the Method 5 filter.  A shutoff valve was placed


between the Method 5 and 25 sample trains to enable the operator


to leak check each section separately.


     For each test, three complete particulate traverses were


made.  A Method 25 sample was extracted during each traverse.


  Federal Register, Vol. 42, No. 160, August 18, 1977.
**
  Federal Register, Vol. 45, No. 194, October 3, 1980.

                              4-5

-------
Method 5 impinger contents were recovered after the third tra-

verse and analyzed for organic and inorganic condensibles with

ether-chloroform extraction.  Impingers were rinsed first with

distilled water, then with acetone to ensure complete recovery of

the condensed organics.  The Method 5 filters and probe rinses

were not recovered until both tests were completed.

     The following leak check procedures were used in conducting

the sampling:

     1.   Prior to each traverse, the Methods 5 and 25 sampling
          trains were leak checked separately before they were
          connected.  The pretest leak check of the combined
          system was conducted with the shutoff valve to the
          Method 25 sample tank closed.  This was necessary to
          prevent loss of tank vacuum prior to sampling.

     2.   The posttest leak check was conducted with the shutoff
          valve between the Method 25 condensate trap and the
          Method 5 filter closed.  This procedure was necessary
          to prevent loss of material collected in the condensate
          trap.  Methods 5 and 25 sample trains were thus leak
          checked separately after each traverse.

     For analysis, the Method 5 filters were desiccated to a

constant weight and then weighed on an analytical balance to

within 0.1 mg.  The probe and front half of the filter holder

were rinsed with toluene after completion of the testing.

     Toluene rinses were shipped to the Emission Measurement

Branch laboratory for analysis.  The impinger contents were

rinsed with distilled water and acetone.  The distilled water

rinse and the impinger contents were analyzed for organic and

inorganic condensibles by ether-chloroform extraction.  After

separation, the organic and inorganic fractions were dryed,

desiccated to a constant weight, and weighed on an analytical

                              4-6

-------
balance.  The acetone rinse was treated in the same manner as the



ether-chloroform extract.  The condensible material from the



acetone rinse was added to the catch from the ether-chloroform



extract to determine the total organic condensibles.



     Method 25 samples were collected by drawing the sample gas



from the back of the Method 5 filter through a dry-ice condensate



trap by means of an evacuated sample tank.  A constant sampling



rate of 85 ml/min was maintained throughout the test.



     The sample tanks were analyzed by injecting the contents



into an analyzer that separates the nonmethane organics from CO,



CH., and C0_; oxidizes organic compounds to CO-; reduces the CO-



to methane; and measures the resulting methane with a flame



ionization detector  (FID).  Condensate is recovered by completely



oxidizing the trap contents, then reducing the C02 to methane for



measurement with the FID.  The total nonmethane organic content



is the sum of the trap and tanks fractions.



     A complete description of the sampling and analytical pro-



cedures is contained in Appendix D.  Calculations including gas



velocity data are shown in Appendix A.  Field and laboratory data



sheets are presented in Appendices B and C, respectively.  Equip-



ment calibration procedures and results are shown in Appendix E.





4.4  GAS STREAM ISOPROPANOL CONTENT



     During each test period, gas samples were collected and



analyzed for isopropanol content.  Samples were collected in



evacuated stainless steel tanks using an 0.635-cm O.D. stainless




                              4-7

-------
steel probe line.  A stainless steel needle valve was placed be-



tween the tank and probe line to control the sampling rate.



     In the laboratory, the sample tanks were pressurized with



nitrogen, and the contents were injected into a gas chromotograph



with an FID for analysis.  Calibration standards of isopropanol



in air were made by injecting known quantities of isopropanol



into stainless steel tanks similar to those used for sampling.



     A precolumn packed with 10 percent methyl silicone oil was



used to prevent the heavy molecular weight ink solvent from



bleeding into the separation column and FID.  Thus, only the



alcohol and light hydrocarbon fractions were quantified with the



FID.



     In calculating the isopropanol concentrations, peaks re-



sulting from all light hydrocarbons were summed.  In the press



dryer, some chemical reactions may occur with the isopropanol.



Thus, it was assumed that all light hydrocarbons observed were



the result of the isopropanol reactions and therefore should be



included with the measured isopropanol.  Appendix C of this



report contains all laboratory data and example chromatograms



to show those peaks which were summed.



     Field data for the isopropanol sampling are shown in Appen-



dix B.  Calculations and results are presented in Appendix A.





4.5  PROCESS SAMPLE ANALYSIS



Printed Signatures



     During the testing at the Lehigh Steck-Warlick Company,



signature samples were collected and analyzed for residual ink




                               4-8

-------
solvent content by means of a soxhlet extraction procedure sug-



gested by a GATF laboratory study.  Signature samples were col-



lected at the beginning, middle, and end of each test run.  Each



sample consisted of about 25 signatures.  Signature samples were



wrapped in oil-free aluminum foil to prevent solvent loss during



shipment to the PEDCo laboratory.  Unprinted paper samples were



collected and shipped in the same manner; extracted, and analyzed



as blanks.



     At the PEDCo laboratory, the top and bottom portions of



each sample pack were discarded.  At least one whole signature



(about 30 to 40 g of paper) from the center of the sample pack



was then shreaded and loosely packed in a soxhlet apparatus.  No



attempt was made to evaluate the percent ink coverage on the



signature, and no attempt was made to separate lightly covered



areas of the signature from the more heavily covered areas.



     The material was then subjected to a four hour soxhlet



extraction using nanograde pentane.  Approximately 350 ml of



pentane were used for each extraction.  After extraction, the



pentane extract was concentrated by evaporating the pentane at



room temperature.  The concentration step reduced the volume of



pentane extract to about 10 ml.  Unprinted paper was extracted



by the same procedure.



     Extracts were analyzed with a gas chromatograph and flame



ionization detector (FID).  The FID was calibrated with standards



of ink solvent in pentane.  Extracts from paper blanks were



analyzed to located and eliminate from consideration those GC




                               4-9

-------
peaks which were attributable to the extractable content of the

paper itself.  Signature extracts were then analyzed, and the

ink solvent content was quantified by comparison to the FID

response of the solvent calibration standards.

     The instrument used for analysis was a Perkin-Elmer 990 gas

chromatograph with a flame ionization detector.  The following

GC conditions were used:

     Column:  4 ft x 1/8 in. O.D. nickel, packed with 3% SP2100
              on 100/120 mesh supelcoport

     Temperatures:  Injection port:  350°C
                    Manifold:  350°C
                    Column:  Ambient, hold 3 minutes then program
                             to 325°C at 12°C/min

     Carrier gas:  Nitrogen at 30 cc/min

Process Ink Samples

     Samples of the process ink were obtained from the manufac-

turer, the Sun Chemical Corporation, and analyzed for density and

percent volatile content by means of the procedures specified in

Reference Method 24*.  These procedures are ASTM Method D-1475-60

for density and D-2369-81 for percent volatiles.  The procedure

for percent volatiles involves heating the sample for 1 hour at

110°C.  In addition, samples were also analyzed by means of an

alternate procedure suggested by printing industry representa-

tives in which the samples were heated for 3 hours at 110°C.  To

determine the relationship between heating time and the measured

percent volatiles, samples were also analyzed by heating for 4-,

5-, and 6-hour periods.  Percent volatile content appears to

level off after 5 hours of heating.
 Federal Register, Vol. 45, No. 194, October 3, 1980.
                               4-10

-------
     All results are included in this report.  Laboratory data



are shown in Appendix C.



Fountain Solutions



     During each test, samples of the fountain solutions, recir-



culating barrel contents, and makeup alcohol were collected to



determine the percent isopropanol in each.  The isopropanol



content in each was determined by specific gravity measurements



taken with a hydrometer.  Laboratory results are presented in



Appendix C.
                               4-11

-------
                            SECTION 5



                        QUALITY ASSURANCE






     Quality assurance is an important facet of stack sampling



because the end goal of testing is to produce representative



emission results.  Quality assurance guidelines provide detailed



procedures and actions necessary for defining and producing



acceptable data.  Three documents were used in this test program



to provide the required guidance to help ensure the collection of



acceptable data and to define data quality that is unacceptable.



These documents were the EPA Quality Assurance Handbook Volume



III  (EPA-600/4-77-8276), the draft PEDCo Environmental Emission



Test Quality Assurance Plan, and the PEDCo Environmental Labora-



tory Quality Assurance Plan.  The last two documents are PEDCo's



general guideline manuals that define the company's standard



operating procedures followed by the emission testing and labora-



tory groups.



     More details on the Quality Assurance procedures are pro-



vided in Appendix F.  Included in Appendix F are details on



quality assurance objectives, data reduction procedures, quality



control checks, performance and system audits, preventative



maintenance, precision, accuracy, completeness, corrective actions,



and quality assurance reports to management.





                               5-1

-------
     Routine standard reference method quality control procedures

were followed throughout this test program.  Included in the

standard procedures are the following:

     0    Calibration of field sampling equipment.  Calibration
          procedures and results are described in detail in
          Appendix E.  All equipment used met established EPA
          requirements.

     0    Train configuration and calculation checks.  Calcula-
          tions from preliminary data were checked on site by the
          EPA Task Manager.  This procedure verified that iso-
          kinetic sampling rates and gas flow rate and moisture
          data were acceptable before further testing was con-
          ducted.  Onsite calculation sheets are included with
          field data in Appendix B.

     0    Onsite quality assurance checks such as sample train,
          pitot tube, and orsat line leak checks.  In addition to
          standard procedures, PEDCo performed the following
          checks during this test.  Dry gas meter calibrations
          were checked on site by means of EPA calibrated criti-
          cal orifice No. 174.  One meter was found to be 13.3
          percent off the expected volume and was not used for
          any testing.  The two meters that were used for the
          testing were 2.5 and 4.6 percent off the expected
          volume through the critical orifice.  Acceptable limits
          for this calibration check are +5 percent of the ex-
          pected volume.  Data sheets for the field meter box
          checks are in Appendix B with the field data.

          Thermocouples were checked on site against a mercury-
          in-glass thermometer on each day of testing.  All
          thermocouples were within 1.5 percent of the absolute
          temperature indicated by the reference thermometer.

     0    Use of designated sampling equipment and analytical
          reagents.  Sampling equipment is described in Appendix
          D.  Blank analytical data for the reagents used in
          sampling are shown along with the laboratory results in
          Appendix C.

     0    Lab analysis procedures.

     0    Internal and external audits to ensure accuracy .in
          sampling and analysis.
                                5-2

-------
     Sampling equipment, reagents, and analytical procedures
for this test follow and meet all guidelines established for
Methods 5* and 25**.
 *
**
Federal Register, Vol. 42, No. 160, August 18, 1977.
Federal Register, Vol. 45, No. 194, October 3, 1980.
                               5-3

-------